e Fun 00 Erufition‘*of ‘ 1“ e1“ .3 ilauea Volcano, Hawaii: | pisodes 1 Through 20, anuary 3, 1983”, Through June COVER Lava pond and low fountain within Puu Oo Crater shortly before onset of vigorous lava fountaining during episode 9. Lava was overflowing spillway, which is outside photograph at lower left. View approximately southward; photograph taken at 1559 H.s.t. September 15, 1983. THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1 THROUGH 20, JANUARY 3, 1983, THOUGH JUNE 8, 1984 ".18. DEPOEFFORY APR 3 3989 "1.; sapasg‘vomy 9389 Fountain (approx 200 m high) emanating from Puu Oo vent, 6.5 hours after beginning of episode 10. Most voluminous flow, com- posed of sluggish pahoehoe, has traveled 300 to 400 m southeastward (toward lower left) from vent. An additional aa-flow lobe (lower right) heads northeast, directly toward camera, within evacuated episode 9 channel, and spatter-fed flows blanket north side of cone. Fountain produced a significant tephra deposit around much of Puu 00. View southwestward; photograph by R.W. Decker, taken at 0730 H.s.t. October 5, 1983. The Puu Oo Eruption of Kilauea Volcano, Hawaii: Episodes 1 Through 20, January 3, 1983, Through June 8, 1984 EDWARD W. WOLFE, Editor U.S. GEOLOGICAL SURVEY PROFESSIONAL PAPER 1463 A comprehensive study of the first 1% years of the most voluminous eruption of Kilauea Volcano in historical time UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1988 DEPARTMENT OF THE INTERIOR DONALD PAUL HODEL, Secretary U.S. GEOLOGICAL SURVEY Dallas L. Peck, Director Any use of trade names and trademarks in this publication is for descriptive purposes only and does not constitute endorsement by the U.S. Geological Survey Library of Congress Cataloging-in~Publication Data The Fun 00 eruption of Kilauea Volcano, Hawaii. (U .8. Geological Survey professional paper ; 1463) "A comprehensive study of the first 1‘I2 years of the most voluminous eruption of Kilauea Volcano in historical time." Bibliography: p. Supt. of Docs.: 19.16:1463 1. Kilauea Volcano (Hawaii)—Eruptions. I. Wolfe, Edward W. 11. Series. QE523.K5P88 1988 551.2‘2'099691 86-600300 For sale by the Books and Open-File Reports Section, U.S. Geological Survey, Federal Center, Box 25425, Denver, CO 80225 PREFACE In contrast to many other of the world’s active volcanoes, Kilauea erupts frequently and with relatively little danger to human life. Furthermore, access to much of the volcano is relatively easy. For these reasons, Kilauea Volcano is an unrivaled location for volcanologic research. Studies of basaltic volcanism there have been carried out for the past 75 years by the staff of the Hawaiian Volcano Observatory (HVO), which has been operated continuously by the US. Geological Survey since 1948. This tradition of volcanic studies was continued dur- ing the Puu Oo eruption, abetted by increasingly comprehensive and sophisticated instrumentation, a continuously developing understanding of Kilauean magmatic processes, and a level of logistic support, provided by helicopters, that was unprecedented. The eruptive activity, which began on January 3, 1983, was episodic: Relatively brief periods of vigorous fountaining and high-volume flow production alternated with longer repose periods. By early June 1984, 20 distinct eruptive episodes had occurred. (At the time of this writing in June 1985, a total of 33 episodes had occurred, and the continuing Puu Oo eruption had become Kilauea’s most voluminous in historical time.) The Puu Oo eruption was intensely monitored by HVO during the first 11/2 years, and a wealth of observational, instrumental, and analytical data were collected. The repetitive style of the eruption provided a superb opportunity to assess a range of eruptive-episode and repose-period behavior. This volume presents the results of the first year and a half of comprehen- sive geologic, geophysical, geochemical, and petrologic monitor- ing and study. Note added in proof—Episodic lava-fountain eruptions con- tinued at Puu 00 through episode 47, in June 1986. In July 1986, as episode 48 was anticipated, a new vent opened approx- imately 3 km downrift from Puu 00, and lava discharge from the Puu Oo vent ceased. Relatively slow, quiet lava discharge from the new vent has continued almost without interruption to the present time (June 22, 1988), building a broad lava shield and an apron of pahoehoe flows that extends across the south flank of Kilauea Volcano to the ocean. CONTENTS [Numbers designate chapters] Page Preface ————————————————————————————————————————————————————————————————— VII 1. Geologic observations and chronology of eruptive events, by Edward W. Wolfe, Christina A. Neal, Norman G. Banks, and Toni J. Duggan ————————————————————————————————————————————————————————— 1 2. Lava samples, temperatures, and compositions, by Christina A. Neal, Toni J. Duggan, Edward W. Wolfe, and Elaine L. Brandt ———————————————————————————————————————————————————————— 99 3. Petrology of the erupted lava, by Michael O. Garcia and Edward W. Wolfe ———————————————————————————— 127 4. Gases from the 1983—84 east-rift eruption, by L.P. Greenland ——————————————————————————————————— 145 5. Constraints on the mechanics of the eruption, by L.P. Greenland, Arnold T. Okamura, and J .B. Stokes —————————————— 155 6. Surface deformation during dike propagation, by Arnold T. Okamura, John J. Dvorak, Robert Y. Koyanagi, and Wilfred R. Tanigawa —————————————————————————————————————————————————————— 165 7. Seismicity associated with the eruption, by Robert Y. Koyanagi, Wilfred R. Tanigawa, and Jennifer S. Nakata —————————— 183 8. Geoelectric observations (including the September 1982 summit eruption), by Dallas B. Jackson —————————————————— 237 ILLUSTRATIONS [Plates are in pocket] PLATE 1. Eruptive fissures in the middle east rift zone of Kilauea Volcano for eruptions from 1961 through 1983, and distribution of flows and vent deposits, measured flow thicknesses, and flow progress for episodes 1 through 3 of the Puu 00 eruption 2—5. Distribution of flows and vent deposits, measured flow thicknesses, and flow progress of the Puu 00 eruption for: 2. Episodes 4 through 7 3. Episodes 8 through 11 4. Episodes 12 through 15 5. Episodes 16 through 20 FRONTISPIECE. Photograph of approx 200—m-high fountain emanating from Puu Oo vent 6.5 hours after the beginning of episode 10 VII 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS By EDWARD W. WOLFE, CHRISTINA A. NEAL, NORMAN G. BANKS, and TONI J. DUGGAN CONTENTS "U "E Abstract Introduction Overview of the first 20 episodes ————————————— Significance of the Puu Oo eruption ——————————— Scope and approach ———————————————————— Methodology ——————————————————————— General observations ————————————————————— Development of vent structures —————————————— Growth of Puu 00 Lava ponds at central vents Fountains at episode 1 fissure vents ——————————— 15 Fountains at central vents Episodes 2 and 3: Fountains at the 1123 vent — — — — 16 Episodes 4 through 20: Fountains at Puu 00 ————— 16 oooooomls-Nmml-i Fissure-vent flows ____________________ 24 Central-vent flews — — — — — — — _ _ _ -_ _________ 26 River-fed flows ____________________ 26 Spatter-fed (rootless) flows ______________ 32 Numerical flow parameters —————————————— 35 Repose-period activity ___________________ 39 Eruptive-episode onsets, endings, and pauses ——————— 39 Later developments ____________________ 41 Chronologic narrative _____________________ 42 Episode 1 (January 3—23, 1983) ______________ 42 Summary of episode 1 ________________ 42 January 3, 1983 ____________________ 42 January 5 and 6, 1983 ________________ 44 January 7 and 8, 1983 ________________ 49 January 8—15, 1983 __________________ 50 January 23, 1983 ___________________ 51 Episode 2 (February 10—March 4, 1983) —————————— 51 Episode 3 (March 21—April 9, 1983) ____________ 54 Episode 4 (June 13—17, 1983) _______________ 60 Episode 5 (June 29—July 3, 1983) _____________ 63 Episode 6 (July 22—25, 1983) _______________ 67 Episode 7 (August 15—17, 1983) ______________ 69 Episode 8 (September 6—7, 1983) _____________ 73 Episode 9 (September 15—17, 1983) ———————————— 75 Episode 10 (October 5—7, 1983) ______________ 76 Episode 11 (November 5-7, 1983) _____________ 78 Episode 12 (November 30—December 1, 1983) ——————— 82 Episode 13 (January 20—22, 1984) _____________ 84 Episode 14 (January 30—31, 1984) _____________ 86 Episode 15 (February 14—15, 1984) ____________ 87 Episode 16 (March 3—4, 1984) _______________ 89 Episode 17 (March 30—31, 1984) _____________ 90 Episode 18 (April 18—21, 1984) ______________ 91 Episode 19 (May 16—18, 1984) ______________ 94 Episode 20 (June 7—8, 1984) _______________ 96 References cited ABSTRACT The Puu Oo eruption began at Napau Crater in the east rift zone of Kilauea Volcano on January 3, 1983. In its first 11/2 years, the eruption produced nearly 240 x 106 m3 of new basalt, built a new 130—m-high cone (Puu 00) at the principal vent, and spread basalt flows over more than 30 x 106 m2 of the rift zone and south flank of the volcano. Several flows entered sparsely populated areas and destroyed 18 dwellings. The Puu Oo eruption continues unabated as of May 1985. The initial outbreak was a fissure eruption. In sporadic eruptions over a 4-day period, the fissure system extended progressively farther downrift nearly 8 km, from Napau at the uprift end to the vicinity of the prehistoric cinder cone Kalalua at the downrift end. Extrusive activity then became localized south of Puu Kahaualea along a 1-km- long segment of the fissure system that erupted intermittently through mid-January. During the next 17 months, 19 brief (9 hours to 12 days) episodes of vigorous fountaining and high-volume emission of lava flows alternated with longer (8—65 days) repose periods. During eruptive episodes, har- monic tremor was at high levels, and rapid subsidence occurred at Kilauea’s summit. During repose periods, harmonic tremor was continu- ous but low, and Kilauea’s summit inflated. New flows and vent deposits accumulated at a fairly steady rate that averaged 13x 106 to 14x 106 m3/mo. Although the volume of lava produced in individual episodes ranged from 2 x 106 to 38 x 106 m3, most episodes produced from 8 x 106 to 14x106 m3. Lava-production rate increased through the series of eruptive episodes, and so comparable volumes of lava were discharged in progressively less time. After episode 1, central—vent eruptions dominated. The primary vent during episodes 2 and 3 was south of Puu Kahaualea. Beginning with episode 4, Puu 00, 1.5 km farther uprift, became the sole eruptive locus. From episodes 2 through 20, the dominant style was one in which a steep pipelike conduit delivered gas and disrupted lava through a vent in the floor of a broad crater within a growing cone of agglutinated spatter. This process formed a fountain that played above the surface of a pond formed of lava that had coalesced from the disrupted and degassed melt. The pond overflowed through a low point in the crater rim to feed one or more long, relatively narrow flows by way of a vigorous, channel- ized river of pahoehoe that underwent a transition to aa several kilometers from the vent. Most such flows were 4 to 8 km long and 100 to 500 m wide; a few were longer, and the longest extended more than 13 km from the vent. The height of the fountain, which reached a maximum of nearly 400 m, was related in part to the lava—discharge rate but was strongly influ- enced by other factors that may have included changing conduit conditions, entrapment of gas during repose periods, and damping by coalesced melt in the pond and, possibly, in the conduit beneath. High, broad-based fountains that maintained a high level of turbulent effer- vescence across the entire crater impeded development of the pond and the efficient lava-delivery system that it supplied. Thus, during periods of higher fountaining, flows were more likely to be relatively disorganized and short; at such times, thick, spatter-fed flows were common. The average velocity of the main, river-fed flows normally ranged from about 50 to 300 m/h, and a few times was from 400 to 500 m/h. The average velocity increased through the series of eruptive episodes, possibly in response to both decreased viscosity and increased discharge rate, which resulted in an increased supply of lava to the individual flows. 2 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 A tendency toward decreasing average flow thickness through the series of eruptive episodes also suggests that viscosity may have decreased as the Puu Oo eruption proceeded. Concomitantly increasing lava temperature, decreasing phenocryst content, and changing lava com- position may all have been related to this apparent change in viscosity. Low-level extrusive activity was common during repose periods. It was dominated by gradual ascent of the column of magma within the open pipe at Puu Oo Crater, and was punctuated by numerous occur- rences of gas-piston activity. INTRODUCTION OVERVIEW OF THE FIRST 20 EPISODES The Puu Oo eruption, in the middle part of Kilauea’s east rift zone, began at 0031 H.s.t. January 3, 1983, at Napau Crater (fig. 1.1). Over the next few days, the eruptive-fissure system extended progressively farther northeastward, and the eruptive locus migrated downrift to the vicinity of the prehistoric cinder cone Kalalua. These events were preceded and accompanied by a large subsidence of Kilauea’s summit (see chap. 6) and a swarm of Shallow earthquakes that migrated northeastward to the vicinity of Kalalua (see chap. 7) as the eruptive dike worked its way downrift through the shallow rocks of the rift zone. Extrusive activity then became localized south of Puu Kahaualea along a 1-km-long segment of the fissure system that erupted intermittently through mid- January. During this first episode, the style was domi- nated by fissure eruptions that produced linear fountains as much as several hundred meters long. Spatter from these fountains built low ramparts adjacent to the erup- tive fissures, and fluid sheets of pahoehoe, which con— verted locally to aa, spread from the erupting vents. After the January fissure eruptions, repetitious central- vent eruptions dominated, first at the 1123 vent (desig- nated by the time of its first eruptive activity) and subsequently at Puu 00 (fig. 1.1). Over the next 17 months (through the period covered by this chapter), there were 19 such eruptive episodes (fig. 1.2; table 1.1). Most were relatively brief occurrences of vigorous fountaining and high-volume production of lava flows, separated by longer periods of relative quiescence. During the eruptive episodes, harmonic tremor was at high levels, and rapid subsidence occurred at Kilauea’s summit; during the repose periods, harmonic tremor was continuous but low, and Kilauea’s summit inflated. Low-level eruptive activity precursory to the major eruptive episodes occurred dur- ing most of the repose periods. Although both the spacing between eruptive episodes and the volume of lava discharged in successive episodes varied somewhat (fig. 1.3), a striking regularity developed. Repose periods ranged in length from 8 to 65 days, but more than 60 percent were from 2 to 4 weeks long. New flow and vent deposits accumulated in the rift zone at an average rate of 13x 106 to 14x 106 m3/mo (value uncor- rected for vesicularity). Although the monthly average was fairly steady, over time the durations of individual episodes tended to decrease, and the lava-discharge rates to increase (fig. 1.4; table 1.1). By the end of episode 20, approximately 240 x 106 m3 of new flows and vent deposits covered 31 x 106 In2 of the central part of the rift zone and south flank of Kilauea. The flows of episodes 2 through 5 overran 15 dwellings in a sparsely populated subdivision, completely crushing and burying them after initially setting them on fire. Three more dwellings were destroyed by the episode 18 lava flow, which passed east of the subdivision and reached another sparsely populated area about a kilom- eter inland from the coast (pl. 1). The central-vent eruptions built large pyroclastic cones at the 1123 vent and at Puu 00. Subsequently, elders of the Hawaiian community at Kalapana named the 1123 vent “Pu‘u Halulu,” in reference to the chantlike sound they heard at Kalapana from the erupting vents. Infor- mal names, “Puu O,” and “O vent,” for the nearby major episode 3 vent (see figs. 1.8, 1.29), were coined because of its proximity in map position to the letter “O” in the label “Lava Flgw of 1965” on the topographic map then in use. The Kalapana elders subsequently chose “Pu‘u ‘O‘o” as the name for the very large new cone that grew at this locality, northeast of Puu Kamoamoa within Hawaii Volcanoes National Park. The ‘o‘o is a native Hawaiian bird, now extinct, that once lived in the eruption area. SIGNIFICANCE OF THE PUU 00 ERUPTION For most of its length, the eruptive-fissure system for the Puu Oo eruption is at the south edge of a 1- to Z-km- wide zone in which repeated eruptions occurred (pl. 1) from 1961 through 1969 (Richter and others, 1964; Wright and others, 1968; Moore and Koyanagi, 1969; Jackson and others, 1975; Swanson and others, 1979). Among the historical fissure systems of the middle east rift, only that of 1977 (Moore and others, 1980), which barely overlaps the east end of the Puu 00 system, is far- ther south. Through episode 1, the 1983 eruption was largely similar in style, duration, and eruptive products to these recent predecessors. The episodic style of central-vent eruption that devel- oped after episode 1 is unique in historical time in the middle east rift zone. However, it is strikingly similar to FIGURE 1.1.—Index map of Kilauea area, showing locations of features described in text. UTM, Uwekahuna tiltmeter; bh, borehole tiltmeters KMM (near Puu Kamoamoa) and KLU (near Kalalua). Inset shows locations of observation stations, designated camps A through E. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS E§m§_ m / m aEmo m2m3m5m2/ 33a / / mcmaflmx onmoEmz :5ch .o:2> w , _ _ J mmmhmfioqg 0.. m 0 Zl =<3<= mZON Huzm ka‘w 9 Q ’2 55.0 55.0 a Q axovo‘ )OO/ smnmz £398me :5 m: .2 B 05 .\/ . / 3.. ‘74 . at: Z \ / . Eta Ezozu %«' moEmoEmv. 33a . / Z/ n. . .IIIII. mYA\\A\1;#J / .330 _x_mw=m=x maufio® :EEsm mmawzx 5:3 foamiwwno ocmo_o> cmszmI I ‘ooamm— :onkoommw .m rom— :Om‘n—am— 4 the first stage of the 1969—71 Mauna Ulu eruption (Swan- son and others, 1979), which took place in the upper part of the rift zone. In 12 episodes during a period of slightly more than 7 months, from May through December 1969, Mauna Ulu produced an estimated 69 x 106 m3 of lava at a fairly steady average rate of approximately 10x 106 m3/mo (value uncorrected for vesicularity; Swanson, 1972). As at Puu 00, the eruptive episodes at Mauna Ulu were characterized by high fountaining and rapid lava discharge. However, average volumes, average recur- rence intervals, and average monthly lava—supply rates were smaller for Mauna Ulu (fig. 1.3). In addition, Mauna Ulu produced more pahoehoe than aa, whereas aa domi- nated after episode 1 at Puu 00. After 12 episodes with high or sustained fountaining, the Mauna Ulu eruption became dominated by a steady slow discharge that built the Mauna Ulu shield and transported large volumes of pahoehoe to relatively great distances from the vent by way of a system of lava tubes (Peterson and Swanson, 1974; Peterson, 1976; Tilling and others, 1987). In con— trast, the episodic style of the Puu 00 eruption has continued through episode 20 and beyond, producing one of the largest vent structures of Kilauea Volcano and creating an extraordinary complex of interleaved aa flows. Like the Mauna Ulu eruption, the Puu Oo eruption has THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 given us an unusual opportunity for systematic study of magmatic and eruptive processes. The Puu Oo eruption continued with remarkable indif- ference to tectonic and eruptive events on nearby Mauna Loa Volcano. A large (M = 6.7) earthquake (Koyanagi and others, 1984; Buchanan—Banks, 1987) occurred beneath Mauna Loa’s southeast flank on November 16, 1983, between episodes 11 and 12 of the Puu Oo eruption. Although this earthquake caused a displacement of near- ly 300 microradians at the Uwekahuna tiltmeter (fig. 1.2), it did not recognizably affect the continuing events at Puu 00. A major Mauna Loa eruption occurred from March 25 to April 15, 1984 (Lockwood and others, 1985), and, for the first time since 1919, the two volcanoes were erupting simultaneously. Episode 17, which occurred on March 30-31 during the Mauna Loa eruption, and episode 18, which occurred on April 18—21 after the Mauna Loa eruption, followed the normal pattern. SCOPE AND APPROACH Our intent is to provide a systematic and thorough account of our observations of the eruptive processes and products during the first 11/2 years of the Puu Oo erup- tion. Other chapters in this volume present concurrent (Inflation) West down East down (10 MICRORADIANS PER DIVISION) (Deflation) SUMMIT TILT, IN MICRORADIANS I I I I I l I : SUMMIT TILT Earthqum + 297.6 I : I I HARMONIC TREMOR PERIODS OF ERUPTIVE ACTIVITY I 1111. I I I 18 19 20 . 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 __1-_|_L__-__L-i #7:! -l .l .I J | l l—l—l .a _| _l I I I I I I I I I I I I I I I I I I DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT Nov DEC 1984 FEB MAR APR MAY JUNE MONTHS IN 1982—1984 FIGURE 1.2.—Relations of Kilauea summit tilt and middle—east—rift-zone harmonic tremor and eruptive activity from December 1982 through June 1984. Summit tilt measured by Uwekahuna tiltmeter; note brief period of reinflation during the large summit subsidence in early January 1983. Relative amplitude of harmonic tremor was approx- imated by averaging signal from a seismic station near eruption site for a 10-minute period once every 6 hours; breaks in plot record data gaps. During episode 19, brief periods of high tremor did not coin- cide with programmed sampling intervals. Occurrences of major erup- tive episodes (solid, full-height bars) and low-level volcanic activity (half-height bars) are shown for Puu 00 eruption, episodes 1 through 20 (numbers). 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 5 TABLE 1.1.—Times of occurrence, durations, areas covered, volumes erupted, and lava-discharge rates for episodes 1 through 20 of the Puu Oo eruption, 1983—81. [For episode 2, "low—level eruption" refers to a period of prolonged slow lava discharge (avg approx 1,400 m3/h) from February 10 to 25, 1983. For episodes 3 through 20, it refers to a period, preceding a major eruption, during which any or all of the following phenomena were observed at the active vent: intermittent gas—piston activity, intermittent spattering, occasional emission of small pahoehoe flows, and presence of a visible lava surface (sometimes partly crusted) in the conduit extending downward from the crater floor. Normally, low—level eruptive activity was confined to the interior of the crater; however, on rare occa- sions, short-lived pahoehoe flows spilled from Puu 00. The "beginning of vigorous eruption" was recognized in the following ways (order indicates priority): (1) direct observation of continuous emission of lava from the crater; (2) time—lapse camera record of beginning of continuous emission of lava from the crater; (3) where spillway activity was not recorded, time—lapse camera record of first appearance of fountaining above the crater rim; (4) beginning of rapid increase in harmonic-tremor amplitude. The "end of vigorous eruption" was recognized in the following ways (order indicates priority): (1) direct observation of cessa— tion of lava discharge; (2) time—lapse camera record of cessation of fountaining; (3) rapid decay of harmon- ic-tremor amplitude. Lava-discharge rate is calculated only for period of vigorous eruption. Except for episode 2, the volume of lava erupted during periods of low—level activity was negligible] First report Beginning End of Duration Lava- . of low-level of vigorous vigorous of Vigorous Area Volume discharge Episode . . . . covered eru ted eruption eruption eruption eruption (106 m2) (10 m3) Eat: (H.s.t.) (H.s.t.) (11.3.t.) (h) (10 m /h) 1983 11 -—- 0031 Jan. 3 ——— --- 4.8 14 ——— 22 1030 Feb. 10 0900 Feb. 25 1451 Mar. 4 174 2.7 14 8O 3 0600 Mar. 21 0100 Mar. 28 0257 Apr. 9 290 7.9 38 130 4 -—- 1025 June 13 1413 June 17 100 2.2 11 110 5 1000 June 29 1251 June 29 0715 July 3 90 3.4 13 140 6 0600 July 21 1530 July 22 1630 July 25 73 2.0 9 120 7 1610 Aug. 8 0741 Aug. 15 1600 Aug. 17 56 3.7 14 250 38 0900 Sept. 2 0511 Sept. 6 0526 Sept. 7 24 2.0 8 330 39 0009 Sept. 14 1541 Sept. 15 1920 Sept. 17 52 2.1 8 150 10 0800 Oct. 2 0106 Oct. 5 1650 Oct. 7 64 2.7 14 220 11 --— 2350 Nov. 5 1845 Nov. 7 43 4.3 12 280 12 1600 Nov. 29 0447 Nov. 30 1545 Dec. 1 35 3.0 8 230 1984 3’413 1100 Jan. 20 1724 Jan. 20 1123 Jan. 22 42 2.6 10 230 314 1123 Jan. 22 1745 Jan. 30 1318 Jan. 31 19 2.1 6 320 15 1000 Feb. 3 1940 Feb. 14 1501 Feb. 15 19 2.2 8 420 16 1900 Feb. 27 1450 Mar. 3 2231 Mar. 4 32 3.2 12 380 17 0910 Mar. 20 0448 Mar. 30 0324 Mar. 31 23 3.0 10 430 18 1340 Apr. 5 1800 Apr. 18 0533 Apr. 21 60 6.6 24 410 519 0830 Apr. 23 0500 May 16 0050 May 18 44 1.4 2 50 20 0800 May 18 2104 June 7 0625 June 8 9 1.6 4 480 1Episode 1 was characterized by intermittent eruptive activity for nearly 3 weeks. Periods of active eruption totaled approximately 99 hours. The last sizable event occurred on January 15; a minor one occurred on January 23. During episode 2, approximately 0.5x106 m3 of basalt was discharged during low—level eruption from Feb- ruary 10 to 25. The remaining 13.6x10 m was discharged during the period of vigorous eruption at a rate of approximately 70,000 m3/h- Flow was partly buried by younger basalt before it could be mapped in detail. Thus, uncertainty is greater in estimates of area, volume, and lava-discharge rate. Fountaining and flow production during episode 13 occurred in two main periods. The first, approximate— ly 31 hours long, was separated by about 5 hours from the 6-hour—long second period. Episode 19 was characterized by low fountain activity and intermittent low—volume overflows from the lava pond within Puu Oo Crater. This activity was interrupted by four 1— to 3—hour-long periods, totaling about 7 hours altogether, of higher fountaining and increased lava discharge estimated at about 100,000 to 200,000 m3/h. 6 geophysical, geochemical, and petrologic studies of the eruptive series, and an interpretative synthesis is given by Wolfe and others (1987). The section below entitled “General Observations” is a topical treatment of eruptive phenomena and products, dealing mainly with the central-vent eruptions of episodes 2 through 20; it is followed by a chronologic narrative of the first 20 episodes. Plates 2 through 20 show the vent deposits and the distribution, thickness, and sequential flow-front positions for the flows produced during each eruptive episode. In addition, graphic summaries of flow progress are given on plates 1 through 5. METHODOLOGY The eruptive zone is in a relatively inaccessible part of Kilauea’s east rift zone. Thus, we relied heavily on heli- copters for monitoring eruptive episodes and repose- period activity. Because of the high cost of helicopter support, we visited the eruption area only intermittently THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 during repose periods, and many times we were not on site at the beginnings of major episodes. We made intensive use of time-lapse cameras, which were kept running most of the time, to supplement our first—hand observations. The resulting film record was helpful in reconstructing the style and timing of activity at the erupting vents, and it provided the basis for the fountain-height summary plots (see figs. 1.21—1.24). Foun- tain heights were measured from images projected onto a computerized digitizing tablet. Scaling for the plots came from measurements made periodically by transit or theodolite, from calibration of the cameras with a target of known size and distance, or from measurement in 35—mm frames from a camera with a lens of known focal length. The timing of such events as the beginning of an eruptive episode varies in quality for several reasons: (1) poor visibility due to inclement weather sometimes interfered with the camera view; (2) not all timing marks were adequately recorded at both the beginning and end of a roll of film; and (3) at longer frame intervals (more than a minute), the camera timing mechanism was sub- 80 60 — 40 MAUNA ULU ERUPTION FIRST STAGE 20 IN MILLIONS OF CUBIC METERS CUMULATIVE ERUPTED VOLUME, MAY JUNE JULY AUG SEPT OCT MONTHS IN 1969 EPISODES 1 THROUGH 20 250 200 PUU OO ERUPTION 100 CUMULATIVE ERUPTED VOLUME, IN MILLIONS OF CUBIC METERS _ a 50 1 2 3 4 5 6 7 8 9 1O 11 12 131415 16 17 18 19 20— O IIIII | :I I I I l cl d cl :I I | I:I:I d :I IZII:L__—‘i l I I l J l l l I I I l I I l I I JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC JAN FEB MAR APR MAY MONTHS IN 1983 AND 1984 FIGURE 1.3.—Cumulative erupted volume, occurrence of major eruptive episodes (numbered solid blocks), and occurrence of low—level volcanic activity (open, half-height blocks) during first stage of 1969—71 Mauna Ulu eruption (Swanson and others, 1979) and episodes 1 through 20 of Puu 00 eruption. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 7 ject to drift that may have led to cumulative errors of several minutes over periods as long as 7 days. Although the times of events recorded in many time-lapse frames are accurate to within a minute or less, reasons 2 and 3 may have led to errors that we surmise could be as large as 10 minutes. A major effort was made to measure lava temperatures and to sample comprehensively. For the most part, we collected samples of melt and quenched them in water to minimize groundmass crystallization. During episode 1, many samples were of molten spatter from the fissure fountains. Later, the central-vent fountains were too dif- ficult to approach, and most samples came from active pahoehoe. Air-fall lapilli were also collected. Details of temperature measurements, sample collection, and chemical analyses are presented in chapter 2, and sam- ple studies in chapter 3. During eruptive episodes, the flow-front progress was monitored periodically by helicopter surveillance. Sketches of the flows at these times, many made under conditions of poor visibility because of smoke from the burning forest and, at times, bad weather, are the basis for the successive flow-front positions shown in plates 1 through 5. Within the Royal Gardens subdivision, observers on the paved streets tracked the flow-front progress directly. Flows and vent deposits were mapped with the help of aerial photographs taken as soon as possible after each major episode. In a few cases, one or two subsequent episodes occurred before the weather was suitable for aerial photography. The photographs were not of photogrammetric—mapping quality, and frames from several different flightlines were commonly required to create a single “pass” relatively free of clouds. Contacts were mapped on the photographs and transferred with an optical comparator to orthophotoquads. Because of planimetric distortions that resulted largely from im- perfect plumb of the camera and the paucity of distinct photoidentifiable features in some of the rain forest, mapped contacts may be locally in error, probably by no more than 100 m at worst and normally much less. The mapping was field checked and is represented in the detailed maps of plates 1 through 5; these maps have been fitted to the published 1224,000-scale topographic quadrangle maps. Thicknesses were measured by handlevel at many points along the flow margins. We tried to avoid places where the observed thickness seemed abnormal—for example, where the flow edge had preferentially piled up against a kipuka, or where there was a large evacuated central channel. The measured thicknesses are shown on plates 1 through 5; they have been omitted for episode 1 because the episode 1 flows in the rift zone are a complex of overlapping flows that are not sorted out on plate 1. The area covered during each episode was measured from the detailed 1124,000—scale map, and the volume of lava was calculated using the measured thicknesses. The lava-discharge rate listed in table 1.1 is the measured volume divided by the number of hours of vigorous eruption. Acknowledgments—The entire staff of the Hawaiian Volcano Observatory (HVO) were involved in field obser- vations, operational support, and other studies (reported elsewhere in this volume) related to the ongoing eruptive activity. We especially thank Jim Griggs, indefatigable staff photographer, for his extraordinary helpfulness and skill in aerial photography, eruption photography, endless darkroom work, and preparation of illustrations. U.S. Geological Survey (USGS) colleagues Jane Buchanan- Banks, John P. Lockwood, Richard B. Moore, and George E. Ulrich, delaying their own project work, each spent many days observing and sampling during eruptive episodes or assisting in eruption-related work during repose periods. Other USGS colleagues from the mainland, as well as university students and professors, in particular from the University of Hawaii, came to help, study, and learn. Our helicopter pilots brought a high level of skill, enthusiasm, and professionalism that contributed immeasurably to the ease of our work. We also received continuing cooperation and assistance from the person- nel of Hawaii Volcanoes National Park. Lu Setnicka’s clerical skill, attention to detail, and unending willingness and good humor were indispensable in completing the manuscript. Many friends invested long days and nights; their enthusiastic support helped us and made our work more fun, and we thank them. Continuing discussion with L.P. Greenland has been invaluable in focusing our thoughts and clarifying our ideas about some of the erup- tive phenomena. 500 I | I I I 400 — ’ — . I 6 300 - — 200 — — LAVA-DISCHARGE HATE, IN THOUSANDS OF CUBIC METERS PER HOUR I :3 0 I I I I I o 1 00 200 300 400 500 DAYS FROM BEGINNING OF 1983 FIGURE 1.4.—Average lava-discharge rate for each eruptive episode (numbers) versus time since the beginning of 1983. 8 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 GENERAL OBSERVATIONS DEVELOPMENT OF VENT STRUCTURES In contrast to the low, linear spatter ramparts that formed during the fissure eruptions of episode 1, the central-vent eruptions of episodes 2 through 20 built large, cratered cones composed of agglutinated spatter, Cinders, and rootless flows. The dominant eruptive style was one in which a broad columnar fountain or one or more small fountains rose above a lava pond within the crater. The major flow was fed by a lava river overflowing from the pond through a low point in the crater rim (fig. 1.5). The central vents quickly became at least partly en- circled by spatter rings built of fallout from the fountain, and further development of the vent structures through successive eruptive episodes resulted largely from upward and outward growth of the original ring structure. Asym- metry occurred, mostly because the prevailing northeast wind tended to concentrate the fallout on the leeward side of the fountain. Growth of the cones resulted primarily from accumula- tion of molten spatter fragments, commonly as large as 1 m in diameter, that formed a fragile agglutinate which built steep and commonly unstable crater walls and cone FIGURE 15—1123 vent erupting during episode 2. Fountains are erupt- ing through a nearly circular lava pond, about 60 m in diameter. A thin nonincandescent skin covers most of pond surface. Base of main fountain is about 40 m Wide. High part of enclosing rim is a steep ram» part of agglutinated spatter, formed from fallout that accumulated on downwind side of fountain. Low part of ring, though partly of pyroclastic origin, is mainly a pahoehoe bulwark built by overflows and by lava that leaked from enclosed pond through short lava tubes. A voluminous lava river debouches from pond through a low place in south rim of enclosing levee. Western part of recently active 0740 cones (at left) is glowing and emitting fume. Fume near right edge of photograph is from easternmost of a discontinuous line of small vents extending uprift from main vent. View southward; photograph by J .P. Lockwood, taken during the afternoon of March 1, 1983. flanks. Fountains higher than about 100 m produced abun— dant fine-grained tephra. At these times, accumulating, highly vesicular air-fall lapilli and bombs (fig. 1.6) formed thick local deposits that mantled the leeward flanks of the cones and extended as thin sheets a kilometer or more downwind. Fine tephra wafted upward by thermal cur- rents commonly ascended hundreds of meters above the higher fountains, and Pele’s hair was carried as far as 45 km from the vent. Changing fountain trajectories commonly caused brief, intense bombardment of a sector of the cone’s flank by spatter. Short-lived, spatter-fed pahoehoe flows normal- ly resulted (fig. 1.7), most of which were subsequently buried by agglutinated spatter or loose tephra. GROWTH OF PUU 00 Although fissure vents had erupted at or near the site of Puu 00 during episodes 1 and 2, Puu Oo emerged as a distinct central-vent edifice during episode 3. The pro- genitor of Puu 00 was a pair of juxtaposed spatter rings enclosing low fountains (fig. 1.8) about 50 m west of the vigorously erupting 0 vent, which had also been active in this area for part of episode 2. By the second day of episode 3 (March 29, 1983), these two spatter rings had coalesced to form a single crater with a low fountain that erupted weakly along with the more vigorous O vent until both shut off on March 30. Episode 3 activity continued for several more days at the 1123 vent. Beginning with episode 4, the newly formed Puu 00 was the sole eruptive locus except for nearby, short-lived fissure vents that erupted weakly along with Puu Oo dur- ing episodes 4 and 11. By the end of episode 4, the main part of Puu 00 was a steep-sided, truncated cone, about 20 to 30 m high and 100 m wide (fig. 1.9); a smaller, satellitic cone had formed on its west flank, where episode 4 fissure vents impinged on the main cone. A steep-walled crater, about 20 m deep, was enclosed within the main cone. During episodes 4 and 5, lava overflowed the crater through a narrow breach that formed a spillway in the south rim and fed flows that extended into the western part of the Royal Gardens subdivision. Near the end of episode 5, part of the northeast rim slumped suffi- ciently to create a low point over which episode 6 lava spilled. This initiated the development of a deep, steep- walled breach that for many months provided egress for lava during eruptions and ingress for geologists between major eruptive episodes. With each successive episode, the cone grew higher and broader (fig. 1.10). Asymmetry was perpetuated by maintenance of the breach, which formed a low place in the northeast rim (fig. 1.11), and by preferential accumula- tion of spatter and air—fall deposits to the southwest, 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY 0F ERUPTIVE EVENTS r’.’ W? or»... FIGURE 1.6.—Puu Oo vent during episode 16. Fountain is 300 to 400 m high. Dark tephra, falling out to northeast, raises a cloud of dust where it strikes flank of Puu 00. Additional fine tephra, visibly incandescent at night, is wafted above denser fountain by updrafts. Pulsating supply to fountain produced upward-surging fronts marked by zones of brighter incandescence. View eastward from camp D; photograph taken at 1659 H.s.t. March 3, 1984. 10 elongating the cone in that direction and forming a topographic peak in the southwestern part of the rim crest. In addition, production of thick spatter-fed flows FIGURE 1.7 .—Molten spatter fragments falling from episode 9 fountain and a thin, fluid, transient spatter-fed flow that formed moments earlier when fountain trajectory was inclined over northwest (right) flank. A still-cooling, earlier spatter-fed flow armors north flank (center). Older thick, spatter-fed aa that extends outward from base of cone forms steep flow front in foreground. Photograph taken at 1400 September 17, 1983, from 100 m north of the cone. Numerals (lower right) indicate date and time. FIGURE 1.8.—Fountains and flows at and near 0 vent early in episode 3. Line of fountains is 100 to 150 m long. 0 vent is at left. Pair of spatter rings in center, each with a low fountain, is at site of Puu 00; they coalesced Within a day to form a single cone and crater. View southward; photograph taken at 0811 H.s.t. March 28, 1983. THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 created a prominent shoulder adjacent to the south flank of the cone (fig. 1.12). Occasionally, gravitational collapse resulted in a net decrease in height in some sections of the cone from one eruptive episode to the next. At the end of episode 19, the crater-rim crest was about 120 m above the local preeruption surface (fig. 1.11). The steep- Walled cone, with average outer slopes of about 30°, was approximately 320 m in maximum diameter and sur- mounted a broad, more gently sloping apron composed of flows and local air-fall deposits. Visible from many distant vantage points, Puu 00 had become the most im— posing topographic feature in the central part of Kilauea’s east rift zone. As the cone grew during successive eruptive episodes, the enclosed crater broadened. Normally, it was approx- imately circular in plan (fig. 1.13), except for a pronounced northeastward bulge during the later episodes from accentuated flaring of the rim adjacent to the spillway. After episode 4, the nearly circular rim crest was 40 to 50 m in diameter. After episode 19 (fig. 1.11), the max- imum diameter of the rim, measured northwest-southeast, was about 120 m, and the distance from the southwest rim crest to the spillway was about 150 m. Episodes 5 through 10 each left a main crater with a relatively flat floor, about 30 m in diameter. (A satellitic crater on the west flank of Puu Oo erupted during episode 5, in addition to the main crater.) The floor was covered with coarse breccia (fig. 1.14) composed of angular blocks from the walls, which rose steeply upward. FIGURE 1.9.—Puu 00 from floor of evacuated southeastern channel after episode 4. Main Puu Oo vent is marked by large spatter cone that stands about 20 m high and encloses a crater about 40 min diameter. Lava feeding main flow spilled through narrow breach visible in south crater rim and formed steep-walled channel in foreground. Satellitic, western Puu Oo cone is visible on left shoulder of main Puu Oo cone. View northward; photograph taken at 1045 H.s.t. June 29, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 11 Our impression was that superficial collapse of the steep walls must have occurred as lava withdrew into the con- duit at the end of an eruptive episode. Normally, after these early episodes, no conduit opening was visible through the rubble. However, two closely spaced, open, vertical pipes, 3 to 4 m in diameter, were visible descend- ing from the crater floor after episodes 8 and 9. Their walls exposed the edges of thin layers of platy and rubbly basalt that probably record successively buried levels of the crater floor (fig. 1.14). The most profound intracrater collapse occurred at the end of episode 10. The resulting crater interior was a com- plex of blocks, including slices as much as tens of meters long, of the crater walls that had collapsed into the depres- SE 850 800 ......... No vertical exaggeration 750 700 . l . I NW Approximate profile of post-episode 4 cone Approximate pre-1983 ground surface (see fig. 1.11) I I I J 300 400 850 800 APPROXIMATE ELEVATION ABOVE SEA LEVEL, IN METERS 7 50 ..... .’././ _____ \ Tow—._._._.~. _ ’ ’ ’ ’ ~ ‘ ‘0‘ %_'—° . , / ’ T No vertical exaggeration ~ ‘9‘ ‘ "‘ ’ , _. _ a ’ a - 6 Approximate pro-1983 ground surface B (see fig. 1.11) 700 ~ I I 1 1 . 1 O 100 200 300 400 APPROXIMATE HORIZONTAL DISTANCE, IN METERS FIGURE 1.10.—Superimposed profiles of Puu 00. A, Viewed from the northeast after episodes 12, 14, 16, 18, and 20. Profiles (identified by episode number) were traced from photographs taken from camp E, located 1.3 km northeast of Puu 00 (see fig. 1.1). After episode 20, summit of Puu 00 stood approximately 130 m above preexisting ground level. B, Viewed from the west after episodes 6, 8, 9, 10, 12, 14, 16, 18, and 20. (Profiles identified by episode number; additional profiles were omitted to preserve clarity of outlines.) Profiles were traced from time—lapse-film frames taken from near camp D, located 750 in west of Puu 00 (see fig. 1.1). Owing to frequent camera moves and several camera focal-length changes, perspective and scale con- trol vary. Thus, control is weak for the precise placement of episode 9 profile relative to episode 8 profile, of episode 14 profile relative to episode 12 profile, and of episode 18 profile relative to episode 16 profile. Scale was determined from photographs of known focal length and distance from cone. In addition, periodic transit measurements of maximum cone height and two detailed topographic surveys served to constrain superpositions. 12 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 sion (fig. 1.15). Grooved and slickensided surfaces low on the inner crater walls indicated that the crater floor had undergone at least 5 to 10 m of subsidence at the end of episode 10. Episode 10 marked a temporary change in the style of venting at Puu 00. The conduit system, which had been delivering lava through the floor of Puu Oo Crater, in- creased in complexity by developing distinct secondary vents at the base of the northwest crater wall and on the west rim. Still greater complexity developed during episodes 11 and 12, when several vents erupted on the rims and flanks of Puu 00, as well as within the crater (fig. 1.16). In addition, lava was slowly erupted during episode 11 from fissures that extended about 200 m uprift and downrift from Puu 00 (pl. 3). The complexity of vent distribution may reflect partial blockage of the shallow crater-floor conduit. Apparently, the blockage began dur— ing episode 10 and was compounded by subsidence of the crater floor and collapse of the interior walls at the end of that episode. Thus, an enlarged, compartmented, and distorted crater formed (fig. 1.17). Episode 13 reestablished a single bowl-shaped, open crater (fig. 1.18). An open conduit that was to remain a permanent feature through subsequent episodes de- scended from the crater floor. After episode 13, this con- duit tapered from about 20 m in diameter at the top to about 10 min diameter at a depth of 25 m, where further visibility was blocked by lava standing in the pipe. From episode 14 on, as much of the pipe as we could see (approx 50 m at most) was vertical, nearly cylindrical, and about 20 m in diameter. The crater floor was broader and more basinlike after episode 14 (fig. 1.13) than in earlier episodes. The basin was largely enclosed by steep crater walls, which, until O 0 «b: I I I O I O x O O . . Camp D \ / / / __ —-{740)/ 0 0.5 1 KlLOMETER I Contour interva| 10 meters FIGURE 1.11.—Topographic map of Puu 00 after episode 19. Map was compiled by R.Y. Hanatani from a survey on May 22, 1984. Dots denote locations of surveyed control points. Hachured line is edge of 1983-84 basalt at time of survey; hachures are on 1983—84 basalt side. Contours on buried pre-1983 surface (dashed, values in parentheses) modified from US Geological Survey Kalalua 7.5-minute quadrangle, 1982. For location, compare with figure 1.1. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 13 episode 19, towered about 40 m above the crater floor. The spillway through the northeast crater rim persisted, and a low arcuate ridge, about 5 m high, separated the shallow basin from the spillway (fig. 1.13). The remarkable FIGURE 1.12.—Puu 00 between episodes 17 and 18. Gone is about 100 m high. Pronounced bulge at right reflects repeated deposition of spat- ter and air-fall deposits and development of thick, spatter-fed flows. View eastward from camp D; photograph taken at 1046 H.s.t. April 15, 1984. O 50 L—_4____A 100 APPROXIMATE METERS FIGURE 1.13.—Puu 00 after episode 14. Bowl-like crater is about 40 m deep. Open, nearly circular, vertical conduit intersecting crater floor is about 20 m in diameter; after episode 13, it was a permanent feature. Photograph by Jay Whiteford, Air Survey Hawaii, taken February 11, 1984. persistence of the geometry and general dimensions of the crater, with almost no change from one repose period to the next until episode 19, suggests that a relatively stable equilibrium between eruptive processes and crater formation had been attained. Minimal collapse of the crater walls after some of these later eruptive episodes reinforces this suggestion. Although significant failure of the walls followed episodes 16 and 18, the walls and floor FIGURE 1.14.—Angular blocks of oxidized agglutinate from crater walls surround an open vertical, 4—m—diameter pipe descending from floor of Puu Oo Crater after episode 9. Photograph taken at about 1000 H.s.t. September 23, 1983. FIGURE 1.15.—Post-episode 10 crater, about 90 m in diameter at the rim. Floor of crater is covered with rubble and collapsed blocks from walls and rim of cone. Fuming areas at base of north crater wall (foreground) and crest of west rim (right) were near sites of promi- nent secondary fountains during episode 10. Open cracks in fuming area high on west rim remained incandescent between episodes 10 and 11. Spillway (S), crossing northeast rim, is choked with rubble that was apparently transported into spillway by last lava exiting crater. View southeastward; photograph taken at 1501 H.s.t. October 30, 1983. Faint numerals (lower right) indicate date and time. 14 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 were largely mantled with smooth, glassy basalt after episodes 13 through 15, and 17. Unlike the preceding episodes that had been character- ized by steady, voluminous lava emission and continuous vigorous fountaining, episode 19 was characterized by intermittent, low-volume overflows from the basin within Puu 00 and only a few brief periods of increased lava emission and vigorous fountaining. The effect of this relatively low level eruptive episode was to nearly fill the crater of Puu 00 with solidified basalt (fig. 1.19), a change that led to elimination of the basin and the associated long, lava-river-fed flows. For months thereafter, higher foun- tains and shorter, thicker, spatter-fed aa flows dominated. LAVA PONDS AT THE CENTRAL VENTS During the central-vent eruptions, lava supplied from beneath the crater floor commonly formed a pond (figs. 1.5, 1.20) that, along with one or more fountains, filled the shallow closed basin within the growing pyroclastic ring. Between eruptive episodes, when no lava was in the crater, we could see that the low point in this ring, which formed a spillway through which the pond overflowed, generally was about 5 to 15 m above the crater floor. At Puu 00, the breach through the crater rim formed a nar- row constriction, so that the surface of the overflowing pond was several meters higher than the rock floor of the spillway, and at times the pond surface appeared to in- cline gently toward the spillway. We estimate that the pond was normally at least 10 to 20 m deep. As Puu Oo slowly grew larger during successive eruptive episodes, FIGURE 1.16.—Puu 00 vent erupting during episode 12. Four general loci of fountaining are visible: (1) spillway area, on northeastern flank; (2) southeast rim of crater; (3) high in northwest sector of Puu 00 (the most vigorous, partly obscured by fume); and (4) low fountain within crater. View westward; photograph taken at 0952 H.s.t. November 30, 1983. Numerals (lower right) indicate date and time. the diameter of the enclosing basin, which, measured at the level of the spillway, gives an approximate limiting diameter for the pond, increased from about 20 to 100 m. FIGURE 1.17.—Puu 00 after episode 12. Irregular shape and compart- mentation of crater were caused by simultaneous eruption of several distinct fountains (fig. 1.16) within crater and on its rims and flanks. Spillway and adjacent evacuated channel are at upper right (arrows). Episode 13 eruption began with gradual filling and overflow of deep compartment adjacent to spillway. View southwestward; photograph taken at 0942 H.s.t. January 10, 1984. Numerals (lower right) indicate date and time. FIGURE 1.18.—Puu 00 after episode 13. Simple, nearly circular crater contrasts markedly with complex crater that preceded episode 13 (fig. 1.17). Lava surface is visible about 25 m down open vertical conduit, which tapers from about 20 m in diameter at the crater floor to 10 m in diameter at lava surface. View southwestward; photograph taken at 1156 H.s.t. January 24, 1984. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUP’I‘IVE EVENTS 15 During Vigorous eruption, lava poured through the breach in the crater wall and normally fed a well- channelized river of pahoehoe. When fountaining was relatively low, we could sometimes see that the smooth lava surface of the spillway extended Without disruption into the crater interior (fig. 1.20), so that the immediate source of the pahoehoe river was the lava pond within the FIGURE 1.19.—Puu Oo Crater before and after filling by episode 19 pahoehoe. A, Deep, bowl—shaped crater after episode 18. Walls and floor are mantled by slumped debris that had temporarily plugged the open, 20-m-diameter pipe. Spillway and steeply inclined chute (see fig. 1.91) descending outer flank of cone are at left. Shorter crater diameter is about 100 m. Arrow marks level to which solid basalt filled crater during episode 19. Helicopter strut and skid visible in lower right corner. View southeastward; photograph taken at 1355 H.s.t. April 23, 1984. B, Puu 00 after episode 19. Deep crater and breach in north- east rim were largely filled during episode 19 by successive pahoehoe overflows from central conduit. Arrow denotes same slumped ledge of crater-rim material as in figure 1.19A. View southeastward; photograph taken at 1056 H.s.t. May 30, 1984. Numerals (lower right) indicate date and time. crater. The relative peacefulness of this pond surface where not turbulently disrupted by the fountain indicated that little, if any, effervescence was occurring and that the pond lava had sufficient strength to maintain its shape adjacent to the vigorously turbulent fountain. Apparent- ly, the lava in the quiet part of the pond had lost most of its dissolved magmatic gas. At times, high, broad, ash-laden fountains issued from the Puu Oo Crater (fig. 1.6), particularly during the early stages of some eruptive episodes. Because of the breadth of these fountains, we could not see the pond at such times. Continuing passage of the lava river through the spillway indicated that melt was accumulating in the basin, but flows produced under such conditions were sometimes more sluggish than when a well-developed pond was present (see subsection below entitled “River- Fed Flows”). FOUNTAINS AT EPISODE 1 FISSURE VENTS Linear fountains, from tens to several hundreds of meters long, erupted from the episode 1 fissure vents. The FIGURE 1.20.—Puu 00 erupting during episode 7. Smooth surface of pond fills crater from base of fountain to spillway, where lava cascades down outer flank (lower left) and supplies main flow. A small tube- fed flow, apparently also fed by lava of the pond, issues from north base of Puu 00 (bottom center). Spatter from 50-m—high fountain feeds flows to southwest (upper right) and northwest (right center). A satellitic crater on west flank is partly visible at upper right; it con- tained an active vent during episode 5 but is not actually erupting here. At times, spatter falling into satellitic crater was so voluminous that a temporary pond formed and overflowed to supply additional lava to southwestern flow. View southward; photograph taken at 0850 H.s.t. August 15, 1983. Numerals (lower right) indicate date and time. 16 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 fountains were mostly low, about 5 to 30 m high. A notable exception occurred at midday on January 7, 1983, when a fountain, 60 to 100 m high, issued steadily from a 500-m-long fissure for a period of about 2 hours. This high fissure fountain occurred simultaneously with extra- ordinarily voluminous lava discharge that was estimated later from the measured flow volume to be 1x106 to 1.5x 106 m3/h. In contrast, normal episode 1 discharge rates ranged from about 0.06x 106 to 0.3 x 106 m3/h and averaged about 0.1 x 106 m3/h. Clots of molten spatter, as large as 1 or 2 m in diameter, fell out of the episode 1 fissure fountains and built low ramparts adjacent to the vents. Very little tephra was transported downwind away from the erupting fissure vents. FOUNTAINS AT CENTRAL VENTS After episode 1, most of the eruptive activity was localized at central vents. The main locus of eruption dur- ing episodes 2 and 3 was the 1123 vent; from episode 4 on, Puu 00, including, on a few occasions, nearby fissure vents, was the sole eruptive locus. In an attempt to characterize the central-vent fountain behavior and in hope of gaining insight into the relations between foun— tain behavior and other eruptive parameters, such as discharge rate and vent geometry, we measured the foun- tain heights in time-lapse film records of the central-vent eruptions (figs. 1.21—1.24). EPISODES 2 AND 3: FOUNTAINS AT THE 1123 VENT Eruption at the 1123 vent during episode 2 began dur- ing the afternoon of February 25, 1983. A spatter ring enclosing a pond with a conspicuous main fountain and a line of discontinuous low fountains (fig. 1.5) quickly formed. For the first 2 days of its activity, this main foun- tain was relatively low, generally about 20 to 40 m above the surface of the lava pond (fig. 1.21). By the morning of February 26, one short flow had been erupted, at a lava- discharge rate of about 40,000 m3/h. The major eruptive activity, mostly accompanied by a 60- to 80-m-high foun- tain at the 1123 vent, began on February 27 and continued until March 4, when episode 2 ended. Lava was discharged at an average rate of about 90,000 m3/h, double that for the earlier flow accompanied by a low fountain. Apparent- ly, this change in fountain vigor reflected a change in the rate of delivery of magma and, thus, of newly exsolved gas to the vent. During episode 3, two major vents were active at the 1123 cone. The northeastern fountain, which was by far the higher (fig. 1.25), was the source of our fountain- height record (fig. 1.23). It erupted within a reentrant in , the growing cone but for most of episode 3 was not con- tained within a distinct crater. There seemed to be little, if any, development of a lava pond. Much of the time, the lower, southwestern fountain, which was within a bowllike crater, was obscured from our View by the growing cone. Generally, the two fountains seemed to behave indepen- dently. We saw no correlation or anticorrelation between them in relative vigor except for a 24-hour period, ending at 2100 H.s.t. April 3, during which emission of spatter was interrupted for short intervals at both vents (see subsection below on episode 3). Lava from the northeastern vent supplied a massive aa flow throughout episode 3 at an average rate of about 75,000 m3/h. Through April 3, little significant lava pro- duction occurred at the southwestern vent. Early on April 4, however, the southwestern vent began to dis- charge lava flows at an average rate of about 130,000 m3/h, continuing until the end of this eruptive episode. A vigorous, well-channelized lava river issued from the crater at the southwestern vent, and although we never got a direct View, we believe that the flow originated from a lava pond within the crater. We judge from the con- tinued steady advance of the flow from the northeastern vent (pl. 1) that the discharge rate there was not signif- icantly affected by the change at the adjacent southwest- ern vent, which resulted in a more than twofold increase in the combined discharge rate. In spite of this striking increase in discharge at the southwestern vent, the foun- tain there remained low. The northeastern fountain, however, greatly increased in vigor (fig. 1.23). Whereas it seldom exceeded 100 min height before April 4, from April 4 through 7 the fountain was higher than 100 m most of the time, with peaks exceeding 200 m, and tephra falls extending more than a kilometer from the vent. This increase in fountain height at the northeastern vent ap- parently reflected increasing overall supply of magma and, thus, increasing delivery of newly exsolved gas to the shallow part of the conduit. As discussed below, we believe that the low fountain within the bowllike crater at the southwestern vent may have resulted from damp- ing by the degassed lava that formed a pond within the crater. EPISODES 4 THROUGH 20: FOUNTAINS AT PUU 00 The fountains at Puu 00 generally emanated from the basin in the lower part of the crater within the growing pyroclastic cone. Exceptions, described below, occurred during episodes 4, 5, and 10 through 12. The range in fountain height was large, from a few tens of meters to nearly 400 m. Unlike the fountains at the 1123 vent during episodes 2 and 3 (figs. 1.21, 1.23), the Puu Oo fountains were com- 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 17 monly highest in the early hours of an eruptive episode hours. Many of the early high fountains, as well as inter- (fig. 1.24). As recorded in the fountain data for episodes mittently high fountains later in an episode (such as those 4, 6, 9, 10, 14 through 18, and 20, the general height, on October 6, 1983, during episode 10 or on March 3, though sometimes varying widely, decayed after the initial 1984, during episode 16) were nearly vertical and broad 200 JlIlllllJlllllllllllllllllllllJllllIlllllllllllllllllllllll'llllllllllllllllllLllllIl 100 Illlllllllllllllllll (D E ,_ 0 Lu 2 25 26 27 28 1 % DAYS IN FEBRUARY AND MARCH ’— I g 200 Illlllllllllllllllllll lllllllllllllIlllllLlllllIllllllllllllillllllll Illlllllllllll I l- l. 100 .5: E- g_ I).— 5-— o_ ‘0- C- Lu— 0 .— DAYS IN MARCH FIGURE 1.21.—Fountain heights of main fountain at 1123 vent during episode 2. Data from a time-lapse camera located at camp A, about 500 m northeast of vent. Datum is approximate base of fountain. Apparent zero values normally reflect data gaps resulting from bad visibility or an inoperative camera; however, they include inter- mittent cessations in fountain activity on March 4. Data plotted continuously from 1200 H.s.t. February 25 to 1500 H.s.t. March 4, 1983. 200 JlllllllllllllllllllllllllllllllllllIlllllllllllllllllllllll HEIGHT, IN METERS a O Illllllllllllllllll 28 29 30 DAYS IN MARCH FIGURE 1.22.—Fountain heights at O vent during episode 3. Record begins at 1322 H.s.t. March 28, approximately half a day after beginning of episode 3. Data from a telephoto time-lapse camera at camp A. Datum is approximate base of fountain. Apparent zero values reflect data gaps resulting from bad visibility or an inoperative camera. Data plotted from 1130 H.s.t. March 28 to 2330 H.s.t. March 30, 1983. 18 HEIGHT, IN METERS THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 200 IIIIIlllllIlIIlIllIllIIIlllIlllllllllllllllllllllIlllllllIlllllllIllIllllllllIllllllIlll 100 29 30 31 1 DAYS IN MARCH AND APRIL 300 IlllIlllllllllllllllllllll llllllllllllllllllllllllllllllllllllllllIlllllllllllllllllllJ 200 llllllllllllllllll 5 o 1 I l 1 2 3 4 5 DAYS IN APRIL 300 |lIlIllllllllllllIIlllIlllIIIllllllllllllIIllIlllIllIlllllllllllllllllIlIllIIllllIllllll 200 100 [IlllIIIllllllllllllllllllllll 7 8 9 DAYS IN APRIL FIGURE 1.23.—Fountain heights of northeastern fountain at 1123 vent during episode 3. Data from a time-lapse camera located initially at camp A and later at camp B. Datum from camp A is approximate base of fountain. Base of foun- tain was not visible from camp B, and so fountain heights shown after 1000 H.s.t. April 1 are minimums. Fountain- ing was intermittent before 1800 H.s.t. March 29. Other data gaps (circled numbers) are explained as follows: 1, Data were lost because of poor visibility or inoperative camera. 2, Camera was moved from camp A to camp B. 3, Fountain was very low, and visibility was partly obscured by fume or water vapor; record of a nearby telephoto camera indicates that low spatter was continuously visible. 4, Fountain was so low that it was out of view at times, and visibility was partly obscured by bad weather. Data plotted continuously from 0500 H.s.t. March 29 to 0500 H.s.t. April 9, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 19 based. The highest ones, generally at least about 200 m high, formed majestic, candlelike pillars that commonly were partly shrouded in a conspicuous cloud of fine tephra (fig. 1.6). Although such fountains generally remained steadily high, the incandescent ejecta that supplied them was emitted in pulses. The spacing between pulses was approximately 1 s; thus, several successive, more bright- ly incandescent fronts could be seen rising upward through the fountain at any one time. Occasionally, the high fountains became abruptly inclined, or they rapidly disintegrated in a confusion of erratic inclined jets that sprayed ejecta over the outer flanks of Puu Oo—as if an obstacle interfering with the upward trajectory had deflected or fragmented the fountain. The lower fountains, typical through the greater part of each episode, generally behaved more irregularly. Many of them seemed to be a composite of short-lived individual jets (fig. 1.26) originating in bursts from different parts of the lava pond that was sometimes visible within the crater. At times, these bursts seemed rhythmic, with new jets rising at intervals from about 1 to 3 s. Possibly because of alternating interference and reinforcement among the jets, the overall fountain height commonly oscillated several times per minute over a large vertical range. Average and maximum fountain heights determined at Puu 00 are shown (fig. 1.27) from the data plotted in figure 1.24. With significant exceptions, discussed below, both values increased during the series of eruptive episodes, and the production of a fountain several hun- dred meters high continued long past episode 20 as a hallmark of the Puu Oo eruptions. ‘Fountain height increased fairly steadily from episodes 4 through 10. The extremely low heights for episodes 4 and 5 may partly reflect the fact that the main Puu Oo Crater was not the sole eruptive locus in those episodes. A line of erupting fissure vents extended a short distance uprift from Puu 00 during the early part of episode 4. Discharge from the westernmost vents soon ceased, and the more easterly ones formed a satellitic vent that was active on the west flank of Puu 00 through episode 5. During the second day of episode 10, high fountaining was abruptly curtailed (fig. 1.24), and new discrete vents formed at the base of the northwest interior crater wall and on the high west rim of the crater. As already sug- gested above in the subsection entitled “Growth of Puu 00,” this event may have recorded partial blockage of the established vent through the crater floor, and the effect may have been exacerbated at the end of episode 10, when subsidence of the crater floor and profound collapse of its walls occurred. Low fountains and complex vent distribution in and near Puu 00 during episodes 11 and 12 indicated that the inferred blockage continued at least through those episodes. Conduit blockage and develop- ment of a complex of subsidiary passages for transmis- sion of lava to the surface seem adequate to explain the diminished fountain heights during and after episode 10. Episode 13 marked a return to a single vent within the crater, as represented by the open pipe that we could see descending from the central part of the crater floor dur- ing all the repose periods after episode 13. Fountain vigor increased from episode 13 through 16, and early high, tephra-laden fountains like those of episodes 9 and 10 resumed in episodes 15 and 16. One could speculate (fig. 1.27) that episodes 15 and 16 marked a return to the trend of evolving fountain behavior that was established during episodes 4 through 10 and interrupted in the later part of episode 10. If so, other perturbations apparently followed episodes 16 and 18, although they are less readily explained. As in episode 10, the crater walls failed markedly at the ends of episodes 16 and 18. However, the disruption within the crater was less severe than after episode 10; and in the ensuing episodes, no subordinate vents opened to suggest that the shallow conduit had been blocked. Whether collapse of the episode 16 and 18 crater walls was a coincidental, super- ficial event or was somehow related to diminished fountain height in the ensuing episode is unclear. Further- more, the episode 19 eruption was abnormal in both volume and style. It was unusually small: Only 2 x 106 m3 of lava was discharged. Unlike the other episodes, episode 19 was characterized by repeated intermittent, low-volume overflows from Puu Oo Crater; these were punctuated by four brief periods of higher volume lava production and moderate fountaining that were, in turn, interrupted by repeated, brief, abrupt pauses in discharge—it seemed like an eruptive episode that could not get fully under way. The trend toward increasing fountain vigor through the series of eruptive episodes at Puu Oo probably reflects evolution of the lava-delivery system. Generally, the rate of lava discharge increased through the 20 episodes (fig. 1.4). Although the relation is complex in detail, we conclude that increasing fountain vigor at Puu 00 was related, at least in part, to increasing discharge rate. We concluded similarly for the episode 2 and 3 fountains at the 1123 vent. L.P. Greenland (oral commun., 1985; see chap. 4) sampled vent-gas emissions from episode 1 through the repose period between episodes 15 and 16. He found that the composition of erupted magmatic gas was constant during that period and inferred that the initial concentration of volatile constituents dissolved in the magma was also constant. Therefore, changes from one episode to the next in the general height of the foun- tain were apparently not related to changes in initial magmatic-gas content. Variation in fountain height might also have been related to changing geometry of the erup- tive orifice. For episodes 14 through 20, however, dur- 20 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 100 IllllllllllllilllllllllIIllllllllLllllIlllllllllllllllllllllllllllllLllllllllllllllllIllllllLlllIlIllllll Episode 4 ® 0 “0° 13 14 15 16 17 230° DAYS IN JUNE1983 100 lllllllllllIllllIIILJJIIJLLJIIlllllllllllllllllllllllllllllllllllllllllllllllllllllllilllllllil : Episodes ® : o__ __ m 1000 29 3° 2 3 1000 a: “I: DAYS IN JUNE AND JULY 1983 Lu 2 E 200 II|ILJIIIIIIllllllllllilllllilllll'llllllllllllllllllllllllllllllllllllllllll 5-: Episode6 C B I 100 O Ll_Llllllllllllllllll :2 o o IIIIIITIIIIIIIIIIIII 25 1900 22 23 24 DAYS IN JULY 1983 100 llllllllllllllllll ||llllll|llllllllllllllllllllllllllllllll Episode 7 llllllll l o | 0500 15 16 DAYS IN AUGUST 1983 FIGURE 1.24.—Fountain heights at Puu 00 during episodes 4 through 20. Data from a time-lapse camera located at camp B for episode 4 and early part of episode 5, at camp C for later part of episode 5 and episodes 6 and 7, at camp D for episodes 8 through 10, and at camp E for episodes 11 through 20 (see fig. 1.1 for camp locations). From camps B, C, and D, height was measured with respect to a reference point on Puu Oo rim. A correction, ranging from 10 m for episode 4 to 25 m for episode 10, was added to compensate for difference in elevation between pond surface and reference point. For each episode, ending and starting times of fountain record are shown in Hawaiian standard time. Except as noted (circled numbers), gaps in the data record periods of bad visibility or an inoperative camera. 1, Data for main (eastern) fountain; camera was activated at 1305 H.s.t. June 13, 1983, at least 2.5 hours after the eruption began. 2, Data for main (eastern) fountain. Apparent zero values during first 3 hours resulted in part from bad weather and in part from fountain being entirely 17 1700 below crater rim. Large data gap on June 30 records transfer of camera from camp B to camp C. 3, Gaps in record of last day reflect concealment of low fountain by crater wall and intermittent bad visibil- ity. 4, Period of apparent low fountaining at about 0700 H.s.t. September 6, 1983, reflects poor visibility caused as Sun rose direct- ly behind fountain. 5, During periods of multiple fountains, only height of main jet was recorded. 6, Values record highest part of a complex of low fountains within Puu 00. 7, First 4.5 hours of record is miss— ing because of inoperative camera. 8, Wide gap records temporary cessation of eruption. 9, Heights of highest fountains, in early part of record, are uncertain because dense tephra, which sometimes ex- tended above camera field of view, prevented direct observation of dense, incandescent fountain. 10, Fountaining was intermittent; there are no gaps reflecting bad visibility or camera inoperativeness. Data are shown only for periods of high fountaining. HEIGHT, IN METERS HEIGHT, IN METERS 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS IllllllllllllllllIllllllllJlJ|||IIIIIIIJJJLIIIlllllllLLlllll b0 F4 Episode9 E 200 IIIIIIIIIIIIIIIIIIIIIIIIIIIIII 200 :_ EpisodeB : : 100 —— 100 :_ o —— o — 0300 6 7 0900 1500 15 80300 DAYS IN SEPTEMBER 1983 DAYS IN SEPTEMBER 1983 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII : Episode10 E 200—— _— q : 100 —— ~— 2 I 010 1. 0101900 DAYS IN OCTOBER 1983 100 IllllllIIIIIIIGDIIIIIIIIIIIlllllllllllllllllllllll IllldllllllIlllllllllllllllllllllll : Episode11 Episode12 .4 0—— 21005 2100 0730 1630 DAYS IN NOVEMBER 1983 DAYS IN NOVEMBER AND DECEMBER1983 200 IlllllllIIJlllllllllillljlllllllllllJIllll1ll|l __ Episode13 Emma” : 100 100 _— o o '_ 1530 DAYS IN JANUARY 1984 1230 1600 1600 DAYS IN JANUARY 1984 FIGURE 1.24.-Continued 22 ing which the fountain height varied widely, the Visible part of the eruptive orifice was an unchanging vertical, nearly cylindrical, 20-m-diameter pipe descending from the crater floor. Superimposed on the general trend toward increasing fountain height was a tendency for higher and generally more tephra laden fountains to occur in the early hours of the eruptive episodes and for fountain vigor to show a progressive decay during the episode. This pattern sug- gests that the lava delivered to the surface early in many of the episodes was slightly more gas rich. Such a phenomenon might be explainable by upward migration and entrapment of exsolved gas during repose periods within the upper part of the magma column in the con- duit system beneath Puu Oo. Escape of the gas would lllllllllllllllllllllll lllllllllllll THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 have been inhibited by a cap of relatively dense melt. This cap could have formed when degassed fountain and pond melt drained back into the conduit at the end of an erup- tive episode (see chap. 5). Elimination of the cap at the onset of an eruptive episode, either by simple extrusion or by erosion and dissemination as new melt rose upward through it, permitted escalating eruption of the melt from the underlying column, the most gas rich first. Maximum fountain height generally increased more rapidly than average fountain height through the series of eruptive episodes at Puu 00 (fig. 1.27). This trend could reflect either increasing effectiveness of the degassed- magma cap as a seal in inhibiting repose-period gas loss from the underlying magma column, or increasing volume of the gas-enriched melt that supplied the early high foun- 11111111111111”an 400 Episode 1 5 ® 300 200 200 100 100 llllllllllllllllllllllIIIIIIlllIIIIIlII lllllllllllllllllllllllllllllllllllllll l w 8 lllllllllllllllllllI'Illllllllllllllllll 3 o o 1200 A 15 DAYS IN FEBRUARY 1984 HEIGHT. IN METERS lllllllllll Episode 16 O O O IIIIIIIIIIIIIIIIIIIllllllll Illll ol 1 | I" ' 4 DAYS IN MARCH 1984 llllllllllllllllllllllllllIllllllllllllllllllllllllllll Episode 17 Episode 18 200 200 100 100 lllIlllllllll IllllllIllll llllllllll llllllllllllll l l O N O 0 (II ‘I’ 0 10500 30 DAYS IN MARCH 1984 3 18 FIGURE 1.24.—Continued llll llllllllll IITTTIIIIII 0930 20 21 DAYS IN APRIL 1984 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 23 tains. The first mechanism is not supported by available measurements of repose-period S02 emissions, and ex- amination of the second mechanism is inconclusive. Average 802 flux was determined at Puu 00 for the repose periods preceding episodes 15 through 20 (see chap. 5). This average flux varied by a factor of about 2 from one repose period to another; however, it showed no correlation, direct or inverse, With the vigor of early fountaining in the next eruptive episode. An anticorrela- tion might be expected if the degree of gas concentration in the upper part of the magma column at the beginning of an eruptive episode reflected the degree of imper- meability of the seal separating the gas-rich melt from the atmosphere between episodes. The eruption data are inadequate for us to confidently and specifically evaluate changes in the volume of the gas- enriched portion of the magma column, because of uncer- tainties that include conduit geometry, distribution of exsolved gas at the time of eruption, perturbation of the shallow reservoir during eruption, and applicability of the average lava-discharge rate to the early, high-fountaining hours of an eruptive episode. Nevertheless, the data permit us to speculate about the volume of gas-enriched melt and the theoretical diameter of the conduit that con- tained it. Greenland and others (see chap. 5) show that most of the exsolved gas present in the magma column at the start of repose periods was in the upper 1,700 m of the column. If we arbitrarily select the upper kilometer of the column as the reservoir in which upward-migrating gas accumulated during repose periods, and assume that the volume of magma involved in the early, high- fountaining part of an eruptive episode represents the ||l||lllllllllllll|lll1 volume of gas-enriched magma in an approximately cylin- drical conduit before the beginning of eruption, we can calculate a hypothetical average reservoir diameter. The best defined early high fountains occurred in episodes 4, 6, 10, 15, 16, 18, and 20; the results for those episodes are listed in table 1.2. Overall, the resulting conduit diameters range from 21 m (episode 10) to 82 m (episode 20). The high-fountaining activity of episode 10 was unusually lengthy and complex (fig. 1.24). For episode 20, it was unusually brief. Because the episode 20 fountain was relatively high throughout the entire episode, we pre- sent a second calculation for which all 9 hours of episode 20 eruption are considered to represent early high foun- taining. Eliminating the two extremes gives a limited range of 1.1 x 106 to 3.4x 106 m3 for the volume of magma involved in the early high fountains, and an estimated conduit diameter ranging from 38 to 66 m—a result agreeing closely with the 50i30-m diameter in- ferred by Greenland and others (see chap. 5) from inter- pretation of gas-flux and geodetic measurements. Overall, then, we recognize a primary pattern of in- creasing fountain vigor that correlates in part with increasing lava-discharge rate, and a secondary pattern of increasingly high, early, tephra-laden fountains that we suspect were supplied by a limited volume of magma in Which exsolved gas became concentrated above initial levels during repose. Major perturbations, incompletely understood, occurred in these patterns after episodes 10, 16, and 18. We suppose that the crater-floor subsidence and crater- Wall collapse at the end of episode 10 reflected a disrup- tion of the shallow part of the conduit system; this Illlllllllllll Episode20 300 200 Episode 19 100 HEIGHT, IN METERS lllll|llllll|l o | 0500 17 DAYS IN MAY 1984 180500 llll|ll|lllllllllllllllllllllllll IllllllllllllllllIIIIIIIIIIIIIIIIII [—[Illllllllllll 0—- 1700 0800 7 -8 DAYS IN JUNE 1984 FIGURE 1.24.—C0ntinued 24 THE PUU 00 ERUPI‘ION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 disruption led to the ensuing low fountains and the com- plexity of the the episode 11 and 12 vents. Unlike most of the other repose periods, in which lava was visible in the pipe or evident just beneath the crater floor for at least one and sometimes many days, visible repose-period activity was minimal or nonexistent before episodes 11 through 13; and the rate of magmatic-gas emission approached zero in the later part of the long repose period preceding episode 13. Collapse within an upper part of the conduit system, reflected by the surficial disruption seen in the crater at the end of episode 10, may have created a plug of porous rubble that impeded the upward rise of erupting magma and forced it to find alternative shallow pathways to the surface. In addition, the perme- able rubble may have occupied the zone normally filled during repose periods by relatively impermeable degassed melt. Thus, gas rising through the underlying magma column may have escaped to the surface more easily than was common during repose periods, and the early high fountains normally driven by trapped gas were minimized or eliminated. The repeated occurrence of a lava pond in the basin, with its surface undisrupted by boiling, and the rapid ex- trusion from the pond of a voluminous pahoehoe river emitting relatively little magmatic gas indicate that magmatic gas was largely depleted in the lava of the pond. FIGURE 1.25.—1123 vent erupting during episode 3. Prominent north- eastern fountain (at right) is about 150 m high. Tephra fallout raises a dust cloud as it strikes surface of growing cone. Southwestern foun- tain (center) is barely higher than its crater rim. Nearly 4 hours earlier, steady, voluminous lava emission began at the southwestern vent, pro- ducing southeast-flowing pahoehoe river visible at far left. At about 0830 H.s.t., lava flow breached south rim of crater by rafting away part of cone’s flank. Part of rafted cone, still' slowly moving when photograph was taken, is to left of center in middle ground; small, active, tube-fed pahoehoe flows issue from its base. View north- westward; photograph by J .D. Griggs, taken at 1113 H.s.t. April 4, 1983. Numerals (lower right) indicate date and time. Greenland and others (see chap. 5) show on theoretical grounds that all newly erupting magma rising Within the upper few hundred meters of the conduit was disrupted by exsolving gas to form a low-density spray of vesicu- lating melt and gas that constituted the fountain. Further- more, they conclude that dense, relatively degassed magma in the basin formed only from coalescence of disrupted melt that had been degassed by fountaining. Once a body of denser, reaggregated melt became a significant element, it may have interacted with the low- density, gas-rich mixture within the conduit and the basin, damping the overall vigor of fountain activity and pro- ducing the more chaotic low fountaining (fig. 1.26) that characterized much Puu 00 activity. Presence of a continually changing body of coalesced melt in the basin and, possibly, at times in part of the pipe may have caused the transient deflections and disintegra- tions of high fountains that we sometimes observed. It also provides a means to account for the seemingly disparate behavior of adjacent fountains. For example, at times during the early hours of episode 10, a high, tephra—laden fountain played side by side in the crater with a much lower, intermittent dome fountain (fig. 1.28). Similarly disparate activity occurred at adjacent but separate vents at the 1123 cone during episode 3 (fig. 1.25); the emission from one vent was gas rich and lava poor, and from the other gas poor and lava rich. Such apparent anomalies may have reflected either temporary or prolonged incursions of coalesced melt from the crater into the upper reaches of the erupting conduit system. F ISSURE-VENT FLOWS Low to moderate rates of lava-discharge that averaged about 0.1x 106 m3/h during the episode 1 fissure erup- tions normally produced thin (1—3 m thick) sheets of pahoehoe, few of which extended more than a few hun- dred meters from the vents (pl. 1). Much of the pahoehoe was Shelly (Swanson, 1973), and its solidifying surface crust became brecciated locally to form slab pahoehoe. In a few cases, longer, more channelized flows converted to aa and produced lobes that extended as far as 2 km from their vents. Some of the lava ponded, forming a crust of relatively smooth surfaced, strong, subhorizontal pahoehoe over areas of commonly 104 to 105 m2. Lava flowing in tubes beneath the surface slab issued slowly at the edge, form- ing a short, steep flow front that normally persisted as a high-standing, encircling rampart when the pond- surface slab subsided after termination of supply. The only high-discharge-rate fissure eruption occurred on January 7, 1983, during episode 1. Over a period of about 3.5 hours, discharge from the easternmost part of 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS FIGURE 1.26.—Complex of relatively low (max approx 40 m high) fountains, consisting of transient individual jets, within Fun 00 Crater during episode 18. Note ponded lava in crater interior near spillway. View southwestward; photograph by J .D. Griggs, taken at about 1200 H.s.t. April 20, 1984. 26 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 the 1983 fissure system (pl. 1) produced a broad, thin flow that extended 5.5 km down the south flank of Kilauea. For 2 to 3 hours at the peak of this eruption, the lava- discharge rate was 1.0 x 106 to 1.5 x 106 m3/h; during this period, the flow front, recognized from the air as pahoe- hoe, sped down the average 3° slope at a rate averaging 1.5 to 2.0 km/h. (These were the greatest discharge rate and greatest sustained flow-advance rate of the 20 erup- tive episodes discussed in this chapter.) As the flow advance slowed down after supply from the vent stopped, the front converted to slab pahoehoe and aa. Observed after it had stopped moving, the flow was aa for most of its length. It had evidently been emplaced in a relatively fluid state; its thickness averaged about 2 m, and it had enveloped standing ohia trees and preserved open tree molds where the trees had burned away. CENTRAL-VENT FLOWS Normally, each central-vent episode produced an elon- gate major aa flow fed by the lava river that debouched steadily through the lowest part of the crater rim (figs. 1.5, 1.26). Subordinate flows were common. Some originated where an overflow from the lava river, general- ly near its source, persisted; others were supplied by rapid or persistent accumulation of molten spatter on the flanks of the cone; some, particularly during episodes 4 and 5 and again during episodes 11 and 12, originated at subor- dinate vents. During episode 3, lava issuing from tubes through the growing 1123 cone formed an apron of pahoehoe adjacent to one sector of the cone. Another apron, of thin sheets of relatively dense, smooth pahoehoe, formed adjacent to the east base of Fun 00, owing to 400 , , 16‘ ' _ 15¢ ‘ I? 200 u, 320 - — }_ Lu 2 — _ Z —‘ 240 — 10' - j— I 18. g - 90 14. — W 170 I 160 — a. O . _ Z 2 _ 6 190 ' .4 E o 13. 8 80 ,_ 7o . o ' I . _ u. 5 | 11 12. 4T1 0 . I l o . . _ O l I l 150 250 350 450 550 DAYS FROM BEGINNING OF 1983 FIGURE 1.27.—Maximum and average fountain heights (average cal- culated from fountain-height data in fig. 1.24) versus time for episodes 4 through 20 (numbered). For episode 19, only maximum is shown. repeated low-volume overflows from the crater during episode 19. The Kilauea central-vent flows described herein share many features in common with the aa lobes produced by the 1984 eruption of Mauna Loa (Lockwood and others, 1985; Lipman and Banks, 1987). In comparison with an average Puu 00 episode (table 1.1), the Mauna Loa erup- tion was lengthy (approx 3 weeks), about 20 times more voluminous (220x 106 m3), and had a generally higher lava-discharge rate (0.5 x 106 to 1.0 x 106 m3/h through the major period of aa-flow production). The Mauna Loa aa extended approximately twice as far (27 km) from the vent as the longest Kilauea flow and about 4 times as far as the major flow in a typical Kilauea episode (table 1.3). Lipman and Banks (1987) concluded that at Mauna Loa, advance of the major flow lobes was terminated by midflow blockages and diversions that resulted from inter- related phenomena, including declining rate of discharge, increasing viscosity of the erupting melt, and physical maturation of the distributary-channel system. In con- trast, discharge was generally steady through normal Kilauea episodes, and most of the major Kilauea flows ad- vanced steadily until they were beheaded, most commonly by the abrupt termination of discharge at the vent. A few major flows stopped advancing because the lava supply was diverted near the vent to a new flow. RIVER—FED FLOWS River—fed aa flows, each one much like its predecessors, were the dominant eruptive product of the central-vent episodes. We recognize in them many of the same for- mative processes, as well as the same aa-flowage zones (with increasing distance downstream, the stabilized chan- nel zone, transitional channel zone, and zone of dispersed flow) described by Lipman and Banks (1987 ) for the 1984 Mauna Loa eruption. Generally, the Kilauea flows were from 4 to 8 km long and from 100 to 500 m wide; the average width of most of the main flows was 200 to 350 m. A few flows were significantly longer; the longest, pro- duced during episode 18, extended more than 13 km from the vent. The average thickness of the river-fed aa flows ranged from about 3 to 5 m. However, the variation in measured thicknesses was much greater; locally, in the distal part, some flows were as thick as 10 m. A major river-fed flow was normally initiated as a lobe of fluid pahoeohoe (fig. 1.29) advancing from its source, the lava river pouring through the breach in the crater rim. Initial levees (Sparks and others, 1976), which formed as lateral spreading diminished away from the axis of the lengthening pahoehoe lobe, quickly localized a central lava channel. Repeated overflows added successive pahoehoe layers to these levees, so that a lava river (fig. 1.30) con- l. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 27 TABLE l.2.—Emptton parameters for early high fountains of episodes 4, 6, 10, 15, 16, 18, and 20, and calculated average diameter of a hypothetical 1-km-long cylindrical conduit [Durations of high fountaining from data in figure 1.24. Average lava—discharge rates adjusted from data in table 1.1 to approximate dense-rock equivalence, assum— ing that average porosity of flows was 25 percent. of high fountaining times average lava—discharge rate. 2(volume/l ,000w) ‘1] Volume is product of duration Reservoir diameter = Duration Average Volume Of Proportion of Calculated of high llava— lava erupted total volume of reservoir Episode fountaining discharge during high erupted lava diameter (h) rate fountaining (percent) (m) (103 m3/h) (106 m3) 4 14 82 1.1 14 38 6 25 90 2.3 33 54 10 32 165 5-3 50 82 15 8 315 2-5 42 57 16 12 285 3.4 38 66 18 8 308 2.5 14 56 120 1 360 .4 12 21 20 9 360 3.2 100 64 1Episode 20 is considered both for early l-hour-long period of highest fountain- ing and for entire episode. fined within a channel of its own making (stabilized chan- nel zone) steadily delivered lava downflow. Such channels, when seen after an eruptive episode, were shaped like box canyons (fig. 1.9); their depth was generally 2 to 5 m, and their width 5 to 25 m. Oftentimes, the evacuated channel from an earlier eruptive episode channelized the flow dur- ing its initial advance (fig. 1.31). The reoccupied channel would soon be bank full, with new overflow levees building along the channel margins, and the new flow front seek- ing its own path downhill. Within a few tens of meters of the vent, maximum velocities measured in the lava river were as much as 10 to 15 ml s, and all estimates of lava discharge were several times larger than those determined from measured lava- flow volumes, apparently owing to the low density of lava near the vent. At 1 km from the vent, the maximum flow velocities were normally about 1 to 3 m/s, and estimated discharge was generally within 1 to 2 times the subse- quently calculated flux. Standing waves were normally present in the zone of high velocity near the vent, and tearing of the thin, incipient pahoehoe skin as it stretched during passage through the waves sometimes revealed a network of large (tens of centimeters in size), vesicle-like openings (fig. 1.32), suggesting that the near-vent lava was a foam honeycombed with voids. Lipman and Banks (1987) noted similar effects at Mauna Loa and documented a downstream increase in the specific gravity of solidified lava samples from about 0.5 near the vent to 1.5—2.5 in the lower reaches of the flows. L.P. Greenland (written commun., 1985) determined porosities of 57 and 46 per- cent for samples dipped from the Kilauea lava river during episode 18 at distances of 70 and 1,000 m, respectively, from the vent. At both volcanoes, the flows became pro— gressively depleted in gas with increasing distance from the vent. At Kilauea, our noses indicated that 802 flux from the flows was minimal; Greenland (see chap. 4) con- cludes that vent and upper—conduit processes had largely FIGURE 1.28.—Puu Oo vent erupting during episode 10. Main fountain is about 60 m high, leans northeastward, and is bombarding spillway with spatter. Just to west of it, a much lower dome fountain, which evolved later to a high jet, also rises from crater. The two fountains may have issued from separate orifices that existed side by side in crater floor before episode 10. View southward; photograph taken at 1058 H.s.t. October 5, 1983. 28 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 1.3—Characteristics of the major episode 1 fissure flow of January 7, 1.983, and of the river- and spatter-fed central-ventflows of episodes 2 through 18 [Average thickness = volume/area; average width = area/length. M, lava-river—fed primary flow of eruptive episode; S, spatter-fed flow] Length . Area Volume from Average Average Average Average . . Episode (106 m2) (106 m3) vent lava f§ux thickness Width veloc1ty Identification (m) (103 m /h) (m) (m) (m/h) 1 2.0 4.0 5,460 1,110 2.0 370 1,400 Fissure flow, January 7, 1983. 2 .3 .9 2,700 40 3.0 110 140 Northeastern flow. 2(M) 1.8 10.7 7,660 90 5.9 240 50 Southeastern flow. 3(M) 1.4 5.4 4,900 80 3.9 290 60 O vent flow. 3(5) 2.2 11.7 3,860 75 5.3 570 30 Northeastern 1123 vent flow, northeastern lobe. 3(5) 1.4 7.0 4,660 75 5.0 370 20 Northeastern 1123 vent flow, southeastern lobe. 3(M) .8 4.0 3,360 130 5.0 240 100 Southwestern 1123 vent flow, lobe 1. 3(M) .3 1.4 1,460 130 4.7 210 --- Southwestern 1123 vent flow, lobe 2. 3(M) 1.9 9.2 7,640 130 4.8 250 90 Southwestern 1123 vent flow, lobe 3. 4(M) 1.8 9.7 7,760 97 5.4 .230 90 --- 5(M) 1.9 8.4 8,420 93 4.4 230 90 Eastern flow. 5 1.1 3.2 5,960 37 2.9 180 80 Western flow. 6(M) 1.7 8.3 6,480 114 4.9 260 80 --- 7(M) 2.7 9.8 6,680 175 3.6 400 110 Northeastern flow. 7(3) .8 3.2 3,320 80 4.0 240 90 Southeastern flow. 8(M) 1.2 4.7 4,360 196 3.9 280 210 ~-- 9(M) 1.8 7.7 5,260 148 4.3 340 100 --- 10(M) 1.4 5.8 4,060 91 4.1 340 60 Northeastern flow. 10(5) 1.3 7.5 3,380 117 5.8 380 50 Eastern flow. 11(M) 3.4 10.4 9,570 242 3.0 360 210 --- 12(M) 1.6 5.0 8,360 143 3.2 190 250 Northeastern flow. 12 .5 1.1 2,300 32 2.1 220 60 Northern flow. 12 .8 1.7 4,440 49 2.1 180 320 Eastern flow. 13(M) 2.3 6.4 7,350 205 2.8 310 210 First flow. 13(M) 1.1 3.2 3,110 525 2.9 350 480 Second flow. 14(M) 1.1 2.5 4,720 129 2.3 230 200 Eastern flow. 14(5) .4 1.2 1,540 64 2.8 260 80 Southeastern flow. 15(M) 1.4 4.9 4,700 257 3.5 300 240 Northeastern flow. 15 .6 1.8 2,940 93 3.2 200 260 Eastern flow. 16(M) 2.2 7.5 7,930 232 3.3 280 240 --- 17(M) 2.4 7.4 10,780 338 3.1 220 490 Eastern flow. 17(5) .4 1.6 1,540 71 4.0 260 150 Southeastern flow. 18 .8 3.0 5,340 123 3.9 150 280 Northeastern flow. 18(M) 3.1 13.6 13,230 226 4.4 230 270 Eastern flow. 18(M) 1.3 3.4 7,350 249 2.5 180 430 Southeastern flow. 18 .8 2.2 4,710 58 2.7 170 130 Southern flow. 1Northeastern 1123 vent, southeastern lobe: area, volume, average thickness, and average width calculated from intersection with northeastern lobe. Distinction between first and second episode 13 flows was uncertain in mapping. volume, flux, and average width of the second flow may have been incorrectly estimated. the second flow from plots showing those variables (figs. 1.43, 1.44, 1.46, 1.47). Consequently, area and, thus, We have, therefore, omitted depleted the original magmatic gas in the lava and that the gases of the flows were mainly plume gases and air in its closed basin was a continuing source of fluid, gas- inflated melt. that had been trapped and mixed With coalescing frag- ments of melt in the fountaining process. The sustained cascade of spatter fragments and the turbulent disrup- tion where the fountain issued from the pond (fig. 1.20) must have churned these gases into the lava supplying the pond, much as an eggbeater whips air into cream or egg white. Thus, as long as the eruption lasted, the pond During some periods of more Vigorous fountaining— for example, the early parts of episodes 6 and 10—a high, broad fountain issued across much or all of Puu Oo Crater, and the lava overflowing the spillway seemed less fluid than normal. At such times, wide flows of spiny pahoehoe grading to a spread near Puu 00. Only after fountain- ing had diminished, pahoehoe issuing through the spillway 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 29 established the typical throughgoing channel system, and the normal elongate, river-fed aa flow then developed. Ap- parently, lava issuing from the crater was less fluid when a high, broad-based fountain prevented development of a stable pond into which air and vent gases could be whipped. As discussed in the next subsection, supply from the fountain with no basin or pond commonly produced FIGURE 1.29.—Pa.hoehoe flows extending from Puu 00 and fissure vents (to right of Puu 00) immediately uprift shortly after beginning of episode 4. Growing spatter ring, enclosing a lava pond and fountain at Puu 00, is approximately 20 to 30 m in diameter. Large, open crater with locally altered rocks, at bottom center just east of Puu 00, is remnant of 0 vent, which was active during episode 3. View southward; photograph by JD. Griggs, taken at 1318 H.s.t. June 13, 1983. Nu- merals (lower right) indicate date and time. FIGURE 1.31.-—Flow from Puu 00 at beginning of episode 9. Pahoehoe has spilled through breach in crater rim and is advancing down evacuated episode 8 channel. A small distributary rejoins main flow at right. Lava flux is estimated at about 10,000 to 20,000 m3/h. View northeastward from crater rim; photograph taken at 1603 H.s.t. September 15, 1983. Numerals (lower right) indicate date and time. thick, slow-moving 23. directly. An additional effect of high fountains was to distribute more tephra outside of the crater, reducing the volume of lava available to supply a river-fed flow. Transient pahoehoe overflows and local sustained leaks from the lava river created a pahoehoe—dominated envel- ope of levees and thin adjacent flows in the near-vent FIGURE 1.30.——Lava river in episode 2 channel, slightly less than 1 km from 1123 vent. Here, channel is about 15 m wide, and maximum velocity, at channel center, is approximately 2.4 m/s. Banks are pahoehoe overflow levees that slope away from channel on their outer flanks. Plates of smooth pahoehoe crust are carried along on moving lava surface. View westward; photograph taken at 1545 H.s.t. March 3, 1983. Numerals (lower right) indicate date and time. FIGURE 1.32.—Standing wave in lava river immediately downstream from Puu 00 during episode 18. Flow is from left to right; wave is probably about 2 m high. Rents in darker pahoehoe skin, formed by stretching during passage across standing wave, reveal large cavities in fluid incandescent lava beneath that give it a honeycombed aspect. Numbers in lower right indicate date and time. View northeastward; photograph taken at 1252 H.s.t. April 18, 1984. 30 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 region (fig. 1.33). Farther downstream, generally 1 or 2 km from the vent, this pahoehoe envelope gave way to an envelope of stagnating aa that contained the still—fluid lava river (fig. 1.34). At times, brief overflows locally mantled the aa adjacent to the river with thin pahoehoe. At a distance generally of about 2 to 5 km from the vent, the river surface underWent a gradual transition to aa (fig. 1.34). Commonly, the position of the pahoehoe-aa transition, once established in the channel, did not migrate appreciably upflow or downflow, and so the toe of the flow tended to extend an increasingly greater distance from the transition as an eruptive episode continued. As at Mauna Loa, a visibly incandescent central channel, of the transitional channel zone, continued its passage within stagnant aa-flow margins (fig. 1.35), although it differed in appearance from the lava of the stable channel upstream. Floating plates of smooth pahoehoe crust (fig. 1.30) within the stable channel zone gave way to abrading clots of basalt; the river surface became more rubbly (fig. 1.36), its velocity slowed, and its incan- descence decreased downstream within the transitional channel zone. We had a few opportunities, when the transitional chan- nel zone extended into the Royal Gardens subdivision, to observe slow downflow movement in which the entire cen- tral part of the flow appeared to move as a unit, separated by a narrow boundary zone from the lateral margin. Where continued movement left a partly evacuated chan- nel within the aa at the end of an episode, we noted that the steep, interior channel walls consisted of aa locally plastered with compacted and smoothed gouge that was grooved parallel to the channel axis (fig. 1.37). In addi- tion to evacuated aa channels in which the walls had FIGURE 1.33.—Proximal part of bank-full episode 11 lava river, with enclosing envelope of new pahoehoe formed by overflows. Distance to Puu 00 is nearly 2 km. Darker aa at flow margins is from an earlier eruptive episode. View southwestard; photograph by J .D. Griggs, taken at approximately 1100 H.s.t. November 7, 1983. obviously been shear boundaries, we commonly found arrays of longitudinal ridges of blocky aa, equivalent to the marginal shear ridges documented by Lipman and Banks (1987) within the transitional channel zone on Mauna Loa. FIGURE 1.34.—Distal part of episode 11 flow. Fluid pahoehoe river (stabilized channel zone) is enclosed within an envelope of stagnating aa, approximately 200 m wide at bottom center. Flow margin in near and middle ground, about 4 km from the vent, underwent essentially no change during remaining 7 hours of eruption. Transition within channel from pahoehoe to aa, visible here where shiny pahoehoe sur- face loses its identity, is approximately 5 km from vent. Advancing flow terminus, obscure in distance at upper left, is about 8 km from vent. View eastward; photograph by J .D. Griggs, taken at approx- imately 1100 H.s.t. November 7, 1983. FIGURE 1.35.—Incandescent aa channel within transitional channel zone of eastern episode 18 flow. Full width of flow in near ground is 300 to 500 m. Between lower right corner of view (approx 8.7 km from vent) and area where incandescent channel loses its identity (approx 1 km distant), enclosing aa is stagnating. Beyond that, in zone of dispersed flow, active zone widens, and flow perimeter becomes in- creasingly active toward flow front, out of view in distance, 11 to 12 km from vent. View southward; photograph by J .D. Griggs, taken at approximately 1142 H.s.t. April 20, 1984. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 31 At the downstream end, the transitional channel zone lost its identity (fig. 1.35) as it graded into the zone of dispersed flow behind the active flow front, which normally advanced at average rates ranging from about 50 to 300 m/h. Unlike the upper flowage zones, in which the active lava passed through an axial region bounded by relatively stagnant, largely solidified lava, the active region widened within the zone of dispersed flow. Thus, the perimeter of this zone became increasingly active in the downstream direction, and some lateral spreading occurred near the front. The major activity, however, was concentrated at the front, and the entire front, tens to hundreds of meters wide and representing nearly the full final width of the flow lobe, advanced more or less uniformly. Sometimes the front divided, forming two or more lobes that advanced separately. At times, one of these lobes captured the bulk of the supply, and progress on the other diminished or stopped. Sometimes the lobes rejoined leaving a kipuka within the flow. Flows of episodes 2 through 5 invaded the Royal Gardens subdivision. More continuous observation of the flow-front advance was possible in the subdivision than in the rain forest between the rift zone and the subdivi- sion. These observations showed that the flow front ad— vanced in pronounced surges (Neal and Decker, 1983), with velocities ranging from about 2 to 30 m/min. Surges at Royal Gardens lasted as long as about an hour, although most were briefer. They were separated by intervals as FIGURE 1.36.—Distal part of the advancing northeastern episode 12 flow. Zone of transitional flow, with its rubbly, incandescent central chan- nel, extends virtually to flow terminus, about 5 km from vent. In near to middle ground, flow is 30 to 50 m wide. Except for avalanching of a few blocks, rubbly aa levees are stagnant; they underwent no fur- ther recognizable lateral spreading as flow advance continued. On its north (left) flank, episode 12 aa impinges on episode 7 aa. Bare, weathered lava at right was erupted in 1963. View northeastward; photograph by RE. Moore, taken about 1700 H.s.t. November 30, 1983. long as 8 hours during which the flow front advanced more slowly or was virtually stagnant. Surging also occurred in flatter terrain closer to the vent. During close monitoring of the episode 16 flow advance, from 4 to 5 km from the vent, brief surges originating across only part of the flow front moved as rapidly as 60 m/min. Advancing surges could commonly be seen from a suitable vantage point while they were still several hun- dred meters upflow of the flow front. Seen from the front, with line-of-sight parallel to the flow surface, an approach- ing surge appeared as a rapidly advancing wall of in- candescent aa, rising several meters above the normal level of the flow surface. With a downward perspective that gave a view of the flow surface (fig. 1.38), a surge was seen as a tapering sheet of highly incandescent lava, emerging upstream from the axial region of the flow and advancing rapidly over the surface of slow-moving, much less incandescent aa. When not accelerated by a surge, the aa—flow front showed only minor incandescence (fig. 1.39A) and advanced slowly as blocks and finer fragments sporadically avalanched down the continually oversteep- ening front. As the surge approached, however, the front rapidly thickened (fig. 1.393), and its advance accelerated. Subsequently, the advancing front thinned, commonly as FIGURE 1.37.—Wall of evacuated channel of episode 4 aa flow near Royal Gardens. Steep wall is plastered by striated gouge composed of fine- ly comminuted basalt left at sheared boundary between more active aa channel fill and bounding levee. Backpack for scale. Photograph taken November 25, 1983. 32 THE PUU OO ERUPI‘ION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 a short-lived, more fluid flow broke out of it (fig. 1.390), or as a rapidly moving sheet of thinner, more fluid lava overtook the thickened aa margin, spilled over it, and became the flow front. During episode 4, collapse of part of the cone at the vent caused a surge that produced overflows in the stabilized channel zone. Approximately 11/2 hours later, a large surge occurred near the distal end of the flow at Royal Gardens about 7 km from the vent. On the supposition that these two events were related, we infer that the pulse of melt traveled the length of the flow at an average velocity of 1.3 m/s, a value in agreement with the lower range of velocities we determined in the stabilized chan- nel zone. Such a transmission would be possible only if a fluid core truly extended through the entire length of the flow. Production of fluid lava at the flow front dur- ing surge events supports this idea. In addition, after each eruptive episode, we invariably found small lobes of dense, spiny pahoehoe that had been extruded from the aa-flow» margins or, in some cases, from the flow tip. Normally, however, we saw no specific upstream events that we could relate to the surges in Royal Gardens. We suspect, therefore, that melt tended to pond beneath the carapace of aa in the middle or lower parts of the flow until the thickness of ponded lava was sufficient for gravitational stress to overcome the yield strength; then, a transient lava surge swept downstream to the flow FIGURE 1.38.—Nighttime surge at Royal Gardens during episode 5. Fluid surge advanced at estimated rate of 21 m/min over preexisting, nearly stagnant aa (lower left, with small points of incandescence.) Photo- graph by R. Seibert, taken at 1821 H.s.t. July 2, 1983. front. Supporting evidence for such a mechanism comes from the observation, during the Mauna Loa eruption, that the belts of shear ridges adjacent to the obvious tran- sitional aa channel inflated and subsided as much as several meters vertically in response to changes of lava level in the channel (Lipman and Banks, 1987). SPATTER—FED (ROOTLESS) FLOWS Brief, intense spatter falls on the steep flanks of the cone produced thin sheets of pahoehoe that armored the cone surface (fig. 1.7). More prolonged falls on the steep flanks produced pahoehoe or 33. flows that extended away from the cone. Pahoehoe apparently occurred where in- tense falls of fluid spatter were rapidly mobilized on the steep slope. During episode 7, intense spatter fall on the west flank of Puu 00 (fig. 1.20) at times filled a non- erupting satellitic crater with a pond that overflowed and supplied a channelized lava river. A prolonged spatter fall on one sector of the cone’s flank sometimes produced a broad and thick spatter—fed aa flow (fig. 1.40) that advanced as much as 3 km from the cone. Such flows did not have well-developed central channel systems and consisted throughout of slowly advancing aa. Near the source, their surfaces were commonly fractured to form slivers elongate perpendicular to the direction of flow; individual slivers could be as large as 5 m across and 50 m long. With continued movement, these slivers broke up to form smaller, more nearly equant blocks. The large fractures are analogous to the tears created by slumping (fig. 1.40), and on the steep cinder—cone flanks, there is probably a continuum from simple slumping to genera- tion of major spatter-fed flows. Likewise, because these spatter-fed flows originate from mobilization of accum- ulating spatter deposits on the flank of the cone, the flows are continuous with the flanks, and so the distinction between the cone and the flows may be obscure. The most voluminous (approx 19 x 106 m3) spatter-fed aa flow was produced during episode 3 at the northeastern vent of the 1123 cone. During approximately 10 days of eruption, the high fountain at this vent produced a multi- lobed aa flow (pl. 1) that ponderously advanced 4 to 5 km from the vent at an average rate of 25 m/h (table 1.3). On average, the flow was about 5 m thick, but locally it was at least 12 m thick where it buried a preexisting spat- ter cone at camp A (fig. 1.41). For most of the eruption, FIGURE 1.39.—Evolution of a surge at Royal Gardens during episode 5. A, Surge, which forms low ridge visible through haze behind aa‘ flow front, approaches nearly stagnant flow front. Rate of approach was estimated at approximately 30 m/min. Photograph by R.W. Decker, taken at 1454 H.s.t. July 2, 1983. B, Just 4 minutes later, flow front has doubled in height, incandescence at its front has in- creased, and accelerated advance has begun. Before it thinned to a fluid lobe, high advancing front shoved truck to vicinity of light—gray sedan before burying them both. Photograph by R.W. Decker, taken at 1458 H.s.t. July 2, 1983. C, After surge overtook aa front, a 1- to 2-m-thick fluid breakout continued down paved street at a rate of 15 m/min. View from farther downslope; photograph by R.W. Decker, taken at 1510 H.s.t. July 2, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPI‘IVE EVENTS 33 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 100 METERS l___|—_J 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 35 the fountain and the flow issued from a reentrant in the growing 1123 cone (fig. 1.42); after the initial 1 to 2 days, there was no closed basin with a lava pond like those that typified most central-vent episodes and produced the river-fed flows. Pahoehoe, that seemed more like a slurry 'on the aa-flow surface, occurred only within a few hun- dred meters of the vent (fig. 1.42); and a short transitional channel, probably contained entirely within a kilometer of the vent, passed, for at least a few days, on the north side of the slowly disappearing camp A spatter cone (fig. 1.41). Otherwise, we saw no evidence of the well— developed channel system that we would see at so many later times, and the resulting flow consisted of coarse, blocky aa throughout. The massive spatter—fed aa flows apparently formed where spatter accumulating from prolonged falls re- mained above solidus temperatures but, unlike the more rapidly mobilized spatter—fed flows discussed earlier, formed a mass with too high a yield strength to flow as thin fluid sheets or lobes on the available slope. This slower mobilization could have resulted from lower rates of accumulation, partial cooling during a high-fountaining trajectory, or lower slope angles. As spatter accumulated in a pile some meters thick, it probably lost part of its entrained gas and became denser, but eventually its mass caused sufficient gravitational stress to overcome its increased yield strength. It then began to flow and con- tinued to do so as long as gravitational instability was maintained by continuing spatter accumulation at the source. NUMERICAL FLOW PARAMETERS Data obtained from our real-time observations and 'postepisode flow mapping include measured length, area, and thickness, and calculated flow-advance rate, volume, average thickness, and lava-flux rate for most of the in- dividual flow lobes produced during the central-vent erup- tions. The fundamental data are given below in the chronologic narrative and in plates 1 through 5, and, along with derived parameters, are summarized in table 1.3. Note that we use the term “lava-flux rate” for the rate of lava supply to individual flow lobes to distinguish that from the overall lava-discharge rate for each episode (table 1.1). Flow length versus volume is plotted in figure 1.43 for the central-vent flows of episodes 2 through 18. Although there is significant scatter, the data show that these two parameters are clearly related. As recognized by Malin (1980) for other historical Hawaiian lava flows, the cor- relation occurs because the flows have a limited range in cross-sectional area. The major flows provide most of the data spread: They range from 2 to 6 m in average thick- ness and mostly from 200 to 350 m in average width (table 1.3). Thickness was probably controlled primarily by viscosity, and width was limited by the dynamics of levee development and channel formation as the flow extended from its point source at the central vent. Com- monly, wider river-fed flows, such as the episode 7 flow (pl. 2), resulted from division of the flow toe into subor- dinate lobes that traveled along nearly parallel paths. Figure 1.43 also implies that the spatter-fed flows have relatively greater cross-sectional areas than other flows of comparable volume. Other workers (Walker, 1973; Wadge, 1978; Malin, 1980) have found, in Hawaii and elsewhere, that flow length is correlated with lava-flux rate. The same is true for these central-vent flows (fig. 1.44). In accord with Malin’s (1980) conclusion for other historical Hawaiian flows, we found that the amount of scatter in the length- flux relation is greater than in the length-volume relation. For central-vent flows of the Puu Oo eruption, the length— flux correlation results primarily from the limited range of flow widths and thicknesses, as does the length-volume correlation. The additional scatter in the length-flux rela- tion apparently reflects largely the relatively great range (from less than 10 to about 200 hours) in the duration of lava supply. Apparent viscosity, discussed subsequently, probably also played a role in controlling flow morphology and, thus, the length-flux relation. Flow-advance plots for each eruptive episode are in- cluded on plates 1 through 5. Except for brief periods in the early hours of some eruptive episodes, the flows tended to have fairly uniform advance rates. Short-term variations in slope in the diagrams are difficult to evaluate; they may reflect either temporary fluctuations in advance rate or errors in flow-front position that resulted from the sometimes—difficult conditions of flow- front reconnaissance. Our overall impression was that normally during an episode, overall discharge rate varied FIGURE 1.40.—Puu 00 (top center) and spatter-fed flows to north, southeast, and west after episode 16. Evacuated and rubbly channel (labeled) for main flow is in upper right. Coarse, blocky, spatter-fed flows originated on flanks of cone, where accumulating viscous spat- ter began moving downslope (arrows indicate direction of movement), deforming under its own weight. More brittle upper part of deform- ing mass fractured into slivers elongated transverse to direction of flow, and continued movement broke the slivers into smaller, more equant fragments. Near-vent parts of small spatter-fed flows at up- per left have been buried by subsequent continued accumulation of pyroclastic materials on flank of cone. Gashes that reflect slumping of the accumulating pyroclastic deposits are conspicuous in south sec- tor. Landscape at left has been smoothed by deposition of an air-fall pyroclastic deposit. Photograph 84.3.16JG120#7 by J .D. Griggs, taken March 16, 1984. 36 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 little, if at all. Therefore, the fairly steady advance of the flow front during an episode reflects generally steady lava discharge and the effectiveness of the channelized flow system in delivering lava to the active distal region of the flow. Short-term variations like those represented by the FIGURE 1.41.—Spatter cone at camp A surrounded by episode 3 aa. Cone, formed in 1965, was 12 m high. Coarse, rubbly aa on south rim (far side) is virtually stagnant. More active lava on north side (in near ground) overtopped north rim of the 1965 cone 2 days later. Subse— quently, cone was completely buried by episode 3 aa. View south- eastward; photograph by BB. Moore, taken on the afternoon of April 1, 1983. ‘ 14 I I I I I I M 12~ - w M 3310— — E M 2 3 M M _ 8_ _ x M MM M g M M :5 M M s e— . — 5 ° ' M " 'M MM 5 3 0 M O 4— M S .. fi.‘ . a M s C 2 ' - SMS 0 I I I I I l o 2 4 6 a 10 12 14 VOLUME, IN MILLIONS OF CUBIC METERS FIGURE 1.43.—Flow length versus volume for central-vent flows of episodes 2 through 18. Data from table 1.3. M, major river-fed flow(s) of each eruptive episode; S, spatter-fed flow; dots, subordinate river- fed flow. surges witnessed at Royal Gardens probably occurred, but most were averaged out by the intermittency of the flow- front determinations (pls. 1—5). The flow-advance diagrams also show no clear effect of differing topographic slope on the advance rate. FIGURE 1.42.—Erupting 1123 vent during episode 3. Pahoehoe distribu- tary system, which extends about 200 m from vent, issues from base of 40— to 50-m-high northeastern fountain. Small tube-fed flows issue from base of 1123 vent at lower left. Episode 2 cones, of 0740 vent, and pahoehoe are visible at lower right. View northwestward; photo— graph by J.D. Griggs, taken at approximately 1230 H.s.t. March 31, 1983. ‘4 I I I I I T M 12 _ _ M E? m 10 — - I— M LLI E g a M M x M M M M g M M I‘ M I— 6 M _ g '— . a ' M o E M M 3 0 M o 4 — M — _l s “' 3 sM . C 2 — ° — 55 M o I l I I I I o 100 000 200 000 300 000 AVERAGE LAVA FLUX, IN CUBIC METERS PER HOUR FIGURE 1.44.—Flow length versus average lava-flux rate for central- vent flows of episodes 2 through 18. Same symbols as in figure 1.43. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 37 ' Average slope angles range from approximately 15° along the axis of the rift zone to about 6° in the Royal Gardens area. Higher than normal early advance rates—for example, in episode 8—reflect rapid advance of the main flow as an elongate lobe of pahoehoe during the early part of an episode before the transition to aa became established. Slower than normal early advance, however, as in episode 16, may reflect lower fluidity of the lava that issued from the crater during periods when the fountain was high and broadly based (see subsection above entitled “River-Fed Flows”), or significant diversion of part of the lava supply to rapidly accumulating tephra deposits from high foun- tains during the early part of an episode. Comparison of the flow-advance diagrams for early episodes with those for later episodes shows that the main flows advanced at higher rates during the later episodes. This relation is summarized in figure 1.45, in which the average advance rate versus date is shown. Also shown is the generally lower velocity of the spatter-fed flows. Three of the later main flows advanced at unusually high average velocities of 400 to 500 m/h. One main flow, which extended from the Puu 00 area to Royal Gardens over a single night during episode 18, was recognized as pahoehoe as its toe moved into the subdivision. Another, the second episode 13 flow, was active for only about 6 hours; its rapid advance rate almost certainly reflects extension of the channelized—pahoehoe channel virtually to the advancing toe. The unusually rapid average ad- vance rate of the third, which was the main flow of episode 17, apparently does not reflect sustained prolongation of fluid channelized pahoehoe to the distal part of the flow. Late in the afternoon of March 30, 1984, after nearly half EPISODE 14 2 3 45 6 7 89101112 1315161718 a; 500 . D . M M o I M E A 400 — - (I) 0: Lu '— I “2’ 300 — — E M ' M '6‘ >: M M M L" zoo — M M - o 9 I; O S I w 100 - M MM M M — 3 M . M s s n: M M “5’| 1 Lu > 85 l < 0 I I I I o 100 200 300 400 500 DAYS FROM BEGINNING OF 1983 FIGURE 1.45.—Average flow-advance rate versus time of eruption for central-vent flows of episodes 2 through 18 (numbered). Same sym- bols as in figure 1.43. a day of eruption, the toe of the episode 17 flow was about 5 km from the vent (pl. 5), and the pahoehoe-3a transi- tion in the channel was 1 to 2 km from the vent. By daybreak the next morning, after the eruption’s end, the aa toe was more than 10 km from the vent; overflows draping the aa-channel margins showed that fluid pahoehoe within the channel had extended as far as 4 km from the vent, but there was no evidence of its having been within the distal half of the flow. For three-quarters of its length, the episode 17 flow followed the axis of the episode 16 flow, and the rapid advance of the episode 17 flow may reflect its containment, over at least part of its length, within episode 16 levees. Thus, inhibition of lateral spreading at the advancing toe may have resulted in an accelerated forward advance. The average velocity of the main flows is correlative with the average lava flux (fig. 1.46), which increased through the series of eruptive episodes (fig. 1.47) in apparent response to increasing overall discharge rate (fig. 1.4). These changes probably reflect progressive enlargement and streamlining of the conduit between the summit reservoir and the vent, as well as progressively decreasing viscosity of the melt in transit to the vent. Unexpectedly, we found that the average flow thickness progressively decreased through the series of central-vent episodes (fig. 1.48). The most likely explanation is that the decreasing average thickness, as well as the increas- ing average velocity, records a decrease over time in the viscosity of the lava. Thermal, compositional, and petrographic data support this explanation. 500 I I I I I I 400 - — 300 - _. AVERAGE VELOCITY, IN METERS PER HOUR ' M - M M M 200 — M M M M _ O . s M _ M _. 100 5 MM In . . s M M m S s o I L I l l I o 100 000 200 000 300 000 AVERAGE LAVA FLUX, IN CUBIC METERS PER HOUR FIGURE 1.46.—Average flow-advance rate versus average lava flux for central-vent flows of episodes 2 through 18. Same symbols as in figure 1.43. 38 Lava temperatures measured through episode 18 are plotted in figure 2.2 (see chap. 2); most are pahoehoe temperatures measured within 1 or 2 km of the vent. Although these data show scatter for each eruptive episode, an unmistakable progressive increase of about 20 °C is indicated. A concomitant compositional change from distinctly differentiated lava in the early episodes to more mafic lava in the later episodes is illustrated in figure 2.33, which plots analyzed Na20+K20 content versus the time of eruption. A similar plot (fig. 1.49) of the abundance of phenocrysts, which are enclosed in a groundmass of glass (see chap. 3), shows that the propor- tion of crystals to melt decreased through the early EPISODE 14 3 45 6 7 89101112 1315161718 2 M 300000 — _ _ M M M _ M M M M - M MM I M 5 MM M 'I” 85 S O I l . . 200 200 000 Z 1 00 000 l AVERAGE LAVA FLUX IN CUBIC METERS PER HOUR 400 0 1 00 300 500 DAYS FROM BEGINNING OF 1983 FIGURE 1.47 .—Average lava flux‘versus time of eruption for central- Vent flows of episodes 2 through 18 (numbered). Same symbols as in figure 1.43. EPISODE 14 2 3 4 5 6 7 8910 1112 1315161718 6 M I s (I) M a: w s E 5 M _ 2 M M 2 M US M M 33 IN M Z 4 — s .. x M S M I 9 I M E. M M I; M .I M < 3 — u N — a: i M ">4 Ms . < M M | 2 I I I I J I O I II I 0 100 200 300 400 500 DAYS FROM BEGINNING OF 1983 FIGURE 1.48.—Average flow thickness versus time of eruption for central-vent flows of episodes 2 through 18. Same symbols as in figure 1.43. THE PUU 00 ERUP’I‘ION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 episodes from a high of about 9 percent; after episode 10, the erupting lava was nearly aphyric. The complete analyses are given in chapter 2, and the petrology is discussed in chapter 3. Field and laboratory studies by Shaw and others (1968) and Shaw (1969) of viscosity relations in Kilauea tholeiite of Makaopuhi lava lake indicate that a 20-°C change within the temperature range of the 1983—84 Kilauea lavas would be expected to yield an inverse change of about 40 percent in viscosity if the melt is crystal free. Addition of crystals with progressive cooling, from none at the 1,200-°C liquidus temperature to about 25 volume percent at 1,120 °C in the solidifying lava lake, greatly exaggerated the viscosity change in the lava lake; a decrease in viscosity, between 1,145 and 1,125 °C, of about an order of magnitude is indicated. In addition, Shaw recognized that the melt itself increased in viscos- ity as its composition changed in response to progressive crystallization. All three effects operated simultaneous- ly during the first 20 episodes as temperature increased, crystallinity decreased, and the bulk composition became less evolved. The ensuing decrease in viscosity, which is recorded by the increasing average velocity and decreas- ing average thickness of the main flows, may be respon- sible in part for the increase in overall rate of lava discharge, which, as discussed above, apparently also con- tributed to the increasing average advance rates of the main flows. Comparison With the 1984 Mauna Loa lava is interest- ing. During that approximately 3 week long eruption, the Mauna Loa lava changed neither in composition nor in its eruptive temperature of about 1,140 °C. Increasing apparent viscosity during the eruption, however, may have reflected increasing crystallinity (Lipman and Banks, 1987), as the microphenocryst abundance increased from less than 0.5 volume percent in the early part of the erup- tion to 20—30 volume percent in the later part. I I I I I I I I I . 3 Lu (J ZI— -:> I1 57 ”5" ”1 33 5 O 1 8> 3.11 2 67 9 Z 3 7 _. LUZ ‘ 77 89 E 77 81011 18 1 3 1617 01 I I I I I 11112] 1314.1516 n7‘8 200 DAYS FROM BEGINNING OF 1983 300 FIGURE 1.49.—Phen0cryst content of lava samples versus time of erup- tion for episodes 1 through 18 (indicated by numbers). Data from Garcia and Wolfe (see chap. 3). 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY 0F ERUPTIVE EVENTS 39 REPOSE-PERIOD ACTIVITY Low-level eruptive activity was Visible between many of the major eruptive episodes. For approximately 2 weeks before the main activity of episode 2, slow and sporadic effusion of about 0.5 x 106 m3 of lava built a low shield and a line of small spatter cones northeast of the 1123 vent (see subsection below on episode 2). Very slow eruption of spatter and pahoehoe also occurred within and near the 1123-vent spatter ring about a week before episode 3 (see subsection below on episode 3). Thereafter, we observed repeated repose-period occurrences of low- level eruptive activity within Puu Oo Crater or, after episode 13, within the open pipe that descended from the crater floor. This activity was dominated by two phenomena: gas-piston activity and gradual ascent of the magma column toward the surface within the conduit below the crater floor. As in the first stage of the 1969—71 Mauna Ulu eruption, the low-level activity at Puu 00 became more continuous with succeeding eruptive episodes (fig. 1.3). A general pattern of summit inflation persisted throughout the repose periods. However, the rate of inflation varied (fig. 1.2), and at times inflation was inter— rupted by brief periods of deflationary tilt change. Har- monic tremor of low amplitude continued in the Puu 00 area through the repose periods. Gas-piston activity much like that described during repose periods in the first stage of the 1969—71 Mauna Ulu eruption (Swanson and others, 1979) was evident many times during the low-level activity at Puu 00, although the volume of rising and falling melt was about a tenth of that typically observed at Mauna Ulu. Some- times we could'observe the gas-piston events directly, either in the lower part of the crater between the early episodes or, between the later episodes, within a bowllike inner crater set into the partly crusted top of the lava- filled pipe. The lava surface, covered by a thin, flexible crust, would gradually rise about 10 to 20 min the bowl or pipe. The lava was apparently uplifted by a gas accumulation buoyantly rising through the upper part of the magma column (Swanson and others, 1979). At times, a low (max 4 m high) dome fountain played on the lava surface, above the conduit that was exposed when the pond drained. Otherwise, this surface was smooth and undisrupted, sug- gesting that the rising lava was relatively depleted in gas. At its maximum elevation, when about 2,000 m3 of melt had accumulated in the bowl, the distended lava surface would be momentarily poised. Then, as the accumulated rising gas pierced the lava surface and began to escape, the lava became agitated and drained rapidly out of sight into the conduit below. Simultaneously, a roaring rush of SOg-rich gas that commonly carried spatter and Pele’s hair was emitted from the collapsing pond and the glow- ing conduit into which it drained. A distinctive burst in harmonic tremor accompanied each brief episode of rapid draining and gas emission. Lava would then reappear, slowly rising in the stem of the funnel, and the process would begin again. The cycles that we observed ranged from about 4 to 20 minutes in duration and, at times, were strongly periodic. Figure 1.50 illustrates this phenomenon. Once the single open pipe became a permanent feature, after episode 13, we could see that the magma-column sur- face within the pipe (fig. 1.51) rose gradually and some- times spasmodically toward the crater floor during each repose period. Our visits to the remote vent during repose periods were intermittent, and visibility down the pipe was commonly limited by blinding fume and, on damp days, dense water vapor. Under the best conditions, we would first see the top of the magma column when it was about 50 m down the pipe; this first sighting generally occurred days to weeks before the next major episode. Following episodes 13, 19, and 20, the top of the column was immediately visible in the pipe (see representation of low-level volcanic activity in fig. 1.2). Sometimes the gently roiled magma surface would be open (fig. 1.51); at other times, it would be crusted, but there was always at least one small glowing vent that emitted gas, spatter, and a few small pahoehoe flows (fig. 1.52). Normally, flows or spatter were emitted periodically from such small vents along with a rush of gas and a burst of harmonic tremor that suggested gas— piston activity below the crust. During the repose periods between the early episodes, when the conduit beneath the crater floor was choked with rubble, similar activity at small vents in the crater floor indicated that the magma column was near the level of the crater floor and that more vigorous eruptive activity was imminent. ERUPTIVE-EPISODE ONSETS, ENDINGS, AND PAUSES We were able .to watch firsthand the beginnings of several major eruptive episodes at Puu 00. The crater would fill with lava sufficient to steadily overflow the spillway. If gas-piston activity had been occurring, it gave way to an open, roiled pond with a low dome fountain. Over a period ranging in length from tens of minutes to about 2 hours, the rate of overflow would progressively increase from 103—104 m3/h to normal lava-discharge rates of at least 105 m3/h. Simultaneously, the fountain would expand from an initial height of less than 10 m to the tens or hundreds of meters typical of vigorous erup— tion. Because of the quiet, relatively slow extrusion and the minimal fountaining during the early part of the erup- tion onset, we had the repeated impression that probably 104 to 105 m3 of degassed magma was normally expelled THE PUU OO ERUPTION OF KILAUEA VOLCANO. HAWAII: EPISODES 1—20. 1983—84 a; O 25 A Eruption 7' Top of conduit HEIGHT ABOVE TOP OF CONDUIT, IN METERS 8 I 1000 1 100 1200 1300 1400 1500 1600 HOURS, H.S.T. JULY 22, 1983 3 Southwest rim Northeast rim / / / Top of conduit \ l V 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 41 from the subsurface conduit before the typical vigorous lava discharge and fountaining began. Most individual fissure-eruption events of episode 1 ended gradually over a period of tens of minutes to several hours. The elongate fountain would diminish in height and become less steady, and discharge would stop along some segments, so that the remaining fountain activity was con- centrated at separated segments or nodes along the erup- tive fissure. Continuous eruption along a segment of the fissure gradually gave way to intermittent eruption that, in the last stages, generally consisted of a series of loud, intermittent gas bursts. These bursts ejected curved, elongate ribbons of lava, several meters long, that spread upward and outward, as if a large bubble had burst within ponded lava in the vent. The central-vent eruptions at the 1123 vent during episodes 2 and 3 also died gradually, as indicated by the progressive decrease in fountain vigor during the last day of each episode (figs. 1.21, 1.23). During this decay, the fountain became so low at times that it disappeared briefly from view. In addition, during the last hour of episode 2, we saw brief cessations of fountaining and lava-flow production. In contrast to the gradual endings of episodes 2 and 3 at the 1123 vent, the endings of eruptive episodes at Puu 00 were relatively abrupt. During some episodes, dis- charge stopped suddenly with no premonitory warning. At least once, we missed the end of an eruptive episode when our attention was briefly diverted from the vent. Ordinarily, however, the fountain diminished in height and became less steady in the last 3 to 10 minutes of the episode, and sometimes during that short period we saw pauses of a few seconds or tens of seconds in fountaining and flow production. Brief gas-bursting events like those that commonly ended the episode 1 fissure eruptions sometimes occurred in Puu Oo Crater during the last moments of an eruptive episode. Normally, the output of lava from Puu 00 was steady until the terminating moments of the episode, after which no further discharge occurred. Significant exceptions, in episodes 13 and 19, are described in detail in the subsec- tions below on these two episodes. In each case, discharge rapidly waxed and waned, often pausing momentarily, so that high fountaining and flow production would abrupt- ly stop and then resume again. In addition, episode 13 discharge stopped for several hours and then resumed. Many such pauses occurred during episode 19, which in- cluded periods of intermittent slow discharge and gentle overflows of lava from Puu Oo Crater. A series of brief pauses was also recorded during a 24-hour period in the middle of episode 3 (see subsection below on episode 3). The repeated brief, abrupt interruptions and resump- tions of vigorous fountaining and flow production seen in episodes 3, 13, and 19 almost certainly reflected inter- ference with subsurface magma transport close to the vent rather than in the summit region. The abrupt or briefly oscillating terminations of eruptive episodes at Puu Oo resembled the pauses that occurred within episodes and, like them, probably were controlled by conditions in the vent region. Furthermore, the onset and ending of deflation at the summit, as recorded by the Uwekahuna tiltmeter, commonly lagged by as much as several hours behind the onset and ending of each eruptive episode at Puu Oo (Wolfe and others, 1987); this relation suggests that the major inflections in the summit tilt record, after the vent became established at Puu 00, were responses to changes manifested first at the vent. LATER DEVELOPMENTS Prolonged pond activity and low discharge during episode 19 largely filled Puu Oo Crater with solidified pahoehoe and led to eventual elimination of the lava pond and the stream of fluid pahoehoe that had issued from it so regularly. Episode 20 was brief and occurred during the night without our Witnessing it. Later mapping showed that nearly half of the erupted volume supplied a massive spatter-fed flow, and the rest formed a lava- FIGURE 1.50.—Gas-piston activity preceding episode 6 at Puu 00. A, Estimated pond depth versus time for July 22, 1983. B, Schematic cross section of crater. Observers on northeast rim of crater recorded 11 filling and draining cycles between 1015 H.s.t. and onset of episode 6 at 1530 H.s.t. Breaks in plot reflect gaps in observation. During drain- ing of crater, lava disappeared from view at a depth of 3 m Within a 2-m-diameter pipe that extended downward from crater floor. It then quickly reappeared in pipe and slowly refilled bowllike crater. Filling normally occurred over an interval of 5 to 22 minutes, drain- ing took 1 to 3 minutes, and lava was out of View in pipe for 1 to 3 minutes. Prolonged filling that began just before 1300 H.s.t. flooded a low point in northeastern part of crater rim. Pond then drained at least once more before a rapid final filling that led to onset of erup- tive episode. Low bench at base of southwest crater wall marks buried site of a vent that built a mound of spatter on preceding day. C, Lava pond near end of filling. Solidified, fresh lava adhering to crater wall just above pond surface records one or more deeper previous ponds. Pond is about 30 min diameter and 10 m deep. A 1- to 2-m-high dome fountain plays on pond surface above conduit. Photograph taken at 1108 H.s.t. July 22, 1983. Numerals (lower right) indicate date and time. D, Lava pond beginning to drain. Crusted surface has become concave and is stretching and pulling apart as lava rapidly withdraws from beneath it. Abruptly increased gas emission has disrupted dome fountain and transformed it into a chaotic spray of spatter fragments. Photograph taken at 1110 H.s.t. July 22, 1983. Numerals (lower right) indicate date and time. E, Floor of evacuated crater after draining. Lava surface is momentarily out of view in glowing 2-m—diameter con— duit descending from floor. Photograph taken at 1059 H.s.t. July 22, 1983. Numerals (lower right) indicate date and time. 42 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 river system that fed three small channelized flows to the northeast. Additional filling of the crater occurred dur- ing episode 20; thereafter, at least through May 1985, the lava-pond, lava-river style of eruption did not recur. Con- sistently high fountains since episode 19 have greatly increased the size of Puu 00 and resulted in the accumula- tion of thick air-fall deposits beyond its downwind flanks. Vestiges of the crater were soon obliterated, and spatter- fed flows developed a thick wedgelike apron of aa surrounding the cone. CHRONOLOGIC NARRATIVE EPISODE 1 (JANUARY 3—23, 1983) SUMMARY OF EPISODE 1 After a 24-hour-long earthquake swarm that migrated from the Mauna Ulu-Makaopuhi Crater area to the Vicinity of Puu Kamoamoa (see chap. 7), eruptive activity began within Napau Crater at 0031 H.s.t. January 3, 1983. Erup- tion continued for 9.5 hours in the early morning of January 3. During that episode, a 6-km-long, discon— tinuous line of erupting fissures extended progressively downrift from N apau Crater to the 07 40 vent (pl. 1), then contracted to the vicinity of Puu Kamoamoa, and finally stopped erupting at 1000 H.s.t. Except for a brief erup- tion at the 0740 vent during the afternoon of January 3, no further discharge of lava occurred on January 3 or 4. Subsequent eruptive activity during episode 1 consisted of intermittent events at several different vents (pl. 1; FIGURE 1.51.—Open pipe intersecting floor of Puu Oo Crater, approx- imately 11/2 days before onset of episode 15. Lava surface is visible about 15 m down pipe. Accumulating spatter has built a 2- to 3-m- wide collar just above lava surface. View to upper right is across Spillway and through deep breach in northeast rim of crater. Spillway surface is about 5 m above top of pipe. View northward; photograph taken at 1126 H.s.t. February 13, 1984. Numerals (lower right) in- dicate date and time. table 1.4). On January 5 and 6, eruptive activity was main- ly concentrated at the 0740, 1123, and 1708 vents south of Puu Kahaualea. Fountain activity and minor lava pro- duction also occurred uprift at Puu Kamoamoa and at two vents uprift of Kamoamoa. Closely following an exten- sion of the earthquake swarm downrift to the vicinity of Kalalua, vents 1 to 2 km northeast of the 0740 vent opened during the morning of January 7 and erupted until the early morning hours of January 8. They produced a spectacular fissure-fed fountain and a voluminous lava flow that advanced down Kilauea’s south flank. From January 8 through 15, eruptive activity was localized south of Puu. Kahaualea, mostly at the 1123 and 1708 vents. Finally, a brief, very small eruptive outbreak occurred about 500 m downrift of Puu Kamoamoa on January 23. Altogether, during a total of approximately 99 hours of eruptive activity over a period of 20 days, episode 1 produced an estimated 14x106 m3 of new basalt that covered an area of 4.8x 106 m2. JANUARY 3, 1983 Guided by earthquake locations determined during the premonitory swarm, a crew of observers arrived early on January 2 at a point midway between Napau Crater and Puu Kamoamoa. There, they awaited the eruptive out- break and were rewarded when fountaining began at the base of the north crater wall of N apau Crater at 0031 H.s.t. January 3. The fountains, however, must have been FIGURE 1.52.—Floor of Puu Oo Crater, approximately 5 hours before onset of episode 14. Lava welling upward in open pipe (similar to that shown in figure 1.51) flooded crater floor and buried top of pipe shown in figure 1.18. Solid lava surface is about 30 m in diameter. Gas (being sampled here), spatter, and intermittently oozing pahoehoe escape from underlying magma column through a 0.5-m-diameter vent pene- trating crust. Most emission takes place during brief episodic bursts that suggest gas-piston activity beneath crust. View southwestward; photograph taken at 1239 H.s.t. January 30, 1984. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 43 TABLE 1.4.—Summary of episode 1 eruptive activity Be innin End Duration Area covered Volume (3.5.3 (H.s.t) “"1“” ' (h) (1.1.2) (106 m3) 0031 Jan. 3 1000 Jan. 3 Napau to 0740 9.48 1.52 3.0 1425 Jan. 3 1521 Jan. 3 0740 .93 .03(7) .05(?) 1123 Jan. 5 1125 Jan. 5 1123 .03 --- —-- 1214 Jan. 5 0955 Jan. 6 0740, 1123, 21.68 -—— -—— 1708. Puu Kamoamoa ----------- .03 .05 Uprift of Puu --— .21 .3 Kamoamoa. 1011 Jan. 6 2049 Jan. 6 0740 10.63 1“40.5 N1.5 on groundl, 20.5 in crevice . 0957 Jan. 7 0959 Jan. 7 January 7 .03 —-— —-— 1030 Jan. 7 1104 Jan. 7 January 7 .57 --— -—— 1111 Jan. 7 1558 Jan. 7 January 7 4.78 --— --- 0740 --- Minor Minor. 1623 Jan. 7 1634 Jan. 7 0740 18 Minor Minor. 1642 Jan. 7 1647 Jan. 7 0740 .08 Minor Minor. 1715 Jan. 7 0432 Jan. 8 January 7 11.28 22.33 25.5 0432 Jan. 8 0502 Jan. 8 0740 .50 Minor Minor. 1443 Jan. 8 1504 Jan. 8 0740 .35 Minor Minor. 1957 Jan. 8 2322 Jan. 8 1708 3.42 ——— ~— 0041 Jan. 9 0330—0730 Jan. 9 1708 5(?) --- —-- 1713 Jan. 9 2100 Jan. 9 1708 3.78 ——- -—- 0502 Jan. 10 0625 Jan. 10 1708 1.38 —-— --- 0759 Jan. 10 1450 Jan. 10 1708 6.85 --— --— 0130 Jan. 11 ~1230 Jan. 11 1708 “41 ——— -—— 0312 Jan. 15 0855 Jan. 15 1123, 1708 5.72 3«0.7 3N3 1830(7) Jan. 23 1930(7) Jan. 23 January 23 1(7) .004 .01 IEstimated total area and volume on January 5 and 6 for the 0740, 1123, and 1708 vents combined. Area and volume for total production of lava from the January 7 vents. 3Estimated total area and volume from January 8 through 15 for the 1708 and 1123 vents combined. less than 30 m high; the observers, looking uprift to the southwest, could see only brightly illuminated fume above the Napau vents. Beginning at 0155 H.s.t., the vent system extended northeastward, forming a progressively lengthening line of segmented eruptive fissures parallel to the axis of the rift zone (fig. 1.53). Although the overall tendency was for downrift (northeastward) extension of the vent system (fig. 1.54), the sequence of vent opening was complicated in detail. In some places, new vents opened uprift of already-active ones. Where less active, the erupting vents spattered weakly; where more active, they produced low, linear fountains, generally about 10 to 30 m high, and fluid lava flows that spread mainly southeastward (fig. 1.55). By 0300 H.s.t., the line of erupting vents was more than 4 km long and had transected the prehistoric cinder cone, Puu Kamoamoa. By 0428 H.s.t., the line had extended more than 5 km northeastward from the initial vent at Napau Crater. At 0740 H.s.t., the easternmost vent of January 3, approximately 6 km downrift from the Napau vents, erupted briefly in dense rain forest south of Puu Kahaualea. By the end of this initial eruptive event, the zone of shallow earthquake activity had extended downrift to the vicinity of Puu Kahaualea; from then through mid- day January 6, shallow earthquakes were concentrated in a zone extending from between Napau Crater and Puu Kamoamoa on the west to the vicinity of the 0740 vent on the east. Although the final extrusive activity on the morning of January 3 was concentrated near the center of the line of erupting vents, the sequence in which the vents stopped erupting was irregular. By the time the easternmost vent began to erupt at 0740 H.s.t., active eruption elsewhere was confined to a 900—m-long stretch centered on Puu Kamoamoa (fig. 1.55). The 0740 vent erupted for only 20 minutes before shutting down at 0800 H.s.t. The vents immediately uprift and downrift of Kamoamoa died at 0850 and 0939 H.s.t., respectively, and by 1000 H.s.t. the first eruptive event of episode 1 was over when the vent at Puu Kamoamoa also stopped erupting. Most of the lava produced in this first event, during the morning of January 3, flowed over basalt from eruptions in the 1960’s (fig. 1.56). The flows of January 3 were short; the longest advanced about 2 km from its vent. The lava was predominantly pahoehoe, although the longer, more channelized flows formed aa (pl. 1). Measured flow thick- ness ranged from about 1 to 3 m; an estimated 3 x 106 m3 of new basalt was emplaced over an area of 1.5 x 106 m2. The vents were marked by deposits of spatter, ranging from a thin veneer where activity was weak and extru- sion minimal, to linear ramparts several meters high in areas of prolonged vigorous eruption. Shortly after 1200 H.s.t. January 3, production of pro- fuse, hot, SOg-rich fume was observed along newly lengthening cracks about 1 km east of the 0740 vent in a locality where new vents would erupt about 4 days later 44 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 on January 7. Also, from 1425 to 1521 H.s.t., the 0740 vent erupted for a second time, producing a 20- to 30-m- high fissure fountain and a small amount of pahoehoe. N 0 further eruptive activity occurred until January 5. JANUARY 5 AND 6, 1983 For nearly 2 days after the eruptive activity of January 3, Kilauea was in repose, and the summit reservoir reinflated slightly. During this period, many of the vents of January 3 emitted copious visible fume and burning gas flares. After summit deflation resumed, accompanied by increasing harmonic-tremor amplitude, the eruption itself resumed on January 5. The major eruptive activity of January 5 and 6 occurred along a 1.1-km-long system of fissure vents south of Puu Kahaualea. This system in- cluded two new fissure vents that filled the gap between the 0740 vent and the nearest January 3 vent to the southwest. The more easterly of these new vents first erupted at 1123 H.s.t. January 5, and the more westerly at 1708 H.s.t. (fig. 1.57). Minor amounts of lava were also discharged several kilometers uprift. The initial eruptive event of January 5 was a 2-minute fissure eruption that began at 1123 H.s.t. at one of the new vents. Then, after a 49-minute respite, vigorous erup- tion began at the 07 40 vent, and for more than 32 hours (table 1.4) the rift zone was in nearly continuous eruption. For most of that time, the 0740, 1123, and 1708 vents erupted, sometimes singly and, at other times, in concert (fig. 1.54). Intermittent, relatively low fountaining and 155°07’30" 155°05’30" | l | 1 19°25. _ EXPLANATION _ Lava type 3 Pahoehoe if) N 9 an 0 VI _ Limit of flow 5 <3 8 WM- Flssure with spatter rampart Puu KahaualeaA ++—H— Open crack Cam A .............. Approximate buried limit of January 3 basalt m p E v a, 8 _ o N g Puu Kamoamoa 8 ’5 l o v a o a: /\ l 9 O (V) l m o Q to o 0': LD N v ‘n T 9 N ' ‘9 8 m m 3 ° 9 l m in V V a) N N O O .— o o l 5 g g VI VI ‘1 3 9 8 $ 2 1.0 o O N 8 m a a '3 ° .' ' Ix m < o o v o N o N v V 5 3 l O A- V o m o "’ l < . Vv u: :z a n 9 w 8 v v to !\ O 8 F :‘2 V V 00 A . “ o v vvv A v r m 8 ' v v v , to m w V — 532 * ’6! 19°23' — g 0 0 V _ 0 AH’% W" t o 1 2 KILOMETERS L I l | l l l FIGURE 1.53.—Episode 1 vents and flows, from Napau Crater to vicinity of Puu Kahaualea, with times (H.s.t.) and locations, deduced from field notes, verbal reports, video tapes, time-lapse film, and photographs, for beginning and end of eruptive activity for each vent active on January 3, 1983. Flows from vent that opened at 0428 H.s.t. and from vents uprift (southwest) of it were mostly emplaced on January 3. Small volume of lava (estimated at max 0.1 x 106 m3) emplaced northeast (downrift) of 0428 vent was buried by lava erupted during later parts of episode 1. 15 14 13 10 DAYS IN JANUARY 1983 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY 0F ERUPTIVE EVENTS 45 I I l I l I I Western - January 7 vents [ Central -—- Eastern (u o-ou E — ° 55 a) E >> > 8 com o E on 0330 <0730 1713—2100 2050270625 0759—1450 0130;1230 0312<0855 19°25' H: 15: Jan Jan 5: 1358-1 527 <2135—2354 0002—0048 010540109 0405—<0605 0630—>0734 <0822 1708-1722 17554800 Jan 5: <1546- >1610 0 6: <172 Puu Kamoamoa Jan 6: 2005v2015 19°23' Western Central Eastern Jan 7:10304104 Jan 7: 0957—0959 Jan 7: 111771450 _ 1408-1558 1111—1450 1715- 8: 0432 15034553 )3 $35“ 1123 vent M 5:1123—1125 am 15'“ 1633—1910 Y“ //‘_‘,.q 6: <1252— >1304 {7‘7 Puu Kahaualea or“? 15: 031240855 Q) Jan 23: 18307—19307 January 7 vents CampA" X Crevnce 0740 vent Jan 5:121471359 1507—1910 :0006—0412 043570732 1011A2049 7:150371536 162341634 1642—1647 :0432—0502 1443-1504 2 KILOMETERS l J l I FIGURE 1.57.—Episode 1 vents and flows, with times (H.s.t) and locations, deduced from field notes, verbal reports, time-lapse film, and photographs of eruptive events, January 5—23, 1983. Closely spaced diagonal pattern shows approximate locations of January 5—6 basalt erupted at Puu Kamoamoa and uprift of Puu Kamoamoa; otherwise, symbols are same as in figure 1.53. 48 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 FIGURE 1.58.—Incandescent vents emitting burning gas. West end of 0740 vent is at far left, and 1123 vent is at center and right center. Cone at far right, about 15 m high, is at east end of 1708 vent. View southwestward; photograph taken at 1814 H.s.t. January 14, 1983. Numerals (lower right) indicate date and time. who FIGURE 1.59.—Lava from 0740 vent, cascading over 10-m-high scarp and disappearing in open crevice. Eastern part of 0740 vent was in- active at this time. Low fountains at 1123 vent, partly obscured by smoke and volcanic fume, are beyond trees at left. Mauna Loa is on skyline. View westward; photograph by J .P. Lockwood, taken at 1304 H.s.t. January 6, 1983. ‘7 FIGURE 1.60.—Basalt flows of January 5-15, remnants of open crevice, and trace of graben that extended northeastward from it. Solid line, graben boundary; bar and ball on downthrown side. Dashed line, open crack. Letters A—D identify features described in text. Hill at camp A is a 1965 spatter cone, about 12 m high. Narrow lobe of January 11 basalt separates cone from rain forest to north (upper right). View westward; photograph 83.1.26JG135A#34 by J .D. Griggs, taken at 0949 January 26, 1983. Numerals (lower right) indicate date and time. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 49 crevice that were still exposed at the end of episode 1. About 100 In to the northeast (C, fig. 1.60), the graben transected January 5—6 basalt with sufficient throw that prehistoric basalt was exposed in the graben walls beneath the new basalt. About 50 m to the north, at locality D (fig. 1.60), a 1—m-wide, open crack, also cutting January 5—6 basalt, disappeared beneath unbroken basalt of January 8—9. Basalt of January 7—8, unbroken except for a nar- row crack, with about 20 cm of dilation, on the north- eastward extension of the northern graben-bounding fault, buried the downrift end of the graben. Extension normal to the rift-zone axis probably also oc— curred as the feeder dike for the 1708 vent approached the surface. New extension cracks were observed cutting January 3 basalt near the uprift end of the 1708 vent late in the evening of January 5. Cracks mapped later in this area (pl. 1) transected January 3 basalt and are visible in photographs taken at about 1100 H.s.t. January 9. The lava flows of January 5—6 consisted of fluid pahoe- hoe that spread in a sheet surrounding the vents south of Puu Kahaualea and that formed thin flows extending southeastward from the vents farther uprift (fig. 1.57). These thin flows, partly on top of January 3 basalt, covered an area of 0.2><106 to 0.3x106 m2; their esti- mated volume was 0.3><106 to 0.4x 106 m3 (table 1.4). Basalt from the 1708, 1123, and 0740 vents was almost completely buried by younger flows during the period January 8—15; it covered an estimated area of 0.5x 106 m2, and its volume on the ground was probably about 1.5x 106 m3. An additional 0.5x 106 m3 is estimated to have disappeared into the crevice. JANUARY 7 AND 8, 1983 Shallow earthquakes, which had been persistent in the eruptive area most of the time since January 2, dimin- ished in frequency at midday January 6. They became abundant again just after 0000 H.s.t. January 7 in the vicinity of Kalalua, farther downrift than the swarm had previously extended. The swarm died late on January 7, and no further abundant shallow earthquakes were recorded in the rift zone through the entire period dis- cussed in this report. Accelerated surface deformation in the same general area began during the evening of January 6 and, by midnight, had caused profound disturb- ance of a borehole tiltmeter (KLU) about 300 m north of Kalalua (fig. 1.1). Daylight on January 7 showed that this tiltmeter was within a northeast-trending zone of new cracks. Immediate remeasurement of a horizontal line crossing the zone of cracks north of Kalalua indicated that approximately 2 m of extension normal to the trend of the rift zone had occurred since the last measurement on January 5. These events apparently recorded emplace- ment of the easternmost segment of the new dike system that had been delivering lava to the surface farther uprift since January 3. Observers stationed between Puu Kahaualea and the 0740 vent felt some of the earthquakes and noticed heavy steaming downrift toward Kalalua during the early hours of January 7. Beginning at about 0230 H.s.t., a glow was recognized in the same area, and minor spattering was seen at about 0430 H.s.t. When daylight came, direct observation showed burned vegetation and a glowing fissure that was emitting fume, but no new flows or vent deposits were found. The eastern part of this new fissure system was in the» center of a shallow, 100- to 150—m-Wide graben with fresh bounding cracks that gaped at least 1 m. A minor eruption on the central part of the new fissure system began at 0957 H.s.t. January 7 (figs. 1.54, 1.57). This eruption was brief, probably no longer than 2 minutes, and produced a low, 60-m-long spatter rampart. Eruption began in earnest on the January 7 vents at 1030 H.s.t. From then until 1104 H.s.t., the western part formed a low (3—10 m high) fountain, about 300 m long. Beginning at about 1115 H.s.t., the central and eastern vents began to erupt. Within the first hour, the 500—m- long fountain at these vents grew to a height of 60 to 100 m, forming a spectacular display; intense fountain- ing continued until 1347 H.s.t., when the high—level activity began to wane rapidly. By 1400 H.s.t., fountain heights decreased to no more than 6 m on the central and eastern vents. At 1408 H.s.t., the western vent, which had been inac- tive since its initial eruption about 3 hours earlier, began erupting again. Within the first few minutes, the western fountain line grew to a height of 50 to 60 m, which soon waned; by 1440 H.s.t., it had decreased in height to no more than 10 m and continued thus. Activity on the still- diminishing central and eastern vents stopped at 1450 H.s.t. At 1503 H.s.t., low fountaining resumed on the cen- tral vent and continued there, as well as on the still- erupting western vent, until 1558 H.s.t., when both vents shut down. The massive midday eruption of the central and eastern vents produced a wide, rapidly advancing flow that ex- tended 5.5 km southeastward (pl. 1) through rain forest on the south flank of Kilauea. For most of its advance, the flow, seen from the air, was wide and consisted of pahoehoe. At 1440 H.s.t., it was about 4.7 km from the vents. Averaged from the time the central and eastern vents began to erupt, about 2.5 hours earlier, the rate of flow-front advance was 1.3 to 1.4 km/h. However, the eruption did not begin at full vigor but was increasing over the first hour, and aerial observations of the flow-front position (pl. 1) during and shortly after the most vigorous part of the eruption suggest that the average advance rate down the south flank at that time was 1.5 to 2.0 km/h; the maximum rate was probably greater than 2.0 km/h. 50 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 When seen again, at 1520 H.s.t., the front had advanced only another 400 m; the rate was slower, and the front consisted of aa and slab pahoehoe. Observed after it had stopped moving, the flow consisted of aa from the area south of Kalalua to the terminus. The aa was thin, general- ly 1.5 to 2.5 m thick; it had enveloped still-standing ohia trees, and it contained vertical molds where the trees had been burned away. N 0 further eruption occurred on the central and eastern vents. However, the western vent erupted at a low to moderate level for an additional 11 hours, from 1715 H.s.t. January 7 to 0432 H.s.t. January 8. The fountain line was relatively low during that period; it was commonly 5 to 15 m high and occasionally reached a height of 20 to 30 m. The line of continuous fountains reached a maximum length of 200 m for a brief period early in the evening. For much of this episode, however, the line was discon- tinuous and the activity sporadic except in the central part, which was the most steady and vigorous. A minor flow, pahoehoe near the vent and aa to the southeast, ex- tended eastward and southeastward on top of the earlier flow from the January 7 vents; it terminated in the area south of Kalalua, about 2 km from its vent. Flows produced by the January 7 vents spread across relatively flat country bounded by a narrow graben, 300 to 400 m southeast of the vents (Holcomb, 1980). These flows crossed the graben west of Kalalua and extended from there down the south flank of the volcano. They covered an area of 2.33x 106 m2 and had a volume of 5x 106 to 6x 106 m3. We estimate that about 4 x 106 m3 of new basalt was emplaced by the midday eruption on January 7 from the central and eastern vents. That event lasted for nearly 5 hours (table 1.4), at an average lava-discharge rate of about 0.8 x 106 m3/h. However, because the greater part of that voluminous flow to the southeast was erupted over a 2- to 3-hr period, the max- imum lava-discharge rate was well in excess of 106 m3/h and could have been as high as 1.5 x 106 m3/h. In contrast, we estimate that the later eruption, from the western vent during the night of January 7—8, produced lava at a rate of less than 105 m3/h. Intermittent, brief (5—34 minute long) eruptions with very low production of spatter and flows occurred several times on January 7 and 8 at the 0740 vent (figs. 1.54, 1.57). Hot, air-rich gases jetting from the 0740 vent also caused occasional minor ejection of spatter and, at one orifice, created a roar that was audible 20 km to the west in Kilauea’s summit region. JANUARY 8-15, 1983 After the last minor eruption from the 07 40 vent dur- ing the early afternoon of January 8, no further episode 1 lava emission occurred there; nor did the January 7 vents, farther downrift, ever erupt again. Intermittent eruptive activity, however, occurred from late on January 8 through early January 15 at the 1708 vent and, in part, at the 1123 vent (figs. 1.54, 1.57; table 1.4). Heated gases, occasionally carrying fragments of incandescent ejecta, continued to jet from the recently active vents. The main activity of the 17 08 vent during this period took place in a series of moderate eruptions that occurred between 1957 H.s.t. January 8 and approximately 1230 H.s.t. January 11. During this 62.5-hour interval, in- dividual eruptive events ranged in length from about 1.4 to 11 hours, and the 1708 vent was erupting about 50 per- cent of the time. Eruptions were characterized by low fountains, generally less than 15 m high. At the east end of the vent, spatter from the fountain built a prominent, 15-m-high cone (fig. 1.58). Pahoehoe or, on occasion, slab- pahoehoe flows moved northward and southward from the vent. Flows to the north tended to be more voluminous; some flowed eastward on the north side of the 0740 vent and covered much of the area of the January 5—6 basalt. On the night of January 8—9, northeast-moving flows onlapped the base of the small 1965 spatter cone at camp A, 300 m north of the 0740 vent. Intense radiant heat from the active pahoehoe-flow edge and concern that camp A might become surrounded by active lava led to a middle-of-the-night evacuation by helicopter, and so no one observed the end of this eruption, which occurred sometime between 0330 and 0730 H.s.t. January 9. A later flow from the same vent on January 11 produced a lobe that passed on the north side of the camp A spatter cone, so that all but the northeast base of the 12-m-high hill was surrounded by new basalt (fig. 1.60). Visibility was so poor during that event that we could estimate only an approximate time, 1230 H.s.t. January 11, for its end. After slightly more than 31/2 days of repose, a final episode 1 eruption in the area south of Puu Kahaualea occurred on January 15. The 1123 vent was the primary lava producer, but the 17 08 vent also erupted. At about 0215 H.s.t. January 15, showers of incandescent frag- ments began to be ejected from the 15-m-high spatter cone that had formed previously at the east end of the 17 08 vent. Reaching heights of 30 to 50 m above the top of the cone, they formed a spectacular fireworks display. After about 15 minutes, the style of eruptive activity evolved from a more or less continuous shower of small glowing fragments to intermittent bursts, accompanied by loud reports, that ejected large chunks and ribbons of spatter, as if large gas bubbles were bursting violently within lava pooled at a shallow level in the vent. Such gas-burst ac- tivity with ejection of clots and ribbons of spatter was typical throughout episode 1 at the end of eruptive events when lava production had largely waned or stopped. The gas-burst activity, alternating with renewed showers of 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 51 small incandescent fragments, continued at the 1708 vent and extended eastward along the 1123 vent, where the activity earlier in the evening had consisted only of con- tinuous emission of flaring gas (fig. 1.58). By 0312 H.s.t., the volcano was truly erupting. A low fountain, about 100 m long, had formed at the 1123 vent, and a low, sustained fountain was also playing at the 1708 vent. By 0315 H.s.t., we could recognize that a lava flow was beginning to move northward from the 1123 vent. This moderately vigorous eruption, with the most promi- nent fountains as much as 20 to 25 m high, continued through the night, producing a pahoehoe sheet that spread southward, eastward, and northeastward. It surrounded the 0740 vent and covered much of the basalt that had been emplaced in previous days between the 0740 vent and camp A. In the days following the January 15 eruptive activity, the vents south of Puu Kahaualea continued to emit gas that formed burning yellowish-orange flares and, at times, small fragments of incandescent ejecta. However, no fur- ther episode 1 eruptive activity occurred in that area, and the 17 08 vent never erupted again. The volume of lava produced by the 1708 and 1123 vents during the period January 8—15 is difficult to estimate because we were unable to map the flows carefully be- tween eruptions. Measurements around the flow edges and in empty tree molds indicate that 2 m is reasonable for the average thickness of any one of these pahoehoe flows. Flows erupted on January 8 and 9 were partly superimposed and formed a widespread sheet that prob- ably covered an area of about 0.5x 105 m2. The area covered by superimposed flows is uncertain, but we esti- mate that 1.5 x 106 m3 of basalt may have been emplaced on January 8—9. The January 15 eruption produced a pahoehoe sheet that covered an area of about 0.4x 106 m2; a reasonable volume estimate for that flow is 0.8 x 106 m3. Although the areal extent of flows produced on January 10 and 11 is the least certain, they did not ex- tend far enough eastward to reach the area between the 0740 vent and camp A. Sketch maps and mapping done after the January 15 eruption suggest that the areas of the January 10 and 11 flows may have been about 0.1x106 and 0.2x106 to 0.3)(106 m2, respectively. An estimated volume of 0.7x 106 m3 seems reasonable and leads to an approximate total of 3 x 106 m3 of new basalt for the period January 8—15. JANUARY 23, 1983 A final, minor eruptive event that occurred on January 23 is included with episode 1. Photographers at camp A, north of the 0740 vent, reported glow and fountaining uprift at about 1910 H.s.t. January 23; a tour-plane pilot also reported seeing eruptive activity from the air at about 1830 and 1930 H.s.t. Reconnaissance the next day showed a small, new vent and flow superimposed on one of the January 3 fissures about 400 m northeast of Puu Kamoa- moa (fig. 1.57). Approximately 9,000 m3 of new lava had been erupted to form a 55- by 70-m pad. Possibly only coincidentally, this event followed an M = 4.4 earthquake that occurred at 1800 H.s.t. beneath Kilauea’s south flank. EPISODE 2 (FEBRUARY 10—MARCH 4,1983) During the 21/2 weeks between the final eruptive activity of episode 1 on January 23 and the onset of episode 2 erup- tion on February 10, 1983, no measurable volume of new lava was discharged on Kilauea. However, the new vents south of Puu Kahaualea remained incandescent, continued to emit burning gases, and, on occasion, ejected small amounts of spatter. In addition to these conspicuous signs of continuing shallow magmatic activity, low-amplitude harmonic tremor persisted in the eruptive zone (see chap. 7), and extension of about 1 cm/d was measured on a line across the trace of the 07 40 vent (see chap. 6). New flows emplaced between February 12 and 14 obstructed this line and ended the measurements. Increased production of spatter was first noticed on February 10. That morning, we discovered two weakly spattering vents at the west end of the 0740 fissure (pl. 1), which was also emitting copious dirty-brown fume. The more active vent had built a new, 6—m-high spatter cone. Less conspicuous fume was also issuing from the eastern 50 m of the 1123 fissure, and the easternmost part of that fuming fissure was incandescent. Additional fume was be- ing emitted from an incandescent fissure east of Puu Kamoamoa in the approximate area where Puu 00 would eventually develop. Except for emission of water vapor, the rest of the episode 1 fissure system appeared to be inactive. By February 12, a second small cone had formed at the west end of the 0740 vent, and a new glowing crack ex- tended tens of meters northeastward of these two cones. The rate of growth of the cones was low; small pieces as well as larger chunks and ribbons of spatter, several tens of centimeters in diameter, were being ejected a few fragments at a time in small intermittent bursts. The glowing crack at the east end of the 1123 vent still per— sisted, and the zone of fume emission had extended uprift along the 1123 fissure so as to include the western part as well. Time-lapse film shows that a brief, low-level spat— tering event occurred at the 1123 vent during the eve- ning of February 12. By February 14, continuing intermittent low-level erup- tion along the 07 40 fissure had extended the line of low spatter cones northeastward over the glowing crack seen on February 12. The cones surmounted a low shield (fig. 52 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 1.61) formed by slow discharge of pahoehoe that spilled from the vents onto the surface of the shield or leaked through lava tubes into its interior. Similar activity con- tinued until February 25, by which time the individual cones, generally about 10 to 15 m high, were juxtaposed in a 170-m-long line that formed a cockscomb—like crest on the flat-topped, 10-m—high shield. We estimate that, over the 15—day period from February 10 to 25, about 0.5 x 106 m3 of lava was extruded to form the cones, the shield, and thin, short pahoehoe flows that extended beyond the limits of the shield. The intensity of eruptive activity increased markedly at about 0900 H.s.t. February 25 and became localized at two vents that formed in the western part of the line of cones (fig. 1.62). The central and eastern parts, which had been active during the preceding night, never erupted again. Steady eruption continued until midafternoon, pro- 13mm" ducing a pahoehoe flow that quickly extended northward toward camp A and then turned northeastward. Although the estimated lava-discharge rate was about 50,000 m3/h, relatively little spatter was ejected. Lava seemed to well out of the more easterly of the two vents, and at the western one, it erupted in a continuous, but somewhat varying, northwest-directed stream (fig. 1.62) that played for hours and resembled a stream of water issuing under slight pressure from a pipe or hose with no nozzle. The style of lava emission at these two vents suggested that the erupting lava was gas poor, possibly owing to shallow subsurface degassing, as indicated by abundant emission of SOg-rich fume at nearby nonerupting vents (fig. 1.61). At about 1430 H.s.t., new fountain activity broke out 100 to 200 m uprift at the 1123 vent (pl. 1), where a 50- to 100-m-long line of fountains formed. The most vigorous part was at the uprift end, where the fountain was about FIGURE 1.61.—Weakly erupting 0740 vent spatter cones surmounting low pahoehoe shield during early. part of episode 2. Dark basalt in foreground is from episode 1. Fault scarp (lower left), with crevice at base, faces south. Numbers in lower right indicate date and time. View southwestward; photograph 83.2.14NB135B#34A, taken at 1634 H.s.t. February 14, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 53 20 m high (fig. 1.21). In the rest of the line, to the north- east, the fountains were no more than a few meters high. At 1518 H.s.t., a new fountain line became active about 900 m uprift of the 0740 vent. Simultaneously, the western part of the 0740 vent stopped erupting and, ex— cept for occasional bursts of spatter during the next 2 days, never erupted again. The vent openings, however, continued to glow and to emit fume throughout episode 2. The new uprift vent formed a line of low fountains, estimated at 5 to 10 m high, that erupted until sometime between 0930 and 1100 H.s.t. February 26, producing a pahoehoe flow that advanced slowly eastward for several hundred meters. An additional erupting vent, the O vent, which was even farther uprift (pl. 1), was first observed at 0140 H.s.t. February 26. Because the O vent was nearly 2 km from camp A, its activity was difficult to watch closely. At first, the activity was probably low and sporadic. However, the O vent was seen erupting vigorously during the early evening of February 26, during the predawn hours of February 27, and at midafternoon on February 27. Observers in passing aircraft estimated that the fountains were 40 to 50 m high. In addition, time-lapse film data indicate that intermittent activity continued until at least about 1900 H.s.t. February 27. As far as we know, the O vent did not erupt at any later time during episode 2. Its eruption formed a 300- to 400-m-wide pad of pahoehoe (pl. 1). The main activity of episode 2 was centered at the 1123 vent. A nearly circular lava pond (fig. 1.5) quickly formed. This pond enveloped the line of fountains erupting from the eastern part of the 1123 vent and initiated the central- fountain, lava-pond, lava-river style of activity that would characterize this series of eruptions for months to come. The main fountain, which dominated the activity, played in the western part of the pond. Its fallout quickly built a prominent rampart of spatter and Cinders that enclosed the pond on the south and southwest; by the end of episode 2, this rampart was about 25 m high. The rest of the enclosing levee, though partly of pyroclastic origin, con— sisted largely of a bulwark of pahoehoe built by overflows and by lava that had leaked from the pond through short lava tubes. The more northeasterly fountains, which were initially low and in a continuous line with the main foun- tain, soon evolved to a discontinuous line of fountains barely rising above the surface of the pond. In addition, a discontinuous line of small fountains opened uprift from the pond at about 0100 H.s.t. February 26 and erupted intermittently and at low levels thereafter. For the first 2 days of its activity, the main fountain at the 1 123 vent was relatively low; its height ranged from about 20 to 60 m above the surface of the pond (fig. 1.21) and was at the low end of this range on February 26. On February 27, the vigor of the fountain increased, and, for about 4 days, the main body of the fountain formed a broad column that was about 60 to 80 m tall most of the time. The intensity of fountaining waned on March 3 (fig. 1.21); the erupting column, when it could be seen through the mist and the volcanic fume that was being blown toward camp A by southerly winds, was about 30 to 50 m tall. Poor visibility on March 4 virtually eliminated any useful time-lapse-camera record. Occasional glimpses showed onsite observers that the fountain was low and varied in behavior, and several times, from about 1000 H.s.t. until it finally shut off permanently at 1451 H.s.t., the fountain apparently shut off or at least dropped to a level so low that it could not be seen, and diminished lava production was reported. Beginning at about 1200 H.s.t., observation of the channel near the vent showed that occasional temporary cessations of fountaining coin- cided with temporary cessations in the supply of lava to the lava river south of the vent. When fountaining would resume, the lava pond would refill to overflowing and reactivate the river. Two successive river-fed lava flows advanced from the 1123 vent (pl. 1). Initially, the lava flowed southward from the vent and then turned northeastward along the same narrow graben that formed a local southeast boundary for episode 1 flows on January 7—8. The episode 2 lava flowed partly on top of episode 1 basalt to the vicinity of Kalalua. In about 19 hours, from the time the 1123 vent opened on February 25 until the active flow terminus was near Kalalua at 0915 H.s.t. the next morning, the flow FIGURE 1.62.—“Firehose” fountain at west end of 0740 vent during episode 2. Line of spatter cones is 170 m long. At left (east) end, tallest cone, which is closest to camera, is about 17 m high and 330 m dis- tant. West end of line is about 400 m from camera. Shiny pahoehoe in foreground is at edge of an active flow from erupting vents. Photograph by J .D. Griggs, taken at 1105 February 25, 1983. Numer— als (lower right) indicate date and time. 54 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 extended about 2.7 km along a relatively flat course at an average rate of approximately 140 m/h. Between the mornings of February 26 and 27, the lava river was diverted about 1 km from the vent, possibly owing to a decrease in lava-discharge rate that permitted the northeastward-flowing lava river partly to freeze. No major active flow lobe was seen between the mornings of February 26, when the northeastern lobe was in place, and February 27 , when the southeastern lobe was first recognized (pl. 1). The fountain height at the 1123 vent was also low during that interval (fig. 1.21). In addition, the rate of summit deflation and the amplitude of har- monic tremor were lower during about the same period. The new distributary lobe, first recognized at 0830 H.s.t. February 27, became the main flow; it was supplied by the channelized pahoehoe river shown in figure 1.5. About 1 km from the vent on March 3, we estimated lava flux in the approximately 15-m-wide channel (fig. 1.30). Maximum velocity (2.4 m/s) was determined by timing the passage, through a measured distance (30 m), of a distinc- tive pahoehoe slab or an object tossed into the most rapidly moving, central part of the lava river. Because velocity decreased to zero at the edges of the channel, we made the simplifying assumption for volume calculations that the average velocity in the channel was half the max- imum velocity. For an assumed depth of 2 m, the esti- mated lava flux was about 130,000 m3/h. Similar estimates of flux close to the vent during the main part of the eruption ranged as high as about 3 times this value. At average velocities of 40 to 70 m/h, the toe of the southeastern flow extended through the rain forest to Royal Gardens, which it reached at 1720 H.s.t. March 2. During its advance through the northwestern part of this sparsely populated subdivision, the aa-flow front burned and crushed one house. The advance rate decreased toward the end of the eruption. Average advance rates of approximately 30 m/h and then 20 m/h are indicated for the last 21 hours of episode 2 (pl. 1). During the 24 hours after the eruption finally stopped, at 1451 H.s.t. March 4, the flow front advanced about another 8 m. Three surges were recognized as the episode 2 flow traversed the corner of the subdivision: at 1745 H.s.t. March 2, 2230 H.s.t. March 3, and about 0730 H.s.t. March 4. The second surge was the best documented. At 2200 H.s.t., a 2-m-high;flow front was moving down Queen Street (pl. 1) at only a small fraction of a meter per minute. This flow front slowly thickened to about 6 m and began to move more rapidly. From 2230 to 2300 H.s.t., it surged ahead about 200 m in 30 minutes and simultaneously thinned to about 2 m. After this surge, the flow resumed slow movement in the forest northeast of the street. In total, vigorous eruption during episode 2 continued from 0900 H.s.t. February 25 to 1451 H.s.t. March 4, a period of nearly 174 hours. In the early part, February 25 to 27 , relatively small amounts of basalt, predominantly pahoehoe, issued from the 0740 vent, the O vent, and an additional vent between these two (pl. 1). The main pro- duction, from the 1123 vent, produced a flow, mostly aa, with a small early lobe to the northeast and a major later lobe to the southeast. The total volume of new basalt erupted during the 71/4 days of vigorous eruption was approximately 13.6 x 106 m3, a value suggesting an over- all. lava-discharge rate of 70,000 m3/h. An estimated 11.3x 106 m3 of basalt composed the major lobe to the southeast, which ranged in thickness from approximate- ly 3 to 10 m. An average lava-discharge rate of about 90,000 m3/h is indicated for the 1123 vent during the 51/4-day period from 0830 H.s.t. February 27 , when the southeastern lobe was first recognized, to the end of the episode. This vigorous episode was preceded by 15 days of slow, intermittent eruption of the 0740 vent that produced about 0.5 x 106 m3 of pahoehoe; an average lava-discharge rate of about 1,400 m3/h is indicated. The total volume of basalt produced during episode 2 was approximately 14x106 m3 over an area of 2.7x 106 m2. EPISODE 3 (MARCH 21—APRIL 9, 1983) The repose period between episodes 2 and 3 lasted 23 to 24 days, during which time the 0 vent glowed and emitted fume. This repose period was briefly interrupted about a week before the beginning of episode 3 by low-level eruptive activity at the 1123 vent on March 21. Harmonic- tremor amplitude gradually doubled in the eruptive zone from 0430 to 0630 H.s.t. that morning, and glow reported- ly was seen over the vent area between 0530 and 0600 H.s.t. Aerial reconnaissance from 1030 to 1100 H.s.t. showed intermittently active fountains, a few meters high, feeding short (max 20 m long) pahoehoe flows within and just west of the ring of spatter that had formed at the 1123 vent during episode 2. Reconnaissance early the next morning showed that the vents were still glowing but that eruptive activity had stopped. As far as we know, it did not resume before the onset of episode 3. Episode 3 began in the early morning hours of March 28. The first report of intense glow in the eruption area was at 0230 H.s.t., and at 0300 H.s.t., the glow was visible from HVO. Harmonic-tremor amplitude had increased slightly above the normal repose-period background on March 27 , and early on March 28 it began to increase rapidly. By 0100 H.s.t., the amplitude had increased about fivefold, and episode 3 was probably under way. Initially, the dominant activity was localized in the vicinity of the O vent, which erupted steadily for nearly 3 days. When we first arrived, just after 0800 H.s.t., March 28, fountains there were issuing along a 100- to 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 55 150-m—long line (fig. 1.8), and active pahoehoe flows ex- tended 50 to 100 m northward and several hundred meters southward. The most voluminous lava production was at the O vent itself (pl. 1), where a columnar fountain rose from a circular lava pond bordered on the west by a rampart of accumulating spatter (fig. 1.8). The fountain ranged in height from about 90 to 130 m into the early hours of March 29 (fig. 1.22); subsequently, it gradually diminished in height to about 50 m before eruption at the O vent stopped at 2019 H.s.t. March 30. Near the center of the line of erupting vents on March 28, two smaller fountains were building a pair of juxta- posed spatter rings containing craters with lava ponds (fig. 1.8). In addition, very small fountains were erupting between this pair and the large eastern vent, and still another small fountain was erupting at the west end of the line of vents. By March 29, the two vents at the center of the line had coalesced to form a single crater within a rim of agglutinated spatter; this crater was the pro- genitor of Puu 00. It continued erupting at a relatively low level until the O vent stopped erupting on March 30; much of the time, its fountain was barely as high as the encircling spatter rampart. The main flow produced at the O vent was a narrow lobe of aa that extended approximately 5 km southeast- ward and advanced at an average rate of about 60 m/h (pl. 1). Its sources were two pahoehoe rivers from the line of vents. The larger river issued from the O vent, at the east end of the line, and the smaller river from the coalesc- ed pair of craters near the center. The flow covered an area of 1.4 x 106 m2 and was mostly 3 to 5 m thick (pl. 1); we calculate its volume to be about 5.4x 106 m3. For an approximate eruption time of 67 hours, this volume gives an average lava-discharge rate of about 80,000 m3/h. Although the O vent was initially dominant, the 1123 vent became the main vent and the biggest lava producer of episode 3. After a hesitant start, the 1123 vent began erupting steadily, forming a spectacular fountain that at times was more than 200 m high (fig. 1.23). For part of episode 3, these two vents erupted together (fig. 1.63). The 1123 vent also produced several thick aa flows, one of which devastated a part of the Royal Gardens subdivi- sion (pl. 1). When we arrived on the first morning of episode 3, shortly after 0800 H.s.t. March 28, two vents immediately west of the ring of agglutinated spatter that had formed at the 1123 vent during episode 2 were glowing and slowly issuing short (approx 20 m long) pahoehoe flows. These same vents, within 30 m of the west base of the spatter ring, had been active during the minor eruption of March 21. In addition, the vent that had erupted within the spat- ter ring on March 21 was also glowing, but no new lava was seen then. The vents just west of the spatter ring, as well as a vent within the ring, erupted sporadically on March 28 and throughout the morning and afternoon of March 29. Dur- ing that time, they produced intermittent low fountains that gradually increased in vigor, so that by late after- noon of March 29, the fountains were about 10 to 20 m high when active (fig. 1.23). Only small flows, close to the vents, were produced. Also during this period of inter- mittent activity, the vents outside of the episode 2 spatter ring built a cone of spatter that coalesced with the west flank of the spatter ring to form a single irregular, grow- ing cone. The two western vents coalesced either late on March 30 or early on March 31. Thereafter, for the rest of episode 3, the ever-growing 1123 cone included two craters, each of which contained an active vent and pro- duced separate flows (pl. 1). Throughout episode 3, the northeastern fountain was the higher of the two (fig. 1.25). Much of the time, the southwestern fountain was low and was obscured from view by the growing cone of spatter and cinder. Thus, our data on fountain activity (fig. 1.23) record the activity only of the northeastern fountain. Except for a unique period on April 2—3, when both fountains apparently went through repeated simultaneous pauses in activity, we recognized no systematic relation, either sympathetic or reciprocal, in their relative vigor. Production of lava became steady at the 1123 vent at about 1800 H.s.t. March 29. Throughout that night until about 0500 H.s.t. March 30, the northeastern fountain FIGURE 1.63.—Simultaneously erupting 1123 vent (near) and O vent (dis- tant). Distance between two active vents is about 1.5 km. Line of low, inactive spatter cones in foreground was built during episode 2 at 0740 vent. Growing 1123 vent buried west two-thirds of this line of cones by end of episode 3. View westward; photograph by J .D. Griggs, taken at 1136 H.s.t. March 30, 1983. Numerals (lower right) indicate date and time. 56 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 was low and steady; its general height gradually increased from about 20 to 30 m, and it fed an aa flow that advanced steadily northeastward (pl. 1). At about 0500 H.s.t. March 30, the fountain increased abruptly in height, and, through March 31, the northeastern fountain was generally from 40 to 80 m high. On April 1, the northeastern fountain at the 1123 vent developed an episodic style of behavior in which its average height gradually rose and fell; inter- vals between the major maximums or minimums ranged from a few hours to more than half a day. Maximum crescendos in vigor of the northeastern fountain occurred on April 4, 5, and 6, when average fountain heights of 150 to 200 m persisted for hours at a time, and occasional peak heights were from 200 to 300 m. Subsequently, the vigor of the fountain gradually diminished, such that by the last day the average height measured during periods of maximum vigor was about 20 to 30 m. During mini- mums late on April 8, the fountain was so low as to be out of view at times; bad visibility also eliminated part of the record. Observers near the vent reported the end of episode 3 activity at 0257 H.s.t. April 9. Except for a 24-hour period on April 2 and 3, the north- eastern fountain played nearly continuously from the time that it became steady on March 29 through the end of episode 3. It became so low as to nearly disappear from view during the early afternoon of April 2 (fig. 1.23). Dur- ing the last day of episode 3, the fountain was so low that at times we could not see it over the flank of the large 1123 cone, although continuing steady harmonic tremor of high amplitude suggests that no break in lava produc- tion occurred. However, from about 2100 H.s.t. April 2 to 2100 H.s.t. April 3, both the northeastern and south- western vents were inactive simultaneously, or at least spatter disappeared from view, 26 times for periods that generally ranged from about 2 to 6 minutes, separated by intervals of normal activity ranging from 20 minutes to 3 hours in length. The northeastern fountain was most- ly 50 to 100 m high during that 24-hour period, and its disappearances and reappearances were abrupt. Although this behavior was not remarkable for the southwestern fountain, which was generally low and nor- mally visible to us only intermittently, it was unusual for the northeastern fountain. Harmonic tremor in the erup- tion zone, which was normally strong and steady, was erratic. Although a one-for-one correlation with the observed pauses in fountaining is not apparent, the tremor behavior was much like the observed fountain behavior. Tremor amplitude repeatedly diminished to very low levels for periods ranging from about 2 to 18 minutes in length. Both of these phenomena suggest episodic inter- ference with the supply of magma to the vent. In addition, the rate of summit deflation on April 3 was temporarily diminished. The average rate of summit deflation on that day, measured in an east-west direction by the Uweka- huna tiltmeter, was about 0.03 microradians per hour, about one-third the normal episode 3 deflation rate of 0.1 microradians per hour. This difference suggests that normal downrift transfer of magma along the conduit from the summit reservoir to the vent was partly impeded. Steady eruptive activity, steady and intense harmonic tremor, and normally rapid summit deflation resumed in the late evening of April 3. The highest fountaining of the northeastern vent ensued in the following days (fig. 1.23), with tephra falls that extended more than a kilometer from the vent. The sporadic activity of April 3 may have been related to an adjustment in the conduit between the summit reservoir and the 1123 vent; steady, rapid dis- charge from the southwestern vent began early on April 4, and the overall production of lava increased markedly. From the beginning of steady eruption of the north- eastern vent at approximately 1800 H.s.t. March 29, a thick, ponderously moving aa flow advanced steadily. Unlike the episode 2 situation, in which the fountain erupted through a lava pond that, in turn, overflowed to supply a well-channelized pahoehoe river, the northeast- ern vent, for much of episode 3, appeared to contain relatively little ponded lava. Instead, the flow may have been largely spatter fed. A short pahoehoe river or distributary system that converted to aa within a few hun- dred meters of the vent (fig. 1.42), appeared to issue from the base of the fountain. During the night of March 29—30, aa supplied by the northeastern vent almost completely surrounded the small 1965 spatter rampart that we had been using as a site for camp A (pl. 1). Over the next several days, the a con- tinued to thicken against the flanks of the 12-m-high hill (fig. 1.41). We last saw the hill late on April 3; subsequent- ly, it was completely buried by episode 3 lava. The flow from the northeastern vent advanced steadily northeastward through April 2 at an average rate of about 30 m/h (pl. 1; table 1.3). This northeasterly advance ap- parently slowed on April 3, and stopped on April 4 near- ly 4 km from the vent after emplacement of a slender lobe northeast of Kalalua. On April 3, some of the lava mov- ing northeastward was diverted to form a small lobe west and south of Kalalua. Advancing at about 10 to 15 m/h, this lobe was in place by late afternoon on April 5. The supply from the northeastern vent was wholly diverted on April 5 to a new lobe of aa that extended near- ly 4 km southeastward between the episode 1 and 2 flows. It advanced at an average rate of about 20 m/h for the rest of episode 3. Lava production at the southwestern vent was minimal before April 4. Early that morning, the vent became steadily active. At about 0725 H.s.t., the crater filled with lava, and a flow poured steadily over the south crater rim (fig. 1.64A). About an hour later, at approximately 0830 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY 0F ERUPTIVE EVENTS H.s.t., a segment of the southern crater wall began to col- lapse, and a narrow breach quickly formed (fig. 1.64B) as the detached sector began moving south. Time-lapse film records show that the collapsing section of the cone was rafted slowly southward during the next several hours. The major portion stopped within 150 m of the cone’s flank (fig. 1.25), but one 50-m-diameter block was transported 400 m (fig. 1.65). The southwestern vent promptly became the source of a well-channelized lava-river system feeding a sequence of flows that advanced southeastward (pl. 1) from April 4 to 9. The first flow extended southward and then southeastward during April 4 and 5; its front stopped about 3.4 km from the vent. Apparently, the flow was beheaded, probably late on April 5, by a second flow that followed the southwest edge of the first. A third flow then broke out near the vent on the east side of the previous two. It captured the lava river and followed the southwest edge of the episode 2 flow toward the Royal Gardens sub- division, which it reached in 2 days. A minor additional flow, pahoehoe instead of the normal aa, was emplaced along the west edge of the previous episode 3 flows early on April 7. It extended about a kilometer south of the vent; most likely it formed from a temporary overflow of the lava river near the vent. The two long flows extending southeastward from the southwestern vent advanced at average rates of 90 to 100 m/h. The last and longest flow reached the northwest end of King Street in the Royal Gardens subdivision at 0956 H.s.t. April 8; the flow front was then about 5.6 km from the vent. A detailed record of the flow-front advance southeastward along that paved street is shown in figure 1.66. A complexity arises because the sector of the flow front on the street stagnated temporarily on April 8, while lobes in the forest on either side continued to advance (pl. 1). The detailed record shows that the rate of flow- front advance before the end of the eruption (excluding the temporary stagnation, which resulted in a short-lived reentrant in the flow front) changed episodically over a measured range of about 40 to 360 m/h. On the pavement, surges with velocities of approximately 150 to 360 m/h were recorded. The data do not rigorously limit the dura- tions of these surges, but a range from about half an hour to more than an hour is suggested. Periods of slower advance, approximately 40 to 100 m/h, alternated with the surges. The highest recorded surge velocities, 183 to 360 m/h, occurred early on April 9, after the flow front had narrowed distinctly (pl. 1). The height of the flow front as it advanced along the paved street was determined from time to time by visual estimate by one of several observers. Most of the time, the flow front was 3 to 5 m high; however, as the sector on the pavement became the locus of most rapid advance after its late—afternoon stagnation on April 8, its front was 57 about 6 m high. During the period of highest velocity, at about 0100 H.s.t. April 9, the narrow advancing front was about 10 m high; subsequently, it decreased in height to approximately 3 to 5 m. The episode 3 lava flow advanced altogether about 2 km southeastward along King Street, and its front came to FIGURE 1.64.—Initial displacement of south flank of southwestern crater of 1123 vent during episode 3. A, 1123 cone shortly after beginning of steady lava production at southwestern vent. Low southwestern fountain plays within crater to right front of prominent northeastern fountain. Lava has filled crater at southwestern vent and is cascading over unbroken rim. View eastward from camp B; photograph by J .P. Lockwood, taken at 0753 H.s.t. April 4, 1983. Numerals (lower right) indicate date and time. B, 1123 vent, % hour later. Narrow, steep breach transects wall of southwestern crater along boundary between stable and failing sectors of crater rim. Segment of cone flank to right of breach is moving slowly southward. Shiny pahoehoe from overflow in figure 1.64A mantles cone on both sides of breach. Small pahoehoe flow advancing over episode 3 tephra is probably an overflow from lava river flowing through breach. Low cones, a few meters high, in right center, formed during episodes 1 or 2. View eastward; photo— graph by J .P. Lockwood, taken at 0839 April 4, 1983. Numerals (lower right) indicate date and time. 58 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 rest about 7.6 km from the vent after destroying six dwell- ings. As a consequence of forming at least one and, possibly, two lobes west of King Street, in addition to the primary lobe on and east of King Street, the flow attained a maximum width of 600 m within the subdivision. After eruptive activity stopped at the vent at 0257 H.s.t. April 9, the flow front advanced an additional 350 m over the next 14 hours (fig. 1.66). Progressively slowing, it ad- vanced another 20 m more over the next 3 to 4 days. 0 L In addition to the major flows from the northeastern and southwestern vents, intermittent slow emission from lava tubes occurred along the south and west base of the 1123 cone. This activity produced a thin apron of pahoehoe flanking the west base of the cone (pl. 1). Tephra, primar- ily from the high northeastern fountain, completely buried this pahoehoe apron. During episode 3, the 1123 vent produced a massive cone (fig. 1.67) about 60 m high, composed of agglutinated my“ 0.5 KILOMEYER l FIGURE 1.65.—Stereophotographs showing 1123 cone (dashed line) and rafted portions of cone (solid line). Southwestern crater (SW) is partly filled by talus from steep crater walls. Northeastern vent (NE), which has no distinct crater, is covered by its own aa. Rootless flows drape west flank of cone. Photographs 83.8.3JG120G#5 and 83.8.3JG120G#6 by J .D. Griggs, taken August 3, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS spatter, rootless flows, and Cinders. Figure 1.65 shows that no distinct crater remained at the northeastern vent and that the vent area was covered by aa. These char- acteristics seem compatible with the observation during this episode that relatively little lava ponded at the north- eastern vent. Apparently, the thick northeastern aa flow was supplied directly, or nearly so, by the high north- eastern fountain. The southwestern vent, which supplied a well-developed lava-river system, contrasts strikingly: A distinct bowl-shaped crater, which must have contained a lava pond, remained there. The thick aa flow from the northeastern vent covered an area of 3.5 x 106 m2; using the thickness data of plate 1 as a guide, we estimate a volume of about 19 x 106 m3. For the 249-hour period from 1800 H.s.t. March 29, when the northeastern vent became steadily active, to the end of episode 3 at approximately 0300 H.s.t. April 9, the average lava-discharge rate from the northeastern vent was about 75,000 m3/h. No significant change in that rate 59 occurred, even in response to the high-volume discharge from the southwest vent from April 4 to 9. Flows from the southwestern vent, also predominant- ly of thick aa, covered an area of about 2.9x 106 m2; we estimate their aggregate volume at about 15x106 m3. The southwestern vent was erupting steadily from about 0730 H.s.t. April 4 to 0300 H.s.t. April 9, a period of 115.5 hours. Thus, its average lava-discharge rate was about 130,000 m3/h, and the combined rate for the 1123 vent from April 4 to the end of episode 3 was about 200,000 m3/h. Including the initial eruptive activity at the O vent, episode 3 continued for approximately 290 hours. Its lavas, almost all aa, covered a total area of 7.9x106 m2 and had an aggregate volume of about 38x 106 m3. The average lava—discharge rate was 134,000 m3/h, but the actual rate ranged from approximately 80,000 m3/h in the early part of episode 3 to 200,000 m3/h in the later part. (57)/' - (77) ' - U) . m _ (75) /End of eruptlon w _ E / 2 7 — .— o (268) =I _ _ g (300)\f\(360I Z — _ _~ /183l P- r— 2 _ g (105) 2 //I63) _ g — ‘5‘} %8) — LL / Lu — /'/ _ g - / 146 <1: (70I// ‘ ’ - p— m — / _ - / o _ C//. / ' (11) ‘ 6 _ ./ (108) 146 - ' _ i )//‘\(48) " (267) _ I | I l l I I I I I I l I I I I I I I I I I I I I l I 1 I I I 1200 1800 0000 0600 1200 1800 2400 TIME, H.S.T. APRIL 8 AND 9, 1983 FIGURE 1.66.—Advance of episode 3 flow front in Royal Gardens sub division from time flow reached pavement at 0956 H.s.t. April 8, 1983, to 1703 H.s.t. April 9, approximately 14 hours after end of eruptive activity. Solid line records advance of flow front southeastward along King Street. Dashed line shows continuing advance of eastern lobe when sector of flow front on King Street temporarily stagnated. Numbers in parentheses are calculated flow-front velocities (in meters per hour) between pairs of data points. 60 THE PUU OO ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 EPISODE 4 (JUNE 13—17,1983) For 64 days after the end of episode 3, activity in the eruptive zone was characterized by weak harmonic tremor and emission of oxidized magmatic gases from fissures, fumaroles, and spatter cones. Slowly diminishing in- candescence persisted in a gaping crack high in the 1123 cone, and by the time episode 4 began, a horizontal distance-measurement line established on Puu Kahaualea on April 21 had shown about 4 cm of extension across the trend of the eruptive fissure just downrift of the 1123 vent. In the succeeding months, no further significant ex- tension across the fissure occurred at this locality. At 1025 H.s.t. June 13, we received a report of foun- taining in the eruptive area from observers in a fixed-wing aircraft. Harmonic—tremor amplitude recorded on a seismometer near Puu Kamoamoa had begun gradually to increase over low repose-period levels, beginning at about 0500 H.s.t. that morning; tremor amplitude peaked at 1100 H.s.t. Arriving at the site at 1145 H.s.t., we found a line of discontinuous fountains, 20 to 40 m high and about 100 to 150 m long, just downrift of Puu Kamoamoa. New vents at the west end of this line were observed open- ing at 1315 H.s.t. At that time, Puu 00, a steep-sided spatter cone at the east end of the active vents, approx- imately 750 m downrift of Puu Kamoamoa, stood 15 to 20 m high and marked the locus of major fountaining and lava discharge (fig. 1.29). A line of less active fissure vents with low spatter ramparts extended uprift 100 m or so FIGURE 1.67 .—60-m—high cone (Puu Halulu) built at 1123 vent during episode 3. Distance to cone is about 1 km. Rootless flows mantle cone’s near flank. Flat, smooth surface adjacent to base of cone is formed of an apron of tube-fed pahoehoe erupted from base of cone and then , mantled by tephra. Episode 3 tephra mantles hummocky pre-1983 lava in foreground. Small building in foreground was used for storage and shelter. View eastward from camp B; photograph taken at 1036 H.s.t. April 29, 1983. Numerals (lower right) indicate date and time. from the growing cone. Three pahoehoe flows, the longest of which had traveled about 500 In from the vent (pl. 2), were being fed by the line of erupting vents. The Puu Oo cone enclosed a crater that was partly filled with a lava pond. Fountains rose 30 to 50 m above the pond surface and were visible above the crater rim throughout the 100-hour-long eruption (fig 1.24). The major lava flow to the southeast (pl. 2) was fed by a lava river that exited the crater where the pond over- flowed through a low point in the crater rim. This breach, situated in the south-southeast wall of the crater, was 3 to 5 m wide and extended from one-half to one-third of the way down from the rim of Puu 00. The lava cascaded over a fall, coalesced at the base of the cone, and commonly produced a spectacular standing wave. By late afternoon on June 13, the lava river had estab- lished a well-developed channel that ranged from about 6 to 20 m in width and was bounded by smooth, stable levees of pahoehoe. Access across the levees to the edge of the lava river was good throughout the entire eruption, and estimates of the width and surface velocity of the river could be made in relative comfort. Lava flux was calculated (table 1.5) as described for the episode 2 chan- nel. On the basis of the appearance of the evacuated chan- nel after episode 4, we use an estimated depth of 3 m near the vent (fig. 1.9) and of 1.5 to 2.0 m 1 km downstream. In addition, observations of large rafted blocks during episode 4 suggest that the 1.5—2.0-m estimate is reasonable. At 1 km downstream, the observed lava flux, 90,000 m3/h, agrees with that determined from mapping (table 1.3). Near the vent, an estimated flux several times larger suggests that there the lava contained a higher pro- portion of entrained gas. In addition to the main flow, two small pahoehoe flows extended northward and southward from the fissure vents immediately uprift of Puu 00 (fig. 1.29). The more active of these two flows extended southward and southeast— ward from the entire line of low fountains. It began as a broad pahoehoe sheetflow with a maximum width of 150 m and was rapidly developing a central channel and lateral levees when observers first arrived. By late morn- ing on June 15, a lobe of this flow had turned eastward to impinge against the main southeastern flow from Puu 00 (pl. 2). A second, slow-moving pahoehoe flow pro- gressed 150 to 250 m northwestward of the eruptive fissure; this flow was fed primarily by low-level fountain- ing and emission of lava from the easternmost of the fissure vents. Over the course of episode 4, the flow developed into a nearly flat roofed, 250-m-diameter lava pond or reservoir. Such roofed ponds commonly formed where pahoehoe was slowly supplied by the vent; they ap- parently contained an interior complex of lava tubes that delivered lava in small streams to the margins or the roof. Pahoehoe buds at the edges of this particular roofed pond 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 61 TABLE 1.5.—Ept'sode 1, lava-river measurements [Width estimated visually during episode 4. Depth esti— mated in evacuated channel after episode 4] Distance Maximum Calculated its: Ngigh D:;)th velocity lava flux fm) (m/s) (103 m3/h) lO-ZO 6-8 3 15 570 30 15 3 3 240 130 20 3 3 320 1,000 7-9 1.5-2 3-4 90 supplied many of the measured temperatures and samples for episode 4 (see chap. 2). For much of episode 4, fountain activity along the fissure vents was characterized by repeated low bursts of fragmented spatter rather than by sustained fountains. Whereas the easternmost fissure vent grew steadily over the first 3 days of episode 4, activity at the other fissure vents decreased rapidly. By the morning of June 15, the cone growing at the eastern fissure vent was similar in size to the cone east of it at the adjacent main vent, and the two had coalesced to form a double cone at Puu 00. During that afternoon, all but two of the fissure vents were gradually buried by flows from the Puu 00 vent. Low bursts of spatter and low-level flow production con- tinued at the more westerly Puu 00 cone until late after- noon of June 15, when the cone became partly roofed and its lava production apparently stopped. The main episode 4 flow, aa except for the proximal 1 to 2 km, advanced at an average rate of 90 m/h (pl. 2). By 1600 H.s.t. June 15, the flow had traveled 5.7 km, and its front was west of the Royal Gardens subdivision op— posite the end of Ekaha Street. For the next 2 days, the flow advanced westward of the subdivision, entering it only locally and causing no property damage. Advance rates while the 2- to 12-m-high aa-flow front traveled downslope just west of Royal Gardens averaged about 1 m/min. As in episodes 2 and 3, however, large fluctuations in velocity and thickness were observed as intervals of surging alternated with periods of near- stagnation of the flow front. A particularly well observed large surge occurred at 1828 H.s.t. June 16. From 1500 until 1828 H.s.t., the front of the episode 4 aa flow at the west edge of the Royal Gardens subdivision was nearly stagnant; the flow itself was approximately 100 m across and was bounded by rubble levees, about 6 m high. Lava output at the vent, 7 km upstream, had been steady until 1700 H.s.t., when partial collapse of the cone caused a temporary but signifi- cant increase in the flux observed in the near-vent chan- nel. At 1828 H.s.t., a large pulse of lava was observed moving rapidly down the deep, nearly stagnant axial chan- nel of the flow just west of Pakalana Street. The surge consisted of a wedge-shaped body of brightly incandes- cent aa, capped by a layer of dark rubble (fig. 1.68). Its surface was 6 to 7 In higher than the tops of the marginal levees, which apparently confined it. Between Pakalana and Pikake Streets, the surge, still within the aa-flow channel, advanced a distance of 420 m in about 12 min- utes, at a rate of 33 m/min. The front of the surge was estimated at about 12 m thick and 70 m wide; over its total estimated length of 1 km, it tapered in thickness to about 1 m. These dimensions suggest a volume of about 400,000 m3 for the observed part of the surge—equivalent, if cor- rect, to about 4 hours of supply from the vent (table 1.3). Subsequent advance of the flow front was confined to a narrow lobe that advanced slowly down the steep slope west of the subdivision. A final small surge occurred at 1700 H.s.t. June 17, nearly 3 hours after the end of episode 4. Draining of the central channel, possibly as surges waned or as the flow front continued to advance at the end of episode 4, left high-standing aa levees west of the subdivision. Locally, the interior walls (fig. 1.37) were plastered by finely comminuted basalt striated parallel to the channel axis. Episode 4 ended suddenly at 1413 H.s.t. on June 17 , approximately 100 hours after it began, with little ap- parent premonitory decay of fountaining or lava output. At the episode’s end, the Puu Oo spatter cone stood about 20 m high. The eastern part of the cone contained a rubble-floored crater, approximately 40 to 50 min diam- eter at its crest (fig. 1.69), bounded by steep, smooth walls. FIGURE 1.68.—Surge west of Royal Gardens, traveling approximately 33 m/min down central channel of episode 4 aa flow at west end of Pakalana Street. Note stop-sign standard partly buried by stagnant aa at edge of episode 4 flow. Photograph by R.W. Decker, taken at 1828 H.s.t. June 16, 1983. THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 0 50 100 METERS I~J*J FIGURE 1.69.—Eastern (EV) and western (WV) craters within Puu Oo spatter cone (45), remaining fissure vents (FV) uprift of Puu 00, and remnant of O vent (3V) of episode 3 and its partial ring of spatter (3s) after episode 4. Vent deposits are delineated by solid lines, dashed where uncertain. Crater rims are delineated by hachured lines; hachures point inward. Active lava flows of episode 5 are visible heading northwestward and southeastward from Puu 00, and heavy fume issues from both eastern and western Puu Oo vents. Photograph 83.6.29JG120A#7 by J .D. Griggs, taken June 29, 1983, several hours after onset of episode 5. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 63 The southeast wall was still partly breached by a V-shaped notch, 3 to 4 m wide at its narrowest point and extend- ing 8 to 10 m down the side of the cone. The floor of the crater was bisected by a crudely linear opening of un- known depth. Fume emanated from this elongate hole, which was oriented subparallel to the axis of the rift zone. On the uprift shoulder of Puu 00, the western vent, approximately 15 m high, was also a steep-walled but part- ly roofed crater. Two small spatter cones marked the sites of part of the surviving fissure vents that had been active until the afternoon of June 15. Extending southeastward from the eastern Puu Oo vent was an evacuated lava channel, box shaped in cross section and approximately 2 to 4 m deep and 10 m wide (fig. 1.9). The floor of this channel was generally smooth; in places, large sections of the wall had collapsed inward. About 1 km downstream, the average depth decreased to 1 to 2 m, and the width increased to about 15 to 20 m. Because the rate of lava discharge from the fissure vents and western Puu Oo vent was relatively low, those vents did not feed vigorous lava rivers, and so no conspicuous evacuated channels remained in the flows that had traveled northward and southward from them. Episode 4 produced approximately 11 x 106 m3 of basalt that covered an area of 2.2x 106 m2; the average lava- discharge rate was 110,000 m3/h. Most of the main, 7.8-km-long southeastern flow was aa, 2.5 to 10.7 m thick and 40 to 400 m wide. A well-defined aa channel persisted for almost the entire aa portion of the flow. The minor flow north of the fissure vents was primarily pahoehoe, several meters thick. The southern fissure-vent flow ultimately extended 2.4 km from the vent and consisted of both aa and pahoehoe, averaging 3.0 m in thickness. EPISODE 5 (JUNE 29-JULY 3, 1983) Observers camped near the site of Puu 00 on the night of June 28—29 noticed a strong glow over the vent. At 1000 H.s.t. June 29, as we remeasured a newly establish- ed horizontal distance-measurement line across the erup- tive fissure just east of Puu 00,1 helicopter pilot Bill Lacy, Jr., reported seeing lava ponding inside the crater. Climb- ing to the southeast rim of the main crater, we saw a thin- crusted pond of lava, like that shown in figure 1.5OB, slowly rising inside the crater. A 1- to 2—m-high dome foun- tain played intermittently in the center of the pond. Fresh high-lava marks visible on the inner crater walls indicated 1Only two sets of measurements were ever made on this short-lived line, which did not sur- vive episode 5. The first was on June 24, and the second just before the onset of episode 5 on the morning of June 29. Approximately 2 cm of extension occurred during that time inter- val between targets at the ends of a 400-m—long line centered over the eruptive fissure. However, it was impossible to determine whether this deformation was due to lateral extension or to local uplift near the eruptive fissure where the measuring instrument was located. that the pond surface had been about 2 m higher at some recent time. At 1214 H.s.t., lava began to ooze from a small hole about 1 to 2 m below the spillway crest. The narrow lava tongue froze, however, before reaching the channel floor, as the pond inside the crater subsided temporarily. The final breakthrough occurred at 1245 H.s.t., when a 1-m- wide pahoehoe cascade appeared at the same opening and reached the evacuated episode 4 channel floor within 1 to 2 minutes. By 1254 H.s.t., a fluid pahoehoe flow, 30 cm thick at the front and thickening upstream, was spilling through the breach in the crater rim and reoccupying the empty episode 4 channel. Advance rates were slow enough for us to sample the first lava down the channel. An in- crease in harmonic tremor in the eruptive area began at approximately 1251 H.s.t., followed by an acceleration of lava discharge and development of a low fountain that rose 5 to 10 m from the pond surface. Activity stabilized by 1300 H.s.t., and the lava channel near the vent was full and often overflowing. A fountain, 10 to 35 m high, with rare bursts to 50 m, played from the surface of the lava pond within the crater. The pond surface was estimated from aerial views to be at the level of the spillway; the pond was probably 20 to 30 m deep. By evening, the fountain had increased in height to about 40 m, and it remained approximately 20 to 40 m high throughout the 90—hour eruption (fig. 1.24). By 1558 H.s.t. June 29, the western vent, high on the uprift flank of Puu 00 (fig. 1.69), became active with low fountaining through a lava pond and production of small flows. Examination of this vent at 1430 H.s.t. had shown no sign of lava; however, very hot air and fume pulsing from an opening at the top of this partly crusted over spatter cone had been detected at 1055 H.s.t. At the uprift base of the western Puu Oo vent, approximately at the site of one of the episode 4 fissure vents, a small pahoehoe flow issued to the south. Whether this was a separate vent or merely a passive outlet for the western Puu Oo vent is uncertain. By the morning of June 30, the small lava and spatter mound that grew around this minor outlet was buried by pyroclastic deposits from the more active western Puu Oo vent. A small pahoehoe flow from the western Puu Oo vent ponded north of the vent during the late afternoon and evening of June 29. In the early morning hours of June 30, it developed a throughgoing channel feeding a local pahoehoe flow that extended northeastward, overrunning camp B (pl. 2). The western Puu Oo vent stopped feeding the north- eastern flow during the afternoon of June 30, after it had traveled 800 m, and began, instead, to feed a well- channelized flow that extended southeastward along the southwest edge of the episode 3 and 4 flows (western flow, pl. 2). This flow advanced at an average rate of about 80 64 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 m/h and took about a third of the total lava discharge of the active vents. By the early evening of July 1, the flow had turned eastward and impinged against the main southeastern flow from the Puu Oo vent. At that time, relatively fluid paheoehoe extended along the channel for approximately three-fourths of the length of the flow. Subsequently, the flow turned southeastward again and continued downslope through the rain forest. The main flow (eastern flow, pl. 2) followed a path toward Royal Gardens just east of the episode 4 flow. It developed a long, stable pahoehoe channel, with a startling hairpin bend 1.5 km from the vent (fig. 1.70). The vigor of the channel seemed nearly identical to that observed during episode 4. The channel was bank full, and no long-lived changes in velocity or flux were observed during the eruption. About 100 m downstream from Puu 00, we estimated a channel width of 7 m and an average surface velocity of 1.25 to 1.5 m/s. For an assumed depth of 3 m, these values suggest a flux of 100,000 m3/h, iden- tical to that determined for the eastern flow later when it was mapped (table 1.3). The transition from pahoehoe to aa in the channel of the eastern flow occurred downstream from the vent at distances of approximately 3.5, 3.5, and 4.2 km for 1325 H.s.t. June 30, 1100 H.s.t. July 1, and 0910 H.s.t. July 2, respectively, translating to 88, 63, and 58 percent of the total flow length at each successive time. For the first 4.5 km, the toe advanced at an average rate of about 170 m/h; it then slowed down to an average rate of 60 m/h. The' higher early rate probably reflects rapid extension of the flow during the first day, when pahoehoe extended near- ly to the toe. The eastern flow approached the subdivision as a 300-111- wide, flat-topped aa front, 3 to 4 m high and moving at about 1 m/min. Just northwest of the subdivision, the flow became confined to a gully, and a narrower lobe, 100 to 200 m Wide, accelerated toward the upper end of the sub- division; it reached Ekaha Street at 1950 H.s.t. (fig. 1.71). After traveling approximately 170 m at an average velo- city of 3.9 m/min, the flow began an extended period of very sluggish advance. During the night of July 1—2, its average forward velocity was 0.5 m/min, and at 0500 H.s.t. July 2 the flow front was essentially stagnant. Movement was sufficiently slight to allow an observer to walk across the entire flow, which was 120 m wide, 2 to 3 m thick, and had a central channel 15 m wide. The lull in activity at the flow front ended at 0620 H.s.t. as a surge reactivated the front, which then advanced across Tuberose Street (fig. 1.71) at a rate of 5.0 m/min. The sudden advance related to this surge, like many others observed in Royal Gardens, lasted less than an hour. Between 0730 and 1500 H.s.t. July 2, at least five surges with durations of approximately 25 to 55 minutes were reported. The most impressive surge occurred at 1453 H.s.t. July 2, following a period of slow advance between Pikake and Plumeria Streets. This surge was first observed 300 m upstream of the front and was ad- vancing at a rate of 30 m/min. Viewed from the front, it appeared to be a wall of aa, about 6 m above the level of the 4-m-high, 100-m-wide preexisting flow (fig. 1.39A). Responding to the approaching surge, the front doubled in thickness, steepened, and became unstable, and its for- ward motion accelerated (fig. 1.393). An abandoned truck was overrun in minutes; before burying it, however, the flow pushed it forward as a bulldozer would for a distance of 40 to 50 m. This thickening and acceleration was followed by breakout of a fluid, thin (1—2 m thick) flow of scoriaceous aa (Lipman and Banks, 1987) from near the base of the main flow (fig. 1.390). This thin lobe moved down Hoku Street at a rate of 15 m/min, destroying another abandoned car and decelerating to signal the end of the surge by 1525 H.s.t. The main body of the flow had thinned to 2 to 3 m and continued to advance at less than 1.0 m/min. One home of the total of eight engulfed in Royal Gardens during episode 5 was destroyed during this surge. The long interval of slow advance that followed the 1453 surge was interrupted at 1821 H.s.t. by another large surge observed passing Pakalana Street. This surge, witnessed primarily from helicopter at night, was a broad sheet of fluid, brightly incandescent lava, about 50 m across and 4 to 5 m high at the front, thinning upstream over a distance of about 800 m (fig. 1.38). As the surge reached the front of the primary flow, a breakout over the lateral levee occurred, sending a short-lived, fluid, thin flow downslope at a rate of 25 m/min. This overflow decelerated rapidly and thickened from 1 to 8 m in 5 minutes. At least three additional surges of smaller magnitude separated by 3- to 7-hour intervals were reported over the next 13 hours in the subdivision. By 0618 H.s.t. July 3, one of these surges had formed a new lobe that traveled just eastward of Hoku Street (fig. 1.71). Fountaining and local harmonic tremor dropped abrupt- ly and concurrently between 0713 and 0720 H.s.t. July 3, approximately 90 hours after episode 5 lava production began. Perceptible movement of the flow toe continued for more than 3 additional hours. A small surge occurred at 0930 H.s.t., and the front continued to creep forward slowly until at least 1030 H.s.t. The vent complex, much like its configuration at the beginning of episode 5 (fig. 1.69), consisted of two juxta- posed spatter cones separated by a wall of spatter. Both cones were steep sided and had bowl-shaped craters. Evacuated, box-shaped lava channels, 3 to 4.5 m deep, extended from the southeast base of the eastern Puu Oo cone and the southwest base of the western cone (see fig. 1.74). The rims of these cones had approximately doubled in height to 40 m during episode 5, and each cone was partly transected by a steep cleft through which the 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS FIGURE 1.70.—Upper reaches of sinuous episode 5 lava river. Erupting eastern vent at Puu 00 is visible at top, 1.5 km from hairpin bend. Western vent and its flow to southeast are also visible. Apparent gap in river, where smoke originates at upper right, is a short roofed segment that developed where river narrowed and accelerated as it flowed over a southeast-facing escarpment. View northwestward; photograph by J .D. Griggs, taken at approximately 1800 H.s.t. July 1, 1983. THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 Eruption stopped (31’s '14 (49) A” (29) ‘12 11 S DISTANCE FROM VENT, IN KILOMETERS 8 — _ ( _ S's/1 _ _ (93) _ — (26) . S - 7. s 8 .. S _ (159) _ s _ 7 — _ 35 . 5‘.) s _ s _ (441) _ ' 4 _ _ (29) _ _ 3. _ (236) ' — o 2 —- 6 — /(308) I1 I I ( I l I I I I I | I I I I I l J I ( I 1 800 0000 0600 1 200 1 800 0000 0600 2400 2400 TIME, H.S.T. JULY 1-3, 1983 EXPLANATION 4 Posiflon of flow (tom— Numbers refer to accompanying table and part A I House covered 0 0.5 1 KILOMETER 1%.I__l 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 67 respective lava rivers had exited. During heavy bombard- ment by spatter late in episode 5, outward slumping of the northeastern flank of the eastern cone created a low area in the rim. Although we did not realize it then, this low part of the rim was to become the persistent spillway of subsequent episodes. At that time, however, the breach on the southeast side was considered the most likely spot for the next spillover if eruptive activity resumed. Approximately 13 x 106 m3 of basalt was erupted dur- ing episode 5; it covered an area of 3.4x 106 m2, at an average lava-discharge rate of 140,000 m3/h. The main flow destroyed eight dwellings in the Royal Gardens sub- division, and four other dwellings were cut off from road access. FIGURE 1.71.—Summary of episode 5 flow advance in Royal Gardens. A, Flow-front advance versus time. Numbers 1 through 14 correspond to successive flow-front positions keyed in figure 1.7lB. Numbers in parentheses are average velocities (in meters per hour). S, surge reported by flow-front observers. Data points do not necessarily cor- respond to beginning or end of surges. Thus, plot shows episodicity of flow advance but not necessarily maximum and minimum flow rates. B, Part of Royal Gardens subdivision, showing flows of episodes 2 through 5 and houses destroyed by episode 5 lava flow. Table below refers to numbered flow-front positions shown in figure (dashes in- dicate no data). Average velocity was computed from distance traveled (as measured from preceding flow-front position along line on map connecting successive flow-front positions) divided by elapsed time. Observed velocity was estimated by flow-front observers at indicated time. Observed velocity Average Thickness velocity Loo al- Tme Comments 1ty (H.s.t.) (m) (tn/min) (m/min) July 1 1 1919 '1—5 0.8 5.0 One lobe out of several had become dominant; flow accelerated. 2 1950 —-— 5.1 ~— Reaches Ekaha. 3 2033 2.5—" 3.9 ' —— Very sluggish (<(1 m/min) most of the night. July 2 M 0620 2.5 0.5 5.0 Surge crosses Tuberose after long stagnant interval; entire flow front 120 m wide. 5 0716 -- 7.11 —— Surge and then decelera‘ tion. 6 08211 -—- 6 -— Surge followed by slow advance. 7 1121 1—2 2.7 8.0 Surge crosses Pikake, then flow thickens to 2—3 In. 8 1315 —— .14 7 Surge after long interval of sluggish advance. 9 1720 3 1.6 .3 Crosses Pluneria. 10 2009 5 6 1.0 —-— 11 2305 8-10 9 5 Flow thickens as it slows. July 3 12 01133 — 0.5 0.1 Crosses Paradise, 20-30 m wide. 13 0635 -— .8 .9 Eruption stops at 0715 H.s.t. 1‘1 1030 -—- .5 -— Flow essentially halted. EPISODE 6 (IULY 22-25, 1983) At 0700 H.s.t. July 20, astronomers on Mauna Kea reported seeing intermittent fountains in the middle east rift zone of Kilauea. Their reports went unsubstantiated and were followed at 0900 H.s.t. by an aircraft observa- tion of fume but no lava at Puu 00. At approximately 0600 H.s.t. July 21, lava was sighted from a passing aircraft inside the eastern crater at Puu 00. At 0900 H.s.t., an HVO observer reported a 5- to 10-m-deep lava pond in the eastern vent. For the next 33.5 hours, low—level erup— tive activity occurred within the crater, accompanied by a fluctuating increase in harmonic tremor and episodic gas-piston activity (fig. 1.50). At approximately 1410 H.s.t. July 22, an extended fill- ing event (fig. 1.50A) progressed so far that overflow seemed imminent. Observers, who had been watching the gas-piston activity from a low area on the northeastern part of the crater rim, left their vantage point, certain that lava would begin its cascade through the breach in the southeast rim of the crater. When they reached a new observation point high on the southeast side of the cone at 1511 H.s.t., however, they found that the pond was completely drained. From this new vantage point, they saw that their former perch on the northeast rim was covered with new cooling lava from the previous high stand of the pond. Lava reappeared on the crater floor after about a minute, and at 1515 H.s.t. a steady, more rapid filling of the pond commenced (fig. 1.50A), accom- panied by a fountain of accelerating vigor which soon indicated that a second retreat would be advisable. At ap- proximately 1530 H.s.t., the 20-m—deep pond began to overflow, first on the northeast side and then, briefly, through the breach in the southeast rim. Fountaining and flow production rapidly increased. Soon lava pouring over the low sector of the northeast rim was supplying a flow to the northeast, and spatter cascading over the north rim, with possibly some pond overflow as well, was feeding a broad, thick, slow-moving, spiny pahoehoe flow that advanced northward and northwestward from Puu 00 (fig. 1.72). A short-lived pahoehoe flow reoccupied the evacuated episode 5 channel to the southeast for 200 m, but supply from the southeastern spillway was soon terminated. The flow to the north and northwest became stagnant after several hours, and, thereafter, all the lava discharged from the crater overflowed the northeast rim and fed the northeastern flow. During episode 6, the early fountain, which rapidly reached 60 m above the surface of the lava pond (fig. 1.24), was strikingly more energetic and sustained than those of episodes 4 and 5. At times, the fountain was directed to the northeast, and heavy spatter fell directly into the pahoehoe river exiting the crater. By 1645 H.s.t., when last seen that day, the flow front had traveled less than 68 THE PUU 00 ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 500 m and appeared to be headed into remote rain forest, where it posed no immediate threat to inhabited areas. The eruption continued steadily throughout the night of July 22—23. At 0300 H.s.t., a strong glow in the direc- tion of Puu 00 was visible from HVO, and the roar of the vent, nearly 20 km away, was audible. There were reports that the fountain was visible from the airport in Hilo, near— ly 40 km north of the vent. Fountain height increased through the first night, reaching a peak of more than 100 m in the early morning hours of July 23 (fig. 1.24). High, steady fountaining con- tinued until about 0830 H.s.t. July 23. The fountain then declined in height sporadically during the rest of the day to a general height of 30 to 60 m that was maintained, with a gradual long-term decline, throughout the rest of the episode. When it was high, the single, vigorous foun- tain was broad and was commonly inclined to the north- east as much as 10°—35° from vertical. The high fountain produced a tephra blanket that fell to the southwest (downwind), mantling the flank of the cone and partly fill- ing the remnant southwestern episode 5 channel adjacent to the southwest flank of Puu 00. During the period of high fountaining, the northeastern flow, mainly aa, continued to be sluggish and failed to develop a throughgoing centralized channel system, even though the lava flux over the northeast crater rim was high. Visual estimates of the flux ranged from 1 to 3 times that seen spilling from Puu 00 during episodes 4 and 5. During the early morning of July 23, the flow was broad, and lobes extended short distances eastward, northeast- ward, and southeastward in the area between Puu 00 and FIGURE 1.72.-—Vigorous fountaining and flow production from Puu 00 1 hour after beginning of episode 6. Spiny pahoehoe in foreground is fed primarily by voluminous spatter cascading down north flank of cone. To left of fountain, lava overflowing low northeast rim sup- plies a flow to northeast. Fountain is approximately 60 m high. View southward; photograph taken at 1637 H.s.t. July 22, 1983. the 1123 cone. At about 1030 H.s.t., after the fountain height had greatly diminished, more rapidly advancing pahoehoe was seen overriding the sluggish aa. A stable channel system developed, and a lobe of this flow con- tinued northeastward to become the major flow of episode 6 (pl. 2). The spillway area on the northeast rim of Puu 00 was a broad zone where lava poured continually from the lava pond. Output consisted of two lava rivers separated by a large pinnacle on the rim (fig. 1.73); they coalesced after descending the steep outer slope of the cone. By the last day of episode 6, the smaller, more northerly spillway had been dammed, and the vent geometry was like that shown in figure 1.20. Supply of lava to the channelized river remained steady, and the flow advanced through the rain forest north of Puu Kahaualea at an average rate of about 80 m/h. Dur- ing the night of July 24—25, the flow front divided into two major lobes that advanced in parallel (pl. 2). At the episode’s end, they were each nearly 6 km from the vent; subsequently, they merged and flowed another half- kilometer. The eruption stopped at about 1630 H.s.t. July 25. Tremor began gradually to decrease at about 1620 H.s.t. and by 1630 H.s.t. had dropped significantly, coincident with shutdown at the vent. In the last few minutes of the eruption, fountain activity became noticeably lower and more intermittent, with loud, pulsing bursts of frag- mented spatter. Episode 6 lava was predominantly aa at the end of the episode. The two-lobed major northeastern flow had ex- FIGURE 1.7 3.—Two spillways and lava cascade feeding flow to northeast during episode 6. Fountain height is approximately 40 m above lava- pond surface. By next day, smaller, more northerly spillway was inactive. View southwestward; photograph taken at 1641 H.s.t. July 24, 1983. Numerals (lower right) indicate date and time. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 69 tended a total of 6.5 km into heavy rain forest, establish- ing a pathway that was to be followed in many subsequent episodes. Remnants of a well-defined central channel were discernible along nearly the entire length of the flow. Pahoehoe overflows as far as 2.75 km from the vent in- dicated that the fluid pahoehoe portion of the channel had extended about 40 percent of the overall flow length. Additional, spatter-fed aa flows with evacuated central channels extended 0.55 km southward and 1.25 km southeastward (pl. 2). In total, episode 6 lava covered an area of about 2.0 x 106 m2 and had a volume of 9 x 106 m3. The episode lasted 73 hours, and thus the calculated lava-discharge rate was about 120,000 m3/h. The Puu Oo cinder and spatter cone had increased its girth and height strikingly during episode 6. The south- west flank was mantled by a thin blanket of tephra (fig. 1.74) from the high fountaining early in the episode. Puu 00 contained a steep-walled, bowl-shaped crater (fig. 1.75A), about 50 m in diameter at its rim crest and 30 to 40 m in diameter at the level of the spillway. An interior bench or terrace, about 10 to 15 m above the floor, skirted the north, east, and south walls. The west wall was steep, unstable, and approximately 25 m high from the floor to the rimtop. At the base of this wall was a fum- ing, incandescent crack and a 5- by 20-m elongate block, about 8 m high, that was mantled by spatter on its top and slickensided on its sides (fig. 1.753). The vertical striations imply that the block was left standing high as lava and rubble of the crater floor subsided at the end of episode 6. The western Puu Oo vent (fig. 1.7 4), active in episode 5, remained quiet during episode 6; it received only occa- sional spatter-fed flows from the episode 6 fountain. It retained definition after episode 6 but was only one-fourth as large as the main part of Puu 00. The Puu Oo spillway was a broad, low area on the north- east crater rim. About 15 m wide (fig. 1.75A), it was smooth and wide enough to use as a landing area for small helicopters. EPISODE 7 (AUGUST 15-17, 1983) On August 8, a pilot reported lava “deep” in the vent. On August 10, observers on the rim and inside Puu Oo Crater saw a small glowing hole at the west edge of the crater floor, where an incandescent crack had been noted on July 26 just after episode 6. A small pahoehoe flow had issued from the vicinity of this hole, and a low collar of spatter surrounded it. By the afternoon of August 11, the floor of the crater was completely repaved with thin new pahoehoe (fig. 1.75A). Sporadic production of small lava flows, accompanied by low-level spattering and occasional cyclic rise and fall of lava visible Within the hole, continued for the next 4 days. Throughout this period, activity was mild enough to permit direct sampling of gases and small lava flows on the crater floor. Observers were absent for the beginning of vigorous lava production on the morning of August 15. An increase in harmonic tremor was recorded at 0709 Est. by a seismometer near Puu 00, and time-lapse film recorded fountains rising above the crater rim by 0741 H.s.t. At 0850 H.s.t., when we arrived at the site, a fountain 60 m high emanated from Puu Oo Crater, which was filled by a lava pond about 20 m deep. The low northeastern part of the rim was again the spillway, and lava from the overflowing pond (fig. 1.20) fed a rapidly advancing pahoehoe flow that followed the path of the episode 6 lava to the northeast. The remnant of the western crater at Puu 00, which had last been active in episode 5, accumulated falling spat- ter so rapidly that it contained an overflowing pond. Thus, a moderately voluminous pond- and spatter—fed flow advanced southwestward from the vicinity of the non- erupting western crater, and a smaller spatter-fed flow traveled northwestward from the north side of Puu 00 (fig. 1.20). The northwestern flow advanced slowly and eventually turned northeastward, traveling a total of 600 m (pl. 2). The southwestern flow received a signifi- cant but varying component of overflow from the second- ary pond and traveled more rapidly, at rates of about 50 to 150 m/h. By morning on August 17, after turning south- eastward and traveling approximately 3 km from the vent, this flow had been beheaded, and it slowly halted. During the heavy spatter fall that produced the rootless northwestern flow, the north rim of Puu 00 became deeply furrowed (fig. 1.76). As we watched, individual pinnacles seemed to grow upward and simultaneously become better defined by development of the bounding furrows, as if the falling fragments were accreting to the pinnacles or eroding the furrows, or both. Eventually the pinnacles became unstable and broke off, falling into the crater or rolling down the outside of the cone. The main, northeastern flow received the major part of the lava output (table 1.3). As in previous episodes, this flow was fed by a vigorous pahoehoe river with narrow overflow levees for approximately the first kilometer. A broad, unimpeded cascade of lava down the spillway on the northeastern flank of Puu Oo Crater supplied this river with overflow from the lava pond. After traveling approximately 1 km over episode 6 basalt during the first hour of the eruption, the flow widened, became aa at its front, and decelerated. It then continued to advance into rain forest south of the episode 6 basalt and, in places, over 1963 basalt, at an average rate of about 100 m/h. The advance of the northeastern flow was complicated by repeated division of the flow front to form subordinate lobes (pl. 2). By the morning of August 16, the front had THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 Ml; WM fum- 100 200 METERS FIGURE 1.74.—Puu 00 and nearby episode 6 flows (6) after episode 6. Evacuated episode 6 channel is conspicuous northeast (to right) of Puu 00. Though fuming heavily, western Puu Oo Crater remained inactive during episode 6. Photograph 83.8.3JG120F#7 by J .D. Griggs, taken August 8, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS divided to form the southern and middle lobes, which traveled side by side, although the southern lobe received more lava supply and advanced more rapidly. Subsequent- ly, each lobe divided again. A slow-moving (approx 30 m/h) northern lobe separated from the middle lobe, and a more rapidly moving (approx 120 m/h) northeastern lobe separ- ated from the southern lobe. Thus, at the episode’s end, the complex northeastern flow was advancing along each of four distinct fronts. FIGURE 1.75.—Interior of Puu 00 between episodes 6 and 7. A, New pahoehoe flows, erupted after episode 6, cover crater floor. Glowing hole (arrow) marks vent at base of steep west wall. Spillway, mantled by shiny episode 6 pahoehoe, is smooth surface near bottom edge of photograph. Large block on crater floor, surrounded by new flows, is 20 m Wide. Smaller western crater, which did not erupt after episode 5, is partly visible at upper right. View southwestward; photograph taken at 1429 H.s.t. August 11, 1983. Numerals (lower right) indicate date and time. B, Large slickensided block on floor of Puu Oo Crater. Person is standing on pahoehoe erupted within Puu 00 between episodes 6 and 7. View southwestward from Puu Oo spillway; photograph taken at 0917 H.s.t. August 14, 1983. Numerals (lower right) indicate date and time. 71 Fountain height varied little during episode 7 (see fig. 1.24); it ranged from about 40 to 80 m over the first day and, thereafter, diminished slowly but steadily to about 30 m near the episode’s end. As in episode 6, fountaining commonly was broadly distributed over the lava pond in the style illustrated in figure 1.26. The fountaining was characterized by irregular jets, many distinctly inclined from vertical. Lava output and fountaining stopped abruptly at approximately 1600 H.s.t. August 17. The episode had produced 14 x 106 m3 of basalt, which covered an area of 3.7 x 106 m2. A duration of 57 hours indicates an average lava-discharge rate of about 250,000 m3lh, then the highest average rate determined for any episode. At the end of episode 7, Puu 00 was a steep-walled cone, almost circular in plan view, containing a bowl-shaped crater, approximately 90 m in diameter at the top (fig. 1.77). The interior walls, partly of rubble, extended downward into a deep, steep-walled hole with little, if any, crater floor. An incandescent crack persisted low in the west wall. A low spillway area again transected the north- east rim of the crater; it led, by way of a steeply inclined notch, down the outside of the cone to an evacuated pahoehoe channel, 8 to 15 m wide. At the base of the cone, where this channel began, a 10- to 15-m-long segment was roofed to form a short tube. The channel retained sharp definition for 600 m downstream, where it evolved into an aa channel that extended nearly to the ends of the FIGURE 1.76.—Furrowing of north rim of Puu 00 by falling spatter dur- ing episode 7. Fountain height is approximately 70 m(f1g. 1.24). View southeastward from near camp C; photograph taken at 0541 H.s.t. August 16, 1983. Numerals (lower right) indicate date and time. THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 0 200 METERS L—___%__I FIGURE 1.77 .—Puu 00 and nearby episode 7 lava flows (7). Photograph 83.9.2JG120A#1 by J .D. Griggs, taken September 2, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS flows. Pahoehoe overflow levees and islands persisted downflow to beyond Puu Kahaualea, 2.7 km from the vent. For the first 2 km, the flow ranged in width from about 100 to 500 m; farther northeast, the maximum width was about 700 m, but individual flow lobes were narrower, generally 150 to 300 m wide. The longest of the four lobes ultimately extended 6.7 km from Puu 00. The western crater at Puu 00, last an active vent dur- ing episode 5, was partly obliterated during episode 7; however, it was still recognizable on the uprift shoulder of Puu 00. The southeastern aa flow, with a broad, poor- ly defined central channel, extended southward and then southeastward 3.3 km from the vicinity of the western crater. EPISODE 8 (SEPTEMBER 6-7, 1983) We became aware of renewed low-level eruptive activity within Puu Oo Crater on September 2. Spatter and small flows from two vents had filled in the deep central depres- sion, forming a crater floor about 12 m below the level of the spillway. One vent had produced a 4-m-high and 6- to 7-m-wide cone of spatter (fig. 1.78) near the center of the smooth and nearly flat new crater floor. A 4—m- diameter opening in the top of the cone was incandescent near the surface and occasionally emitted small spatter fragments to heights of a few meters. Near the west edge of the crater floor, at the site of the incandescent crack, a second vent had produced several short (less than 25 m long) pahoehoe flows and a small amount of spatter. In the process, it had built a rounded lava mound with a tran- sient central hole that was an excellent site for sampling both magmatic gas and melt. This mound was incandes- cent just below a superficial accumulation of agglutinated spatter and small flows. Episodic, vigorous degassing and minor spatter ejection may have reflected the occurrence of gas-piston activity beneath the vent. The vigorously exiting hot gases bulged the cap of agglutinated spatter upward, and small flows were sometimes extruded. In addition, local gas jets occasionally mobilized melt from within the spatter mound to produce small, steep-sided driblet cones. By September 4, continued intermittent low-level erup— tion had filled in parts of the crater floor with approx— imately 300 m3 of small pahoehoe flows. The vent at the west end of the crater floor had grown 3 to 4 m through the accumulation of short flows and spatter. The more central spatter cone continued to emit fragments of spat- ter (max 30 cm diam) to heights of several meters at a rate of a few fragments per minute. On September 5, we discovered an active lava pond, ap- proximately 5 m deep, extending across the entire crater floor. The pond was almost completely covered with a thin crust; in places, small pahoehoe lobes with still- 73 incandescent cracks had flowed over the crust. A high- lava mark on the crater walls and spillway indicated that sometime before 0850 H.s.t. September 5, the pond had reached a depth of about 12 m. No lava, however, had reached above the level of the spillway and exited the crater. A 15-m-long, 1— to 3-m—wide lava stream, estimated to be carrying about 1,000 to 3,000 m3/h, issued from the nearly flooded western vent and traveled across the crusted surface of the lava pond. It disappeared down a hole near the center of the pond at the approximate site of the central intracrater vent of September 2-4. Occa— sionally, the adjacent crusted-pond surface would be broken open or overrun by pahoehoe oozes fed by this lava stream. No net change in the volume of the lava pond or in the vigor of the lava stream was seen during 6 hours of continuous observation. Spattering at the western vent, where the stream originated, was minimal, implying that the lava was relatively degassed. Apparently, deeply stored, gas-rich melt was not participating in this lava cir— culation, which was apparently confined to the upper part of the conduit and storage system beneath Puu 00. At 0511 H.s.t. September 6, the time-lapse camera at camp D recorded the onset of vigorous eruptive activity as fountaining became visible above the rim of Puu 00. By 0730 H.s.t., when observers reached the eruption site, a vigorous fountain was rising approximately 100 m from within Puu 00. It was broadly based, fluctuated little in height, and filled the entire crater. Time-lapse data show a slow but continual increase in fountain height through- out the 24-hour eruption (see fig. 1.24). FIGURE 1.78.—Interior of Puu 00 during low-level activity preceding episode 8. Spatter cone at right is approximately 4 m high. Lava mound at left (circled person for scale) marks a vent that is actively degassing, intermittently producing spatter and small pahoehoe flows, and building little driblet cones. Spillway (out of View to right) is approx- imately 12 m above level of crater floor. View northward; photograph taken at 1920 H.s.t. September 2, 1983. 74 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 The northeastern spillway was once again the locus of overflow from the pond inside Puu 00. The cascade from the spillway supplied a well-channelized river that carried the lava northeastward on top of earlier 1983 flows. When we arrived at 0730 H.s.t. September 6, the flow, entirely of pahoehoe, had reached 1,200 m from the vent, a distance indicating an average advance of more than 500 m/h. This value is a minimum; instead of continuing to advance, the lava had ponded, forming a broad, inflating, roofed pahoehoe pond or reservoir about 1 km northeast of Puu 00. Minor flows broke out to the north from the ponded lava, and by early afternoon on September 6, a more vigorous pahoehoe front had developed from the northeast end of the roofed pahoehoe pond. It traveled north-northeastward into thick rain forest and, for the rest of episode 8, followed the north edge of episode 6 and 7 flows (pl. 3). By about 1600 H.s.t. September 6, the ponded part of the flow was collapsing, apparently being drained as the fluid flow front advanced rapidly north- westward of Puu Kahaualea. The presence of standing trees surrounded by lava as much as 500 m upstream of the front attested to the fluidity of the flow. The front was estimated from helicopter overflights to be about 1 m thick and 80 m wide, traveling at rates of 100 to 160 m/h. An additional flow traveled southeastward from Puu 00 during episode 8. It was primarily spatter fed and, by 1615 H.s.t. September 6, had traveled only 250 m from Puu 00. A moderate tephra fallout carried by tradewinds smoothed the southwest flank of the Puu Oo cone. The upper flank, however, was disturbed between midnight and 0200 H.s.t. September 7, when a section of the southwest rim and flank began to creep downslope, pro- ducing a wedge-shaped rootless flow. Though moving primarily as a gravity slide, the flow was partly fed by occasional falls of heavy fluid spatter. The event left a 30-m-deep scar in the southwest rim and a pronounced bulge in the profile of the southwest flank of the cone (see fig. 1.103). After nearly 24 hours of vigorous lava production, foun- tains began to diminish in height and density at approx- imately 0518 H.s.t. September 7. For the next 7 minutes, a few erratic fountains, reaching 20 to 30 m above the cone, alternated with sporadic, dispersed sprays of ejecta; by approximately 0526 H.s.t., all fountaining had ceased. According to the final aerial reconnaissance, the north- eastern lava flow extended approximately 4.4 km from the vent. At the episode’s end, the flow was largely aa. Owing to poor weather, a comprehensive set of aerial photographs was not obtained until after episode 10, and by then much of the episode 8 flow was buried. Thus, the areal extent and volume of episode 8 lavas are only ap- proximate. Aa thicknesses measured in the episode 8 flow to the northeast ranged from 3.5 to 5.5 m (pl. 3). The proximal part of the flow contained a well-developed evacuated channel, 10 to 20 m wide and 2 to 4 m deep, bounded primarily by aa levees; pahoehoe overflow levees occurred locally. The southeastern rootless flow had ap- parently received an increased volume of spatter during the night of September 6—7. By the episode’s end, the flow had extended approximately 1.7 km southeast of Puu 00 and was primarily aa, 2.7 to 4.0 m thick. An additional small, spatter-fed flow extended 300 m westward of the cone. We estimate that the volume of episode 8 lava was approximately 8x 106 m3, which covered an area of 2.0x 106 m2. This was the smallest volume yet erupted in a single episode during the entire 1983 series. Episode 8 was also the shortest single eruptive episode yet observed, lasting just more than 24 hours. A calculated average lava- discharge rate of about 330,000 m3/h was the highest determined since eruptive activity localized at Puu 00 in June 1983. After episode 8, the vent complex consisted again of a steep-walled cone enclosing a nearly circular crater, ap- proximately 90 min diameter at its rim crest (fig. 1.79). The base of this cone was draped with talus and fractured spatter-fed flows. Although continued slumping re- juvenated the scar in the southwest rim, the primary breach was still in the northeast wall. This breach con- sisted of a vertical cleft or notch, approximately 10 m wide and 5 m above the level of the crater floor. It led downslope outside the crater into an evacuated channel FIGURE 1.79.—Puu 00 after episode 8. Crater diameter is approximately 90 m at rim crest. Spillway is visible in upper left. Two open conduits, about 3 m in diameter, extend steeply downward from crater floor. Steep outer walls of Puu 00 are surrounded in part by talus (lower left) and fractured, spatter-fed flows (top). Fresh scars at right record continued slumping of southwest rim and flank. View eastward; photo- graph taken at 1416 H.s.t. September 9, 1983. Numerals (lower right) indicate date and time. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 75 that followed the axis of the northeastern episode 8 flow. The interior crater walls, except for the southwest sector, which was slumped inside as well as outside, were smooth, spatter mantled, and relatively uniform in height. Two nearly circular conduits, about 3 m in diameter, extended steeply downward from the rough, blocky crater floor. EPISODE 9 (SEPTEMBER 15—17, 1983) Only one week separated episode 8 and the onset of low- level eruptive activity within Puu Oo Crater before episode 9. Shortly after midnight on September 14, a time- lapse camera on the northeast rim of Puu 00 recorded a brief occurrence of low-level spatter emission from the more easterly of the two open conduits on the crater floor (fig. 1.79). At about 1713 H.s.t. September 14, sporadic but slowly accelerating spatter production began again and continued into the morning of September 15. Frag- ments of lava were thrown to maximum heights of about 10 m above the crater floor. At no time during this period was any activity from the more western conduit recorded on film. At 1026 H.s.t. September 15, lava began to well up and out of the central conduit, forming a pond in Puu Oo Crater. By 1132 H.s.t., the pond stood 5 m deep, high enough to send a small pahoehoe flow (total volume, approx 300 m3) cascading over the spillway. It traveled 20 to 30 m before the pond began to drain back into the conduit and the flow was beheaded. By 1312 H.s.t., lava was no longer visible in the crater; only occasional spat— ter and glow in the conduit opening are recorded on film. A second pond began to fill the crater at about 1450 H.s.t.; this time, observers were present on the rim. The low vigor of fountain activity suggests that the lava filling the crater was degassed; a small amount of spatter was being ejected from a 3-m-diameter dome fountain that played on the surface of the thin-crusted lava pond. A drainback began at approximately 1512 H.s.t., and lava poured back down the central conduit in 6 minutes. By 1527 H.s.t., lava reappeared, and the pond again began to rise. At 1541 H.s.t., for the second time, the level of the pond rose high enough to send a small pahoehoe flow over the north- eastern spillway and into the evacuated episode 8 chan- nel (see fig. 1.31). On the basis of estimated channel width, depth, and average surface velocity of 4 m, 1 m, and 1 m/s, respectively, a flux of 10,000 to 20,000 m3/h was esti- mated for this small flow. The lava pond maintained a fairly constant depth of 5 m, while a 3- to 8-m-high dome fountain played continuously above the site of the cen- tral conduit. By 1610 H.s.t., the slow-moving flow had traveled 150 to 175 m down the channel and was easily sampled. The dome fountain remained active and doubled in average height by 1624 H.s.t. Activity steadily in- creased in vigor, and by 1711 H.s.t., after observers left, a camera located 750 m west of Puu 00 at camp D recorded spatter rising above the crater rim. During the night of September 15-16, time-lapse film recorded a fountain emanating from Puu Oo Crater and reaching heights of as much as 200 m (see fig. 1.24). In the early morning, the roar of the fountain was audible 20 km away at HVO, and a large brown fume cloud, visi- ble from HVO during most eruptive episodes, drifted southwestward with the trade wind. Throughout the morning, the fountain was visible from along Hawaii Highway 11 between Hilo and Kilauea’s summit. Although the fountain height varied appreciably throughout episode 9, it tended gradually to decay from general levels of 150 to 200 m in the early part of the episode to about 100 m in the later part. The fountain was broad and commonly erratic in trajectory. At times, two or more separate jets were clearly visible that emerged from the crater in a V-shaped pattern, as if originating from the same point but diverted by some obstacle. Because of the vigorous, low-trajectory fountain jets, spat- ter bombarded almost the entire outside surface of the cone, armoring the flanks with thin sheets of fluid spatter- fed flows (see fig. 1.7). Heavy spatter fall to the west pro- duced short-lived rootless flows that moved over tephra from earlier episodes as well as from episode 9. Light- weight tephra fell in abundance on the southwest flank of Puu Oo—possibly the heaviest observed since the episode 3 tephra fall at the 1123 vent. It smoothed the bulge left after episode 8 but added preferentially to that sector of the cone (see fig. 1.103). The major lava flow of episode 9 exited the spillway and continued northeastward on top of week-old episode 8 basalt. About 1.5 km from the vent, the flow veered slight- ly northward and followed the northwest edge of the episode 8 lava into the rain forest (pl. 3). Although a vigorous, channelized pahoehoe river traced the axis of the flow to the northeast, early lava exiting the vent must have traveled as a broad, thin sheet of pahoehoe; attempts to reach the channel edge on the morning of September 15 were thwarted by a wide area of still-hot pahoehoe, extending several hundred meters northward from the river. By 1200 H.s.t. September 16, the flow had developed two lobes that were traveling northeastward, parallel to each other. The more southerly lobe advanced only 200 to 300 m before stagnating; for the rest of the episode, only the more northern of these two lobes was active. Between 1200 H.s.t. September 26 and 0930 H.s.t. Sep- tember 17, this lobe traveled at average rates ranging from about 40 to 160 m/h along the northwest edge of the episode 8 lava flow (pl. 3). The episode 9 flow was predominantly aa, except for the proximal pahoehoe river. The eruption stopped at 1920 H.s.t. September 17. The conspicuous flow to the northeast, primarily aa, extended 76 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 a total of 5.3 km from Puu 00 (pl. 3). Poor weather again prevented systematic coverage of the eruption area with aerial photography after episode 9. Thus, a volume of 8 x 106 m3 over an area of 2.1 x 106 m2 is based on a com- bination of aerial-reconnaissance sketch maps made dur- ing the eruption and mapping from post-episode 10 photographs, which showed exposed parts of the episode 9 basalt. The estimated average lava-discharge rate based on these numbers is 150,000 m3/h. For several days after the eruption, numerous reports of black clouds or dense black fume emanating from Puu 00 were received at HVO. However, no unusual activity at the vent was ever confirmed. Inspection of the vent on September 23 revealed two open conduits, strikingly similar in position and appear- ance to those observed before episode 9, extending down- ward from the rubble-strewn floor of Puu Oo Crater. The more conspicuous of these two conduits was a cylindrical vertical pipe, 4 m in diameter except for the upper few meters, where it flared outward to a diameter of 6 to 7 m. We could see only 15 to 20 m down the conduit (see fig. 1.14). The upper few meters was partly smoothed and plastered with a thin covering of lava that may have formed during the later stages of drainback at the end of episode 9. Elsewhere in the pipe, the walls exhibited thin, horizontally layered, partly oxidized, platy to rubbly basalt. Although we heard intermittent low-pitched, rumbling exhalations, and heavy fume issued from the conduit, no glow was observed from helicopter or the ground. About 10 m west of this open conduit, a second open hole, elongate in an east-west direction, extended down- ward from the crater floor. This hole was about 1 by 2—3 m at the surface and was brightly incandescent at a depth of about 5 m. Heavy fume also issued from this western conduit. The crater itself was nearly bowl shaped and circular in plan view; its diameter ranged from 30 m at the level of the floor to 100 m at the rim crest, and the spillway stood about 4 m above the general level of the floor. The interior walls of the cone were a combination of steep, primarily smooth, spatter-mantled surfaces to the north and south, and a gentler, slumped, blocky slope on the west. The spillway persisted as a narrow V-shaped notch in the northeast wall, and the rim of the cone descended gradually toward the spillway from either side; thus, the northeast sector of the cone was significantly lower than other sections. An additional low area persisted in the southwest rim, where previous late-stage collapse had been localized. The outer flanks of Puu 00 were mostly steep. Locally, they were a chaotic jumble of slumped blocks below exposed headwalls of agglutinate. Elsewhere, they were smoothed by coherent spatter—fed flows or by tephra deposits that were situated preferentially to the west and southwest. EPISODE 10 (OCTOBER 5-7, 1983) Copious amounts of steam and fume were seen at Puu 00 after episode 9, but no eruptive activity was recogniz- ed until October 2, when a small new lava flow, estimated at 300 to 500 m3 in volume, was observed on the floor of the crater, surrounding a brightly incandescent cen- tral conduit (open pipe, fig. 1.14). This flow had apparently issued from the vent between 0750 and 0800 H.s.t., when an electric tripwire in the crater was cut. When we arrived at 1000 H.s.t., the flow surface was slowly collapsing as lava drained back from beneath it into the open conduit. After that initial small flow was extruded, the vent was relatively quiet through October 4. Observers saw inter- mittent low-level spattering in the open conduit, varying fume production, and a few occurrences of glow over the cone at night. Sometimes the top of the magma column was visible a few meters below the top of the conduit. Harmonic—tremor amplitude and intensity of glow over Puu 00 increased together just after midnight on October 5. At about 0106 H.s.t., several minutes after low-level fountaining had commenced inside the crater, lava began to spill through the deep breach in the northeast rim. By 0200 H.s.t., sporadic fountaining had increased in vigor sufficiently to be occasionally visible over the crater rim from camp D; by 0400 H.s.t., a roar from the eruptive area was audible at HVO, accompanied by a strong glow in the predawn sky. When we arrived at 0730 H.s.t. October 5, we saw the most spectacular fountain thus far at Puu Oo—a single, vertical jet, more than 200 m high (see frontispiece). The fountain was broad, and its base filled nearly the entire bowl-shaped depression within the cinder and spatter cone. Fountaining was high enough that spatter and tephra bombarded almost all sides of the cone, but it was concentrated in the north, northeast, and southeast sectors. The only lava flow of significant volume was a multilobed, tacky pahoehoe flow that had traveled 300 to 400 m southeastward from the northeast base of Puu 00. An additional sluggish aa lobe advanced northeast along the path of the episode 9 pahoehoe channel; slow-moving spatter-fed flows covered a large part of the north flank of Puu 00. We were struck by the low apparent flow out- put and limited flow progress in comparison with previous episodes. The spillway contained a vigorous lava river with an intermittent standing wave near the base of Puu 00. The spillway lava and nearby spatter-fed lava coalesced near the base of the cone and fed the flows to the southeast (eastern flow, pl. 3) and northeast. Unobserved from the air at 0730 H.s.t. but unmistak- able on the ground was still-hot pahoehoe, extending at 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 77 least 1 km to the north and northeast of Puu 00. No longer active, these flows were being overrun by the more slug- gish flow advancing northeastward. Time-lapse-camera data show that an early pahoehoe flow to the north and northeast had occurred. At about 0900 H.s.t., the fountain began decreasing in height, declining by 1100 H.s.t. from more than 200 m to a low of about 40 m. During this interval, the ap- pearance of the northeastern flow also changed. By 1000 H.s.t, a broad, complex pahoehoe flow with an incipient central channel and many anastamosing distributaries was traveling northward and northeastward on top of the earlier aa. Fountain activity began to fluctuate markedly and became more complex after its pronounced diminution between 0900 and 1100 H.s.t. October 5. Although the fountain subsequently was as high as 200 m for some periods, large variations in height and trajectory were superimposed on an overall trend of steadily decaying height (see fig. 1.24), and the development of several distinct fountains within Puu Oo Crater made episode 10 fountain activity some of the more complex observed at Puu 00 up to that time. Shortly before 1100 H.s.t. October 5, a second fountain developed just west of the main jet, then about 60 m high. Initially, it was a smaller dome fountain (see fig. 1.28) that repeatedly disintegrated in low bursts of spatter. For much of the day, however, until just after 0000 H.s.t. October 6, the two fountains fluctuated greatly in height, trajectory, and degree of separation. At times, a single, tall fountain rose vertically for hours, followed abruptly by a rapid decay and appearance of two energetic jets emanating from what appeared to be the central conduit and diverging in a V-shaped trajectory. Commonly, though not in every case, a decrease in the height of the combined fountain was accompanied by a sudden increase in apparent lava output over the spillway. The complexity of fountaining continued to increase. Early on October 6, a third small and possibly indepen- dent fountain emerged adjacent to the northwest interior wall of the crater. This fountain grew rapidly in height to overtake the main fountain. After about 10 minutes, the new fountain lost its identity as the entire crater became filled with low, chaotic fountains, rarely reaching 40 to 70 m above the rim. In harmony with the transition to low, chaotic fountains, the time-lapse film record clearly shows an increase in the flux of lava pouring over the spillway and feeding the northeastern flow. After about an hour, fountaining returned, between 0200 and 0300 H.s.t. October 6, to the simpler system of a single and, at times, double jet that reached a maximum height of 200 m. An apparent decrease in the flow of lava through the spillway accompanied the return to high fountaining. Varyingly high fountaining persisted until nearly mid- day October 6, after which the fountain height decreased and the small, separate fountain returned at the base of the northwest interior wall. This time it was unques- tionably distinct from the other fountains in the crater: It clearly represented a separate new vent. Shortly there- after, another new vent, with its own fountain, opened high on the west interior wall of the crater. These second- ary fountains remained small, generally rising to less than 10 to 20 m above the cone, and built subdued local spat- ter ramparts on the preexisting Puu Oo cone rim. General- ly low fountains, partly on the rim and also in a chaotic array within the crater, persisted with little variation until the eruption’s end on October 7. During this period of low, multiple fountains, output over the spillway, as recorded on time-lapse film, was again higher than during earlier periods of higher fountaining. The lava-channel system for the northeastern flow was slow in becoming well established, and for the first 2 days the flow was mainly sluggish aa that advanced, largely on top of the episode 9 flow, at an average rate of about 50 m/h. The flow was relatively thick and broad, and much lava spread out near Puu 00 (pl. 3). During that time, much of the erupting lava was supplied as fluid spatter, which, along with intermittent overflows from the north- eastern lava river, fed the predominantly spatter fed eastern flow. It, too, advanced slowly, at about 50 m/h. Lower fountain heights in the later part of episode 10 led to a diminished supply of fluid spatter feeding the southeastern flow; the supply of lava to the river feeding the northeastern flow increased concomitantly, and the southeastern flow slowed. The increased supply invigor- ated the lava-channel system, and when we reconnoitered the northeastern flow front at 0730 H.s.t. October 7, the fluid pahoehoe channel had extended to the toe, almost 3 km from the vent. Thus, the flow then extended more rapidly, at more than 100 m/h, as a narrow lobe pushing through the rain forest at the northwest edge of the episode 9 flow. Increased supply to the northeastern lava river also resulted in the formation of a series of stand- ing waves that were conspicuous within 200 m of the vent on October 7. Formed in the 10- to 15-m-wide river, the waves were several meters in amplitude and separated by 15 to 20 m. Low fountaining within Puu 00 continued until at least 1633 H.s.t. October 7, according to a time-lapse-camera record. However, poor weather obscured the View, and we have neither a good film record nor a direct observa- tion of the episode’s end. Harmonic tremor in the erup- tive zone decreased in amplitude rapidly at about 1650 H.s.t., nearly 64 hours after the start of the eruption; we infer that the eruption ended then. Lava from Puu 00 covered an area of 2.7 x 106 m2 with approximately 14x 106 m3 of basalt during episode 10; 78 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 the average lava-discharge rate was about 220,000 m3/h. The northeastern flow extended a total of 4.1 km. Like most long flows from Puu 00, it consisted predominant- ly of aa; its average thickness was about 4 m. A central channel was discernible from the vent to 2.5 km down- flow, and pahoehoe overflows from the channel occurred along much of this distance. Beyond the point where the central channel was recognizable, the flow was entirely aa, and no discernible channel structure was preserved. The southeastern flow, which had been most active dur- ing the first 2 days of episode 10, consisted of two lobes, the longest of which extended a total of 3.4 km from the vent. Both lobes were entirely aa, averaging about 6 m thick. In the zone between the two major flows of episode 10, a broad lava delta extended 900 m eastward and north- eastward of Puu 00. The delta was composed of a com- plex stack of aa and pahoehoe flows emplaced primarily during the first 2 days of the eruption before a channel was well established in the northeastern flow. For the last day of episode 10, this delta was intermittently enlarged by transient spatter-fed flows and a few overflows from the northeastern pahoehoe channel. After episode 10, the Puu Oo edifice was an imposing, broad cinder and spatter cone, 150 to 200 m wide by 300 m long, elongate southward and rising about 80 m above the preeruption surface. The circular crater and slopes that defined Puu 00 were characterized by extremely rough and chaotic terrain (fig. 1.80). Large, wedge-shaped stacks of fractured spatter-fed flows, more than 10 m thick, extended northward and northwestward from the cone. A complex of spatter-fed flows interbedded with and mantled by a thick tephra deposit that had been accumulating since episode 8 caused the southward elongation. The spillway was a broad, rubble-covered low spot still on the northeast rim of the cone. A chute led down the northeast flank of Puu 00 and bifurcated around a large block of cone material that may have been par- tially rafted; the two distributaries then coalesced into a single aa-floored channel that wound its way toward the northeast. The complex interior of Puu Oo Crater was choked with rubble consisting of both boulder-size material and whole slivers of the cone’s walls and rim that had collapsed into the depression after the eruption (see fig. 1.15). Grooved and slickensided surfaces on the inner crater walls in- dicated that the floor of the crater had undergone 5 to 10 m of subsidence at the end of episode 10. One especially large, coherent section of the west and southwest wall had slumped toward the interior of the crater. Its toe formed a steep wall running approximately northwest-southeast through the crater. At the northwest end of the wall, where it intersected the floor of the crater, a low area of intense heat and fume, near the position of one of the secondary fountains during episode 10, persisted through- out the subsequent repose period. A second area of localized fuming and heat emission, near another of the secondary fountains observed during the later part of episode 10, was high on the west-southwest rim of the cone in the zone of detachment of the large collapsed sec- tion. Fuming and incandescence in open cracks persisted there throughout the subsequent repose period. The crater rim above both sites of secondary fountaining had been modified by falling spatter to form subtle, local spat- ter ramparts. Uprift of the crater, running down the tephra-covered west flank of Puu 00 along the strike of the January 1983 eruptive fissure, isolated areas of persistent fuming were visible for most of the repose period following episode 10. This phenomenon occurred after many eruptive episodes, and sometimes incandescent cracks persisted high on the cone. EPISODE 11 (NOVEMBER 5—7, 1933) The repose interval between episodes 10 and 11 lasted 30 days, the longest repose since eruptive activity became localized at Puu 00 in June 1983. During this period, fume was emitted at low levels from the vent complex. Al- though minor glow was reported over the south side of the cone on the nights of October 30 and November 1, no eruptive activity was observed during the repose period. Unlike the previous six episodes, when vigorous eruption was preceeded by hours to days of low-level erup- tive activity within the crater, episode 11 discharge began suddenly. The electric tripwire in Puu Oo Crater was cut between 2350 and 2400 H.s.t. November 5, coincident with the onset of lava emission in the crater. Eruption of lava first signaling the onset of episode 11, however, was not from vents within the Puu Oo cone. The camp D time-lapse camera recorded the initial outbreak west of the main Puu Oo edifice at approximately 2350 H.s.t. It occurred along a 30- to 40-m-long segment in the vicinity of the January 23, 1983, vent (pl. 1), about 200 m northeast of Puu Kamoamoa and about 200 m uprift of the west base of Puu 00. Activity along this zone was weak, consisting of discontinuous low fountains, less than 5 m high, and low to moderate levels of lava emission. Fountaining then began high on the west flank of Puu 00 at about 2356 H.s.t.; nearly simultaneously, fountaining broke out within Puu Oo Crater. A general downrift migration of activating vents continued as at least three distinct vents opened east of Puu 00. This activity culminated at approximate- ly 0107 H.s.t. November 6 as a final small fissure vent opened about 150 m northeast of Puu 00. Most of the extracrater vents paralleled or coincided with the 1983 eruptive fissure. Fountaining associated with eruptive ac- tivity at all vents outside of Puu 00 remained relatively 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS O 100 200 METERS L_____L__————-J FIGURE 1.80.—Puu 00 and nearby episode 10 flows (10) after episode 10. Solid line, flow boundary; dashed where approximate. Crater is approximately 90 to 100 m in diameter. Conspicuous evacuated channel heading northeast from Puu 00 (arrows) fed northeastern flow of episode 10. Eastern flow was mostly spatter fed. South quadrant of Puu 00 is smoothed owing to tephra accumulation since episode 8. Thick pile of spatter—fed flows is Visible on northwest flank. Photograph 83.10.11JG120A#8 by J .D. Griggs, taken October 11, 1983. 80 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983-84 weak throughout the night. Inside Puu Oo Crater, the fountain height quickly reached and was sustained at a maximum of about 50 m (see fig. 1.24). By 0010 H.s.t. November 6, glow over the eruption site was visible from HVO. Broad fountaining was reported from Mountain View (18 km north of Puu 00) at 0200 H.s.t., and fountains remained Visible from Kalapana (14 km southeast of Puu 00) and Mountain View through dawn. At 0730 H.s.t. November 6, when we arrived at the site, active vents extended discontinuously from about 200 m uprift of Puu 00, over the crest and through the crater, and to about 300 m downrift of the cone (fig. 1.81). These vents produced flows that traveled in several directions from the cone (pl. 3). The major flow, supplied by way of the spillway, was traveling northeastward along the path of the episode 10 northeastern flow. It had advanced rapidly since the beginning of episode 11, and at 0730 H.s.t. November 6 it was composed of pahoehoe nearly to its terminus, which was 3.7 km from the vent. For the rest of the day, the flow gradually slowed, widened, and locally entered the rain forest north of the episode 10 flow. Lava discharge into the northeastern pahoehoe river represented about 80 to 90 percent of the total output; the remaining 10 to 20 percent was accounted for by lava discharge from the subordinate vents uprift and downrift of Puu Oo Crater. The small vents east of Puu 00 had low fountains (max 5—10 m high). Lava welled out of these small vents and, along with overflow from the main lava river, fed a slow- moving pahoehoe flow that traveled eastward. Uprift of Puu 00, at least six extracrater vents erupted more vigorously than those downrift. Activity at the smaller of these vents was characterized by repeated bursts that sent fragments to heights of less than 10 m, forming small, coalescing, conical spatter cones. Low-level emis- sion of lava generally occurred from small openings in the bases of these growing spatter structures. The western- most vent (farthest right, fig. 1.81) contained a low (3—5 m high) dome fountain that occasionally disintegrated, con- structing a spatter shell around itself that was open to the north. Lava issuing from these small western vents and, in part, from a more active one high on the west flank of Puu Oo fed overlapping short sheetflows of pahoehoe to the north and west. High on the west flank of Puu 00, the most energetic extracrater vent originated from near the site of the fum- ing and intermittently incandescent area observed there after episode 10. It produced moderately vigorous foun- tains, possibly as much as 20 m high, and fed a cascade that was the primary source for the southern flow (pl. 3), pahoehoe fed by way of a central channel. In addition, occasional northward flows from this vent produced tran- sient, overlapping sheets and tongues of pahoehoe (fig. 1.81). Fountaining from the vent was rapidly construc- ting its own enclosing spatter rim that eventually grew into a separate crater on the Puu Oo cone. The bulk of the fountain activity and lava production, however, was confined to the main Puu Oo Crater. There, multiple fountains, possibly as many as four or more, played from separate vents on the crater floor and interior walls. One prominent fountain at the base of the north interior wall intermittently fed a rootless flow down the north flank of Puu 00. In addition, for the entire episode, FIGURE 1.81.—Discontinuous, 700-m—long line of erupting vents with low fountains during episode 11. Line transects Puu Oo (center) and ex- tends both uprift (right) and downrift (left). Small red spots at far left represent most northeasterly vents. Photographs taken at 0730 H.s.t. November 6, 1983. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 81 several small fountains erupted through the channelized lava river on the northeast flank of the cone below the spillway. Throughout episode 11, fountaining from within Puu 00 remained low (see fig. 1.24), and lava output re- mained steady and high. Owing to the dense jumble of fountains, it was difficult to distinguish a coherent lava pond inside Puu 00. The steady volume of lava spilling down the northeast side of the cone, however, suggested that a significant surface reservoir of lava was contained within the crater. The vigorous pahoehoe river that traveled northeast- ward from Puu 00 contained a remarkable set of stand- ing waves over a zone about 150 to 200 m long (fig. 1.82). The waves first appeared in a stretch of the river about 100 m from the base of Puu 00. At least five major waveforms were visible along this stretch, separated by 20 to 40 m; amplitudes of the waves ranged from approx- imately 1 to 3 m. In the middle of the zone of standing waves, the gradient of the surface of the flowing lava and the enclosing levees changed sharply. The smaller waves were nearer the vent, where the gradient was estimated at 10°. Larger waves occurred at and below the point where this gradient became gentler, possibly 2°. About 90 m downstream of this inflection, the lava river widened and slowed considerably, and waveforms disappeared. In the zone of standing waves, maximum surface velocities of 8 to 10 m/s and a relatively constant average width of 15 m were estimated. Using an assumed depth of 2 to 4 m and an approximate average surface velocity of 4 to 5 ml s, we calculate a flux of about 500,000 to 1,000,000 m3/h. These values are 2 to 4 times that of the average lava flux calculated from the mapped volume (table 1.3), and they presumably represent inflation of the lava by air and vent gases. Posteruption measurements of channel geometry indicated that the real-time estimates of width and depth were minimums. Although the vigor of fountaining within Puu Oo Crater remained steady, the level of activity at the eastern ex- tracrater vents diminished during the night of November 6—7. By dawn on November 7, only low fountains eman- FIGURE 1.82.—-Episode 11 lava river, flowing from right to left, showing part of series of 1- to 3-m-high standing waves described in text. Downstream of waves, lava river widened and slowed significantly as it swung northeastward. Geologist is standing on episode 10 basalt; new overflow levees are visible adjacent to active river. View northeastward; photograph by M.L. Summers, taken November 6, 1983. 82 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 ating from vents near the northeast base of the cone con- tinued to be active and to contribute, along with the vents inside Puu 00, to the northeastern lava river. During the night of November 6—7, the western extracrater vents also shut down gradually. The dome fountain at the westernmost vent (fig. 1.81) remained active until about 0200 H.s.t. November 7. The moderately vigorous vent high on the west flank of Puu Oo shut down at about 0400 H.s.t. November 7 but continued to emit small bursts of spatter intermittently until daylight. Thus, supply to the southern flow was terminated. For most of November 7, fountaining was confined to those vents active within Puu 00 and at its northeastern base; the only active flow was the elongate aa flow that advanced to the northeast. At approximately 1915 H.s.t. November 6, a crack opened in the area west-northwest of Puu 00 (pl. 3) with a report audible to nearby observers over the roar of the fountaining. Although the observers, who were about 30 m from the crack, felt no movement, the borehole tiltmeter (KMM, fig. 1.1), located about 200 m north of the new crack, was jarred off scale. This new crack steamed profusely at first. Oriented N. 60 °E., it was at least 360 m long and recorded a maximum of 0.5 m of extension perpendicular to its strike. The amount of extension and the surface expression of cracking dimin- ished westward and disappeared altogether within episode 1 lava north of Puu Kamoamoa and camp D (pl. 3). To the east, the crack was buried by new lava of episode 11. No subsequent displacement occurred. Episode 11 ended at 1841 H.s.t. November 7 as the remaining vents gradually shut down over a period of several minutes. Fountaining on the northeast flank of Puu Oo stopped first. Then, after several diffuse sprays of fragmental material were emitted, the more central fountains inside Puu Oo stopped erupting. One major flow to the northeast and two minor flows to the east and southeast were produced during episode 11 (pl. 3). The northeastern flow was 9.6 km long and split into two distinct lobes approximately 5 km downflow. The bulk of this flow was aa; however, pahoehoe overflow levees from the active central lava river persisted about 3 km along its length. Likewise, the 2-km-long southeast- ern flow was predominantly pahoehoe in its upper reaches, thickening downstream and evolving into aa in its last 1-km stretch. The eastern flow, 1.4 km long, also consisted primarily of pahoehoe in its proximal part and of aa in the last 600 m. During 43 hours of eruption, 12x106 m3 of new basalt covered an area of 4.3x106 m3; the average lava-discharge rate was 280,000 m3/h. Puu Oo changed strikingly in shape during episode 11 (figs. 1.80, 1.83). Spatter deposits from vents erupting on and near the northeast flank had extended the flank more than 100 m northeastward. In addition, separate craters had grown high on the north and west flanks of Puu Oo. Dense fume emanated from the northern satellitic crater; thus, it is not distinguished in figure 1.83. Incandescent cracks were visible at several places in the spillway cor- ridor that transected the northeast flank of Puu 00. The spillway was no longer a narrow, steep notch; it had become a broad, elongate breach in the northeastern wall of Puu Oo, floored by a smooth, easily scaled ramp of pahoehoe. The vents that had been active uprift and downrift of Puu 00 were marked by 5- to 10-m-high spatter cones or ramparts overlooking the new flows. These vents re- mained hot for many days, but no interior incandescence was ever seen. EPISODE 12 (NOVEMBER (JO-DECEMBER 1, 1983) Between episodes 11 and 12, copious hot, oxidized fume issued from the interior of Puu Oo. Incandescent open- ings persisted in many places throughout the cone com- plex, although the intensity of glow diminished somewhat over the course of the repose period. Most conspicuous of these incandescent areas were those in and near the Puu Oo spillway. The crater rim and internal septa com- posed of agglutinate delineated several chambers within the larger crater. Except for causing some minor collapse within Puu Oo Crater, the M = 6.7 Mauna Loa earthquake of November 16, 1983 (Koyanagi and others, 1984; Buchanan-Banks, 1987) had no significant effect on Puu 00 or on the low- amplitude harmonic tremor in the eruption area during this repose period. Disturbance of the tripwire between 1600 and 1610 H.s.t. November 29, in addition to reports of glow over Puu 00 between 2200 and 2310 H.s.t., suggests that low- level activity preceded the onset of vigorous episode 12 eruption by 6 to 11 hours. Harmonic-tremor amplitude in the eruptive zone began to fluctuate at about 0730 H.s.t. November 29 and remained irregular throughout the day. After a gradual but steady increase beginning at 2300 H.s.t., tremor amplitude increased significantly at 0445 H.s.t., and by 0450 H.s.t., fountaining at Puu 00 was visible from Hilo. Observers arrived at Puu 00 at 0750 H.s.t. November 30. At least four loci of fountaining, all within the crater or high on its rim, could be identified (see fig. 1.16). The northwestern fountain of the group, which was the most vigorous, rose an estimated 50 to 60 m. A second foun- tain, less vigorous than the first, played from a site ap- proximately in the center of the crater. A third major fountain emanated from the spillway, at the northeast end of Puu Oo Crater; this fountain appeared to be erupting from a level lower than the other vents. The fourth major fountain was actually a group of fountains erupting from the southern part of the Puu Oo rim and crater. Good 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 83 views into the crater revealed a complex, multilevel lava Several separate flows issued from Puu 00 that pond, supplying a cascade northeastward over the spill- morning (pl. 4); all were pahoehoe during our first recon- way. Each of the four active vents apparently contributed naissance at 0845 H.s.t. The major flow traveled to this pond. northeastward for the first few hundred meters as a high- 0 100 200 METERS l L 1 FIGURE 1.83.—Puu 00 after episode 11, showing nearby parts of episode 11 flows (11) and spatter ramparts built by small episode 11 vents (FV) to northeast. Solid line, flow boundary; dashed Where approximate. Hachured line, crater rim; hachures point inward. Photograph 83.11.14JG120A#11 by JD. Griggs, taken November 14, 1983. 84 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 velocity, vigorous pahoehoe river confined within the evacuated episode 11 channel; beyond that it slowed and broadened. Spatter from the northwestern fountain and pond overflow fed a pahoehoe flow that advanced north and northeast from the cone. The less vigorous, southern fountains fed several narrow spatter- and pond-fed flows that traveled southward and eastward from Puu 00. The general height of the array of low fountains gradually decreased through November 30 to about 20 to 30 m, and that level was maintained through the rest of episode 12 (see fig. 1.24). A unique variation in Puu 00 fountain behavior occurred on December 1. Beginning at about 0040 H.s.t., a vigorous jet of gas, possibly steam, with entrained tephra became conspicuous at the north- western vent, where, earlier in the eruption, a vigorous lava fountain had played. This jet, about 50 m high, per- sisted throughout the rest of the episode. Periodically, southeast winds caused fallout from the jet to pelt camp D with 2- to 3-cm-diameter tephra. Early on December 1, both the level of the pond inside Puu 00 and the general vigor of fountaining had de- creased. The local fountaining centers had developed individual spatter walls around themselves, so that Puu Oo appeared distinctly chambered (see fig. 1.17). Most of the erupting lava continued to feed the northeastern flow; a lesser amount fed the northern flow, and the eastern flow was stagnant. The average rate of advance of the northeastern flow had declined to about 160 m/h from about 400 m/h on the preceding day, but a reduction in supply to the flow is not implied; the advancing front was 2 to 5 times as wide as on the previous day (pl. 4). At 1545 H.s.t. December 1, after 35 hours of lava pro- duction, a rapid reduction in harmonic-tremor amplitude in the eruptive zone accompanied the cessation of foun- taining at Puu 00. Seen from the air, Puu 00 was an elongate cone, irregular in outline, enclosing at least five distinct chambers separated by conspicuous septa (see fig. 1.17 ). Incandescence was immediately visible in numerous cracks and holes inside some of the chambers; over the succeeding weeks, the openings gradually cooled, and the incandescence diminished. In the northeastern part of the edifice, one large chamber extended downward to a steep- walled pipe, about 30 m across and open for at least 90 m. The walls of this pipe consisted of thin-layered, platy and rubbly basalt that was partially oxidized. No drainback textures could be identified. Spatter and drainback features were, however, identifiable in some of the smaller vent structures clustered inside Puu Oo. Numerous rock- falls within the crater occurred immediately after the eruption and continued well into the subsequent repose period. As was typical of posteruption periods, varying amounts of oxidized fume issued from many parts of the crater. The spillway was an elongate, slightly sinuous, wide chute that ran down the northeast rim of the cone. The main, northeastern flow of episode 12 was extraor- dinarily narrow (less than 100 m) over much of its length (pl. 4). Broad overflow levees composed of pahoehoe occurred along much of the first 4.7 km; beyond that, the flow was entirely aa. The 2—km-long northern flow was characterized by a similar distribution, with a broad pahoehoe field near the vent evolving downflow into aa only. The eastern flows, also pahoehoe evolving down- stream to aa, advanced more than half the distance to Royal Gardens during episode 12. Altogether, 8x106 m: of lava had been erupted at an average rate of 230,000 m /h. Between 7 and 8 km northeast of Puu 00, near the ter- minus of the major, northeastern flow of episode 12, a zone of new, steaming cracks parallel to the east-rift-zone axis was discovered on December 17 by a passing pilot (pl. 4). Subsequent investigation on the ground showed that about 1 m of extension had occurred perpendicular to the strike of the rift zone. At the uprift end, the new cracks disappeared under episode 12 lava; evidence sug- gests that the episode 12 flow may have been broken local- ly by propagation of the new cracks. The lava, however, also was apparently emplaced against crack-related, 1- to 2-m-high fault scarps that did not move subsequently; thus, cracking may have spanned a period of time that encompassed emplacement of the episode 12 flow. The new cracks are within a zone in which numerous older cracks have cut prehistoric basalt. Although these new cracks are approximately on strike with the 1983 erup- tive fissure, they are entirely downrift of the region of shallow earthquakes associated with the January 1983 dike emplacement. Subsequent monitoring near the uprift end of the new cracks showed no further movement. EPISODE 13 (JANUARY 20—22, 1984) Seven weeks of repose separated episodes 12 and 13. The first suggestion of pre-episode 13 low-level activity was a report at 2130 H.s.t. January 12 of a slight orange glow reflected on fume over Puu 00. At 1117 H.s.t. January 20, an HVO observer spotted incandescent cracks in the surface of crusted lava about 50 m down the near- vertical open pipe descending from the large northeastern crater shown in figure 1.17. Between 1300 and 1400 H.s.t., observation from the rim of this crater as well as from a helicopter revealed what was apparently a cascade of lava from an opening in the north wall, deep within the pipe. At about 1545 H.s.t., the lava column began to slow- ly rise, and by 1724 H.s.t. it had reached the level of the spillway and had begun to flow northeastward out of the crater at a rate of approximately 10,000 m3/h. Concur- rently, a 5-m-high dome fountain played discontinuously over the center of the overflowing pond within the crater. The pond surface was roiled, as in a rolling boil in a 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY 0F ERUPTIVE EVENTS 85 saucepan. The vigor of surface agitation and rate of over- flow gradually increased, and by 1740 H.s.t. observers had to retreat from the crater rim. Unlike some previous episodes that reached their max- imum output and fountain height near the start of flow production, episode 13 activity accelerated gradually. By about 2000 H.s.t. January 20, lava output was estimated to be at least 100,000 m3/h in a lava river that flowed northeastward from Puu 00, and the fountain height had slowly increased to 40 to 50 m above the pond (see fig. 1.24). Much different from the complex fountains of the recent previous episodes, this fountain consisted of a single column approximately centered over the position of the large open pipe. None of the peripheral vents of recent episodes was active. This single fountain remained relatively low throughout episode 13 (fig. 1.24), although its height gradually increased to a general maximum of 80 to 90 m. For much of episode 13, the fountain height and the lava flux within the channel close to Puu Oo oscillated in a style we had not seen in earlier episodes. At intervals of about 0.5 to 1 minute, the fountain would diminish to about half its full height and then quickly rebuild to full height. Thus, the general level recorded in fig. 1.24 probably represents neither the maximum nor the minimum but some inter- mediate fountain height. In rapid response to the foun— tain changes, the lava river in the channel near the base of Puu Oo rose and fell. At times, flow through the spillway actually stopped for 10 to 15 seconds. Recovery was always rapid, and surges that filled the channel repeatedly advanced through the first 100 m at about 10 m/s. The surging lava would often overflow the confin- ing banks to produce rapidly moving overflows. Frag- ments of melt thrown out of the surging river on the outsides of sharp bends built ramparts of spatter. Visual estimates suggest that the lava-river flux may have undergone rapid changes as great as threefold. The oscillation in fountain height and lava-river flux apparently recorded repetitive brief interruption in the supply of melt to the vent. It was distinct from more rapid pulsing, at intervals of about a second, that we also saw in the fountain jets. The pulsing output fed a single lava river and flow that moved northeastward (pl. 4), as had all the major flows beginning with episode 6. The flow bifurcated around Puu Kahaualea but rejoined into a single advancing front by midmorning January 21. At 0027 H.s.t. January 22, after 31 hours of eruption, the fountain became intermittent and low for a period of about 10 minutes; eruption stopped entirely at 0041 H.s.t., and the flows were cut off. Eruptive activity resumed at about 0432 H.s.t. with low, intermittent fountaining that lasted until 0502 H.s.t. After another 34 minutes of inactivity, fountaining began anew at 0536 H.s.t. and reached preshutdown levels within about 15 minutes (see fig. 1.24). Output northeastward over the spillway, the only locus of overflow from the lava pond in Puu Oo Crater throughout episode 13, was quickly reestablished and remained vigorous for the rest of the eruption. New lava advanced directly over the earlier episode 13 flow, and the major lobe again traveled in a northeasterly direc- tion. This resumption of activity also resulted in a break- out to the southeast from the northeastern lava river about 700 m from the vent (pl. 4). Beginning at about 1115 H.s.t. January 22, a spasmodic decay of fountaining and output began again at Puu Oo, heralding the end of episode 13, which occurred at 1123 H.s.t. At least five off-on cycles were observed before ac- tivity finally ceased; each cycle lasted about a minute. Off periods were characterized by relative quiet and emission of blue-brown fume from the crater. Helicopter views into the crater immediately after the final disappearance of fountain activity revealed the boiling top of a magma column in a vertical pipe extending downward from the center of the crater floor. Partly crusted at times, the active column remained 0 to 25 m below the level of the crater floor that surrounded the open pipe for the ensuing repose period (see fig. 1.18). The open pipe within the crater (fig. 1.84) was much like the one left after episode 12 that had filled with lava at the beginning of episode 13. About 25 m Wide at the top, the post-episode 13 pipe narrowed downward to about 10 m across at a depth of 20 to 25 m. The crater floor surrounding this conduit was mantled with episode 13 pahoehoe and was generally inclined toward the center of the crater. The interior walls of the crater steepened FIGURE 1.84.—Interior of Puu Oo Crater after episode 13. Top of ver- tical conduit is approximately 20 min diameter where person (circle) is standing. Light fume emanating from conduit is visible in shadowed area at lower left. Spillway, out of view to right, is 5 In higher than ledge on which person stands. View northwestward from southeast crater rim; photograph taken January 25, 1984. 86 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 upward, and the spillway was a steep chute down the northeast side of the cone (see fig. 1.18). Immediately after episode 13, the chute was smoothly paved by pahoe- hoe (fig. 1.85A), but shortly after the eruption ended, blocks and rubble fell into it from the walls (fig. 1.853). In fact, after most eruptions, avalanching of coarse talus into the spillway chute made it a dangerous avenue into the crater. Otherwise, the post-episode 13 crater remained relatively free of rubble for the ensuing repose period. Puu Oo changed significantly in gross shape during episode 13 (see figs. 1.17, 1.18). Tephra fallout from FIGURE 1.85.—Puu 00 spillway and rockfall, approximately 2.5 hours after episode 13. A, Spillway transecting rim of Puu 00 Crater. High- lava mark, which forms a linear boundary between smooth, highly reflective pahoehoe and rougher, less reflective spatter deposits to right of spillway, presumably reflects level of lava that flowed out of crater and down spillway during episode 13. Position of this mark sug- gests that stream of lava was about 10 m thick as it exited crater. Spillway is about 30 m wide. View westward toward crater interior; photograph taken at about 1400 H.s.t. January 22, 1983. B, Similar View to figure 1.86A, several minutes later, showing incandescent blocks that fell from steep north wall. Spillway is about 30 m wide, and so largest of these blocks is about 7 m across. periods of higher fountaining had added some bulk to the south and southwest sides of the cone, healing slump scars and irregularities inherited from episodes 11 and 12. The most radical change, however, was in plan view. After episode 13, Puu 00 was a nearly circular cone enclosing a single central crater. The exterior walls steepened upward from the base of the cone, which was largely sur- rounded by an apron of flows. A deep, sinuous evacuated channel with partly over- hanging levees along its first several hundred meters led away from the northeast base of Puu 00. Its channelized pahoehoe river had supplied the main, northeastern flow during episode 13 (pl. 4). This flow was a composite of the flows produced during each of the two eruptive periods of episode 13. The first flow, emplaced between 1724 H.s.t. January 20 and 0041 H.s.t. January 22, had traveled 7.4 km; it was predominantly pahoehoe for the first 3.5 km and entirely aa for the rest of its length. The second eruptive pulse produced another flow that traveled northeastward on top of the first (pl. 4) for a total distance of approximately 3.1 km. Local aa flows that extended northward and southeastward from the second flow formed as breakouts, on January 22, from the reactivated lava river. Aerial photographs were not obtained until after two additional eruptive episodes had occurred; thus, mapping of the episode 13 lava flow is based in part on helicopter sketch maps made during the eruption. EPISODE 14 GANUARY 30—31, 1984) Throughout the short, 8-day repose period between episodes 13 and 14, the surface of the magma column re- mained visible at depths of 0 to 25 m below the top of the open pipe (figs. 1.18, 1.84) extending downward from the floor of Puu Oo Crater. At times, molten to barely crusted lava was exposed at the surface of this column. However, much of the time the column was capped by a crust of basalt, generally composed of a combination of agglu- tinated spatter and smooth pahoehoe. A small irregular orifice, generally from 0.5 to 3 m across, pierced the crust and served as the vent for minor amounts of spatter, a few small pahoehoe flows, and varying amounts of gas, all emitted episodically in response to gas-piston activity in the magma column. For hours at a time, the gas-piston cycles were rhythmic and brief, 4 t0 6 minutes long, although longer, irregularly spaced events also occurred. The overall rise of the magma column led to flooding of the crater floor sometime between January 27 and 30. This flooding produced a solid crust (fig. 1.52) about 30 m in diameter and 4 m below the level of the spillway. Emis- sion of short (1—3 m long), glassy pahoehoe flows and spat- ter from a 0.5-m-wide vent through the crust near the northeast edge of the pond surface had built a small mound of lava, about 4 m wide and 1 m high. The vent 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY 0F ERUPTIVE EVENTS 87 itself was incandescent, and as we watched it, roaring pulses of gas and minor spatter issued intermittently. At about 1030 H.s.t. January 30, harmonic tremor began a gradual increase that peaked at about 1830 H.s.t. as fountaining from Puu 00 became visible to sailors off the south coast of the island. By 1951 H.s.t., fountaining was visible from the Wahaula visitor center (see fig. 1.1) and from high points in the upper east rift zone. Although a clear view into the spillway was obstructed by vent deposits, time-lapse film from camp E shows that the lava filling Puu Oo Crater reached the level of the spillway and probably began to overflow between 1739 and 1745 H.s.t. The vigor of fountaining gradually accelerated, and by 1848 H.s.t., broad bursts were reaching the top of the cone. The fountain was broad, filling nearly the entire crater. On the morning of January 31, when we arrived at Puu 00, the single broad fountain, approximately centered over the site of the conduit, was about 100 m high. The time-lapse film record (see fig. 1.24) indicates that the fountain quickly built to 200 m in height early in episode 14 and then generally decreased gradually to about 100 m in height at the end of the brief eruption. During that decline, the fountain height fluctuated rapidly. The field observation was that fountain heights fluctuated at inter- vals of 10 to 20 seconds from low levels (10—20 m above the rim of the cone) to higher levels (80-100 m above the rim of the cone). The higher bursts produced short-lived spatter-fed flows down the flanks of Puu 00. Nearly con- tinuous bombardment of the south rim and flank of the cone by spatter produced a large, rootless aa flow that ultimately traveled 1.5 km southeastward of Puu 00 (pl. 4). A smaller spatter-fed flow extended 0.7 km southward. The lava river exiting from the crater over the spillway carried nearly three-fourths of the lava flux. Near the base of the cone, a distributary of the river, diverted north- westward, fed a thick, ponded pahoehoe flow that spread northwest of Puu 00. The main channel fed a complex flow advancing eastward. This time, instead of a single lava river evolving downstream into an aa lobe, a com- plex network of distributary pahoehoe channels emerged from the main channel Within a kilometer of the vent and spread in several directions (fig. 1.86). Several distribu- taries, however, merged near the southwest base of the 1123 cone to produce a flow to the east that evolved downstream to aa and was the most active. By the episode’s end, it had extended 4.7 km from Puu 00 (pl. 4). Other distributaries merged to supply a shorter lobe on the north side of the 1123 cone. After only 19 hours of eruption, lava emission deteri- orated spasmodically between about 1315 and 1318 H.s.t. January 31. Immediately after activity stopped, we had a clear View of the crater interior (fig. 1.87). An incandes- cent circular opening, emitting fume and fine brown par- ticulate matter, marked the central conduit; except for the beheaded flow draining from the spillway chute and glowing collapse along the walls of the cone, no movement was visible. After several minutes, fume filling the con- duit and the interior of the crater obscured the view. Aerial photographs were unavailable until after episode 15; thus, the distribution of episode 14 flows as mapped on plate 4 is approximate in places. When mapped after episode 15, all the exposed episode 14 lava flows were aa except for the thick pahoehoe lobe northwest of Puu 00. The new basalt covered an area of about 2.1 x 106 m2 and had a volume of 6x 106 m3, with an average lava- discharge rate of 320,000 m3/h. After episode 14, Puu 00 again contained a steep- walled, bowl-shaped crater, approximately 40 to 45 m across at the level of the spillway. Extending downward from the floor was a nearly vertical open conduit, about 20 m in diameter, almost an exact replica of the pre- episode 14 conduit (fig. 1.84). The edges of thin lava layers were exposed in its walls. At the top it flared to merge with the sloping crater floor, and the flaring lip was mantled with drainback lava. The interior walls of the crater were lined with a beautiful, delicate, glassy pahoehoe coat and glassy spat- ter bombs. Parts of the wall were furrowed (fig. 1.88), forming a parapet at the rim like that seen on the episode 7 rim (fig. 1.76). A bench that may have represented the high-lava mark of the pond inside the crater during episode 14 was visible on parts of the interior walls (fig. 1.88). Its position suggests that the pond surface stood about 10 m higher than the rock floor of the spillway. In the northeast wall of the crater, the spillway was now a broad, 30- to 40-m-wide cleft. Northeastward, it led down into an evacuated channel that retained definition for about 200 m. The south rim of Puu 00 had been degraded by spatter bombardment into a low, smoothed terrace (fig. 1.87) that would persist through several episodes, providing us with a convenient and safe point of entry and exit. EPISODE 15 (FEBRUARY 14—15, 1984) Although we heard low-pitched, rumbling exhalations from deep within the open pipe on February 1, the day after episode 14 ended, we could see no lava. However, lava was seen deep in the pipe during aerial recon- naissance flights on February 3 and 5. Ground observa- tion on February 7 revealed that a partly crusted magma column stood 45 m below the crest of the spillway. For the next 6 days, this column rose at an average rate of about 4 to 5 m/d. Its crusted surface was broken by a small opening through which spatter was intermittently thrown; the active, roiled lava surface was commonly visible just THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 FIGURE 1.86.—-Puu 00 erupting during episode 14. Complex of distributaries transports fluid lava to north, northeast, and southeast. Fountain at Puu 00 is approximately 100 m high. Altered cinder and spatter cone in foreground is 1123 cone (Puu Halulu); camp E is on right shoulder of this cone. View southwestward; photograph by J .D. Griggs, taken at 1115 H.s.t. January 31, 1984. FIGURE 1.87.—Puu 00 vent, about 8 minutes after end of episode 14. FIGURE 1.88.——Furrowed, spatter-mantled north interior wall of Puu Glowing, 20-m-diameter hole marks location of conduit. Note beheaded 00 after episode 14 (compare fig. 1.76). Wall towers steeply 30 to 40 m lava flow draining from spillway chute and incandescence where local above general level of crater floor. Bench visible in lower part of frame collapse of walls is occurring. Part of helicopter is visible at lower left. may represent a high-lava mark of ponded lava during episode 14; View westward; photograph taken at 1326 H.s.t. January 31, 1984. if so, pond surface was about 10 m higher than the rock floor of the Numerals (lower right) indicate date and time. spillway. Photograph taken February 1, 1984. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 89 beneath the crust. By the morning of February 13, the column surface, which was then free of crust, had risen to within 20 m of the lip of the spillway (fig. 1.51). Be- tween 0700 and 0710 H.s.t. February 14, according to time-lapse film data, the column briefly rose sufficiently high to send a 100- to 150-m-long pahoehoe flow over the spillway and down the evacuated episode 14 channel. Time-lapse film data indicated that low-level fountain- ing commenced inside Puu 00 at about 1940 H.s.t. Feb- ruary 14. Between 1940 and 1943 H.s.t., lava reached the level of the spillway, and sustained overflow began. Foun- taining accelerated quickly along with output. Within an hour, peak fountain heights exceeded 200 m; and in less than 2 hours, at 2133 H.s.t., the maximum height for the episode (350 m) was attained. Thereafter, the single broad fountain centered over the crater declined spasmodically to about 100 m in height at the end of the episode (see fig. 1.24). As in episode 14, the fountain height oscillated widely. High fountains and strong tradewinds combined to pro— duce an extensive tephra fallout downwind of Puu Oo. Tephra fell at least as far southwest as Napau Crater, 4 km uprift of Puu 00 (see fig. 1.1), and evidence of the early high fountains and heavy tephra fall at camp D was conspicuous when we arrived at 0730 H.s.t. February 15. Lightweight, frothy bombs, as large as 50 cm across, littered the ground around the camp, and the canvas tarp covering our shed was still smoldering; it had been burned completely through in places. The major lava flow, fed by lava pouring through the spillway, traveled northeastward from Puu 00. When we first saw it, early on February 15, it was about 2 km long (pl. 4) and almost entirely aa; the pahoehoe-aa transition in the channel was within a few hundred meters of the vent. Subsequently, as the fluid pahoehoe channel length- ened, its envelopment in hot, fresh aa made approach to the central channel difficult. Occasional overflows from the channel were our only source of active pahoehoe for temperature measurements and sampling. An additional aa flow extended eastward. Time-lapse- camera records indicated that this flow had been'fed vigorously during the night of February 14-15 both by spatter and by overflow from the lava cascade down the northeast face of Puu 00. By dawn, however, the flow was being fed only intermittently and was essentially at its final length when we first saw it early on February 15 (pl. 4). The eruption continued steadily until 1458 H.s.t. February 15. Then, over the next 3 minutes, fountaining spasmodically stopped and restarted three times before finally shutting down at 1501 H.s.t. Two conspicuous flows of episode 15 extended north- eastward and eastward (pl. 4) 4.7 km and 2.9 km, re- spectively; both were predominantly aa. In addition, relatively short, thick, spatter-fed aa flows flanked Puu 00 on the west, south, and southeast. The high fountains of episode 15 had deposited much tephra on the southwest flank of the cone. The geometry and general appearance of the bowl-shaped crater were virtually unchanged from its pre-episode 15 condition. Approximately 8 x 106 m3 of new basalt had been erupted over a period of 19 hours, giving the highest average lava-discharge rate thus far—420,000, m3/h. EPISODE 16 (MARCH 3—4, 1984) Conditions at Puu 00 remained relatively stable dur- ing the first 13 days of repose between episodes 15 and 16. High levels of SOg-rich fume emanated from the open pipe, and, except for sighting of incandescence at a depth of 40 to 50 m on February 22 and hearing of low-pitched, rumbling exhalations that may have indicated lava at depth on February 24, no activity was directly observed until February 28. (People camped near the eruption site on February 27 reported glow over Puu 00 at 1900 H.s.t., but this event was not confirmed in the time—lapse film record.) At 1200 H.s.t. February 28, we saw active, partly crusted lava within the pipe, about 30 m below the level of the spillway. Over the ensuing 5 hours, the column rose at a rate of about 1 m/h. By 1700 H.s.t., the lava surface was completely open and vigorously churning. Small amounts of spatter were occasionally emitted from its disturbed surface. During the next 4 days, the column was partly crusted at times. Its behavior alternated between quiescence and rhythmic gas pistoning in which the sur- face rose and fell 10 to 15 m. The lava filled the crater, formed a pond, and began to overflow the spillway at about 1450 H.s.t. March 3. Ini- tially, the fountain was low, 10 to 20 m above the pond surface. The vigor and height of the fountain accelerated gradually, and by 1519 H.s.t., fountaining was high enough to be visible over the rim (about 40 m above the surface of the lava pond) from camp D. At 1700 H.s.t., fountaining peaked at a new record height for Puu 00— approximately 390 m (see fig. 1.24). As in episode 15, the fountain height declined spasmodically through episode 16, although it remained generally high through the first night. By the end of episode 16, the general height was less than 100 m. Glowing tephra, carried by hot updrafts over the vent, was at times wafted to twice the height of the denser part of the fountain. This tephra was especially visible at night, when it caused severe overestimation of fountain heights by observers at distant vantage points. Erratic winds distributed tephra on all sides of the vent. Intermittent- ly, for periods of 10 to 30 minutes during the night of March 3—4, easterly winds dropped incandescent bombs, as large as 20 cm in diameter, on camp D, 750 m west 90 THE PUU OO ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 of the vent. These periods of bombardment made system- atic sampling of tephra easy. However, the aluminum roof of our shed was repeatedly dented and sometimes punc- tured by the larger fragments, and the general clatter of pyroclasts hitting the roof reminded us of an aluminum corn popper at its peak of activity. When high, the single fountain was a broadly based jet of varying trajectory. Within it, at any one time we could see the more brightly incandescent fronts of several distinct pulses rising upward through the fountain (see fig. 1.6). The effect was particularly striking at night. Dur- ing periods of lower fountaining, the fountain decayed into multiple jets of spatter, also varying in trajectory. For several hours before dawn on March 4, the high fountain appeared, on time-lapse film, to be split into two: One fountain was a high, energetic column that leaned to the north, and the other a low, dense fountain that did not rise above the rim of the cone. The situation was similar to one witnessed during episode 10 (see fig. 1.28). Early high fountains produced a thick, broad, spatter- fed aa flow that advanced more than 1 km northward, and several smaller rootless flows that extended southeast- ward and westward (fig. 1.40). The major flow was fed by a vigorous cascade of lava overflowing from the crater through the spillway. This cascade funneled into a nar- row channel for the first several hundred meters, beyond which the pahoehoe river broadened and slowed, heading east and then southeast toward Royal Gardens (pl. 5). The flow advanced mostly on top of earlier 1983-84 flows. Ap- proximately 6 km from its source, the flow split into two lobes, the longer of which extended another 1.6 km across the northeast corner of the Royal Gardens subdivision. This lobe, however, was almost entirely confined to the evacuated aa channel of an episode 2 flow, and so it caused no additional damage to property. Availability of additional assistance during episode 16 gave us the opportunity to monitor the front of the main flow in the middle distance between Puu 00 and the steeper slopes of Kilauea’s south flank at Royal Gardens. In general, the flow front remained relatively fluid and thin (generally less than 4—5 m thick) during this part of its passage. Advance rates measured over short time intervals indicated a typical range in advance rate from 55 to 510 m/h. The faster rates coincided with temporary surges in the flow; these surges were generally accom- panied by thinning of the flow front and commonly by thin (1—2 m thick) breakouts of more fluid lava from the in- terior of the flow. These breakouts would advance for several minutes; briefly, one advanced at a rate of 1 m/s. Fountaining stopped temporarily at 2228 H.s.t. March 4. After a brief renewal of activity, it halted finally at 2231 H.s.t. For the next 10 to 15 minutes, small, diffuse bursts of spatter issued from the open pipe in the crater floor. Burning gases were also seen flaring at the top of the pipe. The main flow, one of the longer in the series, extended nearly 8 km from the vent. It was nearly all aa, although pahoehoe overflows mantled the levees locally within 2 km of the vent. The main flow constituted about 60 per- cent of the total erupted volume, which was 12 x 106 m3, and thick, spatter-fed flows accounted for nearly 40 percent. Altogether, episode 16 lasted 32 hours; the average lava-discharge rate was 380,000 m3/h. On March 8, ground crews entered Puu 00 to find that the bowl-shaped crater interior had the same general form as before. Its appearance differed, however, because coarse rubble from superficial collapse of the walls mantled the walls and floor; broad, delicate pahoehoe sur- faces like those that characterized the crater interior after episodes 14 and 15 were absent. No lava or incandescence was sighted in the conduit until March 20; a significant layer of rubble may have capped the magma column. EPISODE 17 (MARCH 30-31, 1984) During the 25-day repose period between episodes 16 and 17, lava was first sighted in the open pipe at Puu 00 on March 20 at a depth of about 60 m below the level of the spillway. The magma column was largely crusted, and churning lava was visible through a 3- to 4-m-diameter hole in the crust. Intermittent observations for the next 10 days indicated that the column rose slowly, and some gas-piston activity occurred. The major outbreak of Mauna Loa that began early on March 25 had no apparent effect on the activity at Puu 00. The rising lava filled the crater and began overflowing the spillway between 0330 and 0350 H.s.t. March 30, accompanied by intermittent dome fountaining, about 10 to 15 m high. In tandem, harmonic-tremor amplitude in the eruption zone increased slightly at 0330 H.s.t. After four aborted overflows, a final steady overflow of the lava pond occurred at approximately 0448 H.s.t. March 30. Simultaneously, fountaining began to increase, and by 0611 H.s.t. the fountain was visible above the rim of Puu 00 from camp D. Glow over the eruption site had been observed as early as 0515 H.s.t. from HVO, 20 km away. For the first time in 65 years, Mauna Loa and Kilauea were in simultaneous eruption. We arrived at the eruption site at about 1000 H.s.t. March 30. By that time, the fountain centered over the crater was 100 to 140 m high (fig. 1.89). It consisted of a broad complex of short-lived jets of spatter, produced at a rate of about 20 per minute, which created a general impression of rapid pulsation in the overall fountain height. Strong trade winds blew most of the tephra to the southwest and deposited it on the elongate southwest flank of the cone. After a slow initial increase in fountain height during the first 7 hours of the eruption, the height increased abruptly to about 150 m at approximately 0740 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS H.s.t. and then diminished gradually to about 100 m, which was maintained until the episode’s end (see fig. 1.24). When we first saw it at 1000 H.s.t. March 30, a nar- row lava flow, fed by a voluminous torrent of lava cascading over the spillway (fig. 1.89), extended 1.5 km east-northeastward from the base of Puu 00 (pl. 5). In addition, spatter spilling over the south rim of the crater had initiated a spatter-fed flow to the southeast that was then 0.5 km long. Later in the morning, about 2 km from the vent, the main flow turned east-southeastward and began extending in the general direction of the Royal Gardens subdivision. Aa for most of its length, the flow advanced at the extraordinarily high (for Puu Oo flows) average rate of 490 m/h (table 1.3). Along individual segments between our helicopter reconnaissances, average advance rates ranged from about 300 to more than 700 m/h. This high rate may reflect confinement of the flow, for much of its advance, to the topographically confined, axial-channel part of the episode 16 flow. The stable, pahoehoe-bearing channel zone of the episode 17 flow was of normal length. When the flow was 4 to 5 km long, the transition to aa in the channel occurred about 1.5 km from Puu 00; after the episode’s end, we found pahoehoe overflow levees as far as 3.8 km from the vent. Fountaining and flow production ceased together at 0324 H.s.t. March 31, less than 23 hours after episode 17 had begun. At about that time, reports of a shooting star traveling in the direction of Mauna Loa Volcano from the middle east rift of Kilauea reached the media, reviving FIGURE 1.89.—Puu Oo erupting during episode 17. Fountain height is approximately 100 m. Torrent of lava cascades over spillway, feeding a lava river that is locally perturbed by large blocks grounded in chan- nel. Near base of Puu 00 within lava river, standing wave (arrow) several meters high may record presence of another channel obstruc- tion. A second flow, at left, originates from spatter ejected over south rim of crater. View southwestward; photograph taken at 1008 H.s.t. March 30, 1984. Numerals (lower right) indicate date and time. 91 discussion of Pele’s divided attentions during the dual eruption. The main flow extended 10.8 km from the vent; it passed northeast of Royal Gardens and came within 1.7 km of the highway near the coast. Spatter accumulating high on the south flank of Puu Oo fed small flows to the west and a 1.5-km-long flow to the southeast (pl. 5; fig. 1.90). About 10 x 106 m3 of new basalt was erupted at an average rate of about 430,000 m3/h. After episode 17, Puu Oo Crater was about the same size and shape as after episodes 14 through 16. The south wall, over which spatter had poured, feeding flows westward and eastward (fig. 1.90), remained smooth and gently sloping; the adjacent part of the rim was relative— ly low. The north and west walls were high and steep; they funneled directly into the open pipe that was now a per- manent feature of the crater floor. The walls and floor, relatively free of rubble, were largely lined with pahoehoe, as they had been after episodes 13 through 15. New air- fall-tephra deposits mantled the outer flanks of the cone on the south and southwest. EPISODE 18 (APRIL 18-21, 1984) After episode 17, we first detected lava deep in the open pipe on April 5. For the next week, the lava level fluc- tuated between depths of about 10 and 50 m in the pipe, and we were aware of some gas-piston activity. From April 12 to 17 , the lava stabilized at shallower depths of 10 to 25 m down the pipe, and intermittent gas-piston activity continued. Time-lapse film recorded the first occurrence of low fountain activity within Puu 00 Crater at 1751 H.s.t. April 18. By 1800 H.s.t., lava had begun to overflow through the deep breach in the northeast rim of the crater. The intensity of both fountaining and overflow increased gradually. Over the first hour, fountain height increased to about 70 m; it then doubled abruptly at 1859 H.s.t. Thereafter, typically moist east-rift-zone weather ob- scured the camera’s view, but a few clear frames record a very narrow, high fountain reaching more than 200 m above the level of the spillway (see fig. 1.24). At times, this vigorous high fountain fluctuated wildly in inclina- tion and trajectory, showering large sections of the cone with spatter. A heavy tephra fall during the night deposited 10 to 20 cm of lapilli and small bombs in the vicinity of Puu Kamoamoa and camp D, and Pele’s hair fell 20 km west of the vent at HVO. At about 0240 H.s.t. April 19, the fountain height decreased dramatically and, for the rest of the eruption, fluctuated irregularly between about 30 and 150 m. During periods of higher fountain- ing, tephra fallout from the single high, columnar foun- tain was significant, especially on the west and north sides of the vent. Higher fountaining was characterized by a single dominant jet, although at times the fountain split 92 into two divergent jets. At other times, adjacent to the main jet, a lower, less energetic fountain played. During periods of lower fountaining, a more complex assemblage of multiple jets was observed (see fig. 1.26). We arrived at the eruption site at 0700 H.s.t. April 19 to find that our aluminum shelter at camp D had been ripped apart and scattered over several hundred meters downwind. We had often noticed dust devils over the hot tephra accumulating on the flanks of Puu Oo; apparently THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 one had traveled through camp D during the night, tear- ing our metal shed to pieces. During the ensuing day, several more dust devils passed through camp. We subse- quently moved to camp E, on the 1123 cone. Episode 18 produced four long, river-fed aa flows (pl. 5); short, thick spatter-fed flows also advanced northward and westward from the vent. The river-fed flows were all fed by branches of a wide, sinuous pahoehoe river that issued from the spillway in the northeast rim of Puu Oo 100 FIGURE 1.90.—Puu Oo cone (c) and nearby parts of episode 17 lava flows seen through hole in crusted surface of magma column, is visible 200 METERS (17m, main, river-fed flow; 17 s, spatter-fed flows). Glowing lava (arrow), within shadowed, 20-m-diameter open pipe. Thick air-fall tephra, locally overlain by spatter-fed flows, mantles southern part of cone. Solid line, boundaries of cone and flows; dashed where approximate. Photograph 84.4.13JG120D#6 by JD. Griggs, taken April 13, 1984. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 93 Crater. When we arrived at 0700 H.s.t. April 19, a massive spatter-fed flow, apparently fed by high fountains during the night, was slowly extending northward near- ly a kilometer from the vent. The lava flux through the spillway seemed high relative to our recollections of most previous eruptions. The river supplied three major flows at that time, and two of them, extending rapidly north- eastward and eastward from the vent, were entirely pahoehoe. The third, a small flow that extended south- eastward (south flow, pl. 5), advanced slowly and re- mained minor for the entire day. Observations on the ground suggested that supply to this south flow varied; the level of pahoehoe flowing in its central channel fluc- tuated appreciably during the day, decreasing overall. Within the first kilometer of Puu 00, overflows from the lava-river system were voluminous and frequent, produc- ing broad pahoehoe sheets that surrounded the main flow channels. During the night of April 19—20, a thick flow, fed by spatter from high inclined fountains, moved toward camp D (pl. 5). By the morning of April 20, some significant changes had occurred in the lava-river system and its flows. The northeastern flow had been beheaded and was stagnant. The southern flow also was nearly stagnant; the meager amount of lava that continued to feed it had begun to pond and spread laterally within 2 km of the vent. The eastern flow, which had veered southeastward after pass- ing to the north of camp E on the 1123 cone, was still vigorously supplied and was advancing steadily adjacent to the episode 17 flow, more than 10 km from the vent. It was completely contained, however, in its upper reaches, below the level of its wide pahoehoe levees. Our visual impression of the spillway and nearby part of the lava river was that the total output from Puu 00, though still vigorous, appeared to have decreased from the first day. At the time of our last reconnaissance flight for April 20, at about 1720 H.s.t., the eastern flow extended near- ly 12.9 km from the vent and was threatening several houses east of the Royal Gardens subdivision (pls. 1, 5). A minor pahoehoe overflow had diverged from the main channel about 0.5 km from Puu 00; the resulting pahoehoe lobe extended about 0.5 km from the main flow. After our reconnaissance, this minor overflow apparently beheaded the main flow, taking virtually all of the lava supply (table 1.3) and developing into a rapidly moving flow (south- eastern flow, pl. 5) that became the fourth major flow of episode 18. Traveling primarily over earlier 1983—84 flows, this new flow reached Royal Gardens shortly before 0600 H.s.t. April 21. It advanced at a relatively high average rate (430 m/h) during the night, covering 5.4 km in 12.5 hours. Though thin (approx 1.5 m at its edge) and apparently fluid (trees were left standing after its pas- sage, not knocked down as in normal aa flows), the nar- row flow was entirely aa when inspected in the morning. It reached 0.7 km into the subdivision adjacent to the southwest edge of the episode 3 flow; fortunately, it destroyed no homes. Observation of the eastern flow as it entered a sparse- ly developed rural area east of Royal Gardens showed that its advance slowed significantly (pl. 5) in response to its being beheaded near the source. Nevertheless, it still managed to overrun two houses, two vehicles, and some outbuildings before the eruption stopped at 0533 H.s.t. April 21. The flow front, ranging in thickness from 3 to 10 m when active, continued to creep forward slowly for many hours after the eruption stopped, destroying a third house during the afternoon of April 21. The eastern flow, the longest thus far of the entire series of eruptive episodes, was predominantly aa. It was 13.2 km long and came within about 1 km of the ocean and 0.6 km of the coast highway (pl. 5). Pahoehoe overflow levees intermittently bracketed the evacuated central channel for the first 3.5 km of its length. The northeastern and southern flows were also primarily aa, with pahoehoe levees in their near-vent channel areas. The surprise southeastern flow that invaded Royal Gardens had broad pahoehoe margins along its first 1.7 km. After major lava production ceased at 0533 H.s.t. April 21, the time-lapse film record shows that minor bursts of spattering continued at Puu 00 until 0628 H.s.t. At 0830 H.s.t., a crusted lava surface was visible near the top of the pipe, and small amounts of spatter issued from two holes in the crust. Within less than an hour, however, debris collapsing from the oversteepened interior walls of the crater had plugged the pipe opening. It reopened by April 23; apparently, the debris that covered the open- ing collapsed into the pipe. Numerous small, high- frequency seismic events resembling rockfall signatures were recorded on a seismometer near Puu 00 on April 22, including a brief flurry of about 135 events between 0520 and 0545 H.s.t. This seismic activity may have recorded collapse and clearing of the rubble that blocked the magma-filled pipe on April 21. The crater shape established during episode 14 still per- sisted; however, the walls and rim had undergone some superficial collapse, as they had after episode 16. Thus, the floor and walls were a jumble of slumped blocks and finer rubble (see fig. 1.19A). The south crater wall was no longer gently sloping, and entrance into the crater after episode 18 was possible only through the spillway chute, which was a deep, locally overhanging corridor lined with glassy pahoehoe and mantled by debris from posteruption collapse (fig. 1.91). It ran steeply down from the low northeast lip of the crater and abruptly flattened as it emptied into the evacuated channel. Episode 18 lasted 60 hours; it was the longest since episode 5 in early July 1983. Except for episode 3, episode 94 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 18 was the most voluminous: It produced 24 x 106 m3 of new basalt at an average lava-discharge rate of 410,000 m3/h. EPISODE 19 (MAY 16—18, 1984) After episode 18 and continuing through the first half of May, lava at the top of the magma column was inter- mittently visible deep in the Puu 00 pipe. Fume often obscured the View. Bursts of gas and spatter, as well as repeated distinctive, low seismic bursts, indicated gas- piston activity. Beginning at about 0115 H.s.t. May 16, lava apparent- ly rose high enough that intermittent periods of glow and minor spattering within the crater were visible from camp E. At about 0500 H.s.t., the first of a series of short-lived pahoehoe overflows was recorded. For the next 44 hours, overflows lasting from 3 to 30 minutes recurred at ir- regular intervals ranging in length from 4 minutes to several hours. In general, these overflows carried lava at estimated rates of 103 to 105 m3/h. Commonly the rate gradually increased during an individual occurrence. Each overflow was accompanied by coincident higher tremor. The sequence of episode 19 eruptive events is summar- ized in figure 1.92. During periods of overflow, a quiet, partly crusted lava pond filled the crater, and an intermittent low dome foun- tain played over the site of the conduit. The pond would slowly rise until it was high enough to overflow the spillway. At the end of an overflow, the pond would FIGURE 1.91.—Spillway chute at Puu 00 after episode 18. Note person (circle), 1.5 m tall, standing in middle ground for scale. Steep, locally overhanging walls are lined with glassy pahoehoe to a height of about 2 to 4 m, presumably reflecting depth of last flowing lava. Above that, walls are mantled by spatter, some of which has fallen into evacuated chute. View southwestward; photograph taken at 1534 H.s.t. April 27, 1984. Numerals (lower right) indicate date and time. abruptly drop in level, commonly draining completely back into the upper part of the pipe. Such repeated effusive events created a stack of short pahoehoe flows extending eastward from Puu 00. Four separate times, the eruptive activity developed in- to brief occurrences of high fountaining (fig. 1.92), vigorous flow production, and high tremor, similar to the style more typical of Puu Oo eruptions thus far. During these intervals, a fountain as high as 100 m or more (see fig. 1.24) played in the crater, and estimated output over the spillway was 100,000 to 200,000 m3/h. Commonly, the fountain height would fluctuate over a period of 40 to 60 seconds from heights of less than 30 or 40 m to 100 m or more. As in episode 13, output of lava over the spillway, changing rapidly by possibly a factor of 2, fluctuated in concert with the changes in fountain height. In addition, brief interruptions in lava discharge and fountaining oc- curred during longer periods of high fountaining; these interruptions, specifically recorded only during the first high-fountain occurrence, lasted from a few seconds to 4 minutes. Resumption of discharge after each interrup- tion was abrupt, and flow production and fountaining returned to full vigor almost instantaneously. A new flow front would surge rapidly down the channel that had drained moments before. During the high-fountain occur- rences, lava coursed down the channel near the vent and then spread eastward, within 1 to 1.5 km of Puu 00, as broad overlapping sheets, mainly of pahoehoe (pl. 5). Elongate lobes extended 1 to 2 km along the evacuated channels of the eastern and southeastern flows of episode 18. The fourth and last high-fountain event ended early on May 17 (fig. 1.92). Many low-level lava-discharge events occurred during the ensuing 24 hours. A tube system developed close to the vent, and the flows that issued from it commonly formed thin, fluid pahoehoe sheets that spread rapidly; one of them advanced approximately 1 km in 36 minutes, for an average velocity of 1.6 km/h. In sampling and measuring temperatures in this overflow sheet, we were struck by the dense, degassed nature of the pahoehoe. Within several tens of minutes after its emplacement, we could walk on the pahoehoe toes or on the sheet itself without breaking through shelly spots. Repeated effusion of this type constructed an apron of smooth, firm pahoehoe extending about 1 km eastward of Puu 00. During the final hours of episode 19, on the night of May 17—18, overflows and periods of recognizable fountain activity in the vent were smaller in magnitude and more uniform in duration than previously. In fact, the style evolved to one resembling gas-piston activity; the lava column occasionally reached high enough during the rise part of a cycle to produce a short-lived overflow. During nearly 3 hours of careful timing and observation from 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY 0F ERUPTIVE EVENTS 95 camp E between 1930 and 2215 H.s.t., the pattern of individual events tended to shorten over time), sending activity consisted of the following cycle: Increased glow short pahoehoe flows within several hundred meters of and spattering over the conduit signaled the rise of the the vent; drainback of the column would be initiated by lava column; overflow (when the column rose sufficiently vigorous, diffuse bursts of spatter fragments and rapid- high), accompanied by low spattering and dome fountain- ly weakening glow; lava would then remain out of View ing, would last about 2.5 to 7 minutes (the durations of for periods of 5 to 14 minutes; and then the cycle would 0200 0400 0600 0800 1000 1200 1400 1400 1600 1800 2000 2200 2400 0200 I I I I l l I I I I I I I ‘ 1 II 0200 0400 0600 0800 1000 1200 1400 l l I I I I I l I l I I||l|||| . IIIIIIII I I . I IIIIIIIIII 1400 1600 1800 2000 2200 2400 0200 TIME, H.S.T. MAY 16-18, 1984 FIGURE 1.92.—Observed periods of eruptive activity during episode 19. Full-height bars, periods of high fountains and high lava output; half- height bars, periods of low-level activity, including dome fountaining and overflow. Dashed bars, individual film frames of strong glow only that may or may not reflect actual overflows from the crater. Data are from onsite observations, when available, and from time-lapse camera records (of lesser quality in terms of time control and completeness) otherwise; queried periods are times of questionable records. 96 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 begin anew. Episodic bursts of low tremor seen in the seismic record support the interpretation of this type of behavior as reflective of gas pistoning. At about 0030 H.s.t., tremor bursts associated with cyclic rise and fall of the lava column diminished and became highly irreg- ular. The final overflow, marking the last gasp of episode 19, occurred at about 0049 H.s.t. May 18. A helicopter overflight on the morning of May 18 reveal- ed the surface of an active magma column about 30 m down the pipe. This column persisted at shallow levels in the pipe until episode 20. Although fountains during episode 19 were never vigorous or sustained enough to significantly affect the exterior of Puu 00, this new style of generally slow discharge with prolonged rise and fall of a body of lava within the crater drastically modified the shape of the crater (see fig. 1.19). The deep, bowl- to funnel-shaped interior was largely filled by new basalt. Its new raised crater floor was smooth and sloped gently inward toward the open conduit (see fig. 1.19B). Episode 19 lava flows were thin where we could meas- ure them. They covered an area of about 1.4 x 106 m2 and had an estimated volume of about 2 x 106 m3—a minimum value because we may have underestimated the aggregate thickness of episode 19 lava flows near the vent. However, this value agrees reasonably well with the volume we would estimate from our observation of lava discharge during the episode. EPISODE 20 (JUNE 7—8, 1984) An active magma column remained at shallow depths in the open conduit for the entire repose period following episode 19. For the first 10 days, most overflights and ground checks revealed a nearly continuous crust stret- ching across the conduit at a depth of about 15 m. A 3— to 4-m-Wide opening emitted intermittent spatter, Pele’s hair, and bursts of fume and burning gas as gas-piston activity proceeded. On May 29, we thought that lava-pond activity like that of episode 19 might be recurring. Ac- companied by increased harmonic-tremor amplitude in the eruptive zone and slight deflation of Kilauea’s summit, gas-piston activity stopped, and the magma column ap- parently assimilated any existing crust and rose to within 2 m of the top of the conduit. Its surface was open and actively roiled. Occasionally, it overflowed to produce short pahoehoe flows on the adjacent crater floor, and it built a low rampart of spatter around the pipe opening. On May 30, however, the column withdrew a short distance back into the pipe, and the harmonic tremor decayed into the pattern of cyclic bursts indicative of gas- piston activity. Fairly continuous gas-piston activity at shallow levels in the pipe continued through the rest of the repose period. Episode 20 occurred almost entirely at night, and the only direct record of the eruptive activity is on time-lapse film. Continuous, low-level spattering began at 1911 H.s.t. June 7; at 2016 H.s.t., sporadic fountaining, 10 to 30 m high, became visible and slowly increased in vigor. Lava first overflowed the northeast rim and cascaded down the steep spillway chute at 2104 H.s.t.; the fountain was still low, about 15 m. At 2200 H.s.t., with dramatic sudden- ness, the level of output increased markedly; the fountain height increased by a factor of 20 to 30, to more than 300 m, flooding the north and northeast sides of the cone with heavy, fountain-fed spatter and pond overflow. A high but steadily decaying (see fig. 1.24), broad fountain and strong flow production continued uninterrupted until 0624 H.s.t. June 8, minutes after the first HVO observers arrived on the scene. The dying fountain sputtered back on at 0624:30 H.s.t. for 2.5 minutes before ceasing entirely. At 0925 H.s.t., 3 hours after the end of the erup- tion, several small bursts of spatter occurred. A helicopter overflight at 1100 H.s.t. revealed an open lava surface about 30 m down the open pipe; intermittent puffs of fume and bursts of tremor indicated that gas-piston activity had resumed. Episode 20 lasted 9 hours and produced four distinct flows that traveled northwestward and northeastward from the Puu Oo cone (pl. 5). The northwestern flow, of spatter-fed aa, was broad and extended about a kilometer from Puu 00. The three flows to the northeast were river fed. The two longest flows extended to the northeast as narrow lobes consisting of near-vent pahoehoe and distal aa. The longest of these northeastern flows traveled 3.8 km; its evacuated channel was floored with pahoehoe for much of its length. A smaller, bilobed flow to the northeast on top of episode 19 basalt was relatively narrow and predominantly aa. The erupted volume, distributed over an area of 1.6 x 106 m2, was 4 x 106 m3. Although episode 20 was brief and its volume small in comparison with most of its predecessors (table 1.3), it had the greatest average lava-discharge rate, 480,000 m3/h. Episode 20 raised the level of the crater floor about 5 to 10 m relative to the north rim, which had not changed significantly in elevation since episode 18 (see fig. 1.10A). The north rim was about 10 m above the level of the spillway, which was now a broad low in the northeast rim rather than a deep cleft (fig. 1.93). The west and south rims grew upward appreciably during episode 20 (fig. 1.10; compare figs. 1.19B and 1.93). The west rim, which formed the high point, was about 30 m higher than the floor at the spillway; it was about 130 m above the pre-1983 ground surface. The floor still formed a shallow basin; the top of the 20-m-diameter open pipe was about 10 m lower than the spillway surface. Episode 20 added more air-fall tephra to the southwest sector of the Puu Oo cone. 1. GEOLOGIC OBSERVATIONS AND CHRONOLOGY OF ERUPTIVE EVENTS 97 FIGURE 1.93.—Puu 00 after episode 20. Spillway, mantled with pahoe- hoe, is a broad low in northeast rim of crater. Top of 20-m—diameter open pipe is about 10 m lower than spillway surface. Steep west (dis- tant) and south (left) walls grew upward appreciably during episode 20 and partly collapsed afterward, spilling talus onto smooth pahoehoe crater floor. High point on west rim is about 30 m higher than level of spillway and 130 m above pre-1983 ground surface. Spatter falling on north flank of Puu 00 (right) fed a massive flow in that sector. Air-fall tephra deposit on southwest flank is partly visible beyond rim and steep southeast flank of cone. View southwestward; photograph by GE. Ulrich, taken June 14, 1984. Numerals (lower right) indicate year, month, and day. REFERENCES Buchanan-Banks, J .M., 1987, Structural damage and ground failures from the November 16, 1983, Kaoiki earthquake, Island of Hawaii, chap. 44 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: US. Geological Survey Professional Paper 1350, v. 2, p. 1187—1220. Holcomb, R.T., 1980, Preliminary geologic map of Kilauea volcano, Hawaii: US. Geological Survey Open-File Report 80-796, scale 1:50,000, 2 sheets. Jackson, D.B., Swanson, D.A., Koyanagi, R.Y., and Wright, T.L., 1975, The August and October 1968 east rift eruptions of Kilauea Volcano, Hawaii: US. Geological Survey Professional Paper 890, 33 p. Koyanagi, R.Y., Endo, E.T., Tanigawa, W.R., Nakata, J .S., Tomori, A.H., and Tamura, RN, 1984, Kaoiki, Hawaii earthquake of November 16, 1983: A preliminary compilation of seismographic data at the Hawaiian Volcano Observatory: US. Geological Survey Open- File Report 84-798, 35 p. Lipman, P.W., and Banks, N.G., 1987, Aa flow dynamics, Mauna Loa 1984, chap. 57 ofDecker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: US. Geological Survey Professional Paper 1350, v. 2, p. 1527—1567. Lockwood, J .P., Banks, N.G., English, T.T., Greenland, L.P., Jackson, D.B., Johnson, D.J., Koyanagi, R.Y., McGee, K.A., Okamura, A.T., and Rhodes, J .M., 1985, The 1984 eruption of Mauna Loa Volcano, Hawaii: Eos (American Geophysical Union Transactions), v. 66, no. 16, p. 169—171. Malin, MG, 1980, Lengths of Hawaiian lava flows: Geology, v. 8, no. 7, p. 306—308. Moore, J .G., and Koyanagi, R.Y., 1969, The October 1963 eruption of Kilauea Volcano, Hawaii: US. Geological Survey Professional Paper 614-0, p. 01—013. Moore, R.B., Helz, R.T., Dzurisin, Daniel, Eaton, G.P., Koyanagi, R.Y., Lipman, P.W., Lockwood, J .P., and Puniwai, GS, 1980, The 1977 eruption of Kilauea volcano, Hawaii: Journal of Volcanology and Geothermal Research, v. 7, no. 3-4, p. 189—210. Neal, C.A., and Decker, R.W., 1983, Surging of lava flows at Kilauea volcano, Hawaii [abs]: Eos (American Geophysical Union Trans- actions), v. 64, no. 5, p. 904. Peterson, D.W., 1976, Processes of volcanic island growth, Kilauea volcano, Hawaii, 1969—1973: International Association of Volcanol- ogy and Chemistry of the Earth’s Interior, Symposium on Andean and Antarctic Volcanology Problems, Santiago, Chile, 1974, Pro- ceedings, p. 172—189. Peterson, D.W., and Swanson, D.A., 1974, Observed formation of lava tubes during 197 0—7 1 at Kilauea volcano, Hawaii: Studies in Speleology, v. 2, p. 209—222. Richter, D.H., Ault, W.U., Eaton, JP, and Moore, J .G., 1964, The 1961 eruption of Kilauea Volcano, Hawaii: US. Geological Survey Pro- fessional Paper 474-D, p. D1—D34. Shaw, H.R., 1969, Rheology of basalt in the melting range: Journal of Petrology, v. 10, no. 3, p. 510—535. Shaw, H.R., Peck, D.L., Wright, T.L., and Okamura, R., 1968, The viscosity of basaltic magma: An analysis of field measurements in Makaopuhi lava lake, Hawaii: American Journal of Science, v. 266, no. 4, p. 225—264. Sparks, R.S.J., Pinkerton, H., and Hulme, G., 1976, Classification and formation of lava levees on Mount Etna, Sicily: Geology, v. 4, no. 5, p. 269—271. Swanson, D.A., 1972, Magma supply rate at Kilauea volcano, 1952-1971: Science, v. 175, no. 4018, p. 169—170. 1973, Pahoehoe flows from the 1969—71 Mauna Ulu eruption, Kilauea Volcano, Hawaii: Geological Society of America Bulletin, v. 84, no. 2, p. 615—626. Swanson, D.A., Duffield, W.A., Jackson, D.B., and Peterson, D.W., 1979, Chronological narrative of the 1969—71 Mauna Ulu eruption of Kilauea Volcano, Hawaii: US. Geological Survey Professional Paper 1056, 55 p. Tilling, R.I., Christiansen, R.L., Duffield, W.A., Endo, E.T., Holcomb, R.T., Koyanagi, R.Y., Peterson, D.W., and Unger, J .D., 1987, The 1972—74 Mauna Ulu eruption, Kilauea Volcano: An example of quasi- steady—state magma transfer, chap. 16 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: US. Geological Survey Professional Paper 1350, v. 1, p. 405—469. Wadge, G., 1978, Effusion rate and the shape of aa lava flow—fields on Mount Etna: Geology, v. 6, no. 8, p. 503—506. Walker, G.P.L, 1973, Lengths of lava flows, in Guest, J .E., and Skelhorn, R.R., eds., Mount Etna and the 1971 eruption: Royal Society of London Philosophical Transactions, ser. A., v. 274, no. 1238, p. 107—118. Wolfe, E.W., Garcia, M.O., Jackson, D.B., Koyanagi, R.Y., Neal, C.A., and Okamura, A.T., 1987, The Puu Oo eruption of Kilauea Volcano, episodes 1—20, January 3, 1983 to June 8, 1984, chap. 17 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: US. Geological Survey Professional Paper 1350, v. 1, p. 471—508. Wright, T.L., Kinoshita, WT, and Peck, D.L., 1968, March 1965 erup- tion of Kilauea volcano and the formation of Makaopuhi lava lake: Journal of Geophysical Research, v. 73, no. 10, p. 3181—3205. W , , ,, V , w, mil/”TV ,; _ , . ,_ _ f ‘ 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS By CHRISTINA A. NEAL, TONI J. DUGGAN, EDWARD W. WOLFE, and ELAINE L. BRANDT CONTENTS Page Abstract ___________________________ 99 Introduction _________________________ 99 Methods ___________________________ 99 Lava temperatures ____________________ 99 Sampling _________________________ 102 Chemical analyses _______________________ 102 Results __________________________ 102 Explanation of tables _____________________ 102 Vent nomenclature ____________________ 103 References cited _______________________ 103 ABSTRACT A total of 173 lava temperatures were measured and 346 samples were collected during the first 20 episodes of the Puu Oo eruption of Kilauea. Of these samples, 43 were selected for classical chemical analysis. This chapter lists these data and briefly discusses field techniques and results. During the first 11 eruptive episodes, equilibrium temperatures in pahoehoe increased steadily from a range of 1,098—1,125 °C during episode 1 to 1,133—1,144 °C during episode 11. Subsequent measured lava temperatures through episode 20 remained fairly steady within the range of 1,129—1,147 °C. Chemical analyses show a general corre- spondence between progressively changing composition and measured temperatures. INTRODUCTION The Puu Oo eruption of Kilauea Volcano’s middle east rift zone began on January 3, 1983. By June 8, 1984, 20 episodes in the continuing eruption had produced approx- imately 240 x 106 m3 of basalt. Though strikingly similar in style to the first stage of the Mauna Ulu eruption (Swanson and others, 1979), the Puu Oo eruption is unique in Kilauea’s historical record because of the duration of its episodic series, the nature of the resulting lava field, and the size of the principal vent structure. One impor- tant part of the monitoring effort by the Hawaiian Volcano Observatory has been to systematically measure lava temperatures and to collect samples of successive lava flows and other eruptive products. In addition, the major-element composition of samples representing each episode has been determined by classical wet chemical methods (Peck, 1964). This chapter presents a compila- tion of these data sets along with pertinent supporting information. Its purpose is to provide the reader with: (1) a discussion of our strategy, methods, and the condi- tions of sampling and temperature measurements; (2) a careful documentation and listing of all sample, tempera- ture, and compositional data from episodes 1 through 20; and (3) a summary of the results. For more detailed discus- sion of each eruptive episode and the geologic context of specific samples and temperatures, the reader is referred to the eruption narrative and discussion of geologic obser- vations in chapter 1. The petrology of the lavas is dis- cussed in chapter 3. Acknowledgments—Many individuals have generously shared their time, energy, and enthusiasm in the Hawaiian Volcano Observatory’s monitoring program at Puu 00. We thank them all for their assistance. We especially thank Norman G. Banks for his leadership and instruc- tion in thermocouple techniques. METHODS LAVA TEMPERATURES All the lava temperatures reported in this chapter (tables 2.1, 2.2) were obtained using Inconel-clad Chromel- Alumel thermocouples. These probes worked well under the difficult conditions of flowing lava, active vents, and field transport. All episode 1 temperatures and all temperatures in aa flows were measured with 1/4-in.- diameter thermocouples. From episode 2 on, pahoehoe temperatures were measured with a 1Asia-diameter thermocouple, normally preferred because of its minimal thermal mass and, thus, rapid equilibration. Generally, the 1/16-in.-thermocouple was wound around a light steel bar for support, with the tip of the thermocouple extended 10 to 20 cm beyond the end of the supporting bar. Temperatures were read from a direct-reading digital thermometer attached to the thermocouple. Previous use of thermocouple equipment to measure temperatures of basaltic melt was discussed by, for example, Wright and others (1968), Wright and Okamura (1977), and Peck (1978). Techniques for similar temperature measurements in pyroclastic products were described by Banks and Hoblitt (1982). For a more detailed description of specific thermocouple hardware, see the report by Lipman and Banks (1987). After several measurements were made with the Inconel sheath intact, thermal and mechanical stress com- monly rendered the thermocouple useless. However, after we discarded 3 to 6 cm of the faulty end and exposed the internal wires with a pair of strippers, the wires could be 99 100 TABLE 2.1.—Summary of equilibrium temperatures measured during the vigorous parts of episodes 1 through 20 of the Puu Oo eruption of Kilauea Volcano [Temperatures measured in melt during low-level activ- ity between eruptive episodes and in cool, spatter- fed flows are omitted. Two equilibrium temperatures measured in blocky aa 6 km from the vent during epi- sode 3, both lower than 1,000 °C, are also omitted] Number of Average Range of . . . . measured measured Episode equilibrium temperatures temperature temperatures (°C) (°C) 1 8 1,117 1,098—1,125 2 18 1,115 l,111-1,120 3 28 1,120 1,111-1,129 4 9 1,125 1,115-l,l32 5 4 1,127 1,125-1,129 6 4 1,132 1,126-l,l38 7 10 1,135 1,130-1,l41 18 2 1,129 1,128-1,l30 9 ___ --_ ___ lO 6 1,138 1,134-1,142 ll 5 1,140 l,133-1,l44 12 3 1,136 l,l35-1,l37 13 12 1,139 1,129—1,147 l4 2 1,137 l,l36-l,l37 15 2 1,138 1,136-l,l39 l6 8 1,139 l,l35—l,l42 l7 4 1,133 1,131-1,l37 18 6 1,139 1,136-l,l44 l9 2 1,140 l,l38-l,l4l 120 --- ——- -—- 1No equilibrium melt temperatures were obtained during episodes 9 and 20. twisted together by hand and immersed in melt with a high degree of success. This vastly lengthened the life of a single thermocouple wire and made repeated measure- ments of temperatures possible. Also, bare-wire thermo- couples generally equilibrated more quickly than sheathed thermocouples, a trait preferred by most workers. Sheathed and bared thermocouple wires recorded iden- tical temperatures when tested in a furnace over a temperature range comparable to those of our field measurements. Contamination problems leading to erro- neous temperatures, such as those documented by Wright and Okamura (1977), were probably eliminated by the short immersion times (generally several minutes for the 1/4-in.-diameter thermocouple but less than 1 minute for the 1/16—in.-diameter thermocouple). Temperature measurements were most successful in pahoehoe melt that was slow moving and well insulated from exposure to the atmosphere. Thus, ideal situations included oozes through cracks in accretionary levees form- ing at the edges of lava rivers and short, tube-fed toes THE PUU 00 ERUPTION OF KlLAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 at pahoehoe flow fronts or margins. These conditions con- sistently resulted in the highest temperatures obtained in episodes 1 through 20, and presumably most closely approximate true eruption temperatures. A variation of about 5 °C in the temperatures measured in pahoehoe toes at the same locations over periods of several hours prob- ably reflects heterogeneities in the temperatures of melt erupted; no obvious systematic change in subaerial con- ditions (degree of insulation or rate of transport through the evolving network of channels and tubes) can be isolated to explain this variation. However, differing rates, durations, and conditions of transport, superim- posed on possible eruptive thermal heterogeneities, may explain observed variations of 5 to 15 °C in measured pahoehoe temperatures over a period of days. Temperatures of spatter-fed flows, of thin, rapidly cool- ing overflows from lava rivers, and of rapidly moving melt in the highly sheared zone at the edges of leveed lava rivers were all consistently lower than those measured in insulated pahoehoe. For the purposes of this report, a “short-lived overflow from a lava river” is defined as a transient overbank deposit of pahoehoe, generally less than 1 m thick and mobile for periods of no more than a few minutes. “Long-lived overflows” or “breakouts” refer to major spills from the lava river that endured for tens of minutes to several hours. Most of these flows were active long enough to develop into a widespread, locally tube—fed pahoehoe flow. Three temperatures were measured inside vents that were actively degassing and incandescent. Low-level, repose-period eruptive activity that produced short pahoehoe flows or active lava ponds inside the Puu Oo vent was also the source of some of the temperature measurements listed in tables 2.1 and 2.2. In nearly all these situations, the technique for obtain- ing temperatures of melt was similar. After several seconds of preheating by holding the thermocouple adja- cent to exposed melt (several minutes for the 1/4—in.- diameter thermocouples), 3 to 10 cm of the wire was immersed in melt (fig. 2.1) and worked back and forth slowly to maintain fluid-wire contact and to prevent a sheath of solidified lava from forming on the wire. The larger thermocouple was particularly prone to accumulate such a sheath, which would insulate the thermocouple and prevent it from reaching equilibrium with the melt dur- ing the measurement. It was generally easiest to have a team of two people working together to obtain a temper- ature measurement. A thin shield of corrugated aluminum (approximately 3 by 4 ft, with two nylon-rope loops for handles; see fig. 2.1) was effective in reflecting enough heat to protect workers and permit a close approach to the molten lava. One person would immerse and man- euver the thermocouple, attempting to avoid hitting any developing crust that might damage or break the wire, 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS while the other person would read aloud the increasing temperature values. After many minutes with the large thermocouple but commonly within tens of seconds after immersion for the thin one, the thermometer would settle on a single reading or fluctuate within 1 °C in either direc- tion from a single temperature value. The temperature recorded was the value at which the thermometer appeared to stabilize, defined in this chapter as the equi- librium temperature. In some cases, the thermocouple wire would break, or conditions around the probe would deteriorate, and stable temperatures were not obtained. Here, recorded temperatures are considered none- quilibrium, minimum values, denoted with the 2 symbol in table 2.2. Temperatures of aa flows were measured with the more rigid 1/4-in.-diarneter thermocouple. For these measure- ments, the probe was preheated, again by holding the 101 thermocouple inches from the flow surface to lessen the thermal contrast, and inserted into the incandescent matrix between the cooler aa fragments. The thermo— couple was then held in place to be gradually engulfed by the advancing flow. In some cases, at penetrations rang- ing from 0.2—0.3 to 2—3 m, the probe apparently broke through the aa carapace into more continuous melt. Dur- ing episode 3, temperatures of 1,094 and 1,099 °C were determined in an aa flow 6 km from the vent by this means. Those temperatures were 10 to 30 °C lower than those measured within 1 km of the vent earlier in episode 3. During episode 16, temperatures were measured in short-lived breakouts from aa flows. Here, the lava was commonly fluid enough to allow us to use the 1/16-in.- diameter thermocouple; at times, the thermocouple was worked successfully through a relatively thin aa carapace FIGURE 2.1.—Geologists measure temperature of pahoehoe toe about 250 in northeast of Puu Oo. Toe is fed by sluggish pahoehoe flow that branched from main lava river. Flexible, Vle-insdiameter thermocouple is supported by thin but rigid steel rod. Aluminum shield protects workers from radiant heat. Photograph by J .D. Griggs, taken approximately 1200 m. H.s.t. January 31, 1984. 102 into the more plastic flow interior. Temperatures meas- ured under these conditions ranged from 1,135 to 1,138 °C (table 2.2), even at distances as far as 6 to 7 km from the vent. These high temperatures were only 4 to 7 °C lower than the pahoehoe temperatures measured within 2 km of Puu 00 during the same eruptive episode. SAMPLING Samples were collected as soon as possible after temperatures were measured, so as to closely correlate the data in space and time. Most samples were taken by using a rock hammer. Pahoehoe could easily be scooped with the hammer, placed in an empty coffee can or similar container, and quenched with water. Water served the dual purpose of minimizing groundmass crystallization and cooling the glassy sample so that it could be detached from the hammer and handled. Once in the can, the water was allowed to cool the sample and then poured off; the sample was bagged in a heavy canvas sample bag and labeled. At times, a bent steel bar or steel pipe was used to sample melt too difficult to approach within arm’s reach. Aa flows were sampled similarly, although samples were not strictly melt but rather fragments hacked loose with a hammer. Early in the eruption, spatter bombs from low-fountaining fissure vents were collected regularly. Most bombs were collected While still molten and were promptly quenched with water. As the Puu 00 vent became established and the fountains more difficult to ap- proach, hot samples of molten spatter were collected less frequently, and the emphasis in sampling shifted to pahoehoe river and flow samples. We tried to be as comprehensive as possible in sampling melt from successive episodes. Specifically, our objective was to collect samples of each of the following: (1) low- volume, intermittent flows and spatter produced within the vent between most of the major eruptive episodes; (2) lava erupted early in each episode; (3) representative samples throughout the episode; and (4) some of the last melt to be erupted. Our success varied. For most of episode 9, broad, hot, newly solidified sheets of pahoehoe prevented safe access to the lava river, and so we had to settle for samples of air-cooled, solidified basalt. During episode 13, however, we had sufficient help to sample day and night, as well as good access to melt, and we were thus able to measure temperatures and sample overflows from the lava river approximately every 2 hours. In addition to samples of melt or spatter, relatively dense, slowly cooled samples (“dense-rock samples,” table 2.2) from lava flows of each episode were taken for petrographic study. Samples of lightweight tephra pro— duced during high—fountaining events were collected several times during episode 3 and after episodes 9, 12, THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 and 14 through 18. Systematic, repetitive sampling of tephra during episode 16 was reported elsewhere (Paces and Rose, 1984). CHEMICAL ANALYSES RESULTS A plot of lava temperatures against time (fig. 2.2) for episodes 1 through 19 reveals a simple trend. In general, between episodes 1 and 11, average equilibrium temper- atures gradually but steadily increased from 1,117 °C dur- ing episode 1 to 1,140 °C during episode 11 in November 1983 (table 2.1). Thereafter, although our total range of measured temperatures was from 1,129 to 1,147 °C, aver- age eruptive-episode temperatures of 1,133 to 1,140 °C dominated, with little variation. Compositional trends of the erupted lavas, recorded in the variation of major oxides, show a general pattern of progressive enrichment over time of CaO and MgO, with correlative depletion of Fe and incompatible elements. For CaO and NaZO + K20 (fig. 2.3), this compositional change displays a striking, qualitative similarity to the pattern of temperature changes discussed above. Thus, changes over time in lava temperature apparently reflect progres- sive compositional change of the erupting melt. The implications of this apparently sympathetic evolution of temperature and composition are discussed by Wolfe and others (1987). EXPLANATION OF TABLES Most of the information listed in table 2.2 is self- explanatory. The sample identification (for example, 1/84KE14-231F) consists of the following: month and year collected (1/84), the designation “KE”, which indicates Kilauea’s east rift zone, the number of the episode dur- ing which the sample was erupted (14), a unique identify- ing number (231), and either an “F” for flow sample or an “S” for a sample of spatter or air-fall tephra. Under- lined sample identifications flag samples for which classical chemical analyses are listed in table 2.3. In most cases, all the samples from a particular episode are sequential; exceptions are noted, and additional sample identifications cited. Samples collected during low-level activity between eruptive periods bear the number of the succeeding eruptive episode. “Quenc ” refers to whether or not the molten sample was cooled rapidly in water. “Location and comments” contains a brief description of the conditions, setting, and source of the sample or temperature. Information presented under this heading is of varying quality and comprehensiveness; we have 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS attempted here to be concise and yet as complete as the record allows. Further details are provided in chapter 1. VENT NOMENCLATURE Episode 1 eruptive vents referred to in table 2.2 were named after the time or date of their initial activation (for example, 1708, 1123, January 23). The locations of these and other episode 1 vents are shown in figure 2.4. The O vent of episode 3 was named for its proximity on the topographic map to the letter “O” in the word “flow.” Beginning with episode 4, the principle vent struc- ture was Puu O, a cinder and spatter cone that grew just uprift of the O vent and eventually buried it completely (fig. 2.4). Subsequently, elders of the Hawaiian community in Kalapana renamed the two large vent structures of the 1983—84 eruption. The 1123 vent became Puu Halulu, a 103 name which refers to the chanting sound heard in Kalapana during eruptions. Puu O became Puu 00, after an extinct native bird, the ‘o‘o, that once lived in the area. REFERENCES CITED Banks, N.G. and Hoblitt, RR, 1982, Summary of temperature studies of 1980 deposits, in Lipman, P.W., and Mullineaux, D.R., eds., The 1980 eruptions of Mount St. Helens, Washington: US. Geological Survey Professional Paper 1250, p. 295—813. Lipman, P.W., and Banks, N.G., 1987, Aa flow dynamics, Manna Lea 1984, chap. 57 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: US. Geological Survey Professional Paper 1350, v. 2, p. 1527—1567. Paces, J .B., and Rose, W.I., Jr., 1984, Hourly volatile and chemical evolu- tion during Phase 16 of the 1983—84 Kilauea east rift zone eruption [abs]: Eos (American Geophysical Union Transactions), v. 65, no. 45, p. 1130—1131. - 1150 I 1 I 1 1 I I I I 13 11 ,3 1s 11 13 16 7 1° 11 1313 16 19 1140— 13 18 — 1o 13 15 16 18 5 7 10 16 19 12 14 16 17 13 9 7 14 15 1a <7) ‘ 6 7 1o 12 16 _ .J 8 7 10 11 4 17 3 7 13 1717 $1130" 7 13 — 3 4 5 13 8 3 44 e o 3 4 5 Z _ 1 5 _ 1.1.? 1 n: 1 33 D 3 ft 33 13:1120- 2 3 4 - LU 1 2 7 E 2 33 LU 2 4 l— 2 33 < — 1 3 4 _ > 2 3 1 2 2 333 2 3 1110 — _ ”00 I l 1 I l I l 1 I —10 100 210 320 430 540 DAYS FROM BEGINNING OF 1983 FIGURE 2.2.—Equilibrium lava temperatures 21,100 °C measured during vigorous parts of episodes 1 through 19 (no equilibrium temperatures were measured during episodes 9 and 20). Numbers correspond to episodes. 104 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 Peck, D.L., 1978, Cooling and vesiculation of Alae lava lake, Hawaii: episodes 1—20, January 3, 1983 to June 8, 1984, chap. 17 of Decker, U.S. Geological Survey Professional Paper 935-B, 59 p. R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: Peck, LC, 1964, Systematic analyses of silicates: U.S. Geological Survey U.S. Geological Survey Professional Paper 1350, v. 1, p. 471-508. Bulletin 1170, 89 p. Wright, T.L., Kinoshita, W.T., and Peck, D.L., 1968, March 1965 erup- Swanson, D.A., Duffield, W.A., Jackson, DR, and Peterson, D.W., tion of Kilauea volcano and the formation of Makaopuhi lava lake: 1979, Chronologic narrative of the 1969—71 Mauna Ulu eruption of Journal of Geophysical Research, v. 73, no. 10, p. 3181—3205. Kilauea Volcano, Hawaii: U.S. Geological Survey Professional Paper Wright, T.L., and Okamura, R.T., 1977, Cooling and crystallization of 1056, 55 p. tholeiitic basalt, 1965, Makaopuhi lava lake, Hawaii: U.S. Geological Wolfe, E.W., Garcia, M.O., Jackson, D.B., Koyanagi, R.Y., Neal, C.A., Survey Professional Paper 1004, 78 p. and Okamura, A.T., 1987, The Fun 00 eruption of Kilauea Volcano, 11.5 I I I 1 1 1 I l A 1717 18 19 20 12 15 16 1314 18 13 _ 11 _ 12 10 11 5 8 1— - 1 5 7 — LIZJ 10 U a: 5 Lu 4 6 8 o. - 4 _ '— 7 I 3 5 9 g 1 9 Lu 3 10.5 - ‘ — Z T 1 3 '2 Lu 2 1— - 3 — g 2 U o 2 to U ' _ 1 9_5 I 1 I l | | I l l —10 100 210 320 430 540 DAYS FROM BEGINNING OF 1983 FIGURE 2.3.—Weight percent CaO (A) and NaQO + K20 (B) versus time for 41 samples analyzed from episodes 1 through 20. Complete wet chemical analyses are listed in table 2.3. Numbers correspond to episodes. N-aZO+K20 CONTENT, IN WEIGHT PERCENT 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 105 3.5 I l I I I I I I B 3 _. I _ 11 _ 3.0- - 6 .— 7 _ ‘I 8 10 _ g 8 1,1 _ 11 112 1314 15 7 9 10 2 16 17 9 16 17 13 19 _ 18 20 — 25 I I I I I l I 1 —'IO 100 210 320 430 540 DAYS FROM BEGINNING OF 1983 FIGURE 2.3.—Continued 106 THE PUU 00 ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 A Vents of January '———Vents of January 3, 1983 5—6, 8—15. 1983 Vents of January 7—8, 1983 ' Puu Kamoamoa Puu Kahaualea ‘ W t , E t Also January 576 Also January 23 95 em Central as ern . 1708" 740 * ““59 We ,- ,3 v -’*"‘ m ,- :. .1. ._ 05333 Napau Crater ' ' ' 31‘12’ ‘ ’ Kalalua / Also January 3 ,1 .3 0 1 KILOMETER —1 Also January 5 155°15' 155°07'30" 155000 l ! Hawaiian Volcano Observatory/ «9 Kilauea summit Kilauea lki Crater caldera Keanakakoi \ 9 Crater ‘ \‘ o Puhimau *N\_\» Halemaumau (4° Crater “'Onal Pafig\»‘£a 0' "189' A C'ate' "1° H' k “Mary »\_ ’9 (3:21am? Mauna Ulu Q Pauahi Shield Napau Crater 19°30’ — 6‘ Crater 4s R’FT ZONE :2 HAWAll VOLCANOES NATIONAL PARK Makaopuhi Crater Wahaula Visitor Center 471% Puu . Kahaualea ‘buriedl. <—— SOUTH FLANK —>- 0 5 10 KILOMETERS | Camp B Camp E (buried) 1123 vent Puu Halulu 0 l KiLOMETER (buried) 0 vent lburiedl Puu Kamoamoa Puu 00 Camp D PACIFIC OCEAN 19°15'— l FIGURE 2.4.—Locations of vents and features referred to in table 2.2. A, Vents of episode 1 (hachured line) and generalized parts of episode 1 lava flows (shaded). B, Index map of Kilauea’s summit, south flank, and upper and middle east rift zones. Selected prominent historical and prehistoric vents of the east rift zone are shown in black. Area covered by lava flows of episodes 1 through 20 is shown by lined pattern. Locations of principal vents of episodes 3 through 20 are shown in inset A: 1123 vent (Puu Halulu), 0 vent (buried), and Fun 00 (post-episode 20). TABLES 2.2, 2.3 108 THE PUU 00 ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.——Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano [Underlined sample identifications are those for which analyses are listed in table 2.3; ---, sample not collected, time not recorded, or temperature not measured; ?, collector, date, or quench history unknown; do., ditto] Time lemper- Sample Collector Date ature Quench Location and comments (H.s.t.) D (C) 1/83KE1-1F E. Wolfe --------- 01-03-83 0230 -—- yes Small, spiny pahoehoe flow from vent 1 km northeast of northeastern part of Napau Crater rim. Sample collected when solid but still incandescent. 1/83KE1-2F J. Lockwood, 01-03-83 0508 --- 7 Active flow, 20 m northwest of Puu Kamoamoa, R. Moore. approximately 120 m from vent. 1/83KE1-3S R. Moore --------- 01-03-83 0548 —-— ? Spatter from small vent on east side of Puu Kamoamoa. 1/83KE1-4S E. Wolfe --------- 01-03-83 0605 --- yes lblten spatter from vent 2.6 km northeast of northeastern part of Napau Crater rim. --- d0 ------------- 01-03-83 0630 1,113 —-- Pahoehoe from west end of erupting vent, sampled at 0605 H.s.t. (above). 1/83KE1-SS do ------------- 01-03-83 0704 —-- yes Molten spatter from vent 2.6 km northeast of northeastern part of Napau Crater rim. l/83KEl-6S J. Lockwood ------ 01-03-83 0958 ——- yes Spatter from vent on west flank of Puu Kamoa- moa. Sample collected as vent was dying. l/83KEl-7S N. Banks ————————— 01-05-83 1140 —-- no Spatter from 1123 vent. 1/83KE1-8F do ------------- 01-05-83 1240 2},108 yes Flow from 0740 vent, 30-80 m north of vent. 1/83KE1-9F d0 ------------- 01-03-83 1515 1,125 yes Flow from 0740 vent, 50 m north of vent. 1/83KE1-10F E- Wolfe, 01-05-83 1530 1,123 yes Flow from 0740 vent, 100 m north of vent. N. Banks. 1/83KE1-IIS N. Banks --------- 01-05-83 2159 --- yes Molten spatter from west end of 1708 vent. 1/83KEl-1ZS E. Wolfe --------- 01-06-83 1000 --— ? Spatter ejected from 1708 vent during vigor- ous gas emission. 1/83KE1—13F do ------------- 01—06-83 1058 —-- yes Spiny pahoehoe flow from 0740 vent, north of vent. 1/83KE1-14F do ------------- 01-06-83 1240 21,116 yes Pahoehoe flow from 0740 vent, 250 m north of vent. Lava froze to thermocouple. 1/83KE1-15F do ------------- 01—06—83 1736 --— yes Pahoehoe flow near northeast end of 0740 vent. 1/83KE1-165 N. Banks --------- 01-07-83 0930 --— yes Spatter ejected from 0740 vent during vigor~ ous gas emission. 1/83KE1—17S N. Banks, 01-07-83 1100 --- yes Molten spatter from western January 7 vent. J. Buchanan- Banks. l/83KE1-188 N. Banks, 01-07-83 1500 -—— yes Molten spatter from western part of western T. Duggan, January 7 vent. J. Buchanan- Banks. 1/83KE1-195 E. Wolfe --------- 01-07-83 1735 --— yes Molten spatter from western January 7 vent. 1/83KEl-ZOS do ------------- 01-07—83 2005 --- yes Molten spatter from western January 7 vent, from center of lOO-m-long fissure fountain. l/83KE1-215 d0 ------------ 01-07-83 2230 —-- yes Do. 1/83KE1-22F do ------------- 01-08-83 0240 1,124 yes Pahoehoe toe from west end of western January 7 vent, 25 m from vent. 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 109 TABLE 2.2.—-Samptes collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued C. Finn, J. Lockwood. Time Temper- Sample Collector Date (H s t ) ature Quench Location and comments . . . (°C) 1/83KE1-23S N. Banks ————————— 01-08-83 0918 --- yes Spatter ejected from 0740 vent during vigorous gas emission. 1/83KE1-24S do ------------- 01-08-83 1214 —-— yes Do. 1/83KE1-255 do ------------- 01-08-83 1454 —-- yes Do. l/83KE1-26S D. Jackson ------- 01—08-83 2104 --— yes Molten spatter from 1708 vent. 1/83KE1-27F do ------------- 01—08-83 2230 -—- yes Pahoehoe toe of flow from 1708 vent. l/83KE1—28F do ------------- 01-08-83 2249 --- yes Do. l/83KE1-29F do ------------- 01-09—83 0153 --— yes Do. l/83KEl-30F J. Lockwood ------ 01-10-83 0600 --- no Inactive flow, about 300 m north of 0740 vent. Flow probably came from 1708 vent early on January 9. Sample collected when cold. 1/83KE1-315 do ------------- 01-10-83 0720 --- ? Spindle bomb ejected from 1708 vent during vigorous gas emission. 1/83KE1n328 N. Banks --------- 01-10-83 1137 --- yes Molten spatter from 1708 vent. 1/83KE1-335 J. Buchanan- 01-10-83 1346 --- yes Do. Banks. 1/83KE1-34F N. Banks --------- 01-11-83 0315 1,115 yes Flow from 1708 vent, 30 m north of vent. --- do ------------- 01-11-83 0420 1,119 --- Do. 1/83KE1-35F do ------------- 01-11-83 0730 1,115 yes Slabby pahoehoe flow from 1708 vent, about 600 m northeast of vent. 1/83KE1-36F E. Wolfe --------- 01-11-83 0920 --- yes Spiny to slabby pahoehoe flow from 1708 vent, 100 m north of vent. --- do ------------- 01-11-83 1220 1,098 --— sluggish (1 m/s) flow from 1708 vent, in shallow (few tens of centimeters), 2—m—wide channel about 400 m north of vent. 1/83KE1-37F do ------------- 01-11-83 1240 --- 7 Do. 1/83KE1-38F do ------------- 01-15-83 0430 -—- yes Slabby pahoehoe flow from 1123 and 1708 vents, about 500 m north of vents. 1/83KE1-39F do ------------- 01-15-83 0706 --- yes Pahoehoe flow from west end of 1123 vent, 10 m north of vent. l/83KE1-4OF d0 ------------ 01-15-83 0840 2},110 yes Pahoehoe flow from west end of 1123 vent, 30 to 40 m north of vent. Thermocouple froze into flow, reached maximum temperature of 1,110 °C. 1/83KE1-4IS do ------------- 01-17-83 0145 ——— yes Spotter ejected from 1708 vent during Vigorous gas emission. l/83KE1-4ZS R. Moore --------- 01-17—83 1050 —-- yes Do. 1/83KE1—438 J. Judd, 01-17-83 1140 —-— yes Do. R. Moore. 1/83KEl-44S J. Judd, 01-17-83 1220 -—— yes 00. R. Moore, T. Duggan 1/83KE1-4SS do ------------- 01-17-83 1510 —-- yes Do. 1/83KE1-46S L. McBroome, 01-18-83 0425 --- yes Spatter ejected from 0740 vent during vigorous gas emission. 110 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.—Sdmples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Ti Temper- Sample Collector Date me ature Quench Location and comments (H.s.t.) o (C) 1/83KE1-47S N. Banks ————————— 01-20-83 1145 --— yes Spatter ejected from 0740 vent during vigorous gas emission. l/83KE1—4BS L. McBroome —————— 01-21-83 2000 --- no D0. 1/83KE1-49S N. Banks --------- 01-12—83 —-- --- no Spatter erupted on January 3 from Napau Crater vent. 1/83KE1-503 do ------------- 01—12—83 --- ——- no Spatter erupted on January 3 from first vent east of Napau Crater, 0.5 km northeast of northeastern part of Napau Crater rim. l/83KE1-51S J. Buchanan- 01-12-83 -—- ——- no Spatter erupted on January 3 from second vent Banks. east of Napau Crater, 0.8 km northeast of northeastern part of Napau Crater rim. l/83KEl-525 do ------------- 01-12-83 ——— --— no Spatter erupted on January 3 from vent 1.4 km northeast of northeastern part of Napau Crater rim. 1/83KE1-S3S do ------------- 01—12-83 --- —-— no Spatter erupted on January 3 from vent 1.2 km northeast of northeastern part of Napau Crater rim. 1/83KE1-54S N. Banks --------- 01-24-83 --- ——- no Spatter erupted on January 23 from January 23 vent. 1/83KE1-555 do ------------- 01-28-83 —-- ——- no Spatter erupted on January 5—6 from western- most January 5 vent, approximately 1.5 km northeast of northeastern part of Napau Crater rim. 1/83KE1-56S do ------------- 01-28-83 «-- ——- no Spatter erupted on January 5 from January 5 vent at Puu Kamoamoa. l/83KEl—57S do ------------- 02-01-83 —-- --- no Spatter erupted on January 3 from vent approx— imately 0.8 km northeast of Fun Kamoamoa. l/83KE1-585 do ------------- 02-01-83 --- --- no Spatter erupted on January 3 from vent approx- imately 1.1 km northeast of Puu Kamoamoa. (see sample 6/84KE1-276F) 2/83KE2-598 E. Wolfe --------- 02-10-83 1030 -—— no Fresh, glassy Spatter from new 6-m—high spat- ter cone at west end of 0740 vent. 2/83KE2-6OS do ————————————— 02—12—83 1100 -—— yes Molten spatter from west end of 0740 vent. 2/83KEZ-6IS do ------------- 02-12-83 1230 -—— yes Do. 2/83KE2-62F do ------------- 02—14-83 1120 29,113 yes Pahoehoe toe from 0740 vent, within 100 m of vent. ——- do ------------- 02-14-83 1145 2},111 —-- Do. 2/83KE2—63S do ------------- 02—14—83 1440 ——- yes Molten spatter from west end of 0740 vent. ——— N. Banks --------- 02—14-83 1525 1,111 --- Pahoehoe toe from central part of 0740 vent, 30 m south of vent. —-- E. Wolfe ————————— 02-14-83 1535 1,113 ——- Do. --- do ------------- 02-14-83 1550 1,111 --— Do. 2/83KE2-64F do ------------- 02—14—83 1610 1,113 yes Do. ——— N. Banks ————————— 02-14—83 1612 1,113 --— Do. 2/83KE2-65F E. Wolfe --------- 02—15—83 0924 1,113 yes Pahoehoe toe from 0740 vent, within 100 m of vent. 2. LAVA SAMPLE S, TEMPE RATURE S, AND COMPOSITIONS 1 1 1 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Time Temper— Sample Collector Date ature Quench Location and comments (H.s.t.) (°C) —-- N. Banks --------- 02-16—83 0912 1,113 --- Pahoehoe from 0740 vent, 40 m north of vent. 2/83KE2-66F E. Wolfe, 02—16-83 1037 2},109 yes Pahoehoe toe from 0740 vent, 30 m from vent. R. Moore. 2/83KE2-67S do ------------- 02-17—83 1020 —-- yes Molten spatter from 0740 vent, 30 m west of east end. Sample may include some older, cold 1983 spatter. 2/83KE2-68F do ------------- 02-17-83 1038 1,114 yes Pahoehoe toe from 0740 vent, approximately 90 m north of vent. 2/83KE2-69S E. Wolfe --------- 02-18-83 1452 --— yes Molten spatter from 0740 vent, 30 m west of east end. Sample includes some older, cold 1983 spatter. 2/83KE2-7OS N. Banks --------- 02-22-83 1318 1,113 yes Molten spatter from east end of 0740 vent. Temperature measured in spatter-fed pahoehoe flow. —-— do ————————————— 02-22—83 1343 1,113 --- Pahoehoe toe from 0740 vent, near east end of vent. 2/83KE2-71F do ------------- 02-22—83 1415 1,113 yes Pahoehoe toe from 0740 vent, 100 m from vent. --- do ------------- 02—22-83 1425 1,112 --- Pahoehoe from 0740 vent, 100 m from vent. ——- do ------------- 02-22-83 1444 1,119 --- Pahoehoe from 0740 vent, 180 m north of vent. -—- do ------------- 02-22-83 1457 1,117 -—- Pahoehoe from 0740 vent, 50 m from vent. 2/83KE2—7ZS do ------------- 02-23—83 0945 --- yes Molten spatter from 0740 vent. 2/83KE2-738 E. Wolfe --------- 02-23-83 1338 --- yes Molten spatter from east end of 0740 vent. Sample includes some older, cold, 1983 spatter. 2/83KE2-74F C. Milholland, 02—25-83 0741 ——— yes Pahoehoe from west end of 0740 vent. D. Jackson. 2/83KE2-75F do ------------- 02-25-83 1100 -—— yes Pahoehoe from western part of 0740 vent. 2/83KE2—76F N. Banks --------- 02-25-83 1456 1,118 yes Pahoehoe flow from west end of 0740 vent, north of vent. 2/83KE2—77F E. Wolfe --------- 02-25-83 2129 1,116 yes Pahoehoe from 1123 vent, 120 m north of vent. 2/83KE2-78F do ------------- 02-26-83 0110 1,112 yes Pahoehoe toe from 1123 vent, 50 m east of vent. 2/83KE2~79F R. Moore --------- 02—27-83 1335 1,120 yes Pahoehoe toe from western part of 1123 vent, second small vent uprift of main fountain, 20 m north of vent. 3/83KE2-80F do ------------- 03-01-83 0756 1,120 yes Pahoehoe toe in flow from pond, 300 m north- east of 1123 vent. 3/83KE2-815 J. Lockwood, 03-01-83 2146 --- yes Molten spatter from 1123 vent. D. Sundeen. 3/83KE2-82S J. Lockwood ------ 03-02—83 2053 --- yes Do. 3/83KE2-83F E. Wolfe --------- 03-03-82 1830 ——- no Aa from 1123 vent, from east edge of flow on Tuberose Street in Royal Gardens, 6 km from vent. 3/83KE3-84F do ------------- 03-22-83 —-- --- no Small pahoehoe flow erupted the previous morn- ing within episode 2 spatter ring at 1123 vent. 3/83KE3-858 N. Banks --------- 03-28-83 1100 --— 7 Spatter from O vent. 112 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Time TEmper- Sample Collector Date ature Quench Location and comments (H.s.t.) (°C) 3/83KE3-86F N. Banks --------- 03-28—83 1150 1,126 yes Pahoehoe, 150 to 200 m north of O vent. -—- do ------------- 03—28—83 1211 1,126 --- Pahoehoe toe, 100 m north of O vent. 3/83KE3-87S do ------------- 03-28-83 1238 ——- yes Spatter from 0 vent, north of vent. --- do ------------- 03-28-83 1300 1,120 -—- Pahoehoe, 350 m north of 0 vent. -—- do ------------- 03-28—83 1309 1,126 -—- Pahoehoe, 150 m north of 0 vent. 3/83KE3-88F do ------------- 03-28-83 1515 1,129 yes Pahoehoe toe, 150 m west of O vent. --- do ------------- 03-28-83 1525 1,128 —-- Pahoehoe, west of O vent. --— do ------------- 03—28-83 1622 1,123 —-— Pahoehoe, 140 m north of O vent. --- do ------------- 03-28-83 1640 1,123 --— Pahoehoe toe, 130 m north of 0 vent. 3/83KE3-89F E. Wolfe ————————— 03-29-83 1320 1,118 yes Pahoehoe toe, 150 m west of O vent. _-- do ------------- 03-29-83 1330 1,114 --— Spiny pahoehoe toe, 150 m west of 0 vent. --- do ------------- 03-29-83 1332 1,116 --- Do. --- do ------------- 03-29-83 1347 1,118 --- Pahoehoe toe, 150 m west of O vent. --- do ------------- 03-29-83 1403 1,115 --- D0. 3/83KE3-90F do ------------- 03-29-83 1505 1,114 yes Spiny pahoehoe toe, 150 m west of O vent. 3/83KE3-91F do ------------- 03-29-83 1644 --- yes Spiny pahoehoe flow from lava pond at north- eastern 1123 vent, 150 m north of vent. 3/83KE3-92F R. Moore, 03-29—83 2340 29,109 yes Spiny pahoehoe, transitional to as, from F. Trusdell. northeastern 1123 vent, 300 m northeast of vent. 3/83KE3-93F E. Wolfe, 03-30-83 1201 1,121 yes Pahoehoe toe, 80 to 100 m south of 1123 vent. M. Sako. 3/83KE3-94F do ————————————— 03-30-83 1515 --- no Aa from northeastern 1123 vent, 300 m north- east of vent. —-- do ------------- 03-31-83 1021 1,118 -—- Pahoehoe toe, 300 m northwest of 1123 cone. 3/83KE3-95F E. Wolfe, 03-31-83 1028 --- yes Do. B. Stokes. --- E. Wolfe --------- 03-31-83 1107 1,116 ——— Do. --- do ------------- 03-31-83 1130 1,116 —-— Do. 3/83KE3-96F E. Wolfe, 03-31-83 1145 1,122 yes Do. B. Stokes. 3/83KE3-97F do ------------- 03-31-83 1335 1,123 yes Do. --— E. Wolfe --------- 03-31-83 1425 1,121 -—- Do. 3/83KE3—98F E. Wolfe, 03—31—83 1535 1,118 yes Pahoehoe toe, 80 m northwest of 1123 cone. B. Stokes. 4/83KE3-99S J. Lockwood, 04—02-83 1005 --— no Air—fall lapilli (1-2 cm across) that accumu- L. Petersen. lated 1 km west of 1123 vent between 0945 and 1005 H.s.t. 4/83KE3-100F do ------------- 04-02-83 1229 1,122 yes Pahoehoe flow from southwestern 1123 vent. 4/83KE3-101F E. Wolfe, 04-02-83 1325 1,121 yes Slabby pahoehoe from southwestern 1123 vent, N. Banks. 150 m from vent. 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 113 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Ti e Temper- Sample Collector Date m ature Quench Location and comments (H.s.t.) (°C) 4/83KE3-1028 E. Wolfe, 04—02-83 1400 --- yes Spatter from southwestern 1123 vent. N. Banks. 4/83KE3-103F do ------------- 04-02-83 1420 1,114 yes Slabby pahoehoe from southwestern 1123 vent; from 20-m-long flow issuing from south base of 1123 cone. 4/83KE3-104F E. Wolfe --------- 04-02-83 1448 —-- yes Pahoehoe from southwestern 1123 vent; from channel exiting crater to west. 4/83KE3-105F J. Lockwood, 04-02-83 2127 1,118 yes Flow from southwestern 1123 vent, 300 m north- L. Petersen. west of vent. 4/83KE3-106S J. Lockwood ------ 04-03-83 0600 --- no Air-fall lapilli (1—2 cm across) that accumu- lated 1 km west of 1123 vent between 0100 and 0600 H.s.t. 4/83KE3-107F J. Lockwood, 04-04-83 0901 Z},108 7 Flow from southwestern 1123 vent, 150 m north— L. Petersen, west of vent. B. Pedit. 4/83KE3-108F E. Wolfe --------- 04-04-83 1400 1,113 yes Terminus of small spiny to slabby pahoehoe flow from southwestern 1123 vent, 100 m southwest of vent. 4/83KE3-109F do ------------- 04-04-83 1500 1,111 yes Small viscous, spiny pahoehoe flow of south- western 1123 vent, 150 to 200 m south of vent; issuing from base of cone. 4/83KE3-1105 J. Lockwood ------ 04-04-83 1900 --- --- Air-fall lapilli (1-3 cm across) that accumu- lated 1 km west of 1123 vent between 1800 and 1900 H.s.t. 4/83KE3-1118 do ------------- 04-04-83 0700 --- --- Air-fall lapilli (1—2 cm across) that accumu- lated 1 km west of 1123 vent between 0330 and 0700 H.s.t. 4/83KE3-112S do ------------- 04-05-83 1430 --- --— Air-fall lapilli (1-2 cm across) that accumu- lated 1 km west of 1123 vent between 1230 and 1430 H.s.t. 4/83KE3-113F E. Wolfe ————————— 04-05-83 1520 --- yes Cooling but still-incandescent aa from over- flow of pahoehoe river from southwestern 1123 vent, 300 m south of vent. 4/83KE3-114F do ------------- 04-05-83 1550 --- no Dense aa; same overflow, same locality as for preceding sample. 4/83KE3-115F E. Wolfe, 04-07-83 1400 -—- no Spiny pahoehoe flow from southwestern 1123 D. Jackson. vent, emplaced at about 0500 H.s.t. Sample collected about 1 km south of vent. —-- N. Banks --------- 04-08-83 1230 1,094 --- Aa-flow front in Royal Gardens, from south— western 1123 vent, 5.9 km from vent. --- do ------------- 04-08-83 1515 1,099 --- Aa-flow front in Royal Gardens, from south- western 1123 vent, 6.1 km from vent. 4/83KE3-116F E. Wolfe --------- 04-09-83 0850 -—— no Aa-flow front in Royal Gardens, from south- western 1123 vent, 7.5 km from vent. 4/83KE3-117F R. Moore, 04-09-83 1115 -—- ? Hot (not molten) pahoehoe from evacuated chan— F. Trusdell, nel of southwestern 1123 vent, 200 m south D. Jackson. of vent. 6/83KE4-118F T. DUggan, 06-13-83 1530 --- yes Pahoehoe toe of northern flow from fissure B. Talai. vents west of Puu 00, 40 m from vent. 6/83KE4-119F do ------------- 06-13-83 1558 --- yes Pahoehoe toe of northern flow from fissure vents west of Puu 00, 50 to 60 m from vent. 114 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.-—Samples collected and temperatures measured during episodes 1 through 20 of the Puu Oo eruption of Kilauea Volcano—Continued Sample Collector Date Time (H.s.t.) Temper— ature (°C) Quench Location and comments 6/83KE4-120F 6/83KE4-121F 6/83KE4-122F 6/83KE4-123F 6/83KE4-124F 6/83KE4-125F 6/83KEA-126F 6/83KE4-127F 6/83KE4-128F 6/83KEA-129F 6/83KE4-130F 6/83KE4-131F 6/83KE4—1325 6/83KE5-133F E. Wolfe --------- 06-13-83 do ------------- 06-13—83 do ------------- 06-13-83 do ------------- 06-14-83 do ------------- 06-14-83 do ------------- 06-14-83 do ------------- 06-15-83 do ------------- 06-16—83 do ------------- 06-16-83 do ------------- 06-16—83 do ------------- 06-16-83 do ------------- 06-17-83 do ------------- 06-17-83 do ------------- 06-22-83 do ------------- 06-22—83 do ------------- 06-29-83 do ————————————— 06—29—83 1809 1830 2345 0940 1030 1420 1500 1030 1110 1155 1625 1200 1200 1,127 1,120 1,129 1,132 1,127 1,128 39,120 1,128 1,115 1,117 yes yes yes yes yes yes yes (see sample 1/84KE4-207F) 1302 1332 1,127 yes Tube-fed pahoehoe toe from fissure vents west of Puu 00, 30 to 40 m northwest of vent. Crusted pahoehoe toe from fissure vents west of Puu 00, 30 to 40 m northwest of vent. Toe 2 to 3 m long, 20 cm thick, and 60 cm wide. Crust had to be broken to sample melt. Tube—fed pahoehoe from fissure vents west of Puu 00, 30 m north of vent. Tube-fed pahoehoe toe from fissure vents up- rift of Puu 00, 40 to 50 m north of vents. Flow issued from crusted pahoehoe pond de- veloping in flow on north side of vent. Tube-fed pahoehoe bud from fissure vents up- rift of Puu 00, 40 m from Puu 00; from east edge of crusted pahoehoe pond in flow on north side of vent. Tube-fed pahoehoe bud from fissure vents up— rift of Puu 00, 80 m from Puu 00; from northeast edge of crusted pahoehoe pond in flow on north side of vent. Overflow from lava river, Puu Oo vent flow to southeast, 70 m downstream from cascade over a 5— to IO-m-high scarp, 1.1 km from vent. Lava-river sample from Puu Oo vent flow to southeast, 500 m from vent. Lava—river sample from Puu Oo vent flow to southeast, 1.1 km from vent. Thick sheath of solidified basalt grew on probe after temperature reached 1,120 °C. Overflow from lava river in Puu 00 vent flow to southeast, 1 km downstream from vent at base of cascade over preexisting 5- to lO-m- high scarp. Lava-river sample from Puu Oo vent flow to southeast, 70 m downstream of cascade over preexisting 5— to 10-m-high scarp, 1.1 km from vent. Lava-river sample from Puu 00 vent flow to southeast, 80 m downstream of cascade over preexisting 5- to 10-m—high scarp, 1.1 km from vent. Do. Pahoehoe crust from spillway on southeast flank of Puu Oo. Spatter from floor of eastern crater at Puu Oo, erupted late in episode 4. Pahoehoe bud from flow front of eastern flow from Puu 00 at start of episode 5, 60 m southeast of vent in evacuated episode 4 channel. Pahoehoe bud from flow front of eastern flow from Puu 00, 300 m southeast of vent. 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 115 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Time Temper- Sample Collector Date ature Quench Location and comments (H.s.t.) (°C) 6/83KE5-134F E. Wolfe --------- 06-29—83 1335 1,127 yes Pahoehoe bud from flow front of eastern flow from Puu 00, 300 m southeast of vent. 6/83KE5-135F do ------------- 06-29-83 1552 2},127 yes Pahoehoe overflow from lava river, eastern flow from Puu 00, 400 to 500 m southeast of vent. 6/83KE5-136F do ------------- 06-29-83 1735 1,125 yes Pahoehoe bud in northern flow from western Puu 00 vent, 70 m from vent. 7/83KE5—137F do ------------- 07-01-83 1750 2},120 yes Pahoehoe overflow, 150 m from lava river, western flow from western Puu Oo vent. Overflow was thin and froze quickly. 7/83KES-138F R. Moore, 07-02-83 1105 1,129 yes Pahoehoe ooze, west side of Puu Oo cone. Tem- D. Clague, perature measurement was repeated once. J. Eaby. 7/83KE5—l39F do ------------- 07-03-83 0900 -—- no Rootless flow from west base of Puu Oo. 7/83KE5-140F J. Judd ---------- 07-07-83 --- —-- no Cold aa from channel sample of western flow from western Puu 00 vent, 3 km southeast of vent. 7/83KE5—141F E. Wolfe ————————— ? --- ——— no Aa flow from northwest side of eastern Puu 00 vent, 50 m north of vent; probably erupted late in episode 5. 7/83KE5-142F do ------------- 07-13-83 --- --- no Pahoehoe from north edge of ponded northern flow from western Puu Oo vent, 200 m north of vent; erupted June 29 or 30. 7/83KE5-143F do ------------- 07-13—83 --- -—- no Pahoehoe from flow to north and northeast from western Puu Oo vent, 100 m north of vent; erupted June 30. 7/83KE5-144F do ------------- 07-21-83 --- --- no Pahoehoe from terminus of flow to north and 7/83KE6-1455 7/83KE6-146F 7/83KE6-147F 7/83KE6-148F 7/83KE6—149F 7/83KE6-150F 7/83KE6-151F P. Greenland, 07-21—83 R. Tilling. R. Moore, 07-22-83 C. Neal. do ------------- 07-23-83 R. Moore, 07-23-83 C. Neal, L. Petersen. R. Moore, 07—23-83 C. Neal. E. Wolfe ————————— 07-24-83 do ------------- 07-24-83 do ------------- 07—24-83 1400 0935 1215 1338 1107 1058 1250 1,126 1,128 1,138 11,136 2},138 [10 yes yes yes yes northeast from western Puu 00 vent, 800 m northeast of vent; erupted June 30. Spatter in eastern crater at Puu 00; erupted during low-level activity between 1237 and 1304 H.s.t. Edge of rising lava pond in eastern crater at Puu 00. Terminus of viscous, spiny pahoehoe flow, 250 to 350 m northeast of Puu Oo. Pahoehoe flow moving eastward on top of pri— mary aa flow, 1 km from Puu Oo. Pahoehoe flow moving eastward, 1 km from Puu Oo. Pahoehoe overflow from lava river, northeast— ern flow from Puu 00, 700 m northeast of vent. Pahoehoe overflow from lava river, northeast- ern flow from Puu 00, 700 m northeast of vent. Thermocouple quickly developed a sheath of solidified basalt. Pahoehoe overflow from lava river, northeast— ern flow from Puu 00, 700 m northeast of vent. 116 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Sample Collector Date Time (H.s.t.) Temper— ature (°C) Quench Location and comments 7/83KE6-1525 7/83KE6-153S 7/83KE6—154F 8/83KE7-1555 8/83KE7-156F 8/83KE7-157F 8/83KE7-158F 8/83KE7-159F 8/83KE7-16OF 8/83KE7—161F 8/83KE7-162F 8/83KE7-163F 8/83KE7-164F 8/83KE7-165F 8/83KE7-166F J. E. P. E. Judd —————————— 07-25-83 do ------------- 07-25—83 Wolfe --------- 07—25-83 do ------------- 07—26—83 (see samples 11/83KE6-205F and Greenland ————— 08-10-83 Wolfe --------- 08-10-83 do ------------- 08-14—83 do ------------- 08-14-83 do ------------- 08-15—83 do ------------- 08—15-83 do ------------- 08—15—83 do ------------- 08—15-83 do ————————————— 08—15—83 do ————————————— 08—15—83 do ------------- 08—16—83 do ------------- 08-16-83 do ------------- 08-16-83 do ------------- 08-17-83 0656 0730 1300 0800 1645 1322 1332 0930 1110 1141 1245 1615 1645 0755 1305 1503 1200 1,119 1,134 1,135 1,134 1,130 1,136 1,132 1,136 1,135 1,138 yes yes no Spatter from Puu 00. Do. Pahoehoe in northeastern flow from Puu 00, 800 m northeast of vent. Pahoehoe from small channel on north side of Puu 00 that had been draining Puu 00 Crater at 1630 H.s.t. July 25 after the eruption stopped. 11/83KE6—238F) no yes yes yes yes yes yes yes yes yes Spatter fragments on floor of Puu 00 Crater, possibly containing some oxidized post- episode 6 rubble. From low-level activity before episode 7. Small pahoehoe flow on floor of Puu Oo Crater, erupted at 1630 H.s.t. Sample collected from molten interior of crusted toe. Temperature measured directly in vent within Puu 00, 10 minutes before emission of small pahoehoe flow between 1330 and 1332 H.s.t. Probe may not have been in melt. Small, thin pahoehoe flow on floor of Puu 00 Crater; erupted between 1330 and 1332 H.s.t. Viscous, spiny pahoehoe overflow from channel of spatter-fed southwestern flow from Puu Go, in evacuated episode 5 channel, 200 m from vent. Sample collected from different toe 5 minutes after temperature measurement. Pahoehoe overflow on north side of lava river in northeastern flow from Puu 00, 1 kn northeast of vent. Pahoehoe toe in overflow on north side of lava river in northeastern flow from Puu 00, 1 km northeast of vent. Do. Pahoehoe toe in overflow on north side of lava river, northeastern flow from Puu 00, 300 m northeast of vent. Do. Pahoehoe toe at front of flow fed by spatter and, in part, from a secondary pond also supplied by spatter, 200 to 300 m north of Puu Oo. Spiny pahoehoe toe from same flow as above sample, 200 m north of Puu Oo. Viscous tube-fed pahoehoe toe from same flow as above sample, 200 to 300 m from vent. Pahoehoe ooze through accretionaty levee at north side of lava river, northeastern flow from Puu 00, 800 to 900 m northeast of vent. 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 117 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Sample Collector Date (H.s.t.) Temper- ature Quench (°C) Time Location and comments 8/83KE7-167F 9/83KE8-168S 9/83KE8-169F 9/83KE8—17OF 9/83KE8-171F 9/83KE8-172F 9/83KE8-173F 9/83KE8-l74F 9/83KE9-175F 9/83KE9-l765 9/83KE9—l77F 9/83KE9-178F 10/83KE10-179F 10/83KE10-1805 10/83KE10-181F E. Wolfe --------- 08-17-83 do ------------- 09-02—83 do ------------- 09-03-83 do ------------- 09-04-83 do ------------- 09-04—83 do ------------- 09—06-83 do ————————————— 09—06-83 do ------------- 09-07-83 do ————————————— 09-15-83 K. Yamashita ----- 09-16-83 E. Wolfe --------- 09-17-83 do ------------- 09—23-83 do ------------- 10-02-83 P. Greenland ----- 10-02-83 E. Wolfe --------- 10-05—83 1225 1,141 yes (see sample 11/83KE7-239F) 0900 --- no 1030 --- no 0856 1,102 yes 0917 1,120 yes 0910 1,128 yes 1220 1,130 yes 1550 --- no (see sample 11/83KE8-240F) 1630 --- yes 1400 --- no 1100 --- no (see sample 9/83KE9-3AOS) 1000 ——- ? 1000 —-— no 1030 --- yes Slowanoving pahoehoe ooze through accretionary levee at north edge of lava river, north« eastern flow from Puu 00, 900 m northeast of vent. Spatter from low-level eruption within Puu 00. Sample collected adjacent to new 4-m- high spatter cone on floor of crater. Pahoehoe flow on floor of Puu Oo Crater, from western intracrater vent. Temperature inside actively degassing spatter cone over western intracrater vent. Thermo- couple may not have been immersed in melt; sample may be of remelted material. Same as above sample; thermocouple in melt. Temperature measurement was repeated several times. Sample from inside spatter cone, possibly of remelted material. Spiny pahoehoe toe in overflow 100 m from lava river; flow heading northeast from Puu 00, 800 to 900 m from vent. Viscous, spiny tube—fed pahoehoe from north- east edge of crusted pahoehoe pond develop- ing within northeastern flow, 1 km from Puu Oo. Pahoehoe issued from base of 7-m-high levee enclosing pond. Pahoehoe from spillway of Puu 00, 30 m from conduit. Sample was still hot when collect- ed 10.5 hours after eruption. Small overflow from lava river just below a lava fall; flow heading northeast from Puu 00, 30 m from vent. Pele's hair from Puu 00 on Chain of Craters Road, 9.5 km from vent. Pahoehoe from north edge of crusted pahoehoe pond in northeastern flow, 1 km from Puu Oo. Dense rock sample, still warm. Late episode 9 basalt from channel wall, 20 to 30 m northeast of Puu 00. Pahoehoe flow erupted about 0750 H.s.t. on floor of Puu Oo Crater. Some lava from flow interior was draining back into conduit when sample was collected. Spatter on floor of Puu Oo Crater, probably erupted about 0750 H.s.t. Sample may con- tain contaminants of episode 9 rock. Spiny pahoehoe/as from spatter—fed flow, 300 m north of Puu 00. Flow had been active 30 minutes earlier. 118 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu Oo eruption of Kilauea, Volcano—Continued Sample Collector Date Time (H.s.t.) TEmper- ature (°c) Quench Location and comments 10/83KE10-182F lO/83KE10-l83F lO/83KE10-184F 10/83KElO-185F lO/83KE10-l86F 10/83KE10-187F ll/83KEll-188F 11/83KE11—189F 11/83KE11-190F 11/83KE11—191F 11/83KE11-192F 11/83KE11-193F 11/83KE11-194F E. E. E. E. Wolfe --------- 10-05-83 do ------------- 10-05-83 do ------------- 10-05-83 do ------------- 10-05—83 do ------------- 10-06-83 do ------------- 10-06-83 do ------------- 10-06-83 do ------------- 10-07-83 do ------------- 11-06-83 do ————————————— 11-06-83 Wolfe, 11—06-83 C. Neal, M. Summers. do ------------- 11-06-83 Wolfe, 11-06-83 C. Neal. Wolfe --------- 11-07-83 do ————————————— 11—07-83 do ------------- 11-07-83 do ------------- 11—07-83 1319 1407 1438 1610 1341 1420 1050 (see sample ll/83KE10-241F) 0920 0950 1135 1140 1503 1000 1030 1150 1405 31,137 1,138 1,138 1,139 1,135 1,134 23,139 1,142 1,142 1,141 1,144 1,144 1,141 31,135 33,123 1,133 yes yes yes yes yes yes yes Pahoehoe toe in overflow from lava river, northeastern flow from Puu 00, 400 to 500 m north of vent. Pahoehoe toe in broad sheet overflow from lava river, northeastern flow from Puu 00, 1.2 km northeast of vent. Do. Pahoehoe at north edge of sheet overflow from lava river, northeastern flow from Puu 00, 1.3 km northeast of vent. Spiny pahoehoe at northwest edge of overflows moving north from lava river, northeastern flow from Puu 00, 500 to 600 m north-north— east of vent. Viscous pahoehoe bud in overflow from lava river, northeastern flow from Puu 00, 500 m north—northeast of vent. Pahoehoe in fastqmoving overflow from lava river, northeastern flow from Puu 00, 500 to 600 m north-northeast of vent. Temperature measurement was repeated several times. Pahoehoe ooze through accretionary levee at north edge of lava river, northeastern flow from Puu 00, 1.6 km northeast of vent. Active pahoehoe toe on surface of crusted pahoehoe sheet flow from westernmost extra- crater vent, 80 m west of source vent. Pahoehoe toe erupting from beneath crust at edge of broad, crusted pahoehoe sheet flow, 600 m north of vent. Flow was fed by spat— ter cascading over north rim of Puu 00 from vent within northern part of crater. Pahoehoe toe oozing from under solid crust of overflow or seepage from lava river, 50 m from edge of lava river, northeastern flow, 200 to 250 m northeast of Puu 00. Do. Pahoehoe toe at leading edge of rapidly moving, crusted pahoehoe sheet flow from western extracrater vents. Tomperature measured as flow decelerated 125 m uprift of westernmost vents. Pahoehoe from edge of lava river, northeastern flow, 300 m northeast of Puu Oo. Pahoehoe from edge of lava river, northeast- ern flow, 300 m northeast of Puu 00. Large sheath of solidified basalt prevented equi- librium temperature from being reached. Rapidly moving thin pahoehoe overflow, 5 m from lava river, northeastern flow, 300 m northeast of Puu 00. Pahoehoe from edge of lava river, northeastern flow, 4.5 km northeast of Puu 00. 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 119 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu Oo eruption of Kilauea Volcano—Continued Sample Collector Date Time (H.s.t.) Temper- ature (”C) Quench Location and comments ll/83KE11-195F 11/83KE12—196F 11/83KE12—197F 11/83KE12—198F 11/83KE12-199F ll/83KE12-200F 12/83KE12—201F 12/83KE12—202F 12/83KE12-203F 7/83KE5-204F 11/83KE6-205F 1/84KE12-206F 1/84KE4-207F l/84KE13-208F 1/84KEl3-209F 1/84KE13-210F E. Wolfe --------- 11-08-83 C. Neal ---------- 11-30-83 E. Wolfe --------- 11-30-83 do ------------- 11-30-83 do ------------- 11-30-83 do ------------- 12-01-83 1130 no (see sample 11/83KEll-242F) 1030 1140 1300 1450 1600 1110 1130 1300 1440 1,135 21,140 1,137 1,135 31,135 31,144 31,141 yes yes yes yes Slightly oxidized pahoehoe from floor of evac- uated channel, northeastern flow, 250 m northeast of Puu 00. Sample collected 17 hours after eruption. Edge of thin viscous pahoehoe sheet flow on northwest flank of Puu 00, both pond and spatter fed, 250 m north of Puu Oo. Spiny tube-fed pahoehoe, 600 m north of Puu 00. Spiny tube-fed pahoehoe toe, 600 m north of Puu 00. Temperature of 1,138 “C measured at same locality at 1255 H.s.t. Pahoehoe toe at edge of wide, spreading pahoe- hoe sheet flow fed by fountains and pond overflow at southeast rim of Puu 00, 200 m east of Puu 00. Sheet overflow from lava river, eastern flow from Puu 00, 20 to 25 m from channel, 400 m east of Puu 00. Average channel velocity, 2 m/s. Pahoehoe sheet flow heading north from Puu 00, 300 m from vent. Temperature of 2g,141 °C also recorded. Pahoehoe sheet flow heading north from Puu 00, 200 m north of Puu Oo. Solidified pahoehoe pond, 600 m north-north- west of Puu Oo. Dense-rock sample, collect- ed while still hot. Tube-fed pahoehoe toe issuing from edge of crusted lava pond, 600 to 700 m north of Puu 00. Flow from northern vent of Puu 00. (see samples 1/84KE12-206F and l/8AKE12-3415) C. Neal ---------- 07-29-83 E. Wolfe --------- 11—26-83 do ------------- 01-06—84 do ------------- 01—10-84 do ------------- 01-20-84 do ------------- 01-20-84 do ------------- 01-20-84 do ————————————— 01-20-84 1759 2032 2038 2230 no no no no yes yes Dense aa interior of Puu 00 flow in Royal Gar- dens, 7.4 km from source. Spiny pahoehoe outcrop in aa flow, 6.3 km northeast of Puu 00. Dense-rock sample. Spiny pahoehoe that issued from terminus of aa flow, 7 km northeast of Puu 00. Dense-rock sample. Spiny pahoehoe outcrop in aa flow, 7.4 km southeast of Puu Oo. Dense—rock sample. Spiny pahoehoe, solid but incandescent, from 350-m-long flow. Sample collected 250 m northeast of Puu Oo. Pahoehoe oozing through levee at margin of lava river, 300 m northeast of Puu 00. Temperature measured in same locality as pre- vious sample. Long-lived pahoehoe overflow, 25 m from lava river, 300 m northeast of Puu 00. 120 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.—Sarnples collected and temperatures measured during episodes 1 through 20 of the Puu Oo eruption of Kilauea Volcano—Continued Time Temper- Sample Collector Date ature Quench Location and comments (H.s.t.) o ( C) 1/84KE13—211F E. Wolfe --------- 01-21-84 0015 Z},131 yes Long-lived pahoehoe overflow, 30 m from lava river, 250 m northeast of Puu Oo. 1/84KE13-212F do ------------- 01-21-84 0345 1,140 yes Long-lived pahoehoe overflow, 75 m from lava river, 250 m northeast of Puu 00. 1/84KE13-213F do ------------- 01-21-84 0545 --— yes 1h1ten spatter from lava river as it banked high on channel wall, 30 m northeast of spillway. 1/84KE13-214F do ------------- 01—21-84 0635 1,142 yes Front of active, long-lived pahoehoe-sheet breakout from lava river, 300 to 400 m north of lava river, 500 m northeast of Puu 00. 1/84KE13-215F do ------------- 01-21-84 0834 1,141 yes Tube-fed pahoehoe toe in long-lived breakout from lava river, 300 m from lava river, 500 m northeast of Puu Oo. 1/84KE13-216F C. Byers, 01—21-84 0912 1,147 yes Tube—fed pahoehoe toe in long—lived breakout M. Garcia. from lava river, 200 m from lava river, 700 m northeast of Puu Oo. 1/84KE13-217F E. Wolfe --------- 01-21—84 1055 —-- yes Pahoehoe overflow at edge of lava river, 100 m northeast of Puu Oo. 1/84KE13-218F C. Byers, 01-21—84 1206 1,130 yes Tube-fed pahoehoe toe in long—lived breakout M. Garcia, from lava river, 300 m from lava river, 3.5 R. Moore, km northeast of Puu 00. M. Caress. 1/84KE13-219F do ------------- 01—21-84 1215 1,139 yes Do. 1/84KE13-220F do ------------- 01-21-84 1329 1,140 yes Do. 1/84KEl3-221F M. Garcia -------- 01-21-84 1620 --- yes Pahoehoe overflow from lava river, 300 m northeast of Puu Oo. 1/84KE13-222F C. Byers, 01-21-84 2016 —-- yes Pahoehoe overflow, 30 m from lava river, 300 m M. Garcia. northeast of Puu Oo. 1/84KEl3—223F C. Byers, 01-21-84 2208 --— no 1blten spatter from lava river, 250 m north- M. Garcia, east of Puu 00. M. Caress. 1/84KE13-224F do ------------- 01-22—84 0002 --— yes Pahoehoe overflow from lava river, 300 m northeast of Puu 00. I/84KE13-225F E. Wolfe, 01-22-84 0845 1,144 yes Tube-fed pahoehoe toe in several-hour-old C. Neal, breakout from proximal part of lava river M. Garcia, produced during second eruptive period, 250 C. Byers. m north of Puu 00. 1/84KE13-226F E. Wolfe --------- 01-22-84 1050 1,144 yes Pahoehoe toe issuing from under crusted sur- face of actively expanding, long-lived pa— hoehoe sheet overflow from proximal part of lava river, 250 m from Puu 00. 1/84KE13-227F do ------------- 01—22-84 1200 Z},140 yes Tube—fed pahoehoe toe in long—lived breakout from lava river, 300 m north of Puu 00. ——— do ------------- 01-22—84 1212 1,141 --- Do. (see sample 2/84KE13—233F) 1/84KE14—228F do ------------- 01-26-84 -—- --- no Remnant pahoehoe crust from highstand of lava pond inside Puu 00 Crater, erupted between 1700 H.s.t. January 25 and 0900 H.s.t. Janu- ary 26. 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 121 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea Volcano—Continued Sample Collector Time Date (H.s.t.) Temper- ature ("0) Quench Location and comments 1/84KE14-229F 1/84KE14—230F 1/84KE14-231F 1/84KE14-232F 2/84KEl3-233F 2/84KE14-234F 2/84KE15-235F 2/84KE15-236F 2/84KE15-237F 11/83KE6-238F 11/83KE7-239F 11/83KE8-240F 11/83KE10-241F 11/83KE11-242F 2/84KE15-243F 2/84KE16-244S 3/84KE16-2453 E. E. E. E. Wolfe --------- do ------------- Wolfe, C. Neal. Wolfe --------- do ------------- do ------------- do ------------- do ------------- do ------------- do ------------- do ------------- Wolfe --------- 01-30-84 1300 01-31~84 0950 01-31-84 1125 01-31-84 1340 02-10-84 02-10-84 (see 02-15-84 0915 02-15-84 1130 02-15-84 1145 02-15-84 1245 1,136 1,137 yes yes yes no no sample 2/84KE14-342$ 29,129 31,132 1,136 1:139 yes yes yes 2-m-long pahoehoe flow on floor of Puu Oo Cra- ter, from 0.5-m—wide vent through crusted surface of lava pond. Sample was incandes- cent and water cooled. Spiny pahoehoe toe from edge of broad, slow- moving pahoehoe sheet flow supplied by dis- tributary flowing northwest from lava river, 600 m north of Puu 00. Temperature measure- ment was repeated several times. Tube-fed spiny pahoehoe toe from overflow of lava river in eastern flow, 200 m north of lava river, 500 m northeast of Puu 00. Pahoehoe toe in still—active flow that had been supplied by distributary flowing north- west from lava river, 150 m northwest of Puu 00. Sample was taken 30 minutes after erup- tion stopped. Spiny pahoehoe at edge of aa flow, 6 km from Puu 00. Dense-rock sample. Thick ponded pahoehoe, 300 m northwest of Puu Oo. Dense-rock sample. Viscous, spiny pahoehoe toe oozing from edge of aa of northeastern flow from Puu 00, 800 m north of Puu Oo. Viscous, thin pahoehoe overflow, 30 m from lava river, northeastern flow, 500 m north- east of Puu 00. Do. Front of viscous, thin pahoehoe sheet over- flow, 40 m from lava river, northeastern flow, 1 km northeast of Puu 00. (see samples 2/84K315-243F and 2/84KE15—343s) 11-26-83 11-26-83 11-26-83 11-26-83 11-26-83 02-28-84 02-28-84 03-01-84 no no no no no no 1'10 Aa boulder, 5.2 km northeast of Puu 00. Dense-rock sample. Spiny pahoehoe at margin of as flow, 5.9 km northeast of Puu 00. Dense-rock sample. Pahoehoe from ponded area, 200 to 300 m north- west of 1123 vent, 1.2 km from Puu 00. Dense-rock sample. Spiny pahoehoe at edge of aa flow, just south of 1123 vent, 1.3 km from Puu Oo. Dense- rock sample. Spiny pahoehoe surrounded by aa, 5.3 km north- east of Puu 00. Dense-rock sample. Spiny pahoehoe at edge of aa flow, 3.6 km northeast of Puu 03. Dense-rock sample. Spatter collected within Puu 00 during low- level eruptive activity before episode 16. Spatter emitted during the night of February 29-March 1. Sample collected on floor of Puu Oo Crater. 122 THE PUU OO ERUP’I‘ION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983-84 TABLE 2.2.—Sa/mples collected and temperatures measured during episodes 1 through 20 of the Puu 00 eruption of Kilauea, Volcano—Continued Time Temper— Sample Collector Date ature Quench Location and comments (H.s.t.) (°C) 3/84KE16-2468 E. Wolfe --------- 03-02-84 0830 --- no Spatter emitted during the night of March 1—2. Sample collected from terrace 4 m above crusted lava pond in conduit inside Puu 00. 3/84KE16-247F E. Wolfe, 03-03-84 1818 —-- yes Spatter-fed aa flow, 300 m north of Puu 00. C. Neal. 3/84KE16-248F E. Wolfe --------- 03—04—84 0755 >1,120 yes Pahoehoe toe from stagnating flow, 500 m north of Puu 00, that was active between 0600 and 0630 H.s.t., probably spatter fed. First temperature measurement was followed by a second of >1127 “C in a similar locality. 3/84KE16-249F do ------------- 03-04-84 0810 1,139 yes Same toe as for previous sample. 3/84KE16—250F N. Banks --------- 03-04—84 1015 1,135 yes Aa—flow terminus, 2.8 km east of Puu Oo. Tem- perature measured and sample collected in breakouts from molten central zone. 3/84KE16-251F E. Wolfe --------- 03-04-84 1020 1,142 yes Fastqnoving pahoehoe-sheet overflow from lava river, 80 m from lava river, 1.5 km east of Puu Oo. 3/84KE16-252F N. Banks --------- 03-04-84 1205 1,138 ? Fluid breakout, aa—flow terminus, 3.9 km east of Puu 00. Thermocouple penetrated 1.5 m into flow interior. 3/84KE16-253F E. Wolfe --------- 03-04-84 1258 1,141 yes Pahoehoe overflow, 20 to 30 m from lava river, 1.5 km northeast of Puu Oo. 3/84KE16-254F do ------------- 03—04-84 1525 1,141 yes Pahoehoe ooze through crusted pahoehoe at edge of lava river, 1.5 km east of Puu 00. First temperature measurement was followed by oth— ers of 1,134 and 1,136 °C at same locality. 3/84KE16-255F N. Banks, 03—04-84 1613 1,137 yes Edge of aa—flow from Puu 00, 4.9 km east of R. Moore. Puu 00. Temperature measured in and sample collected from 5— to 7-m-thick flow. Probe was worked into plastic interior beneath thin aa carapace. —-- do ------------- 03-04—84 1830 1,137 —-— Fluid breakout, aa-flow terminus, 6.3 km east of Puu Oo. -—- do ------------- 03—04-84 2331 >1,135 --- Fluid, l—m—thick breakout from aa—flow termi— 3/84KE17-256F 3/84KE17-257F 3/84KE17-258F 3/84KE17-259F 3/84KE17-260F (see samples 3/84KE16-262F and E. Wolfe --------- 03-30-84 C. Neal ---------- 03-30-84 E. Wolfe --------- 03-30-84 do ------------- 03-30-84 do ------------- 03-30-84 do ------------- 03-31-84 1050 1120 1310 1415 1700 0745 1,131 1,132 1,137 1,131 nus, 7.5 km from Puu 00. m into flow interior. Probe penetrated 1 3/84KE16—344S) yes no yes Viscous incandescent aa, eastern flow, 1 km northeast of Puu Oo. Spatter—fed aa, flow terminus, 100 m west of Puu 00. Flow 1 m thick and moving 0.7 m/min. Sample was water cooled. Transitional Slabby pahoehoe/a3 overflow from central channel of southeastern flow, 1.5 km southeast of Puu 00. Slabby pahoehoe ooze in overflow at edge of lava river. Same locality as for previous sample. Pahoehoe overflowing edge of lava river, east- ern flow, 300 m northeast of Puu 00. Spiny pahoehoe oozing from underneath crust on evacuated channel floor, eastern flow, 2.0 km northeast of Puu Oo. 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 123 TABLE 2.2—Samples collected and temperatures measured during episodes 1 through 20 of the Puu Oo eruption of Kilauea Volcano—Continued Sample Collector Date Time (H.s.t.) Temper- ature (°C) Quench Location and comments 3/84KEl7—261F 3/84KE16-262F 4/84KE18-263F 4/84KE18—264F 4/84KE18-265F 4/84KE18-266F 4/84KE18—267F 4/84KE18-268F 4/84KE18-269F 5/84KE19-270F 5/84KEl9-271F 5/84KE19-272F E. Wolfe --------- 03-31—84 do ————————————— 03-23—84 do ------------- 04—19-84 do ------------- 04-19—84 do ------------- 04-19-34 do ————————————— 04-19—84 do ------------- 04-19-84 do ------------- 04-20—84 do ------------- 04—20-84 N. Banks --------- 04-20-84 do ------------- 04-20—84 do ------------- 04—21-84 E. Wolfe --------- 04-21-84 do ————————————— 05—16-84 do ------------- 05-16—84 C. Neal, R. Moore, B. Pedit. 05-17-84 0747 (see 0825 0845 0955 1355 1520 1130 1432 2038 2220 0038 1040 (see 0830 1215 1009 1'10 sample 3/84KE17-34SS) 1,139 31,143 1,137 1,140 1,136 1,144 1,136 no yes yes yes yes yes yes sample 4/84KE18-346S) 33,133 1,138 yes yes Spiny pahoehoe outcrOp in eastern aa flow, 2.0 km northeast of Puu Oo. Dense—rock sample. Spiny pahoehoe outcrop in aa flow, 2.2 km east of Puu Oo. Dense—rock sample, collected while still hot. Tube—fed pahoehoe from crusted pahoehoe pond formed of overflows from lava river in northeastern flow, 40 m north of lava river, 500 m northeast of Puu 00. Same locality as for previous sample. Tube-fed pahoehoe from crusted pahoehoe pond formed of overflows from lava river in northeastern flow, 100 m north of lava river, 600 m northeast of Puu 00. Pahoehoe toe issuing from edge of crusted pahoehoe pond formed of overflows from lava river in northeastern flow, 100 m north of lava river, 750 m northeast of Puu 00. Front of 0.5-m-thick, fluid pahoehoe sheet overflow, 150 m north of lava river in northeastern flow, 1 km northeast of Puu Oo. Pahoehoe ooze from crusted overflow of lava river in eastern flow, 40 m north of lava river, 1.1 km northeast of Puu 00. Thin, rapidly freezing pahoehoe overflow of lava river in eastern flow, 3 m north of lava river, 1.5 km northeast of Puu 00. Crust had to be broken to sample melt. Aa—flow front, eastern flow, approximately 13 km from vent. Thermocouple inserted 20 to 30 cm into plastic interior. Temperature measured in moving aa, eastern flow, approximately 13 km from vent. Aa-flow front, eastern flow, approximately 13 km from vent. Thermocouple inserted 20 cm but did not penetrate plastic interior. Spiny pahoehoe surface on floor of evacuated channel of eastern flow, still viscous in cracks, 2.0 km northeast of Puu Oo. Dense- rock sample, water cooled. Pahoehoe discharged from Puu 00 during pond overflow between 0757 and 0822 H.s.t., 150 m northeast of Puu 00. Front of thin (30 Cm thick) pahoehoe overflow from active channel formed during high foun— taining between 0930 and 1230 H.s.t., 10 m from channel, 1 km northeast of Puu Oo. Dense pahoehoe toe at front of flow discharged from Puu 00 during pond overflow between 0937 and 0953 H.s.t., 1 km northeast of Puu 00. 124 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 2.2.—Samples collected and temperatures measured during episodes 1 through 20 of the Puu Oo eruption ofKilauea Volcano—Continued Time Temper- Sample Collector Date ature Quench Location and comments (H.s.t.) (°C) —-- R. Moore --------- 05-17-84 1013 1,141 -—- Same flow and locality as for previous sample. 5/84KE19-273F E. Wolfe --------- 05—28-84 --- --- no Thick pahoehoe slab on channel floor, 100 m east of Puu 00, from last(?) lava overflow from Puu 00 during episode 19. Dense—rock sample. 6/84KEZO-274F d0 ------------- 06-08-84 0715 --- yes Tube-fed, slabby pahoehoe toe at terminus of northern flow, 3.5 km northeast of Puu Oo. --- do ------------- 06-08-84 0750 Zl,l37 --— Do. 6/84KE20-275F do ------------- 06-14—84 1100 --— no Block in Puu Oo spillway. Dense-rock sample. 6/84KE1e276F G. Ulrich -------- 06-26-84 --- —-— no Pahoehoe buried in tephra, 20 m north of west end of Puu Kamoamoa. Dense—rock sample. 9/83KE9-3408 C. Neal ---------- 09-27-83 --- -—- no Lightweight tephra (4—10 cm across) collected at surface, approximately 200 m southwest of Puu 00. 1/84KE12-3AIS do ------------- 01-10-84 —-- ——- no Do. 2/84KE14-3425 do ------------- 02—07-84 ——— --- no Do. 2/84KE15—343S do ------------- 02-21-84 --- --- no Lightweight tephra, large and small fragments, collected at surface, 600 m uprift of Puu 00. Not on main dispersal axis. 3/84KE16-3445 do ------------- 03—08-84 ——- --- no Lightweight tephra collected at camp D, 750 m west of Puu Oo. 3/84KE17-3455 do ------------- 03-30-84 1123 --- no Lightweight part of 40-cm-diameter bomb, prob— ably a product of high fountaining early in episode. Sample collected 100 m southwest of Puu 00 during the episode. 4/84KE18-3A6S do ------------- 04-19—84 1700 --- no Lightweight tephra, including large bomb frag- ments from early high fountains and small fragments from later fountaining. Sample collected approximately 900 m uprift of Puu 00. 125 2. LAVA SAMPLES, TEMPERATURES, AND COMPOSITIONS 270m hw.am £72: mm.oo mien: om.mm NH.OOH five-OOH mm.mm mm.oo~ mm.mm IlllllllamuOH No. #0. NO. No. No. ~0- No. No. No. “0. Ho. IIIIIIIIIIIIO mmwd Ava-om ww.m¢ oTooH mm.mo m1ooH N¢.mo «Ton: No.0“: ~0.mm o¢.oo~ cm.mo IIIIIIIIANuounsm HO. Mo.v HO.V ~O.V ~O.V Houv ~0.v Ho.v Ao.v Ho.v Ho.v IIIII IIIIIIm HMUOH no. 8. .S. .S. .S. 8. .So. .So. .8. 8. 8. -n-nllliluid: Ho.v Ho.v Ho. Ho. Mo. do. Ho. Ho. #0. #0. Ho. IIIIIIIIIIIIIIIMS fio.v Ho. ~O.V HO.V HO.V Ho. HO.V Ho.v AO.V ~0.V ~0.v IIIIIIIIIIIIIII Cu 2. Z. Z. Z. 2. Z. 2. Z. 2. 2. 2. -I-innnln3mom: MN. mm. cN. mN. KN. ow. 6N. AN. 9N. KN. wN. IIIIIIIIIIIIIINO m mm.N ~0.N HN.N om.N N©.N hm.N mm.N NN.N oo.N MN.N M~.N IIIIIIIIIIIIII OMB ~O.V ~0.v ~0.V Ho-v Ho.v ~0.V «AV-v HO.V HO.V Ho.v ~0.V IIIII|IIIIIIIIIO 5 mo. «o. no. no. No. wo. no. No. mo. mo. no. IIIIIIIIIIIIII+ON$ on. Mm. Nm. Tu. Nm. Qu. om. an. Om. cm. wm. IIIIIIIIIIIIiIIONM Vwcu oN.N OM.N RN.N mm.N om.N om.N N<.N ~m.~ N '— 72 — ‘U - 2 O z 5 c — I— — L“ I—‘ z z o —1 I: u 6.8 — z o 0 U l I I 4 11.0 — C — 3 .9 o _ E‘ E m - fi = g a: _ 2 (D 0 V5 2 _ A A E - A —1 10.5 — A A A 1 I I I I I I l I I l J I o 100 W70 10 20 30 40 so so TIME, IN HOURS FIGURE 3.8.—Modal mineralogy and compositional variation over time for lava erupted within the Puu Oo Crater before the beginning of episode 7 (hours less than 0) and discharged from the crater during episode 7 (hours greater than 0). Open symbols refer to phenocryst abundance based on 1,000 points per sample: circles, plagioclase; squares, olivine; triangles, augite. Connected dots are compositions based on microprobe analyses of whole-rock samples fused to glass (table 3.3). Modal compositions from table 3.1. 3. PETROLOGY OF THE ERUPTED LAVA tents in the Puu Oo lava. Between eruptions at Puu 00, the hybrid magma, in a reservoir beneath the vent, evolved by fractionation of a few percent of olivine and augite to produce a limited volume of magma depleted in MgO. The magma thus fractionated was extruded dur- ing the early part of some episodes and throughout other episodes when the erupted volume was less than that of the MgO-depleted volume in the Puu Oo reservoir. This process was certainly established by episode 5. Alternatively, changing melting processes in the mantle may have caused the observed compositional variation. The lava of episodes 11 through 20, as well as some of the lava of episodes 5 through 10, is nearly aphyric (less than 1 volume percent olivine; fig. 3.4; table 3.1) and con- tains more than 7.0 weight percent MgO. Thus, it may have undergone only olivine (plus minor chromite) frac- tionation at crustal depths (Wright, 1971). To remove the effect of varying degrees of olivine fractionation for the following discussion, we have adjusted compositions for olivine control to 12 weight percent MgO. Following Wright (1971), we made these adjustments with a mix- ture of 98.5 weight percent olivine (ngz) and 1.5 weight percent chromite. Hofmann and others (1984) also normalized lava com- positions from the 1969-71 Mauna Ulu eruption to 12.0 weight percent MgO. Plotting the normalized values against time, they noted a progressive decrease in the con- tent of incompatible minor (Na, Ti, K, P) and trace elements. They related this decrease to an increase of ap- proximately 20 percent in the degree of partial melting of the source during the eruption. Normalized compositions of olivine-controlled samples of lava from episodes 5 through 20 also show progressive 0.33 0.30 0.27 Fe/Mg RATIO IN OLIVINE PARTITION COEFFICIENT ore ore l I L 0.7 0.8 0.9 1.0 Fe/Mg RATIO IN WHOLE ROCK 0.15 1 ' FIGURE 3.9.—Fe/Mg molecular ratio of olivine phenocrysts versus whole rock for lava from episodes 5 through 8 (numbered); E, early; L, late. Horizontal lines indicate average analyses (3—4 points) for individual grains; vertical bars indicate range in composition for each sample. Partition coefficient (Fe/Mg in olivine)/(Fe/Mg in rock). 141 compositional changes (fig. 3.10). The contents of incom- patible minor elements (Na, Ti, K, P) decrease, and that of Ca increases; Si and Al contents do not vary system- atically. These results are consistent with a melting model requiring an 8- to 10-percent increase in melting of the source, leading to an increase in the contribution of clinopyroxene to the melt. Trace-element and isotopic analyses of the Puu Oo lavas are now in progress (J .M. Rhodes and A.W. Hofmann, written commun., 1985). These data will allow us to inter- pret the relative importance of each of these processes in controlling the long—term compositional variation of Puu Oo lava. SUMMARY The Puu Oo eruption is one of the most important historical eruptions of Kilauea Volcano because: (1) it is one of the longest (21/2 years and continuing as of July 1985), (2) it is the most voluminous historical eruption of the volcano, and (3) the compositional variation of the lava is large. This compositional variation probably reflects several processes, including crystal fractionation with magma mixing, and/or a progressive increase in the degree of partial melting of the mantle. Compositional variation during episodes 1 through 3 probably resulted from eruption of pockets of magma that had differentiated to varying degrees during storage in the rift zone. Signifi- cant compositional variation between the lava erupted in the early and late parts of episodes 5 through 10 may reflect fractionation of olivine and a lesser amount of augite in a shallow magma chamber beneath Puu 00. An overall progressive increase in CaO and MgO contents, and a decrease in FeOt, T102, NaZO, K20, and P205 contents, may be due to: (1) progressive increase in the proportion of summit magma mixed with rift-zone magma; (2) progressive increase in the degree of partial melting of the mantle, involving an increase in the con- tribution of clinopyroxene to the melt; or (3) a combina- tion of both these processes. REFERENCES CITED Dzurisin, Daniel, Koyanagi, R.Y., and English, T.T., 1984, Magma supply at Kilauea Volcano, Hawaii, 1956—83: Journal of Volcanology and Geothermal Research, v. 21, p. 177—206. Hofmann, A.W., Feigenson, M.D., and Raczek, Ingrid, 1984, Case studies on the origin of basalt: III. Petrogenesis of the Mauna Ulu eruption, Kilauea, 1969—1971: Contributions to Mineralogy and Petrology, v. 88, no. 1-2, p. 24—35. Jezek, P.A., Sinton, J.M., Jarosewich, Eugene, and Obermeyer, CR, 1979, Fusion of rock and mineral powders for electron microprobe analysis, in Fudali, R.F., ed., Mineral sciences investigations, 1976—1977: Smithsonian Contributions to the Earth Sciences, no. 22, p. 46—52. 142 Moore, R.B., 1983, Distribution of differentiated tholeiitic basalts on the lower east rift zone of Kilauea Volcano, Hawaii: A possible guide to geothermal exploration: Geology, v. 11, no. 3, p. 136—140. Moore, R.B., Helz, R.T., Dzurisin, Daniel, Eaton, G.P., Koyanagi, R.Y., Lipman, P.W., Lockwood, J.P., and Puniwai, G.S., 1980, The 1977 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 10.0 - 9.9 — 9.7 — 9.6 12.4 I .— 2 u: o 1: LU IL ’i g .5 012 w _ 3 123 .11 19 z 0 -_ 010 .18 p.— Z 014 .20 ,‘i‘ 12.2 — z .10 .15 8 13' 7. .9 .17 12.1 — ,8 016 I I I I l I .7 — 49 .19 .13 8'9 14 017 9 . .16 49.5 — .1011 .15 .20 O .12 018 010 49.3 — 6. '7 I 1 I 100 300 500 FIGURE 3.10.—Compositi0nal variation versus day of eruption for selected samples of olivine-controlled lava from episodes 5 through 20. Com- positions normalized for olivine control to 12 weight percent MgO, following the procedure of Hofmann and others (1984). Scale bars indicate DAYS FROM JANUARY 3, 1983 eruption of Kilauea Volcano, Hawaii: Journal of Volcanology and Geothermal Research, v. 7, p. 189—210. Peck, L.C., 1964, Systematic analysis of silicates: US. Geological Survey Bulletin 1170, 89 p. Roeder, P.L., and Emslie, RR, 1970, Olivine—liquid equilibrium: Con- 0.46 — 0.44 — 0.42 — 0.40 - 2.40 2.32 — 2.28 - 2:08 - 2.04 — 2.00 - 1.96 100 precision of analyses (Wright, 1971). Diagonal line indicates trend of variation in composition over time. 3. PETROLOGY OF THE ERUPTED LAVA tributions to Mineralogy and Petrology, v. 29, no. 4, p. 275-289. Ryan, M.P., Koyanag'i, R.Y., and Fiske, RS, 1981, Modeling the three- dimensional structure of macroscopic magma transport systems: Applications to Kilauea Volcano, Hawaii: Journal of Geophysical Research, v. 86, no. 8, p. 7111—7129. Wolfe, E.W., Garcia, M.O., Jackson, D.B., Koyanagi, R.Y., Neal, C.A., and Okamura, A.T., 1987, The Puu Oo eruption of Kilauea Volcano, episodes 1—20, January 3, 1983, to June 8, 1984, chap. 17 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: US Geological Survey Professional Paper 1350, v. 1, p. 471—508. Wright, T.L., 1971, Chemistry of Kilauea and Mauna Loa lava in space and time: US. Geological Survey Professional Paper 735, 40 p. 143 1984, Origin of Hawaiian tholeiite: A metasomatic model: J our- nal of Geophysical Research, V. 89, no. B5, p. 3233—3252. Wright, T.L., and Fiske, R.S., 1971, Origin of the differentiated and hybrid lavas of Kilauea volcano, Hawaii: Journal of Petrology, v. 12, no. 1, p. 1—65. Wright, T.L, Swanson, D.A., and Duffield, W.A., 1975, Chemical com- position of Kilauea east rift lava, 1968—1971: Journal of Petrology, v. 16, no. 1, p. 110~133. Wright, T.L., and Tilling, R.I., 1980, Chemical variation in Kilauea erup- tions 1971—1974: American Journal of Science, v. 280—A (Jackson volume), pt. 2, p. 777—793. 4. GASES FROM THE 19833-84 EAST-RIFT ERUPTION By L.P. GREENLAND CONTENTS Page Abstract ____________________________ 145 Introduction __________________________ 145 Experimental _________________________ 145 Sampling sites _______________________ 145 Sampling technique ____________________ 146 Chemical analysis ————————————————————— 146 Results and discussion _____________________ 146 Analyses of gas collections ________________ 146 Equilibrium considerations ________________ 147 Water content _______________________ 149 C/S atomic ratio ______________________ 151 Halogen contents _____________________ 151 Gas content of the magma ________________ 151 Conclusions __________________________ 152 References cited ________________________ 153 ABSTRACT Gases emitted during the January 1988-March 1984 eruptive period of the east rift of Kilauea Volcano had a very low C/ S atomic ratio (0.17). This low atomic ratio is attributable to preemptive degassing in the shallow summit storage reservoir before transport of the magma to the east rift zone. The varying water content, whether total or equilibrated, is the major source of variation in gas compositions; it induces varia- tion in the other species through the constant-sum condition. The varia- tion in water content of the gases is suggested to be due to pressure and temperature conditions at the point where the gases became separated from the magma, and to the chemical kinetics of degassing eruptive magma. Recalculated to an equilibrium assemblage, the gases follow a (log p02)—T relation very close to that directly measured in a Kilauean lava lake. The volatile contents (in weight percent) of the eruptive magma are estimated at: H20, 0.42; S, 0.11; and 002, 0.02. INTRODUCTION Sampling and analysis of eruptive gases from Kilauea Volcano began in 1912 and continued through the now- classic collections of 1917, 1918, and 1919. These samples, all from Halemaumau pit crater, were summarized by J aggar (1940) and have been the subject of many subse- quent studies (Matsuo, 1962; Nordlie, 1971; Gerlach, 1980). Eruptive-gas collections since 1919, however, have been few, sporadic, and of only limited usefulness; these collections were summarized and evaluated by Gerlach (1980). With the installation of gas-analysis laboratory facilities at the Hawaiian Volcano Observatory (HVO) in 1980 and the continuing, intermittent eruptive activity on Kilauea’s east rift zone since January 1983, a thorough sampling and analysis of eruptive gases from Kilauea Volcano has again been possible. This chapter discusses the results of analyses of gas samples collected over the first 14 months of the current eruption. The most recent east-rift eruption began in Napau Crater on January 3, 1983, and was continuing as of May 1985. During the first 5 days of this eruption, the vent system gradually extended about 8 km discontinuously to the northeastward. During January and February 1983, sites of gas emission from eruptive vents and associated ground cracks were scattered along the full length of the fissure; these sites decreased in number until July 1983, when magmatic-gas emissions became localized to a main vent at Puu 00 (see chap. 1). Since that time, gas collections have been possible only on the rare occa- sions when magma rises to the top of the conduit and forms a crusted lava pond within the cone, with small vents emitting gas, spatter, and occasional small lava flows. Full details of the geologic aspects of the eruption are given in chapter 1. EXPERIMENTAL SAMPLING SITES Two general classes of sampling sites are distinguished here: (1) noneruptive vents where magma was not visible, and (2) eruptive vents where magma was actively erupt- ing. Noneruptive sites consisted of posteruptive vents that had emitted lava and (or) spatter in the fissure system (and that occasionally erupted again after sampling), and of ground cracks. Although these ground cracks common- ly were immediately adjacent to eruptive vents, some were as much as 200 m distant. Gases from such nonerup- tive sites presumably represent degassing of the feeder dike (or dike plexus) at some unknown (presumably, less than 500 m) depth. These emissions have not occurred since July 1983, when all activity became localized at the main Puu 00 vent. At eruptive vents, gases were collected within a few meters (or even centimeters) of actively erupting (flow- ing, spattering) lava. In the period January—April 1983, vigorously fountaining vents could occasionally be sam- pled through cracks in the spatter cone. Such samples invariably were highly oxidized, presumably owing to sucking of air into the vent by the chimney effect; none of these analyses are reported here. Since July 1983, the only source of magmatic gases has been the main Puu Oo Crater. Eruptions of Puu 00 have 145 146 been uniformly preceded by a rising of magma in the con- duit to form a lava lake (10-100 m diam) on the crater floor that persists for a few hours to a few days before active fountaining begins. Occasionally, this lava lake has become temporarily crusted over while one or more small vents continue jetting gases, spatter, and small lava flows; these vents provided all the samples described here as being from eruptive vents. SAMPLING TECHNIQUE Gas samples were collected in evacuated bottles, using 1- to 2-m—long lead-in tubes of titanium, stainless steel, or mullite, 1 to 4 cm in diameter. The lead-in tubes were inserted as much as two-thirds of their length into the vent and left for several minutes for the natural gas flow to flush residual air from them. On first usage, the metal tubes were inserted into a vent for as long as 30 minutes to develop an equilibrated corrosion layer that prevented further reaction between tube and gases; no samples were taken during this conditioning process. My experience has shown that 1 to 2 minutes of flushing is adequate before sampling, whereas more than 10 to 15 minutes of flushing of stainless-steel or titanium tubing at temperatures above 1,000 °C results in samples with increased H2 contents, presumably reflecting reduction of water by the metal; all the samples reported here were collected within 5 minutes of tube insertion when stainless-steel or titanium tubing was used. Some samples were collected through mullite lead-in tubes, which are much less reactive than the metals; however, such tubes are discouragingly fragile under rough field conditions. With the sampling technique used here, there appear to be no differences in sample composition related to lead-in-tube composition. The evacuated bottles used for gas collection were similar to those described by Giggenbach (1975). Gases were collected by covering the exit of the lead—in tube with a metal washer to reduce air contamination, inserting the stem of the inverted bottle, and opening the stopcock, whereby the gases bubbled through an aqueous solution. Through July 1983, two collection bottles were used for each sample: (1) one containing 0.05 M A5203 in 6 M HCl, in which HZS immediately precipitated as AS283; 002, CO, and H2 were determined in the headspace, and H20 was determined by the weight gain of the bottle; and (2) one containing 6 M NaOH, in which acidic gases (002, SOZ, HCl, and HF) were absorbed for subsequent determination by wet chemistry; CO and H2 were again determined in the headspace; and H20 was again deter- mined by the weight gain of the bottle. Results from the two bottles were combined to provide a complete analysis. Then, in July 1983, TM. Gerlach (oral commun., 1983) suggested the much simpler technique of using a single THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 bottle containing an ammonia-Cd solution. Thereafter, single samples were collected with a bottle containing a solution of 3 M NH4OH, 0.1 M NH4CH3002, and 0.04 M CdO in which H28 was determined as CdS, acidic gases by wet chemistry in the aqueous solution, CO and H2 in the headspace, and H20 by the weight gain of the bottle. Direct comparison of these two techniques showed no significant differences in the analytical results. Temperatures were measured with a Chromel-Alumel thermocouple. Because of the difficult experimental con— ditions, measured temperatures at the eruptive vents may have been as much as 25 to 50 °C lower than the actual magma temperatures. Extreme precautions to exclude air from the sample bottles were not taken. Experiments have shown that gas collections made in this way are stable for 4 to 6 hours, even in the presence of 95 percent air; however, left over- night before analysis, such samples are useless owing to oxidation effects. Because of the proximity of laboratory facilities and the availability of helicopter transport, gas- chromatographic analyses could ordinarily be completed within 5 hours of collection; on the few occasions when analyses were impossible, the samples were discarded. Aqueous solutions of acidic gases are unaffected by air contamination, except for possible oxidation of 8032‘ to 8042‘; all the S in these solutions was calculated as S02, and these analyses could be performed at leisure. CHEMICAL ANALYSIS 002, CO, and H2 in the headspace of the acidic-solution evacuated bottles, and CO and H2 in the headspace of the alkaline-solution evacuated bottles, were determined by gas chromatography. Instrumental details are described elsewhere (Greenland, in press). 002, 802, HCl, and HF were determined in aliquants of aqueous alkaline solutions of the evacuated-bottle gases by wet chemical procedures described elsewhere (Greenland, in press). HZS was determined by the weight of As283 precipitate in the AszOg-HCI solutions. In the ammonia-CdO solutions, Cd remaining in the supernatant liquid was determined by atomic-absorption spectrometry; the difference from that originally present was attributed to CdS and thus identified with the amount of H28. RESULTS AND DISCUSSION ANALYSES OF GAS COLLECTIONS Analyses of the evacuated-bottle gas collections are listed in table 4.1; the three analyses from January 1983 have been reported previously (Greenland, 1984). The 4. GASES FROM THE 1983—84 EAST-RIFT ERUPTION 147 TABLE 4.1—Analyses of eruptive-gas samples from Kilauea. Volcano, 1983—84 [All values in mole percent] Date H20 H2 C02 C0 302 H25 H01 HF Date H20 H2 C02 CO 502 H25 HC1 HF Noneruptive vents Noneruptive vents--Oonti.nued 1/17/83 85.2 0.85 3.16 0.068 9.58 0.53 0.32 0.26 1/18/83 84.8 .62 2.33 .078 11.30 .39 .21 .26 1/18/83 82.9 .86 2.94 .084 12.30 .49 .20 .26 2/04/83 84.1 .05 2.27 .002 12.63 .65 .22 .09 2/10/83 89.3 .03 1.36 .001 8.16 .83 .22 .10 2/12/83 87.8 .02 1.91 .002 9.32 .62 .21 .14 2/17/83 89.0 .03 1.76 .004 8.50 .60 .08 .08 2/17/83 89.9 .01 .98 .001 8.61 .01 .30 .15 2/17/83 86.3 .26 1.61 .036 11.20 .24 .19 .13 2/23/83 80.3 .39 1.96 .019 16.90 .05 .23 .13 2/23/83 85.3 .02 2.16 .002 11.85 .38 .23 .05 2/24/83 75.9 .39 3.14 .029 19.80 .55 .15 .07 2/24/83 87.8 .02 1.82 .003 9.81 .00 .31 .23 2/27/83 86.4 .05 1.85 .011 11.39 .12 .12 .05 3/01/83 89.7 .02 1.26 .003 8.44 .04 .31 .22 3/01/83 79.7 .10 3.80 .032 16.20 .01 .12 .02 3/05/83 85.1 .00 3.44 .013 10.50 .00 .85 .06 3/05/83 85.8 .00 2.51 .007 11.30 .00 .35 .02 3/08/83 76.3 .00 2.46 .004 20.80 .00 .46 .02 3/08/83 83.6 .00 3.11 .004 12.50 .00 .69 .05 3/08/83 86.8 .03 1.85 .002 10.57 .28 .25 .22 3/22/83 89.4 .00 2.07 .001 8.46 .00 .03 .04 3/22/83 74.9 .03 2.11 .025 22.70 .00 .11 .06 3/22/83 83.3 .05 2.08 .005 14.30 .04 .10 .12 3/28/83 86.9 .01 1.80 .001 10.73 .11 .25 .18 3/28/83 83.9 .02 2.01 .001 13.61 .29 .08 .08 3/29/83 79.6 .07 2.52 .005 17.67 .00 .03 .08 3/30/83 82.4 .03 2.20 .002 15.16 .01 .07 .08 3/31/83 87.9 .02 1.50 .001 10.52 .00 .04 .04 4/01/83 87.9 .04 1.52 .001 10.40 .01 .02 .06 4/04/83 83.8 .03 2.07 .001 13.88 .00 .11 .10 4/05/83 82.4 .03 2.18 .001 15.18 .00 .08 .09 4/09/83 86.0 .03 1.74 .001 11.74 .18 .18 .16 4/09/83 88.1 .00 1.60 .000 9.96 .00 .16 .17 4/11/83 84.2 .03 1.98 .001 13.03 .24 .26 .24 4/13/83 82.7 .04 2.11 .001 14.60 .15 .18 .19 4/19/83 90.3 .01 1.72 .001 7.86 .00 .07 .08 4/21/83 93.5 .00 1.19 .001 5.24 .00 .03 .05 4/28/83 96.3 .00 .70 .001 2.95 .00 .04 .03 4/29/83 93.8 .00 1.03 .000 4.97 .00 .10 .06 4/29/83 96.0 .00 .80 .001 3.14 .00 .05 .04 5/04/83 74.8 .00 1.22 .004 23.77 .00 .16 .07 5/06/83 82.8 .03 2.06 .002 15.00 .00 .06 .02 5/06/83 90.2 .00 .32 .003 9.39 .00 .00 .00 5/13/83 89.5 .00 .69 .004 9.74 .00 .00 .02 large variations in the reduced species H2, CO, and H28 reflect variations in both the temperature and the oxygen partial pressure at which the gases were last in equilib- rium. Many of the samples from noneruptive vents had ample opportunity to react with air and thus be oxidized in the vent system before collection. These sources of variation can be eliminated by considering only the atomic composition of the gases (table 4.2). Much of the variation in major species can be attributed to the constant-sum condition, whereby a variation in water content induces variation in the contents of other gases (Chayes, 1960). The significance of water-content variation in these analyses can be illustrated by a ternary H—C-S atomic diagram (fig. 4.1): In such a plot, points will lie along a straight line extending from the respective apex if all the variation is due to any one of the constit- uents H20, C02, or 802, Which are the major contributors of atomic H, C, and S. The data from the 1983-84 erup- tion (table 4.2) adhere fairly closely to an HZO-control line, with a constant C/ S atomic ratio of about 0.17. This 5/16/83 87.8 .25 1.84 .007 8.96 .65 .34 .20 5/20/83 87.6 .14 1.70 .009 9.07 .53 .72 .26 5/24/83 87.4 .24 1.84 .005 9.37 .51 .40 .24 6/08/83 95.4 .00 .35 .005 4.20 .00 .02 .01 6/08/83 99.6 .00 .35 .000 .00 .00 .00 .01 6/13/83 98-4 .00 .15 .002 1.45 .00 .00 .00 6/14/83 99.6 .00 .34 .001 .00 .00 .00 .01 6/15/83 98.8 .00 .17 .000 .99 .00 .01 .02 6/29/83 99.5 .00 .09 .001 .43 .00 .00 .00 6/29/83 99-6 .00 .38 .003 .00 .00 .01 .00 6/30/83 99.9 .00 .03 .000 .10 .00 .00 .00 Eruptive vents 8/13/83 89.4 0.51 1.46 0.030 8.32 0.31 0.00 0.02 8/13/83 91.1 .36 1.33 .021 7.00 .14 .01 .05 8/14/83 84.5 .38 2.58 .025 12.40 .08 .01 .03 8/14/83 87.1 .56 1.13 .031 10.80 .31 .01 .04 8/13/83 82.7 1.76 1.75 .073 12.70 .72 .10 .21 8/13/83 88.2 .43 1.05 .058 9.68 .43 .05 .14 8/14/83 83-4 1.54 2.78 .086 11.10 1.02 .01 .03 8/14/83 84.0 1.51 2.10 .103 11.80 .43 .02 .06 8/14/83 84.7 2.08 1.36 .089 11.20 .21 .09 .29 9/02/83 91.0 .53 1.24 .014 6.44 .49 .11 .14 9/02/83 86.3 1.18 1.16 .010 10.50 .53 .10 .15 9/02/83 92.4 1 23 1.16 .043 4.43 .44 .13 .15 9/02/83 83.5 .59 1.71 .028 14.10 .01 .04 .08 9/02/83 88.9 .64 1.29 .010 8.22 .64 .10 .16 9/02/83 69.2 .50 1.72 .021 27.10 .72 .15 .55 9/02/83 90.8 .93 1.18 .027 6.63 .22 .11 .15 9/03/83 89.5 1.82 1.18 .056 6.04 1.20 .08 .12 9/03/83 83.5 .56 1.20 .019 13.10 1.33 .12 .15 9/04/83 88.0 1.24 1.38 .044 8.80 .33 .11 .14 9/04/83 91.0 .98 1.39 .031 6.06 .23 .15 .14 9/04/83 90.6 1.24 1.32 .041 6.15 .35 .13 .13 9/04/83 83.3 1.44 1.37 .053 13.40 .26 .11 .12 1/30/84 90.3 1.29 1.62 .065 6.20 .25 .19 .12 1/30/84 89.3 1.35 1.80 .077 6.80 .32 .19 .13 1/30/84 89.4 1.01 1.58 .062 7.23 .43 .16 .14 1/30/84 89.5 1.41 1.53 .074 6.91 .28 .18 .13 1/30/84 87.7 1.39 1.81 .073 8.30 .30 .20 .21 3/02/84 89.6 1.23 1.59 .083 6.74 .45 .11 .22 3/02/84 90.4 1.10 1.46 .078 6.43 .33 .06 .15 result accords With previous suggestions that Hawaiian volcanic gases are commonly admixed With meteoric water (Heald and others, 1963; Nordlie, 1971; Gerlach, 1980). Thus, the variation in the molecular proportions of the gases (table 4.1) is compatible with representation of the samples by a single gas composition that has undergone mixing with varying amounts of meteoric water and atmospheric oxygen and equilibration over a range of temperatures. EQUILIBRIUM CONSIDERATIONS A simple technique for studying equilibrium in volcanic gases is described in detail elsewhere (Greenland, in press). Estimations of apparent equilibrium temperatures from the oxidation reactions of H2, CO, and H28 show that many of the samples listed in table 4.1 approach equilibrium assemblages, whereas others are far from equilibrium. Divergence of volcanic-gas analyses from 148 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 4.2.—Atomic composition of eruptive-gas samples from Kilauea Volcano, 1983—84 [Values for C1 and F exaggerated x1,000] Date H O C S C1 F Date H O C S C1 F Noneruptive vents Noneruptive vents--Continued 1/17/83 174 111 3.23 10.11 320 260 5/16/83 178 109 1.85 9.61 340 200 1/18/83 172 112 2.41 11.69 210 260 5/20/83 178 109 1.71 9.60 720 260 1/18/83 169 113 3.02 12.79 200 260 5/24/83 177 110 1.85 9.88 400 240 2/04/83 170 114 2.27 13.28 220 90 6/08/83 191 105 .35 4.20 20 10 2/10/83 181 108 1.36 8.99 220 100 6/08/83 199 100 .35 .00 0 10 2/12/83 177 110 1.91 9.94 210 140 6/13/83 197 102 .15 1.45 0 0 2/17/83 179 110 1.76 9.10 80 80 6/14/83 199 100 .34 .00 0 10 2/17/83 180 109 .98 8.62 300 150 6/15/83 198 101 .17 .99 10 20 2/17/83 174 112 1.65 11.44 190 130 6/29/83 199 101 .09 .43 0 0 2/23/83 162 118 1.98 16.95 230 130 6/29/83 199 100 .38 .00 10 0 2/23/83 172 113 2.16 12.23 230 50 6/30/83 200 100 .03 .10 0 0 2/24/83 154 122 3.17 20.35 150 70 2/24/83 176 111 1.82 9.81 310 230 2/27/83 173 113 1.86 11.51 120 50 Eruptive vents 3/01/83 180 109 1.26 8.48 310 220 3/01/83 160 120 3.83 16.21 120 20 3/05/83 171 113 3.45 10.50 850 60 8/13/83 180 109 1.49 8.63 0 20 3/05/83 172 113 2.52 11.30 350 20 8/13/83 183 108 1.35 7.14 10 50 3/08/83 153 123 2.46 20.80 460 20 8/14/83 170 114 2.61 12.48 10 30 3/08/83 168 115 3.11 12.50 690 50 8/14/83 176 111 1.16 11.11 10 40 3/08/83 175 112 1.85 10.85 250 220 8/13/83 171 112 1.82 13.42 100 210 3/22/83 179 110 2.07 8.46 30 40 8/13/83 178 110 1.11 10.11 50 140 3/22/83 150 125 2.13 22.70 110 60 8/14/83 172 111 2.87 12.12 10 30 3/22/83 167 116 2.09 14.34 100 120 8/14/83 172 112 2.20 12.23 20 60 3/28/83 174 112 1.80 10.84 250 180 8/14/83 174 110 1.45 11.41 90 290 3/28/83 169 115 2.01 13.90 80 80 9/02/83 184 106 1.25 6.93 110 140 3/29/83 159 120 2.53 17.67 30 80 9/02/83 176 110 1.17 11.03 100 150 3/30/83 165 117 2.20 15.17 70 80 9/02/83 188 104 1.20 4.87 130 150 3/31/83 176 112 1.50 10.52 40 40 9/02/83 168 115 1.74 14.11 40 80 4/01/83 176 112 1.52 10.41 20 60 9/02/83 181 108 1.30 8.86 100 160 4/04/83 168 116 2.07 13.88 110 100 9/02/83 142 127 1.74 27.82 150 550 4/05/83 165 117 2.18 15.18 80 90 9/02/83 184 106 1.21 6.85 110 150 4/09/83 173 113 1.74 11.92 180 160 9/03/83 185 104 1.24 7.24 80 120 4/09/83 177 111 1.60 9.96 160 170 9/03/83 171 112 1.22 14.48 120 150 4/11/83 169 114 1.98 13.27 260 240 9/04/83 179 108 1.42 9.13 110 140 4/13/83 166 116 2.11 14.75 180 190 9/04/83 185 106 1.42 6.29 150 140 4/19/83 181 109 1.72 7.86 70 80 9/04/83 185 106 1.36 6.50 130 130 4/21/83 187 106 1.19 5.24 30 50 9/04/83 170 113 1.42 13.66 110 120 4/28/83 193 104 .70 2.95 40 30 1/30/84 184 106 1.68 6.45 190 120 4/29/83 188 106 1.03 4.97 100 60 1/30/84 182 107 1.88 7.12 190 130 4/29/83 192 104 .80 3.14 50 40 1/30/84 182 107 1.64 7.66 160 140 5/04/83 150 125 1.22 23.77 160 70 1/30/84 183 106 1.60 7.19 180 130 5/06/83 166 117 2.06 15.00 60 20 1/30/84 179 108 1.88 8.60 200 210 5/06/83 180 110 .32 9.39 0 0 3/02/84 183 106 1.67 7.19 110 220 5/13/83 179 110 .69 9.74 0 20 3/02/84 184 106 1.54 6.76 60 150 equilibrium can generally be attributed to excess, meteoric water in the sample and (or) oxidation of the gases, either in the vent or during storage before analysis (Gerlach, 1980; Greenland, in press). Listed in table 4.3 are the results of adjusting the analyses in table 4.1 to equilibrium assemblages, under the assumptions that all disequilibrium is due solely to (1) excess water or (2) oxidation of H2. Nonetheless, about a third of the samples listed in table 4.1 cannot be adjusted to equilibrium by this procedure because: (1) one or more of the gases H2, CO, and HZS were not detected in the analysis, a result suggestive of a highly oxidized sample and (or) a very low temperature of equilibration; (2) esti- mated equilibrium temperatures were higher than 1,200 °C, above the magmatic temperature (see chap. 2), a result suggestive of a poor analysis and (or) contamina- tion of the sample by pyrolysis of organic matter; or (3) estimated water content was less than 30 mol percent, suggestive of a high degree of oxidation of H2 in the sam- ple. The remaining two-thirds of the samples, amenable to this procedure, indicate that most of the samples from eruptive vents and many of those from noneruptive vents (table 4.3) require very small corrections in water content, probably within analytical uncertainty, to be adjusted to 4. GASES FROM THE 1983—84 EAST-RIFT ERUPTION equilibrium assemblages. Attributing all disequilibrium in the samples to oxidation of H2 requires larger relative corrections to H2 than to water content (table 4.3), but the changes, particularly at eruptive vents, are mostly small. Apparent equilibrium temperatures and oxygen partial pressures from the water correction are about the same as from the H2 correction (table 4.3). Oxygen partial pressure versus temperature for these samples relative to the quartz-fayalite—magnetite (QMF) buffer are com- pared in figure 4.2. The gas samples follow a buffer rela- tion that is very close to that of QMF at temperatures above 1,000 °C and that becomes more oxidized at lower temperatures. Fitting a least-squares line to the analytical data yields log p02=(— 1.80x 104)/T+ 3.66, which is very close to the buffer relation observed (Greenland, in press) for Mauna Loa eruptive gases: log p02=(— 1.93x 104)/T+ 4.46. Exclusion of the few points in figure 4.2 that deviate markedly from the others would make the degree of cor- respondence of the Kilauea and Mauna Loa buffers even greater. Sato and Wright (1966) also observed that oxy- gen fugacity in a cooling lava lake is buffered according to the relation log f02=(— 1.86x 104)/T+ 3.73, H C/S:0.17 C S FIGURE 4.1.—Ternary plot of H-C-S composition in evacuated-bottle samples (data from table 4.2). Atomic H has been scaled by a factor of 0.1 for clarity of plotting. Water-control line has a C/S atomic ratio of 0.17. 149 and so the oxidation state of the eruptive gases is ap- parently controlled by a lava-buffer system. WATER CONTENT Of the 29 eruptive vent samples listed in table 4.3, 26 require a less than 10 percent relative change in observed water content to match an equilibrium composition; thus, these samples represent equilibrium assemblages, within experimental uncertainties. In contrast, only 7 of the 25 samples from noneruptive vents (table 4.3) meet this con- straint and thus can be considered approximate equi- librium assemblages; the remaining 18 have been altered by contamination with water and (or) by oxidation. Gases from noneruptive vents have separated from magma at some considerable depth and traversed an unknown length of country rock, and thus had ample opportunity for contamination by meteoric water and ambient oxygen in the vent system, before collection. If oxidation has af- fected these samples, then the actual equilibrium water content is greater than that (64.0 mol percent) obtained by averaging the estimates from water contamination in table 4.3; similarly, if water contamination has been im- portant, then the actual equilibrium water content is less than that (78.9 mol percent) obtained by averaging the seven samples observed to be in approximate equilibrium. — 8.0 I I — 9.0 — 10.0 LOG p02 —11.0 —12.0 | I 7.0 7.5 8.0 5.5 10‘/ TEMPERATURE, IN KELVINS FIGURE 4.2.—Oxygen partial pressure versus temperature for equi- librated compositions of gas samples listed in table 4.3. Circles, devia- tion from equilibrium due solely to water content; X’s, deviation from equilibrium due solely to oxidation of H2. Line indicates quartz- magnetite-fayalite buffer. 150 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 4.3.——Appare’nt equilibrium values in eruptive-gas samples, assuming excess H 2 0 or ox- idation of H 2 [Dashes, samples for which the assumption of oxidation of H2 does not yield an apparent equilibrium] Excess H20 Oxidation of H2 H20 H2 Sample Temper— (mol percent) Temper- (mol percent) ature -—-——-—-——- ature —-—-———-———— (°c) PO2 Esti- Ob- (°C) 1302 Esti— ob— mated served mated served Noneruptive vents 1 1,016 10.48 82.40 85.20 1,018 10.46 1.04 0.85 2 1,090 9.63 71.60 84.80 1,099 9.50 1.35 .62 3 1,066 9.89 77.60 82.90 1,068 9.84 1.20 .86 4 705 14.99 69.80 84.10 709 14.85 .11 .05 5 675 15.77 68.60 89.30 681 15.57 .11 .03 6 703 15.19 50.20 87.80 717 14.77 .14 .02 7 761 14.17 49.40 89.00 777 13.73 .23 .03 8 809 12.21 48.40 89.90 828 11.75 .09 .01 9 1,050 9.94 61.60 86.30 1,065 9.69 .99 .26 10 1,038 9.42 79.00 80.30 1,038 9.41 .42 .39 11 712 14.82 49.60 85.30 725 14.41 .11 .02 12 937 11.25 71.40 75.90 939 11.20 .49 .39 14 907 11.48 44.80 86-40 931 10.99 .37 .05 15 843 12.12 46.80 89.70 864 11.63 .19 .02 16 1,072 8.73 53.80 79.70 1,090 8.43 .32 .10 21 733 14.31 60.60 86.80 743 14.04 .12 .03 24 869 11.53 59.40 83.30 880 11.28 .17 .05 25 700 14.73 46-60 86-90 716 14.25 .07 .01 26 681 15.24 59.40 83.90 689 14.98 .07 .02 28 825 11.72 65.60 82.40 832 11.55 .07 .03 30 791 12.28 82.00 87.90 793 12.23 .06 .04 33 703 14.67 70.60 86.00 708 14.52 .08 .03 35 689 14.99 69.60 84-20 694 14.85 .07 .03 36 700 14.58 75.60 82.70 703 14.51 .06 .04 46 820 13.07 84.40 87.80 821 13.05 .33 .25 Eruptive vents 47 853 12.59 70.40 87.60 860 12.43 0.41 0.14 48 800 13.30 86.80 87.40 800 13.30 .25 .24 57 1,026 10.27 80.80 89.40 1,031 10.19 1.01 .51 58 1,019 10.16 81.80 91.10 1,024 10.08 .81 .36 59 1,005 9.99 81.20 84.50 1,007 9.96 .48 .38 60 1,074 9.72 76.00 87.10 1,080 9.62 1.18 .56 61 1,106 9.57 84.00 82.70 1,105 9.59 1.61 1.76 62 1,144 9.25 60.40 88.20 --- --— --- --- 63 1,043 10.34 84.80 83.40 1,042 10.36 1.39 1.54 64 1,151 9.04 81-80 84.00 --- --- --- -—- 66 930 11.57 88.00 91.00 931 11.55 .73 .53 67 916 11.61 93.00 86.30 914 11.65 .56 1.18 68 1,066 10.10 89.60 92.40 1,067 10.09 1.74 1.23 70 889 12.10 90.20 88.90 889 12.10 .55 .64 71 969 10.87 67.40 69.20 969 10.85 .54 .50 72 1,050 9.95 89.40 90.80 1,051 9.94 1.10 .93 73 1,067 10.30 87.80 89.50 1,068 10.29 2.15 1.82 74 950 11.46 76.00 83.50 953 11.39 .89 .56 75 1,089 9.61 87.00 88.00 1,089 9.60 1.36 1.24 76 1,041 10.09 90.00 91.00 1,041 10.08 1.09 .98 77 1,068 9.93 89.40 90.60 1,068 9.93 1.43 1.24 78 1,148 8.88 83.80 83.30 1,148 8.88 1.38 1.44 79 1,120 9.33 88.00 90.30 1,120 9.32 1.64 1.29 80 1,120 9.38 86.60 89.30 1,122 9.35 1.73 1.35 81 1,095 9.69 83.60 89.40 1,098 9.65 1.67 1.01 82 1,147 9.09 86.60 89.50 1,148 9.07 1.85 1.41 83 1,125 9.26 86.20 87.70 1,125 9.25 1.59 1.39 84 1,131 9.39 83.60 89.60 1,134 9.34 2.06 1.23 85 1,148 9.15 83.40 90.40 —-- --- --- --- 4. GASES FROM THE 1983—84 EAST-RIFT ERUPTION In View of the common meteoric-water contamination of Hawaiian eruptive gases (Gerlach, 1980), the equilibrium water content of noneruptive-vent gases is evidently less than that (85.0 mol percent) of equilibrated eruptive-vent gases; this difference most probably reflects the depth of degassing of the magma (see chap. 5). Comparison of these results with those of other studies of Hawaiian eruptive gases shows this 1983—84 east-rift magma to be water rich. The equilibrium compositions estimated by Gerlach (1980) for the 1918—19 eruption of Halemaumau yield an average water content of 52 mol percent; the equilibrium water content of 1984 Mauna Loa eruptive-vent gases is 56 mol percent (Greenland, in press). However, the Mauna Loa gases from eruptive vents contained 73 mol percent total water; Greenland (in press) attributes the difference between equilibrated and total water at Mauna Loa to the rate of degassing, and so the difference between Mauna Loa and Kilauea magmas may be less than implied by the comparison of equilibrated water contents. The high water content of Kilauea magma may have originated during formation in the mantle. Kyser and O’Neil (1984), however, showed that most submarine basalt erupted from Kilauea’s east rift contains a substantial fraction of assimilated crustal water, and so the high water content of these gas samples may reflect only the shallow storage and transport of the 1983—84 eruptive magma. C/S ATOMIC RATIO The C/S atomic ratio of these gas samples is approx- imately constant (fig. 4.1). Averaging the data from table 4.2 yields C/S atomic ratios of 0.172 and 0.176 for gases from noneruptive and eruptive vents, respectively; thus, unlike water content, noneruptive and eruptive vents are indistinguishable in C/ S atomic ratio. The observation that the C/ S atomic ratio is independent of the depth of degas- sing as represented by these samples implies that 002 and 802 are nearly completely exsolved from the magma at a depth where H20 is only beginning to appear in the gas phase. This conclusion is consistent with suggestions that 002 and S02 are saturated in the magma at the pressure of Kilauea’s summit storage reservoir (Green- land and others, 1984), whereas H20 is considerably undersaturated (Moore, 1965). The very low C/S atomic ratio of this eruptive gas markedly contrasts with that of most basaltic gases, in which carbon is generally more abundant than sulfur (Gerlach, 1982). Gerlach (1982) emphasized the impor- tance of shallow degassing for controlling the C/S atomic ratio of volcanic gases, and Greenland (1984) and Green- land and others (1985) attributed the low ratio in these east-rift gases to prior degassing during storage of the 151 magma in the summit reservoir of Kilauea. The absence of any appreciable change in the C/ S atomic ratio of the gases over the 14-month course of the Puu Oo eruption is consistent with degassing of the magma to equilibrium in the summit reservoir before its transport to the east rift zone. HALOGEN CONTENTS Halogens are unaffected by oxidation and temperature effects, and statistical tests of the data in table 4.1 show no correlation of halogen with water content. Thus, the observed variation in halogen content reflects a real com- positional variation among these gas samples. HCl and HF are highly correlated with each other (99-percent- confidence level) but are uncorrelated with other gas species, a relation suggesting that their variation may be due to magmatic degassing rather than to changes in magmatic composition. The data of Greenland and others (1985) imply that less than 15 mol percent of the Cl and F in the magma is volatilized on eruption, and so very small changes in the efficiency of the degassing process could yield large relative changes in the halogen content of the eruptive gas. The significance of this degassing process is shown by the large difference in Cl/F atomic ratio between samples collected from eruptive vents (mean Cl/F atomic ratio, 0.68) and from vents degassing magma at some shallow depth (mean Cl/F atomic ratio, 2.8). The highest Cl/F atomic ratios from noneruptive vents (max 30) were excluded from figure 4.3, but the noneruptive vents still have much higher Cl/F atomic ratios than the eruptive vents. This change in ratio is due to increasing emission of HF from the eruptive vents. Lit- tle change in the CNS atomic ratio between noneruptive and eruptive vents is shown (fig. 4.4), whereas the F/S atomic ratio is much greater at eruptive than at non— eruptive vents. (Ratios with sulfur are used in figure 4.4 to exclude variations induced by changes in the water con- tent of these samples.) From the data in figure 4.4, H01, like 002 and S02, apparently is degassed from the magma early in its ascent, whereas HF, like H20, is most- ly degassed near the surface. GAS CONTENT OF THE MAGMA Greenland and others (1985) estimated 0.022 and 0.016 weight percent COZ degassed from the eruptive magma, from measurements of C02 emission in the plume dur- ing two episodes of the current eruption. Assuming 85 mol percent H20 in the eruptive gases (as in the equi- librated eruptive-gas samples), a C/ S atomic ratio of 0.17 (see above) yields an HgO/COZ weight ratio of 16. Com- bining this ratio With 0.02 weight percent COZ degassed 152 yields 0.32 weight percent H20 degassed on eruption. The water contents of lava spatter and flowing lava from this eruption are typically less than 0.1 weight percent (see chap. 2). Adding 0.1 weight percent H20 retained by the lava to the volatilized water yields an estimate of the H20 content of the eruptive magma of 0.42 weight per- cent, comparable to Moore’s (1965) estimate of 0.45 i 0.15 weight percent H20 for submarine basalt. The S content of the magma can be similarly estimated. Combining 0.02 weight percent C02 degassed on erup- tion with a C/S atomic ratio of 0.17 yields 0.086 weight percent S degassed on eruption. Assuming 0.02 weight percent S retained by the lava (Moore and Fabbi, 1971; Swanson and Fabbi, 1973; Fornari and others, 1979) yields a preeruptive S content for the magma of 0.11 weight percent. Summing the amounts of H20, C02, and S degassed on eruption and retained by the lava gives a total volatile content of the magma of 0.55 weight percent. Although the weight fraction of gases in the magma is small, their significance is great. With 0.43 weight per- cent of gases exsolved from the melt on eruption, an average molecular weight for the gases of 24, and a magma density of 2.8 g/cm3, we obtain 497 moles of gas per cubic meter of magma. At 1,130 °C and 1 atm pressure, this volatile content equals 57 m3 of gas per cubic meter of magma, corresponding to a system (gas+1iquid) density of 0.05 g/cm3 on eruption. Therefore, it is apparent both that lava constitutes a volumetrically insignificant fraction of a volcanic eruption and that this 20 NUMBER OF SAMPLES o l I 0 1 2 3 4 CUP ATOMIC RATIO FIGURE 4.3.—Cl/F atomic ratio in gases from noneruptive vents (A) and eruptive vents (B). Data from table 4.2. THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 fiftyfold change in system volume must be the overwhelm- ing force in fountaining dynamics. CONCLUSIONS The differing gas compositions found between eruptive and noneruptive vents suggest that degassing of the magma proceeds in stages: At the depth represented by the noneruptive vents, most of the volatilizable C02, S02, and HCl have exsolved, whereas some of the water and most of the halogens remain in the melt; continuing ex- CI/S ATOMIC RATIO I I | I NUMBER OF SAMPLES O 0.01 O 02 0.03 F/S ATOMIC RATIO 004 FIGURE 4.4.—Halogen/sulfur atomic ratios in gases from noneruptive vents (A, C) and eruptive vents (B, D). Data from table 4.2. 4. GASES FROM THE 1983—84 EAST-RIFT ERUPTION solution of dissolved H20 and HF during final rise in the eruptive conduit yields a continuously changing com- position of the eruptive-gas phase. Therefore, there is probably no need to invoke major changes in magmatic composition, meteoric-water contamination, or atmo- spheric oxidation to account for most of the observed chemical variation of these gas samples: The highly dyna- mic, nonequilibrium, eruptive degassing process neces- sarily leads to a varying composition of the gas phase. This conclusion completely accords with Shepherd’s (1921, p. 87) statement about the 1918—19 gas collections from Halemaumau: “We are not here dealing with a mixture of gas which is definite in composition and given off steadily by the magma. Each bubble has its own composi- tion * * *.” Nevertheless, these analyses and those of gases from other volcanoes (Gerlach, 1982; Greenland, in press) show that eruptive gases generally do approximate equilibrium assemblages (except, commonly, for water content). In particular, the oxidation state of Hawaiian gases appears to be controlled by a common, natural buffer system: Both these Kilauea gases and the Mauna Loa eruptive gases (Greenland, in press) follow oxygen-temperature curves very close to those directly measured (Sato and Wright, 1966) in a Kilauean lava lake. It would be of some interest to determine what this system is and how general its ap- plicability might be. These results have implications for any future collec- tions of Kilauea (and, possibly, other) eruptive gases. (1) 0/8 atomic ratios in the gases, which are controlled by degassing of the magma to equilibrium in the summit reservoir, can be expected to be similar to those reported here; an exception would be a long Halemaumau eruption, when summit degassing would coincide with eruptive degassing. (2) The contents of water and halogens, which are unsaturated in the magma at the temperature and pressure of the reservoir, may vary among eruptions, owing to assimilation of crustal water by the magma and to differing compositions of the initial magmas formed in the mantle. (3) Because very little of the halogen content is volatilized on eruption, their variations are better studied in lavas rather than in gas collections. (4) Large variations in the water content of the gases can be ex- pected both within and among eruptions, owing primar- ily to the depth at which gases separate from the magma before sampling, to the rate of magma effusion, to the amount of assimilated crustal water in the magma, and to possible incorporation of meteoric water into the sam- ple. (5) During fissure eruptions, useful gas samples can be collected from low-temperature ground cracks far removed from the actual eruptive site. Many of these collections will be found to have quenched at around 153 1,000 °C, but the depth of degassing must also be con- sidered when evaluating their water content. (6) Because degassing is a dynamic, nonequilibrium process, samples must be collected from as many sites and at as many dif- ferent times as possible. The usual practice of collecting an abundance of samples from a single site within a brief period is expected to result in misleading interpretations. REFERENCES CITED Chayes, Felix, 1960, On correlation between variables of constant sum: Journal of Geophysical Research, v. 65, no. 12, p. 4185—4193. Fornari, D.J., Moore, J .G., and Calk, L.C., 1979, A large submarine sand-rubble flow on Kilauea Volcano, Hawaii: Journal of Volcanology and Geothermal Research, v. 5, no. 3-4, p. 239-195. Gerlach, TM, 1980, Evaluation of volcanic gas analyses from Kilauea volcano: Journal of Volcanology and Geothermal Research, v. 7, no. 3, p. 295—317. ——— 1982, The interpretation of volcanic gas data from tholeiitic and alkaline mafic lavas: Bulletin Volcanologique, v. 45, no. 3, p. 235—244. Giggenbach, W.F., 1975, A simple method for the collection and analysis of volcanic gas samples: Bulletin Volcanologique, v. 39, no. 1, p. 132—145. Greenland, L.P., 1984, Gas composition of the January 1983 eruption of Kilauea Volcano, Hawaii: Geochimica et Cosmochimica Acta, v. 48, no. 1, p. 193—195. 1987, Composition of gases from the 1984 eruption of Mauna Loa Volcano, chap. 30 of Decker, R.W., Wright, T.L., and Stauifer, P.H., eds., Volcanism in Hawaii: U.S. Geological Survey Professional Paper 1350, v. 1, p. 781—790. Greenland, L.P., Rose, W.I., and Stokes, J .B., 1985, An estimate of magmatic gas content and gas emissions from Kilauea volcano, Hawaii: Geochimica et Cosmochimica Acta, v. 49, no. 1, p. 125—129. Heald, E.F., Naughton, J .J ., and Barnes, I.L., Jr., 1963, The chemistry of volcanic gases. 2. Use of equilibrium calculations in the interpreta- tion of volcanic gas samples: Journal of Geophysical Research, v. 68, no. 2, p. 545—557. Jagger, T.A., 1940, Magmatic gases: American Journal of Science, v. 238, no. 5, p. 313—353. Kyser, T.K., and O’Neil, J .R., 1984, Hydrogen isotope systematics of submarine basalts: Geochimica et Cosmochimica Acta, v. 48, no. 10, p. 2123—2133. Matsuo, Sadao, 1962, Establishment of chemical equilibrium in the volcanic gas obtained from the lava lake of Kilauea, Hawaii: Bulletin Volcanologique, v. 24, p. 59—7 1. Moore, J .G., 1965, Petrology of deep-sea basalt near Hawaii: American Journal of Science, v. 263, no. 1, p. 40—52. Moore, J .G., and Fabbi, BR, 1971, An estimate of the juvenile sulfur content of basalt: Contributions to Mineralogy and Petrology, v. 33, no. 2, p. 118—127. Nordlie, BE, 1971, The composition of the magmatic gas of Kilauea and its behavior in the near surface environment: American J our- nal of Science, v. 271, no. 5, p. 417—463. Sato, Motoaki, and Wright, T.L., 1966, Oxygen fugacities directly measured in magmatic gases: Science, v. 153, no. 3740, p. 1103—1105. Shepherd, E.S., 1921, Kilauea gases, 1919: Hawaiian Volcano Obser- vatory Monthly Bulletin, v. 9, p. 83—88. Swanson, D.A., and Fabbi, BR, 1973, Loss of volatiles during foun- taining and flowage of basaltic lava at Kilauea Volcano, Hawaii: U.S. Geological Survey Journal of Research, v. 1, no. 6, p. 649—658. 5. CONSTRAINTS ON THE MECHANICS OF THE ERUPTION By L.P. GREENLAND, ARNOLD T. OKAMURA, and J.B. STOKES CONTENTS Page Abstract ___________________________ 155 Introduction _________________________ 155 Degassing of magma _____________________ 155 Density of the gas-magma system ____________ 155 Base of the eruptive fountain _______________ 156 Effect of drainback ———————————————————— 157 Eruptive-gas composition _________________ 158 Repose-period degassing _________________ 158 Conduit properties ______________________ 159 Lithostatic versus conduit pressure ____________ 159 Conduit diameter _____________________ 160 Rise rate of magma ____________________ 161 Conclusions __________________________ 162 References cited _______________________ 163 ABSTRACT We have combined the data from gas emissions and inflation/defla- tion of Kilauea’s summit to estimate the gas content and rise rate of magma at Puu 00, the vertical gradients of pressure and density in the magma column, and the conduit diameter. Eruptive magma rises through the conduit, estimated at about 50 m in diameter, from the summit reser- voir at an estimated rate of about 0.1 m/s. Most degassing of the magma occurs in the depth interval 500~900 m of the conduit; above 540 m, less than 10 percent of the conduit volume is occupied by melt during episodes of high fountaining. The base of the lava fountain, where the magma column becomes disrupted into a spray of gas and liquid droplets, occurs at a depth of about 500 m. Because of the depth at which foun- taining of the column begins, effusive flow of magma from the vent is impossible during major fountaining events: All the main Puu Oo lava flows during high fountaining are derived from molten droplets coalesc- ing in the crater. The height and vigor of fountaining is controlled by the extent of drainback of degassed magma from the overlying pond and by the rise rate of the eruptive magma; thus, the crater-basin volume, the magma-supply rate, and the conduit dimensions determine the physical appearance of fountaining activity. Repose-period gas emissions represent previously exsolved gases migrating upward; these emissions may be the source of continuous repose-period tremor. The magma col- umn is under less than lithostatic pressure at any depth down to the summit reservoir; the difference is greater than 7.5 MPa at any depth between 350 and 2,500 m. INTRODUCTION The physical mechanics of eruptions has been widely studied (Shaw and Swanson, 1970; Sparks, 1978; Wilson and others, 1980; Wilson and Head, 1981). These studies show the critical importance of conduit diameter, magma rise rate, and magmatic gas content on eruption dynamics—parameters that generally can be only guessed at. Because of the many uncertainties, such studies necessarily are highly general and of only limited ap- plicability to individual eruptions. The recent series of episodic eruptions of Puu 00 have provided an opportunity to estimate these parameters and thus to lay a founda- tion for understanding why and how eruptions take place at one particular basaltic volcano. Our discussion of Puu 00 is restricted to eruptive episodes 15 through 24 and, within each episode, to major high fountaining activity. The sequence of eruptions that have built the massive cone of Puu 00 is fully described in chapter 1. In sum- mary, the magma conduit beneath Puu 00 is connected hydrostatically to the summit reservoir of Kilauea. Dur- ing the eruptive episodes studied here, the conduit opened into the basinlike Puu Oo crater at a height of about 100 m above the preeruptive (1982) ground surface. At the sur- face, the conduit during these episodes was typically about 20 m in diameter. At the end of a given episode, degassed magma in the crater drained back into the conduit. Dur- ing a repose period, magma slowly rose in the Puu Oo conduit as Kilauea’s summit gradually reinflated, and eventually reached the surface and ponded in the crater. With continuing magma inflow into the conduit, the pond overflowed through a spillway, and full-scale fountaining began several hours to a few days later. Data for the erup- tive periods studied here are listed in table 5.1. DEGASSING OF MAGMA Although the weight fraction of gases in the magma is small, their significance to the eruption dynamics is very great. In chapter 4, it is estimated that the gas content of the magma at 1 atm pressure and 1,130 °C is 48 m3/m3, which corresponds to a system density of 0.06 g/cm3 on eruption. This fiftyfold increase in system volume apparently is the overwhelming force in fountain dynamics. In this section, we derive curves for the varia- tions of pressure, density, and gas composition in the Puu Oo magma as a function of depth in the conduit. DENSITY OF THE GAS-MAGMA SYSTEM The current east-rift magma degasses 0.02 weight percent C02, 0.172 weight percent SOg, and 0.23 weight percent H20 on eruption (see chap. 4). Assuming a non- vesicular magma density of 2.8 g/cm3, these values cor- respond to 12.7, 75.3, and 358 mol 002, 802, and H20, respectively, per cubic meter of magma. C02 and 802 155 156 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 5.1.—Pun 00 eruption data [Data for episodes 15 through 20 from chapter 1; data for episodes 21 through 24 from George Ulrich (written commun., 1985). Deflation measurements represent the east—west tilt component recorded at Uwekahuna] Repose Eruption Pava- Total Deflation Deflation . . emission summit . Episode period length . period rate (d) (h) gate3 deflation (h) (urad/h) (10 m /h) (urad) 15 14 19 421 11.6 44 0.26 l6 17 31.5 375 14.8 33 .45 17 26 22 436 10.5 51 .21 18 18 59-5 430 19.5 67.5 .29 19 25 7 314 5.7 51 .ll 20 20 9 467 10.3 29 .36 21 22 8 713 9.4 39 .24 22 8 15 513 13.0 28 .47 23 19 18 528 12.9 24 .55 24 21 20 585 14.3 22 .65 are supersaturated in the magma arriving at the summit reservoir and are rapidly degassed through summit fumaroles; their exsolution leaves the magma saturated in 002 and 802 at the temperature and pressure of the reservoir (Greenland and others, 1984). At any lower pressure, 002 and S02 exsolve to form a separate phase in the magma; we take this saturation pressure to be 55 MPa (at 2- to 3-km depth, depending on the assumed den- sity). From data on submarine basalt, Moore (1965) estimated the saturation pressure of H20 to be 8 MPa. With these values, and assuming a linear solubility curve over this pressure range, we calculate the number of moles n of these gases exsolved as a function of pressure P (in megapascals): n(002) = — O.23P+ 12.7 (1) n(S02) = — 1.37P+ 75.4 (2) n(H20) = — 45.3P+ 362 (3) Sparks (1978) concluded that if magma becomes even slightly supersaturated, nucleation of bubbles is an in- evitable consequence. Thus, these equations yield a realistic estimate of the amount of separate gas phase present in the magma as a function of pressure. From the ideal-gas law, the total volume V (in cubic meters) of gas at 1,130 °C exsolved as a function of pressure (P<8 MPa) is given by V=(450—46.9P)(0.115/P). (4) With a negligible mass of gas and a nonvesicular-magma density of 2.8 t/m3 (= g/cm3), the density p of the melt- gas system is given by p = 2.8/(V+ 1). (5) Substituting equation 4 yields the density as a function of pressure: p = (2.8P)/(5.17 + 0.46P). (6) Thus, the pressure as a function of depth D (in meters) in the conduit is given by P=(0.98pD)+ 1. (7) The density required by equation 7 is the mean density of the overlying column. We computed equation 6 at 0.02-MPa increments of pressure and calculated the in- crements of column height represented by that pressure change from equation 7. Summing the height increments yielded the pressure-density—depth data listed in table 5.2 and plotted in figure 5.1. The mean density of the overly- ing column at any depth was obtained from equation 7 by using the pressure-depth curve of figure 5.1. Note that this procedure considers only the exsolution and expan- sion of gas bubbles as a function of pressure; though appropriate for the magma column, it provides a com- pletely unrealistic description of the physical state of the lava fountain. Our discussion concerns the properties of the magma column and refers to the fountain only for its mass. BASE OF THE ERUPTIVE FOUNTAIN The mean densities of the overlying column as a func- tion of depth are listed in table 5.2. These data indicate that at 940-m depth, where H20 begins to exsolve (8-MPa pressure), the mean density of the overlying column dur- ing eruptions is only about a third that of the undegassed 5. CONSTRAINTS ON THE MECHANICS OF THE ERUPTION TABLE 5.2.—Magma—column properties OVerlying Depth Pressure Density mean HZO/foz (m) (MPa) (g/cm3) density mo'e (g/cm3) ratio 129 0.2 0.10 0.08 4.70 261 .4 .20 .12 4.60 339 .6 .30 .15 4.49 395 .8 .40 .18 4.39 440 1.0 .49 .20 4.28 477 1.2 .58 .23 4.18 508 1.4 .67 .26 4.07 536 1.6 .75 .28 3.96 561 1.8 .84 .30 3.85 584 2.0 .92 .32 3.74 605 2.2 .99 .35 3.63 624 2.4 1.07 .37 3.52 642 2.6 1.14 .39 3.40 659 2.8 1.21 .41 3.29 675 3.0 1.28 .43 3.18 691 3.2 1.35 .45 3.06 705 3.4 1.41 .47 2.95 719 3.6 1.47 .49 2.83 732 3.8 1.53 .50 2.71 745 4.0 1.60 .52 2.59 757 4.2 1.65 .54 2.47 769 4.4 1.71 .56 2.35 781 4.6 1.76 .58 2.23 792 4.8 1.82 .59 2.11 803 5.0 1.87 .61 1.98 813 5.2 1.92 .63 1.86 823 5.4 1.97 .64 1.73 833 5.6 2.02 .66 1.61 843 5.8 2.07 .68 1.48 853 6.0 2.12 .69 1.35 862 6.2 2.16 .71 1.22 871 6.4 2.21 .72 1.09 880 6.6 2.25 .74 .96 889 6.8 2.29 .75 .82 898 7.0 2.33 .77 .69 906 7.2 2.38 .78 .55 914 7.4 2.41 .80 .42 923 7.6 2.45 .81 .28 931 7.8 2.49 .83 .14 939 8.0 2.53 .84 .00 magma; thus, melt constitutes only about 30 percent of the volume of the conduit above 940 m. Sparks (1978) showed that magma becomes disrupted into a spray of liquid droplets when gas amounts to about 75 percent of the total volume. From table 5.2, this gas fraction, cor- responding to a mean density of 0.7 g/cm3, occurs at a depth of 520 m. This depth, at which magma becomes con- verted to spray, can be taken as the base of the fountain. The highest fountain heights at Puu 00 are about 400 m above the local surface; at Puu 00 the actual fountain height is apparently more than double the observed height. 157 A consequence of locating the base of the fountain so far below the surface is that simple effusion of magma from such a vent becomes impossible; the magma is con- verted to spray long before it can reach the surface. Therefore, during major fountaining episodes at Puu 00, the lava pond (and flows therefrom) is formed by coales- cence of liquid droplets falling out of the fountain. Thus, the distinction between main flows and spatter—fed flows at Puu Oo depends on the greater extent of coalescence/ mixing of droplets falling in the crater relative to those falling on the outer, steep slopes of the cone. EFFECT OF DRAINBACK We assume that melt and exsolved gas do not separate in the conduit during eruption (fig. 5.1). The rise of gas bubbles through magma is a complex process (Sparks, 1978), difficult to evaluate. We suggest a magma rise rate in the conduit of about 0.1 m/s during active eruption of lava, a rate that may be sufficiently rapid to prevent rising of bubbles through the melt; the presence of continuous fountains, rather than sporadic gas bursts, at Puu Oo sug- gests this to be the case. In the fountain, where gas becomes the predominant component, some of the de- gassed melt will probably fall back into the magma column rather than being ejected, and so the densities calculated from equation 6 would increase. Furthermore, expected drainback of some of the degassed magma from the overlying pond into the conduit would further increase the densities calculated from equation 6. 10 Density HZO/SO2 MOLE RATIO PRESSURE, IN MEGAPASCALS DENSITY, IN GRAMS PER CUBIC CENTIMETER \Mean density l | I l | 500 600 700 800 900 DEPTH, IN METERS | | 200 300 I 400 0 ' O O 100 1 000 FIGURE 5.1.—Pressure, density, mean density of overlying magma col— umn, and H20/802 mole ratio in the gas phase, as a function of depth in the conduit. 158 If some fraction f of the eruptive magma is degassed, drains back into the conduit, and mixes with rising, new magma, the density given by equation 5 is increased: p =(2.8+2.8f)/(V+ 1+ f), (8) and equation 6 for the variation of density with pressure becomes p =(2.8(f+ 1)P)/[5.17+(0.46 mp]. (9) These equations yield the density of eruptive magma when supply of new magma, mixing of new magma with a constant fraction of degassed drainback magma, and fountaining represent a continuous process. Repeating the calculations above, using equation 9 in place of equation 6, yields a series of curves, depending on the chosen frac- tion of drainback. The effect of drainback can then be seen by plotting (fig. 5.2) the fraction of drainback against the depth of the base of the fountain, that is, where the system density falls below 0.7 g/cm3 and magma disruption begins. Even if all the eruptive magma returns to the con— duit after degassing (loo—percent drainback), the base of the fountain is raised less than 300 m, within the conduit 210 m below the surface. This analysis suggests that, regardless of the extent of drainback of degassed magma into the conduit, (1) actual fountain heights are much greater than observed, and (2) main lava flows from Puu 00 during full-fountain episodes are derived from droplets 0.8 — — 0.6 — — DRAINBACK FRACTION 0.2 — — 0 l 1 1 | o 100 200 300 400 500 DEPTH OF BASE OF FOUNTAIN, IN METERS FIGURE 5.2.—Depth in column at which magma becomes disrupted into spray (forming base of lava fountain), as a function of the fraction of eruptive magma that drains back into the conduit. THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 of magma coalescing in the crater, rather than represen- ting simple effusion of magma from a vent. Equations 8 and 9 depend on the assumption that the degassed drainback magma mixes completely with incom- ing, new magma. More realistically, we expect sinking and mixing of the dense drainback liquid to be slow relative to the influx rate of new magma; thus, an inverse density gradient would be expected to form at the top of the column. Such a gradient would raise the base of the foun- tain in the conduit above that calculated from equations 8 and 9. In the extreme case, formation of a dense plug at the top of the column would prevent fountaining altogether when the pressure at the base of the plug reached 1.5 MPa (table 5.2). We suggest that a density gradient is the major cause of the observed great varia- tion in fountaining height and style at Puu 00. For the eruptive episodes of Puu 00 studied here, extruded lava volumes are generally comparable to estimates of magma- supply rate, a result suggesting that relatively little of the degassed magma drains back into the conduit. Never- theless, our conclusions about the depth of the fountain and the impossibility of effusive flows from Puu Oo depend on the assumption that drainback of degassed magma into the eruptive column is either negligibly small or well mixed; thus, we insist on them only for the periods of full-scale, high-fountaining activity. ERUPTIVE-GAS COMPOSITION Using the pressure-depth curve (fig. 5.1), the HgO/SOZ mole ratio of the gas can be estimated from equations 2 and 3. This estimation is important in interpreting gas collections because, although gases in the fountain can only rarely be directly sampled, nearby vents commonly can provide gas samples. Fountain gas changes composi- tion very rapidly as a function of depth (fig. 5.1), and so the composition of gas samples from adjacent noneruptive vents depends on the depth at which the gas separated from the magma. Gases from nearby vents during this eruption typically have H20/802 mole ratios only half those of gases from eruptive vents (see chap. 4). Therefore, nearby vent gases appear to be escaping from the magma at a depth of 760 m, only about 200 m above the point where H20 begins to exsolve (fig. 5.1). We note that equations 1 and 2 yield a constant 002/802 mole ratio independent of pressure, as has been observed at this eruption (see chap. 4). REPOSE-PERIOD DEGASSING The density at any point in the eruptive-magma column is controlled by the mass of melt in the overlying column. 5. CONSTRAINTS ON THE MECHANICS OF THE ERUPTION When an eruption ends, collapse of the fountain adds a layer of relatively cold and dense material to the top of the magma column. If this material were to sink through the underlying lighter magma, the head pressure on the column would decrease, gases would exsolve and expand, and fountaining would ensue. This fountaining would cause more dense material to sink through the column, inducing more fountaining, and so on. In principle, this process could degas the column all the way to the sum- mit reservoir. Both direct observation and monitoring of seismic activity indicate that, in fact, fountains cease fairly abruptly and only very rarely recommence. This obser- vation is possible only if the dense material falling out of the collapsing fountain is floated by the froth at the top of the column, forming a plug that maintains the head pressure on the column. However, sinking of this dense plug through the underlying magma may provide an explanation for those rare occasions when fountaining activity is sporadic. If all of the fountain mass remains at the top of the magma column, the pressure everywhere in the column (which is due to the mass of overlying material) remains the same after as during eruption. Taking the top of the column when eruption ceases as the base of the previous fountain, the top of the column occurs at 520-m depth and has a pressure of 1.5 MPa (table 5.2). At a density of 2.8 for the degassed fountain material, this pressure implies a 50-m-thick plug (equation 7) overlying the column; thus, the top of the posteruption column (now including the plug) occurs at a depth of 470 m. At the end of eruption, the overlying lava pond also drains back into the conduit, adding to the plug thickness. Because pond dimensions vary for different eruptions of Puu 00, and because the relative contribution of the pond to the thickness of the fountain plug depends on the diameter of the conduit, we can somewhat arbitrarily assume a less-than-70—m addi- tional thickness of the plug due to pond drainback, and take the magma surface to be at a depth of 400 to 470 m (corresponding to an 80,000-m3 pond draining into a 40-m-diameter conduit, values expected to overestimate the pond contribution). We note that any drainback of the pond adds mass to the top of the column and thus in- creases the pressure on the column above that during eruption. During a repose period, the magma column rises in the conduit but this rise does not affect the mass distribution, and so the head pressure on the column remains constant. Because no mechanism exists for reducing pressure on the column, no gases exsolve from the magma during repose periods. Therefore, the gas emissions observed during repose periods must be the result of already-exsolved gases (and gases from new magma injected into the rift) migrating upward. Degas- sing and, thus, fountaining can begin only after the column is raised to the surface and the head pressure is 159 reduced by spreading of the top of the column; the ex- tent of spreading required depends on the thickness of the degassed plug and thus on the volume of the preceding eruptive pond. A problem arises if the top of the magma column is more than 400 m below the surface at the start of a repose period. Using 0.085 MPa/grad for the change in pressure with summit inflation (Decker and others, 1983), a 15-urad summit inflation (typical total repose-period inflation) raises the column only 50 m (equation 7), which is com- pletely inadequate to start a new eruptive episode if the head of the column is at a depth of 400 m. Observation as well as calculation, however, has shown the column head to lie deeper than 50 m. Therefore, because the summit-pressure change during a repose period is inade- quate to renew eruption at Puu 00, the pressure of the summit reservoir must always be greater than required to lift the magma column to the surface. However, because eruptions of Puu 00 are episodic rather than con- tinuous, the connection between the conduit and the reser- voir must be only intermittently open, as suggested by Dvorak and Okamura (1984). If magma is injected into the conduit only sporadically, the only source of con- tinuous motion in the conduit during repose periods is the rise and escape of gas bubbles from the magma. The observation that low-level seismic tremor is continuous during repose periods (see chap. 7) suggests that repose- period tremor reflects the migration of gases in the column, and so tremor magnitude should be related to the volume of gas and the rate of bubble rise. CONDUIT PROPERTIES LITHOSTATIC VERSUS CONDUIT PRESSURE The calculations for figure 5.1 extend only to 8 MPa, the saturation pressure for water. Similar calculations can be made for higher pressures by excluding water from the gases, but in table 5.2 the density of the magma at 8 MPa is about 2.5 g/cm3. Thus, for depths greater than 940 m, it is probably adequate to average this value with an unvesiculated-magma density of 2.8 g/cm3 and to assume a constant density of 2.65 g/cm3. The pressure- depth (P-D) curve below 940 m becomes P=0.265(D—940)+80. (10) Although the pressure in an eruptive conduit is com- monly taken as lithostatic (for example, Wilson and Head, 1981), at 940 m, where water begins to exsolve, the pressure in the conduit is only 8 MPa (fig. 5.1), whereas lithostatic pressure at that depth is 22 MPa. Taking the mean magma density below 940 m as 2.65 g/cm3 (as 160 above) and the overlying lava density as 2.3 (Kinoshita and others, 1963), the ratio of conduit pressure PC to lithostatic pressure P; below 940 m is given by Pc/P; = (0.265D —- 169)/(0.23D + 1), (11) and the difference between lithostatic and conduit pres- sures by Pl—Pc= 170—0.035D, (12) Where D is the depth (in meters). This relation (fig. 5.3) and the pressure-depth data from table 5.2 indicate that the pressure in the conduit is less than lithostatic at any depth down to the summit reservoir. Excess lithostatic pressure is greatest at 900—m depth, amounting to 14 MPa, and is greater than 7.5 MPa at any depth between 350 and 2,500 m; only the strength of the wallrock prevents collapse of the conduit. CONDUIT DIAMETER 802 emissions during eight repose periods (table 5.3) were measured by a correlation-spectrometer (COSPEC) technique (Stoiber and J epson, 1973; Casadevall and others, 1981). From these data, we estimate a repose- period degassing rate of magma that leads to a volume estimate and, thus, the diameter required for a cylindrical conduit. These estimates depend on extrapolating 1 to 3 days of 802 emission data over an entire repose period, even though we would expect emissions to show a rapid initial decline followed by a longer slow decline. Our measurements were generally made in the middle to late part of a repose period and thus lead to minimum esti— mates of degassing rates and conduit diameters. The degassing rate R during a repose period (in cubic meters of magma degassed per hour) can be obtained by dividing the observed number of moles of $02 emitted per hour, N, by equation 2: R=N/(75.4—0.137P). (13) Magma at pressures above 40 MPa can contribute very little to repose-period 802 emissions (equation 13, plotted as fig. 5.4). Magma at pressures below 1.5 MPa represents the degassed plug at the top of the column (ignoring pond drainback) and thus contributes nothing to repose-period 802 emissions. Computing equation 13 at 0.02-MPa incre- ments and averaging these increments over the range 1.5—40 MPa yields the magma-degassing rates for each of the repose periods listed in table 5.3. These degassing rates put constraints on the conduit geometry. Assuming a mean density of 2.65 g/cm3 for the THE PUU OO ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 column below 940 m, a pressure of 40 MPa occurs at a depth of 1,700 m below the top of the column (2,170 m below the surface); allowing for 50 m of degassed magma at the top of the column gives a column length of 1,650 m. The diameter of a cylindrical column calculated from these degassing rates is listed (table 5.3) under the assumptions 20 I l _. n I PRESSURE, IN MEGAPASCALS o I I I I 0 500 1000 1500 2000 DEPTH, IN METERS 2500 FIGURE 5.3.—Excess pressure (lithostatic minus conduit) on magma col- umn as a function of depth in conduit. 60 l I 43— 36 — _ 24— PRESSURE, IN MEGAPASCALS 0 I I I I o 2 4 6 8 I 0 VOLUME OF MAGMA DEGASSED, IN MILLIONS OF CUBIC METERS PER HOUR FIGURE 5.4.—Volume of magma that must be degassed at any pressure to yield 100 t/d (65x 103 mol/h) s02. 5. CONSTRAINTS ON THE MECHANICS OF THE ERUPTION 161 TABLE 5.3).—Estimates of conduit diameter [Conduit diameter from gas emissions calculated on basis of 100- and 30-percent gas escape. Dashes indicate a negative result] Gas emissions Magma supply Repose Diameter Inflation Pérmd SO R Volume (m) (urad) Magma Diameter (episodes) 2 — loss (m) (mol/h) (m3/h) (106 m3) 100 30 ob- Calcu- (106 m3) pct pct served lated 14-15 77,469 1,865 0.63 22 40 7.05 9.94 1 12 60 15-16 37,107 894 .36 17 31 10.47 12.56 .84 52 16—17 59,892 1,442 .90 26 48 9.30 18.94 3.86 111 19-20 51,429 1,240 .60 21 39 9.20 15.84 2 66 92 20—21 65,100 1,568 .90 25 46 7.56 15.60 3.22 101 21-22 19,530 470 .091 8.4 15 5.89 5.53 ——— ——- 22—23 123,690 2,978 1.36 32 59 12.67 14.06 56 42 23-24 97,650 2,351 1.18 30 55 10.51 16.12 2.25 85 that (1) all the gas escapes from the column during the repose period and (2) only 30 percent of the gas escapes to be measured by COSPEC. Because the actual fraction of gas that escapes to the atmosphere during a repose period is certainly less than 100 percent (gas emissions never cease), the conduit diameter based on loo-percent escape can be taken as a lower limit. An independent estimate of the conduit diameter can be based on the summit tilt record. Taking 1.0 grad of inflation of the summit as equivalent to 0.4 x 106 m3 of magma (Dvorak and Okamura, 1984) and assuming that the annual supply of magma to the summit reservoir (110 x 106 m3; Swanson, 1972) is constant over time yields an expected daily tilt change of 0.75 grad at the summit during a repose period. The difference between the observed and expected summit-inflation rates during a repose period provides an estimate of the maximum volume of magma injected into the east-rift conduit (table 5.3). We showed above that the top of the magma column is at least 400 m below the surface at the start of the repose period and must be raised to the surface before an eruption can start. Combining the volume of magma injected into the conduit with the required rise of the column provides the estimate for the diameter of a cylin- drical conduit listed in the last column of table 5.3. Both the estimate of the conduit diameter based on gas emissions and that based on summit inflation apparently have large uncertainties. At best, gas emissions provide a lower limit, and summit inflation an upper limit, for this diameter. Furthermore, the estimate from summit infla- tion refers to the upper 400 m of conduit, whereas the estimate from gas emissions refers to the depth interval 500—2,200 m. Nevertheless, the agreement of these two estimates is sufficiently close for each repose period that the overall average diameter of 501-30 m provides a realistic constraint on the conduit geometry. This estimate is barely compatible, however, with the observed open- ing of the conduit in the crater (typically, approx 20 m diameter); the conduit probably becomes constricted near the surface, possibly throughout the new Puu Oo spatter cone. The conduit diameter for repose period 21—22 could not be estimated from the summit-inflation data because the observed inflation was greater than that expected from assuming of a constant magma-supply rate. This discrep— ancy may indicate a surge of magma from the mantle into the summit reservoir: Repose period 21—22 was among the shortest observed for 25 eruptive episodes. This repose period was also anomalous for very low SOZ emis- sions, leading to a very low estimate of conduit diameter from repose-period gas emissions; the 802 measurement listed in table 5.3 is the median of 63 determinations made 3 days after the end of episode 21, and so we have no reason to discount it. RISE RATE OF MAGMA From the estimates of conduit diameter given above, we can estimate the rise rate of magma in the conduit dur- ing eruption. We assume 20-percent vesicularity for the lava to convert lava-emission rate (table 5.1) to magma— effusion rate. Combining this rate with the minimum conduit diameter estimated from gas emissions yields a maximum rise rate for the eruptive magma. We obtain a minimum rise rate estimate by combining the measured summit-deflation rate during eruption (table 5.4) with the conduit diameter estimated above from summit tilt. The use of independent estimates for conduit diameter and magma-effusion rate assures independence of the two rise-rate estimates listed in table 5.4. The magma-effusion rate estimated from summit deflation is generally less 162 TABLE 5.4.—Estimates of the rise rate of eruptive magma. [-—-, negative number calculated] Gas emissions Magma supply Episode Magma— Rise Deflation Rise effu51on ate rate volume rate 023/5) (m/s) (m3/s) 4 5 o ' ' ‘. " o z _ . _ _ ' 9 ~ . E ‘ ' z ‘ $ _ , , or 1 D '_ ' ' _ 10 'I I I I I I I I I | I'I I I l I I I I | I I 0 5 1O 15 20 DISTANCE, IN KILOMETERS FIGURE 6.9.—Longitudinal section A-A’ (see fig. 6.8 for location) of earth- quake hy'pocenters located within 3 km of plane of eruptive dike. Hypocenter cluster between 2- and 4-km depth extends for nearly 15 km along rift. A group of deeper earthquakes located east of Kalalua may indicate distal end and vertical extent of dike. B RIFT ZONE SOUTH FLANK B, 0 r I+ I I I l I I I I I . 4. 1+ + _ + + + + + _ 4. _ + + I *i" + — ’4‘ V _ +4. '1‘ + _ + 3 + + + + ++ ++ U) _ + #4. + — E ++ 1 ++ + + + + + + I— 5 w ; + + 1 + ~ Lu 4;, + + + t E + + + + + + _ O 7 + *I» + -' + E - + l 1‘" f" + 1 + + + + + + Z * + ++ 1' f — + + ++""+ + 7 I»- + + “l +-* o. 1 + + + ’ u.I + + 4. D 10% + + _ + 1 _ + 1 15 I I I+ I l I I I I l I I I 0 5 10 DISTANCE, IN KILOMETERS FIGURE 6.10.—Transverse section B—B' (see fig. 6.8 for location) of earth- quake hypocenters located in rift zone and south flank. Earthquakes outline a core beneath and slightly to south of eruptive fissures (arrow). Deeper south-flank earthquakes are probably due to increased com- pression caused by intrusion of dike into rift zone. THE PUU 00 ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 southward tilt was again recorded at station KMM. The tilt direction reversed to north side down shortly before the eruption resumed downrift of Puu Kamoamoa (line C, fig. 6.16). The small, short-period aberrations in the station KMM tiltmeter record after 1200 H.s.t. January 5 corresponded to the intermittency of the eruptive ac- tivity and possibly reflected pressure changes in the dike. Except for diurnal tilt changes, no further significant tilt changes occurred at station KMM until early April 1983. KALALUA TILT CHANGES After the first tilt change at Kalalua at 0630 H.s.t. January 2, no significant ground movement occurred at that station for nearly 3 days. The start and stop (lines A and B, fig. 6.17) of the eruption only minimally affected the tiltmeter at station KLU, but shortly before the erup- tion resumed 3 km uprift of the station on January 5 (line C, fig. 6.17), the ground started slowly to tilt northeast- ward. This movement continued until 1800 H.s.t. January 6, when the tilt rate suddenly accelerated, and the tilt direction shifted to the northwest. This large and rapid tilt change persisted into the night and exceeded the dynamic range of the instrument. The next morning, we found that the station was situated in the midst of a set of parallel tensional cracks that had formed during the night. 'IWo remeasurements of a segmented EDM line that crosses the rift zone from Kalalua (fig. 6.3) showed that 2 m of extension occurred between January 5 and January 7, and another 0.6 In between January 7 and January 11 (fig. 6.13). At 1030 H.s.t. January 7, a new fissure 1 km uprift of station KLU started to erupt vigorously (see chap. 1). ANALYSIS AND DISCUSSION PUU KAMOAMOA TILT CHANGES As shown by our model (fig. 6.53), the station KMM north-south tilt component, which was nearly perpen- dicular to the strike of the January eruptive fissure, is sensitive to the vertical movement of the dike top. We interpret the initial south-side-down tilt direction on January 2 at the station KMM, located 600 m north of the eruptive fissure, to be caused by elastic drawdown over an ascending dike. Pollard and others (1983) found that a bimodal ridge-trough-ridge form, with the trough located over the axial plane of the dike, to be the typical cross-sectional profile of the vertical-surface-displacement field over a steeply dipping ascending dike. This form causes tilt toward the dike in the area between the crest and the trough, and tilt away from the dike in the area beyond the crest. As the top of the dike approaches the 6. SURFACE DEFORMATION DURING DIKE PROPAGATION 175 155°10’ 155°05' I I 155°15' I 1 I I I I | I I A ' 6 19°25 3“ 00 .7(6\ 7 _. +++ + +4.,1 + 19°20’ 155°15' 155°10' 155°05' I | I I I I I I I I I I I B ’0 (6+ — + 7’9 + ‘ / + 7c r * 4g ‘ 6‘ ‘7 + * + w& +0; .4: : 4* + +" 1‘ +‘ — I" + 17( ¢ , 7( + a ‘ (x ‘ + ’00 O '7 of - O '79 4. O Q 0 ’I/ 7 I " «7,0 ’9 - 47 ‘7 O 4 + ‘70 47 o v * ¢ 1’7 '74. + + w Ar 6( 70,0 + + ¢ ' o (x + * +£- 5; 4» * + 19°20' ‘ 't + 9- + 0 4 KILOMETERS FIGURE 6.11.—East rift zone, showing epicenters of earthquakes. Fault traces, solid lines; major craters labeled. A, Between 0000 H.s.t. January 5 and 0000 H.s.t. January 7. Note scarcity of earthquakes in vicinity of Kalalua. B, Between 0000 H.s.t. January 7 and 0000 H.s.t. January 9. Note shift in earthquake activity to Kalalua area. 176 ground surface, the ridge crests move toward the dike axis, and this movement causes the tilt direction to reverse in the areas passed by the inward-moving crests. The ratio of the depth of the dike top to the distance between the ridge crest and the dike axis decreases from 1.0 to 0.5 as the dike top nears the surface. Thus, the reversal in tilt direction from south side down (toward the dike axis) to north side down at station KMM at 1730 H.s.t. January 2 indicated that the dike top was between 300 and 600 m of the surface. The reversal of station KMM tilt direction from north to south side down just before the eruption first stopped (line B, fig. 6.16) may have been due to deflation of the area caused by depletion of the available magma supply. This depletion of the rift-zone magma volume resulted from either a rate of effusion higher than the rate of resupply from the summit reservoir or from blockage of the magma-conduit system between the summit region and the rift-zone eruptive area. The immediate reinfla- tion of the summit region shown by the station UWE tiltmeter record (fig. 6.16) suggests that the conduit system between the summit region and the rift zone was temporarily blocked. THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 On January 5, a rapid southward tilt recorded at sta- tion KMM again indicated upward movement of the dike top. As the magma approached the surface and the erup- tion resumed (C, fig. 6.16), the tilt direction reversed to north side down. PUU KAMOAMOA TILT CHANGES AND MODEL FOR DIKE PROPAGATION To derive a propagating dike model that would recreate the tilt changes recorded at station KMM, several param- eters were held fixed, and others were allowed to vary. The site near Makaopuhi where the earthquake swarm began (7 km uprift of station KMM) was fixed as the start- ing point of dike propagation. The starting time was allowed to vary, but only earlier and not later than 0030 H.s.t. January 2. The width of the dike was allowed to be at least 2.6 m, the amount of extension measured from Kalalua on EDM lines across the newly formed cracks on January 7 and 9. From seismic data, the length of the dike was set at 15 km, and the depth was varied between 2 and 4 km. I l I I | l I | I I | I I | I ] I | | I I I — KALALUA I — 15 — | — ‘ PUU KAHAUALEA I ‘ _ I . . _ 2 s ‘ I ' ' E _ PUU KAMOAMOA ° 0 “ . . . 3 1° _ I ° 0. 0 o . . . .. o — >4 0 o 0 . g - o . o. _ ui U . . . ' — NAPA ° 0 fl _ (2) o . ° ‘ C ‘ . o 0 < — o o _ ’— . S . o g ' 33 ' I ° ' ° ‘ o o . 0° . 0 ° 5 — ‘0’. O .0 . O O 0 g — 3 o | o 0 . . MAKAOPUHI ° 0 O 0 _ .0 o — O . O a. I . ' _ . . . _ _° I ° I | | I I I I I I I I I I I I I I I I I I I I 0000 0600 1200 1800 2400 TIME (H.s.t.) FIGURE 6.12.—Earthquake-epicenter migration on January 2, 1983. Note dispersal of leading edge of earthquake pattern after 0600 H.s.t. (dashed line) in area just beyond Napau Crater. Solid line indicates average rate of downrift migration of earthquakes. 6. SURFACE DEFORMATION DURING DIKE PROPAGATION The tilt changes recorded on both components of the station KMM record (fig. 6.18) fit the theoretical displace- ment curves that we derived from Maruyama (1964) for a model dike 3.50 m Wide, propagating both vertically and horizontally at a constant rate. The origin of the model is 7 km uprift of station KMM in the area of Makaopuhi Crater, and the dike top starts from a depth of 2.5 km. Although earthquakes began at 0030 H.s.t. January 2, our model suggests that dike emplacement started at 2100 H.s.t. January 1. This discrepancy may be due to the limitations of our model because the propagation rates are held constant, or it may be similar to the situation on January 6 when the dike started to grow hours before seismic activity increased near Kalalua. The varying rate of upward dike growth is evident in the perpendicular (north-south) tilt component monitoring the vertical as- cent of the dike. Several breaks in slope of the recorded tilt curve cause the data to diverge from the constant theoretical propagation rate. The reversal in tilt direction of the east-west compo- nent at 1030 H.s.t. January 2 (fig. 6.18) marks the time at which the distal end of the growing dike passed sta- tion KMM. If the start of earthquake activity at 0030 H.s.t. is taken as the probable start of dike emplacement, a 0.7-km/h average rate of dike progagation is calculated from the east-west tilt reversal at station KMM, which is located approximately 7 km from the seismically deter- mined point of origin (X, fig. 6.2). The downrift-migration rate of the earthquake swarm (fig. 6.12) also is nearly 0.7 km/h. In contrast, our model produces an average horizon- tal propagation rate of 0.55 km/h, because the origin time is taken to be 2100 H.s.t. January 1. As mentioned earlier, the reversal in tilt direction of the north-south component at station KMM indicates that the top of the dike was between 300 and 600 m below the ground surface at 1730 H.s.t. January 2. The dike did not breach the surface at Puu Kamoamoa until about 0230 H.s.t. January 3 (see chap. 1), a timespan of 9 hours to rise a vertical distance of 300 to 600 m, or a rate of as- cent of 33 to 66 m/h. This rate of ascent for the dike top is nearly consistent with our model, which gives an average rate of 70 m/h of vertical propagation. However, the model does not perfectly fit the data, because the recorded tilt rate varied and the ground probably did not behave elastically as surface tension cracks developed. The constancy of the station KMM north-south tilt com— ponent for 2 hours at the end of January 2 suggests that the dike stopped ascending. It is not unusual for the top of a dike to approach the ground surface and stop without producing an eruption (Pollard and others, 1983). Another estimate of the average rate of upward propagation for the dike top can be calculated from seismic data by the 24-hour timespan between the start of earthquake activity and the eruption. The earthquake data (fig. 6.10) suggest 177 an origin between 2- and 4-km depth, which yields an average upward-propagation rate of about 125 mfh. KALALUA TILT CHANGES The meaning of the tilt changes at station KLU on January 6 (fig. 6.17) cannot be interpreted fully because the tiltmeter was positioned over the axis of the dike, although the reversal of tilt direction from east to west side down could mark the passing of the dike tip or simply be due to inelastic brittle failure of the rock as tension cracks formed. However, a review of the earthquake activity from 0000 H.s.t. January 5 to 0000 H.s.t. January 9 (figs. 6.11A, 6.11B) indicates that brittle failure in the Kalalua area occurred hours after the rapid tilt change began. Thus, the dike tip may have passed station KLU at about 1800 H.s.t. January 6, when the tilt of the east- west component parallel to the strike of the fissures IIIIIIIII IIIIIII [III III I I lIiII KALALUA to K-3 (100 MILLIMETERS PER DIVISION) KALALUA to K-2 HORIZONTAL DISTANCE CHANGE, IN MILLIMETERS llIIIIIII KALALUA t0 K-I | I I I I | 20 10 20 DECEMBER 1982 JANUARY 1983 FIGURE 6.13.——Changes in horizontal distances across axis of rift zone (Kalalua to K-l, K-2, and K3) and in south flank (Q BATH to FORD, fig. 6.2). Positive changes indicate lengthening, and negative changes shortening. Dike intruded between station K1 and K-2 (fig. 6.3). 178 reversed direction. The time delay between the start of seismic activity in the Kalalua area and repropagation of the dike can explain the 3-hour difference between the starting time of our model and the start of earthquakes on January 2. RIF T-ZON E MAGMA STORAGE The initial change in tilt on January 2 at station KLU (fig. 6.15), which is located 5.4 km downrift of the Puu Kamoamoa tiltmeter site, occurred 2 hours after a tilt change was recorded at station KMM (fig. 6.15). This initial tilt change at station KLU is not caused by the surface-displacement field produced by the dike tip. The surface-displacement field moved downrift at the same rate (0.55—0.7 km/h) as the dike tip, but the observed rate of horizontal movement between stations KMM and KLU was 2.7 km/h. One possible explanation for the apparent high rate of horizontal movement is that shortly before 0630 H.s.t., when the dike tip was slightly beyond Napau Crater, it intersected a body of magma stored in the rift zone and caused a pressure pulse to travel through the magma body and affect the tiltmeter at station KLU. The seismic data lend support to this assumption. At about 0600 H.s.t. January 2, downrift migration of the earth- quake swarm (fig. 6.12) terminated in an area just beyond THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 Napau Crater. The preeruption surface-deformation data also suggest accumulation of magma in the rift zone, and petrologic evidence calls for magma-storage bodies in the rift zone for eruption of lava of differentiated composi- tion (see chap. 3). SUMMARY AND CONCLUSIONS Electronic-tiltmeter measurements, previously utilized only in the summit region of Kilauea to monitor the in- flationary state of the volcano, provided invaluable infor- mation about the surface deformation associated with emplacement of a dike into the east rift zone at the start of the Puu Oo eruption. The timing of the onset of initial tilt changes on January 2 indicated that magma first started to move within the rift zone and that the summit later responded at a gradual rate while the dike propa- gated downrift. When the dike again began to move downrift toward Kalalua on January 6, tilt changes were recorded first at Kalalua, then at the summit 4 hours later. This is the first time that magmatic activity has been demonstrated to occur independently on the middle east rift zone, possibly analogous to lower-east—rift activity in 1955 and 1960. In 1960, during the Kapoho eruption, nearly 50 km away from the summit of Kilauea, 4 days elapsed between onset of the eruption and the start of Camp A to 8-1 Camp A to 8-2 LINE—LENGTH CHANGE, IN MILLIMETERS (100 MILLIMETERS PER DIVISION) I amp A to Kahaualea l 20 JANUARY 1 10 FEBRUARY FIGURE 6.14.—Line-leng‘th changes during January and February 1983, across eruptive fissures situated between linear-array EDM stations Camp A and 8-1 (fig. 6.3). TILT CHANGE, IN MICRORADIANS (1 MICRORADIAN PER DIVISION) 6. SURFACE DEFORMATION DURING DIKE PROPAGATION summit subsidence (Eaton and Murata, 1960). However, the summit started to subside Within 2 hours of the 1983 dike emplacement, situated about 15 km away from the summit. These observations imply that rift-zone magmatic activity can be initiated independently of summit in- fluence, and the response time of the summit primarily depends on the distance from the summit and the degree to which the conduit system between the summit and the site of rift-zone activity is filled with magma. Since 1975, the many east-rift-zone intrusions, the last one only 23 days before the January 3 eruption, probably kept the con- duit system between the summit region and the rift-zone eruption site filled with magma and clear of any blockages, so that response of the summit was fairly rapid. The high rate of horizontal movement for the surface- displacement field between Puu Kamoamoa and Kalalua strongly suggests interaction between the propagating IIIIIII 7< E E IIIllIIIIIIIIIIIIIIIIIIlII IIIIIIlIIIlllllll|IIlllllllllIlIIIIllIlIIIIIIIIJ_I__ l I 0000 0600 1 200 TIME (H.s.t.I FIGURE 6.15.—Initial tilt changes (arrows) on January 2, 1983, at four east-rift-zone tiltmeter stations (E SR, PHH, KMM, KLU) and one sum- mit station (SDH). Note that displacement starts (arrows) in upper east rift zone (stas. PHH, ESR) before it does at summit (sta. SDH). Cardinal points indicate direction of downslope tilt. 179 dike and a rift-zone magma-storage body. When station KLU recorded the tilt change, the dike tip was just beyond Napau Crater in the area where the migrating seismic swarm became diffuse. As the dike intersected the magma body, a pressure pulse was probably generated. This pressure pulse was transmitted through the magma body, traveled downrift to the Kalalua area, and caused a signifi- cant change in the perpendicular component of the station KLU tiltmeter. The size of this storage body, extending from Napau Crater to at least Kalalua, is considerable but not unrealistic. The preeruption tilt measurements outline an area of inflation from Napau Crater east to Heiheia— hulu, a distance of approximately 18 km; and Dzurisin and others (1983) calculated the volume of magma stored in the east rift zone since 1956 to be greater than 1 km3. An elastic model was developed to analyze the real-time tiltmeter data. This model provides displacement curves that fit the data but, more importantly, demonstrates the possibility of monitoring the movement of the dike by using tilt-direction reversals. Pollard and others (1983) first proposed the possible use of tilt reversals as a predic- tive tool, and the results of our study have shown that a well-designed array of tiltmeters parallel and orthogonal TILT CHANGE,IN MICRORADIANSHOMICRORADIANS PER DIVISION) JANUARY 1983 FIGURE 6.16.—Tilt changes in summit region (sta. UWE, lower curve) and at eruption site (sta. KMM, upper curve). A, start of eruption; B, temporary end of eruptive activity; C, resumption of eruption; D, increase in rate of subsidence in summit region. Cardinal points in- dicate direction of downslope tilt. 180 to the rift zone can provide the necessary information on movement of a dike. The tilt reversals on the station KMM record marked both passing of the dike tip at the station and ascent of the dike top. The lateral rate of propaga- tion (550—700 m/h) was slower than the 1,700-m/h rate determined by Jackson and others (1975) for the October 1968, east-rift eruption, but our rate agreed with the observed rate (600 m/h) of ground cracking reported by Duffield and others (1982) for the September 1971, southwest-rift eruption. The average rate of ascent for the dike top was determined to be 70 m/h from our model, but the tilt data suggest that this rate varies. As the dike nears the surface, the model becomes ineffective because brittle failure of the ground probably occurs. The movement of large, planar, vertical magma bodies can be monitored with a large network of tiltmeters arranged parallel and perpendicular to the strike of the rift zone. The tilt data must be collected on a real-time basis, and recognition of tilt-direction reversals is critical. TILT CHANGE, IN MICRORADIANS (1OMICRORADIANS PER DIVISION) I l l | 2 3 4 5 6 JANUARY 1983 FIGURE 6.17.—Tiltmeter record from Kalalua (KLU) station. A, start of eruption; B, temporary end of eruptive activity; C, resumption of eruption; D, start of a large, rapid tilt change that exceeded dynamic range of instrument. Cardinal points indicate direction of downslope tilt. THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 REFERENCES CITED Ando, Masataka, 1979, The Hawaii earthquake of November 29, 1975: Low dip angle faulting due to forceful injection of magma: Journal of Geophysical Research, v. 84, no. B13, p. 7616—7626. Dieterich, J .H., and Decker, R.W., 1975, Finite element modeling of surface deformation associated with volcanism: Journal of Geophysical Research, v. 80, no. 29, p. 4094—4102. Duffield, W.A., Christiansen, R.L., Koyanagi, R.Y., and Peterson, D.W., 1982, Storage, migration, and eruption of magma at Kilauea volcano, Hawaii, 1971—1972: Journal of Volcanology and Geothermal Research, v. 13, no. 3—4, p. 273—307. Dvorak, J .J ., Okamura, A.T., and Dieterich, J .H., 1983, Analysis of sur- face deformation data, Kilauea volcano, Hawaii: October 1966 to September 1970: Journal of Geophysical Research, v. 88, no. B11, p. 9295—9304. Dzurisin, Daniel, Koyanagi, R.Y., and English, T.T., 1984, Magma supply and storage at Kilauea volcano, Hawaii, 1956—1983: Journal of Volcanology and Geothermal Research, v. 21, p. 177—206. Eaton, J.P., 1962, Crustal structure and volcanism in Hawaii, in Macdonald, G.A., and Kuno, Hisashi, eds., The crust of the Pacific basin: American Geophysical Union Geophysical Monograph 6, p. 13—29. Eaton, J .P., and Murata, K.J., 1960, How volcanoes grow: Science, v. 132, no. 3432, p. 925—938. Fiske, RS, and Kinoshita, W.T., 1969, Inflation of Kilauea volcano prior to its 1967—1968 eruption: Science, v. 165, no. 3891, p. 341—349. Jackson, D.B., Swanson, D.A., Koyanagi, R.Y., Wright, T.L., 1975, The August and October 1968 east rift eruptions of Kilauea Volcano, Hawaii: U.S. Geological Survey Professional Paper 890, 33 p. J aggar, T.A., and Finch, R.H., 1926, Tilting and level changes at Pacific volcanoes: Pan-Pacific Science Congress, 3d, Tokyo, 1926, Pro- ceedings, p. 672—686. Johnston, M.J.S., 1976, Testing the physical parameters of short baseline tiltmeters intended for earthquake prediction: U.S. Geological Survey Open-File Report 76—556, 11 p. Maruyama, Takuo, 1964, Statical elastic dislocations in an infinite and semi-infinite medium: University of Tokyo, Earthquake Research Institute Bulletin, v. 42, no. 2, p. 289—368. Mogi, Kiyoo, 1958, Relations between the eruptions of various volcanoes and the deformations of the ground surfaces around them: Univer- sity of Tokyo, Earthquake Research Institute Bulletin, v. 36, no. 2, p. 99—134. Mortensen, C.E., Iwatsubo, E.Y., Johnston, M.J.S., Myren, G.D., Keller, V.G., and Murray, T.L., 1977, U.S.G.S. tiltmeter networks, opera- tion and maintenance: U.S. Geological Survey Open-File Report 77—655, 18 p. Pollard, D.D., Delaney, P.T., Duffield, W.A., Endo, E.T., and Okamura, A.T., 1983, Surface deformation in volcanic rift zones, in Morgan, Paul, and Baker, B.H., eds., Processes of continental rifting: Tectonophysics, v. 94, no. 1—4 (special issue), p. 541—584. Ryan, M.P., Blevins, J .Y.K., Okamura, A.T., and Koyanagi, R.Y., 1983, Magma reservoir subsidence mechanics: Theoretical summary and application to Kilauea volcano, Hawaii: Journal of Geophysical Research, v. 88, no. B5, p. 4147—4181. Swanson, D.A., Jackson, D.B., Koyanagi, R.Y., and Wright, T.L., 1976, The February 1969 east rift eruption of Kilauea volcano, Hawaii: U.S. Geological Survey Professional Paper 891, 30 p. Swanson, D.A., Duffield, W.A., Jackson, D.B., and Peterson, D.W., 197 9, Chronological narrative of the 1969—71 Mauna Ulu eruption of Kilauea Volcano, Hawaii: U.S. Geological Survey Professional Paper 1056, 55 p. 6. SURFACE DEFORMATION DURING DIKE PROPAGATION 181 TILT CHANGE, IN MICRORADIANS (10 MICRORADIANS PER DIVISION) JANUARY 1 JANUARY 2 JANUARY 3 FIGURE 6.18.—Tiltmeter record from Puu Kamoamoa (KMM) station of 3.5 m, dike is allowed to breach surface when eruption starts. Final (January 1983), in comparison with theoretical tilt curves (dots) derived length of dike is 15 km and height is 2.5 km. Movement of model dike from an analytical model of a dike propagating both horizontally and starts at 2100 H.s.t. January 1. Cardinal points indicate direction of vertically. Dike originates 7 km uprift from station and moves laterally downslope tilt. at a rate of 550 m/h and vertically at a rate of 70 m/h. With a width 7. SEISMICITY ASSOCIATED WITH THE ERUPTION By ROBERT Y. KOYANAGI, WILFRED R. TANIGAWA, and JENNIFER S. NAKATA CONTENTS Page Abstract ___________________________ 183 Introduction _________________________ 183 Limitations of the data ____________________ 185 Seismic network and data processing _____________ 186 Data, 1982—84 _________________________ 189 Location and migration pattern of earthquakes —————— 189 Depth and magnitude of earthquakes ___________ 191 Rate and frequency of earthquakes ____________ 193 Strong tremor in the east rift zone during eruptive episodes _______________________ 197 Seismic events in the east rift zone between episodes — — — 215 Seismic events at the summit _______________ 215 Chronology of events, 1978—83 ———————————————— 218 Summary and discussion ——————————————————— 220 References cited 234 ABSTRACT Downrift migration of the shallow earthquake swarms that occurred episodically for many months before the east-rift eruption of January 1983 reflected the process of magma movement from the summit to the east-rift storage complex as a preliminary stage to the eruption. Migration of earthquakes at a rate of 0.6 to 0.7 km/h during the final intrusion before the eruption provided a seismic measure for the rate of downrift propagation of magmatic pressure. The spatial distribution of summit, rift-zone, and south-flank earthquakes delineated the magma- transport system. This progression of earthquakes in time, as well as the distribution of earthquakes in space, supports and quantifies earlier concepts of (1) the pattern of magma movement and (2) the structure of the magma- transport system that feeds eruptions at Kilauea. Initially, magma rising from the mantle builds pressure in a small storage complex within a few kilometers beneath the summit. Dictated by the strength of the re- taining caprock and by stress conditions in the east rift zone, the overload of fluid pressure beneath the summit may be tapped along the reser- voir and its vertical conduit within a 10-km-deep structural feature, and directed laterally beneath the east rift in a conduit complex no more than a few kilometers wide. Magma transferred in this rift conduit, in turn, rises nearly vertically through a relatively confined system of fissures to feed the eruptive vents. After the preeruption intrusion or dike emplacement and the early stages of eruption, earthquakes dissipate, and tremor becomes localized. Magma movement is aseismic along certain parts of the vertical con- duit beneath the summit storage system and, eventually, along an extensive length of the lateral transport system linking the summit and the middle-east-rift eruption zone. This decrease and localization of the seismicity, reflecting the increasing efficiency of the magma movement, becomes increasingly apparent during later eruptive episodes. The pro- longed eruption is characterized by episodic transfer of magma batches from the summit to the east rift, presumably controlled by fluctuations of pressure above and below critical levels in the expanded summit-rift system. Continuation of the episodic eruption is indicated by the persistence of tremor localized near the eruptive vents. The tremor signal is intense during active lava output and decreases to a low level during repose times, when amplitudes vary mainly according to minor movement of lava and degassing activity from within the vent. The amplitude of har- monic tremor provides a measure of the vigor of the eruption and the rate of magma movement. Attenuation of tremor as a function of distance identifies the eruptive vent in the east rift as the principal source of radiation, and the summit as a secondary source. The east-rift source of tremor is many times more energetic than the summit source at the height of eruptive episodes. The amplitude of east-rift tremor during vigorous eruption is also influenced by noise generated from high fountaining. INTRODUCTION This chapter summarizes observations of changing seismicity both preceding and accompanying the 1983 east-rift eruption of Kilauea Volcano. We begin our summary with a review of concepts of the seismicity of Kilauea and an account of the aftermath of the summit eruption in September 1982; seismic and ground- deformation signatures of critical intrusive events are listed in table 7.1. A long history of instrumental observations in Hawaii relates ground deformation and shallow earthquakes to magma pressure and eruptions (Eaton, 1962; Decker and others, 1983). At Kilauea and Mauna Loa Volcanoes, gradual inflation of the summit, accompanied by an in- creasing number of earthquakes, commonly leads either to a summit eruption or to rapid deflation of the inflated summit, accompanied by swarms of shallow earthquakes along a rift zone and followed by a flank eruption. Inter- preted in terms of magmatic loading, the seismic data, substantiated by the pattern of ground deformation, outline a dynamic magma-transport system (Koyanagi and others, 1974; Ryan and others, 1981). The relative absence of earthquakes beneath deformational centers that are bounded by seismic zones may be interpreted as regions of low rigidity and concentrations of active magma. The response to magma pressure from the summit reservoir system that generally starts in the south—caldera locality extends episodically into the upper parts of the rift zone bounded by the Koae fault system to the south (Klein and others, 1987). The activity may occasionally extend farther along either rift zone as the earthquake- propagation paths are reoriented northeastward along the east rift zone and southwestward along the southwest rift zone. The Koae faults form the structural boundary that separates the summit region, responding to the movement 183 184 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 TABLE 7.1.—Mirior intrusions of magma at Kilauea, indicated by seismic and tilt data from September to December 1982 [Magma volumes from A.T. Okamura (oral commun., 1985)] Regions intruded by magma, as inferred frOm the seismicity pattern Time interval Volume of intruding magma, as estimated from summit tilt (106 m3) Sept. 26—Nov. 3 Summit to upper east rift Approx 1 zone on September 26; sustained in upper east rift zone from late Sep- tember to about October 20; migration into middle east rift zone from Octo- ber 21 to November 3. Dec. 9-31 Summit to upper east rift Approx 1 zone from December 9 to 10; sustained in upper east rift zone from De- cember 11 to 25; migra- tion into middle east rift zone from December 25 to 31. of magma in the central reservoir complex, from the lower rift zones, influenced by laterally extensive intrusions and tectonic events in the south flank (Duffield, 1975). The seaward displacement of Kilauea’s south flank, in response to magmatically induced dilation of the east rift, was deduced from deformational and seismic data (Swan- son and others, 1976). Structural subdivisions of Kilauea into summit region, east rift zone, and south flank are bas- ed on the geologic classifications proposed by Swanson and others (1976) and augmented with spatial- and temporal-distribution patterns of earthquakes. The unstable regions of Kilauea’s south flank over the past few decades of high-level volcanic activity are outlined by the distribution of aftershocks from an M = 7.2 earthquake near the south coast in November 1975. These after- shocks, concentrated at a depth of about 5 to 10 km, ex- tended laterally over a distance of about 40 km in an elongate zone parallel to the rift system and offset to the south. Tectonic models for the seaward displacement of the south flank, derived primarily from aftershocks of this earthquake, were summarized and refined by Crosson and Endo (1982). Fault-plane solutions of earthquakes and the distribution of aftershocks imply block displacement of the south flank seaward and away from the rift along a subhorizontal zone extending southward from a depth of about 10 km. Incremental displacement is apparently induced by accumulating stresses from magma intrusion into the rift zone. The south-flank slip zone lies above the oceanic crust and dips toward the center of the island (Crosson and Endo, 1982). The increasing gravitational load and southward Widening with depth to about 10 km beneath the rift axis is compensated by increasing sea- ward displacement. Dvorak and others (1986) summarized the pattern of earthquake activity and deformational events associated with magma intrusion from September 1971 to January 1983. Intrusion-related, shallow (less than 5 km deep) earthquakes beneath the east rift were immediately followed by deeper (5—13 km) earthquakes beneath the adjacent part of the south flank. On the basis of their re- cent data, Dvorak and others pointed out that shallow rift intrusions generate compressive strain and that induced stresses from deeper within the rift zone cause extensional deformation of the south flank. Volcanic tremor has been instrumentally documented in Hawaii since 1912. Shimozuru and others (1966) sum- marized various models for the origin of tremor proposed by early investigators, and evaluated their own seismic survey of tremor at Kilauea in 1963. Sources of tremor, influenced by patterns of volcanic activity, were classified according to spectral contents. Their results concurred with earlier findings that harmonic tremor is fundamen- tally related to magma movement and eruptive activity. Aki and others (1977) developed a mathematical model for the magma-transport mechanism, based on driving of magma through a system of connecting cracks by excess fluid pressure. They applied their model to the data col- lected by Shimozuru and others for the brief Kilauea flank eruptions in 1963, and proposed that harmonic tremor is generated by magma movement in a succession of episodic crack extensions and by rapid opening and closing of the narrow channels connecting the fluid-filled cracks. In the 7. SEISMICITY ASSOCIATED WITH THE ERUPTION model of Aki and others, the frequency of the seismic waves is a measure of the length of an activated crack, and the signal strength indicates the rate of magma move- ment through the system of cracks. Using this model, Chouet (1981) obtained a complete representation of ground motion in the near field of the fluid-driven crack. His calculations show the presence of a broad peak in the ground-response spectrum, the characteristics of which depend on source geometry, the bulk modulus of the fluid, medium properties, receiver position, and the component of motion. The basic assumption underlying the models of Aki and others (1977) and Chouet (1981) is that the fluid behaves as a passive element in the source; that is, no acoustic source exists in the fluid. Because of this assump- tion, the source duration predicted by their theoretical models is rather short, and the spectral peaks displayed by the synthetic seismograms are broader than those generally observed for harmonic tremor at Kilauea. Ferrick and others (1982) proposed that harmonic tremor is generated by fluctuating fluid flow. In their ex- periments with hydraulic systems, a disturbance in the steady state of a fluid system was found to cause flow and pressure oscillations in the fluid that would generate oscillating displacement of the conduit wall and elastic waves in the wallrock. A fluid system at rest or gradual- ly changing fluid flow is not expected to produce tremor. Chouet (1985) used this concept to develop a seismic model in which the fluid is included as an active element in the source. In his model, the bandwidth associated with the dominant spectral peak of motion is controlled by the com- bined losses due to viscous attenuation in the fluid and elastic radiation in the solid. When the fluid Viscosity is low, the source acts as a high-Q oscillator, and the ground motion can last a relatively long time. A more refined model of the dynamics of a fluid-driven crack was recently developed by Chouet and Julian (1985), using the same concept of active fluid participation. Our present knowl- edge of the origin of volcanic tremor can be found in Chouet and others (1987). This chapter treats two distinct seismic periods in the eruption chronology. The first part describes the early period of widespread earthquake activity of a highly stressed volcano. The second part emphasizes the activity of harmonic tremor during the later stage of the erup- tion, when the low—stress environment was distinguished by a relative absence of significant earthquakes and by intermittently high eruption rates With strong tremor. Acknowledgments—We dedicate this chapter to the entire staff of the Hawaiian Volcano Observatory, each of whom has played a special part in the volcanic research that contributes to definitive interpretation of the seismic data. Bernard Chouet guided us in our treatment of volcanic tremor. To Ed Wolfe we extend special gratitude for his persistent encouragement, patience, and assistance 185 in completing this chapter. Tina Neal and George Ulrich provided useful comments based on their many hours of field observations of the eruption. LIMITATIONS OF THE DATA The distribution in space and time of earthquake swarms in the shallow crust describes near-surface mag- matic processes, and that of deeper crustal earthquakes outlines the surrounding region stressed by intrusive activity. The dominant frequency and signal strength of harmonic tremor were measured to within the capabilities of the Hawaiian Volcano Observatory (HVO) seismic system, to quantify the source parameters relative to the eruptive process. This compilation of data is intended to serve as a preliminary guide for future detailed analyses of the seismic data associated with the Kilauea volcanic activity in 1983 and 1984. Shallow harmonic tremor associated with the 1983—84 eruption of the east rift of Kilauea varied in amplitude and frequency within the expected range as a function of time and distance from the source. Complex and erratic, high-frequency tremor accompanied the intrusive swarm of earthquakes during the early period of seis- micity and was later replaced by constant, low-frequency tremor that persisted at varying intensity during the pro- longed eruption. The principal tremor in the eruptive zone was followed by weaker tremor and long-period events at the summit. In this chapter, strength of tremor is generally described in terms of micrometers or nano- meters of ground displacement, derived from amplitudes read on seismograms at the dominant frequency and reduced according to instrumental magnification and response. Station corrections obtained from amplitude dif- ferences of local earthquakes and teleseisms were used where station-to-station comparisons of tremor amplitude were made. Although measurements were fairly consis- tent in relative terms, instrumental noise in the seismo- graph system and variations in signal attenuation unique to the station site introduced inconsistencies into the reduction of tremor to actual ground motion. In Hawaii, natural ground noise is particularly high in the frequencies about and below 1 Hz, and noise of about a micrometer in amplitude is common. This noise significantly reduces the detection capability of the HVO seismographs for any tremor in the lower frequency range. The l-s-period seis- mometer used in the standard HVO system also restricts detection of low-frequency signals. Because of structural complexities in the active crustal regions of Kilauea and the feature of harmonic tremor, detailed analysis of tremor requires an extensive expan- sion of our instrumental capability. A network of broad- band and three-component seismometer stations spaced 186 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 less than several tens of meters apart within a few SEISMIC NETWORK AND DATA PROCESSING kilometers of the critical source regions of the east rift and summit would add quality to our data base, and The HVO maintains anetwork of 50 stations covering velocity information from an organized seismic-refraction the Island of Hawaii (fig. 7.1); one station each on the program would enhance our analysis of tremor. Islands of Maui and Oahu provides additional coverage 156° 155°3o' 155° , . . , . Molokai w _ LanalD d Kahoolawe 20° — A Mauna Kea _ A HAWAII A A A 19°30, ‘ Mauna ‘ Loa A NPT Keanakakoi A (1) Crater LUAIKAH A A A Kilauea A 6‘ E85‘ A A PAU MPR - fl 0095 19° - 0 10 20 KILOMETERS ‘ |__._L______| l I n I . I l FIGURE 7.1.—Island of Hawaii, showing locations of seismic stations (triangles) operated by HVO (key stations labeled). Summit calderas are outlined for the active volcanoes Kilauea and Mauna Loa. See Tanigawa and others (1983) for details of station parameters. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION on the northwestern parts of the Hawaiian Archipelago. Seismometer density is highest in the southeastern parts of the Island of Hawaii, where seismic and volcanic activity is centered. The seismometers are mainly short- period vertical instruments operated at high sensitivity to detect low-amplitude seismicity associated with volcan- ism (fig. 7.2). Ten of the stations have three-component seismometers, with the horizontal components operated PERIOD, IN SECONDS 10 5 2 1 0.5 0.2 0.1 0.05 0.02 0.01 - I - I"'| I "l""| l":"" A ' CS.N C5,lOlilz 105? z 9 p— < 9 E 104: o . < E 103: Lu~ 3 0.62.5125 2.5 5 1o :0)“; Hz Hz Hz Hz Hz '-< l | | I I a; 1.0— — :0 an: ma 3m 0.5— _ 1-1: EN w.— (I- 015 . . . .....| . I u;.... 0.1 1 10 100 FREQUENCY, IN HERTZ FIGURE 7.2.—System-response curves for short-period seismographs operated in Hawaii during 1983—84 eruption of Kilauea. Filtered signals from a designated station were assigned to channels 1 through 5 (C1—C5), and normal (N), unfiltered signal was recorded on channel 6 (C6). A, Curve for Type 1 instruments (solid line) applies to standard HVO high-gain vertical components on a Develocorder-based FM system, and curve for modified Wood-Anderson instruments (dot- dashed) applies to lower sensitivity horizontal seismometers at three- component stations. B, Filter system with various center frequencies as shown was designed for a six-channel chart recorder to be used for spectral analysis of specific seismic events. 187 about 12 dB below the gain of the vertical component for measurements of S-wave arrival times and amplitudes. All signals are telemetered by radio to the HVO and recorded on a 1-in.-magnetic-tape recorder; some signals are also recorded on 24-hour rotating-drum recorders and 16—mm-microfilm-strip recorders. Some stations located at the HVO and Hilo on the Island of Hawaii and on Maui and Oahu are maintained independently, and their signals are recorded optically on 24-hour—cycle rotating drums. Hundreds of earthquakes with magnitudes ranging up- ward from a threshold of about 0.1 are detected each day on key stations near active seismic zones in the summit region and rift zones of Kilauea. These tiny events are classified into regional categories on the basis of estimated arrival and amplitude differences, and their hourly and daily numbers are documented as an index of the seis- micity (and state) of the volcano. Short- and long-period events are also distinguished for the summit region. The count of shallow events near detection level is especially sensitive to instrumental magnification, background noise, and reading format and thus is highly approximate, in comparison with the more quantitatively analyzed, larger events. About 2 to 5 percent of the detected earth- quakes exceed 1.0 in magnitude and are sharply recorded at a dozen or more stations. These larger events are selected for hypocenter and magnitude determination by computer (Klein, 1978), using measured arrival times of P and S waves, trace amplitude, and signal duration. Earthquakes that are timed and located total several thou- sand per year and form the primary data base for the quantitative definition of earthquakes and volcanic processes in Hawaii. The location, magnitude, and classification data for all processed earthquakes, as well as for other, distant seismic events, and the instrumen- tal information and highlights of volcanic activity are published in annual summaries of the HVO (for example, Nakata and others, 1984). Seismicity associated with the 1983 Kilauea eruption was classified into short-period (SP) earthquakes, long- period (LP) earthquakes, and harmonic tremor to provide a broad base for the seismic interpretation of volcanic pro- cesses. This classification is based on signature variations routinely relied on to differentiate seismic events at the HVO (Koyanagi, 1982): 1. SP earthquakes, the most common type, occur widely in the southeastern part of the Island of Hawaii, par- ticularly beneath the active volcanoes Mauna Loa and Kilauea. They are heavily concentrated in the crust from about 0- to 15-km depth in tectonic regions under volcanic stress. The shallowest earthquakes, between 0- and 5-km depth, generally coincide with magmatical- ly induced ground—deformation events; their occur- rence defines the locations and times of volcanic activity. The magnitude range of SP earthquakes is 188 wide; correspondingly, the magnitude-frequency parameter (Richter, 1958, p. 359) is commonly low— about 0.5 to 1.5. The seismic signature has a pro- nounced onset of high-frequency waves that attenuate exponentially over time; the dominant frequency changes systematically from about 15 Hz at the onset to less than 1 Hz at the end of the coda. High-frequency body waves are strong for deeper SP earthquakes. Low-frequency and low-velocity surface waves are seen in the shallowest crustal and distant earthquakes, and signal envelopes for these events are generally elongate. 2. LP earthquakes occur only in places of active volcan- ism and suspected magma movement, such as beneath the summit region of Kilauea. They commonly accom- pany harmonic tremor. Their seismic signature and mode of occurrence suggest that these earthquakes may be discrete events that, in some instances, in- crease in number to collectively form harmonic tremor. The frequencies of the seismic waves range from about 1 to 10 Hz and do not substantially change from the start to the end of an individual event. The signal onset is emergent and elongate in comparison with typical SP earthquakes. Magnitude is low and narrow in range, and the poorly defined magnitude-frequency parameter 1) appears to be correspondingly high, rang- ing from about 1.5 to 2.5. 3. Harmonic tremor, which is the seismic indicator of magma movement and volcanic eruptions in Hawaii, is classified into depth categories of shallow, interme- diate, and deep, depending on amplitude differences recorded on the seismic network. In general, tremor signals are sustained in duration and relatively con- stant in amplitude and frequency. In detail, amplitude and frequency constantly oscillate within a limited range at time intervals of a few to about 10 s. Shallow tremor (less than 5 km deep) accompanies eruptions; the recorded amplitude is highest in the active vent area and varies nearly in proportion to the lava-output rate. The frequency of the seismic waves ranges mainly from about 1 to 10 Hz and is sometimes superimposed on lower frequencies. Shallow tremor recorded within about 2 km of the eruptive vent has dominant frequen- cies of 2 to 5 Hz, and at more distant locations (several tens of kilometers away), 1- to 3-Hz signals are com- mon. Bursts of tremor, ranging from minutes to days in duration, sometimes occur independently of eruptive activity. Shallow tremor may accompany intrusions recorded by ground deformation. The attenuation rate of amplitude across the seismic network distinguishes intermediate-depth tremor (mostly 6~12 km) in the lower crustal region beneath the summit from deeper tremor (mostly 30—60 km) in the upper mantle that ex- tends broadly southwestward of Kilauea. The source THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 of tremor in Hawaii is further described by Koyanagi and others (1987). Earthquake P waves are timed to within 0.01- to 0.05-s precision, and in the Kilauea area, where seismometers are spaced about 5 km apart, calculated hypocenter accuracy for crustal earthquakes is commonly better than 1 km. Focal depth is referenced to ground elevation at the earthquake epicenter. The peaked instrument re- sponse means that amplification is strongly influenced by the spectral content of the event. The detection capability of a Type 1 system (see Nakata and others, 1984) varies and is restricted to frequencies from about 1 to 20 Hz. Irregularities in instrumental and ground noise also in- troduce inconsistencies into the reduction of ground motion. For tremor, relative measurements of amplitude at specific stations are consistent to within a factor of about 2, whereas analyses dependent on absolute measurements are confined to order-of—magnitude calculations. Uncer- tainties in the reduction of tremor signals to actual ground motion is introduced by variations of tremor frequency from 1 to 10 Hz, particularly during eruptions, when higher frequencies are detected near the source than at locations more than several kilometers away. In our general procedure, the repeating bursts of amplitude max- ima were read and averaged for 5-minute samples from Develocorder film records, and reduced according to instrumental magnification and apparent frequencies averaged for the amplitude bursts. Station-to-station dif- ferences in background noise and signal amplification were calibrated by comparisons of local-earthquake and teleseismic signals and taken into account in our reduc- tion of tremor amplitudes. To provide more nearly uniform detection capability for earthquakes and harmonic tremor over a wider spectral range, a system based on Wood-Anderson response was recently adapted to the horizontal components of selected stations (fig. 7.2). This system was installed in several frequency-modulated (FM) stations at Kilauea beginning in mid-1983, to improve amplitude measurements used for the determination of earthquake magnitude and tremor readings. A filter system designed by George Kojima at the HVO to record seismic signals at frequencies centered at 10, 5, 2.5, 1.25, and 0.625 Hz was adapted to a six-channel chart recorder. This system is comparable to that used by Bernard Chouet (Chouet and others, 1978; Chouet, 1979) to collect earthquake data near Stone Canyon, Calif. Different stations were monitored at different times, and normalized with unfiltered signals from continuously recorded network stations on Develocorder films or magnetic tape. Tremor, as well as other seismic events associated with the eruption process, was monitored at various stations over different intervals of time. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION DATA, 1982—84 LOCATIONS AND MIGRATION PATTERN OF EARTHQUAKES In the past few years, frequent magma intrusions and eruptions at Kilauea have resulted in high levels of seismicity in the rift zones and adjacent flanks (Nakata and others, 1982; Tanigawa and others, 1983). Similarly, earthquakes in the southeastern part of the Island of Hawaii were numerous in the months before the 1983 eruption (fig. 7.3). Of these earthquakes, crustal events at 5- to 10-km depth in the south flank of Kilauea, believed to occur from seaward displacement of the unbuttressed flank in response to rift intrusions (Swanson and others, 1976), were most persistent. Shallow summit and rift 155° 1 1 fl . I . . . I . . . A EXPLANATION _ Depth, in kilometers 20° — + 0.0 — E] 5.0 O 13.0 A 20.0 Magnitude 0.0 19°— 0 155° EXPLANATION — Depth, in kilometers 200 _ + 0.0 A U 5.0 O 13.0 A 20.0 Magnitude ' 0.0 D40 ' 19° ” o 20 KILOMETERS Q4 I I I I l I I _ I I I l I 189 earthquakes at 0- to 5-km depth occurred during times of sustained inflation and intrusions (fig. 7.4; table 7.1). On the basis of earlier findings (Koyanagi and others, 1974), the shallow events from 0- to 5-km depth delineate an inflation center at the summit and outline linear in- trusion zones projecting southward, southwestward, and southeastward into the rift zones. The deep crustal earth- quakes from September to December 1982 were mainly concentrated in a 20-km—long zone adjacent to the intrusive area in the east rift. The sequence of shallow earthquakes plotted in the summit region indicates the progressive shift to and increase of activity in the upper east rift zone from September to December 1982 (fig. 7.5). The combined plots present an arcuate alignment of hypocenters, with linear zones radiating from the south edge of the caldera near Keanakakoi Crater to the south 155° I I I I I I I I I I I I 1 EXPLANATION _ Depth, in kilometers -+- 0.0 — D 5.0 200 _ HAWAII 0 ”-0 A 20.0 Magnitude 0.0 _ I I0 u 2.0 - CI 3.0 E) 4.0 ‘ D 5.0 I: 0.0 _ 19°— 0 155° EXPLANATION Depth, in kilometers 200 _ t 0.0 — D 5.0 0 13.0 A 20.0 Magnitude 0.0 .I I0 [I 2.0 _ CI 3.0 I: 4.0 ‘ E] 50 D60 _ I I I | | I I I I I I I I 19" ' o 20 KILOMETERS L_|_i FIGURE 7.3.—Locations of earthquakes on the Island of Hawaii for September (A), October (B), November (C), and December (D) 1982. Solid lines, fault traces. 190 that trends into the southwest rift zone, and to the southeast along the upper east rift zone. The increasing rate of earthquakes before and during the early eruption period in January 1983 broadened the seismic zone along the east rift (fig. 7.6). The earthquakes associated with shallow intrusion along a linear 15-km increment of the east rift, the earthquakes associated with shallow collapse at the summit, and the deep crustal earthquakes that are broadly elongate in the south flank are distinguished in this widened seismic zone. The hypocentral distribution of earthquakes from the successive intrusions that outlined the dynamic regions during September 1982 to January 1983 is expanded in map and depth views (figs. 7.7, 7.8). Shallow swarms of earthquakes were concentrated at the summit and along the axis of the east rift. The shallow seismicity beneath the southern caldera area radiated laterally to feed a 3-km- 155°30’ l' | A fir EXPLANATION " Depth . in kilometers + 0.0 Mauna Loa I: 5.0 — an: n Kilauea 2 , ° DOD a, . . / a ” Eb” a 33 / a. an)? ° Dd? a 7 u x/ // - .. D El 5‘ (j o 13.0 ++ + A 20.0 Magnitude 19°20' — - 0-0' 1.0 2.0 155°30’ | ‘ l l EXPLANATION Depth. in kilometers + 0.0 o 13.0 A 20.0 Magnitude 1 9°20' — . n 1.0 / O 10 20 KILOMETERS |__l—l l I l l D 5.0‘ 0.0 ‘ THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 long protrusion to the south and an 8-km-long alignment to the southeast that connects with a 17-km—long exten- sion along the axis of the east rift. The shallow earthquake swarm formed a 2- by 25-km zone centered from 2 to 4 km in depth. Nearly parallel to the rift was a broad zone of deep crustal earthquakes in the adjacent south flank of the volcano. The 30-km-long and 5-km-wide belt of earthquakes were concentrated at depths between 5 and 10 km; the lower limit was near the crustal boundary. The elongate seismic zone that parallels the summit-rift axis is formed by a multiclustered distribution of earthquakes. Regions of decreasing numbers of crustal earthquakes (0—13 km deep) extended obliquely to the major axis of the east-rift and south-flank seismic zone (fig. 7.7). Some of the noticeable gaps occurred across (1) the middle up- per east rift, (2) the bend in the east rift, and (3) the mid- dle east rift south of Puu Kahaualea. 155°30' 155° 1 ' ' I EXPLANATION ‘ Depth. in kilometers 0 + 0.0 ,, U 5.0- , , .15 013.0 - éMauna Loa o [in i, 0 EEE . o Kilauea / u D u@ ./. n / ‘ . ~ “ ° / '. _: A 20.0 Magnitude - 0.0 ‘ 1.0 19°20' 155°30’ EXPLANATION Depth. in kilometers + 0.0 l l l D D 5.0 ‘ O 13.0 Mauna Loa A A 20.0 Magnitude - 0.0 _ u 1.0 19°20’ — c 0“" vgcgix“ 10 20 KILOMETERS l—L_i l I | | FIGURE 7 .4.—Locations of earthquakes in the Kilauea area, southeastern part of the Island of Hawaii, for September (A), October (B), November (C), and December (D) 1982. Solid lines, fault traces. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION The January swarm of east-rift earthquakes migrated downrift, initially at a rate of about 0.6 to 0.7 km/h and episodically thereafter (fig. 7.9). The well-defined linearity formed by the concentration of earthquakes along the leading edge of the space—time diagram during the early morning hours of January 2 defined the downrift migra- tion and rate. The earthquakes then moved farther down- rift erratically until midday. The earthquakes continued within the seismic zone for about a day after the initial migration. From about the middle of January 3, earth- quakes were scattered even farther downrift almost to Kalalua, and then were sustained along a zone between Puu Kahaualea and Napau Crater, assuming a slow, uprift migration pattern at a rate of about 0.06 to 0.07 km/h until the middle of January 6. The final swarm on January 7 extended farther downrift to about 2 km east of Kalalua, again at a rate of about 0.6 to 0.7 ' km/h. 155°18' 155°14' | I A DU’LANATION Depth, in kilometers + 0.0 E! 5.0 19°25’ - 0 130 — Q A 20.0 / / Magnitude // " °'° - u 1.0 D 2.0 19°23’ — .. El 3.0 _ HIIaka El 4.0 o 1 KILOMETER ‘ ® D so i__l ' D I . I I 155°18' 155°14’ | / ' \ | C EXPLANATION Depth, in kilometers + 0.0 o , o I: 5.0 19 25 Kilauea D 0 13,0 — 20. O {5 eanakakoi A 0 Magnitude ‘ 9/ V ‘ + . ' °'° ‘ g. u 1.0 s. . H“ + ‘1: t: D :1 2.0 19°23' — g. .. +2 25“ 1:1 so _ ‘ ‘ i * E! 4.0 0 1 KILOMETER U .. .. HIIaka D 5.0 AA D D D 6.0 191 DEPTH AND MAGNITUDE OF EARTHQUAKES From September 1982 to January 1983, earthquakes located beneath the Island of Hawaii were mainly confined to depths of less than 20 km and had magnitudes of less than 4.2 (fig. 7.10). The 5- to 10—km-deep zone, most noticeably on the south flank, persisted throughout this time interval as the major source of seismic-energy release. These earthquakes increased in number above the already-high background mostly during the early period of eruption from January 2 to 8. Many strong earth- quakes, as large as about M :42, fell in this depth category at seemingly random intervals. Earthquakes shallower than 5 km were abundant beneath the summit and east rift during intrusions in the months before the eruption, as well as in the early erup- tion period. The east-rift intrusions in late September, 1 55°14' ' I EXPLANATION \ Depth, in kilometers + 0.0 155°18’ El 5.0 19°25' _ o 13.0 A 20.0 Magnitude 19°23' 0 1 KILOMETER Q 1 55° 1 4' l EXPLANATION ‘ Depth. in kilometers + 0.0 155°18' 5.0 19°25' U — <) 13.0 A 20.0 Magnitude v 0.0 - u 1.0 19°23' , . +3.“ 0 1 KILOMETER ‘ L___l ‘ | . I . I FIGURE 7.5.—Locations of earthquakes beneath the summit region of Kilauea for 1982. A, 1650 to 1845 H.s.t. September 25. B, 1845 H.s.t. September 25 to 1500 H.s.t. September 26. C, October. D, December. Solid lines, fault traces. 192 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 156° 155° I I I I I I I I I I l l I EXPLANATION Depth, in kilometers 20°- + 0.0 -— El 5.0 0 13.0 A 20.0 HAWAII D Magnitude I 0.0 n 1.0 D 2.0 — CI 3.0 |:| 4.0 ‘ E] 5.0 19° ’ o 10 20 KILOMETERS D 6.0 ‘l ' . I . I . I . I I . . . ' 1 55°30’ 1 55° I ' ' I r a _ B EXPLANATION Depth. in kilometers D + 0.0 19030! _ Mauna Loa U u 5.0 _ . ° 0 13.0 *9- a? nu . U A 20.0 g: ”$3“ / m o Magnitude . 0.0 0 L0 F‘C PACl N 2.0 OCEA 30 Cl C] I] 4.0 20 KILOMETERS [:1 5.0 19°10'— 155°18’ 155°16' 155°14' I I . EXPLANATION ‘ Depth, in kilometers + 0.0 U 5.0 O 13.0 A 20.0 19°25' Magnitude I 0.0 ‘ I: 1.0 I D 3. 19°23 Hiiaka ‘ o 1K|LOMETER ® D M g4 I I I I I l I FIGURE 7.6.—Locations of earthquakes on the Island of Hawaii for January 1983. A, Entire island. B, In the Kilauea area. C, Beneath the summit region of Kilauea. Solid lines, fault traces. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION December, and early January, marked by swarms of shallow earthquakes, were followed within a few days by more subtle increases in the number of deeper crustal earthquakes. RATE AND FREQUENCY OF EARTHQUAKES The cumulative numbers of timed and located earth- quakes larger than M =1.5 are plotted in figure 7.11. Shallow summit and rift earthquakes at 0— to 5—km depth selectively are concentrated at times of sustained sum- mit inflation, summit collapse, rapid rift intrusions, and eruptions. The number of shallow summit and rift earth- quakes of M>1.5 appreciably decreased after the erup— 193 tion in January 1983. Changes in volcanic activity during the later eruptive episodes were reflected seismically only by changes in the number of very small earthquakes, and by variations in the amplitude of shallow harmonic tremor. The rate of deeper south-flank earthquakes, however, was contrastingly high and constant. The relatively high rate during the period of episodic intrusions from September to December 1982 led to a short acceleration in rate in January during the early eruption period, followed by a constant lower rate during the prolonged eruption period in 1983. The number of earthquakes in the south flank decreased further after January 1984. The daily number of small and shallow earthquakes counted in the summit and east rift zone generally responded to the volcanic activity associated with the 1 9°30' 155°20' 155°10' 155° _ . . . . . . . . . . . . . . . . . . . 1g. ’52- ' ”6 _ c/ _ «3 c <> cc ’\,/ ' Cl Kilauea summit cc 7:, '7 7:; h ‘ reglon \ 71 5 o ' 9 a we “4 i o 9/ - °‘ “3» 3‘0 w .- 19°25' 19°20' 19°15’ ' l . . . . . . . I 10 KILOMETERS 4] FIGURE 7.7.—Summit region and eruption area along east rift zone of Kilauea, showing earthquakes less than 15 km deep located with horizontal and vertical standard errors of less than 2 km (see fig. 7.3 for explanation of symbols) and aseismic regions (crosshatched areas) for period September 1982 to January 1983. East rift zone was sometimes separable into upper and middle parts, according to the structural classification of Swanson and others (1976); upper part ex- tends from southeast of the summit region to about Makaopuhi Crater, and middle part includes area from Makaopuhi Crater to Heiheiahulu. Lines A-A', B-B’, and 00’ refer to cross sections normal to middle east rift zone, lines D-D’ and E-E’ to cross sections normal to southeastward trend of summit to upper east rift zone, and lines R-R’ and S—S’ to cross sections along east rift zone and summit to east rift zone, respectively (see fig 7.8). Solid lines, fault traces. Major craters are labeled. 194 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 A . 0 C to I: LlJ ’— I.“ 2 5 — _ O :’ x E ~ U I _ _ '_ 10 n. LLI O l l l l l l | I I O 5 10 15 0 5 10 15 D D' E O +"l 0 4 ++ I t (I) 0 . + I ++ 4 i + L“ l— + 4* E + + D O 65 D u :1 n D x z D D — n D D i 10 — — 101g] — 1o — El ’— 5 u D H 0. Lu 0 I I I 0 l R I 10 O 5 O 5 O 5 10 EXPLANATION R’ Depth, in kilometers Magnitude + 0.0 n 0.0 w E D 5 0 u 10 ,_ . . g —. 13.0 2.0 o O u :1 x A 20.0 [:1 3.0 E . _ [:1 4.0 I p— m l:] 5.0 Lu 0 D 15 0' I | 0 10 20 30 LENGTH, IN KILOMETERS FIGURE 7.8.—Depth sections of earthquake locations in summit region and east rift zone (see fig. 7.7 for locations of section lines). Each section represents a volume of length and depth as indicated and of the following widths: sections A-A ', B-B’, 00’, and S-S’, 5 km; sections D-D’ and E-E', 4 km; and section R-R’, 10 km. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION major eruption that started in 1983 (fig. 7.12). The rate of shallow SP earthquakes that delineate the level of stresses around the summit storage system increased to more than a thousand per day during the summit erup- tion and intrusion in September 1982 and before the east— rift intrusion in December. Later, during the eruption period in 1983 and 1984, the rate of these events ranged from less than a hundred to several hundred per day. The increasescoincided with sustained summit inflations or large, eruption-related deflations; alternating decreases occurred after relatively small deflation events between eruptive episodes. The rate of shallow SP earthquakes delineates the level of stress around the summit storage system. LP caldera earthquakes followed eruption-related deflation of the summit. These events emerged from con- tinuous, intermediate-amplitude harmonic tremor that began during large and (or) rapid deflations; the varia- tion in rate partly reflected the level of continuous tremor at the station NPT. LP events and (or) summit tremor more consistently responded to the amount and increas- ingly rapid rate of summit deflation associated with erup- tive episodes after the first half-year of activity. The pattern of summit tremor and LP events is further 195 described below in the subsection entitled “Seismic Events at the Summit.” During the period before the 1983 eruption, the daily number of east-rift earthquakes dominated by swarms of shallow rift earthquakes marked the intrusions in Sep- tember and December. The total counts have been affected by scattered larger earthquakes in the south flank. After January 1983, many small shocks occurred from various thermal and structural anomalies at the eruptive vent, including degassing and cooling of fresh lava flows. After the summit eruption in September 1982, the number of shallow earthquakes increased in the upper east rift zone, and minor swarms occurred in September- October and December. In both swarms, shallow earth- quakes moved from the summit caldera rapidly into the uppermost east rift zone between Puhimau and Kokoolau Craters. The activity remained in the upper east rift for 2 to 3 weeks and was followed by farther-downrift spreading at a gradual migration rate of about 0.7 km/d (fig. 7.13). The episodic intrusions of magma indicated by the pattern of seismicity extended from the summit to beyond Makaopuhi Crater. 20 I I I I I | / '0'. + ' :2 ‘ ° 9 Lg . . ” E Kalalua ““3. + - , :2. 4-9; . 1., O _ + ++ S +‘ + d 15 + 5/ +‘:| _ ¥ + *+ + + + ‘2 +. E Puu Kahaualea / . . + o ’ D l‘IZJ + *I‘ + 0* ‘ ' .+ <09/ 0 4, f: + . +¢ 0' N ’4 + + o ! ¢+ 5;: . + '- Puu Kamoamoa *' \ * . + d“: + + + * Li + 4- 5 M + \ 1 . 4' D: 10 _ 0 #‘H' .4; \\+ . ‘4?“ ' " I'- + + + + \ o m + o: O\+\ F... o :5 +414. + + 06‘0 O\\ + 4- ° * u 0 0 g $14.0 7 kM/f:\\ ‘ . + 00 u z Napau E! i,» 431‘ s .\ 9 t 4:!- . \+\ < + . o \ 8 5 o + %: ++t+ <2): {ht-E Makaopuhi - * O O 5 a: . w 3. ++. ” Q _ + o - D 0 / I I I I I I 1 2 3 4 5 6 7 DAYS IN JANUARY FIGURE 7.9.—Earthquake locations (see fig. 7.3 for explanation of symbols) along a west-to-east profile of east rift zone of Kilauea as a function of time for the period January 1-7, 1983. Distribution of shallow earthquakes during swarm on January 2 indicates a downrift migration rate of 0.6 to 0.7 km/h (solid line). Thereafter until January 7, distribution implies uprift and, subsequently, downrift movement (dashed lines). Major craters are labeled. 196 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 I I I I I'I I I I I I I 1+I I I I I I I I I I I I 1;! II; + . . x. :4 1 . 3’ “o + ““6“: n+0 a Hat u . + a + . d)” on n l ‘ ”’ op u u up D D 9: ‘13:; n " ¥n°unDE%Uu:5 0° mfigfl’mé a? “Q Eng.” a“; 9”“ .‘I‘Punu' 10-2 m Engfiu gawk “ 5‘ .n “E D U C] ”g u I: o O A o 20 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 5 1O 15 20 25 30 SEPTEMBER 0 I I TI I I I I I L1 I I I I I I |"'1 l I I T I I I I I I I “ . . 3 k. . 3. + '9 w alienated». gm 1'42”» flaw. '1 ~ "t *I h we . '* ‘ I: 4 I: 3 D; an“ Eng 9.5% C: Du ”a an“; ”a u 1:: Q7 5‘ E n Saga" nfiu.§£;% agnsnnnrtimmngaunggn'b ° $‘Li mg Eguugmufinfi°u$ a u u :1 ° 10 1,” ED ° rt El: ”DD 5’ D“ n °%E€D ‘3 ° 15] ‘ o I? I: B o <> 0 20 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 5 1O 15 20 25 30 OCTOBER 2 0 I I I I I l+l I I [Tl L I I I I I I I I rI I I I I I I Lu ~ w . . . v 2 mm m n _ %a nu Rn” n ”in a u m n “:5 E a o I u u M anal.” c5“ “Weenafiafi u‘g “9 “ @Bumfl’ma a —’ 3 El 5: #133 can (I a“ an [F E] D a U Q 10% I, a End—1533 E] a an El: 5;, tb ° D $3 2 D D _ o (f E o % o (L C I.I.I Q 20 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 5 1O 15 20 25 30 NOVEMBER 0 I I I I I 1.1 T ksl++l I’l’l I I I I l»l I lhé If] I I I “it" + + ,.. ' g n, , ‘dhfi,¢:f§ 3:: ”I. .viwpph Q a: 1?}; + o‘ ¢ + O , a "an :q’uq':+ a I: a n “DUE a . C! p a an n ad: EU a flu D a gym-balm E: ”a “3’3: «a 5515"”: a" a “135“ng “i5? 5% M 10.. n U “a o a Cl “ D u C] o D 20 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 5 1O 15 20 25 30 DECEMBER 0 1‘91 ['10] I~.,I I.I I I I I I I I |"["l l‘l I I I I I I I II ”I ‘ ~..; ' , . I: +‘ + s O o9 : :9 O gawuflb "—‘ giudsnflu‘naa' ' a Do I: u ‘7 F I! 51.3qu5 “a an "a 8 on“ DD“ a “I, . flfiaughuguo .“fiy; E- Tonga EU an I: a n D ".51qu a am an o‘w- c» -' q: 10. a; I grub-4° an an -: a: fig: I, .5?“ each “u a U [:1 u - n E (E? D u . a o 0 e E ° ° 20 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 5 10 15 20 25 30 JANUARY FIGURE 7 .10.—Chronologic sequence of earthquake hypocenters beneath the Island of Hawaii in terms of focal depth and magnitude (see fig. 7.3 for explanation of sym- bols). A, September 1982. B, October 1982. C, November 1982. D, December 1982. E, January 1983. 7. SEISMICITY ASSOCIATED WITH THE ERUP’I‘ION The large summit collapse accompanying the January 1983 intrusion produced an anomalous increase in the number of shallow summit earthquakes. In contrast, deep crustal earthquakes were separated randomly in the south flank, except for slight postintrusion increases adjacent to the downrift margins of the intruded zone. STRONG TREMOR IN THE EAST RIF T ZONE DURING ERUPTIVE EPISODES The continuous tremor recorded during magmatic in- trusions and eruptions varied in amplitude according to the apparent rate of magmatic movement and the vigor of the eruption. During the first week of activity in January 1983, high-frequency (5—10 Hz) tremor was very strong beneath the extensive fissure system. The center of maximum tremor accompanying the downrift move- ment of eruptive activity is shown (fig. 7.14) by the relatively larger amplitude of tremor at station MPR on January 2 and, later, downrift at station LUA on January 7 —8. Tremor was strongest during this interval, and the amplitudes recorded during subsequent major episodes were considerably lower. For most of the rest of January, tremor was highest at station PUK, and amplitudes varied generally with changes in eruption activity. The downrift migration of earthquakes, as described previously, is indicated by the hourly count of shallow earthquakes (fig. 7.14). As the eruption continued and lava emission became confined to a single vent system, tremor responded more obviously to the eruption itself and accompanying sum- mit deflation, and less to underground movement of magma between the summit and the site of eruption. During episodes of high lava output, the tremor level in- creased by a factor of at least 10 over that defined by weak background tremor between periods of vigorous eruptive activity (fig. 7 .15). The hourly variations in the amplitude of increasing tremor at the east-rift eruptive site and sum- mit region, the times of major lava outbreaks, and the pat- tern of rapid summit deflation during eruptive episodes 2 through 23 are plotted in figure 7.16. Amplitude changes of the east-rift tremor were generally more gradual dur- ing the early episodes of the eruption than during the later ones. From episode 7 on, changes in amplitude were more abrupt, especially at the start and end of eruptive episodes. Amplitude changes ranged from a slow rate of FIGURE 7.11.—Cumulative number of M >1.5 earthquakes beneath Kilauea from September 1, 1982, to June 15, 1984. A, Shallow (0—5 km deep) summit and east-rift earthquakes, recorded between lat 19°22'—19°28' N. and long 155°14'—155°19' W. B, Shallow (0—5 km deep) east-rift and south-flank earthquakes, recorded between lat 19°16'—19°28' N. and long 154°57'—155°14' W. C, Deep (5—15 km) east-rift and south-flank earthquakes, recorded between lat 19°16'—19°28’ N. and long 154°57'—155°14' W. 197 several percent per hour to rapid changes exceeding an order of magnitude within a few minutes. The pattern of tremor amplitude and duration changed approximately from a continuously long duration of low amplitude, as in episodes 2 and 3, to shorter durations of higher amplitude in later episodes. 400 I, A <—lntrusion 300 +— Summit eruption 200 100 '— — 400 I I I _. e ru ptio n 300 - ‘— Intrusion< NUMBER OF EVENTS 1600 - C _ 1400 1200 1000 -— _ 800 -— _ 600 — _ 400 - — 200 — _ o l I I I I l l I I I I l I I l | | l l I I SEPT NOV JAN MAR MAY JULY SEPT NOV JAN MAR MAY MONTHS IN 1982-84 198 SUMMIT TILT, IN MICRORADIANS NUMBER OF EARTHQUAKES SUMMIT TILT, IN MICRORADIANS NUMBER OF EARTHQUAKES 120 80 40 _. O O CO I _. O O OO 00 IIIIlIIlIIIIlIIIlll 100 120 80 40 O O CO 100 0 1 000 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 llIIIIIIIIIIIIIIIIIIllIllIlllllllllIlIIIlllllIIllIlIIIIIIIIIIIIIIIIIIIIIIIIIIlIIIII|lIIIIIIIIIIIIllIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIilllllllIIllllllIIIIIIIIIIIIIIIIIIIIIIIII W A W ,4 _..___.‘~_..- _._~_ .. __/"‘ M <— Deflation lIIIIlIIIl I I SP LP IIII'IIIIIIIIIIIIIIIIIIIIIIT JULY AUG SEPT OCT NOV DEC IIHIIIIIIIIIIIIIIIHIIIIHIIIIIIIIII”IIIIIIIIIIIIIIIIIIIIIIII“III”IIIIIIIIIIllIllllllll“IIIII‘lIIIlIIIIIIIIIIIIIIIlIIHIIIIlIIHIIIIIIIIIIIIIIIIIIIIIllIIIIIIIIIIIIIIIIIIIIII” _ B _ _‘ L j SP — ‘ r- _ LP I _ A I L . _ UER ' JAN FEB MAR APR MAY JUNE ‘— Deflation FIGURE 7.12.—Daily number of crustal earthquakes in summit region and upper east rift (UER) zone of Kilauea, refer- enced to summit tilt. A, July-December 1982. B, January-June 1983. C, July-December 1983. D, January-June 1984. Shallow summit earthquakes, recorded mainly at station NPT, were classified into short-period (SP) and long—period (LP) events, according to their characteristic signatures. UER earthquakes, recorded at stations KMM, MPR, or PAU, were shallow rifting events west of Kalalua and deeper crustal events in adjacent parts of the south flank. Earthquakes as small as M~O.1 were detected and included in counts. At this level of detection, counts are sensitive to variations NUMBER OF EARTHQUAKES SUMMIT TILT, IN MICRORADIANS NUMBER OF EARTHQUAKES in noise levels caused by weather conditions and volcanic activity. The generally lower number of earthquakes counted during major eruptive episodes is partly attributed to the locally higher seismic background from harmonic tremor. Seismic record was sometimes interrupted by instrumental failure, high winds, and heavy rains. Summit tilt derived from hourly readings of the Ideal Aerosmith east-west-component tiltmeter at Uwekahuna is approximately due to such effects as instrumental drift, climatic interference, and strong earthquakes (Arnold Okamura, written commun., 1986). SUMMIT TILT. IN MICRORADIANS 120 -- 8 I ‘5 I O 7. SEISMICITY ASSOCIATED WITH THE ERUPTION ll||llllIllIIIIlllIllllllIlll|IlllllIIIIllIlIIIIlIllllllIllllIIllllIllIIIIIIIIIlIIIIlII|l|IIHIIlIIlIIIIIIIIIIIIIIIIIIlIllIllIIllllIlIlllIllIIllllIIIIIIIIIIIIIIIIIIIIIIIIIIIIIlIll 6.6 magnitude \ M mW C .— earthquake - <— Deflation _. O O O I 100 _. o o o I OO O O llIlIllllIIlllIlIIII SP I_Illllllll 120 — (I) O I 8 I O JULY AUG SEPT OCT NOV DEC lIllllllIIIlllIIllIIIIlIIIIllIIllllllllIIlIlIIlllIllllIIIIIIIllllllIllIlIllllIlllIIllIIIIIIllllIIIIIIIIIIIIIIllIIIIlIIII|IlllIIIIllllIllIlIllllllIIIllIlllllllIIllIlllIllIIlllllllll WWW IIIIIIIIIIIIIII[IIIIIIIIIIII D _ I .— Deflation I 3 o o I 100 CO lIlIIlIIIlIIlIIIIlIIIIIIIIIII SP UER JAN FEB MAR APR MAY JUNE IllllIl‘lIllIlIIIIl 199 200 PROJECTED LOCATIONS OF VOLCANIC FEATURES KILAUEA HALEMAUMAU PUHIMAU PAUAHI MAKAOPUHI PUU KAMOAMOA KALALUA SEPT OCT NOV DEC THE PUU 00 ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 EXPLANATION Magnitudes Depth:0-5 km - 0.0 I 1.0 I 2.0 I 3.0 l 4.0 .5.0 Depth: 5-15 km 0 0.0 o 1.0 o 2.0 o 3.0 O 4.0 O 5.0 WEST-EAST DISTANCE, IN KILOMETERS MAR JAN FEB FIGURE 7.13.—Earthquake locations along west-to-east profile from summit region to east rift zone of Kilauea within area of figure 7.7 during the period September 1, 1982, to March 31, 1983. The gradual changes in the amplitude of east-rift tremor during early eruptive episodes were accompanied by com- paratively slow changes in tilt and an absence of summit tremor. This pattern advanced to more abrupt and vigorous tremor and tilt activity. Increases in summit tremor were more conspicuous during later eruptive episodes, characterized by rapid deflation events. These summit events are further discussed below in the subsec- tion entitled “Seismic Events at the Summit.” Erratic high—frequency tremor during the intrusion and early eruption period changed to more nearly consistent tremor during later eruptive episodes, when the dominant frequencies constantly alternated from about 1 to 10 Hz at short increments of time ranging from about 1 to 10 s. This pattern of repeated changes in frequency is similar- ly observed for deep tremor episodes that are not asso- ciated with eruptive activity. The absence of correlation between recurring amplitude bursts recorded at different frequencies, particularly at 2.5 and 5.0 Hz (fig. 7.17), in- dicates that the tremor is constantly changing in fre- quency within at least a limited bandwidth. The resulting chain of amplitude bursts most obviously recorded at 2.5 Hz resembles the behavior of high lava fountains, which consists of similarly repeated pulses. Amplitudes read from the bandpass signals at stations MPR and PUK for two selected 5-s intervals, corrected FIGURE 7 .14.—Seismic activity and eruptive events from January 1 to 11, 1983. A, Summit tilt recorded on east-west component of Ideal- Aerosmith tiltmeter at Uwekahuna station (see fig. 7.12). B, Tremor amplitude at stations N PT, MPR, PUK (clipping level, 3.0), and LUA (clipping level, 3.8). Tremor amplitudes were measured and earth- quakes counted for hourly intervals. Tremor amplitudes were reduced to approximate units of micrometers, according to instrumental response for recorded frequency of tremor. C, Number of earthquakes in summit region, along east rift zone from Makaopuhi Crater to Puu Kamoamoa, and along east rift zone from Puu Kamoamoa to Heiheiahulu. Most earthquakes were shallow (less than 5 km deep), small (M 30.5) events. D, Duration of increased eruptive activity. TILT, IN MICRORADIANS TREMOR AMPLITUDE, IN MICROMETERS NUMBER OF EARTHQUAKES _. O 0 0'1 0 10 10 7. SEISMICITY ASSOCIATED WITH THE ERUPTION ..L111.. IIITI Station NPT .uuuln. I'lllllll Station MFR Station PUK Station LUA ....l..|n .l..nn l'llllll Summit region Makaopuhi Crater to Puu Kamoamoa Puu Kamoamoa to Heiheiahulu IIIIIIAII IIIIIIIII 4:4..lnunu 4 6 8 DAYS, IN JANUARY 1983 10 ‘— Deflation .— Downrift <—— Downrift 201 202 THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 for instrumental magnification and background noise, in- amplitude decreased as a function of distance from the dicated varying amplitudes at 0.625, 1.25, 2.5, 5.0, and eruptive vent (fig. 7.19). High-frequency signals were 10.0 Hz (fig. 7.18). High-frequency signals at 2.5 and 5.0 strong in a wide area during the intrusion and early erup- Hz were strong at station PUK, located within 1 km of tion period, but during the later period of centralized erup- the eruptive vent, Whereas lower frequency signals (max tion, tremor amplitude decreased more with distance, 2.5 Hz) were more significant at station MPR, located 6 owing to the shallowness of the source and intense foun- km away. Such measurements represent partial short- taining of the lava. term variations in tremor frequency that collectively Amplitude readings used to determine distance attenua- produce a wider, multipeaked spectrum of frequencies. tion of tremor were first corrected for instrumental Spectral analysis of tremor sampled about 1 km from the magnification, background microseismic noise, and varia- eruptive vent on August 20, 1984, showed varyingly tions in signal level attributed to local ground conditions peaked frequencies from at least 1 to 10 Hz (Chouet and and instrumental installation at each station. Teleseismic others, 1987). Between major episodes of fountaining, the signals from a sufficiently distant source that would nor-_ commonly recorded signal at 2.5 Hz on the vertical- mally be recorded with equal strength across the entire component seismometer at station PUK or KMM served HVO seismic network, and deep local earthquakes, were to differentiate weak tremor from background noise chosen to calculate station corrections. The amplitude of created by lower frequency ocean surf and higher fre- background microseisms for each station was subtracted. quency wind. The station-to-station differences in amplitude from the In addition to changes in tremor amplitude resulting teleseismic and deep events remaining after reduction of from variations in intrusion and eruption rates, tremor background microseisms and instrumental magnification A B lwumw‘lWM am ‘ WM V mwimmw 2.1mm. ‘ v 200* immimwmmmmmmm , , 1000 2 1,111,.w-.,;,,1~p~,m,11y1ww “illilw'l‘ W11“ .mm ' WW ‘ my. «Emmi-I140.mini(FM-{Mia 11-1 mfmhfllflil‘lilihlf ~ aw 11,111 wt;mull-1w» 11 W 2‘ ; 111111111111101|W111 r ‘ 11 Mini (11,11 ‘1 1 l“ 1‘ J “I l11 ‘ v v (3.14 M.“ r. I 1, ., H 1200 .1 . 11.111111111111111» war 11 1‘ ”111.1111 ”My“ >I MH- ‘1 J h . w MIA, M11, 1‘"§“""i":‘tl’l ., 1 1' . , . 11,111.11;.Iw.:.,.4yuli,w}mflwMW,» . 31; 11w,mincwww- \IA‘,-“V‘l‘llfl'1”~l', 1 'u 171 n‘wv'.',\|1,"lw1fl,vb M «(“111qu i."imfln‘mvniiiu.li1.a,1 . 11-va \' 1:111 l' ‘ 1‘ WWlililr‘lw‘lw‘"1‘ w W {M 11mm. N 1mm ”11.1mm “ my!“ ygim'il‘llly‘fifluw , ‘ 11-511};va ' 1 400 r-J‘Mwwv ‘1" 1y?» . . J“ n 1. r ,.: 0200 l .‘IlrJl ,, M'Wn, , 1.1 ‘ F l '1 W 111m- NW ,1 'v“11‘-111‘.~0‘i MAM/W» vrhlew‘lfiw, m. 111.111“, v,»,MM~1.yI,-\HM ‘ WM]? Mil‘mr'n‘f‘fl 5 ‘JW N11-1rifil'fl,.‘ir-wtw v ,‘nflwlw‘wi 1' 14.”,lelw ui W. lliI1H"‘1,’fl'l“i.nmrWI, rill-11,114“. "M 1100115)” ‘M E x w“ 1“. m Mmhw H'vy‘hllm.‘l;‘um J1, "MN ,2 1600 , H‘1.111.111)#1111101 W ‘li‘lv.i.‘1.{lum A 1‘ "31‘" 1““ \ Wm”! 11‘ ,i- m" 1311] ‘11] 1: i )7 1 800 Wmmmwmmmwmmmmm WWWH'M-WWWIWWwwmewWi . ,MHWWMMWWHWMw MW C dWMWMMM ‘vwmwwwmmwww-‘ w.“ 1::.';;.a .1, a, . a «a.» 1.; 1‘ v m 1 11114 L .11 1‘. HIM 1 L 1 ”650 " MW at" . wk. ,1, 1200 ——————_V.——.——-————v——~ 1400 W— 1 MINUTE 7“" I_——__l 1600 —/- 1800 [Hi w, w. 1.. 2000—.—~————-———————— ___.. FIGURE 7 .15.—Parts of seismograms from rotating-drum recorder at station KAH, located 5 km downrift of active Puu Oo vent, showing changes in tremor amplitude during episode 22 on July 8—9, 1984, at constant instrumental sensitivity. A, Tremor amplitude increases after 2000 H.s.t. July 8, at onset of high lava output. B, High—level tremor continues during period of high fountaining and lava production. C, Tremor amplitude decreases to background level after 1000 H.s.t. July 9, at end of eruptive episode. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION were attributed to variations in signal level inherent to local ground conditions and instrumental installation at each station. The station-correction and instrumental- magnification factors were applied in reducing the tremor amplitude read at each station, and the normalized measurements were used to determine attenuation pat- terns and rates as a function of increasing distance from the source. Tremor signals during high-level activity were read from 30- to 120—s samples of digitized magnetic-tape records and from analyses of Develocorder films. Measurements from portable seismographs, normalized to readings from permanent stations located nearby, were used for additional coverage in critical areas near the eruptive vent. Amplitudes for each station measured at the dominant frequency were averaged for the sample interval, normalized according to instrumental magnifica- tion and station correction, and mapped in units of micrometers or nanometers of ground movement. Con- tour intervals were drawn relative to the station of max- imum intensity to outline the areal pattern of attenuation (fig. 7.20). The amplitude pattern may be slightly biased by the seismic-network geometry. The contours indicate that the principal source of tremor is centered near the eruption site, where attenuation outward is radially sym- metrical. Beyond a distance of about 5 km, contours are elongate toward the secondary source near the summit. Tremor was recorded at greater distances in early January 1983 than later in June 1984. The amplitude decayed exponentially with distance at rates that appear to increase as a function of time (fig. 7.21). These decay rates are about comparable to those of crustal earthquakes less than 10 km deep in this region. Tremor was also monitored with a single-component portable system near the eruption area. Such readings as those taken on April 4, 1983, during a period of low and constant eruptive activity, were normalized with readings from a continuously recording permanent station at Puu Kamoamoa; the results are plotted in figure 7.22. The data similarly fit an exponential-decay pattern with a higher rate than that for stronger activity covered on the permanent network stations. Varying rates of signal attenuation with distance were also determined from strong tremor monitored in real time on a filtered system sequentially at about 5-minute intervals for each station in the network. Data were normalized for instrumental magnification and station corrections, as well as any changes in tremor intensity during the sampling interval, by comparison with a con- tinuously recording standard station, such as MPR or KAH (fig. 7.23). Peak amplitudes, read several seconds apart, were averaged over 3 to 5 minutes of chart record for the signals with center frequencies at 0.625, 1.25, 2.5, 5.0, and 10.0 Hz. Normalized amplitudes at the various stations attenuated with distance exponentially and at 203 decreasing rates for the lower frequencies. Microseismic noise was appreciably higher in the lower frequencies, and at the 0.625-Hz center frequency the signal-to—noise ratio was essentially too low to determine its attenuation rate. The analysis was repeated for samples taken earlier dur- ing episode 10 on October 7, 1983, and during episode 16 on March 4, 1984; the results remained reasonably consistent. Hourly readings of tremor amplitude previously made from station MPR seismograms were translated to a directly proportional unit termed “reduced displace- ment,” introduced by Aki and Koyanagi (1981) and used as a basic index of seismic energy generated from deep tremor beneath Kilauea.1 The product of reduced dis- placement and duration of shallow tremor, summit defla- tion, and the volume of lava extrusion (see chap. 1) during eruptive episodes in 1983 and the first half of 1984 are plotted cumulatively (fig. 7.24). These quantities are close- ly related, as seen by their parallel rates over most of the time interval. The cause of the slightly lower rates of tremor and lava extrusion relative to the rate of tilt in May-July 1984 is uncertain; it may be partly due simply to the approximateness of the measurements. The short- term, reduced displacement rates of shallow tremor dur- ing this continuing eruption until July 1984 measured about 80 times higher than the long-term rate of deep tremor beneath Hawaii determined by Aki and Koyanagi (1981). The height of lava fountains commonly associated with eruptive rate was a partial measure of tremor intensity. Obvious correlations existed during large changes in erup- tive activity, as observed consistently at the start and end of eruptive episodes. Fountaining and tremor were both characterized by constant repetitions of amplitude bursts at intervals of a few seconds apart, but many of the minor relative changes in amplitude over longer intervals of minutes or hours did not correlate. This discrepancy sug- gests the involvement of additional factor(s) that we can- not clearly identify, separating the source of tremor and the driving force of lava fountaining. 1Reduced displacement (RD) is a function of tremor amplitude corrected for geometric spreading: A r RD=——-—, 2V5 M where A is the peak-to-peak amplitude, r is the station-to—source distance, and M is the instrumen- tal magnification. This formula is based on deep tremor recorded at relatively short epicentral distances and dominated by body waves. For shallow tremor recorded at relatively large epicen- tral distances, where Raleigh waves predominate, the calculation for reduced displacement was slightly revised by Fehler to accommodate Raleigh waves rather than body waves (Fehler, 1983). The records used in this chapter to determine reduced displacement were taken from stations near the epicentral region, and so Aki’s original formula was applied (Aki and Koyanagi, 1981). If source amplitude is extrapolated from the exponential-decay pattern as expressed in figure 7.21, the reduced displacement determined from the record of station MPR at a distance of 6 km would be about double. In any case, the choice of equations should not seriously affect our aim in determining relative time patterns for the tremor generated. 204 THE PUU OO ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 40_ I I I I l I I I I I _ w I ' : EI—Z _ EpIsodeZ _ -.z - - ass - - 2:2 I I QI—m " - I—‘O 20— _ (En: - 420 - _ LLD— — ' 3015 : I Z . 0- _ I— ”; Z : $83 1.0— — ILE'Z ' : mgfim _ 2 . . 2 : ‘—‘ - g c 2 K . ,: ’2 5 20 _ _ < E D: ' V d E 2 j I Lu 8 2 . O z : 0 h - u; I i u: o - . O D 1.0 — - E I: ' . Lu —I ' . g; o. (I) . _ ._ m _ . E g 0.5— _ Lu ’ - E I I g _ . . E.) 0. 4 a: El 2 Z I O 3 z . - E t _ 0.25 - - LU _l .. . c: o. _ . '- 3 : (No detectable tremor) : 0 DURATION ________— “m .unlu uni Il|I|IIIlllllll|llIIIIIIIIIIIII|IIIIA1| n. ||nll||llII|IIIIIIIAAI|I|IIIIIIIIAIII|Illllllllllllllllllllln . Alllltl“ . lllIlllllnlAIlIl||||II||| 12 13 14 15 16 17 18 DAYS IN JUNE 1983 4o 'I""'l W... "”'”I 'I‘l'l' . . u: i . . >_ . z . Episode 5 : E s s - - 2 I: 2 I ‘ o . _ I— I 1-— 3 O 20 r < n: - .1 E Q - U. D — . 3 w 5 ; g _ I o . . a: g : I o 3 1.0 — —' 2 v: - - E a‘ g . _ E E 0'5 T ‘. E j . ° - : 5 o u: g E I I O 3 Z . _ E I: 0.25 — _ Lu .1 - . 0: CL . . ‘- 2 - (No detectable tremor) ' < . _ o DURATION ________—._____ 28 29 30 1 2 3 4 DAYS IN JUNE AND JULY 1983 1y higher than background tremor, and reduced to ground displacement according to instrumental response at recorded fre- quency of the signal. Note difference in vertical scale to accommodate generally lower amplitudes of summit tremor. Horizontal (time) scale was varied to accommodate the generally shorter and more rapid changes in activity during later episodes. Times of eruptive episodes, defined by horizontal bars at bottom of the plots, correspond to those in chapter 1. 205 206 EAST—< Pr: _ 20 28 29 3E Z 0 (”5 3 1.0 t .J 13.”) 25 05 <1— ' LU 2 2 ‘2 0 32 DZ t—o.25 _l O. 3 o DURATION 40 w _2 r—i‘ =‘o r—< :C 28 2° 22 32 E o u; D D 1.0 : figs 3“: 0.5 LLI E O 5 o 32 DZ : 0.25 _J [L 3 o DURATION THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 ”InIyI”III!IlHlVlllH'lllHlll] nuunn‘nn... .H'nuuvnu'. . n.vn-lunnnInInuununl”nunnu'”Hump-Innunnulvnuunn'nun >— _. \- . — q uuuulyvvu (No detectable tremor) nunnln... —*__ “n...n”hum-nunln...I...lulu“unnulnn.unnl”A“.1.1.1..”nu-“1......”lulu...“....|nnu.u..|..u.....I.1........lnunxuulumnun 20 21 22 23 24 25 26 DAYS IN JULY 1983 ”Innvunul. nu ”luuuunn ........ "'l' .u-nrr-vlnvn1n...In”:luv-.vlrrnnnnylunun” Episode 7 I|III|I||I|||IIA||I n A IIlI'Illlll'll n..|l.n-nl.. IIIIIJJIL 14 15 16 17 18 DAYS IN AUGUST 1983 FIGURE 7 .16.—Continued EAST-RIFT SUMMIT EAST-RIFT SUMMIT DEFLATIONARY SUMMIT TILT, IN MICRORADIANS TREMOR AMPLITUDE, TREMOR AMPLITUDE, DEFLATIONARY TREMOR AMPLITUDE, TREMOR AMPLITUDE, O N U1 7. SEISMICITY ASSOCIATED WITH THE ERUPTION 4o . W'H'HHI 'I' I Vlllll 20 .In.”u..n.Inn-unnlnunnn. Episode 8 Illlllllll|llllllll 1.0 0.5 ||||I|ll|l||ll INMICROMETERS O ‘rnvlu..n ”...-In.- 0.25 unlrluvv—r JIIIIIIII o u DURATION _____— 5 6 7 DAYS IN SEPTEMBER 1983 8 9 4o ...........,.... . ,..... ..... I'””"""" "I .... "'"""'I 20 SUMMIT TILT, IN MICRORADIANS ""‘I .. y... ..l.... . 1...”... . Episode 9 0.5 Il.|..l L . IN MICROMETERS O ... ”HI-...lu... .........l... Ivlv‘l DURATION n... ...Il...” ”AINIHHUHIHH .....I ..... .. Inn-......Inn-”...:lu “unnlnununulu .... ... nun... .I... ... 14 15 16 17 DAYS IN SEPTEMBER 1983 FIGURE 7.16.—Continued 18 19 207 208 EAST-RIFT SUMMIT EAST-RIFT SUMMIT DEFLATIONARY SUMMIT TILT, IN MICRORADIANS TREMOR TREMOR DEFLATIONARY TREMOR TREMOR THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 4o .. .. II ......... I. ........ III ........, .. . ....I. ...... .....m. [.1 I..” Episode 10 20'— uI o _ D ’: : _I o. <0 - E E _' <( I— _ LLI 2 O . I . 2 - - B E : _ D Z . . t ‘ 0.25— _ _I - . fl. . E o u DURATION ___— ”...I.... .I.|. . ........I .. I...........I...........|.. ........I...........1.. ...I.......... .. 4 5 6 7 8 9 DAYS IN OCTOBER 1983 40 _.......................|..,..........m.......|.............. .. "WI 'I'I'I' m I Episode 11 . . Z . . L.‘ S - ‘ .2 E; : g s 20 - - O: ' . E 9 ' r U, 2 . . Z I ‘ I I o - uJ I : D - . 3 1.0 — — I: - . a” g I _ E “,1 0.5 l _ Lu - I E : _ 8 _ . . 2 0. . L15 2 : Z 3 g _ . ': 0.25 — _ _I - _. D. _ . 3 . _ o“ *- DURATION ______ 4 5 6 7 8 9 DAYS IN NOVEMBER 1983 FIGURE 7.16.—Continued ‘IO EAST-RIFT SUMMIT EAST-RIFT SUMMIT TREMOR AMPLITUDE. DEFLATIONARY TREMOR AMPLITUDE, TREMOR AMPLITUDE. DEFLATIONARY TREMOFI AMPLITUDE. IN MICROMETERS 40 20 SUMMIT TILT, IN MICRORADIANS IN MICROMETERS DURATION 40 20 SUMMIT TILT, IN MICRORADIANS 1.0 0.5 0.25 DURATION 7. SEISMICITY ASSOCIATED WITH THE ERUPTION 3 Episode 12 I ....- .unln un- uInuIunulnnnlquuunI-I“I..In...”Jun-unnl...nnunlnunu-nlqunnqun ..... l ........ .IHIIHHIHIIHHH 29 30 1 2 3 4 5 DAYS IN NOVEMBER AND DECEMBER 1983 ' Episode 13 I — ....... .....I ....... I. 1...........I.......... I. ..........I . .. I....LL.....I. .l .... 19 20 21 22 DAYS IN JANUARY 1984 FIGURE 7 .16.—Continued 209 210 EAST-RIFT SUMMIT EAST~RIFT SUMMIT THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 4o_. ............| "l' |,.[ ........ "”"""”'_' >_ g I Episode14 : IF I : <33 _ Z}— < . m -‘ <2 2 . 42 o - 1.1.3 — _ _ gm E _ - Z : 3 0. g = 2 5: 1.0— _ 2': : : EE ‘é’ : : <( E 0.5— _ “J ' : E : _ o _ _ 5 0_ . 0:3 2 I I 03 g _ _ 2t: 0125— _ LLI-J - _ 9:0. _ . I—E _ _ < — . o Afifi DURATION — 29 30 31 1 2 DAYS IN JANUARY AND FEBRUARY 1984 40_ ....... ,1.........'.....................l.........v.........r...|....... ...,..........._ m : Episode15 I >- L 2 _ SEE : P— < . 91— cc - ._— o 20— _ <2 I . . _IE 0 - LL: — [4 Bu) 2 . g . 0’ _ DOE : E: 1.0— _ 2’: Lu_1 CL U) E: g - . < ,_ 0.5— _ w _ E . g . .2 0 [:8 E Z 0:: 2 . 2*: ' 0.25— _ LLI_J - (ID- . PE ‘9 < . . o DURATION _ .l...........l...1.....11l.... .....l. .... ...11. ....-1-1 ........1l.1........ 13 14 15 16 DAYS IN FEBRUARY 1984 FIGURE 7 .16.—Continued DEFLATIONARY TREMOR AMPLITUDE, EAST-RIFT IN MICROMETERS SUMMIT TREMOR DEFLATIONARY EAST-RIFT TFIEMOR SUMMIT TREMOR 7. SEISMICITY ASSOCIATED WITH THE ERUPTION 211 v 40 ”nun” “nun... nun... I “Hun... .nu... . Hun...” nun...” ”nu... n nun... ........ I I ' I I I I Episode 16 1.1.; 20 SUMMIT TILT, IN MICRORADIANS lllllllllllllll AMPLITUDE o ....I.... ....I....I.... .........|. 1.0 0.5 IlllllllllVlll .0 N m ”nun... DURATION |I||xlll|||l||n|I|I|I |IlllllllllllII|I||I||IIIIIII|III||I11|IlI|||ll‘lll|IllIII||| llll lllll||IIIIIIIIIAIIIIIIIIIIII 2 3 4 5 6 DAYS IN MARCH 1984 4o 'l'l ..... ,...........l.......................|....... ...,.I......n._ m ‘ Episode17 I _Z . . ’3 1‘ . :2 . 3 m . go 20— K ' . 32 - ‘ m 2 - . g ' I o' “I . S .2 1.0— — _J 22 - (E 0.5— _ LU 2 . O . 5 0 “DJ E 3 z I: 025— _ .1 . q Q. _ 2 < i - o DURATION _ I‘lllllllllllllllll|llllllx lllllll ...n .11 Illllll IIIAAIIIIIAI|IIIIIJ 29 3O 31 1 DAYS IN MARCH AND APRIL 1984 FIGURE 7 .16.—Continued 212 EAST‘RIFT SUMMIT EAST-RIFT SUMMIT THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 4o ..........., . ”"I'"" ...., '"I'" 'I I I 'l' . ., u) : Episode 18 I E ._‘ <2: ‘ . < :1 a 2 I Z I— < - . 9 I— I: - ~ 1— — o 20 — - < E n: - - -' E U - - b D ’ ' Q (I) E - Z I I o ' . u; I 2 D: O - . O D _ 2 t - LLI —l . 9: CL ‘0 . ,_ 2 n: . < Ii" _ m . E I O . I K g . a: E 5 : O D Z . 2 I: — _ LLI _I . o: n. _ I— : - < _- ‘- - DURATION ———— 17 18 19 20 21 DAYS IN APRIL 1984 23 24 40. 20 DEFLATIONARY SUMMIT TILT, IN MICRORADIANS ”I” ”...-II. ... n... Episode 19 TREMOR AMPLITUDE, IN MICROMETERS TREMOR AMPLITUDE, DURATION ‘— ....I ..... .. III II. ..I... . . . 17 18 DAYS IN MAY 1984 FIGURE 7.16.—Continued 19 20 EAST-RIFT SUMMIT EAST-RIFT SUMMIT DEFLATIONARY Ix...-I....l """"l""" null......-l ..l..IH TREMOR TREMOR DEFLATIONARY TREMOR TREMOR 40 U) .2 5% "< ’25 20 2E 22 82 Z 0 ul 0 2 1.0 I: _I all) 25 ... .AHU // .KPN . V E/ / 19°15’ - / Eruption snte 10 KILOMETERS l_—l FIGURE 7.19.—-Kilauea area, southeastern part of the Island of Hawaii, showing 30-s samples of seismograms of digitized recordings of harmonic tremor at HVO stations (dots) surrounding area of eruptive activity at 1500 H.s.t. January 6, 1983. Records are uncorrected for instrumental magnification. Shaded areas show lava flows produced during initial days of eruption. Major craters and faults are labeled. 220 area between Puu Kamoamoa and Napau Crater (see chap. 6). Tremor amplitude remained generally high dur- ing the eruptive activity on January 5 to 6; it heightened during high fountaining and relatively dropped during low fountaining. After January 6, as the eruption temporar- ily waned, shallow earthquakes in the middle east rift zone increased in number and migrated downrift to about 2 km east of Kalalua; the microearthquake swarm accompanied ground cracking near Kalalua. Lava emission then resumed from fissures extending southeast of Puu Kahaualea. High-level tremor, now centered at station LUA, correlated with high lava output from the eastern- most vents on January 7 and 8. Shallow earthquake swarms related to the principal rift intrusion and summit collapse had essentially ended by January 8. Shallow earthquakes on the east rift associated with subsequent period of the eruption were limited to small events (M<1.0) caused by locally induced thermal and gravitational stresses near the active vents and by fresh lava flows. Seismicity at the summit varied, and minor but periodic inflation-deflation intervals were associated with further eruptive episodes. Major lava pro- duction during episodes 2 through 23 was accompanied by locally recorded tremor, and there were no precursory earthquake swarms as observed before the outbreak of a new eruption. Magma movement within the summit-rift conduit system appeared unrestricted, a behavior that we interpret as defining an open, magma-filled conduit con- necting the base of the summit reservoir to the dike in- trusion beneath the eruptive vent. Dvorak and Okamura (1985) attribute the increases in the rate of deflation at the summit and in tremor amplitude within the east rift with each sequential outbreak between episodes 2 and 7 to an increasing rate of magma flow from the summit to the east rift. The reduction of flow resistance in the con- duit system that accommodated the increased flow rate was considered to be caused by the repeated movement of magma. Seismic parameters for January 1983, the initial month of activity, are outlined for more than 800 earthquakes (table 7.2) and for a continuous record of tremor (table 7.3). The sequence of shallow earthquakes, tremor, summit tilt changes, and eruptive events during the initial 11 days of intense seismicity is summarized in figure 7.14. The progressive downrift movement of shallow earthquakes and tremor was completed during this interval. The timing of and changes in summit deflation, east— rift tremor, summit tremor, and eruptive events for episodes 2 through 23 are plotted in figure 7.16. Tremor in the east rift increases to many orders of magnitude above background during major lava outbreaks. Rapid and sustained deflation was frequently followed by low-level tremor at the summit that reached several times above background level. The early eruptive episodes (2—5) were THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 characterized by gradual changes in low-level east-rift tremor, lava outbreaks that lasted many days, a low rate of summit deflation, and weak (to undetectable) summit tremor. The later eruptive episodes (6—23) featured con- trastingly abrupt and larger changes in these parameters. SUMMARY AND DISCUSSION Seismographs at the HVO continuously monitored the seismicity associated with the 1983 eruption of Kilauea. From the islandwide network of 50 continuously operating stations, the area of volcanic activity was monitored by at least 20 stations spaced about 5 km apart. The nearest. station to the eruption site was less than 1 km away. During the interval from September 1982 to June 1984, several hundred thousand earthquakes were detected in the magnitude range from about 0.1 to 4.2. A total of 4,163 earthquakes of M205 from beneath the Kilauea area were computer processed for location and magnitude. The Kilauea selection numbered 876 events in the shallow summit region at 0- to 5-km depth, 73 in the deep summit region between 5- and 40-km depth, 720 in the shallow east rift zone at 0- to 5-km depth, and 2,494 in the south flank region, mainly at 5- to 13-km depth. These earthquakes, in combination with ground- deformation data, describe the process of magma trans- port from the summit storage complex, through the rift-zone conduit system, to the eruptive vents in the east rift zone. Earthquakes in three distinct hypocentral groups that progressively responded to the volcanic process outlined the magma-conduit system. Shallow summit earthquakes monitored the state of the dynamic summit storage com- plex, shallow swarms of rift earthquakes indicated the summit-to-rift extent of magma intrusion, and deep crustal earthquakes in the adjacent south flank reflected the translation of compressional stresses to Kilauea’s unstable south flank. The spatial distribution of summit, FIGURE 7.20.—Ki1auea area, showing contours of tremor amplitude mapped from 30—5 samples of seismograms from permanent stations (circles), supplemented by records from temporary stations (dots) monitored by portable seismographs. A, 1500 H.s.t. January 6, 1983. B, 0300 H.s.t. June 8, 1984. Peak-to—trough amplitude readings sub- tracted from background noise were reduced to micrometers of ground displacement, according to instrumental response at the signal fre- quency. Intermittent tremor readings from portable seismographs were normalized with continuous records from permanent stations; amplitude distribution was then used to approximate contours at inter— vals of micrometers to proportional fractions of a micrometer, varied to accommodate map scale. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION 155°30' 154°45’ 19°45' A O 1.5 2.0 2,6 3.0 3 5 4 O I 19°15' 155°30' 154°45‘ 19°45' B O 0 19°15' 221 TABLE 7 .2.—Outlrne of seismic parameters for microearthquakes in the east-rift eruption of Kilauea, January 1983 222 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 3 '39“: g g E u. 3 0-0.; rift, and south-flank earthquakes delineated a magma- g g N a g g g: a; m =‘ is 1:0 conduit complex whose maximum dimensions are 3 by 5 55555:,5 by25km- 3 5 ”j .5 5: g e. S ” u: g :3 Our model for the magma-transport system along the E a 5: m. 5.2 ‘L c.’ o . *1 ‘2 .3 ‘3 ; east rift zone based on an inter retation of seismic and n: g u: m o u o c u N m I E - cu H ' ' . 2». :c g: a: 5 g 5 =' 8 g g 5 g corroborating data, as partly shown 1n a profile perpen- 51 ° g: i '2. S E 8 S. 8 s .S 2 g 9% dicular to the strike of the rift zone in figure 7.31, is con- 0 O 9“ sistent with the structure and dynamics proposed by % % Crosson and Endo (1982). The magmatic—conduit complex in ‘3 E ‘3 within the rift, as inferred in the cross section, is no more ‘= “ °° g; >‘ .1 3 2° :2 ' ‘ .3 j g 3 J; §°$ o 3 than 4 km Wide at Its 10-km-depth base, and about 1 km -—1 a) v—l , . 5 .3 “j 3 H “3 E E ‘ Wlde near the surface. The lateral extent at the base is E 3 .5 E 3 E E 3 x E . ‘5 5 e: E : q .5 2 m constrained by the zone of 5- to 10-km-deep earthquakes N N m in the adjacent south flank. The top of the principal con- ; duit is outlined by the narrower zone of earthquakes at o E ‘5 - 2 f3 5 2- to 4-km depth that occur in swarms as the system opens a “ . 55’ S 2 “ 3 i o :3 w m .—4 o i: 56 F4 .2 El 0 3 N “-4 N n) '44 g: (U U m U 0 U “3’ (U U m (I) H F: J: W U (v 'U 3 I: (U N u u ‘14 CD 0 I: D o) hd D D H ,0 >4 «9 g 8 8 10_ .. """'I ..-vur_ p : : I 3 logA =3.60—1.35 log D, 3 {I} >‘ m A : 2 where A is amplitude : ~r< u) u) a o o o and D is distance E h to O .—1 O .. _ _ V3 0) NO If‘ C u-l K: I—‘ O n 3% °’ 5‘3: 2 1 .— .— —_ r—d .1: A m ' ' - w a In m ._1 . . _ 5 355 .5 .5 J. . - - 0.1 — _ _ U) I I I I E E iogA :4.21—l.34 log D, f 3 g 5. iii 5 - where A is amplitude 3 c: u 2 ' and D is distance ' ' .H ‘5’ En; c.’ : °: 0 . . . _ r: c5 as N I II A 3:02;; 9 001 . ......I ........ 21 w a. 2 ' 2 f; $21 ‘7 T \f 2 'l IogA =3.72—1.51 log D, 3 log A =3.9'5—1.61 log D, one 2 v ”j “3 s? :l where A is amplitude _‘ _ where A IS amplitude _ g o 0- and D is distance and D is distance 2 - - _ < .. e ‘ 5 ‘2 ‘2 O (D Q) . . L4 ti 3 o o o I : Q} OJ 0) "l Ln m E 5 8 " m ' Ll — _ Z a. 0-1 .— —_ —_ , c : E E J: I G) O - - . 3 u H m ”-4 _ . _ m m «u u n) m l: u-l 0) vs «a u "-4 54-4 >~. r—4 ,5 ._4 u a. 5.. 'U 0) 5:: -r-4 0) (u l: h 0.: mm wide) us-qu—Imhw-lu 0.01 ‘ ' '| 3.: Z: :3: jjggggaggg 1 100 1 10 100 25‘: .2 E 5' .2 'S 2 3 8 2’3 x 7:” 5 a) S 53" °’ DISTANCE, IN KILOMETERS “’ 3:21 55:; 'S‘”?8"“S.‘°3EB 3 5 3:) 51 51 fi 3 5 3'5 fi 3 3 S 3 .5 FIGURE 7.21.—Tremor am litude as a function of distance from station E P "‘ 8 & PUK in eruption area that re 'stered highest intensity. A, 1500 H.s.t. g1 E 3 January 6, 1983. B, 1500 H.s.t. August 17, 1983. C, 0300 H.s.t. Octo- :; a, : ~21; ber 7, 1983. D, 0300 H.s.t. June 8, 1984. Amplitudes were reduced 5“ E u“: _ E “L: from signals recorded simultaneously during high-level activity at net- .5 g *5 g 3; I work stations 1 to 30 km away. Data points (circles) approximately $1 ‘51 8 E ‘5 fit an exponential-decay pattern, as described by equations for lines 3 m ‘2 (diagonal line) of best fit. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION 223 TABLE 7 .3.—0ntline of seismic parameters for harmonic tremor in the east-rift eruption of Kilauea, January 1983 Epicentral Maximum distance of peak-to-peak Frequency station from T' and amplitude Time of range Principal source Nearest source (km) lme . reduced to maximum (Hz) station duration actual ground amplitude (predominant . East motion (um) range) Summit rift (station) East rift zone Puu Kamoamoa 17 1-5 0100 H.s.t. Clipping level Jan. 5-12 l—lO (2-4) from Napau to (PUK). Jan. 2 to (PUK)- Kalalua Crater. 2400 H.s.t. 2.0 (MPR)1 Jan. 31; 719 hours (continu- ous). Summit caldera —————— North Fit 1 14-21 0100 H.s.t. .2 (NPT) Jan. 3-8 1—5 (2-4) (NPT). Jan. 2 to 0900 l-l.S.t. Jan. 15; 320 hours (intermit— ent). 1Station MPR was located 2 to 10 km from east-rift tremor source. to the surface to accommodate eruptions. Magma from the mantle thus rises steeply to within a few kilometers beneath the summit caldera for temporary storage. Magma is then tapped from the summit reservoir and subvertical conduit below, and directed along the east rift zone. The zone of concentrated earthquakes at 2- to 4-km depth beneath the rift zone outlines the region of active structural changes that accommodate dike emplacements from lateral and vertical movement of magma during rift intrusions. The zone directly beneath this region that lies adjacent to a concentration of 5- to 10-km-deep south- flank earthquakes is relatively aseismic except at the downrift end of the intrusion zone, where rift earthquakes deepen to at least 8 km below the surface. This region may be a relatively passive, semirigid zone that acts as a medium for transfer of magma pressure. Lateral pressure exerted from this 5- to 10-km-deep zone may in- duce compressive stresses and strain release in the adja- cent wallrocks along the south flank and in deep barriers within the rift conduit during intrusions. The top of this zone subjected to less confining stresses would more ac- tively participate in the lateral transfer of relatively fluid and degassed magma. This region may constitute a rela- tively persistent conduit complex that upon overpressuri- zation would episodically feed dike emplacements above. A nearly vertical alignment of eruptive vents above the rift conduit is indicated by the concentric pattern of tremor attenuation from the eruptive vent to distances of about 5 km (fig. 7.20). Although this pattern could also be observed if the seismic source was isolated in a very shallow vent system, part of the energy is probably radiated from at least a few kilometers beneath the vent. Evidence for this interpretation is (l) the constant supply of magma to a deeper magma-transport system and (2) the incomplete harmony between fountain activity and tremor amplitude. The spectrum of peaked frequencies, ranging from at least 1 to 10 Hz, recorded near the ac- tive vent during the height of eruptive activity (Koyanagi and others, 1987) may also be due to variations in source properties that require an extended region beneath the vent. Along the rift zones affected by Vigorous magma move- ment, swarms of shallow earthquakes at about 2- to 4-km depth occur in short episodes (fig. 7.10) to accommodate the rapidly accumulating local stresses. Earthquake activity in the upper kilometer of the summit and rift zones—an isolated region of low velocity and low stress outlined by Hill (1969) and Zucca and Hill (1980) that is highly fractured and constantly deforming—is confined to small events that are generally recorded at an insuffi- cient number of stations to permit standard processing for hypocenter determination. The increasing number of small seismic events in the east rift zone during the erup- tion is mainly attributed to surface and near-surface activity local to the eruption site. Such events include microfracturing and rockfalls along the unstable vent walls, explosive degassing of the magma in the vent, and explosive combustion of methane gas from buried organic material adjacent to active lava flows. Microfracturing from thermally related contractions of fresh lava flows near the monitoring stations also contributed to the in- creasing number of east-rift earthquakes. 224 The clustering of earthquakes indicates variations in stress, probably dictated by nearby magma movement and influenced by regional stresses. Places of few earthquakes along the rift zone may be interpreted as regions Where magma passage is unrestricted. Zones of increasing numbers of earthquakes may be constricted localities where A is amplitude E log A =2.74—1.85 log D, - and D is distance AMPLITUDE, IN MICROMETERS O . '_. I 0.01 I 1 1 I | I l | DISTANCE, IN KILOMETERS FIGURE 7.22.—Tremor amplitude as a function of distance from erup- tive vent during low and localized activity at 1100 to 1535 Est. April 4, 1983. Amplitudes were reduced from signals recorded on a port- able seismograph and normalized with signals from continuously recording, permanent station at Puu Kamoamoa. Data points (circles) approximately fit an exponential-decay pattern, as described by equa- tions for lines (diagonal lines) of best fit. THE PUU 00 ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 Where magma pressure induces compressional stresses in the host rocks, vertically causing earthquakes in the caprock above the magma-transport system, and lateral- ly causing earthquakes within the unstable south flank. The numerous earthquakes in the south flank that occur along a subhorizontal zone, 5 to 10 km deep, in combina- tion with analysis of focal mechanisms, indicate that large blocks of the south flank respond by moving away from the rift axis and slipping southeastward along the prevolcanic ocean floor (Crosson and Endo, 1982). Such relief of stress will allow a temporary accommodation of further rift dilation, and a decrease in south-flank seismic activity, until tensions once more build up in the south flank, owing to continuing intrusion. The earthquake distribution may thus be interpreted in terms of magma pressure and stresses in the country rock that are relieved to accommodate the resulting pattern of magma movement. The episodic bursts of shallow earthquakes concentrated in the Puu Kamoamoa area since 1977 are believed to be associated with disruptions in the conduit system, and in combination with a pattern of rift inflation (Dzurisin and others, 1984) was interpreted to be caused by localization of magma supplied by intrusions. The absence of intrusion-related swarms and inflation farther downrift marked the locality between Puu Kamoamoa and Puu Kahaualea to be the easternmost extent of intrusive ac- tivity. Consequently, the relatively few earthquakes in this specific locality during the early January 1983 activity may have been due to preexisting dike emplacement. Locations of low seismicity along the active rift conduit are interpreted to be regions where magma movement is relatively free, and complementing zones of low-level seismicity extending normal to the rift axis are believed to be due to variations in stress translated tectonically, rather than to injection of magma into the south flank. Isolated earthquakes and aftershock sequences character- ize the seismic activity in the south flank. The absence of localized earthquake swarms and harmonic tremor around aseismic zones in the south flank suggests that these parts are not dynamic regions of magma intrusion comparable to the active regions beneath the axis of the east rift. Irregularities in the spatial distribution of the south-flank earthquakes that accompany major rift intru- sions suggest differences in the rate of seaward displace- ment or, alternatively, aseismic movement of some com- ponents of the south flank (Crosson and Endo, 1982). Some inconsistencies in the detection of small earth- quakes, especially those of M<1.0, are otherwise caused by sustained episodes of high-level tremor that mask earthquake signals. The pattern in overall spatial distribution of the earth- quakes, which extends from the volcanic regime along the rift zone through its tectonic counterpart along the south 7. SEISMICITY ASSOCIATED WITH THE ERUPTION flank, also suggests the influence of regional stresses in addition to the expected local stresses induced by magma intrusion. The aseismic pockets along the east rift zone that continue deep into the adjacent south flank, form- ing elongate zones normal to the orientation of the rift, may be due to neutralized stress conditions developed as a consequence of variations in stress orientation in the surrounding regions. These regions also appear to pro- duce structural accommodations conducive to the develop— ment of magma-storage zones in complementary parts of the rift zone. The magma pockets formed and maintained in these rift—zone anomalies before rapid intrusions, in turn, would tend to interrupt the spatial and temporal distribution of earthquake swarms. Magma pressure would be transmitted freely through these zones of low rigidity, and only a few earthquakes would be generated. At times when downrift movement of magma is effec- tively restricted in the middle east rift zone, uprift pro- pagation of pressure may cause earthquakes to also migrate uprift as the existing dike system tends to widen, or upward diversion of magma by a relative increase of magmatic over gravitational force at the blockage may supply lava to an east-rift eruption. In January 1983, the initial pressure causing earthquakes along the dike system propagated downrift at a rate of about 0.6 to 0.7 km/h 225 before reaching the surface adjacent to the blockage in the conduit. During the subsequent intermittent swarms, earthquakes migrated uprift at a slow rate of about 0.06 to 0.07 km/h and reversed downrift again at a rate of 0.6 to 0.7 km/h, ending east of Kalalua by January 8. Tremor is apparently generated by magma flow in par- ticular localities where pressure either fluctuates or in- creases near conduit constrictions. It intensifies near the surface upon eruption as relatively free flow of impulsively degassing magma impacts the vent walls. Ascending magma is subject to rapid volume increase from depres- surization and release of volatile materials. Thus, the normalization of pressure at the distal end of the intrusion is accommodated laterally by overcom- ing downrift barriers along the preexisting conduit system, upward as dikes extend to shallower zones that are more accessible owing to lower stress conditions, or backward uprift along the active conduit system. If the rate of magma flow into the rift zone is low relative to the strength of downrift barriers and the caprock, the pressure buildup at the terminus would have time to be normalized within the existing fluid system by propa- gating back uprift. This normalization would tend to widen the dike system progressively uprift and cause earthquakes to slowly migrate uprift. As the rate of in- : 10Hz Illllll l l IIIIIII‘ 5H1 log A =2.33-1.7O logD IogA =2.69—1.67 log D 100 lllllll‘ llllllll I RELATIVE AMPLITUDE (A) llllllll Illlllll h) | l I l l | l | | 1 I lll|llll I I 2.5 Hz log A = 2.76—1.36 log D lllllll 1 IIIIli—rrT I 1.25Hz log A = 2.52—0.60 log D _ I‘Flllll I llllllll l Llllllll I l l l Illlllll | l I 1 1O 1 10 1 1O 1 10 DISTANCE (D), IN KILOMETERS FIGURE 7 .23.—Tremor amplitude from filtered signals recorded at various stations and times during episode 23 on July 28, 1984,as a function of distance from Puu Kamoamoa station (KMM), located within 1 km of eruptive vent. Data points (circles) approximately fit an exponential- decay pattern for frequencies centered at 10, 5, 2.5, and 1.25 Hz. Equations describe lines (diagonal line) of best fit. Dashed horizontal lines denote noise level at various frequencies. 226 trusion increases, however, rapid buildup of pressure at the intrusion front would tend to be rapidly normalized by inducing stress relief in the immediate region, and dike emplacement would extend upward and downrift. The relative strength of downrift barriers or a slight decrease in flow rate may reduce the momentum in lateral move- ment downrift and emphasize dilation upward to the surface. Upon eruption, pressure is further reduced near the sur- face as a result of conduit enlargement and lava discharge. The drop in fluid pressure at the vent induces vesicula- tion and accelerates expansion of the rising magma, and the resulting increase in buoyancy also increases the erup- tive intensity, until magma pressure deeper in the con- duit system drops to a critical level. When a sufficient balance in magma pressure remains in the rift conduit after an eruptive episode, a seismically open system is maintained, characterized by persistent low-level har- monic tremor. This condition facilitates more eruptive episodes from the same vent system upon a minimal in- THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983-84 crease in internal pressure. Swarms of earthquakes typical of new eruptions are not required when there is no need for new dikes to form. At the summit, SP earth- quakes increase in frequency with inflation, and harmonic tremor and LP events are correlated with sustained high rates of deflation after the onset of intense eruptive episodes. The tremor activity that accompanies the ac- celerated summit deflation probably reflects adjustments attendant on decompression of magma leaving the sum— mit, and (or) interaction with ground water introduced into the depressurized magma system (Pollard, 1981; L.P. Greenland, oral commun., 1985). The absence of widespread intrusion-related swarms of earthquakes preceding eruptive episodes after the initial dike emplacement and fissure outbreaks, and the general- ly shorter intervals and abrupt changes in tremor ampli- tude in the middle east rift zone responding immediately to changes in eruptive vigor during later eruptive episodes, are attributed to the development of an increas- ingly efficient summit-to-rift transport system. The 40 I I I I I I I I I I I l I I I I I l I 1 of i" O P__l 5 as — T... w — n: \ p—--‘ '- m m 9"— LL 5 2 o ,_ g 9*- 2 Lu 0 _ I 9 U) E E1: 30 _ _ l— D O O — < z [—D m g 8 D U D ‘L’L’J 8 E 25 q 0 a: O 0 _ Tremor ?‘" _ 2 m .. v- “ho-- < I- z _7= P—"D‘" I— “J —_ ~ —— I 2 E < +- - I L“ I— > =’ ‘ E z < l— 20 ‘ 49—_ ' “J Lu .1 ,_ ?_"_ 2 U o s ‘4 °"’ Lu Lu -4 E E o E o——-’ Lava 9-9 <9 <2: 3 D / 9 ______ 4 D D E (D 15 _ —-l ?___l _ O >. J I D (I) X CC .I a . w < .9"— D e 3 Z 0.: —_ 3 2 Lu 9 I, 9"" l— _ I _ E < 10 _ P‘ J ____/ _ u_ D _l _____ / P__l o -' LL 5’ O “J P", I— > D l I U I P “““““ D 5 - / _ O l I n: I I D. l o._—_/ l d/ 0 ,”.’ I I I I I I I I l I I I I I I I I I J F M A M J J A 0 N D J F M A M J J A MONTHS IN 1983 AND 1984 FIGURE 7.24. —Cumulative product of reduced displacement and duration of tremor, cumulative volume of extruded lava, and cumulative defla- tionary summit tilt for periods of high lava production during episodes 1 through 23 as a function of time. Tremor was reduced from station MPR record, assuming a 6- km hypocentral distance. Lava-volume and summit- tilt data from chapters 1 and 6. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION 227 localized tremor that persisted between eruptive episodes varied according to the relative Vigor and pattern of lava is attributed to the continuity maintained in the magma— movement or degassing in the vent. transport system and the active vent. The amplitude of The location of the eruption site initially appears to be low-level tremor near the vent between eruptive episodes dictated by the strength of the retaining roofrock and the ww .,... wvw K V 4 u. “a A... A A... A A 4. - w w w. -r V V v" 'v A .i A AAA AN .44 . A. . M .. A v'vw w v rvv—wvuw y— "wwv .7 V w RELATIVE AMPLITUDE AT 15-MINUTE INTERVALS M 1 MINUTE | l FIGURE 7.25.—Seismograms from a rotating-drum recorder at station KMM, located within 1 km uprift of Puu 00. A, Variations in amplitude of low-level tremor between eruptive episodes on April 28-29, 1984. Tremor at nearly background level (1) and amplitude increases of 3 t0 5 times (2) alternate at intervals of several or more hours. B, Various seismic events during gas-piston lava activity at Puu Oo vent on May 5, 1984: 1, high-frequency microearthquake; 2, rockfall; 3, associated tremor. 228 distribution of fluid pressure, which frequently vary in time and space along the summit and rift magma system. The locus of eruptions along the summit-rift alignment may either shift for successively new outbreaks or remain the same for prolonged eruptions, as with such persistent episodes as the Puu 00 and early Mauna Ulu sequences. Repeated eruptions from the same vent system seem to _. RELATIVE AMPLITUDE AT 15-MINUTE INTERVALS I MINUTE THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 be controlled by changes in the critically balanced pressure system. The sequence of events from gradual inflation of the summit, through impulsive eruption, to rapid summit deflation implies that, once the major out- break is started by critical depressurization of the system, the connecting fluid complex serves as a medium for pressure transfer from the overpressurized summit reser- FIGURE 7.25.—C0ntinued 7. SEISMICITY ASSOCIATED WITH THE ERUPTION voir to the relatively underpressurized conduit region beneath the east-rift vent. A relatively steady and critical- ly lower velocity of magma movement through a wider central part of the rift conduit, once the connection is made, would help to explain the absence of tremor in the active part of the transport system connecting the sum- mit and eruptive vent. Development of an increasingly continuous conduit system may alternatively generate uniquely longer period tremor beyond our instrumental system’s detection capability. Harmonic tremor, at frequencies between about 1 and 10 Hz, varied in amplitude according to the rate of magma flow and the intensity of eruption. The higher frequen- cies dominated near the vent, where tremor was most in- 229 tense. Tremor amplitude attenuated exponentially with distance from the source, and at decreasing rates for lower frequencies. Rapid changes in tremor occurred dur- ing the start and end of most eruptive episodes. Order- of-magnitude changes in tremor amplitude occurred within one to several minutes at these times. Alternative- ly, summit tremor and LP events that occurred after a substantial collapse were much more gradual in buildup and decay, and many times less energetic, than the rift tremor during eruptive peaks. Tremor signals consisting of constant fluctuations over a limited range of dominant frequencies that appeared as amplitude bursts separated seconds apart on the seismo— grams, corresponded to the pulsations of intense lava V “DHW '00:: v ‘ w+ W WWWWMWMWW 1 ‘ - .; 2.2:“ _ A, .7 a: Y I l v ”‘4 a V ‘ ‘1‘ w v “T r “14* W‘ V) _. A h.“ . T" - a _ i . . W W v“ (W 4M“ . .. W E a, - A ., we - M- . ans...“ A “a"... '— Z w . l- l D Z 5. 53 , u‘ r" I l l I .i l r; l H 11‘ WNW I“ L w M. ‘4 W . w “3‘ Wm “ ‘ V firm 0 j ‘ . . IV“ V D ;.n A ‘v .r E l ‘ Jim "W , +1. r v “H" n. ' if}! i y [\n'rfu.‘ 3 2 . , “I" E.“ t" m . w ‘ .r V - 4* ' “W - - w. Z v 2 d ”NH ”W a: l . . . - 1 ' ‘ mmwmmmmww win-WWMM wwwmwmmw 3 'iii‘mwwu'w r335. wwrlmnl Mutt: W MMMM‘W .11.“: WW; tr ll: 1 MINUTE l FIGURE 7 .26.—Seismograms from a rotating-drum recorder at station KMM, located within 1 km uprift of Puu 00, showing various low-level tremor events between major outbreaks of lava at the vent on May 15—16, 1984: 1, constant low-amplitude tremor; 2, tremor burst associated with intermittent lava spattering and degassing; 3, sustained increase in tremor amplitude coinciding with onset of lava fountaining. 230 RELATIVE AMPLITUDE THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 A 0756 H.s.t. 0727 H.s.t. 10 SECONDS I—J FIGURE 7.27.—Low-level tremor burst (T) associated with gas-piston lava activity in eruptive vent, recorded after episode 23 on August 7, 1984. Signal on channel 6 is unfiltered HVO Type 1, and signals on channels 1 through 5 are filtered. A, 0756 H.s.t., station KMM, located within 1 km of the vent. Burst 7. SEISMICITY ASSOCIATED WITH THE ERUPTION fountaining. This pulsating behavior, which is character— istic of both eruption and noneruption tremor, appears to be fundamentally related to the mechanism that drives the movement of magma. The recognizable pulses in lava fountaining and seismograms of tremor average about 5 s apart. A subtle increase in the number of deep—tremor episodes and LP events in 1984 (Koyanagi and others, 1987) sug- gests an accelerated rate of magma supply from the man- tle during the Puu 00 eruption. Calculations based on the volume of extruded lava support this inferred increase in supply rate (Wolfe and others, 1987). The 802 emission in the summit area, indicative of magma influx from depth, rapidly increased early during the intrusive period and remained at a constant high rate thereafter (Green- land and others, 1984). Apparently, the accelerated rate of magma flow that accommodated a decrease in pressure, translated vertically along the transport system, eventual- ly reached the mantle source region more than 50 km RELATIVE AMPLITUDE 0735 H.s.t. 231 beneath the southern part of the Island of Hawaii in ear- ly 1984. A similar correlation of increase in deep tremor and high lava production during the prolonged eruption at Mauna Ulu in 1969—7 4 was noted by Aki and Koyanagi (1981) and later emphasized by Dzurisin and others (1984). The increase in deep tremor associated with the 1983—84 eruption was not evident until many months after the onset of eruptive activity. This delay in deep tremor ac- tivity suggests that the inferred increase in supply rate of magma from the mantle was induced by a pressure decrease in the upper conduit system, an interpretation that favors one of the mechanisms proposed by Dzurisin and others (1984). They suggested that the time—related increase in deep tremor with high lava-production rates, and the relatively rapid reinflation of the summit after large deflation events, can be explained by a process whereby rapid removal of magma from the shallow sum- mit reservoir relieves the load on the hydraulically linked plumbing system to accelerate the rate of magma supply 10 SECONDS |___l was preceded by a high-frequency microearthquake (E). B, 0727 H.s.t., station KMME (east-west component), located within 1 km of the vent. C, 0735 H.s.t., station KMMN (north-south component), located within 1 km of the vent. Peak responses and relative magnifications for filtered channels in figure 7.27A: 1, 0.625 Hz at x2; 2, 1.25 Hz at x10; 3, 2.5 Hz at x26; 4, 5.0 Hz at x60; and 5, 10.0 Hz at x 100. Peak responses and relative magnifications for filtered channels in figures 7 .27B and 7.270: 1, 0.625 Hz at x 1;2, 1.25 Hz at x 1; 3, 2.5 Hz at x 1; 4, 5.0 Hz at x 1; and 5,10.0 Hz at x 1. 232 THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 from the mantle. Episodically varying rates of magma The amplitude of continuous tremor during the Puu Oo supply from the mantle and a close connection between eruption thus far has varied according to the rate of eruptive events, rates of magma supply at depth, and, magma movement implied by the pattern of summit tilt ultimately, rates of melting were inferred by Wright and eruptive activity. The seismic data indicate episodic (1984), on the basis of the petrologic evolution of Hawaiian behavior, ranging from monthlong separations of erup- basalt. In View of how efficiently the summit reservoir tive episodes to shorter recurrences minutes apart dur- and an eruptive vent 20 km away could be magmatically ing gas-piston activity. Such repetitious activity, also linked, thermal and density gradients that continually noted in hydraulic systems elsewhere (Keiffer, 1984), is drive magma from the mantle to the surface could also believed to occur during a relatively steady state of stress develop an increasingly efficient and sensitively balanced in a temporarily isolated and regularly perturbed pressure hydraulic system that would be readily affected by regime. The seismicity and ground deformation accom- remotely induced pressure variations. panying eruptive episodes are localized near the magma 1000 1200 1400 1600 1800 “P f") 5 2000 o: < 1 , 1 2 1 u. 1‘ 1‘ V 1 1 l 1 2200 _ 1 y , m O 2200 .' 1 ‘ i» , 2200 m'ow-fl numb-Mar vmrm‘“ u) “ 1 ‘ 1 11 ‘ ‘ guwrfl» 11.1wwv1 0t 1‘ 1 15¢th CC ,1 . . 1 . l ‘ rm. «9". l D ' I , 1 W“ M . W O 1 1, 1,, 1‘ 1 “m . I 0000 0000 — 1 1 1 y 1 1‘ . oooo ~1 ‘ M- L ‘l IV) v '1‘ I . I ’ \ . , 1 '11 ‘.1, 1 ., . w ‘ ‘, 1 1 ‘ N' , VNcMV- 1 1 1 .1 ,1 1 l ) ~ .— MM 0200 0200 — 1 1 1 11 1r 0200 11 {www- . 11 ~ ) W , V .1- 1. :1 . =~T11 .. W1 «m 0400 0400 -1 1 .. an W 1 _. .W‘ I 1 «w ._ a ‘5' 0600 0500 _._ 1 .W W W I1 « W 0800 — 0800 — j. . ' ‘ ' ‘ N" " "' ”32‘”; . “W " ‘ “" ”“WV‘ 1L, ha§h¢§§$ .171-1‘11109. 3% W‘AJ‘;A‘AA M 1 MINUTE 1___——__1 FIGURE 7.28.—Parts of 24-hour seismograms from a rotating-drum mit. B, Tremor amplitude and number of LP events increase, peak recorder at station NPT, showing changes in tremor amplitude and at about 0000 H.s.t., and gradually decay during and after period of number of long-period (LP) events at summit during and after episode rapid deflation of summit. C, Tremor amplitude and number of more 16 on March 3—6, 1984 (see fig. 7.16). A, Tremor amplitude gradually conspicuous LP events very gradually decrease throughout the day increases at about 2300 H.s.t. during period of rapid deflation of sum- during period of gradual inflation of summit. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION system beneath the summit and eruptive zone. No signifi- cant activity is generated along the flanks to suggest any rapid change in the volcano’s regional stress condition. We speculate that this quasi—steady state is controlled by a critically balanced pressure system which gradually in- creases and rapidly decreases in pressure at time inter- vals dictated by the size of the hydraulically connected system. The stage of depressurization that induces exsolu- tion of volatile materials and vesiculation of the magma accordingly occurs at a higher and more vigorous rate, and is complemented by high-level tremor. The eruptive episodes are characterized by strong tremor with a max- imum amplitude of, at least, an order of magnitude above background that (1) generally lasts a day to several days during high lava production, (2) recurs at intervals of about a month and maintains the pattern in terms of years, and (3) indicates involvement of a 20-km-long fluid system which extends from the summit to the east-rift eruptive zone. The contrastingly minor repetitions of gas- piston activity are characterized by weak tremor with a maximum amplitude of, at least, several times above 233 background that (1) lasts about a minute during vigorous collapse of the magma column in the vent, (2) recurs at intervals about 5 to 15 minutes apart and maintains the pattern for several to many days, and (3) is confined to within several kilometers beneath the active vent. The repetition of eruptive episodes, characterized by a pattern of lava outbreak and high-level tremor followed by summit deflation, suggests that these outbreaks are initiated near the eruptive vent. Depressurization is trans- lated along the rift conduit and back to the summit reser- voir during the eruptive process. High lava production ends when magma pressure drops to a critically low level. The subsequent buildup of magma pressure to a critical level before an eruptive episode involves the summit reservoir, as well as the entire length of the rift conduit system that supplies the eruption. The eruptive interval and the volume of lava produced episodically should therefore be dictated by the volumetric capacity of the summit and rift storage system. Where the summit storage capacity may be relatively fixed, the capacity of the active part of the rift conduit system that feeds the .. Hakka» Tremor (background) LP event RELATIVE AMPLITUDE 10 SECONDS l____J 0740 H.s.t. FIGURE 7.29.—Summit tremor and long-period (LP) events recorded at station NPT after episode 16 at 0740 H.s.t. March 5, 1984. Signal on channel 6 is unfiltered HVO Type 1, and signals on channels 1 to 5 are filtered. Peak responses and relative magnifications for filtered channels: 1, 0.625 Hz at x2; 2, 1.25 Hz at x10; 3, 2.5 Hz at x26; 4, 5.0 Hz at x60; and 5, 10.0 Hz at x100. 234 eruption would depend on the distance between the erup- tive vent and the summit inflation center. The progres- sively longer average intervals and durations of eruptive episodes for Puu 00, in comparison with the upper-east- rift Mauna Ulu sequence in 1969 (Swanson and others, 1979) and the earlier summit sequence at Halemaumau in 1967—68 (Kinoshita and others, 1969), may be controlled by the difference in the length of the active conduit and in the consequent volume of the pressure regime. The length of the conduit system indicated by the distance from the inflation center at the summit to the eruptive vent is 20 km for Puu 00, 8 km for Mauna Ulu, and 2 km for Halemaumau. REFERENCES CITED Aki, Keiiti, Fehler, Michael, and Das, Shamita, 1977, Source mechanism of volcanic tremor: Fluid-driven crack models and their application to the 1963 Kilauea eruption: Journal of Volcanology and Geother— mal Research, v. 2, p. 259—287. Aki, Keiiti, and Koyanagi, R.Y., 1981, Deep volcanic tremor and magma ascent mechanism under Kilauea, Hawaii: Journal of Geophysical Research, v. 86, no. B8, p. 7095—7109. Chouet, Bernard, 1979, Temporal variation in the attenuation of earth- quake coda near Stone Canyon, California: Geophysical Research Letters, v. 6, no. 3, p. 143-146. 1981, Ground motion in the near field of a fluid-driven crack and its interpretation in the study of shallow volcanic tremor: Journal of Geophysical Research, v. 86, no. B7, p. 5985—6016. l I I . I (2370, 7.1) 1000 — — o (I O _ 5 Lu I _ p— I: E 0 ' a 500 — o _ o o o — o 000 o o o o o o o o 0 I 2 3 4 DEFLATIONARY SUMMIT TILT FIGURE 7.30.—Harmonic tremor versus deflationary tilt at summit after each eruptive episode (1-23) in east rift zone from January 1983 to July 1984. Dot denotes data point for exceptionally vigorous first episode, scaled down by half to accommodate rest of plot (actual values, 2,370 and 7.1). THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 1985, Excitation of a buried magmatic pipe: A seismic source model for volcanic tremor: Journal of Geophysical Research, v. 90, no. B2, p. 1881—1893. Chouet, Bernard, Aki, Keiiti, and Tsujiura, Masaru, 1978, Regional varia- tion of the scaling law of earthquake source spectra: Seismological Society of America Bulletin, v. 68, no. 1, p. 49—79. Chouet, Bernard, and Julian, BR, 1985, Dynamics of an expanding fluid filled crack: Journal of Geophysical Research, v. 90, p. 11187—11198. Chouet, Bernard, Koyanagi, R.Y., and Aki, Keiiti, 1987, Theory and discussion, pt. 2 of Origin of volcanic tremor in Hawaii, chap. 45 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: US. Geological Survey Professional Paper 1350, v. 2, p. 1259—1280. Crosson, RS, and Endo, ET, 1982, Focal mechanisms and locations of earthquakes in the vicinity of the 1975 Kalapana earthquake after- shock zone 197 0—197 9: Implications for tectonics of the south flank of Kilauea volcano, island of Hawaii: Tectonics, v. 1, no. 6, p. 495—542. Decker, R.W., Koyanagi, R.Y., Dvorak, J .J ., Lockwood, J .P., Okamura, A.T., Yamashita, KM, and Tanigawa, W.R., 1983, Seismicity and surface deformation of Mauna Loa Volcano, Hawaii: Eos (American Geophysical Union Transactions), v. 64, no. 37, p. 545—547. Duffield, W.A., 1975, Structure and origin of the Koae fault system, Kilauea Volcano, Hawaii: US. Geological Survey Professional Paper 856, 12 p. Dvorak, J .J ., and Okamura, A.T., 1985, Variations in tilt rate and har- monic tremor amplitude during J anuary—August 1983 east rift erup- .! X (D B E South flank B' 0 “I‘m. . . . I . . I |ll\\\\ * HUN 2 — .|H\\\\ _+_ + ‘ + t + g + + + #3; it, 4 — '+ i... + + + ._ Lu °.'. '..+ E +. .4." S 11D]. >4 6 — -::-'.'~ E . a 8 - 321.33 o , . 1o — ____________ 12 J l I I l O 5 10 15 DISTANCE, IN KILOMETERS FIGURE 7.31.—Characterization of east-rift dike complex in and above 2- to 4-km—deep zone of rift earthquakes and in ductile zone below, laterally constrained by distribution of 5— to 10-km-deep south-flank earthquakes. Line B-B’ (see figs. 7.7, 7.8) is normal to rift axis. Ver- tical dashed lines outline shallow region in and above earthquake swarms that actively deforms during magma intrusions; dotted region below outlines relatively passive zone of low rigidity. Entire complex is about 4 km wide at base at about 10km depth and tapers to about 1 km wide when projected to surface above retaining zone marked by earthquake swarms at 2- to 4-km depth. Subhorizontal dashed line denotes principal slip zone dipping toward center of the island, and arrows indicate direction of relative movement of south flank. See figure 7.6 for explanation of symbols. 7. SEISMICITY ASSOCIATED WITH THE ERUPTION tions of Kilauea Volcano, Hawaii: Journal of Volcanology and Geothermal Research, v. 25, p. 249—258. Dvorak, J.J., Okamura, A.T., English, T.T., Koyanagi, R.Y., Nakata, J.S., Sako, M.K., Tanigawa, W.T., and Yamashita, KM, 1986, Mechanical response of the south flank of Kilauea Volcano, Hawaii, to intrusive events along the rift systems: Tectonophysics, V. 124, p. 193—209. Dzurisin, Daniel, Koyanagi, R.Y., and English, T.T., 1984, Magma supply and storage at Kilauea Volcano, Hawaii, 1956—1983: Journal of Volcanology and Geothermal Research, V. 21, p. 177—206. Eaton, J .P., 1962, Crustal structure and volcanism in Hawaii, in Macdonald, G.A., and Kuno, Hisashi, eds., The crust of the Pacific basin: American Geophysical Union Geophysical Monograph 6, p. 13—29. Fehler, Michael, 1983, Observations of volcanic tremor at Mount St. Helens Volcano: Journal of Geophysical Research, v. 88, no. 4, p. 3476—3484. Ferrick, M.G., Qamar, Anthony, and St. Lawrence, W.F., 1982, Source mechanism of volcanic tremor: Journal of Geophysical Research, v. 87, no. 10, p. 8675—8683. Greenland, L.P., Rose, W.I., and Stokes, J .B., 1984, An estimate of gas emission and magmatic gas content from Kilauea Volcano: Geo- chimica et Cosmochimica Acta, v. 49, p. 125—129. Hill, D.P., 1969, Crustal structure of the island of Hawaii from seismic- refraction measurements: Seismological Society of America Bulletin, v. 59, no. 1, p. 101—130. Kieffer, S.W., 1984, Seismicity at Old Faithful Geyser: An isolated source of geothermal noise and possible analogue of volcanic seismicity: Journal of Volcanology and Geothermal Research, v. 22, p. 59—95. Klein, F.W., 1978, Hypocenter location program HYPOINVERSE: U.S. Geological Survey Open-File Report 78—694, 113 p. Klein, F.W., Koyanagi, R.Y., Nakata, J .S., and Tanigawa, W.R., 1987, The seismicity of Kilauea’s magma system, chap. 43 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: U.S. Geological Survey Professional Paper 1350, v. 2, p. 1019—1185. Kinoshita, W.T., Koyanagi, R.Y., Wright, T.L., and Fiske, RS, 1969, Kilauea volcano: The 1967—1968 summit eruption: Science, v. 166, no. 3904, p. 459—468. Koyanagi, R.Y., 1982, Precedure for routine analyses and classification of seismic events at the Hawaiian Volcano Observatory, part 1: U.S. Geological Survey Open-File Report 82—1036, 18 p. Koyanagi, R.Y., Chouet, Bernard, and Aki, Keiiti, 1987, Data from the Hawaiian Volcano Observatory 1969—1985, pt. 1 of Origin of volcanic tremor in Hawaii, chap. 45 of Decker, R.W., Wright, T.L., and Stauf- fer, P.H., eds., Volcanism in Hawaii: U.S. Geological Survey Pro- fessional Paper 1350, v. 2, p. 1221—1258. Koyanagi, R.Y., Unger, J.D., Endo, E.T., and Okamura, A.T., 1974, Shallow earthquakes associated with inflation episodes at the sum- 235 mit of Kilauea Volcano, Hawaii, in Gonzales Ferran, 0., ed., Sym- posium on Andean and Antarctic Volcanology Problems, Santiago, Chile, 1974, Proceedings: Naples, International Association of Volcanology and Chemistry of the Earth’s Interior, p. 621—631. Moore, R.B., Helz, R.T., Dzurisin, Daniel, Eaton, G.P., Koyanagi, R.Y., Lipman, P.W., Lockwood, J .P., and Puniwai, G.S., 1980, The erup- tion of Kilauea Volcano, Hawaii: Journal of Volcanology and Geother- mal Research, v. 7, p. 189—210. Nakata, J .S., Tanigawa, W.R., and Klein, P.W., 1982, Hawaiian Volcano Observatory summary 81, part 1, seismic data, January to December 1981: U.S. Geological Survey report, 77 p. Nakata, J .S., Tomori, A.H., Koyanagi, R.Y., and Tanigawa, W.R., 1984, Hawaiian Volcano Observatory summary 83, seismic data, January to December 1983: U.S. Geological Survey report, 86 p. Pollard, DD, and Aki, Keiiti, 1981, A new source mechanism for volcanic tremor [abs]: Eos (American Geophysical Union Trans- actions), v. 62, no. 17, p. 400. Richter, C.F., 1958, Elementary seismology: San Francisco, W.H. Freeman and Co., 768 p. Ryan, M.P., Koyanagi, R.Y., and Fiske, R.S., 1981, Modeling of the three-dimensional structure of macroscopic magma transport systems: Application to Kilauea Volcano, Hawaii: Journal of Geophysical Research, v. 86, no. B8, p. 7111—7129. Shimozuru, Daisuke, Kamo, Kosuke, and Kinoshita, W.T., 1966, Volcanic tremor of Kilauea Volcano, Hawaii, during July—December, 1963: University of Tokyo, Earthquake Research Institute Bulletin, v. 44, no. 3, p. 1093—1133. Swanson, D.A., Duffield, W.A., and Fiske, RS, 1976, Displacement of the south flank of Kilauea Volcano: The result of forceful intru- sion of magma into the rift zones: U.S. Geological Survey Profes- sional Paper 963, 39 p. Swanson, D.A., Duffield, W.A., Jackson, D.B., and Peterson, D.W., 1979, Chronological narrative of the 1969—71 Mauna Ulu eruption of Kilauea Volcano, Hawaii: U.S. Geological Survey Professional Paper 1056, 55 p. Tanigawa, W.R., Nakata, J.S., and Tomori, A.H., 1983, Hawaiian Volcano Observatory summary 82, seismic data, January to December 1982: U.S. Geological Survey report, 91 p. Wolfe, E.W., Garcia, M.O., Jackson, D.B., Koyanagi, R.Y., Neal, C.A., and Okamura, A.T., 1987, The Puu Oo eruption of Kilauea Volcano, episodes 1—20, January 3, 1983, to June 8, 1984, chap. 17 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: U.S. Geological Survey Professional Paper 1350, v. 1, p. 471—508. Wright, T.L., 1984, Origin of Hawaiian tholeiite: A metasomatic model: Journal of Geophysical Research, v. 89, no. B5, p. 3233—3252. Zucca, J .J ., and Hill, DR, 1980, Crustal structure of the southeast flank of Kilauea Volcano, Hawaii, from seismic refraction measurements: Seismological Society of America Bulletin, v. 70, no. 4, p. 1149—1159. 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPTION) By DALLAS B. JACKSON CONTENTS Page Abstract ——————————————————————————— 237 Introduction ————————————————————————— 237 Self-potential measurements ————————————————— 237 Application to volcanic studies —————————————— 237 Geologic SP sources ——————————————————— 238 Procedure and equipment ————————————————— 238 SP profiles and monitoring arrays on Kilauea ——————— 238 Electromagnetic measurements ———————————————— 239 Application to volcanic studies —————————————— 239 Procedure and equipment ————————————————— 239 Geoelectric changes in the summit region and upper ERZ ————————————————————————— 241 SP changes on the ESR array associated with the September 25, 1982, eruption —————————— 241 SP changes on the ESR array associated with the January 3, 1983, eruption ———————————— 241 CSEM monitoring changes, September 1982 to mid-July 1983 —————————————————— 241 Geoelectrical changes near the 1983 eruption sites — — — — 243 Kalalua SP profile ___________________ 243 Puu Kamoamoa SP profile ______________ 245 Puu Kamoamoa VLF profile _____________ 245 Discussion __________________________ 246 Conclusions __________________________ 250 References cited _______________________ 251 ABSTRACT Self-potential (SP), controlled-source electromagnetic (CSEM), and very low frequency (VLF) electromagnetic data were used to study the September 25, 1982, summit eruption and the first year of the Puu Oo east-rift—zone (ERZ) eruption. Four intrusions into the middle and up- per ERZ closely preceded the onset of the January 1983 eruption. The first intrusion, on September 25, 1982, accompanied the summit erup- tion; it was detected in the ERZ by earthquake-epicenter locations along most of its length, but near its distal end (7 km from the summit) only by an SP monitor. The second, a slow intrusion into the upper ERZ, lasted from the first week of October through mid-November 1982; it was detected by SP changes across the ERZ 7 km from the summit, and by CSEM resistivity changes measured in the summit area. The third and fourth intrusions, on December 9—10, 1982, and January 2, 1983, immediately preceded the onset of the Puu Oo eruption on January 3; they were not detected by any geoelectric monitors. This failure to detect the December and January intrusions is interpreted to mean that little or no fracture continuity existed from the zone of magma transport to the shallow depths measurable by the SP system, and that magma transport was deeper (greater than 2 km) in the near-summit area than was detectable by the CSEM system. The near-coincidence of the January eruptive fissure at the future site of Puu 00 with a major (879 mV) preeruption SP anomaly suggests that fracture continuity to a heat source at depth existed before the eruption. SP increases along a monitoring line near Kalalua, 2 km downrift of the most northeasterly erupting fissure, between December 18, 1982, and January 3, 1983 (15 days), suggest that downrift intrusion had proceeded at least as far as Kalalua by January 3. A small (68 mV), transient SP increase along an electric-field line at Kalalua on January 5 may have been a precursor to an intrusion near the electric-field array about a day later. INTRODUCTION At 0031 H.s.t. January 3, 1983, an eruption on Kilauea’s east rift zone (E RZ) began in Napau Crater; within 5 days it had propagated along a line of fissures that stretched downrift nearly 8 km to within 0.5 km of Kalalua (fig. 8.1). The 1983 eruption events appeared to be primed by upper- and middle-ERZ intrusions that were detected geophys- ically (that is, geoelectrically, seismically, and geodetically) during and after the September 25, 1982, summit erup- tion. This chapter discusses qualitative interpretations of self-potential (SP), controlled-source electromagnetic (CSEM), and very low frequency (VLF) electromagnetic measurements made between September 1982 and the end of the first year of the 1983 middle-ERZ eruption. SELF-POTENTIAL MEASUREMENTS APPLICATION TO VOLCANIC STUDIES SP measurements were begun on Kilauea in 1972 by Zablocki (1980), Who observed that all large positive SP anomalies on Kilauea are related to subsurface localiza- tions of heat and that after an intrusion or eruption, SP- anomaly amplitudes decay slowly while maintaining their characteristic wavelengths. Zablocki attributed this slow SP decrease over time to cooling of an emplaced heat source (for example, a dike). L.A. Anderson (in Dzurisin and others, 1980) concluded from repeated profile measurements across the Escape Road on the upper ERZ that “* * * significant SP increases were recorded near recently active fissures after the June 1976, July 197 6, and February 1977 intrusive events and after the Sep- tember 1977 eruption.” These transient SP anomalies, generally about 1 week long, were interpreted to be related to magma moving beneath Escape Road at depths of 1 to 5 km, on the basis of earthquake-hypocenter loca- tions. These SP changes observed by Anderson were evidently related to magma movement beneath the measured profiles; however, because of the short anom- aly wavelengths (never more than several hundred meters), the magma itself at depths greater than a few 237 238 hundred meters probably could not have contributed directly to the SP changes. Although the exact cause of the transient SP excursions observed in the Kilauea geoelectric studies is unknown, the short-term excursions may be related to short-lived magmatic surges and ensu- ing adjustments in the conduit system. GEOLOGIC SP SOURCES Mechanisms for the generation of large SP anomalies (as much as several hundred millivolts) are probably related to either electrokinetic (streaming potential) or thermoelectric effects. Electrokinetic potentials are generated by fluid flow through a porous medium in response to differential pressures, such as those caused by convection in a geothermal cell, or by steam and, possibly, magma flow in fractures. Thermoelectric poten- tials are generated by temperature gradients across a sec- tion of rock, such as might exist at the boundary between an intrusion and the adjacent wallrock. The ultimate source that drives these mechanisms is subsurface heat, a quantity in plentiful supply at Kilauea. Either thermo- electric or electrokinetic mechanisms can generate the types of positive, monopolar SP anomalies typically associated with intrusions and eruptive fissures at Kilauea. A general discussion of thermoelectric and electrokinetic processes in relation to geothermal areas was presented by Corwin and Hoover (1979), a detailed discussion of thermoelectric SP anomalies produced by dikelike bodies by Fitterman (1983), and a detailed discus- sion of electrokinetic SP anomalies in relation to dikelike bodies by Sill (1984). General discussions of SP effects and anomaly shapes on the Island of Hawaii were presented by Zablocki (1976, 1978b) and Jackson and Sako (1982). PROCEDURE AND EQUIPMENT Two types of SP measurements are made on Kilauea— profiling and monitoring. Profiling measurements are made by advancing a measuring electrode to successive positions along a traverse and reading the electrical poten- tials relative to a stationary reference electrode. SP- monitoring arrays use electrodes permanently sited along a traverse, and electrical potentials are measured relative to a permanently sited reference electrode. The equipment used for SP measurements is simple and requires only a high-impedance millivoltmeter (so that reliable readings can be made where contact resistance to earth ground may be as high as 100—200 k9) and non- polarizing electrodes. Two types of nonpolarizing elec- trodes are used—either copper-copper sulfate for profiling or lead-lead chloride for monitoring. The copper-copper THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 sulfate electrodes consist of a copper strip immersed in supersaturated copper-sulfate solution in a porous ceramic cup. This type of electrode is rugged, but because it must be refilled with solution periodically, it is unsatisfactory for monitoring studies. Lead-lead chloride electrodes con- sist of a lead strip in a solid plaster matrix made by mix- ing plaster of paris with a supersaturated solution of lead chloride. Although lead-lead chloride electrodes are too soft to use for robust profiling unless the plaster matrix is protected with a ceramic cup, they are ideal for monitor- ing studies because they do not need to be recharged with a solution. When measuring profiles, it is desirable to space the reference electrode positions as far apart as practical to keep cumulative errors to a minimum. Measurements are made at smaller intervals between reference electrodes. For example, measurements are commonly made at 100-m intervals, but reference electrode positions are established only every 1.5 km. In the field, first a reference-electrode position is established, and then a second electrode is moved to each measurement site. Readings are made by placing the electrode on the ground and rotating it to make contact with the soil-moisture layer that is generally present a few millimeters beneath the surface. If a reading is not repeatable to within 1 or 2 mV, or is unstable, then contact resistance is probably very high, and readings are discontinued at that site. Errors related to profiling were discussed by Jackson and Kauahikaua (1987), who found that the largest closure errors that have appeared in SP mapping on Kilauea are less than 10 mV/km. SP-monitoring arrays, where the electrodes are per- manently sited, are read by using the same equipment as for profiling. Measurement errors for fixed arrays are related to electrode drift (long term) as the electrodes age and, possibly, to electrode deterioration (see subsection below entitled “SP Changes on the ESR Array Associated with the January 3, 1983, Eruption”). Electrode drift is relatively unimportant for lead-lead chloride electrodes— only about 1 mV/mo (Petiau and Dupis, 1980). SP PROFILES AND MONITORING ARRAYS ON KILAUEA Numerous SP profiles have been established on Kilauea and Mauna Loa, where electrode positions are carefully marked for reoccupation. Only two of these profiles (fig. 8.1), near Puu Kamoamoa (KAM) and Kalalua (KAL) on the middle ERZ, are discussed in this chapter. Profile KAM is 3.5 km long and crosses the ERZ under what is now the edifice of Puu 00. Profile KAL is 1 km long and crosses the ERZ approximately 1.5 km downrift of the most northeasterly eruptive vents (not shown) of the Puu Oo eruption. 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPTION) Only one fixed-array monitor is in operation on Kilauea—on the Escape Road profile (ESR) in the ERZ (fig. 8.1). The ESR array consists of five lead-lead chloride, nonpolarizing electrodes sited along profile ESR (fig. 8.2); voltages relative to a sixth electrode are read at irregular time intervals. Four of these electrodes (68, 69, 73, and 79, respectively, in fig. 8.2) are located adja- cent to fissures that erupted in 1968, 1969, 1973, and 1979. The reference electrode (REF) is located in a relatively stable geoelectric zone; the nearest historical eruptive fissures are 0.5 km to the northwest and south- east. This array has shown changes related to both intru— sions and eruptions (Dzurisin and others, 1980; DB. Jackson, unpub. data, 1979). ELECTROMAGNETIC MEASUREMENTS APPLICATION TO VOLCANIC STUDIES A CSEM monitoring experiment has been run at Kilauea since 1979. The phase and amplitude of the elec- tromagnetic field generated by a transmitter loop is monitored in the summit region and upper ERZ on receiver loops (fig. 8.1). Thus far, all the significant CSEM changes observed at Kilauea (more than a few percent amplitude or a few degrees phase change) have been cor- 155°15' 239 related with intrusions and are interpreted to represent shallow dike emplacement near the receiver loops (Jackson and others, 1985). VLF tilt-angle measurements, which are sensitive to shallow conductivity contrasts in the Earth, have been used at Kilauea since 1970 to map in detail melt-filled lava tubes, to delineate the boundaries of buried lava lakes (Anderson and others, 1971; Jackson and Zablocki, 1981), and to study the relations between SP and VLF anomalies over low-resistivity tabular bodies, such as high-angle dikes and lava lakes (Zablocki, 1978a). On Kilauea, a low- resistivity zone associated with a cooling dike can be iden- tified to about 100-m depth because of its contrast With the more resistant country rock (Zablocki, 197 8a). PROCEDURE AND EQUIPMENT Both CSEM and VLF methods use controlled electro— magnetic sources (as opposed to natural electromagnetic fields); the CSEM system operates in the frequency range 0.1—10 Hz, whereas the VLF system operates in the fre- quency range near 20 kHz. The depth to which a buried low-resistivity body may be identified is primarily a func- tion of the frequency used and of the resistivity of the Earth surrounding the body. As the frequency of the elec- 155°07'30” 19°23'45" Kilauea Iki Kilauea caldera \ \ \_/(' Keanakakoi Crater\ O KKK /. \ Puhimau Crater 8 PUH ,‘ 0“ 0:1 Kokoolau Crater" ESRb?\\ $6 19°17'30" — “3,9" Hiiaka Crater/. «9‘ 90° SYSTEw \ EAST Pauahi Crater ESR Puu Kahaualea \o KNOO 0\ Fun Kamoamoa KAL Kalalua E RIFT ZON 0 PHth Napau Crater Makaopuhl Crater 5 KILOMETERS |__1_L_|__l_l CONTOUR INTERVAL 200 METERS FIGURE 8.1.—Summit and east rift zone of Kilauea, southeastern Island of Hawaii, showing locations of geoelectric monitors and topographic contours. ESR, SP array on Escape Road; REF, reference electrode. SP profiles: KAL, Kalalua; KAM, Puu Kamoamoa. TX, CSEM transmitter loop. CSEM receiver loops: CCR, Cone Crater; KKK, Keanakakoi Crater; OTL, Outlet vault; PUH, Puhimau Crater. ESRbh and PHth, borehole tiltmeters. 240 tromagnetic field decreases and (or) the resistivity of the Earth increases, the depth of investigation increases. The VLF system is shallow looking, investigating depths of tens of meters, whereas the CSEM system is deep look- ing, investigating depths of several kilometers. The CSEM monitoring system at Kilauea uses an ex- tremely low frequency (ELF) loop-loop system1 to measure the amplitude and phase of a transmitted ver- tical electromagnetic field (Jackson and others, 1985) generated by a large (1.5 km on a side) horizontal, quasi— square-wire loop located about 1 km northwest of Kilauea caldera (fig. 8.1). Five frequencies from 0.4 to 8 Hz are transmitted sequentially, using the transmitter loop (Tx, 1Horizontal coils of wire form both the transmitter and receiver loops. THE PUU 00 ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 fig. 8.1). Phases and normalized amplitudes relative to this transmitter are recorded at the receiver coils (for a detailed discussion of CSEM instrumentation, see Cooke and others, 1983). Although several parameters can be measured with the VLF technique, only the inclination (tilt angle) of the elec- tromagnetic field radiated by a 24.8-kHz military-radio transmitter at Seattle, Wash, is discussed in this chapter. Tilt-angle anomalies associated with steeply dipping, low- resistivity tabular bodies are easily recognized by the antisymmetric deflections produced on either side of the body and by their smooth, nearly straight line gradients through the zero point (zero crossover) from the positive and negative peaks (see inset, fig. 8.10). Numerous VLF profiles exist on Kilauea because it is common practice to make tilt-angle readings so as to iden— NE sw .800 I I I IIIII I II I III I I III I I III I I I I I I I ITI I I I I I I I' I I I I I I I I I 700— .4 u, ' | I I t I' I | I I 3 I 2m- I I | I _ 3 II I 2 I I | E II I I | I 3500— II I I I I — Z a II I I I I E | I | a I I I .- I | | I - S I g | I | g 5 II | I I 2 “[300— I . I w — 3 'I | I a: I I o I I I 3200- I I _ I I | 3 II I 5 I I I | I E I I I I I 2100— II o It - 5 E I ‘ ' a I: a I | I To, g 15" m I | I o.- g Ix P _ SI I 5 I "II I T I I Ioolllllll|llllllllllllll llllIIlIlIIIllIIIIII 0 1000 2000 3000 DISTANCE, IN METERS FIGURE 8.2.—Typical self-potential (SP) profile (dots) along Escape Road, showing locations of electrodes (69, REF, XMIT, 73, 79, 68) in ESR array. Open circles show SP differences (increases) between 0935 H.s.t. September 22 and 2225 H.s.t. September 25, 1982. See figure 8.1 for location of profile. 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPTION) tify shallow conductors when SP profiles are first read. Only the Puu Kamoamoa line was read after the start of the Puu Oo eruption. GEOELECTRIC CHANGES IN THE SUMMIT REGION AND UPPER ERZ SP CHANGES ON THE ESR ARRAY ASSOCIATED WITH THE SEPTEMBER 25, 1982, ERUPTION The last SP measurements made on the ESR array (fig. 8.1) before the September 25, 1982, eruption were at 0930 H.s.t. September 22. No unusual SP changes were measured before the eruption; array readings are made only about once every 3 days, and so short-term precur- sors would be recognized only by chance. The eruption began in the southern summit area at 1845 H.s.t., and the ESR array, more than 7 km downrift from the erup- tion site, was read at 2221 H.s.t. and again at 2228 H.s.t. that evening; both data sets were essentially identical. SP changes of 82, 102, 294, 399, and 361 mV were noted on electrodes 69, XMIT, 73, 79, and 68, respectively; the SP increases are plotted in figure 8.2 beneath a typical SP profile at the ESR. The following morning, the array was read at 1130 H.s.t., 5.5 hours after the summit eruption had stopped; all potentials were once again within 2 mV of their values 2 days before the eruption. SP CHANGES ON THE ESR ARRAY ASSOCIATED WITH THE JANUARY 3, 1983, ERUPTION In early October 1982, SP increases began to be notice- able on electrodes 68, 73, and 79 (fig. 8.3). Except for a brief but sharp SP decrease in late October, this trend continued until about November 3, when a rapid SP decrease was recorded on these electrodes, as well as at electrode XMIT. This rapid SP decrease lasted until November 10, after which more a gradual decrease con- tinued until early January 1983. These increasing SP’s in early October coincided closely with slow, nearly aseismic ground displacements in the vicinity of Kokoolau Crater to Escape Road that began on October 6 and con- tinued into early November (see chap. 6, fig. 6.7A). Between the beginning of the eruption on January 3 until mid—March, just before episode 3 that began on March 28, SP changes on the ESR array were near zero (fig. 8.3). Although all voltages on the ESR array began to increase on about March 13, just before episode 3, the increase was most noticeable at electrode 65. The SP peaked during episode 3 and then began to slowly decline. After July 1983, no SP changes occurred that were ob- viously related to the eruption, and those changes that did occur may have been related to electrode deteriora- 241 tion. In mid-August, a 40-mV offset occurred between two measurements on electrode 73. An SP profile run along the array line several days later, in comparison with another in February 1983, showed no offset at electrode 73; thus, this rapid SP shift was presumably caused by electrode deterioration as the electrode aged. Beginning in late November 1983, similar SP changes also occurred on electrodes XMIT, 68, and 79; therefore, even though the array data are presented to the end of the year, they are highly suspect after mid-August. In January 1984, all the electrodes were replaced (they were 3 years old); and as of the time of this writing (June 1985), no similar off- sets have been observed in the array data. CSEM MONITORING CHANGES, SEPTEMBER 1982 TO MID-jULY 1983 On September 27 , 36 hours after the September 25 eruption ended, CSEM monitoring data at two frequen- cies (1.0 and 4.0 Hz) were being collected at the OTL, KKK, and PUH receiver loops in the south caldera and upper ERZ. Previously, the equipment had been under repair. Although no baseline data were available before the September eruption, conspicuous CSEM amplitude changes occurred at these monitoring stations at 1.0 and 4.0 Hz (fig. 8.4); station CCR was not in operation at that time. Amplitude changes continued between early Octo- ber and mid-November, at least in the southern summit area near stations OTL and KKK (the only stations at which data were collected in October). In November, the OCR monitor coil in the southwest rift zone (fig. 8.1) was added to the CSEM array, and smoothly varying amplitude changes over several data points were recorded at 1.0 and 4.0 Hz on the OCR coil into early December. Although the data are noisy, CSEM- amplitude changes on the other monitor coils seem to track those from the OCR coil. It is unclear what these changes may represent. A large (109 grad) summit deflation that began on January 2, 1983, 1 day before the eruption, caused no significant CSEM changes at any of the monitoring stations. Two notable CSEM changes are evident in the data set between early May and mid-July 1984. The first excur- sion in early May, between episodes 3 and 4, appeared most strongly as amplitude and phase changes at stations KKK and PUH (fig. 8.4). Station OTL also showed some change, though less pronounced than those at stations PUH and KKK. There are insufficient data at station CCR to define any changes. The second excursion, before and during episode 5, occurred in late June and early July, when both CSEM amplitude and phase changes were recorded at all four monitoring stations. 242 SELF POTENTIAL, IN MILLIVOLTS I10 MILLIVOLTS PER DIVISION) THE PUU OO ERUP’I‘ION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 Sept. 1982 eruption SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC II [III [IIIIIIIIIIIIIIIIIIIIIII[IIIIIIIIIIIIII <— Data suspect ——>» IXMIT) SP response to Sept‘ 1982 eruption I _ ' I69) I v _ ' I 1 2 3 4 5 6 7 8 9 10 11 12 MONTHS IN 198283 FIGURE 8.3.—Data at five measuring electrodes (69, XMIT, 73, 79, 68) of ESR array from September 15, 1982, to January 1, 1984. Numbered bars, eruptive events. Dashed vertical lines (October 6 and November 3) show begin- ning and end of SP excursion correlated with tiltmeter changes near Escape Road (see fig. 6.7A). 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPTION) 243 GEOELECTRICAL CHANGES NEAR THE 1983 ERUPTION SITES Two SP profiles in the middle ERZ, one near Kalalua and another near Puu Kamoamoa (fig. 8.1), were reoc- cupied after the January 1983 eruption. Each profile had been measured previously and had electrode positions marked for reoccupation. KALALUA SP PROFILE An SP profile was established near Kalalua (fig. 8.1) in January 1979. The southeast end of this profile crosses a 1977 eruptive fissure, and the northwest end passes just a few tens of meters downrift of a 1963 eruptive fissure. The approximate trend of the 1963 fissure as projected across the profile, and the location of the 197 7 fissure on the profile, are shown in figure 8.5. The profile had been IIIIIIIIIIIITIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIlllilllll —0TL —OTL — W W O—C’cW—o 00 — N W4 — NH M— _. Inc-null—I M— __ _ III:IHIHIHIHIHIHIHIMIH IIIIIiI,III.IIIIIIIIIII..IIIIII —KKK — '—KKK — AMPLITUDE, IN PERCENT I1 PERCENT PER DIVISION) WH W _ 2 __Ms W— — — IIllIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIAI—I—IWIIII III III Illl‘llIIITIIIIIIIIIIIIIIIII _PUH : _.‘ _.PUH _ Dilaooooodbdfib W PHASE, IN DEGREES (I DEGREE PER DIVISION) I— OA Hz IIWIIHW .. .1 ._ WMW_ _ IIllIrll’I’TT‘FIAIIIIIIIIIIIIIIIII IIIIIWi—WIIIIIIIIII - IIIII||I|II|IIII IIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIII ccn CCR >—8.0Hz Mow — —8.0Hz WWW—OW .— 6.3Hz W 000% 6.3Hz W M _ W_ _ W“_ 4.0 MW WEDGE 4.0 Hz “analog Q/ocrnofl—u D D3010, LOHZ 10sz Hm W Huh-HP!!- 0'4H1~___.’A—a~A/¢—A W IIIlIIIIIIIlIIIIIIIIllllIIlIII SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY MONTHS IN 1982-83 FIGURE 8.4.—CSEM data, showing amplitude and phase, at five frequencies, for monitors at Outlet vault (OTL), Keanakakoi Crater (KKK), Puhimau Crater (PUH), and Gone Crater (CCR) (see fig. 8.1for locations). CCR monitor was inoperative in September and October 1982. Gaps in records are periods of no data. Numbers (0.4-8.0 Hz) labeled for CCR apply to all graphs. 244 measured five times before the January 3, 1983, eruption. Although some positive SP changes were measured near the 1977 vents on this profile during known periods of middle-ERZ intrusion, changes over the long term had been mostly negative, presumably reflecting cooling of the 1963 and 1977 intrusions. On January 3, 1983, 13 hours after the eruption began and when eruptive fissures had migrated downrift to within 3 km of the SP line, the profile was remeasured. A December 18 profile (the last occupation of the profile before the January eruption) and the January 3 profile, and their differences, are shown in figure 8.5. Between December 18, 1982 and January 3, 1983, SP increases of 12 and 24 mV were measured at data points southeast of the 1963 vent and at the 1977 vent, respectively. NW THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 In an attempt to record subsequent SP changes on the Kalalua profile, a three-electrode array was installed on January 5, with the reference electrode near the center of the profile and the recording electrodes near the 1963 and 1977 vents (see fig. 8.8 for electrode locations). Although the recorders were not operable until 2 days later, readings on the two SP lines were made by using a high-impedance electrometer, with an input filtered to reject radio frequencies, for about 4 hours on January 5. At 1510 H.s.t. January 5, a positive SP excursion began on the electrode near the 1977 vent and peaked at 1543 H.s.t. after increasing by 68 mV (fig. 8.6). By 1600 H.s.t., the SP had returned to its base level and then began to increase again (measurements were stopped because of helicopter scheduling). A mirror image of the 68—mV ex- SELF POTENTIAL, IN MILLIVOLTS (PROFILE DATA,100 MILLIVOLTS PER DIVISION; DIFFERENCE DATA, 40 MILLIVOLTS PER DIVISION) 0 {(—1963 vent | I I I I 0 200 400 1977 vent l I I I I 600 800 1000 DISTANCE, IN METERS FIGURE 8.5.—SP data along Kalalua profile for December 18, 1982 (circles), and January 3, 1983 (dots), and differences between data (diamonds) for January 1983 minus December 1982. Dashed vertical lines show projected positions of 1963 and 1977 eruptive fissures. See figure 8.1 for location of profile. 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPTION) cursion, though smaller in amplitude (approx 12 mV) and negative going (a potential change in the negative sense), is visible on the trace of the 1963 vent electrode. At 1700 H.s.t. January 7, analog SP recording from the 1963 and 1977 electrode arrays was begun, and nearly equal, positive SP changes of about 80 mV on the 1963 and 1977 vent electrodes were detected (fig. 8.7). By January 16, potentials at the 1963 and 1977 vent elec- trodes appear to have decayed to a stable level. Between January 5 and 7, a large (70 mV) SP increase was detected at 140 m on the Kalalua profile (fig. 8.8) that may have been coincident with the SP changes on the 1963 and 197 7 vent electrodes discussed above. This large SP increase was essentially restricted to one data point on the profile at 140 m, as was a 44-mV SP decrease measured on a profile on January 10 (data not shown), and caused no changes at the 1963 vent electrode, only 100 m distant. PUU KAMOAMOA SP PROFILE The Puu Kamoamoa area was a locus of intrusions from November 1978 through 1980, as indicated by tilt changes and numerous earthquake swarms (Dzurisin and others, 1984). A leveling line that crossed What was to become the January 3 eruptive fissure, about 1/2 km northeast of Puu Kamoamoa, was occupied three times between March 1979 and December 1980. During that period, it showed I I I | SELF POTENTIAL, IN MILLIVOLTS I10 MILLIVOLTS PER DIVISION I I I 1300 1400 1500 1600 TIME (H.s.t) FIGURE 8.6.—Transient 68-mV SP anomaly recorded at 197 7 vent (upper curve) and 1963 vent (lower curve) along Kalalua profile at about 1543 H.s.t. January 5. 245 approximately 82 mm of uplift across the zone in which the January 3 vents formed (Hawaiian Volcano Obser- vatory, unpub. data, 1980). In November 1980, a 3.5-km-long SP line was run across the middle ERZ (fig. 8.1), close to the Kamoamoa level- ing line. Two modest SP highs were identified (fig. 8.9) that can be matched to an 1840 fissure covered by 1969 lava and to a mapped fissure (Moore and Koyanagi, 1969, pl. 1), also buried by 1969 lava, where steam was being emitted. By far the largest SP anomaly on the profile, positive 879 mV relative to the reference electrode, occurred over an area at 2,300 m on the profile that showed no surface cracking or steaming but was within approximately 100 m of what would become part of the January 3, 1983, fissure system and the location of Puu 00. On January 28 (25 days after the eruption began), as much of the SP line as could be relocated was reoccupied (fig. 8.9). The January 3 fissure opened about 100 m north- west of the location of the previous SP high, and the peak SP amplitude was 73 mV greater than the previous high 100 m away; however, because it was shifted northwest- ward from the previous peak, it was actually 200 mV more positive than the SP measured at that point in 1980 (fig. 8.9). Complex SP changes also were measured in the vicin- ity of the 1840 fissure, where three positive peaks formed (fig. 8.9). PUU KAMOAMOA VLF PROFILE Before the January 1983 eruption, two VLF tilt-angle anomalies were apparent on the Puu Kamoamoa VLF pro— SELF POTENTIAL, IN MILLIVOLTS (10 MILLIVOLTS PER DIVISION) _ Periods of lava discharge d m 1 I I 1L. 7 8 9 10 11 12 I3 14 15 16 JANUARY 1983 FIGURE 8.7.—SP data recorded at 1977 vent (upper curve) and 1963 vent (lower curve) along Kalalua profile between 1700 H.s.t. January 7 and 2400 H.s.t. January 16. Gaps in records are periods of no data. 246 file (fig. 8.10). One anomaly was associated with the buried 1840 fissure, and the other with the zone of steaming ground between 1.9 to 2.0 km from the northwest end of the profile; however, no tilt-angle anomaly was associated with the preeruption SP maximum. After the January 3 eruption, a high-amplitude VLF tilt-angle anomaly developed over the January 3 eruptive fissure (fig. 8.11) where copious steam was being emitted, and the amplitude of the SP anomaly coincident with the zone of steaming ground also increased. The signature of the preexisting tilt-angle anomaly over the steaming ground also became more complex from interaction with the new high-amplitude tilt-angle anomaly over the eruptive fissure. However, no VLF tilt-angle anomalies accom- panied the positive SP changes in the vicinity of the 1840 fissure. DISCUSSION The SP, CSEM, and VLF electromagnetic measure- ments before and during the first year of the 1983 erup- tion can be used to infer some interesting structural relations when combined with complementary seismic and THE PUU 00 ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 deformation data. Furthermore, viewed in hindsight, they illustrate several instances of phenomena that were ap- parently precursory to intrusive or eruptive events. The very large SP increases at the ESR array asso- ciated with the September 25 summit eruption (figs. 8.2, 8.3) were 7 km from the erupting fissures and at least 1.2 km downrift from the nearest located earthquake at Hiiaka. The migration of earthquakes downrift from the summit during the eruption was related to intrusion. The SP changes that were noted at Escape Road 3 hours and 40 minutes after the eruption onset indicate that a pressure increase, probably related to intrusion, within the magma conduits must have propagated at least as far downrift as Escape Road and suggest that the summit magma plexus was in fluid continuity with the upper ERZ. This pressure increase within the conduit system may have been relieved by intrusion uprift of Escape Road because the SP anomaly had decayed to preeruption levels when the SP arrays were read at 1130 H.s.t. September 26 (5.5 hours after the eruption stopped). The concept of an open fluid core to the upper ERZ in continuity with the summit is not new; it was proposed earlier by Swan- son and others (1976). SELF POTENTIAL, IN MILLIVOLTS (PROFILE DATA, 100 MILLIVOLTS PER DIVISION; DIFFERENCE DATA,4O MILLIVOLTS PER DIVISION) NW SE I l I I I I I I l l I I l I l l | I) I 3 8 | _ E 2 b _ q) o-a U > 8 1’ I E 7’ z | a E 8 c g g m m ._ _ g g 0%, _ '— c— | | | I | LN ‘ I W4 ' | I | _ | I _ | l l I I | 'U a. 9 E g S w e E a 03 II) > C 1\ [\ _ I\ I\ 0') CD .— .— | | l I I o 500 1000 DISTANCE, IN METERS FIGURE 8.8.—SP data recorded along Kalalua profile on January 5 (circles) and January 7 (dots), 1983, before and after a small graben-forming event (near center of profile) that occurred late on January 6, 1983. Diamonds, differences between data for January 7 minus January 5. See figure 8.1 for location of profile. 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPTION) The spatial form of the transient SP changes (fig. 8.2) mimics the shape of the static SP anomalies on Escape Road so closely that it appears that no new shallow in- trusions were emplaced but rather that the SP sources were temporarily strengthened. A pressure increase in the volcanic plumbing system, with very little magma movement, might generate electrokinetic changes (pos- sibly from increased evolution of steam) that could cause the observed changes. For rapid SP variations as these, temporary enhancement of preexisting SP sources is a likely mechanism. Pressure changes accompanied by very little magma movement in the vicinity of Escape Road are also compatible with four other observations relevant to this event. (1) By 1130 H.s.t. on September 26, 5.5 hours after the eruption ceased, no residual SP anomaly remained that was not already present before the erup- tion began. Unless magma was immediately removed from possible intrusions, the SP increases should have per- sisted. (2) No earthquakes (which commonly accompany intrusions) were located as far downrift as Escape Road; the nearest earthquakes were located about 1.2 km uprift, near Hiiaka Crater. (3) Failure of the ESR tiltmeter to 247 respond to any event other than the summit inflation that immediately preceded the eruption (see chap. 6) is evidence that little, if any, new space could have opened to accommodate intrusion in the vicinity of Escape Road during the summit eruption. (4) The striking similarity between the preexisting SP anomalies at Escape Road and the transient anomalies produced during the eruption suggests temporary enhancement of preexisting SP sources without any major changes in the source geometries. On or about October 3, 1982 (1—2 weeks after the Sep- tember eruption), SP’s began to increase on three of the ESR array electrodes (68, 79, 73, fig. 8.3). Simultaneously, seismicity increased near Puhimau (see chap. 7), and tiltmeters ESRbh and PHth abruptly began to register upper-ERZ deformation located approximately between Pauahi and Puhimau Craters (see chap. 6). The close cor- relation in time between the SP increase on ESR array electrode 79 (the largest change at Escape Road) and the tilt rates measured at tiltmeters ESRbh and PHth is shown in figure 8.12. The SP increases measured at Escape Road suggest that the intrusion uprift, inferred NW 1000 I I I I II I I I I I I I I II (I) _ I— _I O _ Z _J =, _ E E _ In ‘ I. 4500— 3 2 g r z o 713 '~= E _ 3 a «i 2 - .2 8 _. 'c '— _ I: .g «5 Lu . m ' ‘5 5 S ‘ g C o g E . H 3 a 8 9 _ o m _ E IIIII—IIIIIIIILIImlIIIIIlIIIlI I 200 I l I l I IQ) I I I I I I I III I .E I All I I I I I I I I I I Em +- E u H — m m — 5.. .E 9 O x W 4 > 100 — O _ <3 5 IT-fi _ o. H _ 22 < If \ 03 or. U L o. I — _. LL d 100 w _ I I I I I I I I I 1 I I I I I I I 1 LI l I I I I I I I I 2000 DISTANCE, IN METERS FIGURE 8.9.—SP data along Puu Kamoamoa profile for April 11, 1980 (circles), and January 28, 1983 (dots). Diamonds, differences between data for January 1983 minus April 1980. January 28 profile, at and southeast of January 3, 1983, eruptive fissure, was on 1983 lava. See figure 8.1 for location of profile. 248 from the seismic and deformation data, was supplying magma elsewhere farther downrift. During the first and second weeks of November, the tilt rates measured at tiltmeters ESRbh and PHth approached zero, and the SP’s at Escape Road began an abrupt decline (fig. 8.3) that lasted until about January 1, 1983. This decline sug- gests that the slow intrusion (or transport) of magma beneath the array probably stopped during the first week of November. In late October, earthquakes near Puhimau decreased, and some earthquakes began to be recorded near Napau. During the first week of November (simul- taneously with the SP decrease at Escape Road), the earthquakes near Napau peaked and then declined abrupt- ly (see chap. 7 )—behavior implying that the magma which had been moving beneath the ESR intruded to near Napau and then stopped. Accompanying the upper-ERZ deformation in October and early November, amplitude changes by as much as 9 percent at the CSEM monitoring coils (fig. 8.4) suggest that magma was also being transported at shallow depths in the summit region near the OTL and KKK monitor coils, as well as at Puhimau (PUH). THE PUU OO ERUPTION 0F KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 Although there were no accompanying earthquake swarms and no obvious correlation with the ongoing ERZ eruption, the CSEM data for May and July 1983 suggest that an intrusion took place in both the summit region and upper parts of the rift zones. These were the last signifi- cant SP changes at the ESR array monitor or on the CSEM monitors that can be correlated with events either preceding or during the first year of the January 1983 eruption. The absence of CSEM response to the 109-urad sum- mit deflation that began on January 2, 1984, is probably related to a deep conductive zone at approximately 2 km depth or deeper in the summit region and upper ERZ, identified in an electromagnetic sounding survey of these areas (Kauahikaua and others, 1986). Scale-model studies (D.B. Jackson and J .P. Kauahikaua, unpub. data, 1983) suggest that this conductive zone acts as a screening layer which masks the effects of conductivity changes at the monitors if these changes occur below the top of the con- ductive horizon. Accordingly, the depth of magma trans- port to the middle-ERZ eruption site that began on January 2, 1983, was apparently greater than 2 km at NW SE 1000 I I I I I I I I I I I I I I I I I 1 I I | I I I I I _ EI _ r ~+ _1 _ U o I “I 2 _ ‘1 j Z N 2 — “ii 2 O //////////f///////////// -500~ <2( .3 A it ~ '3 . . .. . .. 2 _ I— ; HIgh reSIstIVIty Low-reSIstIVIty body 3 Z _I In t = ‘ o _ E n. 5 ‘ 5 _ 9 e _ u; (I) a E 0 .‘L’ 3 “" W O _ g a E °° I 2* — , l I I I I I‘— I I I I I I I I I “ I L I l I I I I I l I I I 8 20 I I I I I B I I I I I I I I I I S I I I I I I I I I I I I I o: _ c 93 E .g m ‘ z 10— 8 - . 2 fl _ <2. “ 2 < o E E s _ I— ._ 10 E — <( I I I I I I I I l I I I I I I I I I I I I I I l I I I 0 1000 2000 3000 DISTANCE, IN METERS FIGURE 8.10.—SP data (dots) and VLF tilt-angle data (squares) recorded along Puu Kamoamoa profile on December 11, 1980. See figure 8.1 for location of profile. Inset shows typical VLF tilt-angle response over a vertical, two-dimensional, low-resistivity body (hachured) in a high- resistivity half-space at shallow depth. 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPI‘ION) least as far as the position of the CSEM monitor just downrift of Puhimau Crater, and occupied a conduit that was relatively unrestricted at least as far as Mauna Ulu, where the first seismic swarm began on January 2. As with the CSEM monitor, no identifiable SP event on the ESR array could be correlated with the onset of the 109-prad summit deflation and the beginning of ERZ intrusion on January 2, 1983. I interpret this absence of correlation to mean that magma transport was through conduits deeper than 2 km (deeper than the slow intru- sion of October-November 1982) beneath the PUH monitor coil and without continuity to the surface at the ESR array. This may have been the same conduit system that fed an intrusion near Pauahi Crater on December 9—10, 1982 (see chap. 7), which also was not detected on 249 the ESR array or the CSEM monitors because it was ap- parently too deep. SP changes between December 18, 1982, and January 3, 1983 (the start of the eruption, with eruptive fissures as far downrift as Puu Kahaualea), on the Kalalua profile (fig. 8.5) suggest that a magma-filled conduit extended as far as the Kalalua area during the early hours of the eruptive activity farther uprift.2 The SP increases may actually have been related to intrusion or, at least, to the magmatic-pressure changes, inferred from borehole- tiltmeter data (see chap. 6), that occurred before the erup- tion began on January 3. A borehole tiltmeter located 2No earthquakes were located near Kalalua at this time, although tremor levels were high and would have masked small events. 1000 I | I I I l I I I I I I I I I I I I I I I I I I w — P _l O _ Z _l =I - 2 E 3 z - 3-: .J 500— ('5 <_t 8 I— _ P 5 (vi .— _ . O 3 g n. ._ ., u. _ a -40 —J .9 LU Io- m - 2 1, — 3 .9 06. , 3 ‘5 30 'c g l _ o 1, _ t; S ‘10 2 U a) D o .1 r. _ w _ I" C V e o 2 g '0 m '2 4 :3 e s g —10 2 < _ —1 ‘20 llllllllLlllllll |lll|lll30 0 500 1000 1500 2000 2 500 3000 DISTANCE, IN METERS FIGURE 8.11.—SP data (dots) and VLF title-angle data (circles) along Puu Kamoamoa profile on January 28, 1983. See figure 8.1 for location of profile. 250 about 400 m uprift of the SP profile registered tilting, down to the south, of about 22 grad at 0840 H.s.t. January 2. At about 1500 H.s.t. January 5, an SP excursion (positive 68 mV) was noted at the 1977 vent electrode (fig. 8.7) located between the center of the Kalalua profile and the 197 7 fissure (fig. 8.8). A mirror image of this excur- sion that appeared at the 1963 vent electrode suggests that the source for this event was close enough to the reference electrode to affect it also (a positive potential that affects the reference will subtract from the poten- tial at a measuring electrode). The excursion lasted about an hour and preceded an episode of ground cracking and formation of a shallow graben that apparently reflected emplacement of a dike during the night of January 6—7. At about 1900 H.s.t. January 6, a borehole tiltmeter 400 m uprift of the Kalalua SP profile went off scale. Daylight on January 7 showed that a shallow graben had formed near the tiltmeter, and new ground cracks extended northeastward, passing within a few meters of the refer- ence electrode for the KAL array. A horizontal-distance measurement indicated that 2.6 m of extension had I I I I I I I I W m > O _ Z SELF POTENTIAL, IN MILLIVOLTS I10 MILLIVOLTS PER DIVISION) I SP response to September eruption I TILT, IN MICROFIADIANS (10 MICRORADIANS PER DIVISION) I I l 1 OCT NOV MONTHS IN 1982-83 I I I SEPT DEC FIGURE 8.12.-East—west components of borehole tiltmeters ESRbh (middle curve) and PHth (upper curve) records, and SP data (lower curve) recorded at electrode 79 on ESR array, during a gradual tilt change in vicinity of Puhimau to Hiiaka Craters from approximately October 6 to November 3, 1982. Gaps in records are periods of no data. THE PUU OO ERUPTION OF KILAUEA VOLCANO, HAWAII: EPISODES 1—20, 1983—84 occurred perpendicular to the strike of the graben and cracks (see chap. 6, fig. 6.13) between January 5 and 7. Apparently, these events recorded shallow intrusion of magma to form a dike north of Kalalua during the night of January 6—7, and eruption from this new feeder dike began a short distance uprift at 1030 H.s.t. January 7 (see chap. 1). SP differences for data obtained between January 5 and 7 (fig. 8.8) on the Kalalua profile show an SP high (approx 250 m Wide) related to the January 6 in- trusion (graben) that is nearly centered on the reference electrode. Because the SP source was nearly beneath the array reference electrode, what seem to be nearly equal, positive-80-mV SP changes on the 1963 and 1977 elec- trodes (fig. 8.7 ) are actually due to a decay of the SP high generated near the reference electrode between January 5 and 7; that is, an SP decrease at the reference electrode appears as an SP increase at the measuring electrodes. The small amount of ground cracking near the reference electrode, in comparison with distinct graben formation only 400 m uprift, may indicate that the dike did not quite reach the SP array. If so, then the rapidly decaying SP anomaly may simply have been caused by a short-lived emission of steam downrift, along the small fractures propagating from the dike tip. SP’s remeasured on January 28 on a profile (fig. 8.9) about 700 m northeast of Puu Kamoamoa (very close to what later became Puu 00) showed that the January 3, 1983, fissure opened within 100 m of a major (879 mV) preexisting SP anomaly that did not coincide with any known structural feature. The relatively small SP increase near the SP profile maximum (200 mV), and the close coin- cidence of the existing SP anomaly with the location of the January 3, fissure, suggest that the eruptive fissure at Kamoamoa probably opened within a fracture zone which had continuity with a heat source within the ERZ before the eruption. A similar relation exists between a large preexisting SP high and ground fracturing on the first day of the 1984 northeast-rift—zone eruption of Mauna Loa (Lockwood and others, 1985). Areas of shallow con- ductivity along the SP profile, probably related to hot water (condensed steam) in fractures, are clearly marked by changes in shallow-looking (to approx 100-m depth) VLF tilt-angle measurements at the 1840 fissure, the steamin' ; ground (fig. 8.10), and the January 3 fissure (fig. 8.11). l" ) VLF tilt-angle anomaly was associated with the preeruptive 87 9-mV SP high (fig. 8.10) near the January 3 fissure. Presumably, any preeruptive conductive zone related to the SP source must have been deeper than 100 m. CONCLUSIONS The January 1983 eruption was preceded by an intru- sion into the upper ERZ during the September 25 8. GEOELECTRIC OBSERVATIONS (INCLUDING THE SEPTEMBER 1982 SUMMIT ERUPTION) eruption, a slow, nearly aseismic intrusion in October- November, and a third intrusion again in December. The October-November and December intrusions apparently set the stage for the January eruption by emplacing magma downright of Escape Road. The site of vent open- ing, near Kamoamoa at least, was already defined by a high-amplitude SP anomaly. The rapid decay of the SP anomaly at ESR during the September eruption, the near- ly aseismic October-November intrusion, and the two SP events at Kalalua several days before a dike was emplac- ed there on January 6—7 all suggest that the conduit down the ERZ was open before the onset of the eruption in January 1983. After the eruption began, the conduit system was continuous to about 0.5 km beyond Kalalua. The absence of CSEM or SP changes that correlate definitively with any eruptive episodes during 1983 sug- gests that the path of magma transport was at least 2 km deep in the upper ERZ (CSEM data) and, probably, within a conduit system that had little, if any, continuity to the surface beneath Escape Road (SP data). REFERENCES CITED Anderson, L.A., Jackson, D.B., and Frischknecht, RC, 1971, Kilauea: Detection of shallow magma bodies using VLF and ULF induction methods [abs.]: Eos (American Geophysical Union Transactions), v. 52, no. 4, p. 383. Cooke, James, Bradley, Jerry, Mitchell, Charles, and Lescelius, Rodger, 1983, A description of an extremely low-frequency loop-loop geophysical system: U.S. Geological Survey Open-File Report 81—1130, 63 p. Corwin, RE, and Hoover. D.B., 1979, The self-potential method in geothermal exploration: Geophysics, v. 44, no. 2, p. 226—245. Dzurisin, Daniel, Anderson, L444, Eaton, G.P., Koyanagi, R.Y., Lip— man, P.W., Lockwood, J.P., Okamura, R.T., Puniwai, G.S., Sako, M.K., and Yamashita, KM, 1980? Geophysical observations of Kilauea Volcano, Hawaii, 2. Constraints on the magma supply dur- ing November 1975—September 1977: Journal of Volcanology and Geothermal Research, v. 7, no. 3—4, p. 241—269. Dzurisin, Daniel, Koyanagi, R.Y., and English, T.T., 1984, Magma supply and storage at Kilauea Volcano, Hawaii, 1956—1983: Journal of 251 Volcanology and Geothermal Research, v. 21, no. 3—4, p. 177—206. Fitterman, D.V., 1983, Modeling of self-potential anomalies near ver- tical dikes: Geophysics, v. 48, no. 2, p. 171—180. Jackson, D.B., Kauahikaua, J .P., and Zablocki, C.J., 1985, Resistivity monitoring of an active volcano using the controlled-source elec- tromagnetic technique: Kilauea, Hawaii: Journal of Geophysical Research, v. 90, no. B14, p. 12545—12555. Jackson, D.B., and Sako, M.K., 1982, Self-potential surveys related to probable geothermal anomalies, Hualalai volcano, Hawaii: U.S. Geological Survey Open-File Report 82—127, 10 p. Jackson, D.B., and Zablocki, C.J., 1981, Interpretation of DC. resistivity, VLF, and total magnetic field intensity measurements on Kilauea Iki lava lake, 197 9—1980: U.S. Geological Survey Open-File Report 81-256, 12 p. Kauahikaua, J .P., Jackson, D.B., and Zablocki, C.J., 1986, Resistivity structure to a depth of 5 km beneath Kilauea volcano, Hawaii, from large-loop-source electromagnetic measurements (0.04—8 Hz): Jour- nal of Geophysical Research, v. 91, no. B8, p. 8267—8283. Lockwood, J .P., Banks, N.G., English, T.T., Greenland, L.P., Jackson, D.B., Johnson, D.J., Koyanagi, R.Y., Rhodes, J.M., and Sato, Motoaki, 1985, The 1984 eruption of Mauna Loa volcano, Hawaii: American Geophysical Union Transactions, v. 66, no. 16, p. 169—171. Moore, J .G., and Koyanagi, R.Y., 1969, The October 1963 eruption of Kilauea Volcano, Hawaii: U.S. Geological Survey Professional Paper 614—C, p. Cl—C13. Petiau, G., and Dupis, André, 1980, Noise, temperature coefficient, and long time stability of electrodes for telluric observations: Geophysical Prospecting, v. 28, no. 5, p. 792—804. Sill, W.R., 1984, Self-potential effects due to hydrothermal convection- velocity cross coupling: Salt Lake City, University of Utah, Depart- ment of Geology and Geophysics report DOE/ID/12079—68, 15 p. Swanson, D.A., Jackson, D.B., Koyanagi, R.Y., and Wright, TL, 1976, The February 1969 east rift eruption of Kilauea Volcano, Hawaii: U.S. Geological Survey Professional Paper 891, 30 p. Zablocki, C.J., 1976, Mapping thermal anomalies on an active volcano by the self-potential method, Kilauea, Hawaii: United Nations Sym- posium on the Development and Use of Geothermal Resources, 2d, San Francisco, 1975, Proceedings, v. 2, p. 1299—1309. 1978a, Applications of VLF induction method for studying some volcanic processes of Kilauea volcano, Hawaii: Journal of Volcano]- ogy and Geothermal Research, v. 3, p. 155—195. 1978b, Streaming potentials resulting from the descent of meteoric water—a possible source mechanism for Kilauean self- potential anomalies, in Geothermal energy: A novelty becomes a resource [abs.]: Geothermal Resources Council Transactions, v. 2, sec. 2, p. 747—748. * :Ls- : PE DEPARTMENT OF_ THE INTERIOR U.S. GEOLOGICAL SURVEY 155°07'30" 155°02'30" 155°07'30" 155°02'3 U" PROFESSIONAL PAPER 1463 PLATE 1 PROGRESS OF LONG SOUTHEASTERN \ \ _, \ I I FLOW OF JANUARY 7 \ \ L/ \g \ ‘ m 6 t x \ \ \t a: I I I I \\ \ s \I Lu \ ‘\, “\ \, I— , 1 - LI.I \» \_\ \X x\ a; 2 \\ K I \ I\ 9 _ T I I x \V Z 3 I I g . / _/ \\ l“ 10/ ' <1 ' E 3 — 5 I / > I» I\ [I . 1\ \‘ \\ LL _ o I I Lu 1 I ‘ , , ,x' 3 .~_ , _ z 25' \1 ' ’ ; 2i ' x\ < ‘ I I ‘\ - I— , x’ \ (I) [I { L L a 00 I l I I 25 1 TIME FROM BEGINNING OF LAVA FLUX , \ \ 3 1 (1115 H.s.t. January 7), IN HOURS Puu Kahauale‘a 1171‘ 1530 apau C'Iglér Napaq Crater , _ Napau Crater EXPLANATION EXPLANATION Basalt of episode 1 through 20, 1983-84 Basalt of episode 1 Pahoehoe Recent eruptive fissures (partly buried by younger basalt) 1983 Puns‘uu ~ Hc‘au Aa +I-H-I+H+I-H-I Fissure with spatter ramparts :l Pre-1983 basalt Contact 1977 (Moore and others, 1980; RB. Moore, written commun., 1985) 19° :rgy /# Kupapau P01 m 1961-69 (Moore and Koyanagi. 1969: Swanson and others, 1976, 1979) H—H——H Open crack House destroyed by 1983-84 lava flow (approximate location provided by Hawaii County Civil Defense Agency) I Approximate position of flow front—Numbers . / indicate date (1983) and time (Hawaii standard 5 time) 1/7:1500 ERUPTIVE FISSURES, 1961-1983, AND BASALT OF EPISODES 1 THROUGH 20 OM 1123 VENT 3,, I 1,, ~ 7, PROGRESS OF EPISODE 3 FLOWS PROGRESS OF EPISODE 2 FLOWS FR U) a: 9 I Ill—J 9 l I I | I \ ”2" _ K ~ Note: Dashed lines indicate unknown flow-advance rate. End of eruption \2 I— 3 I E , x __ U) > Z 6 ‘ E E 6 _ I: O E E E 2 - > § 5 3 E «V Z 5‘ _ V O 3 _ Q? < Z 3 qufi cc <0 I— — 0 LL é U) 42“". SOUTHWESTE N 1123 ENT ‘ m a / R V 190 U _ 88/ // —LOBE 2 _ Z ‘2‘ 19" 25' ‘ < eO I I , I03 0 50 100 I50 3» 25 - 0 I I I 1:: If D 0 50 100 150 / 1 TIME FROM BEGINNING OF EP/ISCLDE, IN HOURS, 35 tit/gum? ? 3 1 :1 . _ e . 47451600. , , TIME FROM BEGINNING OF 1123 VENT ERUPTION, IN HOURS ,1 _ _ _ i _ ’_ , ‘ 4/1:083o ‘ a , "4/4.‘-I§56 - x Ti» '1 ‘I * I m- * "2m 6339M ' __;\_6:o;pj, : I P u Kahaualel ”-7 E ~ . \ _ ‘ 'x 3736243900 {70/ 7“! . * ‘- 5‘ 1\ mam ‘ , , I’ It 1‘” ‘ I , I \ , -\ , - Ffliu KahauaIeafi 8532' “I? _ 3 2 mp A: /' . . ' 4/5: 1700 \1 I; y I, _ [/1 / Camp/A 63 , ’ , \ _\ Northeastérn I1 123 vent (/n ERN FLOW Noemi-73$; \ *Séirthwescér’r/I 1123' M”: _ ' , , ’ \ \fl/Camp B/o \ r‘ 145:1700 _/, 4/7: 1600‘; ,L/ , , . '«3/29: 0915 ~ . 4/ 71V . _ K 4 “0500 45k 2.6, \\./"~/ gibd‘l I 0N I / 4L ,’ ’ {//{// ’ ' . , “ I, , r‘ " EXPLANATION EXPLANATION 215' I v" ’ (2“ Basalt of epiSOde 2 ,z= Basalt of episode 3 ‘ ' / ,. /./’:/fl Flow—Predominantly aa Flow— Predominantly aa I ,/ Spatter/cinder cone - IIIIIIIIIIE hI Spatter/cinder cone H++-I++-I-H-I Fissure with spatter rampart Basalt of episodes 1 and 2 :l Pre-l983 basalt Contact Basalt of episode 1 I: ,4: [:1 Pre-l983 basalt Contact 19° 20’ , ————— Approximate internal flow boundary 5.0 Flow thickness—Measured at edge (in meters) 5.0 Flow thickness—Measured at edge (in meters) Approximate position of flow front—Numbers g} indicate date (1983) and time (Hawaii standard 55'; , time) ' Approximate position of flow front—Numbers indicate date (1983) and time (Hawaii standard time) 2/27: 0830 3/29:I725 ’(Kamakunl Kamakwu I 155°02'30" ,, I , | 155°07'30” 155°02'30" i} INIEHIUH78EDIUGICAL SURVEV, HESIDN, VIRGINIA-498577685797 Base from US. Geological Survey I:24,000 Kalapana, Makaopuhi Crater, Volcano, 1981,- EPISODE 2 SCALE 1:50 000 EPISODE 3 KaIaIua, 1982 155°07’3 U” 1 1/2 0 1 2 3 4 SMILES FFI—iI—qI—qI—q; E- g 1 ‘I 1 .5 O 1 2 3 4 5 KILOMETERS L—I H H H H g I I———T I———————I CONTOUR INTERVAL 20 FEET DATUM IS MEAN SEA LEVEL AREA OF MAP ‘ ERUPTIVE FISSURES IN THE MIDDLE EAST RIFT ZONE OF KILAUEA VOLCANO FOR ERUPTIONS FROM 1961 THROUGH 1983, AND DISTRIBUTION OF FLOWS AND VENT DEPOSITS, MEASURED FLOW THICKNESSES, AND FLOW PROGRESS FOR EPISODES 1 THROUGH 3 OF THE PUU OO ERUPTION can; 1% PROFESSIONAL PAPER 1463 DEPARTMENT OF THE INTERIOR U.S. GEOLOGICAL SURVEY , ” PLATE 3 155°07'30" 155°fl 7'30" 155°02'30" 2 <7 , \ PROGRESS OF EPISODE 8 FLOW _\ ID \2 g 4.5 E 6 , _ '-” / s I— , , x. I- / ‘ LIJ “x L“ / E \\ E / o , / «I 9 / :1 _ I ¥ / r z 3 ‘ / — Z / _ / ~ ,/ ,1/ I—~ / E 1/ E u: 3 - 2 o ‘1 9/16: 16§o I I, O 1 _ _ E , ., I ‘ 5.5’ a: .5 LL 1 y ‘ 19° _ LL I.I.I .. K S. ‘1 \ «, ‘\ 190 25 1 8 1 g . . K . M? 9/16:, or . 25. Z 2 ‘ ‘2 < Note: Dashed line IndIcates 45 , < , , 1 - «1 I )— unknown flow-advance rate. \ ‘ 9/16:1090Q ,, 5 Nofie: Dasned Iin: indicate: ~ - [‘x ‘ /», 1' f/ (L) 0 I , l\ 1 ‘- ¥ 2_ ‘ \ , r’ _ w-a van a r e. 26 . , . ‘1 I x , /’ D 0 un nown o C a _\ ,_’ 9/5,‘ 1799 / , , 1’ D 0 25 50 / o 25 * , \f ' V ” / , // TIME FROM BEGINNING OF _ / \ TIME FROM BEGINNING OF T \ / ’ EPISODE, IN HOURS ’ I, _ EPISODE, IN HOURS ‘ 1 g -- f x ‘ \ :gm arm; ( ' I 5%,”; K, . , 2" ’ ' , , / EXPLANATION EXPLANATION Basalt of episode 8 Basalt of episode 9 Flow—Predominantly aa Flow—Predominantly aa ‘ I Spatter/ cinder cone ' I I ‘1 WI ‘1 ) W Spatter/cinder cone 2 l 19° ,_ Basalt of episodes 1 through 7 , Basalt of episodes 1 through 8 we ” 20' I’m-1933 basalt I: Pre-l983 basalt —— Contact—Dashed where approximate —— Contact—Dashed where approximate 45 Flow thickness—Measured at edge (in meters) 45 Flow thickness—Measured at edge (in meters) mApproximate position 0f “0W front—Numbers 97m) Approximate position of flow front—Numbers ' indicate date (1983) and time (Hawaii standard ' indicate date (1983) and time (Hawaii standard time) time) _\ _ \\ \\ g 6/ l 3.6 x1 1 I— \ ,7 \ — . — ‘ V ‘ z A \ 1.. If: 9 9/ I110: 1615 2 Lu \«J\ «I e 2 , 1 > U) 2 \ L” Q‘ - x ' L n: _ ‘ j \ 2 ,{5’ \ 1 L E ) 2 w 3 — \ 0 Q1} ‘1 2 , O *— \ .1 ' eO / ' 1 u: ”" \ 1 2 {2 //3E 1 LL 2 i ‘1‘ Z L0 “r \‘1 I v ‘ _ 8 3 ,/ ' ‘. 6 _ $5“ _ ’\ . ‘3; . X) ‘ 2 ‘ ' , 157' , 2: 1.5 — - 1' ,2 I— ’\ = ‘ ‘ f t ‘ ‘ ‘ ' ‘ 4% ' - / o \ I _ I , , / I E E 0‘“ _ E o? 2;“ _ 1 ,_1N§. 130/ / IE" SO HERN LOBE ‘ . i 2 Note: Dashed line indicates , . / \ > _ 1 11/6: 0730 , I J, g“ 1 -~ ‘ ' . 3110:0730 0 unknown flow-advance rate. \ ‘ ' 3/ E T \x . " u. _ LN 1 y o l . ‘ , /) 1017; 1400 1 o 04‘ ‘\ ‘ ‘ f3 . \ 19, o 25 50 75 _ _ \ *57 / E «V ‘1 (/1 ,, 25' _ TIME FROM BEGINNING OF EPISODE, IN HOURS 610/7: 07130 . _‘ 8 3 ‘ z ‘ K 2 <5" ‘ ,1 _ ‘ , 3-0 *’ a , 2 £7" \ I _ / 1 \ I f \ » x ; «51/, - < "J \ ,, I, .1 1 2 , I- _ / " r‘ , ’, » / a . . _ 5.7 , o 25 50 , \_ _ _ I, I ._ , ’ I, _ ,, , \ ,1 , , , , ,1 /,fi/ \ TIME FROM BEGINNING OF EPISODE, IN HOURS ,1 V ‘ y 4 , ‘SOU‘FkREASTERN FLOW 43 QM; ' , .1. 4 11/6. 1399 ,r’ j /‘\/V , j , 4191f ‘e 1. Naps.) Crater , ; 'k Napau Crater / I“? s EXPLANATION I / EXPLANATION Basalt of episode 10 Basalt of episode 11 Flow—Predominantly aa Flow—Predominantly aa Spatter / cinder cone Spatter/ cinder cone H+H-I+H-H Fissure with spatter ramparts Basalt of episodes 1 through 9 |:) Pre-l983 basalt Contact Basalt of episodes 1 through 10 ‘:‘ Pre-1983 basalt Contact 19" 20‘ 19° 20’ 45 Flow thickness—Measured at ed e in meters 9 1 ) 44—44—H- Open crack Approximate position of flow front—Numbers indicate date (1983) and time (Hawaii standard a)», time) , 10/7: 0730 4.5 Flow thickness—Measured at edge (in meters) Approximate position of flow front— Numbers indicate date (1983) and time (Hawaii standard time) I ‘ .. ”I a“ - Kamokunl 11/6: 1300 155°07’30" 155°02'30” 155°07'30' 155°02'30" ‘1} INTERIOR—GEOLOGICAL SURVEY, RESTDN, VlflGINIA—ISBB—GBE797 Base from U.S. Geological Survey 1:24,000 EPISODE 10 Kalapana, Makaopuhi Crater, Volcano, 1981; SCALE 1:50 000 EPISODE 1 l Kalalua, 1982 1 1/2 O 1 2 3 4 5 MILES I=L Iv—l I—I I—I I—I L *1 ' *fi 1 a 1 .5 O 1 2 3 4 5 KILOMETERS H H H H H ‘ - I»——-——-——1 - 1 CONTOUR INTERVAL 20 FEET DATUM IS MEAN SEA LEVEL AREA OF MAP DISTRIBUTION OF FLOWS AND VENT DEPOSITS, MEASURED FLOW THICKNESSES, AND FLOW PROGRESS FOR EPISODES 8 THROUGH 11 OF THE PUU OO ERUPTION ’ Pb Ii. "4‘ l 1: DEPARTMENT OF THE INTERIOR ~ PROFESSIONAL PAPER 1463 U.S. GEOLOGICAL SURVEY PLATE 4 155°07’30" 155°02'30" 155°07'30" 155°02'30" _ , , g | x v.\v‘/_\ , , , t, , . , , K 1 _ \_ ‘_ . t _ I /‘ p I , /,-\ / \ K , , , _ | I; Me??? ‘I _ /> , . ‘ \ , _ . “A PROGRESS OF EPISODE 12 FLOWS ‘41 , ‘ \ ‘ ’ a ’ r' ’ “ / i I 2 9 I t\\ I \ ‘x ‘- I x \‘ \ \‘\__ \ Note: Dashed line indicates // \ \ \ ' U) unknown flow-advance rata./ \ I“ “V \‘\ ‘5 ‘N x l— 6 _ / I: I— e \ \ ‘ - Lu / T m ,3 \ 2 / E 0,; \\ . o / O Q)? i I -_J / .J __ Q _ \\ \ g / — 6 1K \ \\ 3 / g o“ ‘x ~ I 2 z — o — ‘ W/ E * "\- x. T ‘2' , F, , J I; a? / 1/ E _ _ \I I / f/ g “$- \ ‘ Q [I > 3 .13 I E 3 _ _ I 1/< 2:;073‘9 I E o L\ O « I ‘ , I .‘ E III I E \ A g - LL __ Q?- _ Lu 3 A R / 190 3 “I U . ‘ ’ 25. ~_ 8 c’o‘ 4 e z _ 9 _ 19° <2( 5 - 41” ff ‘5 ‘ '(7: “WOW - \\‘ t ‘2 5 ‘, a — Og‘iY‘E“ _ PL:U Kahaualea-. \ 5, D I (53’ NORTHERN FLOW , 37 ' , 0o 25 50 [l/ , \\ L , , , ‘1, TIME FROM BEGINNING OF _ o I ‘2”“\‘\‘3°° EPISODE, IN HOURS y 0 25 50 12/1; 08:15 I TIME FROM BEGINNING OF /’i9 EPISODE, IN HOURS 11/30: 1799 11/30: 0845 ' , l.o ulI’Kamoamoa / // , / P Camp D ‘3. _ ug Kamoamoé" EXPLANATION Basalt of episode 13 Flows—Predominantly aa Second flow First flow IIIII EXPLANATION Basalt of episode 12 Flow— Predominantly aa ’{HEEIHIW Spatter/cinder cone MM Spatter/cinder cone 19° _ Basalt of episodes 1 through 11 Basalt of episodes 1 through 12 20’ , , , _ 20’ I:| Pre-l983 basalt I Contact —- Contact—Dashed where approximate H—H—H Open crack 7.0 Flow thickness—Measured at edge (in meters) r, x 70 Flow thickness—Measured at edge (in meters) Approximate position of flow front—Numbers 1 1/22: 0730 indicate date (1984) and time (Hawaii standard , _— Approximate position of flow front—Numbers time) , 11/30: 1700 indicate date (1983) and time (Hawaii standard .. ,» time) EPISODE 12 EPISODE 13 3‘ \ l [\y / r I ‘ ,,,,,,, ‘ . \ ‘ M { .. K _ x \ ‘ K; \ I I F ‘ ' R» V) I PROGRESS OF EPISODE 15 FLOWS PROGRESS OF EPISODE 14 FLOWS - \ / / , \_ _\ , J \ ‘ I \ 6 L \ " g 4 5 \ x \ L _\‘ E g \ w E ’ 2 Lu ‘3 E _ _ \. g S \\ _ x 3 z 3 ' q, ' _ ¥ l \ 211541445 ‘ — ‘Q ‘ \: \ . ‘ ‘ I ,9 F: c)g $.59 Z 21 k' I 0 r / E 5 5:? I: ~ « r o ,. > [5,? $5? $3- — __ _\_\l_\3‘/ > , E ’0 04;?) > ’ , ‘ _/ ‘ l t I O 12,7 Sb ‘0 E _ \ ‘, I , n: I 5 — 036‘ — 0 g I, \I I I / \ , - U- 0S 0: 3 5% I , 2/15; 1030 ' A g , p m IQ‘2 \ L” V» IL ‘4 V I, l _ \\ I . , . , i , , . ' ‘ , u , ’ , / o 25' T ‘ \ L2’ (99$ w _ 5 S“ _ ( I g ‘3'6‘ ‘ . _ ; _ , _ , _ ‘ 19 1 < P9 ‘5’ U ,5: ,5 _ - \ RN11» ; go 2 a, K NOR\THE-ASTE FLOW i 5 x)" < V w . \C' I L . ’ I — 9° l— “I I5 « 35/ f/ D 00 25 (g 5 \';$ ,2 / ' TIME FROM BEGINNING OF \ g 2/139715,‘ EPISODE, IN HOURS _, 00 25 30 / ‘ :- Puu Kahauaiea ; - {I 3 \‘ \ TIME FROM BEGINNING OF PUU Kahaualea‘» ‘ 2’ H 1‘ . _ J} EPISODE, IN HOURS ' J, ‘ I 'x. ‘ SURORDINATE LOBE OFEASTERN FLQIN - . 19° 20' T I // um Kamoa‘moa Eggs! f Y“ /l Napau Crater ,;’\ , ’ I lag/:1 an EXPLA NATION I EXPLANATION 150“” , I, , , ‘7’ , Basalt of episode 14 I , ’ (/14; I Flow—Predominantly aa Spatter/cinder cone Basalt of episode 15 Flow—Predominantly aa Spatter/ cinder cone a .iPunuiw * Hem: Basalt of episodes 1 through 14 I: Pre-1983 basalt Contact Basalt of episodes 1 through 13 |:] Pre-1983 basalt — Contact—Dashed where approximate 19° 20‘ 7.0 Flow thickness—Measured at edge (in meters) 7.0 Flow thickness—Measured at edge (in meters) Approximate position of flow front—Numbers indicate date (1984) and time (Hawaii standard Approximate position of flow front—Numbers , indicate date (1984) and time (Hawaii standard , time) ' 2/1521030 1/31: 1320 ’z’ / Kamokum , _ , , I . , . I 155°07'30" 155°02'30” 155°07'30" 155°02'30" fir INIEHIUHEGEOLOGICAL SURVEY, HESIDN, VIRGIN/(719887685797 Base from U.S. Geological Survey 1:24,ooo EPISODE l4 Kalapana, Makaopuhi Crater, Volcano, 1981; ' SCALE 1:50 000 EPISODE 15 Kalalua, 1982 1 V2 0 1 2 3 4 5 MILES I=l I—I I—I I—I I—i I——-———-I I—-——-—-I ‘ I, . 1 .5 O 1 2 3 4 5 KILOMETERS H H H 1—1 H l————————-l I——l I————l CONTOUR INTERVAL 20 FEET DATUM IS MEAN SEA LEVEL AREA OF MAP DISTRIBUTION OF FLOWS AND VENT DEPOSITS, MEASURED FLOW THICKNESSES, AND FLOW PROGRESS FOR EPISODES 12 THROUGH 15 OF THE PUU OO ERUPTION DEPARTMENT OF THE INTERIOR U.S. GEOLOGICAL SURVEY 155°02'30” 155°07'30" (I? es IE ‘ “"PROEESSIONAL PAPER 1463 “9‘; PLATE 5 PROGRESS OF EPISODE 16 FLOW ‘ - , (V ‘ _ 9 I ‘\ _\ ‘x N, {/5 \ \ X h m \___ i I o: ‘ / ‘ \ Lu I I— x \‘\ ”J 1 :' 6 — — x a K I Z ' ‘I , _ K\ 'I I “J T T 5“; I > \x‘ ‘xx 2 E r 1 a 19:, — LL 3 — — I, \'~ KW“ I) I 25 Lu: \_ }_ I ‘9 _ — \ Q Note: Dashed line indicates \ I IX unknown flow-advance rate. 3 0 I O 25 50 TIME FROM BEGINNING OF EPISODE, IN HOURS ' Camp D Ruu Kamoamoa , » _ , Kamoa é; 0 ,' ~ 3/4: 1200 4.3 3/4: In ' ‘ 3/4: 132.5\ / $184,? ~ fi'SZO ~ ' 3/4: 2036 k ,, 73%; 2330( ‘ ,in EXPLANATION "'/, ' Basalt of episode 16 Flow—Predominantly aa / , I ,, 5 “ Basalt of episodes 1 through 15 I: " I: Pre-1983 basalt I —— Contact f,“ 5.0 Flow thickness—Measured at edge (in meters) Approximate position of flow front—Numbers , , , indicate date (1984) and time (Hawaii standard / time) M , 3/411100 155°07‘30" ‘ ' {I I / \ 0 I» flog ‘ 7 PROGRESS OF EPISODE 17 FLOWS DISTANCE FROM VENT, IN KILOMETERS 2 I End of 9 — eruption _ SUBORDINATE LOBE 0F EASTERN FLOW 3_ _ 0 I O 25 50 TIME FROM BEGINNING_OF EPISODE, IN HOURS w,’ " ,Camp D Ruu Kam‘oanloa . am EXPLANATION Basalt of episode 17 Flow—Predominantly aa I II I III I | IIIIIIIII I, Spatter/cinder cone Basalt of episodes 1 through 16 Contact .' 3/30: 1200 time) 5.0 Flow thickness—Measured at edge (in meters) Approximate position of flow front—Numbers indicate date (1984) and time (Hawaii standard 19° 25' _ 19° 20' EPISODE 16 EPISODE 1 7 I I I PROGRESS OF EPISODE 18 FLOWS Note: Dashed IInes indicate unknown flow-advance rater W 12 — - r’ \// T \ \L (n 7 I: _ _ x I.” x I; 5 . 2 L21 \_‘ 9 9 — § — i 2 4):" < 7 Z _ I3 _ ‘ I—‘ ‘ Z _ Lu 0- > SUBORDINATE 5 6 ' LOBE OF — I O EASTERN FLOW \x I: 3 190 u. 9 /19: 165 25, — Lu — LL _ 4/19:1419 o z _ z I < e 5‘ ‘x I- 0 V ‘9 3 ' ,5? <21 5 J , D to Q? E 2, V «I a i” <‘ o ._ A 3 m ‘ — g o - NORTHEASTERN new a ; \ e / 4 __ // I I ’1 0 L ' 419- 175- \ 0 25 50 75 45 (3‘ R5,: NAT 0 s r' 1 TIME FROM BEGINNING OF 4/19“ 0700 ”3023' OBZESLI BE OF‘\EA\.,_TERN 5pr EPISODE, IN HOURS 4” - * ' Camp D I ‘_ . 4 19: 0700 PUU Kamoamoa I. / ,4»I1"9: 1420‘: . J 4,139.- 1800 0:" ' A/19;:*OZQO 154 , ‘ 5313,, How pended and widened 4/191 1120 / between hiere and the vent K 4Ir1/9- 436 ’ 9n Apr” 20 SOUTHERN new 7 \ 4*“ , “A , EXPLANATION ,le Basalt of episode 18 II ~ “ IIIIIII Flow — Predominantly aa I III Spatter/cinder cone I' I Basalt of episodes 1 through 17 l:' I’m-1983 basalt 19° X 20’ , I C 7/ .L —— Contact—Dashed where approximate /' I ,L , ———— Approximate internal flow boundary 5.0 Flow thickness—Measured at edge (in meters) : , ,j’” —— Approximate position of flow front—Numbers / / 1 4/19: 1130 indicate date (1984) and time (Hawaii standard ,' ' time) / ,Camp D Ruu Kam‘oanlgznaaa dd? EXPLANATION Basalt of episode 20 Flow-Predominantly aa III! I IIIIIIII Spatter/cmaerm Basalt of episode 19 Flow—Predominantly pahoehoe Basalt of episodes 1 through 18 |:l Pre-l983 basalt — Contact—Dashed where approximate 5.0 Flow thickness—Measured at edge (in meters) r _ 19° 25' 19° 20' ‘ Pun ' \quws fl s G s 124000 15500713”, ”WW 155°07'30" Base from U. . eological urvey : , Kalapana, Makaopuhi Crater, Volcano, 1981; EPISODE 18 SCALE 1:50 000 K I I , 1982 aaua 1I—I I—I I—II/Z I—I I—I O 1 3 4 5 MILES i-————-———————I I—-——I I—J 1 .5 O 1 3 4 5 KILOMETERS H H H H H I——I I———I F—I CONTOUR INTERVAL 20 FEET DATUM IS MEAN SEA LEVEL 155°02'30" 1A; INIEfilflfiifiEULDGICAL SURVEY, HESTDN, VIRGINIA719887085797 EPISODES 19 AND 20 AREA OF MAP DISTRIBUTION OF FLOWS AND VENT DEPOSITS, MEASURED FLOW THICKNESSES, AND FLOW PROGRESS FOR EPISODESI6 THROUGH 20 OF THE PUU OO ERUPTION ., ,_ 0‘s ' K PL DEPARTMENT OF THE INTERIOR s PROFESSIONAL PAPER 1463 U.S. GEOLOGICAL SURVEY PLATE 2 155°07'30” 155°02'30" » 155°07'30” , ~ . I . ,. , _ 1 _ , _, 5‘ . I , ._ V “f _-_ I I \y I \, \fi _, \\ I , \\ _\_\ I , ,. 7, L *\ \ _\ I“? 1“ L..\\;/" \\ K :1; / _ - \___ , x,“ \ __ \ \ ,I/ <2 PROGRESS OF EPISODE 5 FLOWS \_ -_ {N l// /’ , TI, \ f. I‘ ‘ V\"\»\ x; 7‘ U) 9 U) 9 I I I ‘ I / K 1 \I _, \\ I \\ D: 2 Lu , I.u I- +- 7 .\ Lu “ ”x. “J — - r 2 I— \ \>\ E 1‘ O O K =I :1 x g 6 Z 6 — _ Z — eye‘s /7 — End of eruptIon _ $V‘ / I: l; 9“ . / Z ,/ 2 g?‘ // LU _ Lu .. _ > A > // 5 2 O 1, O 0: 3 ‘ ‘ cc 3 — — u. ‘\ u. 19" \ .. I.u \ _ III 25’ U /I ‘ U 2 / Z < '— “ I < — _ ('7) «1 i, :7, Note: Dashed line indicates 5 /; I 5 unknown flow-advance rate. 0 I I I I ’9 o I I I X 0 25 50 75 100 125 O 25 50 75 100 ‘\ TIME FROM BEGINNING OF EPISODE, IN HOURS TIME FROM BEGINNING OF EPISODE, IN HOURS ,/«,/ q ._ « , 6/30; 1545 x 6/30: r015 . Puu Oo , Puu Kamoamoa iv 8 f 0 6/153: 1313 6/1341600 4 6/1 4:11} 700 10‘ 5 6115210660 6% 3: A 600 // 7/ 4131f 556/15: 060d 6/15: 10/15 . Na-paqtrexer , «I , ' 7/1jr1’045 / , ff 177/2: 1130 y\\/ 1.. , , - " 7/;81240 “ W ,» EXPLANATION EXPLANATI N Basalt of episode 5 / ,4 7 J Basalt of episode 4 ’ Flow—Predominantly aa Flow— Predominantly aa IIIIIIIII IIII Spatter/cinder cone I I IIIIIIIIIIIIIIIIIII Spatter/cinder cone / I-I-I-I-I-I-I-I-H-I Fissure with spatter ramparts B It I ' d 1th h 4 , asa o episo es roug 19° _ 0 20 19' Basalt of episodes 1 through 3 20 ' ./ I:I Pre-l983 basalt Contact Contact 4.5 Flow thickness—Measured at edge (in meters) w” ' 4.5 Flow thickness—Measured at edge (in meters) — Approximate position of flow front— Numbers 7/3: 0635 Approximate position of flow front—Numbers indicate date (1983) and time (Hawaii standard time) 6/17: 1700 EPISODE 4 EPISODE 5 I K \ . l K, , , r ‘\_ (I \.\ ‘\.’ \ / \\¥ 1 TB PROGRESS OF EPISODE 6 FLOW \ 3/ (D i I E I I I X Y ' ' -\ I— -\ \ \ E I g; SUBOFIDINATE LOBES _\ \ O // 7, ,\ l-U Northeastern ‘\ l :3 6 — ‘ __\\ ‘\\_\ g 6 _ Southern _ ”I W A x ,\ \ \\ \-. x I, -. O \‘ r Z _ _ \ j 3 Middle , \ “x J/ I-— ; , :" ‘ \ E x g N h \. n ‘ j, > I 1 1 // ,_~ ./ °" 9'" KN 8/17: 1600 ‘ 3 _ _ x \V‘ ‘/ t, \\ . . " 1 ‘1 ‘ g 1 {I E 3.8- 3/ :‘ NORTHEASTERN LoBE E <_’ 7 I > x 3} 4.0 I - « 1 LL] _ __ (I CE) 3 _ ‘J x , 3 A8” 7R 0900 A\ 19") _ 3 Note: Dashed line indicates \\__ y‘ , a E \1 8/16: _170C)~ ,. / " :8/1 72/1600 , = _ 190 25 < unknown flow-advance rate. 1’? , . , », u ‘4, WI 6' 1 130 ‘ \ ~ 4 ’ ‘ 25, I— / x j r K . J - g , , u: o I I I 7/25:0815 - I. U I. . .. 3.6 - ‘ - 5 0 25 50 75 100 /_ ' <2: _ 9/16/2300 _ I ., . ~ , , 75II7§01930 600 _ TIME FROM BEGINNING OF EPISODE, IN HOURS \1 {I "3 [if/7 'V ,. ’ ‘ " ~ .42 SOUTHERN LOBE ,./‘ 5 " ' ' * > mm 8/16: .170 ' \ ' _ \ she; 1130 . _, \ f 0 I I \uu Kahaualea ‘ / ‘-. 3. I: - 0 25 50 75 \28/15: 1500 x \ TIME FROM BEGINNING OF EPISODE, IN HOURS 3 3'01 A ,(7/23; 1400 ,=‘ ,v A , (8/15: 1045 , 4'0 , , _\ NORTHEASYERN FLOW I. __ \“ 8/15: 0845 I! 1 m“) \ [I "J I . 7 <5 ‘ 7/23; 14100 22 I i“ 1.5 T ' s' ,4 7 Cam: C ’ 1, " Puu Kamoamoa , Puu Kamoamoa m L “ J Kamoa ( 4 V /16: 06 r SOUTHEAST! R LOW/ {0M5 ’ , ' ”OF 1 /8/16:"1$130_ 45 / ' , e/ngyOO /, I EXPLANATION EXPLANATION B . Basalt of episode 6 i alflalt 0‘ episode 7 Flow—Predominantly aa ow—Predommantly aa Spatter/cinder cone Spatter/cmder cone Basalt of episodes 1 through 6 Basalt of episodes 1 through 5 Pre-l983 ba alt Pre-1983 basalt --L 5 I90 ,- , _ , , Contact 19“ 20' ' ,. —— Contact , ~ , V , , » y- f a, ,. , 20 / , ' ,/ ———-— A roximate internal flow boundar . ‘ ‘/ /// “TQM ,/ ', ; //j, I, 4.5 Flow thickness—Measured at edge (in meters) PP y “"/ ”<7 I ,/ , , 4.5 Flow thickness—Measured at ed e in meters , ’ /’ — Approximate position of flow front—Numbers g ( ) ‘ ——-—— Approximate position of flow front—Numbers 8/16' 1700 indicate date (1983) and time (Hawaii standard , time) ’ 155°07’30” 155°02'30" 155“D7'3l]' 155“02’3[1" fig INTERIOIEGEULDGICAL SURVEY, RESIUN, VIRGINIAsisee—samz Base from U.S. Geological Survey 1:24,000 Kalapana, Makaopuhi Crater, Volcano, 1981; EPISODE 6 SCALE 1:50 000 EPISODE 7 Ka'a'ua: ‘932 1 V2 0 1 2 3 4 5 MILES F—:I I—I 1:: H F—T F———-—-—{ 7 >17 4 ___ W 1 .5 O I 2 3 4 5 KILOMETERS Fl T—I H H H I——T I—-—-—I I—-———I CONTOUR INTERVAL 20 FEET DATUM IS MEAN SEA LEVEL AREA OF MAP DISTRIBUTION OF FLOWS AND VENT DEPOSITS, MEASURED FLOW THICKNESSES, AND FLOW PROGRESS FOR EPISODES 4 THROUGH 7 OF THE PUU OO ERUPTION >7MDEAYS ' I "'4‘ ‘ '[ _ 5‘5 " " "U s BUREAU OF LANDUMANAGEMEN NATIONAL COAL—HYDROLOGY PR0 _ * : 1974*84 , Wm. M?” m3 nfl' ‘ ‘ \ ‘ 3-. “ ' rb’ ;! Ilka; x _ 7 gm, g3 ””11 #11 ”flgu 0" x9 ' i ‘ ”11% _ W S!“ W. .V ,7 ‘ f " Q1 bmfig A aqua ‘ 91M “ilk ' _ U S GEOLOGICAL SURVEY PROFESSIONAL PAPER 1464 Prepared in cooperatlon With the U S Bureau of Land Management “I: W 3;ng 07.1990:- Summary of the U.S. Geological Survey and U.S. Bureau of Land Management National Coal—Hydrology Program, 1974—84 Ediieo.’ by LJ. BRITTON, C.L. ANDERSON, D.A. GOOLSBY, andBP. VAN HAVEREN U.S. GEOLOGICAL SURVEY PROFESSIONAL PAPER 1464 Prepared in cooperation with the U. S. Bureau of Land Management UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1989 DEPARTMENT OF THE INTERIOR MANUEL LUJAN, JIL, Secretary U.S. GEOLOGICAL SURVEY Dallas L. Peck, Director Any use of trade names is for descriptive purposes only and does not imply endorsement by the U.S. Geological Survey. Library of Congress Cataloging-in-Publication Data Summary of the U.S. Geological Survey and U.S. Bureau of Land Management National Coal- Hydrology Program. 1974—84. (U.S. Geological Survey professional paper ; 1464) “Prepared in cooperation with the U.S. Bureau of Land Management.” Includes bibliographies. Supt. of Docs. no. : I 19.16zl464 1. Coal mines and mining—Environmental aspects—United States. 2. Water—Pollution— Environmental aspects—United States. 3. Hydrology—United States. 4. National Coal- Hydrology Program (U.S.) I. Britton, Linda J. II. Geological Survey (U.S.) III. United States. Bureau of Land Management. IV. Series. TD428.CGS86 1989 363.7 '3947 '0973 87-600024 For sale by the Books and Open-File Reports Section, U.S. Geological Survey Federal Center Box 254-25, Denver, CO 80225—0425 CONTENTS Abstract ........................................... Introduction Coal resources, by Wilbur Palmquist ................... Water resources, by Hugh H. Hudson .................. Hydrologic issues related to coal development, by Hugh H. Hudson ...................................... Mine drainage ................................... Water withdrawals ............................... Land disturbance and reclamation .................. Waste disposal .................................. Federal role in coal and water issues, by Hugh H. Hudson History of the coal-hydrology program, by Hugh H. Hudson and Bruce P. Van Haveren ...................... Issues ......................................... Program objectives and information needs ........... Implementation ................................. Hydrologic studies in coal regions, by Linda J. Britten . . . Hydrologic characteristics ........................ Eastern Province—Northern Appalachian region, by David H. Appel ............................ Coal resources .............. ~ ................. Hydrology .................................. Surface-water network ..................... Surface-water characteristics ............... Surface-water quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Eastern Province—Southern Appalachian region, by William J. Shampine ........................ Coal resources ............................... Hydrology .................................. Surface-water network ..................... Surface-water characteristics ............... Surface-water quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Interior Province—Eastern region, by Konrad J. Banaszak .................................. Coal resources ............................... Hydrology .................................. Surfacewater network ..................... Surface-water characteristics ............... Surface-water quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Interior Province—Western region, by Hugh E. Bevans Coal resources ............................... Hydrology .................................. Surfacewater network ..................... Surface-water characteristics ............... Surfacewater quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... scoop-H5 11 13 14 15 19 21 21 22 23 25 26 31 32 32 33 33 35 35 36 37 39 39 39 41 41 42 44 44 44 44 45 47 48 48 48 49 49 50 51 51 51 51 53 53 55 56 57 58 58 58 Hydrologic studies in coal regions—Continued Interior Province—Westem region—Continued Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Northern Great Plains and Rocky Mountain Provinces— Fort Union region, by Orlo A. Crosby and Clarence A. Armstrong .............................. Coal resources ............................... Hydrology Surface-water network ..................... Surface-water characteristics ............... Surface-water quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Northern Great Plains and Rocky Mountain Provinces— Powder River, Bighorn Basin, and Wind River regions, by James F. Wilson, Jr., and Michael R. Cannon Coal resources ............................... Hydrology Surfacewater network ..................... Surface-water characteristics ............... Surface-water quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Northern Great Plains and Rocky Mountain Provinces— Green River and Hams Fork regions, by Neville G. Gaggiani .................................. Coal resources ............................... Hydrology .................................. Surface-water network ..................... Surface-water characteristics Surface-water quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Northern Great Plains and Rocky Mountain Provinces— Uinta and Southwestern Utah regions, by Don Price Coal resources ............................... Hydrology Surface-water network ..................... Surface-water characteristics Surfacewater quality ...................... Ground-water network ..................... Ground-water occurrence ................... Ground-water quality ...................... Coal-hydrology studies ........................ Hydrologic issues related to coal mining ......... Northern Great Plains and Rocky Mountain Provinces— Denver and Raton Mesa regions, by Linda J. Britten and Neville G. Gaggiani .............. Coal resources ............................... Hydrology Surface-water network ..................... Surface-water characteristics Page 58 61 62 63 64 64 65 66 67 68 68 69 70 73 74 75 76 76 78 79 79 79 80 83 85 86 87 87 87 90 90 91 91 94 95 96 97 97 97 98 99 100 100 101 103 104 105 105 105 IV CONTENTS Page Page Hydrologic studies in coal regions—Continued Summary of selected research about the hydrologic effects of Northern Great Plains and Rocky Mountain Provinces— coal mining—Continued Denver and Raton Mesa regions—Continued Discussion .................................. 135 Hydrology—Continued Shallow and deep ground-water flow systems, by Michael Surface-water quality ...................... 107 R. Cannon ................................. 136 Ground-water network ..................... 103 Studies of effects of mining on ground-water flow Ground—water occurrence ................... 108 systems ------------------------------- 137 Ground-water quality ...................... 108 Eastern Province, Northern Appalachian region— Coal-hydrology studies ........................ 109 Eastern Ohio ........................ 138 Hydrologic issues related to coal mining ......... 109 Interior Province, Western region—Macon- Northern Great Plains and Rocky Mountain Provinces— Huntsville 81:98" Missouri ““““ _ """ 139 San Juan River region, by F. Eileen Roybal . . . . 111 Northern Great Plains and Rocky Mountain Prov- Coal resources ............................... 112 inces—West Decker, Montana """"" 140 Hydrology .................................. 113 Summary °f Effects °f surface mining 0“ gmund‘ Surface-water network .................... 113 water ow systems .................. 141 Surface-water characteristics ............... 113 Geochemistry of mine spoils. by Robert E. Davis ..... 142 Surface-water quality ..................... 114 General geochemical processes . . . . l ............ 142 Ground-water network ..................... 115 Evolution Of mine-spOils water quality at selected G r oun d-water occurrence .................. 115 Sites ..... . ....... . ..... ' ................. 144 Ground-water ualit ..................... 116 Abandoned mines, Missouri ................ 144 q Y Coal-h drolo tudies ........................ 116 Abandoned‘mine, Oklahoma """""""" 145 y gy s B Sk Mine M ntana ................... 145 . . . . 1g y - 0 Hydrologic issues related to coal mining ......... 116 - Summary of selected research about the hydrologic effects of West Decker Mine, Montana """"""""" 146 coal mining . 119 ' Seneca Mine, Colorado ..................... 147 Watershed modeling. by Linda G. Stannard and Gerhard Mmi’rtifilizgii :fy 81:11:}:an drainage i1: Kuhn ‘. """""""" . """"""""""" 1 20 Quantity and quality of acid mine drainage ...... 149 Precipitation-runoff modehng system (PRMS) d9 Classification of acid mine drainage ............. 150 scription """"""""""""""""" 120 Nonacid mine drainage ........................ 152 0°“ceptua1 “‘9”th System --------------- 12° Sedimentation, by Randolph s. Parker .............. 155 Model parameters ------------------------ 120 Sediment-data-collection approach and consider- Simulation process ------------------------ 121 ations ................................. 155 Precipitation-runoff modeling system (PRMS) appli- Sediment-data-collection activities .............. 156 cations -------------------------------- 122 Eastern Province ......................... 157 Eastern and Interior Provinces ............. 123 Northern Great Plains and Rocky Mountain Northern Great Plains and Rocky Mountain Provinces ,,,,,,,,,,,,,,,,,,,,,,,,,,, 158 ProVinces ........................... 124 Aquatic biology, by David A. Peterson ............. 160 Discussion .................................. 125 Biological studies in coal provinces ............. 160 Salinity modeling, by Rodger F. Ferreira ............ 126 Eastern Province ......................... 160 Description and application of simple and multiple Northern Great Plains and Rocky Mountain linear-regression models .................. 126 Provinces ........................... 162 Description and application of curvilinear-regression Benthic invertebrates in streams ......... 162 models ................................ 127 Invertebrate drift in streams ............ 165 Description and application of mathematical account- Phytoplankton and periphyton .......... 165 ing models ............................. 128 Summary and additional data needs, by Bruce P. Van Haveren Description and application of mathematical routing and Donald A. Goolsby ......................... 167 models ................................ 132 References cited .................................... 171 ILLUSTRATIONS Page PLATE 1. Map showing coal provinces, selected coal regions, coal areas, and physiographic provinces of the contiguous United States ........................................................................................ In pocket FIGURES 1—4. Maps showing: 1. Eastern Province: coal regions, selected coal fields, and dominant types of coal .................... 4 2. Interior Province: coal regions and dominant types of coal ...................................... 5 3. Northern Great Plains and Rocky Mountain Provinces (part 1): coal regions, selected coal fields, and domi- nant types of coal ..................................................................... 6 4. Northern Great Plains and Rocky Mountain Provinces (part 2): coal regions, selected coal fields, and domi- nant types of coal ..................................................................... 7 FIGURES 22. 23. 24. 25-28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. CONTENTS v Page . Graph showing coal reserves (1974) and demonstrated reserve base (1979, 1980, and 1983) .................. 8 . Graph showing coal production, domestic coal consumption, and coal exports, 1949—83 .................... 8 . Photographs showing reddish-yellow precipitates, “yellow boy,” in streams draining eastern coal-mine areas . . 12 . Photograph showing water draining from a bolt hole in the sandstone roof of an underground mine in Utah . . . 13 . Sketch and photographs showing examples of sediment deposition ...................................... 15 . Photograph showing water in mine pit from dewatering of aquifer in Colorado ............................ 16 . Sketch showing possible effects of mining aquifers ................................................... 17 . Sketch and photograph showing effects of underground coal mining on land subsidence .................... 18 . Typical hydrographs ............................................................................. 27 . Graphs showing: 14. Typical flow-duration curves ................................................................ 28 15. Typical relations of dissolved-solids concentration to specific conductance ......................... 28 16. Typical relations of stream discharge to specific conductance .................................... 29 . Map showing the Northern Appalachian region, coal areas 1—12, and sampling sites ....................... 31 . Typical hydrographs showing variation of stream discharge in the Northern Appalachian region ............. 34 . Graph showing maximum, median, and minimum specific conductance at selected surface-water-quality sites in the Northern Appalachian region .................................................................. 35 . Map showing the Southern Appalachian region, coal areas 13—24, and sampling sites ...................... 40 . Typical hydrograph showing variation of stream discharge for Collins River near McMinnville, Tennessee, during the 1977 water year in the Southern Appalachian region ........................................... 42 Graph showing typical flow-duration curves for selected streams in the Southern Appalachian region ........ 43 Graph showing maximum, median, and minimum specific conductance at selected surface-water-quality sites in the Southern Appalachian region .................................................................. 43 Map showing the Eastern region, coal areas 25—35, and sampling sites .................................. 47 Graphs showing: 25. Typical flow-duration curves for selected streams in the Eastern region ........................... 49 26. Mean annual stream discharge for selected streams in the Eastern region ......................... 49 27. Maximum, median. and minimum specific conductance at selected surface-water-quality sites in the Eastern region ............................................................................... 50 28. Relation of annual sulfate load to area of surface-mined land for part of the Eastern region .......... 50 Map showing the Western region, coal areas 36—42, and sampling sites .................................. 54 Graph showing typical flow-duration curves for selected streams in the Western region .................... 56 Graph showing maximum, median, and minimum specific conductance at selected surface—water-quality sites in the Western region .............................................................................. 57 Map showing location of areas of coal-hydrology studies done in the Western region ....................... 59 Map showing the Fort Union region, coal areas 45—47, and sampling sites ............................... 62 Typical hydrographs showing variation of stream discharge for selected streams during the 1980 water year in the Fort Union region ............................................................................ 65 Graph showing maximum, median, and minimum specific conductance at selected surface-water-quality sites in the Fort Union region ............................................................................ 66 Graph showing summary of medians and ranges of specific conductance at 26 sites for different-sized drainage areas in the Fort Union region ...................................................................... 67 Map showing the Powder River, Bighorn Basin, and Wind River regions, the Bull Mountain coal field. coal areas 48—51, and sampling sites ..................................................................... 73 Typical hydrographs showing variation of stream discharge for selected streams in the Powder River, Bighorn Basin, and Wind River regions ....................................................................... 77 Graph showing maximum, median, and minimum dissolved-solids concentrations at selected surface-water-quality sites in the Powder River, Bighorn Basin, and Wind River regions ................................... 78 Map showing location of areas of coal-hydrology studies conducted in the Powder River, Bighorn Basin, and Wind River regions ................................................................................ 80 Map showing the Green River and Hams Fork regions, coal areas 52—54, and sampling sites ................ 85 Typical hydrographs showing variation of stream discharge for selected streams during the 1978 water year in the Green River and Hams Fork regions ............................................................ 88 Graph showing maximum, median, and minimum dissolved-solids concentrations at selected surface—water-quality sites in the Green River and Hams Fork regions .................................................. 89 Map showing the Uinta and Southwestern Utah regions, coal areas 55—58, and sampling sites .............. 95 Graph showing typical flow-duration curves for selected streams in the Uinta and Southwestern Utah regions . 98 Graph showing maximum. median, and minimum specific conductance at selected surface-water-quality sites, water years 1979—84, in the Uinta and Southwestern Utah regions ........................................ 99 Map showing the Denver and Raton Mesa regions, coal areas 59 and 61, and sampling sites ................ 104 Graph showing typical flow-duration curves for selected streams in the Denver and Raton Mesa regions ...... 106 Graph showing maximum, mean, and minimum specific conductance at selected surface-water-quality sites in the Denver and Raton Mesa regions ................................................................ 107 Map showing the San Juan River region, coal areas 60 and 62, and sampling sites ........................ 111 Typical flow-duration curves for selected streams in the San Juan River region ........................... 114 VI FIGURES 52. 53. 54. CONTENTS Graph showing maximum, median, and minimum specific conductance at selected surface-water—quality sites in the San Juan River region ........................................................................ Schematic diagram of the conceptual watershed system and its inputs ................................... Map showing location of selected US. Geological Survey coal-basin modeling studies, 1984 ................. 55—58. Graphs showing: 71- 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 73. 74. 75. 76. 77. 78. 55. Simulated dissolved-solids loading to the Yellowstone River from the Tongue and Powder Rivers, Montana, 1978 water year ....................................................................... 56. Difference between dissolved-solids concentrations estimated from the two-variableregression model and dissolved-solids concentrations measured at site 09188500 Green River at Warren Bridge near Daniel, Wyoming ............................................................................ 57. Difference between dissolved-solids concentrations estimated from the multiplevariable—regression model and dissolved-solids concentrations measured at site 09188500 Green River at Warren Bridge near Daniel, Wyoming ............................................................................ 58. Estimated monthly mean dissolved-solids loads at site 09188500 Green River at Warren Bridge near Daniel, Wyoming ............................................................................ Diagram of a simple stream network with nodes and node numbers for the model developed by Parker and Norris (1983) ...................................................................................... Graph showing a comparison of modeled mean monthly dissolved-solids concentrations for existing conditions and conditions associated with short-term anticipated mining at Middle Creek at mouth, node 15 ............ Map showing location of Rosebud Creek. Montana, and reaches simulated by the model developed by Woods (1981) Simplified flow chart of model for calculating monthly dissolved-solids concentration in five reaches of Rosebud Creek, Montana .................................................................................... Diagram showing ground-water flow system of a typical drainage basin ................................. Map showing coal provinces and location of selected minesite study areas ............................... Schematic section of ground-water occurrence and flow at a watershed in Jefferson County, eastern Ohio ..... Generalized hydrogeologic section through an abandoned surface mine, Macon-Huntsville area, Missouri ...... Generalized hydrogeologic section, West Decker Mine, Montana ........................................ Map showing coal provinces and location of selected mine-spoils study areas ............................. Diagram showing generalized section of hypothetical mine area and major hydrologic and geochemical processes Graph showing titration curves of untreated and oxidized ground-water samples containing different iron ratios Graphs showing similar titration-curve characteristics provided by different samples collected from: 71. Seep (site 2), Clarion County, Pennsylvania ................................................... 72. Well 4K, Clarion County. Pennsylvania ...................................................... 73. Well C—2, Clarion County, Pennsylvania ...................................................... Graph showing relation between percent of area disturbed and sediment yield for the Eastern Coal Field of Kentucky Map showing locations of coal areas that have coal-area hydrology reports that include discussion of benthic in- vertebrate and (or) algal data .................................................................. Graph showing relation between number of invertebrate taxa per sample and mean concentration of dissolved manganese .................................................................................. Graph showing composition, density, and diversity of benthic invertebrates from the Belle Fourche River, upstream and downstream from a coal mine in northeastern Wyoming ........................................ Diagram showing change in mean percentage composition of benthic-irwerbebrate taxonomic groups in the downstream direction in Trout Creek, northwestern Colorado .................................................. TABLES TABLE 1. Status of US. Geological Survey coal-area hydrology reports ......................................... , ..... 2. Coal beds, reserves, historic production, and average as-reoeived analysis of major coal reserves for each State in the Western region ........................................................................................... 3. Active coal mines and coal production by county, lower Tertiary coal beds, Powder River, Bighorn Basin, and Wind River regions and Bull Mountain coal field ................................................................. 4. Technical hydrologic-mvestigations reports compiled as part of the US. Geological Survey and US. Bureau of Land Manage ment coal-hydrology program in the Uinta and Southwestern Utah regions ................................ 5. Selected Precipitation-Runoff Modeling System (PRMS) parameter definitions ................................. 6. Acid mine drainage classes ............................................................................ Page 115 121 123 127 128 129 129 130 132 134 135 136 137 138 139 141 143 144 151 152 153 154 158 161 162 163 164 Page 25 55 75 100 122 150 CONTENTS VII CONVERSION FACTORS The inch-pound units in this report may be converted to SI (International System) units by using the follow- ing conversion factors. Multiply inch-pound units By To obtain SI units acre 4,047 square meter acre-foot 1,233 cubic meter acre-foot per square mile 476 cubic meter per square kilometer cubic foot per second 0.028317 cubic meter per second cubic foot per second per 0.01093 cubic meter per second square mile per square kilometer cubic yard 0.7646 cubic meter foot 0.3048 meter gallon per day 3.785 liter per day gallon per minute 3.785 liter per minute inch 25.40 millimeter mile 1.609 kilometer pound per cubic foot 16.02 kilogram per cubic meter square mile 2.590 square kilometer ton 0.9072 megagram ton per square mile 0.3503 megagram per square kilometer Temperature in degree Celsius (°C) may be converted to degree Fahrenheit (°F) by using the following equation: °F = 9/5 °C + 32. Temperature in degree Fahrenheit (°F) may be converted to degree Celsius (°C) by using the following equation: °C = 5/9(°F - 32). SUMMARY OF THE U.S. GEOLOGICAL SURVEY AND U.S. BUREAU OF LAND MANAGEMENT NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Edited by LJ. BRITTON, C.L. ANDERSON, D.A. GOOLSBY, and B.P. VAN HAVEREN ABSTRACT During the decade 1974—84, the U.S. Geological Survey and the U.S. Bureau of Land Management cooperated on investigations to collect information and to study hydrologic processes related to development and mining of federally owned coal. In addition, the U.S. Geological Survey conducted similar investigations related to nonfederally owned coal. As a result of these nationwide investigations, a large quantity of hydrologic information and data has been collected and compiled in more than 500 reports. This report summarizes the major findings and accomplishments that have resulted from data-collection activities, hydrologic studies, and research concerned with the effects of coal mining on water resources. This summary report includes: (1) A description of the Nation’s coal- and water-resource issues related to coal development, including history, objectives, and design of the coal-hydrology program; (2) a summary of the hydrologic information collected in the major coal provinces and published in more than 500 reports and journal articles; and (3) a summary and application of results obtained from topical studies undertaken throughout the program, including discussions on watershed modeling. salinity modeling, ground-water flow systems, geochemistry of mine spoils, mine drainage, sedimentation, and aquatic biology. A detailed coal-hydrology reference list concludes the report. INTRODUCTION During 1974, the U.S. Geological Survey and the U.S. Bureau of Land Management began cooperative coal- hydrology investigations to collect information and to study hydrologic processes related to development of some federally owned coal resources. With passage of the Surface Mining Control and Reclamation Act in 1977, Congress directed the U.S. Geological Survey to expand the coal-hydrology studies to other coal prov- inces nationwide and to include nonfederally owned coal. Results from these investigations were to be used in making coal-leasing decisions to help minimize effects Manuscript approved for publication August 20, 1986. of coal mining on water resources. As a result of these investigations, a large quantity of hydrologic informa- tion and data has been collected and published, and ma- jor accomplishments related to the understanding of water-resource and coal-hydrology issues have been realized. Results of these studies have been published in more than 500 reports. However, until now a con- solidated summary of the major findings of these nationwide investigations has not been published. A na- tionwide summary will be of considerable use to Federal and State agencies concerned with land-use planning and with regulation and enforcement of land-use and resource-development plans. In addition, an important result of the many coal-hydrology investigations was to indicate additional hydrologic information and research needs that warrant further study. This report describes the Nation’s coal- and water- resource issues related to coal development; the history, objectives, and program design of the Federal coal- hydrology program; and the Federal government’s role in coal- and water-resource issues. The report sum- marizes the major findings and accomplishments that have resulted from the nationwide studies concerned with the effects of coal mining on water resources. Finally, this report includes a discussion concerning ap- plication of the results obtained from the many multi- disciplinary studies that were undertaken during the Federal coal-hydrology program regarding land-use planning, coal leasing, and land reclamation. This report is based primarily on information con- tained in a series of reports that generally characterize the hydrology of individual coal areas throughout the Nation. These reports are referenced, as is other perti- nent literature on the various subjects discussed, and the reader is urged to consult the literature for a more thorough discourse. 1 COAL RESOURCES By WILBUR PALMQUIST‘ Coal is the most abundant fossil fuel in the United States. Knowledge of the quantity, quality, and loca- tion of coal resources is necessary for meeting energy needs and for solving potential environmental issues, which include air, water, and land degradation. The basic unit of coal is a bed, seam, or vein. One or more beds meeting minimum criteria of thickness, quali- ty, extent, and depth of burial constitute a coal field. One or more coal fields are combined into coal regions, which are further grouped into the coal provinces shown on plate 1 (modified from Fenneman and Johnson, 1946; Trumbull, 1960). Coal occurs in at least 39 States and underlies 13per- cent of the land area in the United States. The major- ity of the coal deposits occur primarily in 17 States and are divided among 11 Eastern and 6 Western States. In most areas, the coal-bearing rocks and enclosed coal beds are in broad, shallow, structural basins or synclines. In the Eastern Province, the coal generally is classified as high-volatile bituminous (fig. 1) and oc- curs less than 3,000 feet below the surface. In the In- terior Province, the coal generally is classified as high-volatile bituminous (fig. 2) and occurs less than 2,000 feet below the surface. In the Northern Great Plains Province, the coal generally is classified as sub- bituminous and lignite (fig. 3) and occurs less than 1,500 feet below the surface. In the Rocky Mountain Province, the coal generally is classified as subbituminous and high-volatile bituminous (fig. 4) and sometimes occurs at great depths—more than 6,000 feet deep in the Uinta and Southwestern Utah regions of Colorado and Utah, as much as 15,000 feet deep in the Green River region of southwestern Wyoming, and 20,000 feet deep in the Wind River and Bighorn Basin regions of central and northern Wyoming (pl. 1). However, shallower coal beds also exist in the Rocky Mountain Province. The U.S. Bureau of Mines and the U.S. Geological Survey are responsible for estimating and analyzing the Nation’s coal resources. During 1974, Averitt (1975) estimated that at least 4 trillion tons of coal existed in the United States; this estimate did not include coal deposits deeper than 6,000 feet or coal on continental shelves. Averitt (197 5) showed the reserve base, the quantity of coal deemed to be economically and legally available for mining, to be 434 billion tons. Because of increasing needs for more accurate infor- mation about coal resources, a detailed and uniform classification system has evolved (Wood and others, 1983). The current system defines coal reserves as all coal that is currently economically recoverable and coal lU.S. Bureau of Land Management. resources as coal that is potentially recoverable but sub- ject to more favorable economic conditions and develop- ment of advanced technology. The U.S. Energy Information Administration of the ' U.S. Department of Energy annually determines the demonstrated reserve base, a continuation of a U.S. Bureau of Mines function redelegated to the U.S. Department of Energy during 1977. The demonstrated reserve base, which is equivalent to the reserve base determined by the U.S. Geological Survey, is the ton- nage of coal that can be extracted economically at the present time and the tonnage of coal that potentially could be extracted in the future using advanced tech- nology and improved economic conditions. The demon- strated reserve base changes with time, decreasing as coal is mined and increasing as new coal is discovered. The U.S. Energy Information Administration deter- mined the demonstrated reserve base of coal in the United States to be 489.5 billion tons on January 1, 1983—55.5 billion tons more than the U.S. Geological Survey’s estimate (Averitt, 1975) of 434 billion tons for 1974 (fig. 5). Of the estimated coal reserves, 24.5 billion tons can be recovered at producing mines that were ac- tive at the end of 1984 (U.S. Energy Information Ad- ministration, 1984). During the year 1900, coal provided 93 percent of all the energy used in the United States (Perry, 1983). By 1972, coal was providing only 17.3 percent of the N a- tion’s energy because of increased use of oil and gas. The oil embargo of 197 4 and the consequent quad- rupling of oil prices caused a shift to coal for energy pro- duction. By 1983, coal was providing 22.1 percent of the Nation’s energy (fig. 6). There is enough coal in the demonstrated reserve base to meet all the energy needs of the Nation for more than 10 centuries at the present rate of consumption. Although the dollar price of coal historically has increased, overall coal costs relative to wages have decreased (Simon, 1981). Increased coal de- mand and the relatively inexpensive cost of coal from the Western United States have caused production of Western United States coal to increase substantially. Coal production in the Eastern United States has re- mained relatively stable, even though transportation costs have increased. The Federal Government may own as much as 60 per- cent of the coal in the Western United States, but it owns very little coal in the Eastern United States. The U.S. Bureau of Land Management has identified about 17 million acres of land in the Western United States that contain coal deposits. Federally owned coal occurs beneath 11.5 million acres of land in the Western United States, of which 4.8 million surface acres are under 3 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 80° 35° WARRIOR AND PLATEAU COAL FIELDS SOUTH CAROLINA CAHABA COAL FIELD ALABAMA o 100 200 300 400 MILES IL | I | | | I o 100 200 300 400 KILOMETERS EXPLANATION - ANTHRACITE LOW-VOLATILE BITUMINOUS COAL HIGHVOLATILE BITUMINOUS COAL FIGURE 1.——Eastern Province: coal regions, selected coal fields, and dominant types of coal (modified from US. Department of the Interior, 1975). COAL RESOURCES 5 MISSOURI 35° ARKANSAS EXPLANATION - ANTHRACITE LOW-VOLATILE BITUMINOUS COAL HIGH—VOLATILE ‘y BITUMINOUS COAL (I) 1. d i l < 0 LI I: Ill 0 OCT NOV DEC JAN FEB MAR APR - MAY JUNE JULY AUG SEPT WATER YEAR 1,800 i‘u MEDIAN OF ANNUAL MEAN B I'uLi 1.600 — DISCHARGES FOR 1931-60 . V — 0 WATER YEARS ' > " Drainage area, 5 1.400 — I MEAN DISCHARGE FOR 1974 - 496 square miles — 8 WATER YEAR ’ 1,200 — ._ a E E g 0 1,000 — — LIJ '1 m < 800 — — I I: U E g) 600 — — o E 400 — -— < E ,_ 200 — _ a: OCT Nov DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT THE YEAR WATER YEAR FIGURE 18.—Typica1 hydrographs showing variation of stream discharge in the Appalachian region. A, Discharge and precipitation for Coal River at Tornado, West Virginia, site 1, for October 1, 1978 to September 30, 1979 (modified from Ehlke and others, 1982b). B, Seasonal pattern of discharge for Little Beaver Creek near East Liverpool, Ohio, site 2 (modified from Roth, Engelke, and others, 1981). Loca- tion of sites shown in figure 17. EASTERN PROVINCE—NORTHERN APPALACHIAN REGION 2,400 O: u.r CL 2,200 —— 4 — 3 w Lu 2 2,000 — _ E .03 — 1 800 — ‘- U) u.I . ° 23 5 Lu 1,600 — — — Lu 2 a: 1,400 — _ g 8 ui D 1 200 e 3 ’ 9 '— _ _ E < 1,000 I 8 3:.) 800 D |_ 11 —‘ 2 DJ 10 o g 600 — — o E 6 8 12 CE) u.l 400 — I 7 I — _ U 8 200 — 3 I — e I 5 0 I EXPLANATION 3 SITE NUMBER MAXIMUM I MEDIAN MINIMUM SITE NUMBER IN FIGURE 17 SITE NAME AND DRAINAGE AREA 3 Sandy Run at Van Ormer, Pennsylvania (2.80 square miles) 4 Crowley Hollow near Keating, Pennsylvania (2.67 square miles) 5 Upper Sheriff Run near Lynch, Pennsylvania (4.44 square miles) 6 Service Creek near Shippingport, Pennsylvania (4.50 square miles) 7 Naylor Ditch near Sebring, Ohio (3.96 square miles) 8 Thomas Fork at Santiago, West Virginia (4.50 square miles) 9 No name tributary Wills Creek near Conesville, Ohio (3.00 square miles) Stewart Creek at Baldwin, West Virginia (3.33 square miles) 11 Packs Branch at Packs Branch, West Virginia (4.61 square miles) Trace Fork at Ruth, West Virginia (2.82 square miles) DATA IN FIGURE REPRESENT 3T0 7 SAMPLES PER SITE DURING 1979~81 FIGURE 19,—Maximum, median, and minimum specific conductance at selected surface-water-quality sites in the Northern Appalachian region. mining. Stream discharge is the main factor affecting the sediment yield of a given watershed. Sediment concentrations and loads generally are largest during high flows and smallest during low flows. Increased pre cipitation not only increases stream discharge and its 35 ability to transport sediment, but also increases erosion and, therefore, the supply of transportable material. Average annual suspended-sediment yields for large streams in the region range from 20 to 800 tons per square mile (Schneider and others, 1965, sheet 8). This range may not include the extremes that occur in smaller tributaries in the region. Yields tend to increase in a southerly direction, ranging from 20 to 250 tons per square mile in the northern part of the region to 100 to 800 tons per square mile in the southern part. The glaciated area in the northwestern corner of the region is an exception to the above statement; here, yields range from 100 to 800 tons per square mile. Sediment yields of streams are affected by numerous other factors, including physiography, soils, climate, and land use. Land-use activities that disturb the land surface, such as surface mining, construction, agriculture, and silviculture, increase erosion and sedi- ment yields. Although the network of daily suspended- sediment sites is not adequate to correlate sediment load with specific land uses, active surface mining is known to produce some of the largest rates of erosion. In west-central Pennsylvania (coal area 3) the synoptic- site data indicate that for any given instantaneous unit stream discharge, the instantaneous suspended- sediment discharge may vary by a factor of 1,500 with no appreciable change in land use (Herb and others, 1981a). GROUND—WATER NETWORK The 1983 ground-water observation-well network in the region consisted of 83 wells that were measured on a routine basis. Water-level and water-quality data also are available for many thousands of short-term project wells. These data consist of one or more samples and were collected as part of specific studies or projects. GROUND—WATER OCCURRENCE Two principal types of aquifers underlie the region— unconsolidated alluvial and glaciofluvial deposits and consolidated bedrock aquifers composed of sedimentary rocks. Unconsolidated aquifers are the best sources of ground water for municipal and industrial uses in the region. Water production from wells in these aquifers depends on permeability, areal extent, saturated thick- ness of the sand and gravel materials, and proximity of wells to rivers. The quality of water in alluvial aquifers generally is suitable for most uses but often requires some treatment. In some locations, however, 36 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 wastes from chemical and industrial plants have con- taminated local ground water. Major sources of ground water in the Appalachian Plateaus physiographic province part of the region (pl. 1) are the Pennsylvanian aquifers. The Upper Penn- sylvanian aquifers consist of nearly horizontal layers of mostly shale and thin interbeds of fine-grained sand- stone, siltstone, coal, and limestone. The Lower Penn- sylvanian aquifers consist mostly of massive coarse-grained sandstone and shale, siltstone, coal, and thin limestone beds. The primary permeability of the Pennsylvanian bedrock aquifers generally is negligible. Ground water flows through and is stored in joint systems, fractures, and bedding planes and in carbonate rocks in dissolu- tion channels. These aquifers commonly are very local in extent. In some areas, these local aquifers are perched and are under a single hilltop. Mississippian aquifers in the southeastern part of the region are similar in lithology and permeability to the Pennsylvanian aquifers but are gently to moderately folded. Parts of the sandstones are saturated and con- fined by overlying and underlying shales. The aquifers can yield moderate to large quantities of water. The predominantly carbonate Greenbrier Limestone of the Mississippian aquifers has good potential for large—scale withdrawal of ground water. Fracture open- ings in these strata generally are enlarged by solution, and wells that penetrate enlarged openings may have large yields. Farther to the east, in the Valley and Ridge physio- graphic province part of the region (pl. 1), the aquifers are faulted and compressed into steep folds that great- ly affect the occurrence and movement of ground water. In these areas, ground-water conditions are more variable than in the rest of the region. The principal car- bonate units of the Devonian and Ordovician aquifers and some of the massive sandstone units of the Devo- nian aquifers have potential for providing large quan- tities of ground water. The carbonate units also are sources of springs and have large yields (as much as 15,000 gallons per minute) that supply small water- supply systems and light industry. GROUND—WATER QUALITY Quality of water in the Pennsylvanian aquifers generally is suitable for most purposes. However, the water usually varies from soft (60 milligrams per liter calcium carbonate) to very hard (as much as 400 milligrams per liter calcium carbonate). In some loca- tions, specifically where drainage from coal mines is a source of recharge to underlying aquifers, the water may be very hard (1,300 milligrams per liter calcium carbonate) and may contain large concentrations of iron (180,000 micrograms per liter), manganese (9,900 micro- grams per liter), sulfate (2,500 milligrams per liter), and chloride (2,300 milligrams per liter). Brine underlies freshwater in most areas of the Appalachian Plateaus physiographic province part of the region (pl. 1), but generally it is located deeper than is accessed by usual drilling practices (300 feet in valley areas). Water quality of the Mississippian aquifers general- ly is suitable for most uses. Hardness and locally large iron concentrations (greater than 300 micrograms per liter) are common issues. Because of sink holes and large solution openings that may be in direct hydraulic con- nection with sources of contamination in outcrop areas, the carbonate unit is very susceptible to biological and chemical pollution. COAL-HYDROLOGY STUDIES Economic interest in coal resources and hydrologic issues associated with coal mining resulted in several coal-hydrology studies in the region. For example, in February 1972, the most destructive flood in West Virginia’s history occurred when a coal-waste dam col- lapsed on Buffalo Creek in the southwestern part of the State. Davies and others (1972) reported the hydrology and engineering geology of the disaster. The large loss of life, human suffering, and property damage focused attention on an aspect of coal hydrology that often is overlooked. Many of the coal-waste dams in existence at that time were not engineered dams and simply were unsafe. Several studies that investigated the effects of various mining practices on the hydrology of small basins were done in the region. Reed (1980) determined that acidity of water in a drain in an area in northern Pennsylvania that previously had been deep-mined, but now was being strip-mined, had increased nearly 600 percent during a 3-year period. The acidity of the water in the two other drains in the same area increased by 100 and 45 percent. The effects of underground mining and mine collapse on areal hydrology were determined by Hobba (1981) at one site where the mined bed of coal is topographical- ly above major streams and at two other sites where the mined bed of coal is below major streams. The min- ing and associated subsidence cracks increase hydraulic conductivity and interconnection of overlying water- bearing rock units that, in turn, cause increased infiltra- tion of precipitation and surface water, decreased evapotranspiration, and increased base flows in some small streams. Gaining and losing streams occur in deep-mined areas, depending on local conditions. Mine pumpage and drainage can cause diversion of water EASTERN PROVINCE—NORTHERN APPALACHIAN REGION 37 underground from one basin to another. Aquifer tests indicated that near-surface rocks have greater trans- missivity in a mine-subsided basin than in unmined basins. Increased infiltration and circulation of ground water through shallow subsurface rocks increased dissolved-solids loads in streams, as did treated and un- treated contributions from mine pumpage and drainage. A study in cooperation with the U.S. Bureau of Land Management was started in 1981 in West Virginia to calibrate deterministic rainfall-runoff models for various land-use conditions in coal areas of the region and to develop water-quality regression models for simulating water-quality constituent concentrations in West Virginia. As a continuation of studies in the coal areas of Alabama by Puente and others (1982), studies in West Virginia indicate that water-quality properties and constituents such as specific conductance, sulfate con- centrations, noncarbonate hardness/total hardness ratio, and magnesium/calcium ratio can be used to iden- tify streams substantially affected by coal-mine drainage (Celso Puente, U.S. Geological Survey, oral commun., 1985). Concentrations of these chemical con- stituents varied greatly on a statewide basis. However, the water quality of streams draining coal areas under- lain by the same rock type were similar, and the coal areas were delineated into two distinct geochemical zones. A study was done from 1979 to 1980 to monitor the water quality of streams within the coal-mining areas of western Maryland and adjacent areas of Pennsyl- vania and West Virginia. The report (Staubitz, 1981) contains streamflow, water-quality, and biological data for various river basins in the Eastern Province. Ground-water conditions that occurred during coal strip mining in two small watersheds in eastern Ohio are described in a report by Helgesen and Razem (1981). Water levels in the top aquifers declined as mining in- creased near the watersheds. Depletion of the top aquifer was indicated by decreased stream base flow and by increased mineralization after mining. Helgesen and Razem (1981) concluded that no immediate substan- tial effects of mining were evident on ground-water levels or ground-water quality beneath the strippable coal. An assessment of water quality in streams draining coal-producing areas in Ohio is reported by Pfaff and others (1981). A reconnaissance of water quality at 150 sites and a study of 4 small basins indicated that acid mine drainage generally occurred where abandoned drift or strip mines were located; areas characterized by reclaimed or active strip mines indicated few occur- rences of acid mine drainage. The preimpoundment water quality of the Tioga River basin (fig. 17), Pennsylvania and New York, is described in reports by Ward (1976, 1981). Water quali- ty in the Tioga River is degraded by acid mine drainage entering the river downstream from strip- and deep- mined coal areas. Diel measurements indicated that acid mine drainage has decreased biological activity in the Tioga River (Ward, 1981). Relations between selected water-quality constituents were developed for the sam- pling sites throughout the basin. Downstream trends also were analyzed and reported. A study in the Tug Fork basin (fig. 17), Kentucky, Virginia, and West Virginia, used a rainfall-runoff model to determine if land-use changes associated with sur- face mining in the basin affected basin streamflow characteristics (Doyle and others, 1983). The model was calibrated and verified for two periods, one represent- ing 1980 land use and one representing 1950 land use. Statistical tests made for the two periods indicated no difference in streamflow characteristics at any of the locations. In addition, analyses were made to determine if future increases in surface coal mining might affect basin stream discharge. The modeling results indicated that increasing mining in an upland watershed by as much as 200 percent had little effect on stream dis- charge (Doyle and others, 1983). Additional coal-hydrology reports for counties, basins, States, or parts of the region that contain in- formation about flood frequency, runoff characteristics, water quality, ground water, and water use are avail- able. Two recent compilations of coal-hydrology studies and sources of data for this area, prepared in coopera- tion with the U.S. Bureau of Land Management, are by Grason (1982) and Cochran and others (1983). HYDROLOGIC ISSUES RELATED TO COAL MINING Coal-mining activities may affect all aspects of water resources in the Northern Appalachian region. Runoff characteristics of streams may be changed. Changes are more noticeable during low flows when, depending on local mining and geology, mining may cause a stream to lose or gain water (Hobba, 1981). The chemical quality of water in more than 6,000 miles of streams in the region has been identified as be- ing substantially affected by coal-mine drainage. In many of the affected stream reaches, the water is not suitable for most uses without expensive treatment. The pH of water draining from mined areas commonly ranges from 2.0 to 5.0 in the northern part of the region where rocks generally contain few calcareous minerals and coal contains a substantial quantity of sulfur. In contrast, the pH of mine drainage often is neutral or alkaline in the southern part of the region where cal- careous minerals are common and coal contains little 38 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 sulfur. Specific-conductance values and concentrations of total iron, total manganese, dissolved sulfate, and dissolved solids usually are larger in mined areas than in unmined areas. Sources of sediment in surface mines include newly cleared areas, haul roads, mine-spoil piles, and newly reclaimed land. Strip-mine spoil is a mixture of freshly exposed sandstone, limestone, shale, and soil. Mine spoil rapidly weathers and decomposes into unconsolidated particles that are easily eroded. If a mine site is not reclaimed, mine-spoil piles may remain sources of large sediment yields for many years. However, after a mine site is properly reclaimed, erosion decreases substan- tially, and sediment yield from the mine site is a short- term issue (Staubitz and Sobashinski, 1983). Ground-water resources may be affected by surface and underground mining. However, underground min- ing and related mine collapse usually cause the largest changes. If a mine collapses, subsidence fractures de- velop along or parallel to weak zones such as existing joints and fractures. The effect on hydrology apparently is the greatest near the land surface. Hobba ( 1981) indi- cated that the underground mining process may cause: 1. Lowered water tables above the underground mine and drying up of shallow wells; 2. Fluctuations as much as 100 feet annually of water levels in some wells; 3. Increased infiltration of precipitation, causing decreased evapotranspiration and resulting in higher base flows or increased leakage into mines; 4. Interbasin transfer of ground water because of mine pumpage or drainage; 5. Increased dissolved-solids concentrations and gen- erally more acidic conditions; and 6. Large underground voids that can store large vol- umes of water. Some of the changes in hydrology caused by mining , are beneficial. Numerous underground coal mines, for example, store plentiful supplies of potable ground water, and many are being used as sources of public supply. Lessing and Hobba (1981) determined that 72 public water systems in West Virginia pump more than 7 million gallons of potable water per day from aban- doned coal mines to supply 81,600 people and various establishments. EASTERN PROVINCE—SOUTHERN APPALACHIAN REGION 39 EASTERN PROVINCE—SOUTHERN APPALACHIAN REGION By WILLIAM J. SHAMPINE The Southern Appalachian region, an area of about 42,500 square miles, extends in a southwesterly direc- tion from Kentucky through ’Ibnnessee into central Alabama and northwestern Georgia (fig. 20) and in- cludes coal areas 13—24 (table 1). The region is within four physiographic provinces (pl. 1). From east to west these provinces are Valley and Ridge, Appalachian Plateaus, Interior Low Plateaus, and Central Lowland (Fenneman and Johnson, 1946). The Valley and Ridge province is characterized by long, steep-sided ridges separated by northeast-trending valleys. The Appala- chian Plateaus province generally is a rolling upland area, and it has a steep escarpment separating it from the lower altitudes of the Valley and Ridge province The Interior Low Plateaus province is characterized by near- ly level to gentle slopes, and it has an escarpment along the western edge separating it from the Central Lowland province. The Central Lowland province is characterized by gently sloping lowlands and many knobs or rounded hills. The major stream systems in the region include the Kentucky, Licking, Clinch, Holston, ’Ibnnessee, Coosa, and Cahaba Rivers (pl. 1, fig. 20). The region has a moist, temperate climate, and mean annual precipitation ranges from about 40 inches in parts of Kentucky to 60 inches in 'lbnnessee. Precipita- tion is fairly well distributed throughout the year, although some areal variations can be noted during the rainy season. October consistently is the driest month of the year throughout the entire region. During the rainy season in the northern half of the region, the wet- test month of the year is July and the second wettest is March. In the southern half of the region, the wettest month of the year is March and the second wettest is January or February. Thunderstorms occur throughout the year but are most frequent during the spring and summer months, The Valley and Ridge physiographic province part of the region is underlain by Ordovician and Cambrian rocks that predominately are carbonate rock, siltstone, shale, and some sandstone. Karst topography formed by solution of carbonate rocks is common in this province. The Appalachian Plateaus physiographic province part of the region is underlain by gently dipping sandstone, shale, siltstone, and coal of Pennsylvanian age; these rocks have a thickness of about 1,500 feet. Below the rocks of Pennsylvanian age are the rocks of Mississip- pian age, predominately limestone, calcareous shale, and siltstone. The Interior Low Plateaus physiographic prov- ince part of the region is underlain by carbonate rocks of Mississippian age, which average 600 to 700 feet in thickness. These rocks are underlain by shale of Devo- nian age and limestone of Ordovician age. The Central Lowland physiographic province part of the region is underlain by limestones of Ordovician age. Land in the Southern Appalachian region ranges from flat farmland to poorly accessible, heavily forested mountains. Most of the area is rural, although some ma- jor urban areas are included. Land use is estimated to be about 80 percent forest, 15 percent agriculture (cropland and pasture), 3 percent urban, and 2 percent mining. Although quantitative water-use data are sparse throughout the region, some use patterns are apparent. The large-volume users are industry and public supply, the majority of which use water from surface sources, streams, and reservoirs. Most of the rural supplies use ground water pumped from local wells. COAL RESOURCES Coal in the Southern Appalachian region is high- volatile bituminous (fig. 1) and occurs in rocks of Penn- sylvanian age in the Appalachian Plateaus physio- graphic province. More specifically, in the northern part of the region, most of the mineable coal occurs in the Breathitt Formation and, to a much lesser extent, in the Lee Formation. In the southern part of the region, most of the mineable coal occurs in the Pottsville Formation. The occurrence of coal throughout the region is charac- terized by multiple coal seams of relatively wide lateral extent and varying thicknesses, ranging from 6 inches to as much as 19 feet. However, most of the seams are relatively thin, typically less than 5 feet in thickness. The number of coal seams varies locally. Typically, there are about 25 coal seams in an area; however, as many as 60 coal seams have been identified in the Cahaba coal field (fig. 1) in Alabama. The coal seams occur at in- tervals ranging from a few feet to more than 300 feet. More than 4,200 coal mines are located in the South- ern Appalachian region. Most of these mines are located in Kentucky, but mining is done throughout the region. TWO-thirds of the mines are surface mines, but under.- ground methods also are used to mine the coal. The method of mining varies according to the geology and topography of the mine site. Only limited quantitative data are readily available about coal production in the region. In Kentucky, more than 140 million tons of coal were mined during 1978; most of the coal was produced in six counties: Bell, Buchanan, Harlan, Perry, Pike, and Wise (fig. 20). HYDROLOGY Increased national interest in coal during recent years has caused substantial increases in the quantity of 40 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM. 1974—84 83° 35° 34° wuss e xaaxwgn‘axb‘ / w I \ R5; his ,« mu. ' K Mil/Ref o; TALBOT x, 3 “tom.“ v i \ guiier 32° Base from U. S. Geological 0 50 100 MILES Survey 1:2,600,000 United States base map 0 50 100 KILOMETERS EXPLANATION COAL REGION 16 COAL-AREA BOUNDARY—Number refers to local areas (Cochran and others, 1983) .6 SAMPLING SITE AND NUMBER—See figures 21-23 for site descriptions FIGURE 20.-The Southern Appalachian region, coal areas 18—24, and sampling sites. hydrologic data collected in the Southern Appalachian impetus for the increase in data collection came from region. Additional work has been funded through a passage of the Surface Mining Control and Reclama- variety of sources at national and local levels. One major tion Act of 1977 , which established environmental laws EASTERN PROVINCE—SOUTHERN APPALACHIAN REGION 41 associated with mining. This Act established a need for expanded data bases to ensure compliance. As coal became a more important source of energy, the issue of how coal mining might affect water resources also became more important. A need devel- oped to assess the basic hydrologic information that was available and to supplement deficiencies in the data base as needed. Coal-area hydrology reports have been published for all 12 of the coal areas (13—24) in the Southern Appalachian region (table 1). SURFACE-WATER NETWORK The Surface Mining Control and Reclamation Act of 1977 resulted in an increase in the quantity of hydro- logic data collected after 1977 in the Southern Appa- lachian region. During 1976, stream-discharge data were collected at 281 sites in the region. After passage of the Act in 1977, a total of 320 new sites were established to meet specific data requirements associated with new coal-related studies. Most of these sites were intended for short-term use, although a few have been maintained to study the long-term effects of mining. The greatest quantity of data collected consisted of a continuous record of water stage and a calculation of the daily stream discharge at a site. The period of record for continuous-record data at specific sites extends as far back as the late 1800’s, although a record of 20-to-30-years duration is more typical. For many sites, however, the data were collected for a specific study, and the period of record may be for a few years only. Data also were collected at partial record sites where 8 to 12 instantaneous-discharge measurements were made annually during a period of several years. One of the largest increases in the data-collection pro- gram occurred in the area of water quality. There were 209 active water-quality sites in the region during 1976, and an additional 535 sites were established after 1976. The data collected include measurement of major cat- ions and anions, physical properties, trace elements (both dissolved and sorbed on the bottom materials), and some biological data. Most of the water-quality sampling was done synop- tically to provide generalized areal coverage. Most of the specific studies began during the late 1970’s in- cluded a series of samples collected on a periodic basis, commonly monthly, and at least one site where a con- tinuous record was made of the temperature, pH, and specific conductance. Because of the importance of sediment, there was a substantial increase in the number of sediment-data- collection sites in the region after passage of the Surface Mining Control and Reclamation Act of 1977. During 1976, there were only 35 sites where sediment data were being collected. After 1976, the number of sites in- creased to 213. The most common type of data collected was suspended-sediment concentrations, although particle-size analyses were done on some samples. SURFACE-WATER CHARACTERISTICS A stream discharge hydrograph of a stream located near the middle of the Southern Appalachian region is shown in figure 21. The flow variability illustrated by this hydrograph is typical of streams throughout the region. The general shape of the hydrograph would be similar for basins of varying sizes, although the absolute values of the stream discharge would increase as the basin size increases. Figured on a unit basis, many surface-water charac- teristics are relatively uniform throughout the region. For example, the average annual discharge of a stream is equal to about 2 cubic feet per second per square mile of drainage area; thus, a stream in a drainage basin of 100 square miles would have an average annual stream discharge of about 200 cubic feet per second. Low flows are affected by several factors that are dif- ficult to measure quantitatively, such as the storage and transmission capacity of the rocks of the area, the per- viousness of the soil, and the type and density of vegeta- tion. Low-flow frequency is expressed as the lowest average stream discharge for a given number of con- secutive days for a given recurrence interval. The 3-day, 20-year low flow and the 7-day, 10-year low flow are common indices. Specific low-flow data are available from the individual coal-area hydrology reports listed in table 1. One generality that can be made, however, is that most streams draining less than about 100 square miles in the region approach zero discharge dur- ing the low-flow season. Techniques have been developed for estimating the magnitude and frequency of floods at gaged and un- gaged sites throughout the region. These techniques are represented by generalized regression equations from which estimates can be made of flood flows at any site. These equations are reported by McCabe (1962) and Hannum (1976) for Kentucky, Randolph and Gamble (1976) for Tennessee, and Peirce (1954), Gamble (1965), Heine (1973), and Olin and Bingham (1977) for Alabama. If the flows are arranged according to frequency of occurrence and are plotted as a flow-duration curve, the resulting curve shows the integrated effect of the various factors that affect runoff in the basin. Typical flow-duration curves for two streams in the region are shown in figure 22. The curves are plotted in unit stream discharge so that a more direct comparison can be made. The shapes of the curves illustrate the differences in flow characteristics between streams draining rocks of 42 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 50,000 | | l I I /i0 1 i | I l - Maximum instantaneous discharge, — 45,800 cubic feet per second 0 E 0 10,000 — Drainage area, 640 square miles — u.I _ _ w .— 0: Lu — _ D. l— _ _ LIJ u.I — - LL L_J _ _ m D U _ a g Average discharge, 1925-79, ui 1,175 cubic feet per second (D _ _ 0: < I U ‘2 Q __ ___ - ___... _ 2 1,000 —— — < I.” E _ _ U) ' _ Minimum instantaneous discharge. /120 cubic feet per second 100 I I I I I I I I I I I OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT WATER YEAR FIGURE 21.——'l‘ypical hydrograph showing variation of stream discharge for Collins River near McMinnville, Tennessee, site 1 (fig. 20), during the 1977 water year in the Southern Appalachian region (modified from May and others. 1981). Pennsylvanian age that contain coal reserves and rocks of other ages in this area. Richland Creek is typical of streams draining uplands of the Appalachian Plateaus physiographic province that contain coal reserves. The flow-duration curve for Richland Creek is steep at stream discharges less than 1 cubic foot per second per square mile because of the small ground-water contribu- tion to the stream. South Chickamauga Creek, however, is typical of the Valley and Ridge physiographic prov- ince, and its flow-duration curve has a flatter slope on the lower end, indicating larger yields from the ground- water system. The slope of the curves at the upper end essentially is the same, indicating that the high-flow runoff per square mile is similar for both streams. SURFACE—WATER QUALITY The range and median of specific-conductance meas- urements for selected streams in the central part of the region are shown in figure 23. These data are divided EASTERN PROVINCE—SOUTHERN APPALACHIAN REGION 43 1° llllll\lilllill| ||||l| IIIIIl llJIIl l Illllll ll|II|I SITE 2 South Chickamauga Creek _ near Chickamauga, Tennessee Drainage area: 428 square miles / STREAM DISCHARGE, IN CUBIC FEET PER SECOND PER SQUARE MILE 0.1 _— __ _ SITE3 Z : Richland Creek near : ' Dayton, Tennessee ‘ Drainage area=50.2 square miles 0001lllIlIlIlIIlIlllllIll 0.01 0.1 1 10 50 90 99 99.99 PERCENTAGE OF TIME INDICATED STREAM DISCHARGE WAS EQUALED OR EXCEEDED FIGURE 22.—Typical flow-duration curves for selected streams in the Southern Appalachian region (modified from Hollyday and others, 1983; see figure 20 for location of sites). into groups that clearly illustrate the effect of human activity. The ranges illustrated also are typical of the entire region, although localized sources of pollution have caused the specific conductance occasionally to be as much as 26,000 microsiemens per centimeter. Water-quality issues associated with coal mining in the region include increased concentrations of sulfate, manganese, and iron. Acid waters and large concentra- tions of trace elements that typically are associated with coal mining generally are not issues in the region. Sulfate concentrations in streams draining undisturbed basins typically range from 20 to 40 milligrams per liter. Streams draining areas disturbed by coal mining, how- ever, typically contain sulfate concentrations ranging from 100 to 2,000 milligrams per liter. Concentrations of total recoverable manganese in streams in the region range from 0 to about 25,000 micrograms per liter, although the median values generally are less than 100 micrograms per liter. Total recoverable iron concentra- tions in streams range from 0 to 510,000 micrograms URBAN RURAL COAL AREA AREAS AREAS UNAFFECTED AFFECTED El 500 BY MINING BY MINING 6 E E i 4 5 E ‘3 400 — 5 2 E E 13 T < Lu LU 5 U U 300 2 12 14 3 E 8 — 7 8 I _ D D. l.l..l . é ‘0 5 U E Lu 200 — _ 2 e a u. _ — N W 100 — _ 8 o t‘ 9 10 Q. I < 11 ‘0 2 I I I 2 0 EXPLANATION 2 SITE NUMBER MAXIMUM I MEDIAN MINIMUM SITE NUMBER IN FIGURE 20 SITE NAME AND DRAINAGE AREA 2 South Chickamauga Creek near Chickamauga, Tennessee (428 square miles) 4 Wolttever Creek above Ooltewah, Tennessee (24.5 square miles) 5 South Chickamauga Creek at Chickamauga, Tennessee (428 square miles) 6 West Chickamauga Creek near Kensington, Georgia (73.0 square miles) 7 Sequatchie River near College Station, Tennessee (154 square miles) 8 Sequatchie River near Whitwell, Tennessee (402 square miles) 9 Long Branch near Hinkle, Georgia (373 square miles) 10 Daniel Creek at SR 143 near Trenton, Georgia (4.80 square miles) 11 Bear Creek at SR 157 near Durham, Georgia (798 square miles) 12 Rock Creek at SR 170 near Durham, Georgia (080 square mile) 13 Rock Creek below SR 170 near Durham, Georgia (0.80 square mile) 14 Rock Creek at Nickajack Road near Hinkle, Georgia (7.40 square miles) FIGURE 23.—Maximum, median, and minimum specific conductance at selected surface-water-quality sites in the Southern Appalachian region (modified from Hollyday and others. 1983). per liter and have a median value greater than 1,000 micrograms per liter in the northern part of the region and less than 500 micrograms per liter in the southern part. Basins draining surface-mined areas generally yield 5 to 10 times as much sediment as basins draining un- disturbed areas. Although suspended-sediment yields vary considerably because of topography, mining activ- ity, rainfall intensity, and so forth, unmined areas in the region commonly yield 500 to 3,000 tons per square mile 44 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 per year, and mined areas commonly yield 3,000 to 20,000 tons per square mile per year. It also is typical for 60 to 70 percent of the total annual sediment load at a given site to be transported during one major storm during the year. GROUND—WATER NETWORK The effects of mining on ground water are not as obvious as the effects on surface water, but they do ex- ist. The network of data collected about ground water in the region also was increased in response to the in- creased interest in coal hydrology. During 1976, water- level or water-quality data were collected at 102 wells in the region. After 197 6, data were collected at an ad- ditional 261 wells. Because conditions affecting ground water are not as variable as those of surface water, the frequency of data collection is not as often. Water levels may be measured daily, but more typically, monthly or even annually is sufficient. Water-quality samples may be collected as infrequently as annually. Water-quality analyses generally measured the concentrations of the major cations and anions. GROUND—WATER OCCURRENCE Fractured sandstones and conglomerates supply most of the water for wells in the region. Yields of wells in these rocks range from 5 to 300 gallons per minute, although most wells yield only 10 to 20 gallons per minute. Wells in the carbonate rocks that overlie or underlie the coal areas have yields ranging from 1 to 3,000 gallons per minute; most of these wells yield less than 30 gallons per minute. Water levels in wells throughout the region are not affected by mining except for wells located close to min- ing operations. In some instances, the area immediate- ly adjacent to a mine is dewatered to prevent flooding in the mine. The dewatering process is temporary, how- ever, and would cease at the time of termination of the mining and reclamation operations. Coal mining in the Southern Appalachian region seems to have less effect on the ground-water resources than it does on the surface-water resources. This may be because of fewer underground mines than surface mines but also may be an indication of less available data about ground-water resources to properly define the extent of problem areas compared to the more available data about surface-water resources. GROUND—WATER QUALITY The quality of water from ground-water systems in the region typically is very good; dissolved-solids concentrations generally are less than 500 milligrams per liter. Ground water that is moderately to very saline occurs in the region, although it tends to be sporadic and localized. The quality of \Water from wells in the Valley and Ridge physiographic province generally is suitable for drinking. However, contamination of ground water is a potential issue throughout those areas of this province that are underlain by carbonate rocks. These carbonate rocks, primarily limestones and dolomites, are subject to dissolution along fractures, joints, and bedding planes (Hollyday and others, 1983), which can cause large, interconnected conduits that facilitate the rapid, extensive spread of contaminants. Water from most wells in the Appalachian Plateaus physiographic province is a soft to moderately hard, mixed type (calcium bicarbonate, sodium bicarbonate, or calcium sulfate) and contains relatively small concen- trations of dissolved solids. Locally, acidic water may occur and some large concentrations of manganese, iron, and chloride have been reported (May and others, 1981). Water from wells in the Interior Low Plateaus physio- graphic province generally is a moderately hard to hard, calcium bicarbonate type and contains moderate con- centrations of dissolved solids. Large iron or manganese concentrations are not a widespread issue, but locally, large concentrations of both constituents have been reported (May and others, 1983). Water from most wells in the Central Lowland physio- graphic province is a very hard, calcium bicarbonate type and contains moderate concentrations of dissolved solids. Iron and manganese concentrations generally are within maximum contaminant levels for drinking water (U.S. Environmental Protection Agency, 1986c). COAL—HYDROLOGY STUDIES Economic interest in coal resources and potential hydrologic issues associated with coal mining have prompted several hydrologic investigations in the re- gion. The most intensive work has been done in the coal areas of Alabama where the U.S. Bureau of Land Man- agement is responsible for managing extensive Federal Mineral Ownership lands. The U.S. Bureau of Land Management has been involved in cooperative studies with the U.S. Geological Survey in Alabama from 1976 to present (1985). These studies have included forma- tion and compilation of hydrologic data bases on mine- able lands, assessment of hydrologic changes resulting from mining, and computer simulation modeling. Knight and Newton (1977) reported that the degrada- tion of water quality is the most serious and widespread coal-mining-related issue in Alabama. Their work as- sesses the extent of the issue in Alabama and describes EASTERN PROVINCE—SOUTHERN APPALACHIAN REGION 45 work needed to further define the issue and to provide the data needed to develop potential solutions. Data in- dicate that water draining from mined areas common- ly has a pH that ranges from 2.1 to 5.0, generally has large sulfate and dissolved-solids concentrations, is hard to very hard, and may contain objectionable concentra- tions of iron. Puente and Newton (1982) analyzed climatic, physio- graphic, hydrologic, and land-use data for 67 basins in the Warrior coal field and derived equations for assess- ing water quality in streams that drain unmined areas and mined areas that have been reclaimed. Their data indicate that water-quality effects in mined basins reclaimed under present mining laws are less severe than in basins reclaimed under previous systems. In addition, dissolved-solids concentrations varied in response to the age of the mine. Generally, the concen- trations maximized after about 7 years of mine opera- tion and returned to pre-mining levels after about 15 years of mine operation. Some recent work, as yet un- published (L.J. Slack, US. Geological Survey, written commun., 1987) indicates that reclamation efforts in Alabama are effective and will shorten this time cycle. An assessment of hydrologic conditions in potential coal-lease tracts in the Warrior coal field (fig. 1), Alabama, was done by Puente and others (1982). Climatic, physiographic, hydrologic, and land-use data were analyzed to derive relations for assessing and predicting water quality in streams that drain mined and unmined areas. An equation was derived estimating specific conductance. The independent variables in- cluded stream discharge, percentage of basin mined, channel distance between stream sampling site and mined area, and relative age of mined areas. By using additional equations, based on relations between specific conductance and other constituents, estimates can be made of water-quality variables commonly used as mine indicators, such as hardness and, dissolved- solids and sulfate concentrations. Using these relations, hydrologic assessments of the coal-lease tracts were made. Based on limited verification data, these assess- ments proved to be reasonably accurate. The effects of future mining activities in tracts also were estimated. The methods used to estimate future effects on surface- water quality were described and examples were in- cluded. In addition, Kidd and Hill (1983) prepared a report summarizing the major coal-hydrology publica- tions and project activities related to hydrology in the Warrior and Plateau coal fields of Alabama (fig. 1). Lake Tuscaloosa, a water supply for Tuscaloosa, Alabama (fig. 20), is located in a drainage basin contain- ing areas that are being surface mined for coal. Although only about 5 percent of the basin has been mined, Cole (1984) has reported that there has been a small increase in the dissolved-solids concentration of the lake since the beginning of mining. Water draining Cripple Creek basin, a mined basin, contributes an esti- mated 310 tons per square mile per year of dissolved solids to the lake. Water draining Binion Creek basin, an unmined basin, contributes an estimated 50 tons per square mile per year of dissolved solids to the lake. Cole (1984) also reported that in some instances, natural factors affecting sediment deposition in Lake Tusca- loosa, such as steep overland and channel slopes, may cause more sedimentation in the lake than is caused by the disruption due to coal mining. Bradfield (1986a, 1986b) has investigated the effect of coal mining on the benthic macroinvertebrate com- munity in Tennessee. Analysis of variance tests in- dicated significant trends toward decreased number of taxa, number of organisms, and sample diversity at sites that have relatively poor water-quality conditions. These trends indicate significant differences in benthic invertebrate communities at sites with increasing evidence of the effects of land use. More than 50 reports have been published describing the hydrology of mined areas in Alabama, Tennessee, and Kentucky. In addition to the coal-area hydrology reports, ongoing studies in Tennessee and Kentucky in- clude the collection of a large volume of hydrologic data that is used to formulate and compile data bases about mineable lands. Data are being collected on basins with premining, active mining, and postmining conditions, although reports have not yet (1985) been prepared describing the results of this work. HYDROLOGIC ISSUES RELATED TO COAL MINING Coal-mining activities in the Southern Appalachian region have not produced any unusual or unexpected hydrologic issues. The characteristics of streams drain- ing basins containing extensive areas of surface min- ing have been modified somewhat. If the vegetation in the area is stripped, runoff and the speed with which it moves may increase, thus affecting the magnitude and timing of flood flow. By lowering the water table in an area, the ground-water contribution to a stream will decrease during periods of low flow; a stream may contain less water than during premining periods, or it even may go dry. The most significant water-quality issue in the region associated with mining is the increase of suspended- sediment loads in streams adjacent to surface mines. The increase may be fivefold to tenfold and may have a serious deleterious effect on the streams and subse- quent users of the water. However, retention ponds and reclamation efforts are successful and can help ameli- orate the extent of the issue. 46 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 As expected, the concentrations of selected dissolved constituents are increased by coal mining. The largest concentration changes occur with sulfate and iron, which are derived from solution of the pyrite (iron sulfide) associated with coal. Other increases include manganese and possibly some trace elements associated with the suspended sediment. In general, increases in the concentrations of sodium, chloride, calcium, and magnesium in the water due to mining are not substan- tial enough to restrict potential uses of the water. INTERIOR PROVINCE—EASTERN REGION 47 INTERIOR PROVINCE—EASTERN REGION By KONRAD J. BANASZAK The Eastern region, an area of about 48,500 square miles in Illinois, southwestern Indiana, and western Kentucky (fig. 24), includes coal areas 25—35 (table 1). N 0 report was done for coal area 26, so that area is not CLIN‘ION ,/ * scowr a 2mm REE a Base from U. S. Geological Survey 0 122,500,000 United States base map A v / WWW mm» discussed in this regional section. The region is located within two physiographic provinces (pl. 1); the Interior Low Plateaus comprises the southeastern part of the region, and the Central Lowland comprises the majority Mr, Hon _ -./ HéwR "/ 50 100 MILES o 50 100 KILOMETERS EXPLANATION COAL REGION COAL-AREA BOUNDARY—Number refers to local areas (Cochran and others, 1983) .2 SAMPLING SITE AND NUMBER—See figures 25 and 27 for site descriptions FIGURE 24.—The Eastern region, coal areas 25—35, and sampling sites. 48 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 of the region. The land is flat to gently rolling, although some ruggedly rolling hills occur. Three major rivers, the Ohio, the Wabash, and the Illinois (fig. 24) drain the area, and the Mississippi River is just west of the region. Precipitation ranges from about 34 inches per year in the northwestern part of the region to 50 inches per year in the southeastern part. The region is a structural basin of predominantly Paleozoic sedimentary rocks generally overlain by a layer of glacial materials. The basin is asymmetric, and the deepest part occurs in extreme southwestern Il- linois. Rocks are marine and shoreline deposits of alter- nating sandstones and carbonates with some shales from the Cambrian through Mississippian age. The Pennsylvanian is represented by sequences of terrige- nous sandstones and coals alternating with sequences of marine shales and carbonates. The coal-bearing Penn- sylvanian rocks indicate a major change in rock types and relation of land to sea. The glacial material (drift) that occurs in the region is important to mining. In the northern and eastern parts of the region, the drift can be more than 200 feet thick, which makes the coal less accessible for mining. The region was last covered during the Wisconsinan glacial advance. The pre-Wisconsinan glacial debris generally is less than 50 feet thick. Major rivers are located in channels filled with glaciofluvial sands and gravels that generally are about 30 to 100 feet thick. Because of the general topography, more than 70 per- cent of the land is used for agriculture, especially row crops, and about 15 percent is forest. Generally, to the north and west, the land is used more for agriculture; to the south and east, the land is forest. For example, in coal area 25 (in the northwestern part of the region), land use is 64 percent crops, 14 percent pasture, 11.5 percent forest, and 4.5 percent urban (Zuehls and others, 1981a). In coal area 34 (in the southeastern part of the region), land use is 40 percent crops, 16 percent pasture, 35 percent forest, and 30 percent urban (Quinones and others, 1983). Major cities in the region are Belleville, Carbondale, Champaign-Urbana, Danville, Decatur, Peoria, and Springfield in Illinois; Evansville, Terre Haute, and Vincennes in Indiana; and Madisonville and Owensboro in Kentucky (fig. 24). Water-use data for the region are neither comprehen- sive nor uniformly categorized. It is apparent from the partial record that is available that industry (including electrical power cooling) is the major user of water in the region and that surface water is the major source. In contrast, rural water usage is for public and domestic supply, and the major source is ground water. COAL RESOURCES Coal in the Eastern region is high-volatile C through A bituminous coal, which is of high heating value (11,500 to more than 14,000 British thermal units per pound) (fig. 2). The coal classification depends on the maximum depth of burial and the length of time at that depth. Generally, the heating value of the coal increases toward the south, reaching a maximum value around the Ohio River at the Illinois-Kentucky border. Coal resources of the Eastern region are contained entirely in rocks of Pennsylvanian age. Most major coal beds are located in the middle of the Pennsylvanian section; some major coal beds are located in the lower part of the Lower Pennsylvanian. Only minor production oc- curs in the upper part of the Pennsylvanian section. The thickness of the Pennsylvanian is as much as 3,500 feet in one small area in western Kentucky. Goals from the Eastern region generally are high-sulfur coals (more than 2 percent sulfur). The total coal resource in the Illinois and Kentucky parts of the Eastern region is at least 224 billion tons, distributed as 186 billion tons in Illinois and 38 billion tons in Kentucky. Indiana has 17 billion tons of coal reserves. Production for the three States during 1980 was about 129 million tons: 64 million tons in Illinois, 40 million tons in Kentucky, and 25 million tons in In- diana (Indiana University, 1983). HYDROLOGY The major objective of the coal-hydrology program in the Eastern region was to quantify the effects of sur- face mining on surface water. The types of studies related to coal hydrology undertaken throughout the region during the coal-hydrology program may be categorized into area assessments, modeling of small watersheds, and special studies. Area assessments and extensive data-collection efforts aided in the publication of coal-area hydrology reports (table 1). Modeling of small watersheds provided a means of assessing the direct effects of coal mining as well as other land uses. Special studies considered issues having particular characteristics, such as sewage sludge amendments to coal-mined lands. SURFACE—WATER NETWORK Presently (1985), stream discharge and stage data are collected at approximately 175 sites, water-quality data at approximately 165 sites, and sediment data at ap- proximately 25 sites in the region. The data-collection sites were increased to 275, 800, and 40, respectively, during the Federal coal-hydrology program, although none of the additional sites remain as parts of the per- manent network. The period of record generally is in tens of years for water quantity and less than 10 years for water quality. The original long-term sediment sites INTERIOR PROVINCE—EASTERN REGION 49 are on large streams where apportioning sediment loadings to various land-use practices is difficult. The reader is referred to table 1 and to the US. Geological Survey District offices in the respective States for in- formation regarding any particular site. SURFACE—WATER CHARACTERISTICS Typical flow-duration curves are shown in figure 25, illustrating the similarity of larger streams in the region whether unmined or mined (Quinones and others, 1983). A report by Martin and others (1987) indicates that stream discharge extremes are more moderate, high flow is not as peaked, and low flow is more sustained I.IJ =I EI'OOOEIIIIII|||ll|l||||llll Lu _ D: - _ < _ _ 8 _ _ a) I — 33 100:— —: D. : : D _ I Z O _ _ U — _ (“A I 10 :\\ \\ ‘ SITE3 —: E E \\ (UNMINED) E '— - _ LLI - _. LU LL — _ U a ‘I .— —_ D 3 : U I I E _ _ . ' \\ ' < °-‘ :— T\\\ —: I : \ : 8 ; \ \\ SITE 1 3 D F \WIUNMINED) _ E * \ ‘\ ‘ $0.0, ||l||||||l||l|I\ |l||| E 0.01 0.1 1 10 50 90 99 99.99 m PERCENTAGE OF TIME INDICATED STREAM DISCHARGE IS EQUALED OR EXCEEDED EXPLANATION SITE NUMBER IN FIGURE 24 SITE NAME AND DRAINAGE AREA 1 South Fork Panther Creek near Whitesville, Kentucky (58.2 square miles) 2 Tradewater River at Ulney, Kentucky (255 square miles) 3 Little Cane Creek near Creal Springs, Illinois (1.45 square miles) 4 Bankston Fork near Crab Orchard, Illinois (1.90 square miles) FIGURE 25.—Typical flow-duration curves for selected streams in the Eastern region. Data for sites 1 and 2 are from Quinones and others (1983); data for sites 3 and 4 are from Brabets (1984). in small watersheds that have been mined than in un- mined watersheds. Brabets (1984) showed these flow characteristics for small streams in the Illinois part of the region (fig. 25). Using data from the existing network of surface- water sites, many estimations and conclusions can be made. Estimates of mean annual stream discharge in the region are based primarily on drainage area. These estimations are applicable to the streams in the region for which long-term stream-discharge records are available. A relation for estimating mean annual stream discharge based on 38 years of discharge records for 23 surface-water-discharge sites that have drainage areas ranging from 5.54 to 9,549 square miles in coal area 25 is shown in figure 26 (Zuehls and others, 1981a). This relation is typical for streams throughout the region. In addition, stream discharge extremes, including 7-day, 10-year high flow, 7-day, 10-year low flow, and peak discharges, can be estimated using equations developed from many years of discharge records at surface-water-discharge sites in the region and are reported in the coal-area hydrology reports (table 1). SURFACE—WATER QUALITY Generally, surface mining affects surface-water qual- ity. The change almost always is a greater concentration of dissolved solids, as indicated by specific-conductance values (fig. 27). Specific-conductance data shown in figure 27 are reported for selected stream sites in coal areas 25 and 35. For all sites measured in the two areas, specific conductance ranged from 85 to 2,720 micro- siemens per centimeter at sites upstream from mining and 160 to 9,200 microsiemens per centimeter at sites 101000 IIIIIIII I IIIIIII lIIIIIll I I TIIIIII I IIIIII_ I lllllll I lllll 1 .000 l llllllll I lIlllIll 100 PER SECOND I l lllllll l llIllIII 10 | I IITllll l lllllllI STREAM DISCHARGE. IN CUBIC FEET J Illlllll I IIIIIIII I IIIIlIII I I IIIUII 1 10 100 1,000 10,000 AREA, IN SQUARE MILES llLlll 100.000 FIGURE 26.—Mean annual stream discharge for selected streams in the Eastern region (modified from Zuehls and others, 1981a). 50 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974—84 LARGE OTHER STREAMS UPSTREAM DOWNSTREAM STREAMS FROM MINING FROM MINING 10,000 I Lu )- I4 Lu __ 2 9,000 -- m ,2 2 Lu 0 8,000 — fl 0: u.) 0. g U) 7,000 — -~ in 2 2 U) .1 % 8 6,000 —- — O in a: Li.) 9 3:: 5,000 — — 2 (D 4 2 “cl “Jr L0 4,000 *- — U N z I— E < 3,000 — ’3 — U 9 D D 11 2 2,000 — — O 6 U 12 0 l g 1,000 —— 5 3 7 s — o l 1 I I u.) 0. l2 '“ "’ o EXPLANATION 5 SITE NUMBER MAXIMUM I MEDIAN MINIMUM SITE NUMBER IN FIGURE 24 SITE NAME AND DRAINAGE AREA 3 Little Cana Creek near Creal Springs, Illinois (1.45 square miles) 4 Bankston Fork near Crab Orchard, Illinois (190 square miles) 5 Spoon River at Seville, Illinois (1,636 square miles) 6 Big Muddy River at Murphysboro, Illinois (2,169 square miles) 7 Indian Creek near Wyoming, Illinois (62.7 square miles) 8 Turkey Creek near Fiatt, Illinois (115 square miles) 9 North Fork Saline River near Texas City, Illinois (173 square miles) ‘IO Lusk Creek near Eddyville, Illinois (42.9 square miles) II Big Creek near Bryant, Illinois (40.3 square miles) 12 Snakeden Hollow near Dahinda, Illinois (9.82 square miles) 13 South Fork Saline River near Carrier Mills, Illinois (147 square miles) I4 Sugar Creek near Stonefort, Illinois (35.4 square miles) DATA IN FIGURE REPRESENT 5 TO 67 SAMPLES PER SITE DURING 1970-79 FIGURE 27 .—Maximum, median, and minimum specific conductance at selected surface-water-quality sites in the Eastern region. Data are from Zuehls and others (1981a, 1981b). downstream from mining. Several studies (Toler, 1982; Wilber and Boje, 1983; Brabets, 1984) indicate that representations of statistical values for a selected 250 N O C I I _. (n o I | 100 — — ANNUAL SULFATE LOAD, IN THOUSANDS OF TONS S I I l | l o 10 20 30 40 50 so SURFACE-MINED AREA, IN SQUARE MILES FIGURE 28.—Relation of annual sulfate load to area of surface-mined land for part of the Eastern region (modified from Toler, 1982). constituent are not as useful as calculations of sulfate loads, for example, in evaluating water quality of surface-mined land in the region (fig. 28). Generally, streams draining mined areas also have larger concen- trations of dissolved trace elements, such as iron, man- ganese, nickel, and zinc and smaller values of pH. Glacial geology also has an effect on stream quality. According to a report by Wilber and others (1985), streams drain- ing Wisconsinan glacial material in Indiana generally have larger values of pH, greater alkalinity, and larger calcium concentrations than streams draining areas of bedrock or pre-Wisconsinan glacial material. At sites where sediment data are not obtained con- tinuously, it is difficult to determine the total loading of suspended sediment. Wilber and others (1985) con- cluded that on the basis of present, incomplete data, agricultural land and mined land cannot be distin- guished from each other based on sediment yield, but that sediment yields are substantially larger from these lands than from forested land. GROUND—WATER NETWORK The U. S. Geological Survey network for collection of ground-water data consists of about 30 wells where ground-water-level information is obtained on a con- tinuous basis and about 10 wells where ground-water- level information is collected on a semiannual basis. Although many other wells were drilled for specific proj- ects, they have been abandoned. Thousands of wells have been drilled in the region, but only limited data are available from local agencies. INTERIOR PROVINCE—EASTERN REGION 51 GROUND—WATER OCCURRENCE The major aquifers in the region are unconsolidated glacial deposits; sandstones, coals, and limestones of Pennsylvanian age; limestone and dolomite of Missis- sippian through Silurian age; and sandstone of Cam- brian age. The shallow aquifers are used mainly for domestic supplies. In the southern part of the region, large-capacity wells generally are developed in the deep sandstones. In the northern part of the region, large capacity wells generally are developed in outwash or glaciofluvial channel deposits. The bedrock units generally are poor water-yielding aquifers, the best being the fractured sandstones and then the coals themselves (Banaszak, 1980). The overly- ing glacial material, especially sands and gravels in the Wisconsinan till and the glaciofluvial channel deposits, can be productive aquifers (yielding as much as 500 gallons per minute) and capable of supplying cities and towns with water supplies. GROUND—WATER QUALITY Generally, ground water in the region is hard to very hard, and dissolved-solids concentrations increase with depth. In the southwestern part of the region, alluvial aquifers had a median dissolved-solids concentration of 323 milligrams per liter, and bedrock aquifers had a me- dian concentration of 519 milligrams per liter. In the eastern part of the region, alluvial aquifers had a me- dian dissolved-solids concentration of 316 mflligrams per liter, and bedrock aquifers had a median concen- tration of 391 milligrams per liter. Shallow waters generally are calcium bicarbonate type, whereas deep waters can be sodium bicarbonate or sodium chloride types. COAL-HYDROLOGY STUDIES Area assessments made during the Federal coal- hydrology program provided the general hydrologic and geologic description of the Eastern region. Wilber and Boje (1983) determined that concentrations of metals and other trace elements sorbed onto the clay fraction of streambed materials at 69 sites in coal areas of In- diana. They concluded that concentrations of aluminum, arsenic, cobalt, iron, nickel, and selenium were substan- tially greater for sediments from mined watersheds than from agricultural or forested watersheds. In Illinois, Brabets (1984) concluded that variability of discharge was less for mined than unmined areas, but that dis- solved solids, calcium, and sulfate concentrations were greater for mined areas. Earlier, Toler (1982) determined the same result for sulfate and further noted that relatively large concentrations of dissolved aluminum, arsenic, chromium, copper, iron, manganese, and zinc commonly occur in mine drainages where sulfate con- centrations are larger than 2,000 milligrams per liter. In Kentucky, Davis and others (1974) have determined that deep (as much as 1,000 feet) sandstone aquifers contained freshwater. Ground-water studies have in- dicated the importance of fractures in bedrock to ground-water flow (Banaszak, 1980). A modeling effort has been directed at assessing the cumulative effects of discharge waters from several mines on a single receiving stream. A theoretical method (Bobay and Banaszak, 1985; Bobay, 1986) has been developed for predicting the chemical effects in surface water caused by the oxidation of iron and manganese. This method explicitly solves chemical reactions on a kinetic basis and mixing reactions on a thermodynamic basis. The method needs to be cali- brated using onsite data collected for that purpose and then verified. Patterson and others (1982) studied the use of sludge irrigation (the sludge used was 5 percent solids) for land reclamation, and they concluded that no difference in ground-water or surface-water quality could be attrib- uted to the application of sludge. The effects of dewater- ing surface-mine pits have been studied by Weiss (1984) and Weiss and others (1986). They determined, on the basis of a finite difference model, dimensionless values that are used in simple equations applied to hydrogeo- logic settings typical of the Eastern region. Banaszak (1985) reported on hydrogeologic effects of a hypothet- ical coal mine in Indiana and concluded that a properly conducted mining operation will have minimal effects external to the mine. Finally, Wilber and others ( 1985) conclude that: (1) pH levels of streams draining agricul- tural, forested, and reclaimed mined watersheds gener- ally range from 6.3 to 8.8; (2) pH levels in streams draining unreclaimed mined watersheds are more vari- able and range from 3.8 to 7.9; (3) boron, iron, manga- nese, nickel, and zinc concentrations generally are larger in mined watersheds than forested and agricultural watersheds; and (4) iron and manganese concentrations are less in reclaimed mined watersheds than in unre- claimed watersheds. HYDROLOGIC ISSUES RELATED TO COAL MINING The major issue of the coal-hydrology program was to quantify the effects of surface mining on surface- water quantity and quality. It has been substantiated that peak flows generally are decreased, low flows are increased, and water quality is degraded in streams draining reclaimed mined land. Evidence from studies in Indiana indicates that even moderate reclamation 52 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 efforts can yield substantial improvements. Additional The effect of dewatering of mine pits on ground water data collection and analysis are needed to determine can be predicted using new techniques. These effects whether regulation of reclamation efforts will make im- are increaSed permeability and storage but degraded provements to the extent needed. quality of water. INTERIOR PROVINCE-WESTERN REGION 53 INTERIOR PROVINCE—WESTERN REGION By HUGH E. BEVANS The Western region, an area of about 85,000 square miles, is composed of parts of Arkansas, Iowa, Kansas, Missouri, Nebraska, and Oklahoma (fig. 29) and includes coal areas 36—42 (table 1). Although no reports were done for coal areas 36 and 37, data and results from these areas are included in this regional discussion. The region is drained by the Des Moines, Missouri, and Arkansas Rivers and by other western tributaries to the Mississippi River. Three physiographic provinces (pl. 1) occur in the region: (1) The Central Lowland, plains and low hills with local relief ranging from 100 to 300 feet, extends over most of the region; (2) the Ozark Plateaus, tablelands with local relief ranging from 300 to 500 feet, borders the eastern edge of the region; and (3) the Ouachita, high hills and low mountains with local relief ranging from 500 to 3,000 feet, occurs in the ex- treme southeastern part of the region. Most of the region has a warm, temperate, rainy climate except for Iowa, which has a cold, temperate, snowy climate. Mean armual precipitation increases from northwest to southeast, averaging about 28 inches in Iowa and about 50 inches in Arkansas. Most precip- itation occurs as rain during the grong season; spring and fall are the wettest. Droughts can occur any time of the year but are most severe during the summer when evapotranspiration rates are greatest. A veneer of sedimentary rocks overlies Precambrian crystalline rock in the region (Eardley, 1951). Most structural features developed during the Paleozoic age when land movements caused the formation of broad basins and arches. During Mississippian, Pennsylva- nian, and Permian time, advances and retreats of shallow seas resulted in sequences of marine strata alternating with sequences of nonmarine strata. Subse- quent erosion has removed most sedimentary rocks from arches, but thick deposits remain in three promi- nent basins—the Forest City basin (which underlies southeastern Nebraska, south-central Iowa, northeast- ern Kansas, and northwestern Missouri), Cherokee basin (which underlies southeastern Kansas and north- eastern Oklahoma), and Arkoma basin (which underlies east-central Oklahoma and west-central Arkansas). These basins contain the Pennsylvanian coal resources of the Western region. Bedrock of Pennsylvanian age, primarily marine shales and layers of limestone alter- nating with layers of coal, is exposed at the surface in most of the region. Along the eastern edge of the region, erosion has removed rocks of Pennsylvanian age, and rocks of Mississippian age (limestone, dolomite, and sandstone) crop out. Along the northwestern edge of the region, rocks of Permian age (shale and limestone) and Cretaceous age (sandstone and shale) overlie rocks of Pennsylvanian age. Rocks of Cretaceous age, sand and weakly cemented sandstone, occur at the surface at the extreme southern tip of the region in Oklahoma. The region has been glaciated generally north of the Kansas River in Kansas and the Missouri River in Missouri. Loess and glacial drift of Pleistocene age have been deposited over bedrock in these areas. Terrace and other alluvial deposits of Holocene age occur in major stream valleys, especially in the Kansas, Missouri, and Arkansas River valleys. Some of the most agriculturally productive soils in the world have developed in this region under the native vegetation, which is primarily bluestem prairie and, in scattered areas, oak and hickory forest. Crop produc- tion, including wheat, sorghum, corn, and soybeans, is the predominant land use in Nebraska, Iowa, northeast- ern Kansas, and northwestern Missouri (Anderson, 1970). Most of the remaining region is cropland mixed with grazing land although forests are intermixed with cropland and grazing land along the eastern edge of the region. Principal urban areas include Des Moines, Iowa; Topeka, Kansas; Kansas City, Missouri; Kansas City, Kansas; St. Joseph, Missouri; Tulsa, Oklahoma; and Fort Smith, Arkansas. Principal water uses in the region are electric-power generation, public supplies, industry, and irrigation. Although large volumes of water are used for electric- power generation, most of the water is returned to the streams. Excluding electric-power generation, 60 per- cent of the total volume of water used in the region is for public supplies, 23 percent is for industry, and 17 percent is for irrigation. Surface-water sources provide 56.5 percent and ground-water sources provide 43.5 per- cent of the water used. COAL RESOURCES The Western region has a demonstrated coal reserve base of about 16 billion tons (Averitt, 1975). The demon- strated coal reserve base consists of coal beds 28 inches or more thick that are located under less than 1,000 feet of overburden. These reserves occur in rocks of Pennsylvanian age. Most of the land containing the reserves is privately owned; however, in Oklahoma, fairly large acreages of coal lands are on Indian reserva- tions. The coal reserves are mostly medium- to high- volatile bituminous (fig. 2), except in extreme east- central Oklahoma and west-central Arkansas where low-volatile bituminous and semianthracite reserves occur. Coal beds, reserves, historic production, and 54 r ‘1 ( 1 - wimbssujnv, . \ [XS . 1 ‘s ,L ’L bi—+—{% \h “5‘10 jcdin'rifl N y . m I o -. , « W‘tanmm «mgw \' I Maniac" K’ «m {‘I DE‘EWAKKF ‘3 1:3 k ‘f\ .q I \ M » t 5" ’_ Z’Qk‘é‘q +4; .— _r_ ‘1 POLK‘I,» 7 Jami?“ My“? § j’s WM} M75 cfluf ”’ :2 ‘ Base from U. 8. Geological Survey 122,500,000 United States base map 0 50 50 100 KILOMETERS .5 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 EXPLANATION COAL REGION COAL-AREA BOUNDARY— Number refers to local areas (Cochran and others, 1983) SAMPLING SITE AND NUMBER— See figures 30 and 31 for site descriptions FIGURE 29,—The Western region, coal areas 36—42, and sampling sites. INTERIOR PROVINCE—WESTERN REGION average as-received analysis of major coal reserves for each State in the Western region are in table 2. Approximately 1.3 billion tons of coal have been pro- duced in the Western region (table 2). Coal production began in 1840 and peaked about 1920. Production generally declined from the 1920’s through the 1950’s. Since 1960, coal production has increased in Missouri and Oklahoma and peaked in these States during the late 1970’s. Prior to the 1930’s, most coal was produced by underground mining. Since the 1960’s, almost all pro- duction has been from surface mines. HYDROLOGY The generalized objectives for the coal-hydrology pro- gram in the region were to collect hydrologic data 55 relevant to coal mining and to use these data in con- junction with available data and reports to describe the quantity and quality of surface- and ground-water resources and the hydrologic effects of coal mining. To fulfill these objectives, the hydrologic-data network was expanded, coal-area hydrology reports were prepared, and interpretive hydrologic studies were undertaken. Coal-area investigations used available information from reports and data files to describe: (1) Physical features (climate, geology, and physiography); (2) coal mining (mining methods, coal production, locations of mined areas, and coal reserves; (3) cultural features (population, land use, and water use); and (4) hydrology (hydrologic data base, streamflow and water-quality characteristics, ground-water occurrence and quality, and hydrologic effects of coal mining). TABLE 2.—Coal beds, reserves, historic production, and average as-received analysis of major coal reserves for each State in the Western region [Btu, British thermal units; --——, data not available] State Coal beds (millions of short tons)1 production (millions of short tons)l Coal reserves Historic coal Moisture Average as-received analysis Fixed Ash Sulfur carbon (range, (range, (range, in in in percent) percent) percent) Volatile matter (range, in percent) Heating value (range, in Btu’s per pound) (range, in percent) Nebraska2 Nodaway. Elmo, ........ Wamego, Lorton, and Honey Creek. Laddsdale, Carruthers, . . White Breast, Wheeler, Bevier, and Mystic. 10 >1 Iowa3 2,885 370 Mineral, Bevier, ........ 290 Mulberry, and N odaway. Kansas4 1,388 Missouri5 Rowe, Drywood, Weir- . . . 9,488 350 Pittsburg, Tebo, Mineral, Fleming, Croweburg, Bevier, Mulky, Summit, Lexington, and Mulberry. Oklahoma6 Lower and Upper ....... 200 Hartshorne, McAlester (Stigler, Secor, Mineral, Morris), Croweburg, and Iron Post. 1,294 Arkansas7 Lower Hartshorne ...... 665 100 and Paris. 17-35 26-34 21—41 8—20 1—6 4,400—9,700 4—13 31—37 35—47 11-29 4—8 8,800—10,500 5-10 —-—- —-—— 10—14 3—7 11,100—12,600 31—40 39—54 9—19 10,500-12,300 17—45 44—74 4—16 1—6 11,000—14,400 1.8—2.5 12-18 69-75 7-12 1—3 13,000-14,100 1Data from Averitt, 1975. 2Data from Burchett, 1977. 3Data from Landis and Van Eck, 1965; Avcin and Koch, 1979; Hatch and others, 1934. 4Data from Brady and Dutcher, 1974; Brady and others. 1976; Ebanks and others, 1979. 5Data from Wedge and others, 1976; Robertson and Smith, 1981. 6Data from Friedman, 1974. 7Data from Haley and others, 1979; Howard, 1984. 56 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974-84 SURFACE—WATER NETWORK To provide hydrologic data for coal-hydrology investi- gations in the 1970’s, approximately 115 stream- discharge and surface-water-quality sites, and 50 strip- mine-pond, water-quality sites were established in the Western region. Nearly all of these data-collection sites have been discontinued because the investigations sup- porting them have been completed. However, long-term data are available for many other sites. During 1982, the US. Geological Survey collected hydrologic data at 191 continuous-record stream-discharge and reservoir sites and at 51 surface-water-quality sites (chemical, sediment, and (or) biological data) in the Western region. High- and low-stream-discharge, miscellaneous stream discharge, and miscellaneous surface-water—quality data also were collected. SURFACE—WATER CHARACTERISTICS In the region, stream discharge varies over time in response to daily, seasonal, and longer term fluctuations in climatic factors (precipitation and evapotranspira- tion); discharge also varies spatially in response to dif- ferences in climate, topography, and geology. Daily variations in stream discharge are caused by variations in precipitation. When sufficient precipitation occurs and produces overland runoff, discharge increases to a peak and then decreases as precipitation and overland runoff diminish and cease. Seasonal variations in stream discharge primarily occur because of differences in precipitation and evapotranspiration. Stream discharge is greatest during the spring and fall when precipita- tion quantities are large and evapotranspiration rates are least. The least discharge generally occurs during the summer when precipitation quantities are small and evapotranspiration rates are greatest. Average annual runoff generally increases from north- west to southeast in response to increasing precipita- tion in that direction (Busby, 1966). Flow-duration curves (fig. 30) indicate the percentage of time that a specified stream discharge was equaled or exceeded. The part of the flow-duration curve that shows stream discharges that are equaled or exceeded only a small percentage of the time (during high flows) represents discharge provided by overland runoff. The part of the flow-duration curve showing stream discharges that are equaled or exceeded most of the time (during low flows) represents discharge provided mainly by base flow from ground water. The flow-duration curves for the Verdigris River (site 1), Petit Jean River (site 2), and Big Hill Creek (site 3) are characteristic of streams draining rocks of Penn- sylvanian age (mainly shale). The steep slopes of the 3°|l||l||||llll|l W/ / / I I I I II .- STREAM DISCHARGE, IN CUBIC FEET PER SECOND PER SQUARE MILE _o 0.01 1 10 50 90 99 PERCENTAGE OF TIME INDICATED STREAM DISCHARGE WAS EQUALED OR EXCEEDED EXPLANATION SITE NUMBER IN FIGURE 29 SITE NAME AND DRAINAGE AREA 1 Verdigris River near Claremore, Oklahoma (6,534 square miles) 2 Petit Jean River near Booneville, Arkansas (241 square miles) 3 Big Hill Creek near Cherryvale, Kansas (37.0 square miles) 4 Spring River near Ouapaw, Oklahoma (2,510 square miles) FIGURE 30.—-Typical flow—duration curves for selected streams in the Western region. Data for sites 1, 3, and 4 are from Marcher, Kenny, and others (1984); data for site 1 are from Bryant and others (1983). curves indicate that stream discharge primarily results from surface runoff, is quite variable, and is not well sustained by base flow. The Verdigris River has more stream discharge per square mile during low-flow periods than the other streams draining rocks of Penn- sylvanian age because its drainage basin has more alluvium, which provides larger volumes of base flow. INTERIOR PROVINCE—WESTERN REGION 57 The generally greater discharge of the Petit Jean River is because of greater precipitation. The smaller streams generally have more stream discharge per square mile during high flows than large streams because smaller basins have steeper slopes and because an individual storm is more likely to cover completely a small drainage basin than a large one. The slope of the flow- duration curve representing the Spring River (site 4) is much flatter than those curves representing the other streams. The Spring River drains rocks of Mississippian age (limestone, dolomite, and sandstone), which provide greater volumes of base flow than do rocks of Penn- sylvanian age. Average-, high-, and low-flow characteristics have been determined for stream-discharge sites that have sufficient discharge records, and techniques have been developed for estimating these characteristics at un- gaged sites. These characteristics, techniques, and additional information about stream discharge are avail- able in coal-area hydrology reports for the region (table 1). For example, in coal area 38, equations for estimating average annual flows and flood peaks for selected recur- rence intervals have been developed (Detroy, Skelton, and others, 1983). Also, in coal area 38, most streams that drain areas less than 50 square miles will cease to flow for 7 consecutive days during 50 percent of the years. SURFACE-WATER QUALITY The quality of water in streams in the region generally is dependent on the source of water providing the streamflow. Streamflow provided by ground-water discharge (base flow) usually has larger concentrations of dissolved solids than streamflow that results from overland runoff, because ground water has been in con- tact with minerals in the rocks for a relatively long time. As streamflow increases from overland runoff, concen- trations of dissolved solids are diluted. The types of dissolved constituents in streams during base flow indicate the mineralogy of aquifer materials. Streams draining areas where limestone or dolomite crop out generally have relatively large concentrations of calcium, magnesium, and bicarbonate. Streams drain- ing areas where shale crops out can have relatively large concentrations of sodium, chloride, and sulfate. General- ly, measured dissolved-solids concentrations in streams are less than 500 milligrams per liter and range from less than 100 milligrams per liter in Arkansas to more than 2,550 milligrams per liter in the Canadian River basin in the extreme southwestern part of the region in Oklahoma (Rainwater, 1962). A comparison of specific-conductance values for selected streams drain- ing unmined and mined areas is shown in figure 31. UPSTREAM DOWNSTREAM FROM MINING FROM MINING 4,500 13 4,000 -— " —‘ 12 n: 7’ Lu .— u.I 2 3,500 — ._ U l E [L 3.000 — “ (I) (I) E 2 U) 0 2,500 — — O 0‘ (£3 E a 2 8 ‘g 0 2,000 — — Lu (2) (lg 11 .— < l ,_ < 8 1 500 10 — o ' 7 8 Z O U 2 E 1,000 - _ ‘— (J E U) 6 5 500 — 9 _ I _L O EXPLANATION 5 SITE NUMBER MAXIMUM I MEDIAN MINIMUM SITE NUMBER IN FIGURE 29 SITE NAME AND DRAINAGE AREA 5 Unnamed tributary to Mulberry Creek near Amoret, Missouri (5.42 square miles) 6 Blue River near Staniey, Kansas (46.0 square miles) 7 Indian Creek at Overland Park, Kansas (26.6 square miles) Tomahawk Creek near Overland Park, Kansas (239 square miles) 9 North Sugar Creek tributary 3, below La Cvgne Lake, Kansas (1.96 square miles) 10 Mulberry Creek at Mulberry, Missouri (16.2 square miles) 11 Cox Creek tributary near Mulberry, Kansas (8.00 square miles) 12 Cox Creek 1 mile south of Arcadia, Kansas (30.0 square miles) 13 Deer Creek near Hallowell, Kansas (700 square miles) 14 East Cow Creek at Frontenac, Kansas (7.50 square miles) FIGURE 31.—Maximum. median, and minimum specific conductance at selected surface-water-quality sites ,in the Western region. Data are from Detroy, Skelton, and others (1988) and Marcher, Kenny, and others (1984). 58 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 ‘During high streamflow, there are increases in con- centrations of suspended sediment and some constitu- ents, such as iron and manganese, that are sorbed to suspended sediment. Mean annual suspended-sediment concentrations of streams in the region increase from less than 270 milligrams per liter in the southern and eastern parts of the region to more than 6,350 milli- grams per liter in the western part (Rainwater, 1962), generally because the eastern part has more forest and woodland and the western part has more cropland. GROUND—WATER NETWORK To provide ground-water data for coal-hydrology investigations in the 1970’s, approximately 100 ground- water-level and (or) ground-water-quality sites were established in the Western region. Nearly all these data- collection sites have been discontinued because the in- vestigations supporting them have been completed. During 1982, the US. Geological Survey collected ground-water—level data at 122 observation wells in the region. Miscellaneous groundwater-quality data also were collected. GROUND—WATER OCCURRENCE Ground-water resources are limited in most of the region. Although ground water occurs throughout the area and there is ample precipitation to provide recharge, well yields are controlled by hydrogeologic properties of aquifers. The following discussion is based on State summaries presented in US Geological Survey (1985). Bedrock of Pennsylvanian age, primari- ly fine-grained shale with some limestone, is present at or near the surface throughout most of the region. Wells in rocks of Pennsylvanian age generally yield less than 50 gallons per minute; wells in these rocks are shallow and frequently go dry during droughts. Cretaceous sandstones in the southwestern and northwestern parts of the region are productive aquifers and commonly yield 50 to 500 gallons per minute. Carbonate-rock aquifers of Cambrian, Ordovician, and Mississippian age, primarily limestone and dolomite with some sand- stone, occur along the eastern edge of the region and commonly yield 50 to 100 gallons per minute. Alluvial aquifers of clay, silt, sand, and gravel are present in valleys of the Missouri, Kansas, East and West Nishna- botna, Platte, Grand, Thompson, Chariton, Des Moines, Osage, Canadian, and Arkansas Rivers, and yields com- monly exceed 1,000 gallons per minute. Glacial-drift aquifers and buried alluvial-valley aquifers occur primarily north of the Kansas and Missouri Rivers. Water yields from these aquifers may exceed 500 gallons per minute in some parts of the region. Alluvial and glacial-drift aquifers are recharged easily by precip- itation and surface drainage. GROUND—WATER QUALITY Water from rocks of Pennsylvanian age, which occur at the surface in most of the region, can have large con- centrations of sodium, chloride, and sulfate if the water is from a shale formation. Calcium and bicarbonate generally are the major dissolved constituents in water from limestone formations. In or near outcrop areas where bedrock aquifers receive recharge, concentrations of dissolved solids usually are less than 1,000 milligrams per liter. As the strata dip toward the west, the ground water becomes more mineralized and generally is not used. The Cretaceous sandstone aquifers yield calcium magnesium sulfate type water in the northwestern part of the region and sodium or calcium bicarbonate type water in the extreme southwestern part of the region. The limestone and dolomite aquifers of Cambrian, Or- dovician, and Mississippian age generally yield calcium bicarbonate type water along the eastern edge of the region. Water from the alluvial aquifers generally is a calcium magnesium bicarbonate type, and concentrations of dissolved solids generally are less than 500 milligrams per liter. Glacial-drift aquifers and buried alluvial-valley aquifers that occur north of the Kansas and Missouri Rivers yield calcium bicarbonate type water. Concen- trations of dissolved solids generally are less than 500 milligrams per liter in water from shallow wells, but con- centrations usually increase with depth. Alluvial and glacial-drift aquifers are very susceptible to contamina- tion from surface sources. COAL—HYDROLOGY STUDIES Reports of US. Geological Survey coal-hydrology studies are available for Iowa, Kansas, Missouri, and Oklahoma (fig. 32 and table 1); some results are de- scribed briefly here. These reports contain hydrologic information necessary to determine the effects of coal mining on water resources. Several of the reports pres- ent hydrologic data collected during the investigation; the other reports are interpretive and either provide areal descriptions of the physical setting, coal-mining activities, cultural features, hydrology, and hydrologic effects of coal mining and (or) present methods used to identify, assess, or predict various hydrologic effects of coal mining. The water quality of coal-mining areas in south- central Iowa was investigated in cooperation with the Iowa Department of Environmental Quality. The report INTERIOR PROVINCE—WESTERN REGION 59 94° 1° - ~ EXPLANATION COAL REGION “L "3“ LOCATION OF COAL-HYDROLOGY \ STUDIES ‘ V 1 Estimating stream-aquifer interactions in coal areas of eastern Kansas (Bevans, 1986) 2 Baseline water quality of Iowa’s coal region (Slack, 1979) 3 Statistical summaries of water-quality data for streams draining coal-mined areas, southeastern Kansas (Bevans and Diaz, 1980) 3 A procedure for predicting concentrations of dissolved solids and sulfate Ions in streams draining areas strip mined for coal (Bevans, 1980) 4 Applications of remote-sensing techniques to hydrologic studies in selected coal- mined areas of southeastern Kansas (Kenny and McCauley, 1983) 5 Quality-of-water data and statistical summary for selected coal-mined strip pits in Crawford and Cherokee Counties, southeastern Kansas (Pope and Diaz, 1982) 6 Physical and hydrologic environments of the Mulberry coal reserves in eastern Kansas (Kenny and others, 1982) 6 Hydrologic responses of streams to mining of the Mulberry coal reserves in eastern Kansas (Bevans, 1984) 7 Physical environment and hydrologic characteristics of coal-mining areas in Missouri (Vaill and Barks, 1981) 8 Preliminary appraisal of the hydrology of the Rock Island area, Le Flore County, Oklahoma (Marcher and others, 1983b) 9 Preliminary appraisal of the hydrology of the Stigler area, Haskell County, Oklahoma (Marcher and others, 1982) C§OPER / 10 Preliminary appraisal of the hydrology of wow " the Red Oak area, Latimer County, \ Oklahoma (Marcher and others, 1983c) 11 Preliminary appraisal of the hydrology of the Blocker area, Pittsburg County, Oklahoma (Marcher and others, 1983a) 12 Hydrology of an abandoned coal-mining area near McCurtain, Haskell County, Oklahoma (Slack, 1983) .3 FAULKNER ‘3:ch gem.» mm 34 Base from U. S. Geological 0 50 100 MILES Survey 12,500,000 United States base map 0 50 100 KILOMETERS FIGURE 32.—Location of areas of coal-hydrology studies done in the Western region. 60 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 of this investigation provides surface-water-quality data representing high-, average, and low-flow conditions for White Breast, English, and Cedar Creeks (Slack, 1979). The U. S. Geological Survey made a coal-hydrology in- vestigation of an area underlain by strippable reserves of coal in east-central Kansas. The results of this in- vestigation were published in two reports. The first report describes the physical and hydrologic environ- ments of the study area based on available data and reports (Kenny and others, 1982). The second report evaluates the effects of abandoned coal mines on areal surface-water quality and the effects of an active strip mine on streamflow and water-quality characteristics of small streams (Bevans, 1984). Results of this investi- gation indicate that most streams contained dissolved- solids concentrations less than 500 milligrams per liter. However, streams draining abandoned coal-mined areas contained larger concentrations of sulfate than streams draining unmined areas; and a small stream draining an active strip mine had less high flow and more low flow, a 41 percent larger load of dissolved solids, a 244 percent larger load of sulfate, and a 25 percent larger load of suspended sediment than a small stream drain- ing a nearby unmined control basin. Stream-aquifer interactions in coal areas of eastern Kansas were investigated by Bevans (1986). Basin con- stants (equivalent to the aquifer transmissivity divided by the product of the aquifer storage coefficient and the squared average distance from the stream to the ground-water divide) were developed by analyzing the slopes of base-flow recession curves. The basin con- stants then were used in conjunction with discharge records to estimate ground-water recharge, storage, and discharge that resulted from one or more periods of recharge and to estimate the rate of evapotranspiration. A regional regression equation was developed for estimating basin constants for ungaged basins. Channel-geometry techniques for estimating stream- flow characteristics were applied to surface-mined areas in a report by Osterkamp and Hedman (1979). Simple- and multiple-power-function equations relating channel configuration and channel-material data to mean and peak stream discharge were developed for perennial streams in the Central and Western United States and for ephemeral streams in the Western United States. The equations for perennial streams also probably can be applied in the Eastern United States. The Kansas Department of Health and Environment and the US. Geological Survey cooperated in a coal- hydrology investigation in Crawford and Cherokee Counties, Kansas. Reports published from this investi- gation present: (1) Statistical summaries of water- quality data and regression equations that relate stream discharge and specific conductance to concentrations of selected chemical constituents (Bevans and Diaz, 1980); and (2) regression equations that can be used to predict instream concentrations of dissolved solids and sulfate from the percentage of drainage basin that was strip mined (Bevans, 1980). An investigation of the Cherry Creek basin in Chero- kee County, Kansas, by the US. Geological Survey in cooperation with the Kansas Department of Health and Environment and the Kansas Geological Survey, ap- plied and evaluated remote-sensing techniques for coal- hydrology studies. Color and color-infrared aerial photography were used with simultaneously collected water-quality samples to identify cause-and-effect rela- tions between land, water, and vegetation disturbances. Types and extent of vegetation on abandoned and reclaimed mine lands, drainage patterns, point sources of acid mine drainage, and recharge areas of under- ground mines were determined from aerial photography (Kenny and McCauley, 1983). The Office of Surface Mining Reclamation and En- forcement cooperated with the U. S. Geological Survey in an investigation of water-quality characteristics of coal-mine strip pits in Crawford and Cherokee Counties, Kansas. The report of this investigation contains a statistical summary of water-quality data collected from the strip pits and regression equations that relate specific conductance, concentration of dissolved solids, and acidity to concentrations of selected chemical con- stituents (Pope and Diaz, 1982). The US. Geological Survey investigated the coal- mining areas in Missouri. The report of this investiga- tion describes the physical environment, coal-mining practices, general hydrology, and the 1980 hydrologic- data base for the north-central and western coal-mining regions. Water in streams draining unmined areas generally had pH values near neutral (7.0) and concen- trations of dissolved solids less than 400 milligrams per liter. However, water from some streams affected by coal-mine drainage had pH values less than 4.0 and con- centrations of dissolved solids greater than 1,000 milli- grams per liter (Vaill and Barks, 1981). The hydrology of an abandoned coal-mining area near McCurtain in‘Haskell County, Oklahoma, was investi- gated by the U.S. Geological Survey. The report presents hydrologic data collected during the investiga- tion, an evaluation of the water resources of the area, and an appraisal of the probable effects of reclamation (Slack, 1983). Analysis of water-quality data from a stream draining abandoned and reclaimed surface mines indicated that concentrations of dissolved constituents, principally sulfate but also calcium, magnesium, sodium, chloride, and alkalinity, were larger in reaches of the stream that drained abandoned mines. Also, in- stream concentrations of these constituents increased INTERIOR PROVINCE-WESTERN REGION 61 as the area of abandoned mines drained by the stream increased. The US. Geological Survey, in cooperation with the US. Bureau of Land Management, investigated water- resources effects of coal mining on Federal coal lands in Oklahoma. Reports containing hydrologic data, descriptions of the physical settings, preliminary ap- praisals of hydrology, and probable hydrologic effects of coal mining have been published for the Blocker area in Pittsburg County (Marcher and others, 1983a), the Stigler area in Haskell County (Marcher and others, 1982), the Rock Island area in Le Flore County (Marcher and others, 1983b), and the Red Oak area in Latimer County (Marcher and others, 1983c). Results of these investigations indicate that streams draining unmined parts of these areas had mean concentrations of dis- solved solids ranging from 50 milligrams per liter in the Red Oak area to 322 milligrams per liter in the Rock Island area. Streams draining mined parts of the areas had mean concentrations of dissolved solids ranging from 132 milligrams per liter in the Red Oak area to 1,766 milligrams per liter in the Stigler area. Concen- trations of iron and manganese were larger downstream from areas of old and recent strip mining in the Red Oak area than in unmined parts of the area. Information about other coal-hydrology investigations in the Western region that are not yet completed, or for which reports are not yet published, can be obtained from US. Geological Survey district offices in this region. HYDROLOGIC ISSUES RELATED TO COAL MINING Surface and underground mining of coal disturbs the hydrologic environment, often affecting the quantity and qualityof surface and ground water in the region. The cleariifg ofland prior to surface mining causes in-i’ creased runoff and erosion, which increases concentra- tions and loads of suspended sediment in receiving streams. Sediment ponds constructed to intercept run- off from active surface mines regulate discharge by decreasing flows. However, concentrations and loads of suspended sediment in streams draining active mines are larger than normal because colloidal clay particles often do not settle out if detention times in the sediment ponds are relatively short and flocculating agents are not used (Bevans, 1984). As mining proceeds, excava- tion exposes unweathered bedrock to physical and chemical weathering. In most of the region, excess acidity generated from weathering of iron sulfide minerals increases the weathering of calcite and dolomite, leaving increased concentrations of dissolved solids (sulfate, bicarbonate, calcium, and magnesium) in solution. However, in shaft- mined areas where carbonate rocks are not exposed dur- ing mining, excess acidity decreases the pH, thereby releasing iron, manganese, aluminum, lead, zinc, and other metals into solution (Kenny and others, 1982). Me- dian values and ranges of specific conductance often are greater in streams draining mined basins. A comparison of specific-conductance values for selected streams draining unmined and mined basins is shown in figure 31. Specific conductance and concentrations of dis- solved solids increase as the percentage of drainage basin that is mined increases (Bevans, 1980). If an aquifer is disturbed during mining, water will be pumped from the mine, and local water levels may decline. After mining ceases, the ridge-and-furrow topography and strip pits of abandoned surface mines and the mine shafts and sinkholes above collapsed tun- nels of abandoned underground mines intercept runoff and increase recharge, thereby decreasing high flow and increasing base flow in streams that drain the mines. Reclaimed surface mines, which have been graded to ap- proximate the original topography and have been planted in grass, probably will not affect ground-water levels or flow in adjacent streams within the region. Small streams draining mined areas in the region often are contaminated during low-flow conditions by base flow that has large concentrations of sulfate and other dissolved constituents or by acid mine drainage. During runoff, the contamination is diluted (Bevans, 1980). Active and abandoned surface mines will con- tribute large quantities of sediment to receiving streams until vegetation is reestablished. Reclaimed surface mines yield quantities of sediment comparable to un- disturbed areas. Water-quality degradation in streams that results from coal mining usually is only a local issue because runoff and base flow from unmined parts of a drainage basin dilute the coal-mine drainage. 62 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974—84 NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES—FORT UNION REGION By ORLO A. CROSBY and CLARENCE A. ARMSTRONG The Fort Union region (fig. 33), an area of about 45,000 square miles in east-central Montana (about 14,000 square miles), western North Dakota (about 30,000 square miles), and northwestern South Dakota (about 1,000 square miles), includes coal areas 45—47 (table 1). About 41,000 square miles of the region is in the Missouri River drainage basin and about 4,000 square miles is in the basins of the Des Lacs and Souris 50° ‘07)» 106° 105° 104° 103° 102° 101\0 < \1000 99° ' « l\ \\KL \L .a ‘ v i ,\,W .2“ {31’ l \ . l ° TN “f"§‘/\éfi\w ’ L’x Ass'mwel» \ " c ‘ ”M \\ Smucfl'oni \_\\ ., "'33" \ ‘\ \ mm“ a ‘\ ,IIWKQANADA \‘\ (Es: ‘¢\. V 440 (l \ ._ , Base from U. 8. Geological 0 50 100 MILES Survey 12,500,000 United States base map 0 50 100 KILOMETERS EXPLANATION COAL REGION 45 COAL-AREA BOUNDARY—Number refers to local areas (Cochran and others, 1983) .2 SAMPLING SITE AND NUMBER—See figures 34 and 35 for site descriptions FIGURE 33,—The Fort Union region, coal areas 45—47. and sampling sites. NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES—FORT UNION REGION 63 Rivers in northwestern North Dakota. Although gen- erally included in the basin of the Missouri River, the Coteau du Missouri, a part of the Great Plains prov- ince in northwestern North Dakota (pl. 1), contributes little or no surface flow to the river basin, although it contains many prairie potholes, sloughs, and small lakes. The region is in the Great Plains and Central Lowland physiographic provinces (pl. 1). The Coteau du Missouri, a part of the Great Plains physiographic province, is a series of elongated hills, ridges, sloughs, and prairie potholes that is bounded on the north by a scarp that grades into the Central Lowland physiographic prov- ince. The Central Lowland physiographic province in the region is characterized by a northeastward-sloping plain with a few low, rounded hills and shallow depressions. Mean annual precipitation ranges from 12 inches in the western part of the region to about 18 inches in the east-central part. About 65 percent of the annual precipitation occurs during May through August. Most of the precipitation that falls during June, July, and August is the result of thunderstorm activity (U.S. Weather Bureau, 1962). Mean annual snowfall is about 36 inches. Mean annual lake evaporation ranges from about 34 inches in the northeastern part of the region . to about 45 inches in the southwestern part (National Oceanic and Atmospheric Administration, 1982). The entire region is located Within the ovate Williston basin. The axis of the basin trends north-northwest, and the deepest part is in eastern McKenzie County, North Dakota. Maximum thickness of the sedimen- tary rocks of Cretaceous and Tertiary age is about 15,000 feet. The oldest formation exposed in the area is the Cre- taceous Bearpaw Shale or the equivalent Pierre Shale. The overlying marine Fox Hills Sandstone and the con- tinental Hell Creek Formation that overlies the Fox Hills Sandstone also crop out in scattered locations in Montana. The Fox Hills Sandstone generally is con- sidered the deepest formation that could yield fresh- water; it consists of sandstone and interbedded siltstone, shale, and sandy shale. The Hell Creek For- mation is composed of interbedded sandstone, clay- stone, and lignitic shale. Glacial drift, primarily till and glaciofluvial deposits, covers the northeastern part of the region. Glacial till, where present, ranges in thickness from 0 to as much as 600 feet. Glaciofluvial deposits are variable in thick- ness but can be as much as 400 feet thick. Alluvium con- sisting of clay, silt, sand, and gravel is as much as 40 feet thick and occurs in channels and flood plains of present-day streams. At present (1985), land use in the area is about 52 percent rangeland, 38 percent cropland, 1 percent woodland, and 9 percent other uses (primarily urban and developed areas). About 86 percent of the land surface is privately owned; about 9 percent is federally owned and administered primarily by the US. Forest Service as National Grasslands; and about 5 percent is owned by the State. About 35 to 40 percent of the coal in the region is federally owned, about5 percent is owned by the States, and the rest is privately owned. Coal ownership is shown in detail on maps available from the US. Bureau of Land Management (1974, 1975, and 1977). The population of the region is about 280,000 (U.S. Bureau of Census, 1981). Cities with populations of more than 5,000 are Bismarck, Dickinson, Mandan, Minot, and Williston, North Dakota, and Glendive and Sidney, Montana. Thermoelectric-power generation and irrigation ac- count for about 95 percent of the water used in the region. About 97 percent of the total water used is ob- tained from surface-water sources and about 3 percent is obtained from ground-water sources. COAL RESOURCES Coal in the study area is classified as lignite (fig. 3), and much of it is near the land surface where it can be mined economically using surface-mining techniques. Estimates of coal reserves in the region vary greatly. Strippable lignite reserves in Montana were estimated by Slagle and others (1984) to be about 7.3 billion tons, about 20 percent of previous estimates made by the US. National Research Council (1981a). Similarly, strip- pable reserves in North Dakota (fig. 33) were estimated to be about 4.2 billion tons by Pollard and others (1972), about half the estimate of the US. National Research Council (1981a). Pollard and others (1972) did not in- clude in their estimate any coal beds less than 5 feet thick, even though they might overlie the principal bed to be mined. In 8 of the 16 coal deposits evaluated in North Dakota, they also did not include any coal that might occur beneath more than 50 feet of overburden. Current limits on overburden stripping in the State are about 150 feet. In attempting to correct the estimates of Pollard and others (197 2) for the deficiencies noted, the North Dakota Geological Survey during 1981 esti- mated reserves at about 15 billion tons of strippable coal in the State (North Dakota Geological Survey, 1981), about twice the US. National Research Council (1981a) estimate. No strippable coal reserves occur in South Dakota. Coal mining in the region began during the late 1800’s. Early production was from small underground mines that provided fuel for local consumption. When 64 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 equipment became available, strip mining began and increased until about 1940, when 2,230,000 tons were produced in the region—800,000 tons from underground mines. Total production of coal generally increased at an irregular rate to about 8.5 million tons from July 1975 through June 1976. Production from underground mines decreased and had ceased by 1966. From July 1979 through June 1980, about 17 million tons of coal was produced from 13 coal mines in North Dakota and 1 coal mine in Montana. A history of coal mining in North Dakota has been chronicled by Oihus (1983). Most of the coal presently (1985) is being used for the production of electricity, but some is used for the manu- facture of synthetic natural gas, charcoal briquettes, drilling-mud additives, and organic solvents. The larg- est production of coal is in Mercer, Oliver, and western McLean Counties, North Dakota, where about 10 mil- lion tons of coal is mined annually to supply nearby powerplants. HYDROLOGY Many data-collection changes occurred in the region as part of the cooperative U.S. Geological Survey and U.S. Bureau of Land Management coal-hydrology pro- gram. New stream-discharge and surface-water-quality sites were added, and some historical sites were expanded. Because of funding decisions, most of the stream- discharge and surface-water-quality sites have been discontinued. The data that were collected constitute a substantial data base from which changes due to pres- ent or future coal mining may be substantiated; however, detailed studies probably will be needed to determine the rate and extent of the changes that may occur. During 1983, the U.S. Geological Survey in cooperation with the North Dakota Public Service Com- mission established two programs in which eight of the previously established stream-discharge and surface- water-quality sites are monitored at 6-week intervals. Monitoring at these sites is expected to continue until mining at upstream sites ends, thus providing a nearly complete record of the effects of mining activities in the basins being monitored. Water-quality sampling began in the coal areas as early as 1945, but it was not until the mid-1970’s that the need for a much larger water-quality data base was recognized. At about the same time the coal-area studies were being initiated, an expanded suite of water-quality determinations also was made at most coal-area sites. The expanded suite of constituents collected included onsite determinations of specific conductance, pH, alkalinity, dissolved oxygen, and water temperature and laboratory determinations of common ions, nutrients, and sediment. In addition, trace elements, pesticides, radioactive constituents, microbiological and biological contents, and some organic compounds were deter- mined on an irregular basis. SURFACE—WATER NETWORK The surface-water data-collection network in the region started with the establishment of a few continuous-record stream-discharge sites during 1903. The number of continuous-record sites has fluctuated but generally increased with time to the present (1985) number. The first sites were established as part of a water-accounting system and were located on the Mis- souri and Souris River main stems and on the major tributaries to the Missouri River. Additional sites soon were established on these tributaries and other streams as the demands for water increased. The early estab- lished sites provide a long-term record from which surface-water statistical information can be obtained. The long-term records also can be used to extend incomplete or short-term records using correlation techniques. However, regulation, storage, or diversion on many streams has rendered invalid the usage of parts of long-term records to define surface-water characteristics. Before 1970, most of the continuous-record stream- discharge sites were established on the larger perennial streams to meet some specific water-management need. During the 1970’s, the continuous-record stream- discharge measurement site network was expanded to obtain data to evaluate the hydrology of the region. Generally, sites established during the 1970’s were established in response to the U.S. Bureau of Land Management’s energy-related responsibilities. As a result, 10 continuous-record stream-discharge sites and 15 crest-stage stream-discharge sites were added to those in Montana. Forty-five continuous-record stream- discharge sites were added to those previously estab- lished in North Dakota. One continuous-record stream- discharge site was added in South Dakota. Miscellaneous stream-discharge measurements (generally low flow) were made at 104 sites in the region. Many of the measurements were made as part of 13 small-area coal studies. These stream-discharge measurements were used to determine the ground-water contribution to streamflow. Interest in low-flow characteristics has resulted in periodic measurements of low-flow discharge on many large and small streams for 1 or more years. Because of the ephemeral nature of the streams and the varied sources of low-flow discharge, correlation of low flow between sites is poor, and onsite measurements are the only dependable source of information. NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES—FORT UNION REGION 65 A network of crest-stage stream-discharge sites was established along a number of streams during 1954—55. Most of these sites in North Dakota were discontinued during 1973. Data from these sites, in addition to data from the continuous-record stream-discharge sites, were used to develop flood-frequency and magnitude rela- tions. Operation of the crest-stage stream-discharge sites also resulted in the collection of a large quantity of periodic stream-discharge information. Many new methods, such as those described by Crosby (1975), were developed to estimate surface-water characteristics from these limited data. Most of the data collected at continuous-record and crest-stage stream-discharge sites are available in computer-usable form. The data collected since 1965 also are available in annually published US. Geological ' Survey reports, “Water Resources Data for Montana,” “Water Resources Data for North Dakota,” and “Water Resources Data for South Dakota.” Water-quality sampling by the US Geological Survey began as early as 1945. The program to sample the major streams for water-quality properties and com- mon ions was continued for 5 or 6 years at most sites. A major effort was made during the late 1970’s and early 1980’s to acquire a more complete water-quality data base. Water-quality data have been collected at 135 sites in the study area. The data collected at these sites are variable with regard to time and duration of collec- tion. Also, there are data for additional miscellaneous, one-time-sample, water-quality sites in the US. Geo- logical Survey files. The data are available in pub- lished form in annual reports of the US. Geological Survey, in the US. Environmental Protection Agency’s STORET computer files, and, since about 1950, in the US. Geological Survey’s WATSTORE computer files. Water-quality determinations were made at 68 sites in the region as part of the US. Geological Survey and US. Bureau of Land Management coal-hydrology pro- gram. A complete list of the constituents that were determined for Montana can be found in Slagle and others (1984) and for North Dakota in Croft and Crosby (1987) and Crosby and Klausing (1984). Statistical evaluation of water-quality data during the program was completed by Haffield (1981). A common practice for several years has been for US. Geological Survey investigators to obtain water- temperature and specific-conductance measurements whenever a discharge measurement is made. These data also are stored in the computer files. Prior to 1976, most sediment data were collected on the main-stem Missouri River or near the mouths of the major tributaries. Since 1976, mainly in response to in- creased energy development, sediment data also have been collected at many other sites. The data generally are determinations of the sediment concentrations at the time of discharge measurements. In addition, some particle-size analyses were made. SURFACE—WATER CHARACTERISTICS Stream discharge varies greatly within the area. Flows in all unregulated streams have large seasonal variations, and the largest flows occur during spring as a result of snowmelt and rainfall. Stream discharge in the Missouri River is not as variable because of the effects of upstream storage reservoirs. Further infor- mation can be obtained by referring to Slagle and others (1984), Croft and Crosby (1987 ), and Crosby and Klaus- ing (1984), and (or) by contacting the US. Geological Survey offices in Helena, Montana, and Bismarck, North Dakota. Daily-flow hydrographs (fig. 34) indicate the seasonal variation in stream discharge during 1980 for the Yellowstone River (site 1) and for the Redwater River (site 2). The hydrographs indicate the effects of snowmelt and rainfall on the discharge in a large perennial stream (Yellowstone River) and atypical prairie stream (Redwater River). Increased discharge in the Yellowstone River during May and June and in the Redwater River during March and April is the result of snowmelt. A method and equations for estimating average annual flows using basin characteristics have been 1°°'°°° I I I I I I I I I I I 10,000 W SITE 1 1,000 __ Yellowstone River _ near Sidney, Montana (Drainage area, 69,103 S'TE 2 _ square miles) Redwater River _ at Circle, Montana (Drainage area, 547 square miles) \ 100 — STREAM DISCHARGE, IN CUBIC FEET PER SECOND | l l | l l l l | 0.01 OCT NOV DEC JAN FEB MAR APR MAYJUNEJULYAUGSEPT WATER YEAR FIGURE 34.—Typica.l hydrographs showing variation of stream discharge for selected streams during the 1980 water year in the Fort Union region (modified from Slagle and others, 1984; see figure 33 for location of sites). 66 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 developed for ungaged sites in central and eastern Mon- tana (Omang and Parrett, 1984). Data for calculating mean available flows for all stream-discharge sites in coal areas 46 and 47 are available in reports by Croft and Crosby (1987) and Crosby and Klausing (1984). Flood estimates can be made from available data for gaged and ungaged streams in the region. The most reliable estimators of future floods generally are the fre- quency analyses of stream-discharge records. Regres- sion equations for estimating floods on ungaged streams draining less than 100 square miles are in a report by Crosby (1975), and equations for ungaged streams draining 100 square miles or more are in a report by Patterson (1966). SURFACE—WATER QUALITY Dissolved-solids concentrations vary greatly in water from all streams except the Missouri River. The most common dominant cations are calcium, magnesium, and sodium, and the dominant anions are bicarbonate, sulfate, and chloride. Large dissolved-solids concentra- tions can be objectionable because of possible physiological effects, mineral taste, or economic con- straints associated with their removal. Numerous standards have been established for dissolved solids. Generally, it is desirable to have con- centrations of dissolved solids less than 500 milligrams per liter for public water supplies. During snowmelt runoff or high runoff from thunderstorms, most streams in the region have dissolved-solids concentrations less than 500 milligrams per liter. During periods of low flow, dissolved-solids concentrations in water from many of the streams will exceed 1,300 milligrams per liter, an approximate concentration at which the water will acquire a mineralized taste. Water containing dissolved-solids concentrations in excess of 2,500 milligrams per liter has only limited use; however, livestock will tolerate larger concentrations of dissolved solids under most circumstances (National Academy of Sciences, National Academy of Engineer- ing, 1973). Dissolved-solids concentrations of water in many streams will exceed 2,500 milligrams per liter dur- ing extremely low flows when ground-water discharge is the primary source of water. Medians and ranges of specific conductance at selected surface-water—quality sites are shown in figure 35. A summary of medians and ranges of specific con- ductance at 26 sites for different-sized drainage areas upstream and downstream from mining is shown in figure 36. The median specific conductance of streams in small basins downstream from mining is greater than that of streams in similar unmined basins. However, the median specific conductance of streams downstream UPSTREAM DOWNSTREAM I 9,000 FROM MINING FROM MIN N6 8 I __ _ u.I 8,000 — I.— u.) E '2 Lu 7,000 — — U I Lu 0. (é) m 6,000 — - w 2 2 (I) E d 8 0 5,000 — 4 — a a —— _ Lu __ E a: .7 Z 8 4000 F _ _. D ' Lu 3 a I— 5 I<5 < 3000 — 3 — U I D D Z O 0 2,000 -—— — 9 E U E.‘ (I) 1,000 — J O EXPLANATION 4 SITE NUMBER MAXIMUM I MEDIAN MINIMUM SITE NUMBER IN FIGURE 33 SITE NAME AND DRAINAGE AREA 3 Elm Creek near Golden Valley, North Dakota (82.0 square miles) 4 North Creek near South Heart, North Dakota (40.8 square miles) Coal Creek near Stanton, North Dakota (15.8 square miles) Spring Creek below Lake lie at Dunn Center, North Dakota (116 square miles) 7 Norwegian Creek near Belfield, North Dakota (39.8 square miles) 8 Buffalo Creek tributary near Gascoyne, North Dakota (15.7 square miles) FIGURE 35.—Maximum, median. and minimum specific conductance at selected surface-water-quality sites in the Fort Union region. Data are from Crosby and Klausing (1984). from mined and unmined basins greater than 50 square miles in size is essentially the same. Unpolluted streams draining undisturbed basins in the region generally will have alkaline water. Values of pH in the streams commonly range from about 6.5 to NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES—FORT UNION REGION 67 UPSTREAM FROM MINING DOWNSTREAM FROM MINING 9,000 _6_IAI 7,500 — — 6,000 _s_(A) _4_(B) 4,500 — 3 (cr _3_(C) 4 (5) 3,000 -— — 1,500 '— — SPECIFIC CONDUCTANCE, IN MICROSIEMENS PER CENTIMETER AT 25 DEGREES CELSIUS EXPLANATION 6 NUMBER OF SITES DRAINAGE AREA, IN MAXIMUM SQUARE MILES I MEDIAN (A) 10-50 MINIMUM (BI 50-90 (C) Greater than 90 FIGURE 36.—Summary of medians and ranges of specific conductance at 26 sites for different-sized drainage areas in the Fort Union region. 8.5. Even though a substantial degree of basin disturb- ance occurs because of mining, the surface waters generally remain alkaline. The surface waters are buf- fered by carbonate minerals in the prairie soils, and stream pH decreases to less than 7.0 only during major precipitation or when the site is located immediately ad- j acent to acid sources such as lignite outcrops. Oxida- tion of sulfur species in the coal-mining areas generally will cause a decrease in pH. However, the prevalence of soils that have moderate buffering capacity makes acid mine drainage unlikely. Sulfate concentrations in streams generally are discharge dependent; concentrations in most streams are less than 250 milligrams per liter during runoff from snowmelt or from intense thunderstorms, and these con- centrations are more than 500 milligrams per liter at low flows when streamflow is dominated by ground- water discharge. The variation of sulfate concentrations with discharge makes detection of changes due to min- ing problematical without a record of concentration variations throughout a varied range of discharges. The network of surface-water sites in the region provides a substantial data base for providing this background information. The coal-bearing formations in the area generally con- tain enough carbonate minerals to rapidly neutralize any acids that may form. Hence, trace elements general- ly are not dissolved in quantities large enough to exceed drinking-water standards (US. Environmental Protec- tion Agency, 1986a, 1986c). Iron, which is not known to be toxic, is one possible exception. Many water samples contained dissolved-iron concentrations larger than 300 micrograms per liter, which is the recom- mended limit. Most of this excessive iron, however, is believed to be in a colloidal form that may have passed through the 0.45-micrometer filter that was used to process water samples prior to analysis of dissolved chemical species. Locally, boron concentrations may be excessive for some sensitive plants; but, generally, con- centrations of boron are less than that which might damage most crops that are grown in the region. At some sites, standards for other trace-element concen- trations have been exceeded for short intervals during base flow, but no deleterious effects from any of the above excessive trace-element concentrations have been reported. A poor correlation exists between the sediment con- centration and measured discharge for small basins in the area. The poor correlation is because of the great variability of factors such as soil types, soil conditions (frozen,.thawed, degree of saturation, and tillage), land use, precipitation intensity, rapidity of snowmelt, and the time of sampling in relation to hydrologic events. Almost all the avaflable water-quality data are from areas of undeveloped energy resources. It would be extremely difficult, if not impossible, to extrapolate these data to estimate the effects of mining or other land-use changes. It is unlikely that cumulative effects of energy development will be detectable at downstream sites on major streams for many years. GROUND—WATER NETWORK The ground-water network provides general water- level and groundwater-quality data for most of the 68 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 region. The network of ground-water observation wells being monitored is reviewed and updated periodically. The network as of September 30, 1981, comprised 395 wells. The frequency of measurement can vary from one annual measurement to a continuous record. Lithologic logs are available for all observation wells. Various geophysical logs also commonly are available. At least one chemical analysis of water is available for most wells. Chemical quality routinely is monitored at several of the wells. Many other wells have been constructed by the US. Geological Survey and various cooperators but are not part of the network. Information on these observation and test wells is available from the US. Geological Survey computer storage files and in pub- lished US. Geological Survey annual State reports. Some data concerning several thousand private wells scattered throughout the coal areas also are available. GROUND—WATER OCCURRENCE Bedrock aquifers consisting of sandstone, lignite, and clinker underlie the entire region and are the only sources of ground water throughout most of the region. Yields of wells developed in these aquifers vary depend- ing on the thickness of the aquifer and the rate at which water moves through the aquifer. Wells scattered throughout the region produce water from aquifers in alluvial, terrace, or glacial sand and gravel. Most stream valleys in the region contain some alluvium, but generally only the larger valleys contain sufficient alluvium to yield water. The major glacial-drift and alluvial aquifers were deposited in river valleys formed by meltwater from glaciers. These valleys range from 0.25 to 2 miles inwidth and may be as much as 300 feet deep. In addition, a few preglacial river valleys contain sand and gravel aquifers. The preglacial drainages range from 0.5 to 5 miles in width and may be as much as 400 feet deep. Ground water also is obtainable from isolated sand and gravel lenses in the glacial till. These lenses seem to be distributed randomly both laterally and vertically. Water levels in shallow aquifers less than 200 feet deep indicate that the ground-water flow patterns follow the land surface, and flows are from the topographic high areas toward the nearby drainages. Water levels in the deeper aquifers indicate that the water is under artesian pressure, and regional flow generally is from the southwest to the northeast. Because of artesian pressure, wells flow when drilled in or near the bottom of the valleys in the Missouri, Yellowstone, or Little Missouri River basins. Flows from wells in these valleys have decreased the artesian pressure beneath these river valleys so that part of the regional flow has been diverted toward these rivers. GROUND-WATER QUALITY Quality of water from wells developed in alluvial and glacial-drift aquifers generally is suitable for most uses. Dissolved-solids concentrations range from 159 to 1,000 milligrams per liter. Calcium generally is the principal cation present, but magnesium or sodium may be significant locally. Bicarbonate is usually the dominant anion present, but sulfate may dominate locally when sodium is abundant. The quality of water from shallow bedrock aquifers is quite variable. Included in this group are most lignite and sandstone aquifers that may be affected by mining. Dissolved-solids concentrations, which range from 286 to as much as 9,700 milligrams per liter, commonly in- crease with depth. Sodium and calcium generally are the principal cations present, but magnesium sometimes is abundant in waters dominated by calcium. Sodium dominance over calcium generally increases with depth because of cation exchange of divalent cations for sodium on sodic smectites in the siltstones and clay- stones of the 'llertiary and Cretaceous formations. Bicar- bonate and sulfate are the dominant anions present. In parts of aquifers affected by combined pyrite oxidation, gypsum dissolution, and cation exchange, local sulfate concentrations may be as much as 6,500 milligrams per liter. Water from aquifers at depths greater than about 200 feet in the Tertiary and Cretaceous formations generally has smaller dissolved-solids concentrations than water from shallower bedrock aquifers. Furthermore, dissolved-solids concentrations generally are greater in the Tertiary aquifers in the Fort Union Formation than in the Cretaceous aquifers, but chemical compositions are similar. Sodium and bicarbonate ions dominate. Calcium and magnesium concentrations tend to be small because of cation exchange for sodium in the overlying siltstones and claystones. Sulfate usually is a minor constituent, and concentrations generally are less than 100 milligrams per liter because of sulfate reduction as the water moves downward. Dissolved- solids concentrations in the aquifers range from 610 to 10,200 milligrams per liter and increase from southwest to northeast. Wells in this group supply most of the ground water used in the region. Although the water generally is soft (less than 60 milligrams per liter calcium carbonate), large dissolved-solids concentra- tions in addition to large sodium-adsorption ratios make the water unsuitable for irrigation except locally near the outcrop areas. Water from bedrock aquifers general- ly is usable for domestic, livestock, and some industrial NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES—FORT UNION REGION 69 use, but the large concentrations of dissolved solids commonly exceed the criteria and regulations estab- lished by the U.S. Environmental Protection Agency (1986b, 1986c). Large fluoride concentrations have been detected in water from some wells, and these concen- trations may restrict water used for domestic supply. Additionally, some water is unsuitable for use by people on sodium-restricted diets because of large sodium concentrations. COAL-HYDROLOGY STUDIES Many investigations concerning coal or water resources in the Fort Union region have been completed. Most of the studies completed prior to the mid-1970’s were related to coal thickness, ash content, or some other characteristic of the coal. Hydrologic studies generally were not oriented toward coal and water rela- tions, but much valuable data concerning well yields and the quality of water in coal beds were collected. Many of these older studies have been referenced by Slagle and others (1984), Croft and Crosby (1987), and Crosby and Klausing (1984). Many of the investigations since about 1975 have been oriented toward coal and hydrologic relations. Many of the investigations made by the U.S. Geological Survey were initiated to aid the U.S. Bureau of Land Management in their management and leasing respon- sibilities (Crawley and Emerson, 1981; Armstrong, 1982; Horak, 1983a, 1983b). The studies also contrib- uted to an expanded data base that, in turn, made it possible to determine some environmental effects and to project probable future environmental effects of coal mining. Other investigations, principally done in cooper- ation with the State agencies, have been oriented toward understanding the hydrologic and hydrochem- ical processes that may be affected by mining and documenting these effects at selected sites. Many of these investigations have been listed by Cochran and others (1983). Extensive investigations, such as those by Moran and Cherry (1977) and Groenewold and Rehm (1980), have identified mining and reclamation practices that will maximize postmining reclamation and land use. Reports of other studies, such as those by Sandoval and others (1973) and Sandoval and Gould (1978), discuss methods of returning the landscape to maximum productivity for cropland or rangeland, which were the principal land uses prior to mining. To develop a predictive means of simulating the effects of land-use changes such as opening a mine on previously farmed land, rainfall-runoff models were applied to typical small basins (Emerson, 1981, 1988). Because basin characteristics have not been regionalized, extension of these models to other un- mined or mined sites has not been possible. Moran and Cherry (1977 ) reported temporal increases in saline seeps associated with mining. The overall flow regimen probably would be altered only slightly by min- ing unless impoundments or other alterations are made on the main-stem streams. Van Voast and others (1978) and Groenewold and others (1983) indicated that leachates of spoils material at several sites throughout the region are enriched two to five times in dissolved-solids concentrations relative to median dissolved-solids concentrations in unmined aquifers. Thus, runoff from spoils or reclaimed areas, or both, probably will have increased salinity. Moran and others (1979) stated that the dissolved- solids concentration of a stream affected by mining is approximated by the mass-balance mixing of spoils water and natural ground water in proportion to the percentage of the drainage basin occupied by mine spoils. If only a small percentage of the drainage basin upstream from a monitoring site is occupied by mine spoils, no change in water quality would be observable at the site. Attempts to assess basin-wide effects of mining have been few. To date (1985), perhaps the most successful assessment was a model developed by Woods (1981) to determine the effect of mining at multiple sites in south- eastern Montana on the dissolved-solids concentration in the Tongue River. Because of the complexities and uncertainties involved, most attempts to assess the cumulative effects of mining on the hydrologic system have resembled the qualitative approach of Lumb (1983). Typical cones of ground-water heads in the vicinity of mines in North Dakota and Montana have been described by Groenewold and others (1979, 1983), Davis (1984b), and many others. Because typical strip-mining processes produce a zone of relatively large hydraulic conductivity at the base of reclaimed spoils (Winc- zewski, 1977), most aquifers destroyed by mining are restored effectively during reclamation. Rehm and others (1980) have stated that the restored aquifers have hydraulic properties within the range exhibited by unmined lignite and sandstone aquifers in the area. Furthermore, because spoils and reclaimed soils have greater infiltration capacities than undisturbed over-burden, and because water-table depressions pro- duced by mine operations induce large horizontal gra- dients in hydraulic head, water levels in most reclaimed aquifers approach premining levels within a few years of mine cessation. Restoration of premining water levels has been documented extensively at several North Dakota mine sites by Groenewold and others (1979, 1983). 70 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM. 1974—84 Because increased quantities of soluble salts seem to characterize spoils relative to unmined overburden, most efforts at predicting the quality of spoils water have used ultimate column leaches (Hood and Oertel, 1984) or batch mixing solutions (Davis, 1984b). Dif- ficulties relating experimental results to water-rock reaction ratios in disturbed and undisturbed aquifer systems have caused investigators to use empirical ratios between experimental analyses and measured spoils-water quality in one locality to predict spoils- water quality elsewhere. Resultant large uncertainties in these predictive methods need not obscure their util- ity as a management tool at sites having no previous history of mining. Geochemical studies of spoil-pile samples from mines in the region indicated that leachate from spoils was more alkaline and contained considerably more soluble- salt material than did the leachate from natural topsoil from nearby locations (Sandoval and others, 1973; Groenewold and others, 1983). Sodium-adsorption ratios for spoil extracts ranged from 2 to 64; most ratios were large enough to indicate limited revegetation potential. Other analyses (Power and others, 1974) indicated that Paleocene shales from depths greater than 30 feet con- tained considerable exchangeable ammonium nitrogen. When these shales were exposed to the atmosphere, nitrification of exchangeable ammonium occurred that resulted in increased soil nitrate content after mining. Groenewold and others (1983) reported ground-water nitrate emichments at several mine sites and attributed them to a combination of the above process and nitrate released by explosives used in mining operations. Investigations of the processes affecting ground- water quality in mined and unmined shallow bedrock aquifers in the Fort Union region have been done by Moran and others (1978) and Groenewold and others (1983). These investigations documented a consistent set of interrelated chemical reactions. Mining Operations accelerate these chemical reactions, which already are operative in the natural environment. Acceleration of sulfide-mineral oxidation and dissolu- tion of generated gypsum catalyzed by cation exchange are the principal reasons that spoils water contains sulfate and sodium. Although the quality of ground water in the vicinity of surface mines apparently has deteriorated to some extent, the mineral content of deteriorated water generally remains less than or equal to the maximum concentration that occurs in some wells prior to mining. However, Rahn (1975) and Van Voast and others (1978) indicated that median concen- trations of dissolved solids in spoils water can be 60 to 73 percent greater than those in domestic and livestock wells in the vicinity of mines. Moran and others (1978) and Groenewold and others ( 1983) determined that median concentrations of sulfate in spoils water were two to five times greater than those in wells unaffected by mining. From 1955 to 1967, uranium was obtained from uraniferous coal in eastern Billings, northeastern Slope, and western Stark Counties, North Dakota, by first stripping the overburden from the coal, then covering the coal with waste oil and old tires and burning the coal onsite, transporting the ash by truck to nearby kilns for further concentration of the uranium in the ash, and subsequent shipping of the ash by train to uranium- ore processing plants in Colorado and New Mexico. Once mining ceased, seepage from the lignite aquifers flooded mine pits, enabling uranium and associated trace-element constituents in residual ash and spoils to contaminate the ground water. Because infiltration in spoils and reclaimed land is depression concentrated and fracture controlled, in- filtrating water contacts a finite quantity of disturbed soils and geologic materials that are a source for solu- ble contaminants. Thus, available mineral salts in mined land can be removed by leaching infiltration in a rela- tively short period of time. Groenewold and Murphy (1983) have reported that water in spoils older than 25 years is not appreciably more concentrated in dissolved solids than water in adjacent, undisturbed aquifers. Thus, disturbance of unreclaimed spoils can produce groundwater-quality degradations from unleached salts as severe as those initially produced by mining. Projection of hydrologic and hydrochemical con- sequences observed at mine sites to downgradient, offsite locations has been hindered by limited infor- mation about the hydraulic properties of spoils and coal and by the complexity of the chemical reactions involved. Available data about hydraulic properties for the aquifer materials have been collected by Rehm and others (1980), but very little is known about anisotropy and fracture control of flow. Furthermore, little is known about the potential of downgradient aquifers to cleanse themselves of mine wastes by ion exchange on coal and clay minerals and by sulfate reduction. HYDROLOGIC ISSUES RELATED TO COAL MINING Hydrologic issues that have occurred in the region as a result of mining have been few. Isolated issues that have been identified are associated with: (1) Modifica- tion of land-surface topography and stability; (2) in- creased sediment discharge; and (3) alteration of the quantity and quality of surface water and ground water available in mining areas. However, most of these issues may be ameliorated by application of appropriate min- ing and reclamation procedures. NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES—FORT UNION REGION 71 Because mining changes the configuration of the land surface. postmine landscapes sometimes are not suit- able for return to original land uses. For instance, if sufficient spoils are not available to restore all the land- scape to elevations above the water table, it may be necessary for some of the land to be established as wetlands for use by migratory waterfowl or to construct lakes for recreational use. In the worst example of reclamation, unreclaimed spoils may be suitable only for wildlife habitat; or, in the instance of underground mining, caving may produce a pitted land surface that is of limited usefulness and that is potentially dangerous. The effects of mining on the quantity and timing of runoff are dependent on mining practices, but these ef- fects usually are very minor. Some realignment of small tributaries may be needed, but the overall drainage areas should remain virtually unchanged. The watershed area subject to active mining at any one time is too small to substantially 'alter the flow magnitudes except possibly during periods when there is only base flow. Alteration of existing stream chan- nels to intercept and divert surface runoff within the mining area can cause alterations in existing flow regimes downstream. Local erosion and transportation of sediment may increase, at least initially, because of removal of vege- tative cover from land contributing runoff to the streams. The sediment yield will depend on the mining practices—the development of sedimentation ponds, the stability of diversion channels, and the speed with which vegetative cover can be reestablished (Gilley, 1980). However, State and Federal laws require containment of sediment and runoff within the mine site, so offsite effects should be minimal. Surface waters within mining areas typically are degraded by increased concentrations of sulfate, which is consistent with leachate studies. Locally, decreases in pH and increases in concentrations of hydrogen sulfide, sodium, bicarbonate, iron, and fluoride also may occur. However, the quantity of recharge of streams by runoff will be decreased by increased infiltration in spoils and reclaimed soils and containment of runoff within mined areas. Water that unavoidably flows into or precipitation that falls directly on the mined area can be pumped or otherwise routed to impoundments where excessive sediment loads and objectionable chemical concentrations may be ameliorated by settling or discharged during high-flow periods. Because runoff and surface-water flow could be readily managed by routing techniques or construction of impoundments, surface runoff should produce little change in the chemical quality of water diverted around active min- ing areas. If necessary, mines could chemically treat water leaving mine sites to minimize further changes in surface-water quality. Despite control of surface factors that might affect surface-water quality at mine sites, some degradation of surface-water quality downstream from mine areas may be observed because of seepage of degraded ground water into the surface water. To accurately determine the effects of strip mining on water quality, a premin- ing data base needs to be established in areas of pro- posed mining. Resampling of wells after mining has begun could indicate if there is any alteration of water chemistry because of the mining. However, most site- specific studies have had limited or no premining data; so water-quality changes attributed to mining in the studies were inferred from temporal and spatial changes as mining expanded or began in new areas. Effects of mining on shallow ground-water systems are not nearly so manageable using design features as those on surface-water systems. Removal of the coal could disturb enormous volumes of earth, some of which may have been a local aquifer. In some instances, ef- fects of mining excavations could extend into a regional flow system. In any event, all disturbed earth, whether it is saturated or unsaturated, is a medium for the move- ment of subsurface water, and removal of the coal disturbs the natural flow regime. Any aquifers within or above the coal could be destroyed at the mine site during mining and could cause decreases in hydraulic head for some distance from the mine. In most in- stances, local wells destroyed by mining can be replaced in deeper strata. The distance to which the drawdown extends and the rate at which it spreads depends on the magnitude of recharge and discharge fluxes, the prox- imity of the mine to the ground-water recharge and discharge areas, and the hydraulic characteristics of the aquifers and adjacent materials. Normal dragline stripping procedures result in the replacement of materials in approximate reverse order of their original state except for the topsoil. Near- surface material is replaced at the base of the spoil piles and deeper sediments on the surface. Therefore, sedi- ment that was deeply buried and at equilibrium with a reducing environment would be replaced in an environ- ment of rapid oxidation and weathering; similarly, sedi- ment that had been exposed to surface oxidation would be replaced in a reducing environment, often below the water table. Resultant oxidation, weathering, and alter- ation reactions could produce large quantities of solu- ble salts that might affect infiltrating water. Because strata exposed in mine pits are exposed to the same chemical processes that affect spoils, these strata also may be a source of soluble salts to infiltrating water. Because sulfide minerals such as pyrite contain abundant quantities of trace-element contaminants, 72 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM. 1974-84 oxidation could release trace elements to the hydrologic system. However, the quantity and nature of the clay minerals in the Paleocene Fort Union Formation would indicate that transport of trace elements would be retarded greatly by adsorption and cation exchange. Furthermore, the large bicarbonate concentration in most ground and surface waters would produce rapid precipitation of most trace elements even though present in small concentrations. Thus, the potential for contamination of ground water by concentrations of trace elements toxic to people would seem to be small. POWDER RIVER, BIGHORN BASIN, AND WIND RIVER REGIONS 73 NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES— POWDER RIVER, BIGHORN BASIN, AND WIND RIVER REGIONS By JAMES F. WILSON, JR., and MICHAEL R. CANNON The Powder River, Bighorn Basin, and Wind River Montana and northern Wyoming (fig. 37). Coal areas regions, in addition to the Bull Mountain coal field, com- 48—51 (table 1) are located in this larger area, although pose an area of about 60,000 square miles in southern important coal resources in the southernmost part of 1 48°30 E 109° 108° 1c7° 103° {mlBalo 2 "‘ \ a” ‘\ r~ , . 3;)"; f 3A $3 1‘. {}_as.:\._ / 155 / H, »>( ‘356 : o , up] 47 33”??ij ,,,,, I 11;»: ‘s-ngs/ygn _ T “’1 Bulb wast“; r::§jv»g,,€ \ . ‘_ 3,231: ‘55 $172:qu /%&de 45° 44°? 43° r ,4 ‘ ,, :?\I/ V ///fl/f , new" /‘ n ,v , / \ 1 m\ “ away. \ 42° ’ ‘ I, Base from U. S. Geological 0 l 50 100 MILES Survey 122,500,000 United States base map \ 0 50 100 KILOMETERS EXPLANATION COAL REGION 50 COAL-AREA BOUNDARY—Number refers to local areas (Cochran and others, 1983) .2 SAMPLING SITE AND NUMBER—See figures 38 and 39 for site descriptions FIGURE 37.—The Powder River, Bighorn Basin, and Wind River regions, the Bull Mountain coal field, coal areas 48-51, and sampling sites. 74 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 the Powder River region are outside the boundary of coal area 50. The regions are located in three physio- graphic provinces. From east to west these provinces are the Great Plains, Wyoming Basin, and the Middle Rocky Mountains (pl. 1). The regions are located entirely within the Missouri River drainage basin. About 83 per- cent of the regions are located in the Yellowstone River basin; the remainder of the regions are located in parts of the Musselshell, Cheyenne, and North Platte River basins (fig. 37). Climate is typified by small annual precipitation, cold winters, and hot summers. Throughout most of the plains and intermontane basins, average annual precip- itation is 12 to 16 inches. Most precipitation occurs from showers and thunderstorms during. April through August. Precipitation in the mountains generally aver- ages 20 to 30 inches; much of the precipitation that sus- tains streamflow throughout the year occurs as snow in the mountains during November through April. Exposed bedrock units range from crystalline rocks of Precambrian age to sedimentary rocks of late Ter- tiary age. Unconsolidated deposits of Quaternary age occur in many stream valleys. The bedrock units delineate the major structures of the area—mountain uplifts and large, synclinal basins. Rocks of Precam- brian age are exposed in the cores of major uplifts, and progressively younger rocks through Cretaceous age outline the uplifts and basins. The centers of the basins are characterized by widespread deposits of rocks of early Tertiary age. About 50 percent of the land is privately owned, about 10 percent is owned by Indians [Crow and N orth- ern Cheyenne (in Yellowstone, Bighorn, and Rosebud Counties, Montana) and Wind River (in Hot Springs and Fremont Counties, Wyoming) Reservations], about 5 percent is owned by the States of Montana and Wyo- ming, and the remaining 35 percent is owned by the Federal Government. Federal lands are administered by the US. Bureau of Land Management and the US. Forest Service. Except for the lands of the Indian Reser- vations, the Federal Government owns most of the minerals beneath the surface. Agriculture—rangeland and cropland, both irrigated and nonirrigated—is the primary use of land. Irrigated croplands are located mainly along the valleys of the largest rivers. Forests cover 10 to 15 percent of the regions. The principal industrial use of the land is devel- opment of mineral resources, chiefly coal, oil, and gas. Surface coal mines occupy only a very small percent- age of the regions; lands reclaimed after the coal has been extracted generally are returned to rangeland or nonirrigated cropland. The regions generally are sparsely populated. The principal population centers are Billings, Montana (about 84,000), and Casper, Wyoming (about 59,000) (US. Bureau of Census, 1981). The population of Camp- bell County, Wyoming, which includes Gillette, approx- imately doubled to 24,000 from 1975 to 1980 in response to increased mining of coal. During 1980, about 95.5 percent of the water used was from surface-water sources and about 4.5 percent was from ground-water sources. Irrigated agriculture was by far the largest use, 98 percent of which was from surface-water sources. Most of the irrigation occurs along the Yellowstone, Clarks Fork, Wind/Bighom, Tongue, and Powder Rivers. The flows of these rivers and their tributaries generally are fully appropriated. COAL RESOURCES Coal is abundant in the regions, particularly in the Powder River region. Coal deposits are contained in rocks of Late Cretaceous and early Tertiary age. Most of the coal is federally owned; however, in the northern part of the regions, there is a complex pattern of Fed- eral, State, and private ownership. Coal on the Indian Reservations generally is held in trust for the tribes. At least 12 formations of Late Cretaceous age have coal; however, the most economically important coals are in the Eagle Sandstone and Judith River Forma- tion in the Bighorn Basin region in Montana and in the Meeteetse and Mesaverde Formations in the Wind River and Bighorn Basin regions in Wyoming. Most coal fields are located near the flanks of the structural basins where the deposits are steeply dipping and af- fected by local folding and faulting. In Montana, the Upper Cretaceous coal beds are relatively thin, ranging from less than 1 foot to a max- imum of 6 feet in thickness. In Wyoming, the coal beds generally are 4 to 6 feet thick, but a few are 15 to 30 feet thick. The coals generally are classified as high- volatile bituminous (figs. 3, 4) and have a smaller sulfur content (0.5 percent or less) and a higher heat content (about 10,000 British thermal units per pound) than the subbituminous coals of these regions. The Upper Cretaceous coal beds generally are recov- erable only by underground mining. During the early 1900’s, there were underground mines in Montana and Wyoming, but production was small (10 to 200 tons per day). At the present time (1985), there is no mining of Upper Cretaceous coal beds in these regions. The Fort Union Formation (Paleocene) and Wasatch Formation (Eocene) are the principal coal-bearing for- mations of early Tertiary age; the deposits underlie most of the plains areas of the Powder River region. Ma- jor coal beds within the Fort Union Formation can be traced for many miles, some for more than 50 miles. The Tongue River Member of the Fort Union Formation POWDER RIVER, BIGHORN BASIN, AND WIND RIVER REGIONS 75 contains the largest reserves of all the coal-bearing rocks; most of the coal presently mined is in this unit. Reserves of lower Tertiary coal in the regions are very large. Recent estimates of strippable coal are 33 billion tons in the Montana part of the regions (Matson and Blumer, 1973) and 24 billion tons in the Wyoming part (Lowry, Wilson, and others, 1986, p. 34). Total coal reserves in Wyoming exceed 500 billion tons. The estimated reserves for Wyoming probably will change as the result of a recently completed, extensive coal- mapping program by the US. Geological Survey. Generally, the lower Tertiary coal beds are nearly flat lying, are at or near the surface, and are uninterrupted by folding or faulting. The thickness of coal beds in one coal field in the Bighorn Basin region in Montana ranges from 3 to 12 feet. In contrast, in the Powder River region, the Healy coal bed (Wasatch Formation) near Buffalo, Wyoming, is more than 200 feet thick, and the extensive Wyodak-Anderson coal bed (Fort Union For- mation) near Gillette in northeastern Wyoming is 25 to 175 feet thick, averaging about 70 feet. Most of the Ter- tiary coals of these regions are classified as sub- bituminous (figs. 3, 4), although some coal in the extreme northern part of the Powder River region is lignite. Similarly to the Upper Cretaceous coal beds, the subbituminous Tertiary coals have a small sulfur content—less than 1 percent in the Wasatch and less than 3 percent in the Fort Union. Heat content of the coals generally is about 8,200 British thermal units per pound, although 10,000 British thermal units per pound is reported for coals in the Bull Mountain coal field. Strip mining of the lower Tertiary coal beds, par- ticularly in the Powder River region, is extensive. The coal beds first were mined during the late 1800’s in many of the coal areas. Much of the early mining was underground, and production was small in comparison with that of the present (1985). As of 1984, there were no underground mines, but 26 active surface mines pro- duced more than 140 million tons in Montana and Wyoming (table 3). All mining is from the Fort Union Formation, except in Converse County, Wyoming, where mining is from the Wasatch Formation. Nearly all the coal produced is used for thermoelectric power generation in at least 16 States. Campbell County, Wyoming, leads the Nation in coal production with 14 mines, including the Nation’s two largest, Black Thunder and Belle Ayr. All 14 mines are in the Wyodak coal bed. Two additional mines are under construction in Campbell County, and several more are planned in Montana and Wyoming. Most of the foregoing information about coal resources, unless otherwise noted, was obtained from the coal-area hydrology reports by Slagle and others (1983, 1986), Lowry, Wilson, and others (1986), and TABLE 3.—Active coal mines and coal production by county, lower Tertiary coal beds, Powder River, Bighorn Basin, and Wind River regions and Bull Mountain coal field [Sourcesz Montana Department of Labor, written commun, 1984; Wyoming State Inspector of Mines, written commun., 1985] Location Number Production of mines of mines (millions of tons) Montana (1983) Big Horn County .................. 4 16.32 Musselshell County ................. 2 .02 Rosebud County ................... _2 12.12 Subtotals ........................ 8 28.46 Wyoming (1984) Campbell County ................... 14 106.80 Converse County ................... 1 3.34 Sheridan County ................... 2 2.52 Hot Springs County ................ _1 .04 Subtotals ........................ E 112.70 Totals .............................. 26 141.16 Peterson and others (1987). The original sources of the information are cited in those reports. HYDROLOGY Information about the hydrology of the Powder River, Bighorn Basin, and Wind River regions was inadequate, in view of the impending very large increase in coal strip mining in the eastern half of the region. Two general- ized objectives for the coal-hydrology program in these regions have remained unchanged since 1974: (1) '1b in- crease knowledge about the availability of water; and (2) to provide information for assessing the effects of coal development on water resources on a regional basis and on a site-specific basis. The term “coal development” in- cludes mining and transportation of coal, mine reclama- tion, coal conversion, and increased population. A twofold approach was used: (1) Hydrologic-data net- works were expanded or otherwise enhanced; and (2) in- terpretive hydrologic studies were begun. The two types of activities, which were coordinated, were confined mainly to the Powder River region. Regionally, there was a substantial base of information upon which to build, largely because of long-term interest in water sup- plies in the semiarid areas of the regions. Locally, the information base was sparse to nonexistent; for exam- ple, except for flood hydrology, little was known about the hydrology of small basins that have ephemeral streams typical of most of the areas underlain by strip- pable coal. Data bases and techniques for estimating flood discharges and hydrographs for small ephemeral streams were available from previous studies and proved invaluable to the coal-hydrology program. 76 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 SURFAC E—WATER NETWORK Collection of systematic (continuous or recurrent) records of discharge in these regions began during the 1890’s at a few sites. Collection of daily or monthly water-quality and suspended-sediment samples did not begin until 1946. Most stream-discharge, water-quality, and suspended-sediment sites were located on the largest streams and their principal tributaries. The data were needed by the U.S. Geological Survey to inventory and assess the surface-water resources of the regions and by State and other Federal agencies to plan and develop surface-water resources. Over the years, the network has been expanded substantially. Sampling frequency and types of analyses of water- quality samples vary at a site and differ among sites. Onsite determinations of temperature, pH, and specific conductance of water commonly are made. At some sites, determinations may include dissolved oxygen and turbidity. Laboratory analyses usually include common dissolved ions (salinity) but also may include trace elements, nutrients, pesticides, and radiochemical and biological constituents. During 1974, there was considerable expansion of the surface-water network for coal hydrology and of the funding from three main sources: (1) The U.S. Geo- logical Survey’s program to describe the water resources of the coal regions; (2) the U.S. Bureau of Land Management’s program to obtain information for the orderly and environmentally sound development of Federal coal resources; and (3) the U.S. Environmental Protection Agency’s energy-related program for ex- panded water-quality monitoring at network sites. The coal-hydrology program included a 4-year (197 7—81) ex- periment in which the U.S. Geological Survey con- tracted with a private engineering firm to install and operate a number of stream-discharge, water-quality, and suspended-sediment sites in the regions. The data collected by the contractor met or exceeded U.S. Geo- logical Survey accuracy standards, but the activity was determined to be more costly than if done by U.S. Geological Survey personnel; therefore, the experiment was terminated. Pairs of stream-discharge sites were installed during 1974 at 16 locations to measure gains or losses of discharge in streams crossing outcrops of the Madison Limestone in the Powder River region in northeastern Wyoming. This project was done in cooperation with the Wyoming State Engineer and the Old West Regional Commission. The number of pairs of sites was decreased during 1976; operation of the remaining sites continued through 1982. Many of the sites were located at remote, high-mountain locations, making access difficult. Time and funding constraints precluded construction of cableways for making high-water measurements. Therefore, dye-dilution techniques were modified for measuring discharge at the remote sites. The results were considered adequate for computing the high-water discharge at the sites. As of 1982, approximately 290 stream-discharge sites, 230 water-quality sites, and 190 sediment sites had been operated at one time or another in these regions. Most water-quality and sediment sites are located at stream- discharge sites so that dissolved-solids concentrations and suspended-sediment loads can be computed. The period of record at these sites ranges from 1 year to many decades. Since 1980, the network has decreased steadily as Federal funding for coal-hydrology investi- gations has decreased. However, most of the sites added after 1973 have 5 or more years of record, thus pro- viding valuable documentation of conditions with which future conditions can be compared. In addition to continuous-record stream-discharge sites, the network includes partial-record stream- discharge sites and miscellaneous stream-discharge sites. Partial-record stream-discharge sites are operated to obtain supplemental flood-frequency data. Funded mainly in cooperation with the Montana and Wyoming Highway Departments, most partial-record stream- discharge sites are used to record only the annual peak discharge, but a few specially instrumented sites were used to record complete hydrographs of rainfall and runoff on small ephemeral streams. A total of 117 partial-record stream-discharge sites have been oper- ated in the regions, but most have been discontinued. Approximately 300 documented miscellaneous stream~discharge sites are in the regions, particularly in the Powder River region. These are sites where one- time or occasional measurements of discharge have been made or chemical or biological samples collected. Com- monly, data collection is done on a synoptic or areawide basis, such as reconnaissance measurements of low flow or water quality. SURFACE—WATER CHARACTERISTICS Although about three-fourths of the area of the regions consists of plains drained by ephemeral and in- terrupted streams, most of the runoff occurs in the few large perennial streams that transect the regions and in their smaller perennial tributaries that originate in the mountains. Discharges in all streams have large seasonal and year-to-year variations (fig. 38); the largest discharges occur during the spring as a result of snowmelt and rainfall. Plains streams, many of which do not flow most of the year, can have large flows of short duration because of thunderstorms. Characteris- tics commonly used to describe or assess surface-water POWDER RIVER, BIGHORN BASIN, AND WIND RIVER REGIONS 77 .a O O O | I I | | I I I | | I Ephemeral stream A (Site 1) 100 — '— 8 | I 1’°°°|||||||l||l Perennial stream 100 __ (Site 2) STREAM DISCHARGE, IN CUBIC FEET PER SECOND 8 l sasssrszszs: D U- : W 02 W E<§gq Z 6 5' — 2,000 — — (I) Q 8 0 1,000 — 7 I — . I __ 0 I EXPLANATION 4 SITE NUMBER MAXIMUM I MEDIAN MINIMUM SITE NUMBER IN FIGURE 37 SITE NAME AND DRAINAGE AREA 4 Clarks Fork Yellowstone River near Belfry, Montana (1,154 square miles) 5 Powder River at Sussex, Wyoming (3,090 square miles) 6 Powder River at Moorhead, Montana (8,088 square miles) 7 Clear Creek below Rock Creek near Buffalo, Wyoming (322 square miles) 8 Squirrel Creek near Decker, Montana (33.6 square miles) 9 Hanging Woman Creek near Birney, Montana (470 square miles) 10 Otter Creek at Ashland, Montana (707 square miles) 11 East Fork Armells Creek near Colstrip, Montana (97.3 square miles) 12 Little Powder River below Corral Creek near Weston, Wyoming (204 square miles) 13 Belle Fourche River above Dry Creek near Piney, Wyoming (594 square miles) 14 Caballo Creek at mouth near Piney, Wyoming (260 square miles) 15 Little Thunder Creek near Hampshire, Wyoming (234 square miles) FIGURE 39.—Maximum, median, and minimum dissolved-solids con- centrations at selected surface-water-quality sites in the Powder River, Bighorn Basin, and Wind River regions. a few milligrams per liter to more than 100,000 milli- grams per liter. Annual sediment yields from basins in these regions range from less than 10 to more than 1,000 tons per square mile. Equations are available for esti- mating sediment yields from small ungaged watersheds POWDER RIVER, BIGHORN BASIN, AND WIND RIVER REGIONS 79 in eastern Montana (Lambing, 1984). Suspended sedi- ment in streams of the Powder River structural basin of southeastern Montana was summarized by Litke (1983). Valuable sediment data were collected at many sites during the widespread flooding of May 1978 (Parrett and others, 1984); a maximum daily suspended-sediment load of 2.81 million tons was recorded in the Powder River at Arvada, Wyoming (Sheridan County), on May 20, 1978. The load in the Powder River at Arvada dur- ing 197 8 was 16.3 million tons; the 24-year average an- nual load is 4.7 million tons. In contrast, the 5-year average annual load in the Yellowstone River at Billings, Montana, is only 1.7 million tons. Sediment data for small ephemeral streams are scarce. Calibration and operation of automatic-sampling devices necessary to collect the data is expensive and time consuming. Such data are available, however, for one experimental basin (drainage area, 0.8 square mile) in Wyoming. GROUND—WATER NETWORK Information for about 3,200 wells in the Montana part of the regions and about 4,300 wells in the Wyoming part is stored in the computer files of the US Geolog- ical Survey. The information for each well includes some or all of the following: ownership and use, completion data, geologic data, water levels, well yield, and water quality. From the 1920’s to 1974, most ground-water data were collected during hydrologic studies to assess the availability of water. The information was needed mainly for agricultural, municipal, and domestic use. In response to increased coal development beginning in 1974, large quantities of ground-water data have been collected by industry, universities, and government agen- cies, mostly for studies of the hydrology and possible environmental effects in relatively small areas, such as a proposed or existing mine, but also for identifying base- line conditions. Much of the data collected in Campbell County, Wyoming, consists of water levels in the Wyodak-Anderson coal bed and in the overburden. These data are centralized through the Gillette Area Ground- water Monitoring Organization (GAGMO), which is mainly composed of mining companies. Water levels were reported for more than 780 wells at 21 mine sites during 1981, resulting in a compilation of data for most of the area close to the coal outcrop in Campbell County (Gillette Area Groundwater Monitoring Organization, 1983). The US. Geological Survey continues to measure water levels and collect water samples for chemical anal- yses from selected wells. Most samples are analyzed only for common ions; some analyses include trace elements. Samples for analysis of radiochemical constituents mainly are from areas of known uranium deposits. Re- cent reports for Montana include data for test wells (Lev- ings, 1981b; Wood, 1984), for water-supply wells (Slagle and Stimson, 1979; Levings, 1981a), and for water quali- ty (Lee, 1979). Reports for Wyoming include water-level data (Ragsdale, 1982), records of wells in alluvium (Wells, 1982), and water-quality data (Wells and others, 1979; Wells, 1982; Larson and Daddow, 1984). GROUND—WATER OCCURRENCE Paleozoic aquifers, such as the Mississippian Madison Limestone, Pennsylvanian 'Iensleep Sandstone, and the Pennsylvanian and Permian Minnelusa Formation, may yield more than 950 gallons per minute. However, the large yields are dependent on secondary permeability that is not present everywhere. Water has not been developed from these rocks in much of the area because of the great depth at which they occur and the uncer- tainty about the quality of the water. Where the rocks occur within 2,000 feet of the surface, which is only near the uplifts, the dissolved-solids concentration commonly is about 500 milligrams per liter. Upper Cretaceous and lower 'Ibrtiary aquifers, such as the Fox Hills—Hell Creek (or Lance) and Fort Union— Wasatch, are the most extensively used. Aquifers within the formations primarily are sandstones but may in- clude coal beds, particularly in the 'Iertiary aquifers. Most wells are drilled for stock or domestic supplies, which generally can be obtained from wells less than 500 feet deep. Because the rocks are more than 500 feet thick in some places, yields greater than 500 gallons per minute can be obtained from deep wells. The quality of the water in shallow wells generally is a sulfate type and has dissolved-solids concentrations in excess of 2,000 milligrams per liter. Water from wells more than 250 feet deep generally is a bicarbonate type and dissolved-solids concentrations are about 1,000 milligrams per liter. Quaternary alluvial deposits that will yield adequate quantity and quality of water for even the relatively small supplies required for stock or domestic use gener- ally occur only along streams that originate in and near the mountains. Elsewhere the alluvium may be too thin or too fine grained, or the quality of the water may be too poor for most uses. Alluvium occurs in the flood plains and as capping on terraces, but the deposits cap- ping the terraces may not have adequate saturated thickness to yield water to wells unless irrigation is prac- ticed on the terraces. GROUND—WATER QUALITY Water quality, as characterized by the quantity and type of dissolved constituents, differs greatly within and 80 between the aquifers. Dissolved-solids concentration, the most commonly used indicator of water quality, ranged from 150 to 16,500 milligrams per liter in samples from approximately 2,000 springs and wells. Average concentrations in various aquifers ranged from about 1,500 to about 3,300 milligrams per liter; about 90 percent of all samples had concentrations exceeding 500 milligrams per liter. Dominant ions vary throughout the regions. In north- eastern Wyoming, dominant ions in aquifers in rocks of Paleozoic age are calcium and bicarbonate at or near the outcrops but are more likely to be sodium and sulfate or chloride at increasing distances from the out- crops. Water from shallow wells in the Upper Creta- ceous and lower Tertiary aquifers has calcium, sodium, and sulfate; water from deeper wells has sodium and bicarbonate, similar to most wells in the western part of the region in Montana. The dominant ions in water in southeastern Montana are calcium, magnesium, sodium, and bicarbonate in shallow wells drilled at topo- graphically high areas, sodium and sulfate in shallow 471100 - 1090 I 108° if--. -‘. ,5. {5 1°7°_, SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM. 1974-84 wells drilled in valleys, and sodium and bicarbonate in deeper wells (Lee, 1980). Dissolved trace-element concentrations generally are small in most aquifers; the notable exceptions are iron and manganese, which occur in large concentrations in water from many wells. In northeastern Wyoming, for example, dissolved manganese concentrations exceeded 50 micrograms per liter in 43 percent of the wells sampled, and dissolved iron concentrations exceeded 300 micrograms per liter in 35 percent of the wells. Large concentrations of boron, selenium, and strontium occur in a few wells throughout the regions. COAL—HYDROLOGY STUDIES Many studies and reports that characterize the water resources or that evaluate potential effects of coal min- ing in these regions, mostly in the Powder River region, have been completed since 1974 (fig. 40); a few remain- ing reports are in progress. With few exceptions, the studies were done as part of either the U.S. Geological ‘106° 105° ll «~~\ . J; 104° 420'“ x- I“ . , Base from U. S. Geological 0 Survey 12,500,000 United States base map 0 50 100 MILES 100 KILOMETERS FIGURE 40 (above and facing page).—Location of areas of coal-hydrology studies done in the Powder River, Bighorn Basin, and Wind River regions. POWDER RIVER, BIGHORN BASIN. AND WIND RIVER REGIONS EXPLANATION AREAS SHOWN ON MAP COA L—AREA BOUNDARY SMALL AREAS AND NUMBER 81 AREAS NOT SHOWN ON MAP (Study areas extend beyond 48 Slagle and others (1986) mapped area) 49 Slagle and others (1983) MONTANA MONTANA 50 Lowry, Wilson, and others (1986) 1 Cannon (1982) Druse and others (1981) 51 Peterson and others (1987) 2 Ea‘fn‘m (113:3) Lambing (1984) OTHER LARGE AREAS : M:Vol:r::)nds3(1982); Cary (1984) tee (1979' 1980)1 8 MONTANA 5 McClymonds (1984a) ee'and others ( 9 1) Levmgs (1981b) Dockins and others (1980) 6 McClymonds (1984b) Lewis and Hotchkiss (1981) Ferreira (1984) 7 McClymonds (1985) Lewis and Roberts (1978) 8 U. 5. Bureau of Land Management (1975b) Miller (1979‘ 1981) Knapton and Ferreira (1980); 9 U. S. Bureau of Land Management (1977a) Omang and others (1982) . . . . thke (1983) 10 U. 5. Bureau of Land Management (1978) Slagle and Stimson (1979) : : 2 : : : Slagle and others (1985) 11 U. 5. Bureau of Land Management (1982) Stoner and Lewis (1980) 12 Davis (1984b) (2 areas) Wood (1984) Woods (1981) WYOMING WYOMING WYOMING 13 Boner and others (1976) (Numerous sites Armentrout and Wilson (1987) 7 is m area shown and in 3 areas outside Daddow (1986b) mapped area) Druse and others (1981) ‘i” 4‘ 14 Druse (1982) (3 areas) Glover (1984) 1 Peterson (in press) 15 Jordan and others (1984) Larson and Daddow (1984) A7 7 16 Larson (1985) Lenfest (1985) 17 Lenfest (1987) (2 areas) Lowry (1981’ 18 Lowry and Rankl (1987) Peters“ (1988) 19 Naftz (1985) (2 areas) Rankl and Lowry (in press) 20 Rankl (1982) (3 areas) Wells (1982’ 21 Rankl (1987) 22 Wangsness (1977); Ringen and others (1979) Survey’s coal-hydrology program or the U.S. Bureau of Land Management’s Energy Minerals Rehabilitation Inventory and Analysis (EMRIA) program. The studies were of three general types: (1) Appraisals or summaries for large areas within the region; (2) descriptions or assessments of small areas, usually of lease-tract size; and (3) studies of hydrologic processes. , Surface-water quality was assessed in several‘areal studies. The chemical quality and low flow of many streams in the Powder River region were documented during 1977—7 8 by Druse and others (1981). Statistical summaries of surface-water quality at sites in the coal areas were made for southeastern Montana by Knap- ton and Ferreira (1980) and for northeastern Wyoming by Peterson (1988). Larson (1985) summarized the chemical quality of the North Platte River in the vicini- ty of Casper, Wyoming; this stream reach is suscepti- ble to indirect effects of coal development in northeastern Wyoming. Additional studies by Lee and others (1981) documented the chemical quality of base flow of Otter Creek, Tongue River, and Rosebud Creek in southeastern Montana. Peterson (in press) studied benthic invertebrate communities in streams in north- eastern Wyoming. Invertebrate data are useful to eval- uate possible changes in stream environments caused by land disturbance; for example, Peterson (in press) determined that the average density of invertebrates in the Belle Fourche River downstream from a coal mine was larger than that upstream from the mine. Studies of large areas included determination of the areal extent and hydrogeologic properties of significant aquifers. Lewis and Hotchkiss (1981) delineated and described five hydrogeologic units in the Wasatch-Fox Hills sequence in Montana and Wyoming. Lewis and Roberts (1978) and Stoner and Lewis (1980) did sim- ilar studies in eastern and southeastern Montana. Miller(1979, 1981) mapped the thickness and con- figuration of the base of the Fox Hills—lower Hell Creek aquifer and described the availability of water from selected aquifers in the Powder River region. Daddow (1986b) mapped the potentiometric surface of the Wyodak-Anderson, the coal bed mined exten- sively in northeastern Wyoming. Ringen and Daddow (in press) demonstrated that the alluvium of the Powder River valley is not a significant aquifer. Lowry(1981) and Rank] and Lowry (in press) indi- cated that regional movement of ground water in shallow aquifers in northeastern Wyoming is not substantial in comparison with local movement. Dis- charge by evapotranspiration from alluvial aquifers was 82 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 estimated by Lenfest (1985) at 12 sites in Wyoming and Montana. Two regional studies of major aquifer systems contributed important information for assessing the availability of large supplies of ground water for coal development: the Madison Limestone Study (US. Geological Survey, 1975) and the Northern Great Plains Regional Aquifer-System Assessment (US. Geological Survey, 1979). Although not part of the coal-hydrology program, these studies were closely coordinated with that program. The areas of the two studies were nearly identical, encompassing eastern Montana, northeastern Wyoming, and western North Dakota and South Dakota. The Powder River region is within the areas of the two studies. Most of the information used in the two studies was obtained from data for municipal, industrial, and private water wells and from oil-and-gas wells; however, the U.S. Geo- logical Survey drilled three additional test wells for the Madison study and one well for the Northern Great Plains study. The studies produced a series of reports (for example, MacCary and others, 1983; Henderson, 1984). A preliminary description of the hydrology of the Madison Limestone was prepared in coopera- tion with the Montana Bureau of Mines and Geology and the Wyoming State Engineer (Swenson and others, 1976). Also, an assessment of losses and gains in discharge of streams crossing outcrops of the Madison was done in cooperation with the Wyoming State Engineer and the Old West Regional Commission (Boner and others, 1976; Wyoming State Engineer’s Office, 1976). The physical features, resources, and hydrology of four large areas delineated by major drainage basins were summarized by Slagle and others (1983, 1986), Lowry, Wilson, and others (1986), and Peterson and others (1987). The reports partly fulfill the requirement of the Surface Mining Control and Reclamation Act of 1977 (Public Law 95—87) for information to be used by regulatory agencies in assessing the potentialhydro- logic effects of proposed mines. The potential effects of surface mining on the hydrol- ogy of the following coal tracts were studied as part of the US. Bureau of Land Management’s EMRIA pro- gram in Montana: Otter Creek (US. Bureau of Land Management, 1975b), Bear Creek (US. Bureau of Land Management, 1977a), Hanging Woman Creek (US Bureau of Land Management, 1978), Pumpkin Creek (US. Bureau of Land Management, 1982), Prairie Dog Creek (McClymonds, 1982), Cook Creek (Cannon, 1982), Snider Creek (Cannon, 1983), Greenleaf-Miller (Levings, 1983), Corral Creek (McClymonds, 1984a), West Otter Creek (McClymonds, 1984b), and Horse Creek (McClymonds, 1985). The locations of these sites are shown in figure 40. Several additional site studies are scheduled for completion in the near future. Study results indicate that the effects of coal mining on ground water range from negligible effects in areas where coal beds are located above the water table to substantial problems associated with aquifer dewatering and leaching of soluble materials from mine spoils in areas where coal beds are saturated. In almost all areas, shallow aquifers removed by mining are not the only available water supply; wells drilled to deeper, unaf- fected aquifers could replace water supplies lost because of mining. In a study of the White Tail Butte area in north- eastern Wyoming, which included an EMRIA site, Lowry and Rankl (1987) described the effect of large areas of exposed clinker beds (coal altered by igneous intrusion) and the effect of changes in infiltration (as might be caused by reclamation, for example) on runoff from small basins. Lowry and Rankl (1987) also con- cluded: (1) That the area is a discharge area for part of the regional ground-water flow, although the volume of water moving regionally is small compared to that mov- ing locally; and (2) that, based on water-quality samples, the discharging water had moved only a short distance through the bedrock. Water in aquifers below the coal is adequate for the present land use, which is livestock grazing. Two studies were made of the hydrologic effects of an abandoned, nonrehabilitated surface mine in the Hid- den Water Creek area near the Wyoming-Montana State line. The mine was operated from 1944 to 1955. Wangsness (1977) determined that, in comparison with ponds outside the mined area, ponds within the mined area had smaller concentrations of dissolved oxygen, larger concentrations of major ions, and less diverse biological communities. Ringen and others (1979) deter- mined that the effect of mining on sediment yield was greater than the effect on chemical quality of water; sediment accumulated in a pond in the mined area at a rate more than 11 times that in a pond in a nearby unmined area. A study of potential hydrologic effects of active sur- face coal mining at the Belle Ayr and Caballo Rojo mines in Campbell County, Wyoming, was done in cooperation with the Wyoming Department of Environ- mental Quality. Preliminary findings, based on model- ing of discharge in Caballo Creek, were that there is little, if any, change in discharge caused by mining (Jordan and others, 1984). A report (Bloyd and others, 1986) provides a detailed description of the results of the surface-water modeling as well as a general descrip- tion of ground-water flow in the area. Digital-model simulations also were used in two studies in Montana. Slagle and others (1985) used POWDER RIVER, BIGHORN BASIN, AND WIND RIVER REGIONS 83 models to depict maximum drawdown of shallow ground water as a result of mine dewatering. Ferreira (1984) modeled the cumulative effects of mining and agriculture on concentrations of dissolved solids in Rosebud Creek; the simulations indicate that irrigation accounts for a larger cumulative percentage of dissolved-solids concentration than present (1985) min- ing, but that planned, full-scale mining would account for a larger percentage than irrigation. The objectives of several studies done in southeastern Montana and northeastern Wyoming were to describe or simulate various hydrologic processes. The general purpose of such studies, usually done at one or a few sites, is to use the knowledge gained in studies at other locations that have similar characteristics but that lack data. The studies are discussed in the following paragraphs. Three studies of hydrologic processes were concerned with discharge. A procedure using storage analysis of discharge records to assess the long-term water supply of an ephemeral stream was developed by Glover (1984); the method overcomes deficiencies of previous methods developed for perennial streams by accounting for zero- flow periods and the large day-to-day variability of dis- charge of ephemeral streams. Cary (1984) tested the US. Geological Survey’s Precipitation-Runoff Model- ing System (PRMS) to develop, calibrate, and verify a watershed model for use in simulating hydrologic proc- esses in small basins. Rankl (1982) developed an em- pirical infiltration model for estimating runoff from specified design storms in small basins that have ephemeral streams; the method uses incipient-ponding curves based on rainfall-runoff data and information from soils maps. In a follow-up study that estimated runoff from rainfall and infiltration, Rankl (1989) extended his previous work to other small ephemeral- stream basins that have a variety of soil types, and verified a physically based equation for defining the in- filtration parameters. The problem of predicting the source and quantity of sediment discharge from small ephemeral-stream basins was studied by Rankl (1987). Rankl collected rainfall, discharge, and sediment-concentration data from a very small basin (0.8 square mile) and used the PRMS to evaluate the physical processes that affect sediment production. For a given storm, sediment load and peak discharge were strongly correlated, a potentially useful means to evaluate the effects of surface coal mining on erosion and sedimentation. ' The interrelation of discharge and shallow ground water was the subject of two studies. Rankl and Lowry (in press) determined that vertical movement of ground water was restricted and that discharge occurred at out- crops not necessarily at stream level and in insufficient quantities to reach perennial streams. Also, data indi- cate that base flow, present in only a few streams, is related to local conditions rather than to a regional flow system. Lenfest (1987) investigated the function of alluvium in stream channels in recharging bedrock aquifers; the use of discharge losses between stream- discharge sites to estimate recharge resulted in under- estimation, in comparison with an analytic solution of a ground-water flow equation. Geochemical processes that control or affect the qual- ity of ground water in the Fort Union Formation were investigated by Lee (1980). Water quality was deter- mined to be related to mineralogy, distance along a ground-water flow path, ion exchange, and anaerobic bacteria that reduce sulfate to sulfide. The process of sulfate reduction in ground water was investigated fur- ther by Dockins and others (1980). Water quality and geochemical processes affected by mining were the subject of three studies. Woods (1981) developed a computer model for assessing the effects of various combinations of mining and agricultural development on potential increases in dissolved-solids concentrations in streams as a result of leaching of over- burden materials used to backfill surface coal-mine pits. In a study of geochemical processes that affect the quality of water in mine spoils, Davis (1984b) compared the quality of water in mine spoils with that in undis- turbed aquifers and developed a method of predicting future water quality in mine spoils and the effects of mine spoils on the hydrologic system. The geochemical processes in reclaimed mines in Wyoming is being studied by Naftz (1985) to determine the differences be- tween the quality of water in undisturbed aquifers and that in spoil aquifers. HYDROLOGIC ISSUES RELATED TO COAL MINING At the beginning of the coal-hydrology program it was recognized that one issue was a general lack of knowledge about the hydrology of small ephemeral- stream basins. The most significant issue of all, how- ever, was that no large, inexpensive supplies of water were readily available even though the Powder River region has abundant coal. That fact was the rationale for using air cooling, rather than less expensive water cooling, at the Wyodak steam-electric plant in Camp- bell County, Wyoming—the world’s largest (1983) plant of that kind. In another example, a plan to use ground water (20,000 acre-feet per year) for a coal-slurry pipeline from northeastern Wyoming to Arkansas was aban- doned, partly because of opposition to transporting water out of the area. The scale of present (1985) and planned mining in the Powder River region is large; therefore, some regional 84 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974—84 consequences of mining can be expected. The largest regional effect on water resources probably will be caused by the increases in population and land use, rather than by mining activities. Increased use of ground water in the population centers may cause over- withdrawals. Generally, however, regional effects of coal development on quantity of ground-water flow are unliker Erosion, sediment deposition, water-level declines, and water-quality degradation are potential local issues associated with mining. Erosion and sediment deposition, potential issues in any watershed affected by mining, generally are controlled at all active mine sites as required by State and Federal surface-mining regulations. Erosion is controlled by proper grading and revegetation of areas affected by mining, and sediment runoff is controlled through the use of set- tling ponds. However, channel scour can occur down- stream from settling ponds if the discharged water has suspended-sediment concentrations that are sub- stantially smaller than the natural, or premining, concentrations. Mine pits that intersect the water table will lower water levels near the mine and have the potential to af- fect nearby wells and springs. Water-level declines have been measured near surface mines in Montana and Wyoming; however, such effects generally are limited to areas within 1 or 2 miles of the mine. Lowering of the water table near the Wyodak Mine in Campbell Coun- ty, Wyoming, which began operations during the 1920’s, had not extended westward more than 1,500 feet by 1979. Where surface mining begins at the edge of the coal mine and extends some distance beyond the point where the water table is intercepted, the water levels in adjoining aquifers may be lowered. Although infiltration, runoff, and aquifer properties can be engineered in a variety of ways, the most likely effect of reclamation will be to insert into the existing hydrologic system a unit having completely different hydrologic properties. The changes could be either beneficial or detrimental; for example, flooding may be decreased downstream from reclaimed areas that have increased capacity for infiltration, but useful low-to- average flows also may be decreased. If recharge is increased, however, low flows could be increased by an increase in ground-water discharge to the stream. Possibly the most important long-term issue associ- ated with surface mining is the potential for degrada- tion of water quality because of the leaching of soluble materials from mine spoils. Chemical analyses of spoil- derived water from mine areas in Montana have indi- cated that the dissolved-solids concentrations of water in spoils statistically are greater than those in undis- turbed aquifers near the mines. Acid mine drainage is not an issue in these regions, however, because of the abundance of carbonate minerals and the large buffer- ing capacity of the natural waters. The hydrologic effects of underground mines were not studied as part of the coal-hydrology program in these regions because largescale underground mining is not planned. In an analysis of the effects of past under- ground mining in the Sheridan, Wyoming, area, Dunrud and Osterwald (1980, p. 41) described environmental issues such as subsidence, diversion or pollution of sur- face or ground water, and a variety of adverse effects of underground mine fires. They stated (p. 43—44) that such issues are much easier to assess and control in sur- face mining. On the positive side, water supplies have been developed from wells bored into flooded mine work- ings (Lowry, Wilson, and others, 1986, p. 8). The Green River and Hams Fork regions (fig. 41), an area of about 46,900 square miles located primarily in southwestern Wyoming and northwestern Colorado but also in parts of (table 1). Major coal fields outside the regions also are included in this discussion; these include the Hanna and Rock Creek coal fields in Wyoming and the North Park and Danforth Hills coal fields in Colorado (fig. 4). The 40° 39° ' GREEN RIVER AND HAMS FORK REGIONS NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES— GREEN RIVER AND HAMS FORK REGIONS By NEVILLE G. GAGGIANI 85 regions are located in parts of the Middle Rocky Moun- tains, Southern Rocky Mountains, and Wyoming Basin physiographic provinces and the Uinta Basin section Idaho and Utah, include coal areas 52—54 of the Colorado Plateaus physiographic province (pl. 1). “K (2/ i \ ~ \\(/ \) “‘- /'\ . _ \W (>4! Base from U. S. Geological I joo MILES Survey 112,500,000 United States base map 0 so 100 KILOMETERS EXPLANATION COAL REGION 52 COAL—AREA BOUNDARY—Number refers to local areas (Cochran and others, 1983) .2 SAMPLING SITE AND NUMBER—See figures 42 and 43 for site descriptions FIGURE 41.—The Green River and Hams Fork regions, coal areas 52—54, and sampling sites. The major rivers draining the regions are the Green and North Platte Rivers (fig. 41). Major tributaries of the Green River are the Yampa and White Rivers and major tributaries of the North Platte River are the 86 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Laramie and Medicine Bow Rivers. Other smaller drainages in the regions include the Great Divide Basin, where there is no surface flow into or out of the basin, and the Bear River, which drains the western edge of the regions and flows into Idaho. The major rivers and many of the tributaries originate in granitic mountains and flow into and through sedimentary basins between mountain ranges. The regions are semiarid and have warm summers and cold winters. Parts of the regions vary from alpine in the higher altitudes of the mountain ranges to arid in parts of the Green River basin in Wyoming and the White River basin in Colorado and Utah. Precipitation in the regions is quite variable. In the high mountains, average annual precipitation is as much as 60 inches and occurs mainly as snow. In the semiarid basins, average annual precipitation ranges from less than 7 to more than 16 inches and occurs mainly as rain. Rocks of Precambrian age to Quaternary age are ex- posed in the Green River and Hams Fork regions. The tectonic activity of the Laramide orogeny during the Late Cretaceous Period consisted of major faulting, folding, uplift, and subsidence, which formed the basic structure of the synclinal basins and anticlinal moun- tains. It was the last major episode of tectonic activity in the regions. Erosion by water, Wind, and glaciation further shaped the physiographic features of the regions by exposing mountain peaks composed of rocks of Precambrian age and filling in basins with alluvial and aeolian deposits of Quaternary age. Bedrock in most areas of the regions consists of sedimentary deposits of Cretaceous and Tertiary age. Most of the coal deposits occur in these rocks. Outside of the sedimentary basins, where the mountains have been pushed up by tectonic activity, most of the bed- rock is crystalline rock of Precambrian age. Economic deposits of coal occur in units of the Cretaceous Mesaverde Group, the Lance and Medicine Bow For- mations, the Cretaceous and Tertiary Evanston and Ferris Formations, and the Tertiary Fort Union, Wasatch, Hanna, and Coalmont Formations. The major land use in the regions is livestock graz- ing. Other land uses are timber harvesting, farming, recreation, mineral mining, and urban use. More than 60 percent of the land in the regions is rangeland that is either privately owned or owned by the Federal or State Government. The majority of cropland is used to grow hay for livestock or to improve grazing on pasture- land. Where possible, cropland is in a low area near a river so that the land can be flood-irrigated by canals. Irrigated and nonirrigated cropland compose less than 10 percent of the land area. Timberland (forested land) compose less than 10 percent of the land west of the Continental Divide (in the Rocky Mountains) and about 30 percent of the land area east of the Continental Divide. Fourteen percent of the land in the regions is forested; generally these areas are located on the slopes of the mountain ranges. - The regions have a small population density, and there are few urban areas. According to the 1980 cen- sus, the regions have less than 262,000 people. The two major urban areas in the regions are Laramie in south- eastern Wyoming (population 24,410) and Rock Springs (population 19,485) in southwestern Wyoming (Kuhn and others, 1983; Lowham and others, 1985). The major water use in the regions is irrigation of crops. Both surface- and ground-water sources are used, but more surface water is used than ground water (Lowham and others, 1985). In the Green River basin, the Bear River basin, and the Great Divide Basin of Wyoming, more than 80 percent of the water is used for irrigation. Other uses include residential and indus- trial, public water supplies, and stock watering. COAL RESOURCES Coal-bearing rock formations underlie most areas of the two regions; the largest continuous area underlain by coal deposits is in the Green River region. The ma- jor coal-bearing formations in the regions are in the Mesaverde Group, which contains coal beds 3 to 20 feet thick. Most of the coal occurs in the Williams Fork and Iles Formations of the Mesaverde Group. Other eco- nomic deposits of coal occur in the Lance Formation, which contains coal beds generally 0.5 to 10 feet thick, the Fort Union Formation, which contains coal beds that are as much as 40 feet thick, and the Wasatch For- mation, which contains coal beds that also are as much as 40 feet thick. Three coal fields are located outside the boundaries of the regions. The Hanna coal field (fig. 4), in Carbon County, Wyoming, is the largest of the three. Coal deposits occur in the Hanna, Ferris, and Medicine Bow Formations and in the Almond Formation of the Mesa- verde Group. The Hanna Formation has 32 coal beds that are more than 5 feet thick and 8 coal beds that are 20 to 40 feet thick. The Ferris Formation has 28 mine- able coal beds that generally are 5 to 10 feet thick and 3 coal beds that are 25 to 40 feet thick. The Medicine Bow Formation has three coal beds of economic impor- tance; the thickest of these is 9 feet. The Almond For- mation has coal beds 5 to 10 feet thick that are suitable for surface mining. Little is known about coal beds in the Rock Creek coal field (fig. 4), which is located in Albany and Carbon Counties, Wyoming. A few coal beds occur in the Pine Ridge Sandstone of the Mesaverde Group and in the Hanna Formation. The North Park coal field (fig. 4) in GREEN RIVER AND HAMS FORK REGIONS 87 Jackson County, Colorado, has coal deposits in the Coal- mont Formation. Three coal beds of economic impor- tance, ranging in thickness from 3 to 80 feet, occur in this formation. Coal in the regions is of moderate heating value and has comparatively large moisture content but has small concentrations of sulfur. The coal ranges from sub- bituminous to high volatile B and C bituminous (fig. 4). The strippable coal reserves in the regions have been estimated to be about 10.5 billion tons. Total coal production from the 14 coal fields in the Green River and Hams Fork regions is reported to be more than 660 million tons. Most of this production was low-sulfur subbituminous coal and was from surface and underground mines. During 1983 in coal area 52, there were 7 active surface mines and no active underground mines; underground mining ceased during 1982. Dur- ing 1980 in coal areas 53 and 54, there were 10 active surface mines and 8 active underground mines, but by 1983 there were only 5 active surface mines and no ac- tive underground mines. HYDROLOGY Two general objectives for the coal-hydrology pro- gram in the regions were: (1) To increase knowledge about the availability of water; and (2) to provide infor- mation for assessing the effects of coal development on water resources on a regional basis and on a site-specific basis. . At present (1985), all the coal mined in the Green River and Hams Fork regions is from surface mines. Potential issues using this method of mining are related to the changes in the quantity and quality of surface runoff caused by changes in soil infiltration rates and vegetative cover. After surface mining ceases, reveg- etation is difficult in arid-to-semiarid areas. Also, changes to the natural landscape will affect the hydrol- ogy of an area in some way. The extent of that effect will vary from site to site depending on the following factors: . Mining and reclamation methods; . Slope of land being mined; . Types of soil and rock; . Quantity of precipitation; . Quality of ground water and surface water; and . Rate of water movement. As part of the coal-hydrology program, the hydrologic- data-collection network was expanded or otherwise enhanced, and interpretive hydrologic studies were undertaken. This twofold approach was to indicate existing and potential issues discussed above and to evaluate data requirements to address the factors af- fecting potential hydrologic effects from coal mining. @mACDNr—t SURFACE—WATER NETWORK Surface-water and water-quality data are available for 951 sites, of which 195 sites are presently active (1981 for coal area 54; 1982 for coal areas 52 and 53). The earliest records for measured discharge were for the Laramie River during 1890 at Woods Landing in Laramie County near the Colorado-Wyoming State line and for the Green River during 1891 at Green River, Wyoming. The Laramie River measurement probably was related to water rights for irrigated cropland (Lowham and others, 1985). The historical surface-water network includes 391 continuous-record stream-discharge sites, 57 water- quality sites, 130 sites where discharge and water- quality data were collected, 57 peak-discharge (crest- stage) sites, and 316 miscellaneous sites where one or more discharge measurements or water-quality samples were made as part of interpretive studies. The number of sites added since the coal-hydrology program began during 1974 are 119 continuous-record stream-discharge sites and 185 water-quality sites. Water-quality data collected at the sites include dissolved-solids concentrations, specific conductance, alkalinity, pH, and concentrations of sulfate and selected dissolved trace elements. Collection intervals varied from only one sample at some miscellaneous sites to monthly or daily samples at some water-quality sites. Generally, the ephemeral streams in the arid and semi- arid basins of the regions where coal mines usually are located do not have long periods of record. Suspended-sediment data are available for 161 sites in the Green River and Hams Fork regions, of which 83 presently are active (1981 for coal area 54; 1982 for coal areas 52 and 53). The longest period of record and earliest record of a suspended-sediment measurement are from a site located on the Green River near the town of Green River, Wyoming, that has records from 1951 to the present (1985). Many of the sites have only 1 year of record. The frequency of collection varies; most sites have weekly, monthly, or quarterly records. Some sites have‘daily records where samples are obtained either by using automatic pumping samplers or by having a local observer collect the samples. SURFACE—WATER CHARACTERISTICS Most of the discharge in the regions occurs in streams draining mountainous areas and originates from snow- melt. Most of the annual discharge occurs during spring and early summer. The exact time of the snowmelt varies within the regions and is controlled by temper- ature, altitude, slope, aspect (the direction the slope is facing—if the slope is facing north, the melting will be 88 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974-84 the slowest), and vegetative cover. At the White River downstream from Meeker, Colorado, 60 percent of the discharge occurs during May, June, and July, whereas at the Yampa River near Maybell, northwestern Colo- rado, 70 percent of the annual discharge occurs during April, May, and June. During late summer, fall, and winter, flows primarily are the result of ground-water inflows. Minimum discharges occur from January through March. Intermittent and ephemeral streams that originate in and drain the arid to semiarid basins have extended periods of no flow. Because flow for these streams is supplied by summer rainstorms, ground-water dis- charge, and springs, there is not enough water to sus- tain flow throughout the year (fig. 42). Estimates of average annual discharge and high flows at selected frequencies can be made at ungaged streams using relations developed at gaged streams. Average annual discharge can be estimated at ungaged streams using predictive equations developed for Colorado (Liv- ingston, 1970) and Wyoming (Lowham, 1976). High-flow and floodflow frequencies and magnitude can be estimated using equations developed for Colorado, Wyoming, Utah, and Idaho for the respective areas in the regions. Floodflows usually are caused by summer rainstorms or rain occurring on snow in the spring. Flows in the small ephemeral streams, where the coal generally is mined, are the most difficult to estimate. Craig and Rankl (1978) developed relations especially for these types of streams in Wyoming. Similar rela- tions also are presently being developed for streams in Colorado (Livingston and Minges, 1987). Low-flow fre- quencies are not predictable in the regions because perennial streams are affected by regulation and diver- sion and ephemeral streams are dry for most days each year. SURFACE—WATER QUALITY Results of chemical analyses of water samples col- lected throughout the regions indicate that the quality of water in the mountainous streams is good but degrades to fair or poor in many of the plains streams (Lowham and others, 1985). Lakes and reservoirs in the regions have had occasional algal blooms. Data collected from streams, lakes, and reservoirs include water temperature, specific conductance, pH, alkalinity, and concentrations of dissolved solids, sulfate, phosphorus, and trace elements, such as arsenic, boron, cadmium, chromium, copper, iron, lead, manganese, mercury, selenium, and zinc. Biological and bacteria samples also were collected. Dissolved-solids concentrations generally were small (less than 500 milligrams per liter) in streams in and 6°° l‘l”'l"l I I I I I II Snowmelt from high altitudes 500 — and from occasional rainstorm ._ runoff 400 — — 300 —- __ Snowmelt from 200 — low altitudes _ Low flows \ are largely due 100 — to ground-water D Z O U u.I (O n: u.I D. l- u.I ”J U. 2 m a inflows\ g 0 I I I ’ I I I I I I I I g5°IIIIIIIIIII n: — Site 2. — E 4 (Drainage is mostly __ U 0 — in the plains) ‘13 _ D Rainstorm runoff\ 2 3o — __ :5 E — Snowmelt runo(f\ - 0’ 20 — __ E - _ 3 1° _ No flow occurs — __ during much _ of the year i o I I I I I I I I i- > U 2 m I E >- I.u >— (g .— U o in < Iu < a. < z _I 3 n. °z°a“2<222 U 2 to u: :1: Lu >. 0 l- o z D -: “L E < 2 a 2 < 33‘ WATER YEAR FIGURE 55.—Simulated dissolved-solids loading to the Yellowstone River from the Tongue and Powder Rivers, Montana, 1978 water year (modified from Knapton and Ferreira, 1980). Equation 4 becomes linear (as in eq. 1) when log trans- formations are made on variables C and Q. Equation 4 assumes a constant year-round relation; however, there are many sites that do not follow this assumption. Seasonal shifts in the relation between esti- mated and measured dissolved-solids concentration are exemplified by dissolved-solids residuals from a site in the Green River basin of Wyoming, as shown in figure 56 (DeLong, 1977). In this example, the positive residuals during one period are balanced against nega- tive residuals during another period. To account for their seasonal effects, DeLong (1977) incorporated a time variable into coefficients A and B in equation 4 using the following functions: Loglo A = Bo‘+ Blsin(at) + 32 cos(at), and (5) where a = 0.987 degree per day or 0.0172 ra- dian per day; t = day of the water year; and Bo through B5 = regression coefficients. Regression coefficients Bo through B5 are determined by a multiple-variable-regression technique referenced by DeLong (1977). The improved residual plot for the Green River at Warren Bridge near Daniel, Wyoming, is shown in figure 57. 128 100 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974-84 a: o I 60—— 40— 20 — o 0' . .20 _. .40 _ ~60 _ DIFFERENCE BETWEEN ESTIMATED AND MEASURED DISSOLVED-SOLIDS CONCENTRATIONS, IN PERCENT O .80 _ m I I I I | l | I l | | OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT WATER YEAR FIGURE 56.—Difference between dissolved-solids concentrations estimated from the two-variableregression model and dissolved-solids concentrations measured at site 09188500 Green River at Warren Bridge near Daniel, Wyoming (modified from DeLong, 1977). Equations 4, 5, and 6, in addition to equation 3, have been incorporated into a computer program to produce dissolved-solids-load hydrographs for a period of years, as shown for site 09188500 Green River at Warren Bridge near Daniel, Wyoming (fig. 58). Loads estimated at several sites in a stream can provide quantitative in- formation about the quantity of dissolved solids gained in intervening reaches. By evaluating the chemical com- position of dissolved-solids gain, the source of dissolved solids can be delineated (DeLong, 1977). DESCRIPTION AND APPLICATION OF MATHEMATICAL ACCOUNTING MODELS Regulatory agencies often need to assess the cumula- tive effects of several coal mines operating in several different tributaries that eventually flow into a main- stem stream. Parker and Norris (1983) developed a model to assess cumulative effects of mining on Trout Creek drainage and a reach of the Yampa River in Col- orado. The model consists of a series of nodes on the stream network that are used to sum water quantity and quality throughout the system. Various mining plans can be inserted into the model to compare the dif- ferent cumulative effects at downstream sites. The algorithm for the model is an accounting pro- cedure that sums water quantity and quality in monthly time steps from one or more upstream nodes to a down- stream node. The model has input nodes, internal nodes, and output nodes (fig. 59). The input nodes (nodes 1, 2, and 3) are the most upstream nodes where the sum- mation process starts. Internal nodes (nodes 4, 5, and 6) provide for input changes to the system. These changes can be point sources of water from dewatering activities or diffused sources of water such as drainage from a coal-spoil pile within the reach upstream from the node. An output node is any node at which there is a need to determine the model estimates through time and to examine differences in these estimates with various anticipated mining activities. The most down- stream node (node 6 in fig. 59) usually would be an out- put node. If the cumulative effects of coal mining in the area upstream from nodes 4 or 5 (fig. 59) are of interest, nodes 4 or 5 also could be output nodes. SALINITY MODELING 129 100 I 80 — _ 60 -— _ 40 — — 20- — DIFFERENCE BETWEEN ESTIMATED AND MEASURED DISSOLVED-SOLIDS CONCENTRATIONS, IN PERCENT D j; I I I I I I I I I I I 400 OCT NOV DEc JAN FEB MAR APR MAY JUNE JULY AUG SEPT WATER YEAR FIGURE 57.—Difference between dissolved-solids concentrations estimated from the multiple-variable-regression model and dissolved- solids concentrations measured at site 09188500 Green River at Warren Bridge near Daniel. Wyoming (modified from DeLong, 1977). At any node. the surface-water-quantity component, where Q, = incremental stream discharge (increase or which is mean monthly stream discharge in cubic feet decrease) within the reach; per second, is calculated by the equation: a and b = the regression coefficients from simple linear regression; and Q8 = stream discharge at some nearby + Qr’ I7) streamflow-gaging sites. n EQu u=1 Qt: where Qt = stream discharge at node i; Qu = stream discharge at adjacent nodes im- mediately upstream from node i; n = number of adjacent nodes immediately upstream from node i; and Q, = incremental stream discharge (increase or decrease) within the reach between node i and adjacent nodes immediately 2'000 I I I l I l I I I Load estimated from the 1,000 — multiple-variable-regression model\ _ DISSOLVED-SOLIDS LOAD, IN TONS PER DAY upstream. o 1 . I 1 1970 1971 1972 1973 1974 1975 The estimate of incremental stream discharge within WATER YEAR the reach can be obtained by reading the data or by esti- mating the data by the equation: FIGURE 58.—-Estimated monthly mean dissolved-solids loads at site 09188500 Green River at Warren Bridge near Daniel, Wyoming Q, = a + st, (8) (modified from DeLong. 1977). 130 2 INPUT MODE 3 INPUT MODE 1 INPUT NODE 04 INTERNAL NODE, OUTPUT NODE, OR BOTH ’5 INTERNAL NODE, OUTPUT NODE, OR BOTH p6 INTERNAL NODE, OUTPUT NODE, OR BOTH W FIGURE 59.——Modified diagram of a simple stream network with nodes and node numbers for the model developed by Parker and Norris (1983). In the model, several stream reaches have an upstream and a downstream node at a streamflow- gaging site. In these situations, Q, could be measured directly, and measured stream-discharge data were used. In situations where measured data were not avail- able, Q, initially was established at zero and modified by altering the regression coefficients in equation 8 dur- ing calibration. For each anticipated mining activity, the Mined Land Reclamation Division of the Colorado Department of Natural Resources estimated the quan- tity of water discharging to the stream. At each node the surface-water-quality component, mean monthly dissolved-solids concentration in milli- grams per liter, is calculated by the mass-balance equation: (2 QuCu) + QrCr C- = L , (9) (ElQu) + Q]- where C,- = dissolved-solids concentration at node i; n = number of nodes immediately upstream from node i; Q, = stream discharge at nodes immediately upstream from node i; SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Cu = dissolved-solids concentration at nodes immediately upstream from node i; Q, = incremental discharge; and C, = dissolved-solids concentration associated with the incremental discharge within reach. The dissolved-solids concentration within the reach (0,) is obtained from the linear-regression equation: C, = (log. (10) Initial estimates of C, are obtained from measured data at each node. For input nodes, the measured data indicate the actual value of C, because C, is the in- tegrated dissolved-solids concentration for the total reach upstream from that node. However, for internal and output nodes, measured data do not indicate dissolved-solids concentration for the reach between nodes; they indicate an integration of dissolved-solids concentration upstream from each node. Thus, the measured data are not a direct estimate of 0,. Final estimates of the regression coefficients in equation 10 were obtained in the calibration process. For each anticipated mining activity, the Mined Land Reclamation Division estimated dissolved-solids con- centration for water discharging to the stream from the mining activity. During mining operations, the concen- tration of dissolved solids was 2,860 milligrams per liter for combined surface- and ground-water discharge. For postmining situations, no discharge from a surface- water source was assumed, and the dissolved-solids con- centration was estimated at 3,200 milligrams per liter from a ground-water source only. There can be instances when the stream discharge at the upstream node or nodes is greater than the stream discharge at the next node downstream. The dissolved- solids concentration can be decreased in proportion to the water quantity lost, assuming that water lost in the reach is lost to ground water and, therefore, that the water lost removes the associated dissolved solids. However, some of the dissolved solids assumed lost to ground water may remain on the bed and banks of the stream channel to be removed during the next high flow. In addition, water lost to evapotranspiration leaves the associated dissolved solids in the streamflow. To accom- modate these problems, a calibration factor was added to increase the dissolved-solids concentration. This factor was adjusted during the model calibration. Thus, in a losing reach, C, is decreased to the minimum value and adjusted upward by: Qi Ci=Cl n l E., 11 2Q, l ( I u=1 where E,- = calibration coefficient 21.0. SALINITY MODELING Stream-discharge data that are input at the various input nodes are obtained either from continuous stream- discharge data or through various regression relations using nearby sites that have continuous stream- discharge data. For water-quality data, linear-regression equations were obtained between the logarithm of in- stantaneous stream discharge and the logarithm of dissolved-solids concentration. These equations were in- put directly into the model for each input node. For each output node that has mean monthly stream- discharge data (either measured or extrapolated), a linear-regression equation between the logarithm of in- stantaneous stream discharge and the logarithm of dissolved-solids concentration, in milligrams per liter, was obtained from data available, at the sites. Using these equations, a dissolved-solids concentration was obtained for each mean monthly stream discharge. Calculation of the load of dissolved solids (tons per month) was obtained from: L = Q-C-K-Nm, (12) where L = dissolved-solids load, in tons per month; Q = mean monthly stream discharge, in cubic feet per second; C = dissolved-solids concentration, in milli- grams per liter at the mean monthly stream discharge; K = 0.0027, a conversion constant; and Nm = number of days in the month. The calculated values for the period of record are used as the measured values and are compared to modeled values for calibration and error analysis. Calibration of the model was made so that modeled outputs of stream discharge, dissolved-solids concen- tration, and dissolved-solids load closely matched measured data at the output nodes. Altered model parameters were the regression coefficients in equations 8 and 10 and the coefficient E, in equation 11. During calibration, an attempt was made to decrease the mean square error for each variable throughout the total 72 months the model was run (Parker and Norris, 1983). The error function uses the differences between the logarithms of measured and predicted values. The mean square error is: MSE = 52 + $2, (13) where MSE = mean square error; 3-52 = square of the mean of the differences between the logarithms (base e) of measured and model prediction for 131 each model variable for each month; and s2 = variance of the differences of the loga- rithms (base e) between the observed and model prediction for each model variable for each month. In this equation, the first term (In) is the bias from the true mean zero and the second term (s2) is the variance. During calibration, the attempt is made to decrease the bias to zero with a minimum variance (Parker and Norris, 1983). The calibrated model used streamflow quantity and quality in Trout Creek and the Yampa River from Oc- tober 1975 to September 1981.Calibrated model output was compared to output from the model that had been perturbed by adding increased stream discharge or dif- ferent dissolved-solids concentrations resulting from an- ticipated mining. Estimates of actual discharge of water through a mine to the receiving stream and the associ- ated dissolved-solids concentration were provided by the Mined Land Reclamation Division of the Colorado Department of Natural Resources. The model was run for a series of anticipated mining activities during which short-term and long-term effects were studied. Short-term effects include surface— and ground-water effects, such as discharge from sediment ponds, discharge from underground mine workings, and discharge of affected waters from shallow ground-water systems that would occur during the mining operation and for a short time following reclamation. The natural flow patterns of the effected ground-water systems are disrupted by mining, and surface and ground waters are mixed. Increased evaporation losses from the sediment ponds are assumed to be offset by increased runoff from disturbed areas. The long-term effects of mining occur after: (1) Dis- turbed areas have been reclaimed successfully; (2) surface- and ground-water systems have had sufficient time to reach equilibrium; (3) sediment-control struc- tures have been removed, and the quantity and qual- ity of runoff from the reclaimed areas have returned to premining conditions; and (4) mine-spoils aquifers and underground mine workings have resaturated, and ground water passing through the disturbed area dis- charges to its premining discharge areas. The quantity of ground-water flows would equal premining quan- tities, but the quality would be degraded. A comparison of modeled dissolved-solids concentra- tion for existing conditions and for conditions asso- ciated with short-term anticipated mining at Middle Creek at mouth (node 15), is shown in figure 60. In the short-term anticipated mining, the greatest change in model variables was indicated for Middle Creek at 132 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 1'6°° I I I I I EXISTING CONDITIONS 1,400 ———— ANTICIPATED MINING 1 ,200 1,000 800 600 400 200 MEAN MONTHLY DISSOLVED-SOLIDS CONCENTRATION, IN MILLIGRAMS PER LITER | | | I I I I o I I I I I I 20 25 30 35 7 NUMBER OF MONTHS 40 45 50 55 60 65 70 75 FIGURE 60.—A comparison of modeled mean monthly dissolved-solids concentrations for existing conditions and conditions associated with short-term anticipated mining at Middle Creek at month, node 15 (modified from Parker and Norris, 1983). mouth (node 15) and Fish Creek (node 19). The mean stream discharge is increased by 31 percent at node 15 and decreased by 1 percent at node 19. This primarily is an indication of anticipated dewatering activities directly upstream. The mean monthly dissolved-solids concentration increases by 316 milligrams per liter at node 15 and by 98 milligrams per liter at node 19. The total monthly load of dissolved solids increases to vary- ing degrees at all output nodes. A long-term version of this anticipated mining indicates similar increases in dissolved-solids-concentration values. DESCRIPTION AND APPLICATION OF MATHEMATICAL ROUTING MODELS A computer model was developed for determining dissolved-solids load in the Tongue River in Montana, to evaluate cumulative potential effects of strip min- ing on dissolved-solids concentration at any given number of locations in a stream (Woods, 1981). This model then was modified for use on Rosebud Creek in Montana (Ferreira, 1984). Located within the drainages of both streams are irrigated agricultural areas and several areas of Federal coal that potentially are avail- able for leasing. The model consists of a monthly mass- balance routing of stream discharge and dissolved- solids load down the main stem of each stream and can account for dissolved-solids load coming from each mine area. The main stem of each stream is divided into several reaches. Once the locations of each reach are established as part of the model, the model user can vary the mined acreage, dissolved-solids concentrations in mine spoils, and quantity of irrigated acreage in each reach. The user then can study relative changes SALINITY MODELING in the dissolved-solids concentration from mining and agriculture as they affect the water quality in each reach. All hydraulic components are accounted for in the model in an effort to provide versatility. The mass balance of discharge between the upstream and downstream ends of each reach is computed by the equation: QOUT = QIN + QP ‘ QE ‘ QET + QGW QT ' QSI + QRI ‘ 901 + QIRF ' QOL, (14) where all units are in acre-feet per month, and QOUT = stream discharge at downstream end of reach; Q IN = stream discharge at upstream end of reach; Q P = precipitation received on stream surface; QE = evaporation loss from stream surface; Q ET = evapotranspiration from riparian vegetation; QGW = ground-water inflow or outflow; QT = discharge from tributaries; QSI = volume of stream discharge stored as ice; QR I = volume of stream discharge released from me; Q D I = volume of stream discharge diverted for irri- gation; Q IR F = volume of irrigation return flow; and QOL = volume of other water losses. The mass balance of dissolved solids between the upstream and downstream ends of each reach is com- puted by the equation: + DSLIRF + DSLM — DSLOL’ where all units are in tons per month, and DSLOUT= dissolved-solids load at downstream end of reach; DSL IN = dissolved-solids load at upstream end of reach; DSLGW = dissolved-solids load in ground-water inflow or outflow; DSLT = dissolved-solids load input by tributary streams; DSL D I = dissolved-solids load diverted by irrigation flow; DSL IR F = dissolved-solids load returned by irrigation flow; DSL M = dissolved-solids load input by mining; and DSLOL = dissolved-solids load removed with other water losses. The dissolved-solids concentration at the downstream 133 end of the reach is calculated using the following equation: -2811, _ Q'f where DSC = dissolved-solids concentration, in milli- grams per liter; DSL = dissolved-solids load, in tons per month; Q = stream discharge, in acre-feet per month; and f = a factor (0.00136) that converts the prod- uct of acre-feet and milligrams per liter to tons. DSC (16) Other equations and factors used to calculate values for variables contained in equations 14, 15, and 16 are obtained from climatological data, ground-water data, continuous stream-discharge measurements, base-flow measurements, and channel-geometry data. Many of the same regression equations mentioned previously are used to calculate hydrologic variables for each reach. Some of these equations are incorporated in the model; others are used to calculate constant values used in the model as block data. For Rosebud Creek, initial stream discharge and dissolved-solids concentrations are input at the down- stream end of reach 1 (fig. 61). These values are affected directly by input of dissolved solids from mining and water losses from irrigation if acreage involved in these two activities is larger than what presently exists in the drainage of reach 1. The resulting values at the down- stream end of reach 1 then are used as input for the upstream end of reach 2. Within reach 2 and each successive reach, gains and losses to stream discharge and dissolved-solids load are accounted for algebraically. The model step is monthly, and each simulation is for one calendar year (fig. 62). In the model, monthly traveltime of stream discharge and dissolved-solids concentration within each reach and from the headwaters to the mouth are instantaneous. Simulated monthly streamflow and dissolved-solids load generally were within the 95-percent confidence limits of the mean monthly values calculated for Rosebud Creek at the mouth near Rosebud, Montana. From May through September, the simulated mean monthly streamflows varied by no more than 15 per- cent of the historical mean values. Except for January, May, and December, the simulated mean monthly dissolved-solids loads varied by no more than 13 per- cent of the historical mean values. Simulations based on mining that occurred during 1983 indicated that irrigation return flows composed 134 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 107° 106° I I 46° * r> . / a? NORTHERN Q . 90"“; O C caow m?» CHEYENNE & ’% Ago/v INDIAN MONTANA Map area RESERVATION ongue River Reservoir Deckero 45° ———___M_(MFANA — _ _____J WYOMING (I) 1'0 2'0 3‘0 MILES (I) 1'0 2'0 3'0 KILOMETERS 06295113 EXPLANATION COMBINATION STREAM-DISCHARGE AND WATER-QUALITY STATION AND NUMBER —- u — BASIN BOUNDARY FIGURE 61.—Location of Rosebud Creek, Montana, and reaches simulated by the model developed by Woods (1981) (modified from Ferreira. 1984). SALINITY MODELING Reads data by user: Initial dissolved- solids concentration in upstream reach, area mined, dissolved-solids concentration of mine leachata, and area irrigated. READ DATA Writes values input by user and initial conditions in upstream reach. Sets program to first month and advances to each successive month until twelfth month. YES J=1 > Sets program to first reach and J: 5 advances to each successive reach J =J+1 > N0 until filth reach. YES Sums the gains and losses of flow SUM FLOW in each reach. Sums the gain and losses of dissolved- SUM 05" solids load (DSL), in tons, in each reach. SUM FLOW Calculates the dissolved-solids concen- —— =' DSC tration (DSC), in milligrams per liter, SUM DSL in each reach. WRITE Prints streamllow, dissolved-solids load, RESULTS and dissolved-solids concentration for downstream and of each reach. —'—‘—\1/ CALCULATE Calculates monthly mean and standard STATISTICS deviation, minimum, maximum, and percentage values. Prints summary of results. WHITE SUMMARY ( STOP ) FIGURE 62.—Simplified flow chart of model for calculating monthly dissolved-solids concentration in five reaches of Rosebud Creek. Montana (modified from Ferreira, 1984). 135 a larger cumulative percentage of dissolved-solids con- centration (about 3 percent in reach 5) than mining (about 0.4 percent in reach 5). However, when all areas are mined simultaneously, the cumulative per- centage resulting from irrigation in reach 5 (2.5 percent) would be smaller than that resulting from mining (14.7 percent). Using the Tongue River model, Woods (1981) re- ported that the simulated mean annual dissolved-solids concentrations would increase by 4.8 percent if all potentially available, federally owned coal were mined. Even with this increase, the dissolved-solids concentra- tion in the Tongue River still would be suitable for irri- gation supply (Woods, 1981). DISCUSSION All models could be improved with more data. How- ever, budget and time constraints would make some ad- ditional data collection impractical. For single-equation models, additional data that indicate extreme hydro- logic events or additional variables such as the time factor used by DeLong (1977) could be beneficial in evaluating the effects of mining. For the more complex models, better estimates of mine-derived dissolved- solids load added to the stream, and more accurate delineation of reaches receiving this load would improve the prediction capabilities of the model. Presently (1985), dissolved-solids loads from mined areas are estimated from water-quality changes meas- ured in mine spoils. Mine-spoil water-quality changes do not account for chemical changes that could occur enroute from the mine to the stream. The stream reaches generally defined as those receiving ground- water flow from a mined area are located on a straight line downgradient from the mine. However, because ground water could be moving downgradient through the alluvium but parallel to the streamflow for some distance before entering the stream, the actual reaches receiving the dissolved-solids load could be different from those specified in each model. Depending on where different mines are located, actual flow paths of mine- spoil water could make a considerable difference when predicting accumulated effects of mining in a given area. 136 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 SHALLOW AND DEEP GROUND-WATER FLOW SYSTEMS By MICHAEL R. CANNON The ground-water flow system within a drainage basin is composed of a continuous series of ground-water flow paths that originate in areas of ground-water recharge and terminate in areas of discharge. Ground-water flow systems largely are controlled by the topography of the basin and the hydraulic conductivity of the porous soils and rocks through which ground water moves. The to- pography of the basin establishes the potential energy available to a unit mass of water as it moves from a recharge area in the higher part of the basin to a dis- charge area in the lower part. The hydraulic conductiv- ity establishes the rate at which a volume of ground water will move through an area of porous soils and rock when affected by the hydraulic gradient (gradient of potential energy). A typical ground—water basin may contain several aquifers and confining beds and ground-water flow paths of many lengths (fig. 63). 'Ibth (1963) suggested that ground-water flow systems can be classified into three types: (1) A local flow system, which has its recharge area at a topographic high and its discharge area at an adjacent topographic low; (2) an intermediate flow system, which is characterized by one or more topo- graphic highs and lows located between its recharge and discharge areas; and (3) a regional flow system, which has its recharge area at the major topographic high and its discharge area at the bottom of the basin. The subdivision of ground-water flow into local, in- termediate, and regional flow systems is somewhat ar- bitrary. The size of an area studied often determines how a hydrologist will categorize the flow systems observed in the study area. For instance, in a study of the entire Mississippi River basin, a certain flow system FR”), fl: __ .\ M GROUNDWATER FLOW SYSTEM / \ . _, // Flow lines \\ ’ / \_——_——_—__—__— may be designated as a local system, whereas in a study of a small tributary basin to the Mississippi River, the same ground-water flow system may be designated as an intermediate system. For the purposes of this report, local flow systems are those that actively circulate ground water and rapidly respond to changes in rates of recharge and discharge. Local flow systems have relatively short traveltimes, transport a large part of the ground water within a basin, and usually provide most of the base flow to small streams and rivers. Because of their shallow depth and active circulation of ground water, local flow systems are most easily affected by coal-mining activities. Intermediate flow systems have relatively slow rates of flow and long flow paths but respond faster than regional flow systems to long-term changes in recharge or discharge. Discharge from intermediate flow systems usually is to the medium-sized and largest streams in the basin. Coal-mining activities, such as mine-pit dewatering, can affect intermediate flow systems by decreasing available recharge or increasing discharge from aquifers near the mine pit. Decreases in water levels in an aquifer containing intermediate flow systems generally are small but can have widespread and long- term effects. Regional flow systems have slow, deep circulation of ground water, have the longest flow paths in the basin, and react slowly to changes in recharge and discharge. Regional flow systems generally contain a vast volume of water in storage but discharge this water slowly as base flow to large rivers or to coastal areas. Water in regional flow systems commonly is more mineralized than water in shallow, local flow systems because of the ischarge EXPLANATION FLOW SYSTEM Intermediate Regional E MILLENNIA .— Confining bed :‘_.—_-'.‘_.—._—_.—.—.-_—_*_—‘_"_—.—_—_—_—_—.._—_—_—_-_—._-_—_—_: FIGURE 63.—Ground-water flow system of a typical drainage basin (modified from Heath, 1983). SHALLOW AND DEEP GROUND—WATER FLOW SYSTEMS long residence time and long flow paths of the regional flow system. Because of their great depth of flow, large volume of water in storage, and slow rate of flow, ef- fects of coal mining on regional flow systems generally are not severe or are not observed unless the flow system is stressed for a long period of time. STUDIES OF EFFECTS OF MINING ON GROUND—WATER FLOW SYSTEMS The US. Geological Survey, as part of its coal- hydrology program, has invested considerable time and Northern Great Plains and Rocky Mountain Provinces Pacific Coast Province Interior Province 137 effort in studying the effects of the surface mining of coal on ground-water systems. Studies have been made in almost all States Where coal is mined by surface methods. Studies range from assessments of basic ef- fects such as aquifer dewatering and water-level declines in wells, to detailed studies of the hydraulic properties of mined areas and observed changes in local and re- gional flow systems. To illustrate how the surface min- ing of coal can affect the ground-water hydrology of a basin, mine-site studies made by the US. Geological Survey are presented for three different coal provinces (fig. 64). EXPLANATION 4. MINE-SITE STUDY AREA AND SITE NUMBER 1 Jefferson County, Ohio 2 Muskingum County, Ohio 3 Macon-Huntsville area, Missouri 4 West Decker, Montana FIGURE 64.—Coa.l provinces and location of selected mine-site study areas. 138 EASTERN PROVINCE, NORTHERN APPALACHIAN REGION— EASTERN OHIO Several small watersheds in eastern Ohio were studied to assess the effects of the surface mining of coal on the hydrologic systems (fig. 64) (Helgesen and Razem, 1981; Razem, 1983, 1984; Weiss and Razem, 1984). One of the study sites was a 29-acre watershed in Jefferson County that was drained by a perennial stream. The premining watershed was underlain by nearly flat-lying interbedded shale, sandstone, limestone, and coal of the Pennsylvanian and Permian Systems. Shaly underclay beds below two major coal seams formed bases for two perched aquifers (fig. 65). The perched aquifers repre- sented local flow systems. Recharge to the upper perched aquifer was from local precipitation, and discharge was to springflow, seepage, and evapotranspiration near the coal outcrop. Addi- tional discharge from the upper perched aquifer was to leakage through the underclay. Recharge to the lower perched aquifer was from leakage from the overlying perched aquifer and from direct infiltration of precipita- tion where the upper clay layer was absent. Water in the lower perched aquifer moved from the recharge area near the basin divide toward the discharge area at the mouth of the basin. Additional discharge was to leakage through the clay at the base of the aquifer where it A DUNKARD GROUP Underclay ,, Middle aquifer W SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 eventually was drained through the mined Pittsburgh No. 8 coal bed. During mining of the upper coal bed (Waynesburg N o. 11 coal bed in fig. 65), water levels in the two perched aquifers declined abruptly. Springflow from the top aquifer, which supplied base flow to the stream, de- creased and eventually ceased completely (Razem, 1984). By the time mining was completed, most of the upper perched aquifer had been removed and replaced by mine spoils that were regraded to approximate the premining topography. Wells installed in the mine spoils initially were dry, but most indicated a saturated thickness of 3 to 4 feet after 1 1/2 years. In the part of the upper perched aquifer not removed by mining, water levels did not recover to premining levels because of improved drainage of the aquifer toward the mined area. In the second perched aquifer, which was not mined, water levels were as much as 40 feet higher than premining levels. The higher water levels were the result of the removal or disturb- ance of the clay layer during mining and the resultant increased leakage from the mine spoils to the lower aquifer. Surface mining of the upper coal bed had several ef- fects on the local ground-water flow. The rate of infiltra- tion of precipitation into the mine spoils seemed to be less than infiltration rates to the undisturbed basin. The EXPLANATION Deep aquifer (dry) F“— MONONGAHELA FORMATION —--> DIRECTION OF WATER MOVEMENT UNSATURATED ZONE SATURATED ZONE FIGURE 65.—Schematic section of ground-water occurrence and flow at a watershed in Jefferson County, eastern Ohio (from Razem, 1984). SHALLOW AND DEEP GROUND-WATER FLOW SYSTEMS decrease was caused by the destruction of the soil struc- ture and compaction of the sofl overlying the mine spoils. Discharge from the upper aquifer decreased in volume, and the upper aquifer no longer discharges to springs or seeps. Discharge from the upper aquifer (now a spoils aquifer) occurs as downward leakage into the lower perched aquifer. The mine spoils also had a larger hydraulic conductivity and storage than the premining bedrock aquifer. Similar geologic conditions and effects of mining on ground-water flow systems were observed at a 43-acre watershed in Muskingum County, Ohio (fig. 64) (Weiss and Razem, 1984). The premining watershed was char- acterized by nearly flat lying sedimentary rocks of the Pennsylvanian System. Underclay beneath the two major coal beds formed bases for perched ground-water flow systems, producing three separate aquifers under- lying the watershed. Mining in the basin removed the upper coal bed and the top perched aquifer and replaced the bedrock with spoil material. Water levels in the spoils are at a much lower altitude than existed in the premining aquifer because of a larger hydraulic conduc- tivity in the spoils, areal variations of the hydraulic characteristics of the confining bed, and a slower rate of recharge from precipitation caused by removal of 139 vegetation and compaction of topsoil. Recharge to the middle aquifer decreased by 25 percent because of decreased leakage from the overlying aquifer and less recharge from precipitation than in the premining basin. Discharge from both the upper and lower perched aquifers decreased following mining, and springs fed by discharge from the upper perched aquifer were trans- formed to a zone of seeps. INTERIOR PROVINCE, WESTERN REGION— MACON—HUNTSVILLE AREA, MISSOURI The Macon-Huntsville area of north-central Missouri (fig. 64) has been the site of extensive coal mining since the late 1800’s (Hall and Davis, 1986). Almost all coal produced there since the mid-1960’s has been mined by surface methods. Because of the hilly topography of the area, contour mining is the most commonly used min- ing method; overburden is removed in successive strips that follow the contour of the land surface. Spoils are placed in strips in the mined-out areas as the mining operation advances (fig. 66). Highwall lakes, which oc- cupy the last mine cut, are a characteristic feature of the mined area after reclamation or abandonment. FIGURE 66.-——Generalized hydrogeologic section through an abandoned surface mine. Macon-Huntsville area, Missouri (modified from Hall and Davis, 1986). 140 Shallow aquifers of the area occur within alluvium, glacial drift, and bedrock. The bedrock is of Pennsylva- nian age and is composed of shale, limestone, sandstone, and coal. Recharge to the shallow alluvial aquifer oc- curs by infiltration of precipitation and by lateral and probably vertical flow from adjacent aquifers. Recharge to glacial drift occurs by infiltration of precipitation and lateral flow from adjacent aquifers. Based on a ground- water flow model, a possible range of annual recharge to glacial drift is 0.2 to 5.2 inches. Recharge to bedrock aquifers probably occurs by infiltration of precipita- tion and vertical flow from glacial drift. The rate of recharge to bedrock is unknown but probably is small. Natural discharge from the shallow-bedrock and glacial-drift aquifers generally is to alluvium along the stream valleys; discharge from alluvium generally is to streams. Aquifers have developed in the mined areas where coal and overburden have been replaced with mine spoils consisting of a heterogeneous mixture of glacial drift and broken bedrock. Recharge to the spoils aquifer is from precipitation and lateral and vertical flow from adjacent aquifers, although precipitation probably is the major source of recharge. The spoils are fairly per- meable and generally more permeable than the alluvial, glacial-drift, and bedrock aquifers; the potential rate of infiltration ranges from 0.2 to 0.6 inch per hour. Long-term effects of surface mining on the shallow flow systems primarily result from the replacement of glacial drift and bedrock with more permeable spoils. The porosity of the spoils also is greater than the porosi- ty of the undisturbed overburden materials. The greater porosity and permeability have resulted in greater rates of recharge to mine-spoil aquifers than to nonmined aquifers of the area. NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES—WEST DECKER, MONTANA The West Decker Mine in Montana (fig. 64) produces more than 5 million tons of coal per year from the com- bined Anderson-Dietz 1 coal beds in the Tongue River Member of the Paleocene Fort Union Formation. The principal shallow coal aquifer in the area is the Anderson-Dietz 1 coal aquifer (Davis, 1984b). Other shallow aquifers are the Dietz 2 coal aquifer, clinker- and-alluvium aquifer, and mine-spoils aquifer (fig. 67). Important factors affecting the hydrologic regime of the West Decker Mine area are the semiarid climate that limits the available recharge, the very permeable clinker beds that provide maximum infiltration from the limited recharge, and the location of the mine near the Tongue River valley that is a major discharge area for ground-water flow systems. SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 The Anderson-Dietz 1 coal aquifer is about 50 feet thick in the West Decker area and is a source for stock and domestic water supplies. Recharge to the Anderson- Dietz 1 coal aquifer is by infiltration of precipitation, by upward leakage from the Dietz 2 coal bed, and by lateral flow from aquifers north and west of the mine site. A large part of recharge to the ground-water system is from infiltration of precipitation in areas of outcropping clinker (Davis, 1984a). Natural discharge from the Anderson-Dietz 1 coal aquifer primarily is to the Tongue River or to alluvium and clinker in the Tongue River valley (fig. 67). The West Decker Mine is located in the discharge area of an intermediate or regional flow system; some of the water discharging from the Anderson-Dietz 1 coal aquifer may originate as recharge in the topographic highs 10 to 15 miles west of the mine. Mining of the Anderson-Dietz 1 coal bed has caused water levels to decline in shallow aquifers near the mine and has modified the location of the discharge area for the coal aquifer. After 1 year of mining, water levels in the Anderson-Dietz 1 coal aquifer had declined more than 50 feet at the mine pit, about 20 feet within 1A mile north, west, and south of the mine, and 10 feet or more at a distance of 11/2 miles southwest of the mine (Van Voast, 1974). After 3 years of mining, the cone of depres- sion caused by the mine pit had extended farther to the north, south, and west of the mine. Water-level declines of about 3 feet were measured during 1 year of obser- vation at a well more than 2 miles west of the mined area (Van Voast and Hedges, 1975). East of the mine, water-level declines were much less extensive because of recharge induced from the Tongue River Reservoir and nearby alluvium and clinker. In the area east of the mine, the ground-water gradient reversed, and ground water now flows from the river toward the mine. Water levels in wells completed in the unmined Dietz 2 coal aquifer also declined as a result of increased vertical leakage upward to the Anderson-Dietz 1 coal bed. Placement of overburden material behind the ad- vancing mine pit has created a spoils aquifer composed of a rubble zone at the base of the spoils that generally is confined by relatively impermeable clay in the overly- ing spoils. Therefore, recharge to the spoils primarily occurs as lateral flow from the Anderson-Dietz 1 coal aquifer, and infiltration of precipitation is not con- sidered substantial. The spoils aquifer has an average hydraulic conductivity of 2.8 feet per day, a median hydraulic conductivity of 1.8 feet per day, and an effective porosity of 0.4 percent (Davis, 1984b). The hydraulic conductivity of the spoils aquifer is some- what smaller than the hydraulic conductivity of local coal aquifers, which have an average hydraulic con- ductivity of 5.2 feet per day and a median hydraulic SHALLOW AND DEEP GROUND—WATER FLOW SYSTEMS WEST Perched \L \l/ water table>,.‘ Local coal Recharge from clinker —————— / Spring Dischargeto\ mine pit 141 EAST Mine spoils Tongue River Reservoir Dietz 2 coal FIGURE 67.—Generalized hydrogeologic section, West Decker Mine, Montana. conductivity of 2.2 feet per day. The effective porosity of the spoils aquifer is very similar to that of the coal aquifers, which have an effective porosity of at least 0.3 percent. SUMMARY OF EFFECTS OF SURFACE MINING ON GROUND—WATER FLOW SYSTEMS Hydrologic studies of surface-mine sites have indicated that surface mining of coal almost invariably has some effect on the flow of ground water at or near the mine site. Because surface mining alters the land surface only to relatively shallow depths, the mining primarily affects local flow systems within shallow aquifers. Decreases in water levels in an aquifer contain- ing intermediate flow systems generally are small but can have widespread and long-term effects. Regional flow systems can be affected by mining; however, ef- fects on water levels or flow rates within regional flow systems generally are not severe or are not observed unless the flow system is stressed for a long period of time. The effects of surface coal mining on ground-water systems can be classified into short-term effects, those that occur during mining and reclamation operations, and long-term effects, those that are permanent or con- tinue long after final reclamation. The most common short-term effects of surface coal mining on ground water include dewatering of shallow aquifers, decreases in discharge to springs and small streams, and lower- ing of water levels in wells. These short-term effects generally dissipate, at least in part, after mining is com- pleted and local flow systems become reestablished within the mined area. Long-term effects of surface mining include the destruction of springs and shallow aquifers, alteration of recharge rates, some local decreases in static water levels, and changes in the rate and direction of ground- water flow. Long-term effects almost always are caused by the replacement of overburden materials with spoils materials that have considerably different hydraulic characteristics than the original overburden. The change in the shape of the land surface after mining also contributes to long-term changes in the location of ground-water recharge and discharge. 142 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 GEOCHEMISTRY OF MINE SPOILS By ROBERT E. DAVIS Surface mining of coal entails removal of overburden, the material above the coal, generally in successive fur- rows or strips. After the initial strip or cut is removed, the broken and mixed overburden is placed in a previous cut. In some areas, the overburden or coal, or both, are aquifers. These aquifers are destroyed by mining. How- ever, the replaced overburden, called mine spoils, may become saturated, thus producing a new aquifer. At some surface coal mines, the last cut may be left open and eventually may fill with water, forming a highwall or final-cut lake. One of the potential hydrologic effects of surface coal mining is a change in the quality of ground water asso- ciated with the saturated mine spoils from premining conditions. The effects vary between mine areas and are dependent on the geochemical processes that predom- inate in an area. These processes are dependent on the source and quality of recharge to the spoils, the miner- alogy of the spoils, and the biological activity in the spoils. A number of studies of the geochemical processes related to mine-spoils aquifers have been done in the central and western United States. The studies sum- marized in this report were done in Montana (Davis, 1984b), Colorado (Williams and Hammond, 1988), Oklahoma (Slack, 1983), and Missouri (Hall and Davis, 1986). Locations of the major coal provinces and selected mine-spoils study areas are shown in figure 68. The effects of surface mining generally observed in these studies include increases in the concentrations of dissolved solids and certain trace elements. Acidic pH values, generally associated with coal mining in some parts of the eastern United States, were not observed. GENERAL GEOCHEMICAL PROCESSES The geochemical processes that result in the observed changes in chemical composition of ground water in the mine spoils initially occur in the soil zone and unsatu- rated zone. During periods of infiltration of rain or snowmelt, carbon dioxide gas (002) from the at- mosphere and organic decay reacts with the infiltrating water (H20) forming carbonic acid (H2003): 002 + H20 "2‘ H2003. (17) The resulting acid dissociates and results in the for- mation of bicarbonate ions (HCO3‘), carbonate ions (0032-), and hydrogen ions (H+): H2003 : H+ + H003", (18) HCO3‘ .— H+ + 0032-. (19) The effect of these reactions is a slight increase in the concentrations of bicarbonate and carbonate and the production of a slightly acidic environment that is con- ducive to dissolution of carbonate minerals such as calcite (CaCO3) and dolomite [CaMg(CO3)2]: CaCO3 + H+ .—: Ca2+ + H003" (20) CaMg(CO3)2 + 2H+ :: Ca2+ +Mg2+ +2HCO3‘ . (21) Reactions 20 and 21 result in an increase in the concen- trations of calcium (Ca2+), magnesium (Mg2+), and bicarbonate ions and a decrease in the concentration of hydrogen ions. The slightly acidic environment resulting from reac- tions 18 and 19 also can cause hydrolysis of feldspar minerals, such as anorthite (CaAlSiZOS), albite (N aAlSi308), and microcline (KAlSi308), and the forma- tion of clay minerals such as kaolinite [AlZSi205(OH)4]: CaAlSi208 + 2H+ + H20 ->.A.12s1205(0H)4 + Ca“, (22) 2NaAlSi308 + 2H+ + 9H20->A1231205(0H)4 +4H4Si04+2Na+, (23) 2KA1Si308 + 2H+ + 9H20 +A1281205(0H)4 +4H4Si04+2K+. (24) The result of reactions 22, 23, and 24 is an increase in the concentrations of calcium (Ca2+), sodium (Na+), and potassium (K+) and a decrease in the concentration of hydrogen ions. However, in a system containing car- bonate minerals, reactions 22, 23, and 24 probably do not have a large effect on the system. Sulfate ions (8042‘) may be added to solution as a result of oxidation of sulfide minerals such as pyrite (FeSz) or by dissolution of sulfate minerals such as gyp- sum (CaSO4-2H20): FeSZ+1502+14HZO:4Fe(OH)3+SSO42'+16H+, (25) Caso4 - 2H20: Ca2+ + so42- + 2H20. (26) The oxidation process may be catalyzed by sulfur- oxidizing bacteria such as Thiobacillus ferroxidans, GEOCHEMISTRY OF MINE SPOILS Northern Great Plains and Rocky Mountain Provinces 143 EXPLANATION 4‘ MINE AREA—Numeral corresponds to mine areas in the following list 1 Abandoned mines, Missouri 2 Abandoned mine, Oklahoma 3 Big Sky Mine, Montana 4 West Decker Mine, Montana 5 Seneca Mine, Colorado FIGURE 68.—Coal provinces and location of selected mine-spoils study areas. Ferrobacillus ferroxidans, and Thiobacillus thioxidans. The source of oxygen (02) in reaction 25 probably is at- mospheric. The sulfate ions produced in reaction 25 either are transported in solution to the aquifer or are precipitated as gypsum, probably as a result of evapotranspiration. The precipitated sulfate minerals may be dissolved later and transported to the aquifer by deeply percolating recharge water. The hydrogen-ion concentration or acidity resulting from reaction 25 is buffered by additional dissolution of carbonate minerals and feldspar minerals, if present, as in reactions 20, 21, 22, 23, and 24. Consequently, concentrations of calcium, magnesium, sodium, potassium, and bicarbonate may be increased. Adsorption and ion-exchange reactions also may be important processes in soil and in unsaturated and satu- rated zones. Generally, the most prevalent reaction in- volves calcium and sodium and may be expressed as: Na2X + Ca2+ = CaX + 2Na+, (27) where X represents the solid-host species, such as a clay mineral or organic material. If proceeding from left to 144 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 right, reaction 27 results in an increase in sodium con- centration and a decrease in calcium concentration, which forces additional dissolution of carbonate min- erals to maintain equilibrium. Magnesium and potas- sium also may be involved in reactions similar to reaction 27. During anaerobic conditions, sulfate may be reduced to sulfide by bacteria: $03 + ZCH20 (organic material) sulfate-reducing HS" + HCO3 + H20 + 002. (28) bacteria Sulfate-reducing bacteria Desulfovibrio desulfuricans are known to exist in at least some coal-mining areas (Dockins and others, 1980). Reaction 28 results in a decrease in sulfate concentration and an increase in bicarbonate concentration. The reactions above illustrate the general geochemical processes that may occur in the mine spoils; therefore, 600 — 1 a,b,c.d,e,f,g,h 500 — Material overlaying 2 df h coal (overburden) c, ' '9' table or . 400 —__W_——a‘9l—-———-—p°_’entlometric surfac ‘\ e Material underlying coal 200 — VERTICAL DISTANCE, IN FEET; DATUM IS ARBITRARY 8 O l 100 ‘— ground water in the mine spoils primarily will contain some combination of calcium, magnesium, sodium, potassium, bicarbonate, and sulfate ions. The concen- trations and types of ions are dependent on which of the geochemical processes predominate. Possible major hydrologic and geochemical processes are shown diagrammatically in figure 69. EVOLUTION OF MINE-SPOILS WATER QUALITY AT SELECTED SITES ABANDONED MINES, MISSOURI Dissolved-solids concentrations in water from wells in or near mine spoils at three abandoned, unreclaimed mine sites in north-central Missouri (mine area 1, fig. 68) ranged from 1,890 to 4,660 milligrams per liter and averaged 2,900 milligrams per liter (Hall and Davis, 1986). The mine spoils consist of glacial drift of Quater- nary age and bedrock of Pennsylvanian age disturbed by mining. The drift is composed of poorly sorted sand, Mine spoils 1 a,b,C.d,e,f.g,h —- Highwall lake a.b,c,d,e,f.g,h,i \ EXPLANATION -—> GENERAL DIRECTION OF GROUND- WATER FLOW POSSIBLE HYDROLOGIC PROCESSES Infiltration of precipitation Flow through unsaturated zone Horizontal flow through saturated zone bWNd Vertical flow from underlying aquifer POSSIBLE GEOCHEMICAL PROCESSES a Carbon dioxide production from oxidation of organic matter Oxidation of sulfide minerals Dissolution of carbonate minerals Hydrolysis of feldspar minerals (DQOO' Precipitation of sulfate minerals because of evapotranspiration Dissolution of sulfate minerals Ion exchange and (or) adsorption Precipitation of carbonate minerals 30-. i Sulfate reduction Processes resulting from disruption and exposure of spoils during mining. These processes probably affect only the initial volumes of water passing through the spoils * FIGURE 69.—Generalized section of hypothetical mine area and major hydrologic and geochemical processes. GEOCHEMISTRY OF MINE SPOILS silt, and clay and some well-sorted sand lenses. Dissolved-solids concentrations in water from the drift ranged from 239 to 1,280 milligrams per liter and aver- aged 602 milligrams per liter. The bedrock is composed of shale, limestone, sandstone, and coal. Dissolved- solids concentrations in water from the bedrock ranged from 580 to 883 milligrams per liter and averaged 754 milligrams per liter. Specific constituents observed in greater concentra- tions in water from mine spoils include calcium, mag- nesium, sulfate, iron, and manganese. Water from the mine spoils is a calcium magnesium sulfate type, whereas water from the drift is either a calcium magne- sium bicarbonate or calcium magnesium sulfate type; water from the bedrock is either a sodium bicarbonate or calcium bicarbonate type. Water from highwall lakes is similar to water from the mine spoils. The major minerals in the mine spoils are quartz, calcite, dolomite, and various nonsodic clays. Pyrite, gypsum, and several iron-oxide minerals also are present. Climate in the area is humid. Recharge to the drift occurs by infiltration of precipitation and lateral flow from adjacent aquifers. Recharge to bedrock probably occurs by infiltration of precipitation and vertical flow from the drift. The rates of recharge to the drift and bedrock are unknown. Recharge to the mine spoils is from infiltration of precipitation and flow from adjacent aquifers. Precipitation probably is the major source of recharge, both directly and indirectly as flow from highwall lakes. Mass-transfer calculations using the computer pro- gram BALANCE (Parkhurst and others, 1982) indicate that the main geochemical processes that result in the observed water quality in the mine spoils may include oxidation of pyrite, dissolution of gypsum, dissolution and precipitation of calcite, dissolution of dolomite, con- sumption of oxygen gas, consumption and release of carbon dioxide gas, precipitation of iron oxide minerals, and release of relatively small quantities of sodium ions by exchange or adsorption. Saturation indices deter- mined using the computer program WATEQF (Plum- mer and others, 1976) indicate that water from all sources are near saturation with respect to quartz, calcite, and dolomite. Water from wells in the drift and the bedrock is undersaturated with respect to gypsum; water from wells in or near the mine spoils is saturated with respect to gypsum. The difference probably results from oxidation of sulfide minerals and dissolution of gypsum in the unsaturated zone on surfaces freshly ex- posed by mining. Ages of the mine spoils at the three mines differ. At these mines, spoils were emplaced during 1940, during 1952, and during 1968. Statistical comparisons indicate 145 that differences in general water quality exist between the three mine-spoils areas, but the differences cannot be attributed to age. Therefore, the major changes determined in water quality occurred within 12 years or less and have persisted for more than 40 years. Most of the geochemical processes that result in the measured increases in dissolved-solids concentrations occur in the unsaturated zone. Therefore, any attempts to improve water quality in the mine spoils probably would benefit from decreasing the volume of water recharging the mine spoils by vertical infiltration. ABANDONED MINE, OKLAHOMA Ground water in saturated mine spoils at an aban- doned, unreclaimed coal mine in eastern Oklahoma (mine area 2, fig. 68) predominantly is a sodium sulfate type (Slack, 1983). The average dissolved-solids concen- tration was 1,990 milligrams per liter in 57 water samples from 6 wells completed in the mine spoils. Water from the wells is similar in quality to water in two nearby highwall ponds. Except for dissolved solids, iron, manganese, and sulfate, constituent concentra- tions generally do not exceed drinking-water regulations of the US. Environmental Protection Agency (1986a, 1986c). The coal that was mined was part of the Pennsylva- nian Hartshome Sandstone, which is overlain by the Pennsylvanian McAlester Formation. The Pennsylva- nian McAlester Formation mainly consists of silty shale and several sandstone layers. X-ray diffraction analyses of mine-spoils material, which contains the Hartshome Sandstone and the McAlester Formation, indicate that the spoils predominantly consist of quartz, kaolinite, chlorite, illite, and mica. Also generally present, in quan- tities less than a few weight percent, were smectite, calcite, feldspars, pyrite, gypsum, and siderite. The mechanisms of recharge were not discerned. How- ever, because of a relatively humid climate, recharge from infiltration of precipitation probably is a major factor. The geochemical processes that result in the spoil water quality probably include oxidation of sulfide minerals, dissolution of sulfate minerals, dissolution of carbonate minerals, hydrolysis of feldspar minerals, and ion exchange or adsorption. BIG SKY MINE, MONTANA Dissolved-solids concentrations in water from satu- rated mine spoils at the Big Sky Mine in southeastern Montana (mine area 3, fig. 68) average about 3,700 milli- grams per liter, whereas dissolved-solids concentrations in water from coal (subbituminous C) aquifers in the area average about 2,700 milligrams per liter (Davis, 1984b). 146 The specific constituents for which increases in concen- tration were measured include calcium, magnesium, potassium, bicarbonate, and sulfate. Water from both aquifers is mainly a calcium magnesium sulfate type. The coal is part of the Tongue River Member, which in this area is sandier than near the West Decker Mine (see “West Decker Mine, Montana” section). Over- burden samples from an area near the Big Sky Mine generally contained, in general descending order of abundance, quartz, dolomite, kaolin, feldspars, mica, and calcite. Samples associated with mine spoils also contained gypsum. The clay fraction generally consisted of kaolinite and illite. Climate in the area is semiarid. Recharge to the coal aquifers is by infiltration of precipitation in areas of hydraulically connected clinker and, probably to a lesser extent, by infiltration of precipitation through the sandy overburden. The quality of ground water in the coal aquifers primarily is determined by dissolution of silica, sulfate, and carbonate minerals in the unsaturated zone. Because of the lack of sodic clay in the overburden, ex- change of calcium and magnesium for sodium is not predominant. Sulfate reduction probably decreases the sulfate concentration and increases bicarbonate concen- tration (Dockins and others, 1980), although this proc- ess apparently is not predominant either. Recharge to the mine-spoils aquifer occurs as lateral flow from the coal it replaces and as vertical infiltration of precipitation. The main geochemical processes that result in the increased dissolved-solids concentration in the spoils include oxidation of pyrite, dissolution of the resultant gypsum, and dissolution of carbonates. Saturation indices determined using the computer program WATEQF (Plummer and others, 1976) sub- stantiate the geochemical processes listed above. Water from the lignite aquifers is saturated or supersaturated with respect to quartz, calcite, and dolomite and satu- rated to slightly undersaturated with respect to gypsum. Water from the mine spoils generally is satu- rated to supersaturated with respect to quartz, calcite, dolomite, and gypsum. The general state of saturation or supersaturation of water from the mine spoils in- dicates that the dissolved-solids concentration is near maximum and will not increase substantially assuming solute sources and temperature do not change. In summary, the increases in dissolved-solids concen- tration in water from the mine-spoils aquifer at the Big Sky Mine primarily result from oxidation of sulfide minerals, dissolution of the resultant sulfide minerals, and limited dissolution of carbonate minerals. The dissolved-solids concentration in water in the mine spoils probably will not change in the near future, although the concentration may decrease if the mine- spoils water flows through a coal aquifer. SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974-84 WEST DECKER MINE, MONTANA Dissolved-solids concentrations in water from saturated mine spoils at the West Decker Mine in southeastern Montana (mine area 4, fig. 68) average about 2,500 milligrams per liter, whereas dissolved- solids concentrations in water from coal (subbituminous C) aquifers in the area average about 1,400 milligrams per liter (Davis, 1984b). The specific constituents for which increases in concentration were measured include calcium, magnesium, sodium, potassium, bicarbonate, and sulfate. Water from both aquifers is mainly a sodium bicarbonate type, although water from the mine spoils tends to have a larger percentage of total milli- equivalents per liter of calcium plus magnesium and of sulfate plus chloride. The coal aquifer is part of the Tongue River Member, which in this area consists of calcareous shale, siltstone, sand and sandstone, and coal beds. Overburden samples from an area near the West Decker Mine primarily con- tained quartz, layer silicates, and the clay-mineral groups smectite, illite, chlorite, and kaolin. Also present, in general descending order of abundance, were feld- spars, carbonates such as calcite and dolomite, calcium siderite and siderite, and pyrite. Gypsum was detectable in only two of the samples. Climate in the area is semiarid. Recharge to the coal aquifer is by infiltration of precipitation in areas of hydraulically connected clinker, which is baked, fused, and fractured rock that results from burning of underly- ing coal beds. Recharge to the coal aquifer also occurs as vertical leakage from an underlying coal bed. The chemistry of ground water in the coal aquifer predom- inantly is determined by the dissolution of silica, sulfate, and carbonate minerals in the recharge areas. The sulfate minerals probably result from oxidation of pyrite in the recharge areas. Most of the calcium and mag- nesium from dissolution of the carbonate minerals is exchanged for sodium with either sodic clays in the clinker or with the coal itself. Sulfate reduction prob- ably decreases the sulfate concentration and increases the bicarbonate concentration (Dockins and others, 1980). Recharge to the mine-spoils aquifer primarily occurs as lateral flow from the coal it replaces. Infiltration of precipitation through the mine spoils probably is not a substantial source of recharge because of the large concentration of clay in the mine spoils. Therefore, the geochemical processes that result in increased dissolved-solids concentrations probably occur in the saturated zone. These processes probably include dissolution of gypsum formed from oxidation of pyrite in the overburden during mining, dissolution of car- bonates, and ion exchange or adsorption. GEOCHEMISTRY OF MINE SPOILS Saturation indices determined using the computer program WATEQF (Plummer and others, 1976) sub- stantiate the geochemical processes listed above. Water from the coal aquifers is supersaturated with respect to quartz, undersaturated to saturated with respect to calcite and dolomite, and undersaturated with respect to gypsum. Water from the mine spoils is super- saturated to saturated with respect to quartz, calcite, and dolomite and undersaturated with respect to gyp- sum. The undersaturation with respect to gypsum for mine spoils and lignite aquifers indicates limited quan- tities of pyrite and gypsum in the overburden and mine spoils or a lack of surface recharge through the over- burden and mine spoils, or both. The general state of saturation or supersaturation with respect to other minerals for water in the mine spoils indicates that the dissolved-solids concentration is near maximum and will not increase substantially assuming solute sources and temperature do not change. In summary, increases in dissolved-solids concentra- tion in water from the mine-spoils aquifer at the West Decker Mine primarily result from oxidation of limited quantities of sulfide minerals and dissolution of the resultant sulfate minerals, dissolution of carbonate minerals, and cation-exchange or adsorption reactions. The dissolved-solids concentration in water in the mine spoils probably will not change in the near future, although the concentration may decrease if the mine- spoils water flows through a coal aquifer. SENECA MINE, COLORADO Investigations at the Seneca Mine in northwestern Colorado (mine area 5, fig. 68) were limited to defining water movement, water chemistry, and geochemical processes in the upper 6 feet of the unsaturated zone (Williams and Hammond, 1988). Material above the two coal beds being mined consists of soils developed on material derived from the Cretaceous Iles Formation and Williams Fork Formation. These formations vary 147 in composition from sandstone to shale. The clay frac- tion of the sediments consists primarily of kaolinite and illite and has very little smectite. The area has a semiarid climate and receives about 16 inches of precipitation annually. Recharge to all shallow aquifers in the mine area is from infiltration of precipitation. Recharge to the undisturbed system is estimated to be one-half inch per year. Recharge to the mine spoils ranges from 2 to 6 inches per year because of a greater infiltration rate near the land sur- face. However, recharge to the mine spoils is decreas- ing because of continuing natural compaction of the spoils. Water in the upper 6 feet of the spoils is a calcium magnesium sulfate type. Dissolved-solids concentration ranged from an average of 3,960 milligrams per liter during 1978 to an average of 3,560 milligrams per liter during 1979. The average concentrations of the major ions were 460 milligrams per liter for calcium, 370 milli- grams per liter for magnesium, 111 milligrams per liter for sodium, 2,540 milligrams per liter for sulfate, and 224 milligrams per liter for bicarbonate. The primary geochemical processes that affect the water quality in the unsaturated zone are dissolution of carbonate minerals, oxidation of sulfide minerals, and dissolution of gypsum that results from oxidation of sulfide minerals. Carbonic acid, that results from ab- sorption of carbon dioxide from the atmosphere, decay- ing organic matter, and plant respiration enhances the dissolution of carbonates. Pyrite oxidation, which also enhances carbonate dissolution, occurs as new reaction surfaces are exposed to an oxidizing environment dur- ing disruption by mining. The sulfate produced may be precipitated as gypsum during periods when evapo- transpiration rates are rapid and later may be dissolved by deeply percolating recharge water. Saturation indices determined using the computer program WATEQF (Plummer and others, 1976) indicate that the water generally is saturated or supersaturated with respect to calcite, dolomite, and gypsum. 148 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 MINE DRAINAGE By ARTHUR N. Orr In the Eastern United States, acid mine drainage is the most acute issue associated with coal mining. In the Western United States, nonacid (saline) drainage is an issue, although it is much less severe and far more local than acid mine drainage is in the Eastern United States. Acid mine drainage results from milling in a humid en- vironment and diminishes in severity as precipitation lessens in a westward direction and as the sulfur con- tent in coal decreases (coals in the Western United States generally contain less sulfur than coals in the Eastern United States). The most substantial production of acid mine drain- age occurs in the Appalachian regions of the Eastern United States, an area that encompasses about 111,000 square miles from northern Pennsylvania to central Alabama. The acid mine drainage within these regions primarily occurs in Pennsylvania, Ohio, West Virginia, and Kentucky, possibly because more than 90 percent of the coal is located in these States. Also, more coal has been mined in the Northern Appalachian region than in the Southern Appalachian region, which has resulted in more pyritic material being disturbed and subject to oxidation in the Northern Appalachian region than in the Southern Appalachian region. Acid mine drainage does not seem to be a significant issue in the Western United States. Wentz (1974a) reported that acid mine drainage was not an issue in Colorado and believed this was because of the minimal sulfur content of coal in the Western United States. Wentz (1974a) stated that 65 percent of the bituminous, subbituminous, and lignite reserves in the United States, which occur primarily west of the Mississippi River, contain 1 percent or less total sulfur, whereas 43 percent of the reserves in the Eastern United States contain more than 3 percent total sulfur. The infrequent occurrence of acid mine drainage in the Western United States also can be attributed to relatively deep soils that contain calcareous material, small quantities of annual precipitation, and, conversely, large evapotranspiration levels and buffered ground waters. In contrast, the Eastern United States has coals that have a large sulfur content, annual precipitation nearly twice as much as that of the Western United States, and soils that are shallow and relatively acid. PRODUCTION OF ACID MINE DRAINAGE The chemical characteristics of water that drains from a coal-mine site are dependent on the geologic, hydro- logic, and topographic features that are associated with the mine site. When the rock material overlying the coal seam is non-calcareous and the material or the coal seam contains fine-grained pyrite (iron sulfide), oxygen and water can combine to oxidize the pyrite and produce iron sulfate and sulfuric acid. Pyrite oxidation produces a strongly acidic solution. This acidic environment may cause dissolution of large concentrations of aluminum and manganese and toxic metals such as chromium, cadmium, lead, copper, zinc, and nickel in the mine drainage. These metals will remain dissolved as long as the solution remains strongly acidic. The production of acid that is associated with coal mining primarily is attributed to the oxidation of sulfur. Even though sulfur occurs in coal in organic and in- organic forms, it is the inorganic, pyrite sulfur that is the most reactive. Further, Caruccio and others (1976) have reported that not all pyrite sulfur is acid produc- ing. Pyrite that has grain sizes larger than 400 micro- meters or that has crystals of cubical or triangular shape are inert and, therefore, are nonacid producing. Conversely, pyrite that morphologically occurs as clusters of crystal spheres approximately 25 micro- meters in diameter, called framboidal pyrite, is very reactive and acid producing. Furthermore, the distribu- tion of framboidal pyrite has been determined to be relatively abundant in coal formed in the marine, brackish-water environment as opposed to coal formed in the freshwater environment. Stumm and Morgan (1970) summarized the major mechanisms believed to be involved in the production of acid mine drainage: The dissolution of pyrite and the oxidation of sulfide to sulfate, Fes2 + £02 + H20 :: Fe” + 2so4-2 + 2H+; (29) the oxidation of ferrous iron, Fe” + £02 + H+ :: Fe+3 + %H20; (30) the oxidation of pyrite by ferric iron, Fes2 + 14Fe+3 + 8H20 :: 15Fe+2 + 2804‘2 + 16H+; and (31) the precipitation of ferric iron, Fe+3 + 3H20 :: Fe(OH)3 + 3H+. (32) Nordstrom (1977) stated that acid mine drainage is the most acidic of all weathering reactions primarily MINE DRAINAGE because of the oxidation of sulfur to sulfate and dis- cussed the following sequence of reactions that describe the decomposition of pyrite. Pyrite is abiotically oxi- dized to ferrous iron and sulfur: Fe82 + $02 2H+ : Fe” + $2 + H20. (33) Ferrous iron is rapidly oxidized and hydrolyzed at ap- proximately neutral pH and equalizes the change in pH shown by equation 33: Fe+2+ £02 + gHzo : Fe(OH)3 + 2H. (34) The sulfur product shown by equation 33 is oxidized to sulfate and is the reaction that decreases the pH: 82 + 302 + 2H20 : 2804-2 + 4H+. (35) Equations 33 and 35 can be combined to yield equation 29, and equations 30 and 32 can be combined to yield equation 34. As the pH decreases, ferric iron con- centration increases, and ferric iron becomes the domi- nant oxidizer of pyrite when pH is less than 3 (eq. 31). Oxidation of sulfur to sulfate, however, causes nearly all the acidity. The quantity of ferric iron available for the oxidation of pyrite would be minimal at a pH of 3 in a purely abiotic system. However, microbial catalysis (the oxidation of ferrous iron by Thiobacillus ferrox- idans) increases the normally slow reaction rate by several orders of magnitude. According to Kleinmann and others (1981) the acid- production sources for coal-mine drainage occur in a three-stage sequence that is dependent on the activity of Thiobacillus ferroxidans and solution Eh (oxidation- reduction potential) and pH. During the first stage, reactions represented by equations 29 and 34 are the mechanisms involved. Reaction 29 proceeds abiotical- ly and by direct biotic oxidation. Reaction 34 proceeds abiotically but lessens as the pH decreases. Solution chemistry during this stage is indicative of a pH in ex- cess of 4.5 (slightly acidic) and large sulfate and small iron concentrations. A change from the abiotic to the biotic oxidation of ferrous iron initiates the second stage. The mechanisms operative for the second stage are the same reactions involved during the first stage. Reaction 29 continues to be driven abiotically and biotically. Reaction 34, however, proceeds at a rate determined primarily by biotic activity. The solution chemistry indicates the in- creased acidity. The pH now ranges from 2.5 to 4.5, the solution contains large quantities of sulfate, and acid- ity and total iron concentrations are increased. Total iron is represented by a small ferric/ferrous iron ratio. A decrease in Fe(OH)3 precipitation and a concurrent 149 increase in iron solubility at a pH less than 3 results in increased ferric iron activity. The third stage begins as the pH approximates 2.5. The mechanisms operative during the third stage in- volve reactions represented by equations 30 and 31. Reaction 30, the oxidation of ferrous to ferric iron, is entirely bacterially mediated and affects the rate of reaction 31. The combined effects of the bacterial ox- idation of ferrous iron, reduction of ferric iron by pyrite, and formation of ferric oxyhydroxides and ferric sulfate determine the steady-state activity of ferric iron, while the availability of ferric iron is limited by the rate of reaction 30. The solution chemistry now would have a pH of 2.5 or less (very acidic) and large concentrations of sulfate and total iron. Iron is composed of a large ferric/ferrous iron ratio. As previously discussed in this section, the produc- tion of acid primarily occurs because of the oxidation of pyritic sulfur by ferric iron to ferrous sulfate. The resulting iron sulfate and associated aluminum sulfate salts are an additional source of acidity. Dissolution of these salts in distilled water results in a solution pH of about 2.0. These salts generally occur in areas of acid mine drainage; however, the occurrence of these salts only recently has been ascertained. It also is possible that much of the present-day acidic discharge is derived from past mining processes that produced these acid iron and aluminum salts. QUANTITY AND QUALITY OF ACID MINE DRAINAGE Underground mines below the ground-water table usually have a continuous discharge, while underground mines located where the water table fluctuates above and below the mine level produce intermittent dis- charge. Mines need to be dewatered during the active mining phase, which partially dewaters the ground- water system in the mine area and increases the discharge in the adjacent surface-water flow system. For some horizontal coal seams and for coal seams that dip away from the mine opening, mining causes dewatering of the overlying strata. When mining ac- tivities cease, the mined cavity eventually will flood (because of not being pumped) and ground water will return to premining levels. If the coal seam dips toward the mine opening, the ground-water flow system will dewater the mine cavity continuously by gravity drain- age, regardless of the mining status (US. National Research Council, 1981a). In contrast, discharge from surface mines generally is intermittent and usually coin- cides with precipitation. Evaluation of effects of acid mine drainage on the water quality of streams is very subjective. For example, in a summary of surface-water quality for the 150 Eastern Province (Wetzel and Hoffman, in press) report that of 1,500 sites that were sampled at least 4 times during 1972—82, the median values at only 25 sites either equaled or exceeded the criteria for pH, total iron, total manganese, and dissolved sulfate established by the Federal Water Pollution Control Administration (1968) to delineate deterioration by acid mine drainage. The criteria for all four constituents may not have been simultaneously exceeded in a single analysis; however, the median for a minimum of four values for total iron, total manganese, and dissolved sulfate and the minimum value for pH equaled or exceeded the level. These are stringent criteria. Criteria for acid mine drainage1 Water-quality property or constituent pH (units) ...................... less than 6.0 Total iron (micrograms per liter) . . . greater than 500 Total manganese (micrograms per liter) greater than 500 Dissolved sulfate (milligrams per liter) greater than 75 1Established by Federal Water Pollution Control Administration (1968). Biesecker and George (1966) determined that 194 sites (61 percent) were measurably affected by acid mine drainage based on studies of 318 sites in the Appala- chian region that have drainage areas larger than 100 square miles. This determination was based on samples that contained concentrations of sulfate greater than 20 milligrams per liter. If the previous sulfate criterion of greater than 75 milligrams per liter is used, the number of sites decreases from 194 to 124. Further- more, if the pH criterion of less than 6.0 is used, the number of sites affected decreases to 61. If the cri- terion used by the Appalachian Regional Commission (1969) listed below is applied to the 1965 sulfate data, the number of sites decreases further to 41. Thus, instead of there being 61 percent of the 318 sites considered affected by acid mine drainage, there only would be 13 percent of the sites affected by acid mine drainage. SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Criteria for acid mine drainage1 Water-quality property or constituent pH (units) ...................... less than 6.0 Acidity/alkalinity (milligrams per liter) net alkalinity, less than 20 Hardness (milligrams per liter) ..... greater than 250 Suspended solids (milligrams per liter) greater than 250 Dissolved solids (milligrams per liter) greater than 500 Total iron (micrograms per liter) . . . greater than 1,500 Aluminum (micrograms per liter) . . . greater than 500 Manganese (micrograms per liter) . . greater than 1,000 Sulfate (milligrams per liter) ....... greater than 250 1Appalachian Regional Commission (1969). A new concept for evaluating the effect of acid mine drainage on a stream or river system is net alkalinity. This concept, used by Rozelle and others (1976), pro- vides an assessment of the acid-alkaline balance within the system. The net alkalinity is defined as the differ- ence between the alkalinity of a sample determined by titration using a standard acid to a pH of 4.5 and the acidity determined by titration using a standard base to a pH of 8.3. If the alkalinity is greater than the acid- ity, net alkalinity is positive, pH generally will exceed 6.0, and alkaline conditions prevail. Positive net alkalin- ity provides a measure of the alkaline reserve or capaci- ty of the system to neutralize acid input. If the acidity is greater than the alkalinity, the net alkalinity is nega- tive, the stream or system is considered acid and usual- ly, but not necessarily, has a pH of less than 6.0. To illustrate the usefulness of the concept, a stream that receives 100 pounds per day of positive net alkalinity could be expected to assimilate or neutralize 100 pounds per day of acid input without decreasing the pH to less than 6.0. CLASSIFICATION OF ACID MINE DRAINAGE A classification devised by Hill (1968) that differen- tiates acid mine drainages based on their most prom- inent chemical characteristics is listed in table 6. Although there are four numbered classes, three classes TABLE 6.—Acid mine drainage classes [Modified from Hill, 1968] Water-quality Class 1, Class 2, Class 3, Class 4, property or acid discharges partially oxidized and not oxidized constituent oxidized and (or) neutralized and neutralized and (or) alkaline neutralized pH (units) ................................ 2—4.5 3.5—6.6 6.5—8.5 6.5—8.5 Acidity (CaCOa) (milligrams per liter) ......... 1,000—15,000 0—1,000 0 0 Ferrous iron (milligrams per liter) ............ 500—10,000 0-500 0 50-1000 Ferric iron (milligrams per liter) .............. 0 0—1,000 0 0 Aluminum (milligrams per liter) .............. 0—2,000 0-20 0 0 Sulfate (milligrams per liter) ................. 1,000—20,000 500—10,000 500—10,000 500-10,000 MINE DRAINAGE are identified by oxidation designations: partially ox- idized, oxidized, and not oxidized. These designations seem to be related to the form of the iron. According to Nordstrom (1977), acid mine drainage should be well suited to relating oxidation-reduction measurements to the ferrous/ferric iron redox potential. 151 Another way to classify acid mine drainage is to use dual-acidity titration curves (Ott, 1986). Even though dual-acidity titration curves do not directly measure the redox potential, the shape formed by these curves is de- pendent on the ferrous/ferric iron ratio as shown in figure 70. 9 l l l l A B / 8 — / — — l — // WELL T / WELL U 7 — April 27, 1984 l/ — — April 24, 1984 — U) ': _ __ _. Z 3 Z I. _ _ _ a. 3 Fe“ (3.0 milligrams per liter) Fe+2(558 milligrams per liter) _ Fe“ (14 milligrams per liter) Fe+3 (372 milligrams per liter) =0.21:1 / =1.5:1 2 I I I I | I I 0 250 500 750 1,000 0 1,000 2,000 3,000 4,000 5,000 ACIDITY AS CALCIUM CARBONATE, ACIDITY AS CALCIUM CARBONATE, IN MILLIGRAMS PER LITER IN MILLIGRAMS PER LITER 9 l I I / EXPLANATION C // UNTREATED—Unoxidized, raw sample 8 wm / — ----- °i"2:f:':f::‘13“21°2.:°°“‘ April 24, 1934 / y g p x' e 3" ea 6 7 / __ :2 2 6 — D Z L 5 — I O. 4 _ / /,_// Fe+2 (270 milligrams per liter) 3 :// Fe+3 (3.0 milligrams per liter) =90 2 1 2 I I I I I 0 250 500 750 1,000 1,250 1,500 ACIDITY AS CALCIUM CARBONATE, IN MlLLlGRAMS PER LITER FIGURE 70.-—Titration curves of untreated and oxidized ground-water samples containing different ferrous/ferric (Fe+2/Fe+3) iron ratios. A, well T; B, well U; and C. TR 2 well 1. 152 The acidity titration curve of the untreated sample from well T (fig. 70A) is nearly superimposed on the titration curve of the oxidized sample treated with hydrogen peroxide (H202). The iron composition of the untreated sample is about 20 percent ferrous iron and about 80 percent ferric iron. The dual curves (fig. 7 OB) that represent a sample from well U, whose iron compo- sition is 60 percent ferrous iron, are somewhat diverg- ent, while the dual curves (fig. 700) that represent a sample from TR 2 well 1, whose iron composition is about 99 percent ferrous iron, are completely divergent. The samples as shown in figure 70 can be classified as follows: 70A, oxidized; 7 OB, partially oxidized; and 7 00, not oxidized. Another use for acidity titration curves is as “finger- prints” that seem to change little with time (figs. 71—7 3) as long as substantial geochemical changes do not oc- cur (Ott, in press). For example, the curves remain similar even though a 10-month period of time elapsed between sample collections (fig. 71). Also, changes in SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 ionic concentration due to dilution (fig. 73) do not seem to cause the characteristic shape of the curve to change. The mine-drainage classification by Hill (1968) dis- cussed at the beginning of this section was based on the concentrations of the principal chemical constitu- ents as determined in the laboratory. Mine discharge also can be classified onsite by estimating the concen- trations of chemical constituents from onsite determina- tions. In addition, Lovell (1973) determined that a reasonable correlation between specific conductance and sulfate concentration exists. Ott (1986) used regression equations derived from acidity titration curves to esti- mate concentrations of ferrous iron, ferric iron, alumi- num, and total acidity. NONACID MINE DRAINAGE As indicated by Hill’s mine-drainage classification (Hill, 1968), all mine drainage is not acidic. Nonacidic drainage may occur where pyrite is: (1) Massive rather 9 I I I I I I I A B / January 16, 1984 / October 30, 1984 / 8 — LABORATORY VALUES /_' '— LABORATORY VALUES / — pH 2.6 pH 2.8 Specific conductance (1,950 microsiemens per centimeter) Iron (51 milligrams per liter) Aluminum (49 milligrams per liter) / Sulfate (1,280 milligrams per liter) / / pH, IN UNITS Specific conductance (2,760 microsiemens per centimeter) / Iron (112 milligrams per liter) / Aluminum (58 milligrams / per liter) Sulfate (1,380 milligrams per liter) I I | I 2 I I I I 0 100 200 300 400 ACIDITY AS CALCIUM CARBONATE, IN MILLIGRAMS PER LITER 500 0 200 400 600 800 ACIDITY AS CALCIUM CARBONATE, IN MILLIGRAMS PER LITER 1 ,000 EXPLANATION UNTREATED— Unoxidized, raw sample ————— OXIDIZED—Treated with 30 percent hydrogen peroxide and heated FIGURE 71.—Similar titration-curve characteristics provided by different samples collected from seep (site 2), Clarion County, Pennsylvania. A, Sample collected January 16, 1984; and B, Sample collected October 30, 1984. (Fifty—milliliter samples titrated with 0.1 Normal sodium hydroxide.) MINE DRAINAGE 153 9 I I I | | I I I I A January 17, 1984 // 3 October 30, 1984 a — — — .1. LABORATORY VALUES // LABORATORY VALUES / pH 2.3 / pH 3.0 // 7 _ Specific conductance (3,288 microsiemens per centimeter) / Specific conductance (2,240 / Iron (702 milligrams per liter) / — — microsiemens per centimeter) / — Aluminum (167 milligrams per liter) // "on ”-48 milligrams per liter) // (D Sulfate (3,696 milligrams per liter) / AlumInum (61 mIlIIgrams / l: / _ _ per liter) / _ g Sulfate (1,390 milli- / z / grams per liter) / _ / / If _ / _ / ‘3' / /’/ 2 I I I | I I I I | I 0 500 1,000 1,500 2,000 2,500 3,000 3,500 0 200 400 600 800 1,000 ACIDITY AS CALCIUM CARBONATE, ACIDITY AS CALCIUM CARBONATE, IN MILLIGRAMS PER LITER IN MILLIGRAMS PER LITER EXPLANATION UNTREATED — Unoxidized, raw sample ————— OXIDlZED—Treated with 30 percent hydrogen peroxide and heated FIGURE 7 2.—Similar titration-curve characteristics provided by different samples collected from well 4K, Clarion County, Pennsylvania. A, Sample collected January 17, 1984; and B, Sample collected October 30, 1984. (Fifty-milliliter samples titrated with 0.1 Normal sodium hydroxide.) than fine grained or framboidal; (2) of small sulfur con- tent; or (3) overlain by calcareous material. Based on results from leaching studies and onsite observations, Caruccio and others (1976) concluded that pyrite mor- phology was considerably different among similar sulfur-content pyrite samples that did and did not pro- duce acid. The nonacid-producing pyrite was massive and most grains were larger than 400 micrometers. Smaller grained cubical or triangular pyrite crystals that range from 5 to 10 micrometers also were deter- mined to be inert. Williams and others (1982) deter- mined positive exponential relations among both acidity and sulfate and total sulfur concentrations in laboratory leaching studies. These relations indicate that acid is not produced in shale that has small sulfur content or in coal that has small pyrite content. Based on labora- tory and onsite observations, Williams and others (1982) determined that the equivalent of 1 foot of limestone is sufficient to inhibit or neutralize all the acid produced from the oxidation of 10 feet of brackish shale that has a 3 percent sulfur content. As long as alkalinity exceeds acidity in percolating mine water, the only major effect on mine discharge ex- posed to acid-producing pyrite is a larger than normal sulfate concentration. The catalyzing iron bacteria are excluded, and the iron and possibly the aluminum solubility is kept to a minimum. Where a calcareous overburden exists over a- coal, a nonacidic drainage from a mine area may occur; how- ever, this is not always the situation because carbonate solubility is dependent on the pH and the partial pres- sure of carbon dioxide (002). According to Caruccio (1973), water in contact with carbonates at a partial pressure equal to 10‘3-5 atmospheres do not generate sufficient alkalinity to effectively neutralize acid pro- duction. The solubility of carbonates is much less than that of sulfides at atmospheric conditions, and thus greater acidities than alkalinities can be produced, which can result in an acidic environment. In order for percolating mine water to benefit from the acid- neutralizing properties of bicarbonate, the water re- quires large partial pressures of CO2 before contacting carbonate material. This requirement can be met by con- ditions in the soil. Soil air contains a C02 content that 154 SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974—84 9 l l l I I I I I l l I l I a A September 18, 1934 October 31, 1984 / _ / / _ / m _ I: z D _ Z / =3 — / ~ // / _ / _ // /// ”’/ 2 _ _ 2”, _ 1 . | . I . I . . I . I . I I 0 1,000 2,000 3,000 4,000 0 1,000 2,000 3,000 4,000 ACIDITY AS CALCIUM CARBONATE, IN MILLIGRAMS PER LlTER LABORATORY VALUES pH 4.7 Specific conductance (5,800 microsiemens per centimeter) lron (1,464 milligrams per liter) Aluminum (43 milligrams per liter) Sulfate (2,919 milligrams per liter) ACIDITY AS CALCIUM CARBONATE, IN MILLIGRAMS PER LITER LABORATORY VALUES pH 4.7 Specific conductance (5,590 microsiemens per centimeter) Iron (1,370 milligrams per liter) Aluminum (23 milligrams per liter) Sulfate (6,270 milligrams per liter) EXPLANATION UNTREATED— Unoxidized, raw sample ————— OXIDIZED—Treated with 30 percent hydrogen peroxide and heated FIGURE 7 3.—Simi.lar titration-curve characteristics provided by different samples collected from well C—2, Clarion County, Pennsylvania. A, Sample collected September 18, 1984; and B, Sample collected October 31, 1984. (Fifty-milliliter samples titrated with 0.1 Normal sodium hydroxide.) ranges from 0.3 percent to about 10 percent. In general, as 002 increases, oxygen content decreases. In nearly all instances, the 002 content increases with soil depth because 002 is heavier than the other gases and there- fore sinks into the lower layers. Thus, the porosity and moisture content of the lower soil horizons and the underlying rock formation determine whether 002 ac- cumulates in the subsoil and unsaturated zone or whether it disperses into underground fractures or other passageways. SEDIMENTATION 155 SEDIMENTATION By RANDOLPH S. PARKER At the time of the initiation of the Surface Mining Control and Reclamation Act of 1977 (PL 95—87), there was an immediate interest in “natural” suspended- sediment production in the areas of coal mining and in the changes in suspended-sediment production resulting from the mining activity. Perhaps no other phase of the hydrologic cycle received such attention during 1978. One reason for this attention was the engineering re- quirements for designing structures to contain sediment produced on the mine site. Unfortunately, sufficient data were not available to appropriately answer many of the design questions. Therefore, the importance of suspended-sediment-data collection was established early in the program, but the actual collection of these data was extremely slow. The slow start can be attributed, in part, to the typ- ical issues faced in any sediment study. Streamflow- gaging sites designed to collect suspended-sediment data are very expensive and manpower intensive. Suspended-sediment concentrations are quite vari- able in time, and these concentrations usually are related to a variety of other hydrologic, climatic, and land-use factors. To understand these relations, many of these other factors also need to be monitored. This monitoring tends to increase the cost of sediment-data collection. Attempts to predict changes in the suspended- sediment concentration and load that occur because of coal mining need to rely substantially on the knowledge of how other components of the water balance change. Whether the infiltration rate increases or decreases in a reclaimed area directly affects the quantity of erosion and initial transport of sediment. Changes in the magni- tude of the evapotranspiration component by coal min- ing directly affects the volume of available surface runoff and that, in turn, affects available energy to detach and transport sediment particles. Knowledge of changes in the quantity of sediment are related direct- ly to the knowledge of changes in the other components of the water balance. Unfortunately, the changes in these components are not well defined. SEDIMENT—DATA-COLLECTION APPROACH AND CONSIDERATIONS There are many difficult decisions on where within a watershed to locate sediment-data collection sites. One approach, monitoring the outlet of a basin in which a variety of land uses occur upstream, enables only the definition of the sediment characteristics of the integrated effects of these land uses. A second ap- proach, monitoring segments of a watershed in order to partition the effects within a watershed, multiplies the cost of data collection and manpower needs. The first approach would not, it seemed, provide the necessary predictive capabilities. The second approach was far too expensive and manpower intensive to be practical. Certain problems are not unique to sediment-data col- lection or research. Results of disturbance by coal min- ing on the water resources within a watershed are difficult to study because of the tenuous nature of coal economics. Coal mining is subject to the economics of energy production. Coal is not mined and stockpiled, but it is removed and land is disturbed as the demand for coal increases. Because of this situation, there is little control of land use during the study. Because of the inability to control land use within a watershed dur- ing the period needed to collect hydrologic information, sediment studies have incorporated a multiwatershed approach. By using this approach, land-use changes and the response of the water resource are carefully documented. By compiling a sufficient number of dif- ferent basins and land-use situations, the response of the water resource to various land-use changes may be identified. This technique necessitates monitoring many different basins. Although there are some merits to the multiwater- shed approach, there are some disconcerting factors that indicate basic problems with this method. Primari- ly, there is the myriad of hydrologic responses possible within a basin because of the temporal interactions among coal mining and the weather and climate. For example, the effects of coal mining on sediment load during high flow are different than during low flow, and if particular stages of the mining activity are completed during the low-flow season, effects may be lessened. In addition, there is the spatial interaction between the ex- tent of mining activity and where this activity is located within the watershed. Thus, when the area of disrup- tion within a watershed is doubled, the sediment load does not necessarily double. Synoptic studies can supply much of the background information about an area and even indicate changes that occur because of mining activity. However, suspended-sediment data need to be collected during high flows, and this necessitates data collection only at certain times of the year. Rainfall simulators have been developed and tested on rehabilitated slopes in order to facilitate data collection (Lusby and ’lby, 1976), but 156 there are problems with the capability of this technique to model natural precipitation. There is another aspect to suspended-sediment data collection that concerns the disruption of downstream channels as a result of changes in sediment input upstream. An increase in suspended sediment can result in storage of the sediment in the downstream channel, which causes increased bank erosion. A decrease in sedi- ment from upstream can result in degradation of the bed and banks of the downstream channel. Downstream channel changes may occur from a single upstream disruption. Because effects can occur far downstream from changes in the sediment input upstream, difficul- ties in monitoring are increased. In Tennessee, for ex- ample, suspended-sediment yields have increased as much as 200 times since coal mining began. Instead of a stable, armored channel that was narrow and lined with mature trees, the channel downstream from min- ing became more than twice as wide and had numerous gravel bars, and the banks showed indications of recent erosion (Osterkamp and others, 1984). Examples of sediment storage and subsequent flush- ing occurs in various coal-mining areas in the Eastern United States. The number of years necessary to remove this stored material is unknown. In the Western United States, there is some concern that decreases in sediment resulting from entrapment in sediment ponds will result in bed and bank scour to downstream channels. SEDIMENT-DATA-COLLECTION ACTIVITIES Data collection of suspended sediment at streamflow- gaging sites has continued near many coal-mining sites throughout the United States. These sediment- data-collection sites have been funded through a variety of sources. Some of the early sediment-data collection began as a cooperative program between the U.S. Bureau of Land Management and the US. Geological Survey. During the mid-1970’s, the US. Bureau of Land Management needed to collect hydrologic data to assess the hydrologic effects of coal mining in Western States where Federal coal was to be leased. This cooperative program began as a long-term modeling effort, and a number of streamflow-gaging sites for sediment-data collection were established within small watersheds. The US. Geological Survey provided financial sup- port for collection of sediment data in 19 small basins in the Eastern and Interior Provinces (Kilpatrick and others, in press). This set of data provides information about sediment from rainfall-runoff events in the small basins from Pennsylvania to Alabama. These basins SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 ranged from 0.17 to 5.08 square miles in area and in- cluded basins with and without mining activity. Sedi- ment data were collected primarily from 1981 through 1983, although some basins have different periods of record. Additional funding of streamflow-gaging sites for sediment-data collection was provided by the US. Geo- logical Survey’s contract streamflow-gaging-sites program. This program established a network of streamflow-gaging sites in regions of actual or an- ticipated coal mining by contracting the operation of the sites and data collection to private companies. The purpose of this program was to serve as a system of environmental monitoring to assess the effects of coal mining and associated reclamation on the surface waters of the given region. Because of this purpose, many of the streamflow-gaging sites in this program were located downstream on larger tributaries and not downstream from individual coal-mining operations. In some regions, sediment data were available from previous work. Summarizing these data into general equations provided estimating techniques for analyz- ing effects from coal development. One such study was done in southeastern Montana (Lambing, 1984). This study examined sediment-yield data for 121 sites including reservoir sedimentation surveys, sedi- ment sampling at streamflow-gaging sites, and indirect estimation of sediment yields based on physical char- acteristics of the basin. Multiple regressions were used to estimate sediment yields for small ungaged basins. In many areas, sufficient data are not available to assess premining hydrology or to evaluate potential changes to the hydrologic regime as a result of mining. To provide the necessary hydrologic evaluation in these instances. some methods have been summarized (Frickel and others, 1981; Hadley and others, 1981; Shown and others, 1981, 1982). Methods described in- clude estimating erosion and sediment yield in the par- ticular area by using the Universal Soil Loss Equation and reservoir sedimentation surveys. Because of the interaction between the various com- ponents of the water balance and sediment movement, watershed models have been used to investigate this complex system. Such models enable sequences of algorithms to be linked together for water and sediment movement. For example, PRMS (Precipitation-Runoff Modeling System) has programs to detach soil particles from the upland slopes and transport sediment down- stream. In PRMS, sediment detachment and movement on the overland flow planes is done only during storms and is the sum of the rainfall detachment rate of sedi- ment and the overland flow detachment rate of sedi- ment (Leavesley and others, 1983, p. 37). Each of these SEDIMENTATION components is computed using equations described by Smith (1976) and Hjelmfelt and others (1975). The sedi- ment computed from the overland flow planes is trans- ported through the channel system as a conservative substance. Sediment detachment and deposition in the channel are not included in the model. Thus, sediment removed from the overland flow planes is transported directly through the channel network to the mouth of the system. Each of these components has a number of parameters that need to be determined. Hydrologic data collected from two watersheds have been used in the PRMS modeling system in Wyoming. Modeling has been done for surface water and sediment (Rankl, 1987). One of these watersheds is a natural basin, and the other is an artificial drainage basin established on a reclaimed area. Rankl uses the model to help identify sediment-source areas within the water- sheds and to provide a basis for comparison between the mined and unmined areas. Modifications to PRMS have been done to calculate scour through rilling on disturbed areas and deposition within diversion terraces (Reed, 1986). Within this same modification, there is an allowance to route eight differ- ent size classes of sediment depending on the distribu- tion of the soil material and to route the water-sediment mixture through a sediment pond. These algorithms have been tested on several watersheds in Pennsylvania. Use of models undoubtedly will increase because they enable the user to examine many components of the water balance and to identify their interactions. These interactions become extremely important when at- tempting to identify changes that occur from mining activity. Improvements in PRMS have been suggested by Carey and Simon (1985). A sizable data base for use in modeling has been developed as a result of the cooperative efforts between the US. Bureau of Land Management and the US. Geological Survey, but analysis of this data base has only just begun. EASTERN PROVINCE Studies done in the Eastern Province indicate sedi- ment yields that range from 23 to 900 tons per square mile per year (Harkins and others, 1981; Quinones and others, 1981; Ehlke and others, 1982b). Although this is a fairly large range in sediment yields, the disruption of a watershed by human activities increases sediment yields dramatically. Basically, these increases are the result of removal of the protective cover of vegetation. There are problems with identifying the changes in sedi- ment yield associated with a particular land use. In many basins, coal mining occurs concurrently with log- ging, agriculture, road construction, and urban development. 157 There are several examples of the changes in sediment yield because of coal-mining activity, although more data are needed to provide a better definition of the ac- tual changes. In West Virginia, a comparison of two basins, one mined and the other unmined, indicated the mined basin had a sediment yield 240 times greater than the unmined basin within a unit-discharge range of 2 to 10 cubic feet per second per square mile (Ehlke and others, 1982a). Another comparison in West Virginia was between an unmined basin and a basin where 20 percent of the area was mined. The unmined basin had a sediment yield of 730 tons per square mile, at a unit discharge of 10 cubic feet per square mile. The mined basin had a sediment yield of 3,650 tons per square mile for the same unit discharge (Ehlke and others, 1982b). In coal area 16 in Virginia and Tennessee (pl. 1), a com- parison of an unmined basin with a basin in which 9 per- cent of the watershed has been disturbed by mining indicates a substantial increase in sediment yield. At a unit discharge of 3 cubic feet per square mile, the un- mined basin had a sediment yield of 7 tons per square mile. At the same unit discharge, the mined basin had a sediment yield of 1 10 tons per square mile. Additional- ly, the unmined basin had sediment concentrations that ranged from 1 to 7 milligrams per liter, whereas con- centrations in the mined basin ranged from 34 to 1,030 milligrams per liter (Hufschmidt and others, 1981). In another comparison in Tennessee, the mined basin of New River (382 square miles) had 20 times more sedi- ment during 1 year than the contiguous mined basin of Clear Fork (272 square miles) (Parker and Carey, 1980). This difference occurred even though the drain- age area of New River is not that much larger than the drainage area of Clear Fork, and even though only about 7 percent of the New River basin is mined as compared to 1 percent of the Clear Fork basin. Hubbard (1976) reported sediment yields of 300,000 tons per square mile in an extensively mined area of Alabama. Undisturbed areas in this region of Alabama have sediment yields that range from 20 to 800 tons per square mile (Harkins and others, 1981, p. 40—41). These reported differences in sediment yields indicate that sediment concentration and sediment yield- in- crease with the onset of mining. These increased sedi- ment yields may be of short duration because of reclamation. In addition, the sediment concentrations that occur in the smaller disturbed basins often are decreased downstream by dilution from larger receiv- ing streams (Harkins and others, 1980). Flint (1983, table 12, p. 32—33) computes a regional average sediment yield (in tons per square mile) for the Eastern Coal Field of Kentucky by using 28 selected streamflow-gaging sites. The computed regional aver- age for sediment yield is 486 tons per square mile, and 158 the standard deviation is 460 tons per square mile. This regional average is shown in figure 74; the shaded area represents plus and minus one standard deviation of this regional average. In this same coal region, Curtis and others (1978) monitored a number of extensively mined basins. Comparing eight of these basins during January 1, 1974 through December 31, 1975 with respect to area disturbed by coal mining indicates a definite trend (fig. 74). The sediment yield from these basins ranged from 732 to 21,000 tons per square mile Where more than 3 percent of the area is disturbed, there is a rela- tion between area disturbed and sediment yield that can be described in the power form of the equation: Y = 370 01-50, (36) where Y = mean annual sediment yield, in tons per square mile; and D = area of the basin disturbed, in percent. Basins that have less than 3 percent of the area dis- turbed by mining may be within the variability of the sediment yield of the region (fig. 74). Perhaps there is a threshold of disruption within these basins below which little evidence of sediment-yield increase can be identified. If this is true, more data will be needed to adequately identify these relations. In other basins in Kentucky, the change in sediment yield with changes in discharge indicates that the rela- tion shown in figure 74 may not be so direct. In Ken- tucky, comparisons of the Red River and Goose Creek in coal area 14 to the extensively mined Troublesome Creek in coal area 14 indicate that during low and medium flows, substantially more sediment is generated from the mined basin than from the unmined basins, but that during high flows, less sediment is generated from the mined basin than from the unmined basins (Quinones and others, 1981, p. 54—55). This determina- tion is made from a comparison of the sediment-rating curves for these basins, and these curves are defined using 30 to 40 data values. Therefore, these determina- tions are somewhat tenuous, but they do indicate addi- tional factors that need to be defined and studied to fully understand sediment movement. NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES This area encompasses a variety of climatic regimes, and these climates affect sediment movement. In the northern areas from Montana and North Dakota through the central Rockies of Colorado, stream dis- charge results primarily from snowmelt. In this type of environment, the larger stream discharges occur SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974—84 1001000 _ | I I I I I | I I I I | | | I I I l I | I 'i.‘ ' — 2 Lu _ a: 10,000 : —: ‘3’: : o : O - _ (n - . o _ a: - _ Lu 0. h . _ (I) g o _ . _ |_ 1,000 :——__——___:___——___———_._—______: Z I I —~ _ Regional mean _ D — _ _I u_.| - _ : — .5 — E E a E 100 _— 1, __ a I 'E 2 LU : ‘1? : (n _ g _ _ U, _ 10 I l l I | I I II I I I I I I | I] I | | l I l | I 0.1 1 10 100 AREA DISTURBED, IN PERCENT FIGURE 74.—Relation between percent of area disturbed and sedi- ment yield for the Eastern Coal Field of Kentucky (modified from Curtis and others, 1978, p. 18). A comparison is made with the regional mean sediment yield computed by Flint (1983, p. 32). during the snowmelt season of March through July, and most of the sediment is transported during this period. Frozen soils that exist at the onset of snowmelt can in- hibit infiltration and lead to increased streamflow, but this condition can prevent soil movement and thus decrease sediment input (Slagle and others, 1984). Differences identified in sediment yields in these snowmelt regions generally are ascribed to changes in geology. For example, two areas mapped in Montana with respect to sediment yield have three categories: less than 10, 10 to 100, and greater than 100 tons per square mile. These three categories are mapped primari- ly with respect to geology (Slagle and others, 1983; Slagle and others, 1984). In parts of Montana, Wyoming, and northern Colo- rado, the form of the precipitation is affected by altitude. At the lower altitudes, intense summer thun- derstorms are important in sediment transport. At the higher altitudes, sediment transport is related to stream discharges that occur during the snowmelt season. The effect of altitude in ultimately affecting sediment move- ment is increased because there usually is a change in SEDIMENTATION the geology with increasing altitude. An example of this situation occurs in coal area 54, an area along the border between Wyoming and Colorado (Kuhn and others, 1983). The lower altitude areas have a cover of short prairie grass and the bedrock is sandstone, siltstone, and shale. The higher altitudes have a forest cover, and the bedrock is granite, schist, and gneiss. Estimates for the sediment yields for these two areas are reported in acre-feet per square mile. In order to maintain common units for sediment yield in this report, it is assumed that the sediment has a specific weight of 80 pounds per cubic foot (Guy, 1970, p. 33). Given this assumption, the higher altitudes have sediment-yield estimates of 1.7 to 44 tons per square mile, while the lower altitudes have estimates of 17 to 190 tons per square mile. These ranges are different by an order of magnitude. Further to the south, in the coal regions in Colorado, Utah, and New Mexico, most of the annual precipita- tion occurs as intense summer thunderstorms. These storms are infrequent but can move large quantities of sediment. Again, geology is a major factor in affecting the differences in sediment yields. For example, in a study in central Utah, estimates of sediment yield range from 3 to 5,200 tons per square mile. The larger yields are from the lowlands where the bedrock is predomi- nantly shale and sandstone. The higher altitudes con- sist of the Wasatch Plateau and the Book Cliffs where the bedrock is primarily sandstone, shale, and lime- stone, and the estimates of sediment yield generally are smaller (Lines and others, 1984). Because of the infrequent nature of precipitation in areas of intense summer thunderstorms, sediment yields calculated from observed data are rare. Instead, 159 many estimates for sediment yield are derived from in- direct methods that account for sediment-yield changes based on geology, slope, vegetation, and land use. An example of such a method is the PSIAC method (Pacific Southwest Inter-Agency Committee Water Manage- ment Subcommittee, 1968). Results from this method are reported in acre-feet per square mile and are con- verted in this report to tons per square mile using the above assumption of a specific weight of 80 pounds per cubic foot. Using the PSIAC method in northwestern New Mex- ico, the area was divided into three categories: (1) Mesa tops, clinker hills, and dry lake beds, which have sedi- ment yields as much as 700 tons per square mile; (2) low- altitude badlands and gullied alluvial plains, which have intermediate sediment yields of 700 to 2,000 tons per square mile; and (3) moderate-to-steep badlands, which have sediment yields estimated at 2,000 to 5,500 tons per square mile (US. Bureau of Land Management, 1977b). In southeastern Colorado, movement of sediment pri- marily is produced by intense summer thunderstorms. The geology of an area dramatically affects changes in sediment concentration and sediment load. Headwater streams transport substantially less sediment where they drain areas of limestone and igneous and metamor- phic rock of Precambrian age. In lower reaches under- lain by the Raton and Poison Canyon Formations, suspended-sediment concentrations range from 50,000 to 200,000 milligrams per liter during peak flows, and sediment yields can range from 300 to 2,840 tons per square mile (Abbott and others, 1983). 160 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 AQUATIC BIOLOGY By DAVID A. PETERSON Aquatic biota often are used as indicators of the qual- ity of the aquatic environment, and this quality may indicate the effects of coal mining. The biota are adapted to the environmental conditions in the stream; different conditions support different biota. Because their lifespan ranges from months to years, the organ- isms inhabiting streams reflect conditions within the recent past that might not be discernible in an instan- taneous sample. The use of benthic invertebrates and algae as water-quality indicators has been well docu- mented (Wilhm and Dorris, 1966; Larimore, 1974; Lowe, 1974). During water-quality studies of coal-mining areas, the US. Geological Survey has sampled benthic invertebrates more frequently than algae. Benthic invertebrates that occur in streams generally are immature insects such as mayfly nymphs (Ephem- eroptera), dragonfly and damselfly nymphs (Odonata), caddisfly larvae (’Irichoptera), and midge, blackfly, deer- fly, and other true fly larvae (Diptera). Some insects such as beetles (Coleoptera) and true bugs (Hemiptera) can be aquatic during both immature and adult stages. Aquatic invertebrates that are not insects include snails, leeches, aquatic earthworms, and crustaceans. Benthic inverte- brates occur within or on the substrate of the stream bottom, including submerged objects such as vegeta- tion and logs. Drifting invertebrates are those that are temporarily suspended in the streamflow current. Two common methods for sampling benthic inverte- brates are the Surber-sampler and the kick-net methods. The Surber sampler delineates an area of 1 square foot; substrate within the square-foot area is agitated caus- ing invertebrates to float downstream into a catch net. The kick net is a dip net, hand held downstream from the area where the user is agitating the substrate by kicking. Surber samplers are used for quantitative studies, whereas kick nets are used for qualitative studies. An Ekman grab sampler with spring-loaded jaws is useful for sampling pools. Drifting invertebrates in streams are sampled by placing a net in an undis- turbed riffle for a given length of time. Algae sometimes are collected during water-quality studies from either the periphyton or the phytoplank- ton. Periphyton refers to algae attached to the substrate, whereas phytoplankton refers to algae suspended in water. Through photosynthesis, algae and larger aquatic plants are the primary producers in the aquatic food chain; they also provide habitat for other aquatic organ- isms. Periphyton commonly are sampled by scraping natural substrates or by placing artificial substrates in the stream for 30 days or more. Phytoplankton are collected by immersing a water-sampling bottle in the water at the desired location and depth. BIOLOGICAL STUDIES IN COAL PROVINCES Biological data have been reported for some coal areas in the United States. Locations of coal areas that have coal-area hydrology reports that include discussion of benthic invertebrates and (or) algal data are shown in figure 75. Several interpretive studies done in these provinces are described in the following sections. In addition, biological surveys done by State agencies are described in some of the coal-area hydrology reports for these provinces. The reports about coal areas 25—42 in the Interior Province generally did not contain biological information. EASTERN PROVINCE Effects of coal mining on aquatic biota are much dif- ferent in the Eastern United States than in the Western United States because of the occurrence of acid mine drainage in the Eastern United States. Indicators of acid mine drainage are concentrations of dissolved sulfate larger than 75 milligrams per liter, total iron and manganese concentrations larger than 500 micrograms per liter, pH less than 6.0, and acidity concentrations larger than alkalinity concentrations (US. Department of the Interior, 1968). Effects of acid mine drainage on benthic invertebrates were noted during biological reconnaissance in several Eastern States from 1979 to 1980. Kick-net samples col- lected from some of the streams in Pennsylvania yielded no benthic invertebrates, whereas others did not con- tain a biological community (two or more species of ben- thic invertebrates), as defined by the US. Office of Surface Mining (1979). Inverse relations between numbers of benthic-invertebrate taxa and indicators of acid mine drainage were noted. Herb and others (1981b, p. 48) determined that streams in coal area 5 (fig. 75) that had no benthic invertebrates had a mean dissolved- sulfate concentration of 162 milligrams per liter, whereas streams that had five or more benthic inverte- brate taxonomic orders had a mean dissolved-sulfate concentration of 24 milligrams per liter (significant at the 95 percent confidence level). Similar reports on coal areas 1, 2, 3, and 5 (table 1) also noted poorly developed (diversity and size) benthic-invertebrate communities associated with acid mine drainage. AQUATIC BIOLOGY NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES NORTH DAKOTA NEBRASKA COLORADO KANSAS ARIZONA NEW MEXICO OKLAHOMA EXPLANATION REPORT INCLUDES BENTHIC— INVERTEBRATE OR ALGAL DATA REPORT DOES NOT INCLUDE BENTHICv INVERTEBRATE OR ALGAL DATA 161 EASTERN PROVINCE INTERIOR PROVINCE WESTERN EASTERN REGION REGION NO REPORT PREPARED FOR COAL AREA —-—13— COAL-AREA BOUNDARY AND NUMBER FIGURE 75.-—Locations of coal areas that have coal-area hydrology reports that include discussion of benthic invertebrate and (or) algal data. The benthic-invertebrate surveys in Tennessee and adjacent States during 1980—81 indicated that water quality of the streams ranged from excellent to poor. The Shannon-Weaver diversity index and a biotic index were used to describe the health of the benthic- invertebrate communities in coal areas 16, 17, 19, 20, and 21 (table 1). The Shannon-Weaver diversity index measures the diversity of the community; larger values indicate diverse, well-balanced communities under a minimum of stress (pollution); small values indicate stressed communities, dominated by a few taxa (Wilhm and Dorris, 1966). The values of the diversity index ranged from 1.81 to 3.82. The biotic index measures the benthic-invertebrate community using pollution- tolerance ratings of the taxa. Statistical techniques were used to determine the ef- fects of water quality in streams of Tennessee on the benthic-invertebrate communities. Using analysis of variance procedures, Bradfield (1986b) reported signif- icantly fewer taxa, smaller densities and diversities, and a larger percentage of Diptera in streams that have pH values less than 6.0 and streams that have relatively large values of dissolved constituents and specific con- ductance (associated with land-use effects, including coal mining), than in streams that have pH values larger than 6.0 and relatively small values of dissolved con- stituents and specific conductance (relatively not af- fected). The dissolved-constituent concentrations were calculated using mean concentrations of manganese, iron, and sulfate. The decreased number of taxa, densi- ty, and diversity adversely affects the ability of the benthic-invertebrate community to process instream detritus and decreases the food source for higher trophic levels. Relations between the characteristics of the benthic- invertebrate communities and water quality also were tested using multivariable regression techniques (Brad- field, 1986b). Correlation coefficients between number of taxa and density with dissolved-constituent concen- trations, specific conductance, and pH were less than 162 0.55. This indicates that the selected water-quality variables accounted for less than 30 percent of the variability in the number of taxa and density. The cor- relation is poor partly because community characteris- tics are not linearly related to water-quality variables. Dissolved-manganese concentrations accounted for more of the benthic-invertebrate-community variabil- ity than the other water-quality variables. Dissolved- manganese concentrations were negatively correlated with number of taxa (fig. 76), density, diversity, and per- cent Ephemeroptera; the dissolved-manganese concen- trations were positively correlated with percent Diptera. Dissolved-manganese concentrations, number of taxa, and percentage of Diptera and Ephemeroptera may be useful indicators of the effects of coal mining on streams in Tennessee. In addition, nonparametric cluster anal- ysis of benthic-invertebrate samples collected from Ten- nessee streams during 1982—83 indicate a species group of Ephemeroptera that may be useful in identifying the effects of coal mining (Bradfield, 1986a). NORTHERN GREAT PLAINS AND ROCKY MOUNTAIN PROVINCES The effects of coal mining on water quality of streams in the Northern Great Plains and Rocky Mountain Provinces, such as increased concentrations of dissolved solids and turbidity, have the potential to affect stream biota. Changes in the quantity or timing of water in a stream also may affect the biota, because much of the coal in the Western United States is mined in semiarid areas. BENTHIC INVERTEBRATES IN STREAMS The relation between specific conductance and benthic-invertebrate communities was examined by Klarich and Regele (1980). They reported an inverse relation between diversity-taxa richness of the benthic- invertebrate community and specific conductance in streams in southeastern Montana. Stream communities were divided into categories of large, moderate, and small diversity-taxa richness, based on Margalef and Shannon-Weaver diversity indices and the ratio be- tween the number of taxa collected and the number of taxa expected for a given number of samples. Streams with large diversity-taxa richness (relatively good qual- ity) had a mean specific conductance of 1,135 micro- siemens per centimeter, compared to streams with moderate diversity-taxa richness that had a mean specific conductance of 2,580 microsiemens per cen- timeter, and streams with small diversity-taxa richness that had a mean specific conductance of 3,215 micro- siemens per centimeter. Benthic-invertebrate density SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 50 I llllllll I llllllll lllllllll I II o _ . _ o LLI o a" 40— . o o . _ E o < _ ' ' _ tn : . ‘ .’ I o o- o o E 30— C. O. . O O . . _ o u o. O . O.. 0: . E o o n o o _ — o o 00.0 no to o- < o u a. O. n l— .3 , r: :2. .3 . 3 (”5201. :.:...'...'. — o to on... o. D: . a .. on . .oo . o m —.. . I . : O ...C. Q. _ g . : .=:: ' . . D 10 —. : .. o. . :0: o — Z : 3 .:o : so a o. o o so on o _ C. .I. I. . . :C _ O C. O O o : . o o a 0 I llllllll I lllllll I I PIIQIll l 'Illlll 1 10 100 1,000 10,000 DISSOLVED MANGANESE, IN MICROGRAMS PER LITER FIGURE 76.—Relation between number of benthic-invertebrate taxa per sample and mean concentration of dissolved manganese (from Bradfield, 1986b). The negative slope of the regression line indicates that the number of taxa per sample decreases as the dissolved- manganese concentration increases. was not related to specific conductance. These findings were based on Surber samples collected from natural substrates in 35 streams. Klarich and Regele (1980) also discussed the equitability and percentage similarity of the benthic-invertebrate communities and compared samples from natural and artificial substrates. Benthic-invertebrate communities in the southern Powder River basin, the major coal-mining area in northeastern Wyoming, were examined during 1980 and 1981. Average benthic-invertebrate density and median Shannon-Weaver diversity in the Belle Fourche River, an ephemeral stream, were smaller upstream from a coal mine than downstream from the mine (fig. 77). Dif- ferences between the sites may be the result of inter- mittent flow from a tributary, diversion of the stream around the mine, infrequent discharge pumped from the mine pit, or a combination of these factors (Peterson, in press). The samples were collected from pools at the two sites using an Ekman grab sampler. In this study, the benthic invertebrates were identified to the genus level; identifications to species level are necessary to better ascertain whether community-composition changes occurred that were not detected at the genus level. Effects of coal mining on water quality and benthic invertebrates, when present, may be difficult to sepa- rate from other factors such as agriculture. In north- western Colorado, Britton (1983) studied benthic- invertebrate communities in six streams upstream and AQUATIC BIOLOGY 3,000 3 UPSTREAM 2,500 — _ 2,000 — — 2 1,500 — — 1 ,000 500 3,000 3 DOWNSTREAM 2,500 SHANNON-WEAVER DIVERSITY INDEX 2,000 1,500 NUMBER OF BENTHIC INVERTEBRATES PER SQUARE FOOT 1,000 500 0 o o o o o o o o ‘— .— eo~ coa co co on no no co co co ouc- mN c» CDA m- m a) m a)- a: '12 ‘12 F. F." P.“ P. F.“ ta '_" ta no ‘—" .— ‘ BE “.32 2'3 §E 3% 5 F2 "2 9% 2% >a >5 Lu [—5 |— a: ma. CED. E IE U) 0’: 2 U) I g— gv 3 35 m Ed gfi $§ 5:98 23 a 0 o E 29 $93 <' <" :3 D “I: ”J 3 E E < < B] 8 g m D EXPLANATION DIPTERA OTHERS EPHEMEROPTERA O————-—O DIVERSITY TRICHOPTERA FIGURE 77.——Composition, density, and diversity of benthic invertebrates from the Belle Fourche River, upstream and downstream from a coal mine in northeastern Wyoming (from Peterson, in press). 163 164 downstream from coal mines. Benthic-invertebrate com- munities changed in the downstream direction, but the changes could not be attributed solely to mining. For example, benthic-invertebrate density and specific con- ductance in Trout Creek were larger downstream from coal mining (sites Tr-2 and Tr-3) than upstream from mining (site Tr-l), but a small change in substrate unrelated to coal mining also was noted. Community composition changed in the downstream direction from a well-balanced community of several functional groups to a community composed largely of Dipteran Chiro- nomids (fig. 78). 0.0-3.7 0.8—3.7 SITE Tr-I SITE Tr-2 SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Changes in streamflow type such as those that might be caused by addition or depletion of water because of coal mining may result in changes in the composition of the benthic-invertebrate community. The relation of flow type to the benthic-invertebrate community was reported by Peterson (in press) in a study of perennial, intermittent, and ephemeral streams in northeastern Wyoming. The benthic-invertebrate communities that inhabit perennial streams generally were similar to each other because of the taxa adapted to flowing water, and communities of ephemeral streams generally were similar to each other because of the taxa adapted to EXPLANATION RANGES IN PERCENTAGE COMPOSITION FOR ALL SAMPLING DATES 3.9-43.3 Ephemeroptera (mayflies) Diptera (true flies) Trichoptera Icaddisflies) Coleoptera (beetles) . Hydracanna (water mites) Oligochaeta (worms) SITE Tr—3 FIGURE 7 8.—Change in mean percentage composition of benthic-invertebrate taxonomic groups in the downstream direc- tion in Trout Creek, northwestern Colorado (from Britten, 1983). The sites are labeled in downstream order, from Tr-l to Tr-3. AQUATIC BIOLOGY standing water, based on J accard coefficients of com- munity similarity and a cluster diagram. Benthic- invertebrate communities of intermittent streams did not form a comparable cluster. INVERTEBRATE DRIFT IN STREAMS Invertebrates drifting in streams have been studied because of their potential importance in estimating time for invertebrate recolonization of streams reclaimed following mining and also as a measure of the inverte- brate community. Rates of drift generally were larger in perennial streams than in intermittent or ephemeral streams of northeastern Wyoming, indicating that a reclaimed or disturbed reach in a perennial stream would be recolonized faster than a reclaimed reach of an intermittent or ephemeral stream (Peterson, in press). A study of invertebrate drift in two streams of northeastern Wyoming during 1977 indicated behav- ioral (voluntary) drift of invertebrates in a stream at normal stage, compared to catastrophic drift in a stream flooded by rainfall (Wangsness and Peterson, 1981). The catastrophic drift contained many types of invertebrates that normally do not occur in drift. The catastrophic drift may aid in distributing invertebrates, as well as alleviating the effects of scour during floods. P HYTOPLANKTON AND P ERIPHYTON In Scofield Reservoir, Utah, concentrations of mer- cury originating from coal particles have caused con- cem for water users but have not exceeded current State standards (Stephens, 1985). The study of the effects of coal mining on the reservoir was done in coop- eration with the US. Bureau of Land Management. N onpoint sources contributed most of the pollution to the reservoir, particularly phosphorus and nitrogen. 165 Large nutrient concentrations were associated with late summer blooms of the bluegreen algae Aphanizomenon flos—aquae and Anabaena flos—aquae and resultant fish kills. Biota of strip-mine ponds abandoned 20 years previ- ously were compared to native ponds in northeastern Wyoming by Wangsness (1977). Phytoplankton, periphyton, and benthic-invertebrate communities were less diverse in strip-mine ponds than in native ponds. Sloughing of the banks of strip-mine ponds may have been a factor in the differences. Samples of phytoplankton and bacteria were collected from 12 reservoirs in eastern Montana during a study done in cooperation with the US. Bureau of Land Management (Ferreira and Lambing, 1984). The study evaluated the suitability of the reservoirs for fish propa- gation, waterfowl habitat, livestock watering, and recreational use. Periphyton in streams of southeastern Montana were surveyed by Bahls (1980). Specific conductance within the range of 239 to 6,400 microsiemens per centimeter did not seem to have an overriding or adverse effect on the structure of periphyton communities. At values less than 6,400 microsiemens per centimeter, diatom diver- sity increased with specific conductance. Previous work in Montana by Miller and others (1978) indicated decreased diatom diversity associated with saline seeps and specific conductance values greater than 7,500 microsiemens per centimeter. A study in southeastern Montana (Bahls and others, 1984) demonstrated the use of diatoms of the periphyton and phytoplankton as indicators of water quality. Many of the diatom species are useful as in- dicators of dissolved solids, suspended sediment, and temperature, which are three variables likely to be af- fected by surface mining and related activities (Bahls and others, 1984). SUMMARY AND ADDITIONAL DATA NEEDS By BRUCE P. VAN HAVEREN5 and DONALD A. GOOLSBY During the early 1970’s, knowledge of hydrology and mining relations was obscured by uncertainties and, perhaps, a few misconceptions. For example, many scientists surmised that large volumes of water would be needed for reclamation processes in addition to that derived from natural precipitation. Today (1985), near- ly all reclamation is done without supplementary water. Predicted large-scale withdrawals of surface water for use in coal-fired powerplants and coal-gasification and liquefaction operations have not materialized because these industries have not developed in the Western United States at the rate once expected. During 1973, the National Academy of Sciences did a study of the rehabilitation potential of coal lands in the Western United States (National Academy of Sciences, 1974). In their report, the National Academy of Sciences discussed the adequacy of hydrologic data in the semiarid Western United States. The report stated that, for the areas being considered for coal development, serious gaps existed in the quantitative data about surface-water quality and aquatic biota. In addition, streamflow data for tributary streams and ground-water data were seriously deficient in the semiarid coal regions of the Western United States. The National Academy of Sciences (1974) report indicated the need for two distinct levels of hydrologic investigations—regional and site specific. Both levels of investigation were incorporated into the U.S. Geolog- ical Survey and U.S. Bureau of Land Management coal- hydrology program. Regional-scale appraisals would enable evaluation of the effects of several proposed mining operations in a large drainage basin and would include bench-mark stations for long-term monitoring of hydrologic changes. Site-specific investigations would answer questions about hydrologic changes that result from specific mining operations. Site-specific data also were needed to design measures to mitigate any onsite or offsite effects on water resources and to prepare reclamation plans. By 1985, the occurrence of water resources and natural hydrologic processes in the major coal regions were better understood, and the general direction and magnitude of hydrologic changes resulting from surface mining can now (1985) be predicted more accurately. Throughout all coal regions, the most significant ac- complishment of the coal-hydrology program was the expansion of the hydrologic data base. That data base, in addition to results from site-specific hydrologic 5U.S. Bureau of Land Management. investigations, provided for a much greater understand- ing of hydrologic processes in coal regions. The surface- and ground-water hydrology of small areas (potential coal-lease tracts) was investigated thoroughly in the Green River and Hams Fork, Powder River, Bighorn Basin, Wind River, and Fort Union coal regions. Specific attention was focused on small water- sheds less than 30 square miles in area. The results of those investigations aided development of mining and reclamation “best management practices” that max- imize postmining landform stability and minimize disruption of surface hydrologic processes. In addition, much was learned about the hydrology of alluvial fill in small headwater valleys. This information has con- tributed to a better understanding of the hydrologic function of alluvial valley floors. Site-specific studies done in Montana indicated that wells and shallow aquifers removed by mining may not be the only available water supply; wells drilled to deeper, unaffected aquifers could replace water supplies used during mining. Where saturated coal beds are mined, substantial problems may occur because of aquifer dewatering and leaching of soluble material from mine spoils. In the larger watersheds, more than 30 square miles in area, attention was focused on characterizing surface- water and ground-water chemical quality. Because much of this data collection has preceded any extensive development of coal resources, an important data base now exists. These baseline data will be invaluable for future monitoring of effects of coal development on water quality. Studies indicate that local decreases in water tables will occur within 1 to 2 miles of surface mines, but regional effects of coal mining on ground water are unlikely. The largest regional effect on water resources in the Powder River basin probably will be caused by increases in population and land-use changes, rather than by mining activities. An accomplishment of the coal-hydrology program applicable to all coal regions was the increased knowl- edge of the areal extent and hydrogeologic properties of key aquifers. A better understanding of the relation between coal and ground-water systems now (1985) ex- ists. For example, studies indicate that the alluvium of the Powder River valley is not a significant aquifer, and that the regional movement of ground water in shallow aquifers in northeastern Wyoming is not as important as local movement. Primarily, the local flow systems within shallow aquifers will be affected by mining rather 167 168 than the regional flow systems. In all the coal regions, the general effects of mine dewatering now (1985) can be predicted. In the Fort Union region, the principal geochemical processes that affect ground-water quali- ty were identified, and the effects of mining on these processes Were determined. For many surface-mining situations, the length of time that water of deteriorated quality will exist in mine spoils can be determined. In terms of funds expended, nearly 15 percent of the total coal-hydrology-program funds was spent on devel- oping predictive methods and simulation models. This effort was coincident with the general increase in the use of computers for large-scale, data-base manipula- tion and system simulation. For example, equations were developed using channel geometry and basin- characteristic techniques to predict average annual runoff. Methods to indirectly determine peak discharge also were developed for all the coal regions in the Western United States. A model to simulate surface hydrologic processes in small watersheds, including ‘ sediment yield, was developed and calibrated for use in the Western, Fort Union, Powder River, Bighorn Basin, Wind River, Green River, Hams Fork, and San Juan River regions. In the Eastern coal province, substantial progress was made in determining specific relations between min- ing and hydrologic processes. For example, mining generally decreased peak flows and increased low flows. The discharge of conservative chemical constituents from small streams in the Appalachian regions was cor- related with the age of coal mines in the contributing watersheds. A study in Alabama indicated that the dissolved-solids concentration will reach a maximum after about 7 years of mine operation and return to premining levels after about 15 years of mine operation. The effects of mining on stream-water quality were less in mined basins that were reclaimed under present reclamation laws than in basins reclaimed under previous systems. One of the most important long-term issues associ- ated with surface mining in the Western United States is the potential for degradation of water quality by leaching of soluble salts from mine spoils. Water samples obtained from mine spoils in Colorado and Montana have indicated consistently larger dissolved- solids concentrations than samples obtained from nearby undisturbed aquifers. Salinity models were developed to predict the effect of leachates from mine spoils on the dissolved-solids concentration of receiv- ing streams. Modeling of dissolved solids in Rosebud Creek, Montana, has indicated that irrigation return flow accounts for a larger percentage of dissolved-solids loading than does mining at present (1985) levels. However, during full-scale mining, the percentage due SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 to mining would exceed that of irrigation. Most of the geochemical processes that result in increased dissolved-solids concentrations occur in the unsaturated zone. Therefore, it would be advantageous to minimize vertical infiltration of water through mine spoils. Acid mine drainage is not expected to become a ma- jor issue in the Western United States because coal from that area typically has a small sulfur content. However, coal regions in the Eastern United States have considerable acid mine drainage due to a larger sulfur content in the coal, larger precipitation quan- tities, and decreased alkalinity concentrations in the soil and overburden. The soundness of coal-leasing decisions that involve water resources depends, in part, on the technical ade- quacy of the hydrologic data used in support of those decisions. Decisions based on insufficient or poor qual- ity data have a great degree of uncertainty about them. The ability to make sound leasing decisions, in which water-resource issues are at stake, has improved as a result of the coal-hydrology investigations. At the land- use planning stage, the US. Bureau of Land Manage- ment uses hydrologic data to identify water-resource values and potential water-related issues. Prior to 1975, it was difficult to predict even the nature of potential coal development and water-resource issues; the direc- tion and magnitude of the effects on water resources could not be predicted at that time. By 1980, a substan- tial hydrologic data base had been acquired in the coal regions. Although some monitoring data were available from existing mines, the US. Bureau of Land Manage- ment still did not have a definitive concept of probable effects, particularly ground-water related effects. These data uncertainties contributed to the deferral of many land-use planning decisions (US Congress, 1984). However, by 1983 the data collected and the investi- gations done during the previous 8 years had begun to produce valuable information that was sufficient to identify and address the major water-resource issues related to coal leasing at the land-use planning stage. For example, baseline water-quality data for the eastern half of Wyoming now (1985) are considered adequate for assessing effects of future mining. One of the major decision points in Federal coal leas- ing has been the unsuitability criteria, which are envi- ronmental suitability tests by which all potential coal-lease tracts are reviewed before lease approval (US. Bureau of Land Management, 1985). Four of the twenty criteria apply to water reSources: flood plains, municipal watersheds, National resource waters, and alluvial valley floors. Identification and mapping of flood plains is a detailed, site-specific process, best suited for analysis at the mine-planning stage. Flood-frequency analysis techniques were developed that enable determination SUMMARY AND ADDITIONAL DATA NEEDS of the 100-year flood flows for streams, including ephemeral drainages, in the major coal regions. Further- more, sufficient methods now exist to predict the effects of surface-mine development on downstream flood plains. Municipal watersheds may not be mined unless it is determined that mining will not impair the quantity or quality of a public water-supply system (US. Bureau of Land Management, 1985). Sufficient information is available for all the coal regions to evaluate the poten- tial effects of surface coal mining on a municipal water- shed. Because detailed development data are needed for hydrologic analysis purposes, those evaluations would need to be made at the mine-planning stage. National resource waters are high-quality surface waters identified in State water-quality-management plans. Mining is to be excluded from Federal lands hav- ing such waters. A one-quarter mile buffer zone also is to be included, but this zone may be decreased or eliminated if the management agency can show it is not necessary for protection of the resource. Again, from a hydrologic standpoint, this situation needs to be analyzed in detail using specific mine plans. Disruption of the surface- and ground-water hydrol- ogy of alluvial valley floors is of particular concern in the Powder River, Bighorn Basin, Wind River, Green River, Hams Fork, Fort Union, Denver, and Raton coal regions. Alluvial valley floors often support grass-hay production that is crucial to ranching and farming operations. Water is supplied either by flood irrigation or by natural subsurface means. Land owners are con- cerned that any disruption of hydrologic systems that are associated with alluvial valley floors will decrease the productivity of their hay meadows. Given detailed mine plans, hydrologists should be able to predict with reasonable certainty the hydrologic changes that may occur because of surface mining. However, at this time (1985), sufficient information about the hydraulic prop- erties of mine spoils is not available to enable the pre- diction of hydrologic effects of mine reclamation and the resultant consequences for an alluvial valley floor system that depends on shallow aquifers for its subirrigation. In summary, it seems that the hydrologic data base is adequate in most of the coal regions to address water- resource issues at the land-use planning and lease. approval stages. However, less information is available to prescribe specific regulations for water-resource pro- tection at mine sites. In addition, the ability to predict cumulative effects of and prescribe regulations for multiple coal-development sites within a large water- shed is not fully developed at this time (1985). Sufficient hydrologic knowledge presently (1985) exists to prescribe measures that will maximize the 169 rehabilitation of the water resources affected by min- ing. But, because of limited numbers of studies that describe hydrologic responses of mine-land rehabilita- tion, it is more difficult to predict long-term hydrologic trends of mined and rehabilitated lands. This knowledge eventually will be derived from ongoing studies of post- mining hydrologic systems. At the beginning of the coal-hydrology program, many needs existed for hydrologic information. These needs and the data-collection programs, hydrologic in- vestigations, and research activities designed to meet these needs have been described in previous sections of this report. Many of these needs have been met, and much has been learned during the decade of the coal- hydrology program. Data have been obtained to define baseline hydrologic conditions in coal areas where little or no prior data existed, and much has been done to im- prove the understanding of hydrologic principles and processes as they pertain to coal mining. However, some needs have not been fully met and, as in many scien- tific efforts, knowledge gained during the coal- hydrology program has generated new questions and needs that were not known at the beginning of the pro- gram. A summary of some of the more important re- maining needs for coal-hydrology information follows. * Additional information is needed about the hydraulic properties of mine spoils and coal aquifers such as anisotropy and fracture control of flow, to enable prediction of downgradient, ground-water flow patterns and their effects on water quality. * Additional research is needed to determine how chemical and microbiological processes such as ion exchange, sulfate reduction, and mineral equilibria can affect the recovery potential of contaminated aquifers that are downgradient from mine spoils. * Additional work is needed to develop techniques to predict and quantify the cumulative hydrologic effects, particularly in terms of water quality and sediment, of multiple mines in the same drainage basin. A determination of cumulative effects is re- quired by the Surface Mining Control and Recla- mation Act and is needed by the Office of Surface Mining Reclamation and Enforcement to permit new mines and to renew permits for existing mines. * Additional information and analysis is needed in the Interior coal province to determine if the post- mining water quality in areas reclaimed to Surface Mining Control and Reclamation Act standards is substantially different from unreclaimed areas. * During the final years (1982—84) of the coal- hydrology program, an intensive effort was begun to obtain precipitation, runoff, suspended- sediment, and water-quality data for 19 small 170 a: SUMMARY OF THE NATIONAL COAL-HYDROLOGY PROGRAM, 1974—84 basins of varying land uses and hydrologic settings. Land use ranged from forested areas that had no mining to intensively mined basins. The hydrologic information was collected on a storm-event basis using automatic monitoring and sampling equipment. Curtailment of funds for coal-hydrology studies has prevented detailed analysis and interpretation of these data. Addi- tional analysis of the data base is needed to deter- mine how accurately models and other techniques developed during the coal-hydrology program can predict effects of mining on sedimentation and water quality. Techniques and models that are suc- cessful in predicting effects would be very useful to regulatory agencies. In the Western United States, several case studies are needed to document long-term hydrologic changes caused by surface mining. Ideally, these studies would begin prior to mining and continue during and after the mining process. The primary purpose of these studies would be to provide hydrologic information to test, calibrate, verify, and refine concepts, hypotheses, and models about the hydrologic effects of mining and to determine the effectiveness of reclamation efforts. A number of studies begun during the coal-hydrology pro- gram obtained hydrologic information prior to and during mining. However, no studies in the coal provinces in the Western United States have fol- lowed mining operations through the entire cycle from premining to postreclamation. REFERENCES CITED REFERENCES CITED Abbott, P.0., Geldon, A.L., Cain, Doug, Hall, A.P., and Edelrnann, Patrick, 1983. Hydrology of area 61, Northern Great Plains and Rocky Mountain Coal Provinces, Colorado and New Mexico: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—132, 99 p. Alley, W.M., Britton, L.J., and Boyd, E.L., 1978a, Reconnaissance evaluation of water resources for hydraulic coal mining, Crested Butte coal field, Gunnison County, Colorado: U.S. Geological Survey Open-File Report 78-938, 23 p. __1978b, Reconnaissance evaluation of water resources for hydraulic coal mining, Grand Hogback coal field, Garfield and Rio Blanco Counties, Colorado: U.S. Geological Survey Open-File Report 78—885, 37 p. Amuedo and Ivey, 1974, Regional coal resources study of the Trinidad- Raton basin, Colorado and New Mexico: Denver, Amuedo and Ivey Engineering Consultants, unpublished report. Anderson, J .R., 1970, Land use, in National Atlas of the United States of America: U.S. Geological Survey, p. 158—159, scale 1:7,500,000. Andrews, ED, 197 8, Present and potential sediment yields in the Yampa River basin, Colorado and Wyoming: U.S. Geological Survey Water-Resources Investigations Report 78—105, 33 p. Appalachian Regional Commission, 1969, Acid mine drainage in Ap- palachia: Washington, DC, 126 p. Argonne National Laboratory [prepared by Systems Consultants, Inc.], 1982, Energy and water resources: Washington, D.C., U.S. Department of Energy, 322 p. Armentrout, G.W., Jr., and Wilson, J .F., Jr., 1987, An assessment of low flows in streams in northeastern Wyoming. U.S. Geological Survey Water-Resources Investigations Report 85-4246, 30 p. Armstrong, CA. 1982, Evaluation of the hydrologic system in the New Leipzig coal area, Grant and Hettinger Counties, North Dakota: U.S. Geological Survey Open-File Report 82-698, 41 p. Arnold, E.C., and Hill, J .M., compilers, 1981, New Mexico’s energy resources, ’81: Santa Fe, New Mexico Energy and Minerals Department, 62 p. Avcin, M.J., and Koch, D.L., 1979, The Mississippian and Pennsylva- nian (Carboniferous) systems in the United States [Iowa]: U.S. Geological Survey Professional Paper 1110-M—DD, p. M1—M13. Averitt, Paul, 1975, Coal resources of the United States, January 1, 1974: U.S. Geological Survey Bulletin 1412, 131 p. Babu, S.P., Barlow, J .A., Craddock, L.L., Hidalgo, R.V., and Friel, EA, 1973, Suitability of West Virginia coals to coal-conversion processes: Morgantown, West Virginia Geological and Economic Survey Coal-Geology Bulletin 1, 32 p. Bahls, L.L., 1980, Salinity and the structure of benthic algae (periphyton) communities in streams of the southern Fort Union Region, Montana: Helena, Montana Department of Health and Environmental Sciences, 35 p. Bahls, L.L., Weber, E.E., and J arvie, J .O., 1984, Ecology and distribu- tion of major diatom ecotypes in the Southern Fort Union Coal Region of Montana: U.S. Geological Survey Professional Paper 1289, 151 p. Banaszak, K.J., 1980, Goals as aquifers in the Eastern United States, in Symposium on Surface Mining Hydrology, Sedimentology. and Reclamation, Proceedings: Lexington, University of Kentucky, p. 235—241. ___1985, Potential effects on ground water of hypothetical surface coal in Indiana: Ground Water Monitoring Review, v. 5, no. 1, p. 51-57. Bauer, D.P., Rathbun, R.E., and Lowham, H.W., 1979, Traveltime, unit concentration, longitutional dispersion, and reaeration characteristics of upstream reaches of the Yampa and Little Snake Rivers, Colorado and Wyoming: U.S. Geological Survey Water- Resources Investigations Report 78—122, 66 p. 171 Bevans, H.E., 1980, A procedure for predicting concentrations of dissolved solids and sulfate ions in streams draining areas strip mined for coal: U.S. Geological Survey Water-Resources Investiga- tions Open-File Report 80—764, 17 p. _1984, Hydrologic responses of streams to mining of the Mulberry coal reserves in eastern Kansas: U.S. Geological Survey Water-Resources Investigations Report 84-4047, 30 p. ___1986, Estimating stream-aquifer interactions in coal areas of eastern Kansas, in Subitzky, Seymour, ed., Selected papers in the hydrologic sciences 1986: U.S. Geological Survey Water-Supply Paper 2290, 154 p. Bevans, H.E., and Diaz, A.M., 1980, Statistical summaries of water- quality data for streams draining coal-mined areas, southeastern Kansas: U.S. Geological Survey Hydrologic Data Open-File Report 80-350, 42 p. Bevans, H.E., Skelton, John, Kenny, J.F., and Davis, J.V., 1984, Hydrology of area 39, Western Region, Interior Coal Province, Kansas and Missouri: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—851, 83 p. Biesecker, J .E., and George, J .R., 1966, Stream quality in Appalachia as related to coal-mine drainage, 1965: U.S. Geological Survey Cir- cular 526, 27 p. Blanchard, P.J., 1984, Ground-water conditions in Kaiparowitz Plateau area, Utah and Arizona, with emphasis on the Navajo Sandstone: Utah Department of Natural Resources Technical Publication 81, 75 p. Bloyd, R.M., Daddow, P.B., Jordan, P.R., and Lowham, H.W., 1986, Investigation of possible effects of surface coal mining on hydrology and landscape stability in part of the Powder River structural basin, northeastern Wyoming: U.S. Geological Survey Water-Resources Investigations Report 86-4329, 101 p. Bobay, K.E., 1986, Theoretical technique for predicting the cumulative impact of iron and manganese oxidation in streams receiving discharge from coal mines: U.S. Geological Survey Water- Resources Investigations Report 86-4039, 29 p. Bobay, K.E., and Banaszak, K.J., 1985, Theoretical technique for determining the cumulative impact of iron and manganese oxida- tion in streams receiving coal-mine discharge, in Symposium on Surface Mining Hydrology, Sedimentology, and Reclamation: Lex- ington, University of Kentucky, p. 105—114. Boner, F.C., Lines, G.C., Lowry, M.E., and Powell, J.E., 1976, Geohydrologic reconnaissance and measurement of perennial streams crossing outcrops of the Madison Limestone, northeastern Wyoming, 1974: U.S. Geological Survey Open-File Report 75—614, 63 p. Borland, J .P., 1970, A proposed streamflow-data program for New Mexico: U.S. Geological Survey Open-file Report, 71 p. Bower, DE, 1985, Evaluation of the precipitation-runoff modeling system, Beaver Creek basin, Kentucky: U.S. Geolological Survey Water-Resources Investigations Report 84—4316, 39 p. Brabets, T,.P., 1984, Runoff and water-quality characteristics of surface-mined lands in Illinois: U.S. Geological Survey Water- Resources Investigations Report 83—4265, 78 p. Bradfield, A.D., 1986a, Evaluation of coal-mining impacts using numerical classification of benthic invertebrate data from streams draining a heavily mined basin in eastern Tennessee: U.S. Geological Survey Water-Resources Investigations Report 85—4289, 59 p. 1986b, Benthic invertebrate population characteristics as affected by water quality in coal-bearing regions of Tennessee: U.S. Geological Survey Water-Resources Investigations Report 84—4227, 19 p. Brady, L.L., Adams, D.B., and Livingston, N.D., 1976, An evalua- tion of the strippable coal reserves in Kansas: Lawrence, Univer- sity of Kansas, Kansas Geological Survey Mineral Resources Series 5, 40 p. 172 Brady, LL, and Dutcher, L.F., 197 4, Kansas coal—A future energy resource: Lawrence, University of Kansas, Kansas Geological Survey Journal, 28 p. Britton, L.J., 1983, Reconnaissance of benthic invertebrates from tributary streams of the Yampa and North Platte River basins, northwestern Colorado: U.S. Geological Survey Water-Resources Investigations Report 83—4191, 73 p. Brogden, RE, and Giles, TR, 1977, Reconnaissance of ground-water resources in a part of the Yampa River basin between Craig and Steamboat Springs, Moffat and Routt Counties, Colorado: U.S. Geological Survey Water-Resources Investigations Report 77 —4, scale 1:120,000. Brooks, Tom, 1983, Hydrology and subsidence potential of proposed coal-lease tracts in Delta County, Colorado: U.S. Geological Survey Water-Resources Investigations Report 83-4069, 27 p. Bryant, C.T., Lyford, F.P., Stafford, KL, and Johnson, D.M., 1983, Hydrology of area 42, Western Region, Interior Coal Province, Arkansas: U.S. Geological Survey Water-Resources Investigations Open-File Report 82—636, 62 p. Burchett, R.R., 1977, Coal resources of Nebraska: Lincoln, Universi- ty of Nebraska, Nebraska Geological Survey Resources Report 8, 185 p. Busby, M.W., 1966, Annual runoff in the conterminous United States: U.S. Geological Survey Hydrologic Investigations Atlas HA—212, scale 1:7,500,000. Cannon, M.R., 1982, Potential effects of surface coal mining on the hydrology of the Cook Creek area, Ashland coal field, southeastern Montana: U.S. Geological Survey Open-File Report 82—681, 30 p. 1983, Potential effects of surface coal mining on the hydrology of the Snider Creek area, Rosebud and Ashland coal fields, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 82—4051, 28 p. Carey, W.P. and Simon, Andrew, 1985, Physical basis and potential estimation techniques for soil erosion parameters in the precipitation-runoff modeling system (PRMS): U.S. Geological Survey Water-Resources Investigations Report 84-4218, 32 p. Carpenter, DH, 1983, Technique for estimating magnitude and fre- quency of floods in Maryland: U.S. Geological Survey Water- Resources Investigations Open-File Report 80-1016, 119 p. Caruccio, F.T., 1973, Characterization of strip-mine drainage by pyrite grain size and chemical quality of existing groundwater, in Hut- nik, Russell, and Davis, Grant, eds., Ecology of the reclamation of devastated lands: New York, Gordon and Breach Scientific Publishers, v. 1, p. 193—226. Caruccio, F.T., Geidel, Gwendolyn, and Sewell, J .M., 1976, The character of drainage as a function of the occurrence of framboidal pyrite and ground water quality in eastern Kentucky: Washington, DC, National Coal Association and Bituminous Coal Research, Sixth Symposium on Coal Mine Drainage Research, p. 1—16. Cary, LE, 1984, Application of the U.S. Geological Survey’s precipitation-runoff modeling system to the Prairie Dog Creek basin, southeastern Montana: U.S. Geological Survey Water- Resources Investigations Report 84—4178, 95 p. Chaney, T.H., Kuhn, Gerhard, Brooks, Tom, and others, 1987, Hydrology of Area 58, Northern Great Plains and Rocky Moun- tain Coal Provinces, Colorado and Utah: U.S. Geological Survey Water-Resources Investigations Open-File Report 85—479, 103 p. Chow, Ven Te, ed., 1964, Handbook of applied hydrology: New York, McGraw-Hill, various pagination. Christensen, R.C., Johnson, EB, and Plantz, G.G., 1986, Manual for estimating streamflow characteristics of natural-flow streams in the Colorado River Basin in Utah: U.S. Geological Survey Water- Resources Investigations Report 85—4297, 38 p. Christensen, RC, and Plantz, G.G., 1985, Streamflow characteristics SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 of the Colorado River Basin in Utah through September 1981: U.S. Geological Survey Open-File Report 85-421, 674 p. Cochran, B.J., Palmquist, Will, Van Haveren, B.P., Tamberg, Nora, and Goolsby, D.A., 1983, Coal hydrology bibliography: Lakewood, Colo., U.S. Bureau of Land Management and U.S. Geological Survey, 448 p. Cole, ER, 1984, Effects of coal mining on the water quality and sedimentation of Lake Tuscaloosa and selected tributaries, North River basin, Alabama: U.S. Geological Survey Water-Resources Investigations Report 84-4310, 75 p. Covay, K.J., and Tobin, R.L., 1981, Quality of ground water in Routt County, northwestern Colorado: U.S. Geological Survey Open-File Report 80-956, 38 p. Craig, G.S., J r., and Rankl, J .G., 1978, Analysis of runoff from small drainage basins in Wyoming: U.S. Geological Survey Water- Supply Paper 2056, 70 p. Crawley, M.E., and Emerson, D.G., 1981, Hydrologic characteristics and possible effects of surface mining in the northwestern part of West Branch Antelope Creek basin, Mercer County, North Dakota: U.S. Geological Survey Water-Resources Investigations Report 81—79, 73 p. ‘ Croft, M.G., and Crosby, O.A., 1987, Hydrology of area 46, Northern Great Plains and Rocky Mountain Coal Provinces, North Dakota: U.S. Geological Survey Water-Resources Investigations Open-File Report 84-467, 80 p. Crosby, O.A., 1975, Magnitude and frequency of floods in small drainage basins in North Dakota: U.S. Geological Survey Water- Resources Investigations Report 19—7 5, 43 p. [Available only from National Technical Information Services, Springfield, Va., as PB—248 480.] Crosby, O.A., and Klausing, R.L., 1984, Hydrology of area 47, North- ern Great Plains and Rocky Mountain Coal Provinces, North Dakota, South Dakota, and Montana: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—221, 93 p. Curtis, W.F., Flint, RR, and George, F.H., 1978, Fluvial sediment study of Fishtrap and Dewey Lakes drainage basins, Kentucky, Virginia: U.S. Geological Survey Water-Resources Investigations Report 77-123, 92 p. Daddow, P.B., 1986a, Ground-water data through 1980 for the Hanna and Carbon basins, south-central Wyoming: U.S. Geological Survey Open-File Report 85—628, 91 p. 1986b, Potentiometric-surface map of the Wyodak-Anderson coal bed, Powder River structural basin, Wyoming, 1973-84: U.S. Geological Survey Water-Resources Investigations Report 85—4305, scale 1:250,000. Dames and Moore Engineering Consultants, 1978, Surface and ground water monitoring programs for Kaiser Steel Corporation, York Canyon Mine, Raton, New Mexico: Denver, 0010., Report for Kaiser Steel Corporation by Dames and Moore Engineering Con- sultants, 1 v. Danielson, T.W., ReMillard, M.D., and Fuller, RH, 1981, Hydrology of the coal-resource areas in the upper drainages of Huntington and Cottonwood Creeks, central Utah: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—539, 85 p. Danielson, T.W., and Sylla, D.A., 1982, Hydrology of coal-resource areas in the southern Wasatch Plateau, central Utah: U.S. Geo- logical Survey Water-Resources Investigations Report 82—4009, 66 p. Davies, W.E., Bailey, J .F., and Kelly, D.B., 1972, West Virginia’s Buf- falo Creek Flood—A study of the hydrology and engineering geology: U.S. Geological Survey Circular 667, 32 p. Davis, R.E., 1984a, Example calculations of possible ground-water inflow to mine pits at the West Decker, East Decker, and pro- posed North Decker mines, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 84—4199, 31 p. REFERENCES CITED 1984b. Geochemistry and geohydrology of the West Decker and Big Sky coal-mining areas, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 83—4225, 109 p. Davis. R.W., Plebuch, R.V., and Whitman, H.M., 1974, Hydrology and geology of deep sandstone aquifers of Pennsylvanian age in part of the Western Coal Field region. Kentucky: Lexington, Ken- tucky Geological Survey Report of Investigations 15, 24 p. DeLong, L.L., 1977, An analysis of salinity in streams of the Green River basin, Wyoming: U.S. Geological Survey Water-Resources Investigations Report 77-103, 32 p. 1978. Predicting effects of coal development on surface-water salinity. Green River basin, Wyoming, Annual Meeting, American Geophysical Union, San Francisco, 1978 [abs.]: EOS, v. 59, no. 12, p. 1067. __1979, Predicting effects of coal development on surface-water salinity, Green River basin, Wyoming—Wyoming Mining Hydrol- ogy Seminar, April 1979: Laramie, University of Wyoming, 1 v. ___1985, Water quality of streams and springs, Green River basin, Wyoming: U.S. Geological Survey Water-Resources Investiga- tions Report 82-4008, 36 p. Detroy, M.G., Skelton, J ohn, and others, 1983. Hydrology of area 38, Western Region. Interior Coal Province. Iowa and Missouri: U.S. Geological Survey Water-Resources Investigations Open-File Report 82-1014, 85 p. Dockins, W.S., Olson. G.J., McFeters, G.A., Turback, 8.0., and Lee. R.W., 1980, Sulfate reduction in ground water of southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 80—9, 13 p. [Available only from National Technical In- formation Service. Springfield, Va., as PB—80 221 971.] Doelling, H.H., 1972, Central Utah coal fields—Sevier-Sanpete, Wasatch Plateau, Book Cliffs, and Emery: Utah Geological and Mineralogical Survey Monograph 3, 571 p. Doelling, H.H., and Graham, R.L., 1972. Southwestern Utah coal fields—Alton, Kaiparowits Plateau, and Kolob-Harmony: Utah Geological and Mineralogical Survey Monograph 1, 333 p. Doyle, W.H., Jr., Curwick, P.B., and Flynn, K.M., 1983, A flood model for the Tug Fork basin, Kentucky. Virginia, and West Virginia: U.S. Geological Survey Water-Resources Investigations Report 83—4014, 87 p. Driver, N.E., Norris, J .M., Kuhn, Gerhard, and others, 1984, Hydrology of Area 53, Northern Great Plains and Rocky Moun- tain Coal Provinces, Colorado, Wyoming, and Utah: U.S. Geological Survey Water-Resources Investigations Open-File Report 83-765, 93 p. Driver, NE, and Williams. RS, 1986, Hydrogeology of and poten- tial mining impacts on strippable lignite areas in the Denver aquifer, east-central Colorado: U.S. Geological Survey Water- Resources Investigations Report 84—4366, 39 p. Druse, S.A., 1982, Verification of step-backwater computations on ephemeral streams in northeastern Wyoming: U.S. Geological Survey Water-Supply Paper 2199, 12 p. Druse, S.A., Dodge, K.A.. and Hotchkiss, W.R., 1981, Base flow and chemical quality of streams in the northern Great Plains area, Montana and Wyoming, 1977—78: U.S. Geological Survey Water— Resources Investigations Open-File Report 81—692, 60 p. Dum'ud, GR, and Osterwald, F.W., 1980. Effects of coal mine sub- sidence in the Sheridan, Wyoming. area: U.S. Geological Survey Professional Paper 1164, 49 p. Eardley, A.J., 1951, Structural geology of North America: New York, Harper, 624 p. Ebanks, W.J., Jr., Brady, L.L., Heckel, P.H., O’Connor, H.G.. Sander- son, G.A., West. RR, and Wilson, F.W., 1979, The Mississippian and Pennsylvanian (Carboniferous) systems in the United States [Kansas]: U.S. Geological Survey Professional Paper 1110-M— DD, p. Q1—Q30. Ehlke, T.A., Bader, J.S., Puente, Celso, and Runner, G.S., 1982a, Hydrology of Area 12, Eastern Coal Province, West Virginia: U.S. 173 Geological Survey Water-Resources Investigations Open-File Report 81—902, 75 p. Ehlke, T.A., Runner, G.S., and Downs, S.C., 1982b, Hydrology of Area 9, Eastern Coal Province, West Virginia: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—803, 63 p. Ehlke, T.A., and others, 1983, Hydrology of Area 10. Eastern Coal Province, West Virginia: U.S. Geological Survey Water-Resources Investigations Open-File Report 82—864, 73 p. Emerson, D.G., 1981, Progress report on the effects of surface min- ing on the surface-water hydrology of selected basins in the Fort Union coal region, North Dakota and Montana: U.S. Geological Survey Open-File Report 81—678, 28 p. ___1988, Surface-water hydrology of Hay Creek watershed, Mon- tana, and west branch Antelope Creek watershed, North Dakota: U.S. Geological Survey Water-Resources Investigations Report 88-4038, 111 p. Engelke, M.J., Jr., 1978, The biology of Salt Wells Creek and its tributaries, southwestern Wyoming: U.S. Geological Survey Water-Resources Investigations Report 78-121, 82 p. Engelke, M.J., Jr., Roth, D.K., and others, 1981, Hydrology of Area 7. Eastern Coal Province, Ohio: U.S. Geological Survey Water- Resources Investigations Open-File Report 81—815, 60 p. Federal Water Pollution Control Administration, 1968, Water quali- ty criteria—Report of the National Technical Advisory Commit- tee to the Secretary of the Interior: Washington, D.C., U.S. Government Printing Office, 234 p. Fenneman, N .M., 1931, Physiography of western United States: New York, McGraw-Hill, 534 p. Fenneman, N.M., and Johnson, D.W., 1946, Physical divisions of the United States: U.S. Geological Survey map, scale 1:7 ,000,000. Ferreira, R.F., 1984, Simulated effects of surface coal mining and agriculture on dissolved solids in Rosebud Creek, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 84-4101, 60 p. Ferreira, R.F., and Lambing, J .H., 1984. Suitability of water quality for fish propagation, waterfowl habitat, livestock watering, and recreational use at 12 reservoirs in eastern Montana: U.S. Geolog- ical Survey Water-Resources Investigations Report 84—4085, 96 p. Fitzgerald, K.K., Peters, CA. and Zuehls. E.E., 1984, Hydrology of Area 29, Eastern Region, Interior Coal Province, Illinois: U.S. Geological Survey Water-Resources Investigations Open-File Report 82—858. 70 p. Flint, R.F., 1983, Fluvial sedimentation in Kentucky: U.S. Geological Survey Water-Resources Investigations Report 83—4152, 75 p. Flippo, H.N., Jr., 1977 . Floods in Pennsylvania: Harrisburg, Penn- sylvania Department of Environmental Resources Water- Resources Bulletin 13, 59 p. 1982, Technical manual for estimating low-flow characteristics of Pennsylvania streams: Harrisburg, Pennsylvania Department of Environmental Resources, Water-Resources Bulletin 15, 86 p. Frenzel, P.F., 1983, Simulated changes in ground-water levels related to proposed development of Federal coal leases, San Juan basin. New Mexico: US. Geological Survey Open-File Report 83—949. 63 p. Freudenthal, P.B., 1979, Water-quality data for the Hanna and Car- bon basins, Wyoming: US. Geological Survey Open-File Report 79-1277. 41 p. Frickel. D.G.. Shown, L.M.. Hadley. R.F., and Miller. R.F., 1981. Methodology for hydrologic evaluation of a potential surface mine, Red Rim site, Carbon and Sweetwater Counties, Wyoming: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—75, 65 p. Friedman, S.A., 1974, Investigation of the coal reserves in the Ozarks section of Oklahoma and their potential uses—Final report to the Ozarks Regional Commission: Oklahoma City, Oklahoma Geolog- ical Survey. 117 p. 174 Friel, E.A., Ehlke, T.A., Hobba, W.A., Jr., Ward, S.M., and Shultz, RA, 1987, Hydrology of Area 8, Eastern Coal Province, West Virginia and Ohio: U.S. Geological Survey Water-Resources In- vestigations Open-File Report 84—463, 78 p. Gabelman, J .W., 1956, Tectonic history of the Raton Basin region, in Guidebook to the geology of the Raton Basin, Colorado: Denver, Rocky Mountain Association of Geologists, p. 35—39. Gaggiani, N.G., Britton, L.J., Minges, DR, and others, 1987, Hydrology of area 59, Northern Great Plains and Rocky Moun- tain Coal Provinces, Colorado and Wyoming: U.S. Geological Survey Water-Resources Investigations Open-File Report 85—153, 124 p. Gamble, CR, 1965, Magnitude and frequency of floods in Alabama: Montgomery, Alabama Geological Survey Reprint Series 11 [reprinted by permission of Alabama Highway Department, HPR 5], 42 p. Gaydos, M.W., and others, 1982a. Hydrology of area 17. Eastern Coal Province, Tennessee and Kentucky: U.S. Geological Survey Water- Resources Investigations Open-File Report 81—1118, 75 p. 1982b, Hydrology of area 19, Eastern Coal Province, Tennessee: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—901, 75 p. Geldon, A.L., in press, Ground-water hydrology of the Central Raton Basin, Colorado and New Mexico: U.S. Geological Survey Water- Supply Paper 2288. Geldon, A.L., and Abbott, P.O., 1985, Selected climatological and hydrologic data, Raton Basin, Huerfano and Las Animas Coun- ties, Colorado, and Colfax County. New Mexico: U.S. Geological Survey Open-File Report 84—138, 268 p. Gillette Area Groundwater Monitoring Organization, 1983, 1982 GAGMO annual report: Gillette, Wyo., AMAX Coal Company, compiler, unpaginated. Gilley, J .E., 1980, Runoff and erosion from mined lands in western North Dakota, in Soil Conservation Society of America Sym- posium, Proceedings—Adequate reclamation of mined lands: Bill- ings, Montana, March 26—27, 1980, p. 51-518. Glover, KC, 1978, A computer program for simulating salinity loads in streams: U.S. Geological Survey Open-File Report 78—884. 35 p. 1984, Storage analysis for ephemeral streams in semiarid regions: U.S. Geological Survey Water-Resources Investigations Report 83—4078, 55 p. Goetz, CL. 1981, Preliminary analysis of historical streamflow and water-quality records for the San Juan River basin, New Mexico and Colorado, in Wells, S.G.. and Lambert, Wayne, eds., En- vironmental geology and hydrology in New Mexico: Santa Fe, New Mexico Geological Society Special Publication No. 10, p. 21—25. Goetz, C.L., Abeyta, CG, and Thomas, E.V., 1987, Application of techniques to identify coal-mine and power-generation effects on surface-water quality, San Juan River basin, New Mexico and Col- orado: U.S. Geological Survey Water-Resources Investigations Report 86—407 6, 80 p. Grason, David, 1982, A presentation and evaluation of the hydrologic information available for the major Federal coal lands in seven eastern States—Sources of available information and a plan for future work: U.S. Geological Survey Open-File Report 82—525, 335 p. Griggs, R.L., and Northrop, 8A., 1956, Stratigraphy of the plains area adjacent to the Sangre de Cristo Mountains, New Mexico, in Guidebook of southeastern Sangre de Cristo Mountains, New Mexico: New Mexico Geological Society Annual Field Conference, 7th, 1956, Guidebook, p. 134—138. Groenewold, G.H., Hemish, L.A., Cherry, J .A., Rehm, B.W., Meyer, G.N., Clayton, LS, and Winczewski, L.M., 1979, Geology and geohydrology of the Knife River basin and adjacent areas of west- central North Dakota: Bismarck, North Dakota Geological Survey Report of Investigation 64, 402 p. SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Groenewold, G.H., Koob, R.D., McCarthy, G.J., Rehm, B.W., and Peterson, W.H., 1983, Geological and geochemical controls on the chemical evolution of subsurface water in undisturbed and surface- mined landscapes in western North Dakota: Bismarck, North Dakota Geological Survey Report of Investigation 79, 151 p. Groenewold, G.H., and Murphy, EC, 1983, Development of a hydrogeologic and hydrogeochemical data base for abandoned lands—Phase 1: Grand Forks, North Dakota Mining and Mineral Resources Research Institute Report 83—1, 15 p. Groenewold, G.H., and Rehm, B.W., 1980, Instability of contoured surface-mined landscapes in the Northern Great Plains—Causes and implications, in Adequate Reclamation of Mined Lands Sym- posium, Billings, Montana, 1980, Proceedings: Ankeny, Iowa, Soil Conservation Society of America, p. 2,1—2.15. Guy, H.P., 1970, Fluvial sediment concepts: U.S. Geological Survey Techniques of Water-Resources Investigations, book 3, chap. Cl, 55 p. Hadley, R.F., Frickel, D.G., Shown, L.M., and Miller, RR, 1981, Methodology for hydrologic evaluation of a potential surface mine—The East Trail Creek basin, Big Horn County, Montana: U.S. Geological Survey Water-Resources Investigations Open-File Report 81-58, 79 p. Haffield, N .D., 1981, Statistical summaries of streamflow and water- quality data for streams of western North Dakota, 197 7 -80: U.S. Geological Survey Open-File Report 81—1066, 78 p. Hains, GR, 1973, Floods in Alabama—magnitude and frequency: Montgomery, Alabama Highway Department, 39 p. Haley, B.R., Glick, E.E., Caplan, W.M., Holbrook, DR, and Stone, CG, 1979, The Mississippian and Pennsylvanian (Carboniferous) systems in the United States [Arkansas]: U.S. Geological Survey Professional Paper 1110—M—DD, p. 01-014. Hall, DC, and Davis, RE, 1986, Ground-water movement and ef- fects of coal strip mining on water quality of high-wall lakes and aquifers in the Macon-Huntsville area, north-central Missouri: U.S. Geological Survey Water-Resources Investigations Report 85-4102, 112 p. Hannum, G.H., 1976, Technique for estimating magnitude and fre- quency of floods in Kentucky: U.S. Geological Survey Water- Resources Investigations Report 76—62, 70 p. [Available only from National Technical Information Service, Springfield, Va., as PB-263 762.] Harkins, J .R., and others, 1980, Hydrologic assessment, Eastern Coal Province, Area 23, Alabama: U.S. Geological Survey Water- Resources Investigations Open-File Report 80—683, 76 p. Harkins, J .R., and others, 1981, Hydrology of area 22, Eastern Coal Province, Alabama: U.S. Geological Survey Water-Resources In- vestigations Open-File Report 81—135, 72 p. 1982, Hydrology of area 24, Eastern Coal Province, Alabama and Georgia: U.S. Geological Survey Water-Resources Investiga- tions Open-File Report 81—1113, 79 p. Hatch, J.R., Avcin, M.J., and Van Dorpe, RE, 1984, Element geochemistry of Cherokee Group coals (Middle Pennsylvanian) from south-central and southeastern Iowa: Iowa City, Iowa Geological Survey Technical Paper 5, 108 p. Heath, RC, 1983, Basic ground-water hydrology: U.S. Geological Survey Water-Supply Paper 2220, 84 p. Heimes, F.J., Moore, GK, and Steele, T.D., 1978, Preliminary ap- plications of LAN DSAT images and aerial photography for deter- mining land-use, geologic and hydrologic characteristics—Yampa River basin, Colorado and Wyoming: U.S. Geological Survey Water-Resources Investigations Report 78-96, 48 p. Hejl, H.R., Jr., 1980, Preliminary appraisal of ephemeral-streamflow characteristics as related to drainage area, active-channel width, and soils in northwestern New Mexico: U.S. Geological Survey Open-File Report 81—64, 15 p. REFERENCES CITED 1982, Hydrologic investigations and data-collection network in strippable coal-resource areas in northwestern New Mexico: U.S. Geological Survey Open-File Report 82-358, 32 p. __1984, Use of selected basin characteristics to estimate mean annual runoff and peak discharges for ungaged streams in drainage basins containing strippable coal resources, northwestern New Mexico: U.S. Geological Survey Water-Resources Investiga- tions Report 84-4260. 17 p. _.in press, Application of the precipitation-runoff modeling system to the Ah-Shi-Sle—Pah Wash watershed, San Juan Coun- ty, New Mexico: U.S. Geological Survey Water-Resources In- vestigations Report 88—4140. Helgesen. J .O., and Razem, A.C., 1981, Ground-water hydrology of strip-mine areas in eastern Ohio (conditions during mining of two watersheds in Coshocton and Muskingum Counties): U.S. Geolog- ical Survey Water-Resources Investigations Open-File Report 81-913, 25 p. Hem, J .D., 1985, Study and interpretation of the chemical character- istics of natural water [3d ed.]: U.S. Geological survey Water- Supply Paper 2254, 263 p. Henderson, Thomas, 1984, Geochemistry of ground water in two sand- stone aquifer systems in the Northern Great Plains in parts of Montana, Wyoming, North Dakota. and South Dakota: U.S. Geological Survey Professional Paper 1402-0, 84 p. Herb, W.J., 1981, Technical manual for estimating mean flow characteristics of Pennsylvania streams: University Park, Penn- sylvania Department of Environmental Resources Water- Resources Bulletin, 1 v. Herb. W.J., Brown, DE, Shaw, L.C., and Becher, A.E., 1983a, Hydrology of Area 1, Eastern Coal Province, Pennsylvania: U.S. Geological Survey Water-Resources Investigations Open-File Report 82-223, 88 p. Herb, W.J., Brown, DE, Shaw, L.C., Stoner, J .D., and Felbinger. J .K., 1983b, Hydrology of Area 2, Eastern Coal Province, Pennsylvania and New York: U.S. Geological Survey Water-Resources In- vestigations Open-File Report 82—647, 93 p. Herb, W. J ., Shaw, L. C., and Brown, D. E., 1981a, Hydrology of Area 3, Eastern Coal Province, Pennsylvania: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—537. 88 p. 1981b, Hydrology of Area 5, Eastern Coal Province, Penn- sylvania, Maryland, and West Virginia: U.S. Geological Survey Water-Resources Investigations Open-File Report 81-538, 92 p. Hill, R.D., 1968, Mine drainage treatment—State of the art and research needs: Cincinnati, Ohio, Federal Water Pollution Con- trol Administration, BCR68-150, 102 p. Hjelmfelt, A.T., Piest, RR, and Sexton, K.E., 1975, Mathematical modeling of erosion on upland areas: Congress of the 16th Inter- national Association for Hydraulic Research, Sao Paulo, Brazil, 1975, Proceedings, v. 2, p. 40—47. Hobba, W.A.. Jr., 1981, Effects of underground mining and mine col- lapse on the hydrology of selected basins, West Virginia: Morgan- town, West Virginia Geological and Economic Survey Report of Investigation 33, 109 p. Hollis, Robert, 1982, The coalstream pipeline project, in Kilpatrick, G.A., and Matchett, Donald. eds.. Water and energy, technical and policy issues: New York, American Society of Civil Engineers, p. 38-42. Hollyday. ER, and others, 1983, Hydrology of area 20. Eastern Coal Province, Tennessee, Georgia, and Alabama: U.S. Geological Survey Water-Resources Investigations Open-File Report 82—440, 81 p. Hood, J .W., and Fields, F.K, 1978, Water resources of northern Uinta basin area, Utah and Colorado: Salt Lake City, Utah Department of Natural Resources Technical Publication 62, 75 p. Hood, W.C., and Oertel, A.O., 1984, A leaching column method for predicting effluent quality from surface mines, in 1984 National 175 Symposium on Surface Mining Hydrology, Sedimentology, and Reclamation, Lexington, Ky.: Lexington, University of Kentucky Press, p. 271-277. Horak, W.F., 1983a, Hydrology of the Wibaux-Beach lignite deposit area, eastern Montana and western North Dakota: U.S. Geological Survey Water-Resources Investigations Report 83—4157, 89 p. 1983b, Water resources of the Rattlesnake Butte area, a site of potential lignite mining in west-central North Dakota: U.S. Geological Survey Water-Resources Investigations Report 83—4228, 53 p. Howard, J .M., 1984, Arkansas, in 1984 Keystone coal industry manual: New York, McGraw—Hill Mining Publications, p. 481—483. Howard, W.B., 1982, The hydrogeology of the Raton Basin, south- central Colorado: Bloomington, Indiana University, unpublished M.A. thesis, 95 p., appendices A—K. Hubbard, E.F.. 1976, Sedimentation in Lake Tuscaloosa, Alabama: U.S. Geological Survey Open-File Report 76-158, 35 p. Hufschmidt, P.W., and others, 1981, Hydrology of area 16, Eastern Coal Province, Virginia and Tennessee: U.S. Geological Survey Water-Resources Investigations Open-File Report 81-204, 68 p. Indiana University, 1983, Coal resources fact book, in Illinois Basin coal planning assistance project, v. 1 of 4: Bloomington, En- vironmental Systems Application Center, 323 p. Jarrett, RD, and Veenhuis, J .E., 1984, An evaluation of rainfall-runoff data for the Denver Federal Center, Lakewood, Jefferson Coun- ty, Colorado: U.S. Geological Survey Water-Resources Investiga- tions Report 84—4050, 29 p. Johnson, DR, and Metzker, K.D., 1982, Low-flow characteristics of Ohio streams: U.S. Geological Survey Open-File Report 81—1 195, 292 p. J ohnson, R.B., 1961, Coal resources of the Trinidad coalfield in Huer- fano and Las Animas Counties, Colorado, in Contributions to economic geology: U.S. Geological Survey Bulletin 1112-E, p. 129-180. J ordan, P.R., Bloyd, R.M., and Daddow, RB, 1984, An assessment of cumulative impacts of coal mining on the hydrology in part of the Powder River structural basin. Wyoming; a progress report: U.S. Geological Survey Water-Resources Investigations Report 83-4235, 25 p. Kenny, J .F., Bevans, HE, and Diaz, A.M., 1982. Physical and hydro- logic environments of the Mulberry coal reserves in eastern Kan- sas: U.S. Geological Survey Water-Resources Investigations Report 82—4074. 50 p. Kenny, J .F., and McCauley, J .R., 1983, Applications of remote-sensing techniques to hydrologic studies in selected coal-mined areas of southeastern Kansas: U.S. Geological Survey Water-Resources Investigations Report 83-4007, 33 p. Kidd, R.E., and Bossong, GR, 1986, Applications of the precipitation- runoff model in the Warrior coal field: U.S. Geological Survey Open-File Report 85—678. 65 p. Kidd, R.E., and Hill, T.J., 1983, A summary of selected publications project activities, and data sources related to hydrology in the Warrior and Plateau coal fields of Alabama: U.S. Geological Survey Open-File Report 82—913, 80 p. Kiesler, Jay, Quinones, Ferdinand, Mull, BS, and York, K.L., 1983, Hydrology of area 13, Eastern Coal Province. Kentucky, Virginia, and West Virginia: U.S. Geological Survey Water-Resources In- vestigations Open-File Report 82-505, 112 p. Kilpatrick, F.A., 1984, Coal hydrology program of the U.S. Geological Survey, in Houghton, R.L., and others, eds., Symposium on the Geology of Rocky Mountain Coal: Bismarck, North Dakota Geo- logical Society. p. 80. Kilpatrick, F.A., and others, in press, Coal basin modeling for hydrologic impact assessment, Part A—General description of hydrology, geology. and data collection: U.S. Geological Survey Water-Resources Investigations Report 85-4123. 176 Kircher, J .E., 1982, Sediment transport and source areas of sediment and runoff, Big Sandy River basin, Wyoming: U.S. Geological Survey Water-Resources Investigations Report 81—72, 51 p. Kircher, J .E., Choquette, A.F., and Richter, B.D., 1985, Estimation of natural streamflow characteristics in western Colorado: U.S. Geological Survey Water-Resources Investigations Report 85—4086, 28 p. Kirkham, R.M., and Ladwig, LR, 1980, Energy resources of the Denver and Cheyenne basins, Colorado—Resource characteristics, development potential, and environmental problems: Denver, Colorado Geological Society Environmental Geology Series 12, 258 p. Klarich, D.A., and Regele, S.M., 1980, Structure, general character- istics, and salinity relationships of benthic macroinvertebrate associations in streams draining the southern Fort Union Coal Field Region of southeastern Montana: Billings, Montana Depart- ment of Health and Environmental Sciences, 148 p. Kleinmann, R.L.P., Crerar, DA, and Pacelli, RR, 1981, Biogeochem- istry of acid mine drainage and a method to control acid forma- tion: Mining Engineering, v. 33, no. 3, p. 300—305. Knapton, J .R., and Ferreira, RR, 1980, Statistical analyses of surface- water-quality variables in the coal area of southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 80—40, 128 p. Knight, A.L., and Newton, J .G., 1977, Water-related problems in coal- mine areas of Alabama: U.S. Geological Survey Water-Resources Investigations Report 76—130, 56 p. [Available only from National Technical Information Service, Springfield, Va., as PB—271 527.] Kuhn, Gerhard, 1982, Statistical summaries of water-quality data for two coal areas of Jackson County, Colorado: U.S. Geological Survey Open-File Report 82—121, 23 p. Kuhn, Gerhard, Daddow, P.B., Craig, G.S., Jr., and others, 1983, Hydrology of Area 54, Northern Great Plains, and Rocky Moun- tain Coal Provinces, Colorado and Wyoming: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—146, 95 p. Lambing, J .H., 1984, Sediment yields in eastern Montana—Summary of data and proposed techniques for estimating sediment yields from small, ungaged watersheds: U.S. Geological Survey Water- Resources Investigations Report 84—4200, 45 p. Lambing, J .H., and others, 1987, Hydrology of area 43, Northern Great Plains and Rocky Mountain Coal Provinces, Montana: U.S. Geological Survey Water-Resources Investigations Open-File Report 85—88, 95 p. Landis, ER, and Van Eck, O.J., 1965, Coal resources of Iowa: Iowa City, Iowa Geological Survey Technical Paper 4, 141 p. Larimore, R.W., 1974, Stream drift as an indication of water quality: American Fisheries Society Transactions, v. 103, no. 3, p. 507 —517 . Larson, LR, 1985, Water quality of the North Platte River, east- central Wyoming: U.S. Geological Survey Water-Resources In- vestigations Report 84-4172, 85 p. Larson, LR, and Daddow, R.L., 1984, Groundwater-quality data from the Powder River structural basin and adjacent areas, northeast- ern Wyoming: U.S. Geological Survey Open-File Report 83—939, 56 p. Larson, LR, and Zimmermann, E.A., 1981, Water resources of upper Separation Creek basin, south-central Wyoming: U.S. Geological Survey Water-Resources Investigations Report 80—85, 69 p. Leavesley, G.H., Lichty, R.W., Troutman, B.M., and Saindon, LG, 1983, Precipitation-runoff modeling system—User’s manual: U.S. Geological Survey Water-Resources Investigations Report 83—4238, 207 p. Lee, R.W., 1979, Ground-water-quality data from the northern Powder River basin, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Open-File Report 79—1331, 55 p. SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 1980, Geochemistry of water in the Fort Union Formation of the northern Powder River basin, southeastern Montana: U.S. Geological Survey Water-Supply Paper 2076, 17 p. Lee, R.W., Slagle, SE, and Stimson, J .R., 1981, Magnitude and chemical quality of base flow of Otter Creek-Tongue River, and Rosebud Creek, southeastern Montana, October 26—November 5, 197 7: U.S. Geological Survey Water-Resources Investigations Report 80—1298, 25 p. Leist, D.W., Quinones, Ferdinand, Mull, D.S., and Young, Mary, 1982, Hydrology of area 15, Eastern Coal Province, Kentucky and Ten- nessee: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—809, 81 p. Lenfest, L.W., Jr., 1985, Evapotranspiration rates at selected sites in the Powder River Basin, Wyoming and Montana: U.S. Geolog- ical Survey Water-Resources Investigations Report 82-4105, 23 p. 1987, Recharge of shallow aquifers through two ephemeral stream channels in northeastern Wyoming: U.S. Geological Survey Water-Resources Investigations Report 85—4311, 38 p. Lessing, Peter, and Hobba, W.A., Jr., 1981, Abandoned coal mines in West Virginia as sources of water supplies: Morgantown, West Virginia Geological and Economic Survey Circular C—24, 18 p. Levings, G.W., 1981a, Selected hydrogeologic data from the Northern Great Plains area of Montana: U.S. Geological Survey Open-File Report 81—534, 241 p. 1981b, Selected drill-stem-test data from the Northern Great Plains area of Montana: U.S. Geological Survey Open-File Report 81—326, 20 p. _1983, Potential effects of surface coal mining on the hydrology of the Greenleaf-Miller area, Ashland coal field, southeastern Mon- tana: U.S. Geological Survey Water-Resources Investigations Report 82—4101, 31 p. Lewis, B.D., and Hotchkiss, W.R., 1981, Thickness, percent sand, and configuration of shallow hydrogeologic units in the Powder River basin, Montana and Wyoming: U.S. Geological Survey Miscella- neous Investigations Series Map I-1317, scale 1:1,000,000, 6 sheets. Lewis, B.D., and Roberts, RS, 1978, Geology and water-yielding characteristics of rocks of the northern Powder River basin, southeastern Montana: U.S. Geological Survey Miscellaneous In- vestigations Series Map I—847—D, scale 1:250,000, 2 sheets. Lines, G.C., 1985, The ground-water system and possible effects of underground mining in the Trail Mountain area, central Utah: U.S. Geological Survey Water-Supply Paper 2259, 32 p. Lines, G.C., and Glass, W.R., 1975, Water resources of the thrust belt of western Wyoming: U.S. Geological Survey Hydrologic In- vestigations Atlas HA-539, various scales, 3 sheets. Lines, G.C., and Morrissey, D.J., 1983, Hydrology of the Ferron Sand- stone aquifer and effects of proposed surface-coal mining in Castle Valley, Utah, with a section on Stratigraphy, by T.A. Ryer, and a section on Leaching of overburden, by RH. Fuller: U.S. Geo- logical Survey Water-Supply Paper 2195, 40 p. Lines, G.C., and others, 1984, Hydrology of Area 56, Northern Great Plains and Rocky Mountain Coal Provinces, Utah and Colorado: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—38, 69 p. Litke, D.W., 1983, Suspended sediment in selected streams of south- eastern Montana: U.S. Geological Survey Water-Resources In- vestigations Report 82—4087, 52 p. Livingston, R.K., 1970, Evaluation of the streamflow-data program in Colorado: U.S. Geological Survey Open-file Report, 76 p. 1981, Rainfall-runoff modeling and preliminary regional flood characteristics of small rural watersheds in the Arkansas River basin in Colorado: U.S. Geological Survey Water-Resources In- vestigations Report 80—112, 48 p. [Available only from National Technical Information Service, Springfield, Va, as PB—81 224 313.] REFERENCES CITED Livingston. R.K.. and Minges, DR. 1987, Techniques for estimating regional flood characteristics of small rural watersheds in the plains region of eastern Colorado: U.S. Geological Survey Water- Resources Investigations Report 87—4094, 72 p. Lovell. H.L., 1973. An appraisal of neutralization processes to treat coal mine drainage: Washington, D.C., Environmental Protection Agency Technology Series, EPA—670/2—73—093, 364 p. [Available only from National Technical Information Service, Springfield, Va., as PB—231 249/4GA.] Lowe, R.L., 1974, Environmental requirements and pollution tolerance of freshwater diatoms: Cincinnati, Ohio, U.S. Environmental Pro- tection Agency. EPA—670/4-74—005, 333 p. Lowham, H.W., 1976, Techniques for estimating flow characteristics of Wyoming streams: U.S. Geological Survey Water-Resources Investigations Report 76-112, 83 p. 1978, An analysis of stream temperatures, Green River basin, Wyoming: U.S. Geological Survey Water-Resources Investiga- tions Report 78-13, 50 p. 1982, Streamflows and channels of the Green River basin, Wyoming: U.S. Geological Survey Water-Resources Investiga- tions Report 81-71, 81 p. _1988, Streamflows in Wyoming. U.S. Geological Survey Water- Resources Investigations Report 88—4045, 78 p. Lowham, H.W., DeLong, L.L., Collier, K.R., and Zimmerman, E.A., 1982, Hydrology of Salt Wells Creek—A plains stream in south- western Wyoming: U.S. Geological Survey Water-Resources In- vestigations 81-62, 52 p. Lowham, H.W., DeLong, L.L., Peter, K.D., Wangsness, D.J., Head, W.J., and Ringen, B.H., 1976, A plan for study of water and its relation to economic development in the Green River and Great Divide basins in Wyoming: U.S. Geologiéal Survey Open-File Report 76—349. 110 p. Lowharn, H.W., Peterson, D.A., Larson, L.R., Zimmerman, E.A., Ringen, B.H.. and Mora, KL. 1985, Hydrology of area 52, Rocky Mountain Coal Province, Wyoming, Colorado, Idaho, and Utah: U.S. Geological Survey Water-Resources Investigations Open-File Report 83-761, 96 p. Lowry, M.E., 1981, The relative importance of regional and local ground-water systems in the Powder River structural basin, Wyoming and Montana [abs]: Laramie, Wyo., Tenth Annual Rocky Mountain Ground Water Conference, 1981, Proceedings, p. 71. Lowry, M.E., and Rankl, J.G., 1987, Hydrology of the White Tail Butte area, northern Campbell County, Wyoming: U.S. Geological Survey Water-Resources Investigations Report 82—4117, 55 p. Lowry, M.E., Wilson, J .F., Jr., and others, 1986, Hydrology of area 50, Northern Great Plains and Rocky Mountain Coal Provinces. Wyoming and Montana; U.S. Geological Survey Water-Resources Investigations Open-File Report 83—545, 131 p. Lumb, A.M., 1983, Procedures for assessment of cumulative impacts of coal mining on the hydrologic balance: U.S. Geological Survey Open-File Report 82—334, 50 p. Lusby, G.C. and Toy, T.J., 1976, An evaluation of surface-mine spoils area restoration in Wyoming using rainfall simulation: Earth Sur- face Processes, v. 1, no. 4, p. 375—386. MacCartney, J .C., and Whaite, R.H., 1969, Pennsylvania anthracite refuse—A survey of solid waste from mining and preparation: U.S. Bureau of Mines Information Circular 8409, 77 p. MacCary, L.M., Cushing, E.M., and Brown, D.L., 1983, Potentially favorable areas for large-yield wells in the Red River Formation and Madison Limestone in parts of Montana, North Dakota, South Dakota, and Wyoming: U.S. Geological Survey Professional Paper 1273—E, p. E1—E13. Marcher, M.V., Bergman, D.L., Slack, L.J., Blumer, SR, and Goe- maat, R.L., 1987, Hydrology of Area 41, Western Region, Interior Coal Province, Oklahoma and Arkansas: U.S. Geological Survey Water-Resources Investigations Open-File Report 84-129. 86 p. 177 Marcher, M.V., Bergman, D.L., Stoner, J .D., and Blumer, S.P.. 1983a, Preliminary appraisal of the hydrology of the Blocker area, Pitts- burg County. Oklahoma: U.S. Geological Survey Water-Resources Investigations Open-File Report 81-1187, 48 p. ____1983b, Preliminary appraisal of the hydrology of the Rock Island area, Le Flore County, Oklahoma: U.S. Geological Survey Water-Resources Investigations Report 83-4013, 35 p. 1983c. Preliminary appraisal of the hydrology of the Red Oak area, Latimer County, Oklahoma: U.S. Geological Survey Water- Resources Investigations Report 83—4166, 44 p. Marcher, M.V., Huntzinger, T.L., Stoner, J .D., and Blumer, SR, 1982, Preliminary appraisal of the hydrology of the Stigler area, Haskell County, Oklahoma: U.S. Geological Survey Water-Resources In- vestigations Report 82—4099, 37 p. Marcher, M.V., Kenny, J .F., and others, 1984, Hydrology of Area 40, Western Region, Interior Coal Province, Kansas, Oklahoma, and Missouri: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—266, 97 p. Martin, J .D.. Duwelius, RR, and Crawford, C.G., 1987. Effects of surface coal mining and reclamation on the geohydrology of six small watersheds in west-central Indiana: U.S. Geological Survey Open-File Report 87-210, 99 p. Matson, RE, and Blumer, J .W., 1973. Quality and reserves of strip- pable coal, selected deposits, southeastern Montana: Helena, Mon- tana Bureau of Mines and Geology Bulletin 91, 135 p. Maura. W.S., 1985, Selected trace-element data for streams in the southern Yampa River basin, northwestern Colorado: U.S. Geo— logical Survey Open-File Report 85-192, 154 p. May, V.J., and others, 1981, Hydrology of area 18, Eastern Coal Prov- ince, Tennessee: U.S. Geological Survey Water-Resources In- vestigations Open-File Report 81—492, 78 p. May, V.J., and others, 1983, Hydrology of area 21, Eastern Coal Prov- ince, Tennessee, Alabama, and Georgia: U.S. Geological Survey Water-Resources Investigations Open-File Report 82—679. 92 p. McCabe, J .A., 1962, Floods in Kentucky—Magnitude and frequen- cy: U.S. Geological Survey Information Circular 9, 196 p. McCain, J.F., and Jarrett, RD, 1976. Manual for estimating flood characteristics of natural-flow streams in Colorado: Colorado Water Conservation Board Technical Manual 1. 77 p. McClymonds, N .E., 1982, Hydrology of the Prairie Dog Creek drainage basin, Rosebud and Big Horn Counties. Montana: U.S. Geological Survey Water-Resources Investigations Report 81—37, 64 p. [Available only from National Technical Information Service. Springfield, Va., as PB—82 224 213.] ______1984a, Potential effects of surface coal mining on the hydrology of the Corral Creek area, Hanging Woman Creek coal field, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 83—4260, 53 p. _____1984b, Potential effects of surface coal mining on the hydrology of the West Otter area, Ashland and Birney-Broadus coal fields, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 84—4087, 70 p. 1985, Potential effects of surface coal mining on the hydrology of the Horse Creek area, Sheridan and Moorhead coal fields, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 84-4239, 61 p. McWhorter, D.B., Rowe, J .W., Van Liew, M.W., Chandler, R.L., Skogerboe, R.K.. Sunada, D.K., and Skogerboe, G.B., 1979, Sur- face and subsurface water-quality hydrology in surface mined watersheds, Part 1: U.S. Environmental Protection Agency, EPA-600/7-79-193a, 215 p. [Available from National Technical Information Service. Springfield. Va., as PB-80 142 003.] Miller, E.M., 1978, Techniques for estimating magnitude and frequen- cy of floods in Virginia: U.S. Geological Survey Water-Resources Investigations Report 78-5, 83 p. Miller, M.E., Bergantino. R.N., Vermel, W.M., Schmidt, S.A., Botz, M.K., Bahls, L.L., and Bahls, P.A., 1978. Regional assessment 178 of the saline seep problem and a water-quality inventory of the Montana Plains: Butte, Montana Bureau of Mines and Geology Report 42, 24 p. Miller, W.R., 1979, Water resources of the central Powder River area of southeastern Montana: Butte, Montana Bureau of Mines and Geology Bulletin 108, 65 p. 1981, Water resources of the southern Powder River area, southeastern Montana: Butte, Montana Bureau of Mines and Geology Memoir 47, 53 p. Moran, S.R., and Cherry. J .A., 1977, Subsurface-water chemistry in mined-land reclamation—Key to development of a productive postmining landscape: Second Annual General Meeting of the Canadian Land Reclamation Association, Edmonton, Alberta, 1977, Proceedings, p. IV.1—IV.29. Moran, S.R., Cherry, J.A., Rehm, B.W., and Groenewold, G.I-I., 1979, Hydrologic impacts of surface mining of coal in western North Dakota: National Symposium on Surface Mining Hydrology, Sedimentology, and Reclamation, Lexington, Ky., 1979, Proceed- ings, p. 57—65. Moran, S.R., Groenewold, G.H., and Cherry, J.A., 1978, Geologic, hydrologic, and geochemical concepts and techniques in over- burden characterization for mined-land reclamation: Grand Forks, North Dakota Geological Survey Report of Investigation 63, 152 p. Morrissey, D.J., Lines, G.C., and Bartholoma, SD, 1980, Three- dimensional digital-computer model of the Ferron Sandstone aquifer near Emery, Utah: U.S. Geological Survey Water- Resources Investigations Report 80-62, 109 p. [Available only from National Technical Information Service, Springfield, Va., as PB-81 128 449.] Myers, R.G., and Villanueva, E.E., 1986, Geohydrology of the aquifers that may be affected by the surface mining of coal in the Fruitland Formation in the San Juan basin, northwestern New Mexico: U.S. Geological Survey Water-Resources Investigations Report 85-4251, 41 p. Naftz, D.L., 1985, Assessment of postmining ground-water quality at Wyoming coal mines [abs]: Gillette, Wyo., Second Hydrology Symposium on Surface Coal Mining in the Northern Great Plains, 1985, Proceedings, p. 28. N aten, R.W., and Fuller, R.H., 1981, Selected biological characteristics of streams in the southeastern Uinta basin, Utah and Colorado: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—644, 38 p. National Academy of Sciences, National Academy of Engineering, 1973, Water quality criteria, 1972: U.S. Environmental Protec- tion Agency Report EPA-R3—73—033, 594 p. 1974, Rehabilitation potential of western coal lands—A report to the Energy Policy Project of the Ford Foundation: Cambridge, Mass, Ballinger Publishing, National Academy of Sciences, Na- tional Academy of Engineering, 198 p. National Oceanic and Atmospheric Administration, 1982, Monthly normals of temperature, precipitation, and heating and cooling degree days 1951—80: Asheville, N.C., Climatography of the United States no. 81 (by State). Nielsen, G.F., ed., 1984, Keystone coal industry manual: New York, McGraw-Hill Mining Publications, Mining Information Services, 1.388 p. Nordstrom, D.K., 1977, Hydrogeochemical and microbiolog'cal factors affecting the heavy metal chemistry of an acid mine drainage system: Palo Alto, Calif., Stanford University, Ph. D. disserta- tion, 210 p. N orris, J .M., 1987, Surface water-quality characteristics in the upper North Fork Gunnison River basin, Colorado: U.S. Geolog- ical Survey Water-Resources Investigations Report 86—4152, 42 p. SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Norris, J .M., and Maura, W.S., 1985, Water-quality data for streams in the upper North Fork of the Gunnison River, Colorado: U.S. Geological Survey Open-File Report 85-190, 122 p. Norris, J .M., and Parker, RS, 1985, Calibration procedure for a daily flow model of small watersheds with snowmelt runoff in the Green River coal region of Colorado: U.S. Geological Survey Water- Resources Investigations Report 83—4263, 32 p. North Dakota Geological Survey, 1981, Energy development in west- ern North Dakota, in Gerhard, Lee, ed., North Dakota Geological Survey newsletter: Grand Forks, N. Dak., p. 3-4. Office of the Federal Register, National Archives and Records Serv- ice, 1982, The United States government manual 1982/83: Washington, D.C., U.S. Government Printing Office, 913 p. Oihus, C.A., 1983, A history of coal mining in North Dakota, 1873—1982: Grand Forks, North Dakota Geological Survey Educa- tional Series 15, 100 p. Olin, DA, and Bingham, R.H., 1977, Flood frequency of small streams in Alabama: Montgomery, Alabama Highway Department, HPR 83, 44 p. Omang, R.J., and Parrett, Charles, 1984, A method for estimating mean annual runoff of ungaged streams based on basin character- istics in central and eastern Montana: U.S. Geological Survey Water-Resources Investigations Report 84—4143, 15 p. Omang, R.J., Parrett, Charles, and Hull, J.A., 1982, Mean annual runoff and peak-flow estimates based on channel geometry of streams in southeastern Montana: U.S. Geological Survey Water- Resources Investigations Report 82-4092, 33 p. Oriel, SS, and Mudge, M.R., 1956, Problems of lower Mesozoic stratigraphy in southeastern Colorado, in Guidebook to the geology of the Raton Basin, Colorado: Denver, Rocky Mountain Association of Geologists, p. 19—24. Osterkamp, W.R., Carey, W.P., Hupp, C.R., Bryan, B.A., 1984, Move- ment of tractive sediment from disturbed lands: Proceedings of the Conference on Water Resource Development, American Socie- ty of Civil Engineers, Hydraulic Division, p. 59—63. Osterkamp, W.R., and Hedman, ER, 1979, Discharge estimates in surface-mined areas using channel-geometry techniques, in Sym- posium on Surface-Mine Hydrology, Sedimentology and Reclama- tion, Proceedings: Lexington, University of Kentucky Bulletin 119, p. 43-49. Ott, A.N., 1986, Estimating iron and aluminum content of acid mine discharge from a north-central Pennsylvania coal field by use of acidity titration curves: U.S. Geological Survey Water-Resources Investigations Report 84—4335, 25 p. _in press, Dual acidity titration curves—Fingerprint, indicator of redox state, and estimator of iron and aluminum content of acid mine discharge and related waters, in Subitzky, Seymour, ed., Selected papers in the hydrologic sciences: U.S. Geolog'cal Survey Water-Supply Paper 2330. Pacific Southwest Inter-Agency Committee Water Management Sub- committee, 1968, Factors affecting sediment yields, in Factors affecting sediment yield and measures for the reduction of ero- ' sion and sediment yield: Water Management Subcommittee, PSIAC, 10 p. Parker, R.S., and Carey, W.P., 1980, The quality of water discharg- ing from the New River and Clear Fork basins, Tennessee: U.S. Geological Survey Water-Resources Investigations Report 80-37, 52 p. Parker, R.S., and Norris, J .M., 1983, Simulated effects of anticipated coal mining on dissolved solids in selected tributaries of the Yampa River, northwestern Colorado: U.S. Geological Survey Water- Resources Investigations Report 83—4084, 72 p. 1989, Simulation of streamflow in small drainage basins in the southern Yampa River basin, Colorado: U.S. Geological Survey Water-Resources Investigations Report 88—4071, 47 p. REFERENCES CITED Parkhurst, D.L., Plummer, L.N., and Thorstenson, D.C., 1982, BALANCE—A computer program for calculating mass transfer for geochemical reactions in ground water: U.S. Geological Survey Water-Resources Investigations Report 82—14, 29 p. [Available only from National Technical Information Service, Springfield, Va., as PBe82 255 902.] Parrett, Charles. Carlson, D.D., Craig, G.S., J r., and Chin, E.H., 1984, Floods of May 1978 in southeastern Montana and northeastern Wyoming: U.S. Geological Survey Professional Paper 1244, 74 p. Patterson. G.L., Fuentes, R.F., and Toler, LG, 1982, Hydrologic characteristics of strip-mined land reclaimed by sludge irrigation: U.S. Geological Survey Water-Resources Investigations Report 82—16, 34 p. Patterson, J .L., 1966, Magnitude and frequency of floods in the United States, Part 6A, Missouri River basin above Sioux City, Iowa: U.S. Geological Survey Water-Supply Paper 1679, 471 p. Peirce, L.B.. 1954, Floods in Alabama—Magnitude and frequency: U.S. Geological Survey Circular 342, 105 p. Perry, Harry, 1983, Coal in the United States—A status report: Science, v. 22, no. 4622, p. 377—384. Peterson, D.A., 1988, Statistical summary of the chemical quality of surface water in the Powder River coal basin, the Hanna coal field, and the Green River coal region, Wyoming: U.S. Geological Survey Water-Resources Investigations Report 84—4092, 109 p. ___in press, Invertebrate communities of small streams in north- eastern Wyoming: U.S. Geological Survey Water-Resources In- vestigations Report 85-4287. Peterson, D.A., Mora, K.L., Lowry, M.E., Rankl. J .G., Wilson, J .F., Jr., Lowham, H.W., and Ringen, B.H., 1987, Hydrology of Area 51, Northern Great Plains and Rocky Mountain Coal Provinces, Wyoming and Montana: U.S. Geological Survey Water-Resources Investigations Open-File Report 84-734, 73 p. Pfaff, C.L., Helsel, D.R., Johnson. DP, and Angelo, G.G., 1981, Assessment of water quality in streams draining coal-producing areas in Ohio: U.S. Geological Survey Water-Resources Investiga- tions Open-File Report 81-409, 98 p. Plantz, G.G., 1983, Selected hydrologic data, Kolob-Alton-Kaiparowits coal-fields area, south-central Utah: U.S. Geological Survey Open- File Report 83-871, 28 p. _1985, Hydrologic reconnaissance of the Kolob, Alton. and Kai- parowits Plateau coal fields, south-central Utah: U.S. Geological Survey Hydrologic Investigations Atlas HA—684, scale 1:250,000. Plummer, L.N., Jones, BR, and Truesdell. A.H., 1976, WATEQF— A FORTRAN IV version of WATEQF, a computer program for calculating chemical equilibrium of natural waters: U.S. Geological Survey Water-Resources Investigations 76-13, 63 p. [Available only from National Technical Information Service, Springfield, Va., as PB—261 027.] Pollard, B.C.. Smith, J .B., and Knox, 0.0., 1972, Strippable lignite reserves of North Dakota—Location, tonnage, and characteristics of lignite and overburden: U.S. Bureau of Mines Information Cir- cular 8537, 37 p. Pope. L.M., and Diaz, A.M., 1982, Quality-of-water data and statistical summary for selected coal-mined strip pits in Crawford and Cherokee Counties, southeastern Kansas: U.S. Geological Survey Open-File Report 82-1021, 28 p. Power, J .F., Bond, J .J ., Sandoval. F.M., and Willis, W.O., 1974, Nitrification in Paleocene shale: Science, v. 183, p. 1077—1079. Price, Don, 1984, Map showing selected surface-water data for the Huntington 30 x 60-minute Quadrangle, Utah: U.S. Geological Survey Miscellaneous Investigations Series Map I—1514, scale 1:100,000. Price, Don, and Arnow, Ted, 1974. Summary appraisals of the Na- tion’s ground-water resources—Upper Colorado Region: U.S. Geo logical Survey Professional Paper 813—0. p. 01-040. 179 Price, Don, and Miller, L.L., 1975, Hydrologic reconnaissance of the southern Uinta basin, Utah and Colorado: Salt Lake City, Utah Department of Natural Resources Technical Publication 49, 66 p. Price. Don, and others, 1987 . Hydrology of Area 57, Northern Great Plains and Rocky Mountain Coal Provinces, Utah and Arizona: U.S. Geological Survey Water-Resources Investigations Open-File Report 84—068, 63 p. Puente, Celso, and Atkins, J .T., 1986, Simulation of rainfall-runoff response in small coal-mined and in undisturbed watersheds in West Virginia: U.S. Geological Survey Open-File Report 86-321, '106 p. Puente, Celso, and Newton, J .G.. 1982, Hydrogeology of potential min- ing areas in the Warrior coal field, Alabama: U.S. Geological Survey Open-File Report 82-105, 117 p. Puente, Celso, Newton, J .G., and Bingham, B.H., 1982, Assessment of hydrologic conditions in potential coal-lease tracts in the War- rior coal field, Alabama: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—540, 43 p. Quinones. Ferdinand, Mull, D.S., York, Karen, and Kendall, Victoria, 1981, Hydrology of area 14, Eastern Coal Province, Kentucky: U.S. Geological Survey Water-Resources Investigations Open-File Report 81-137, 82 p. Quinones, Ferdinand, York, K.L., and Plebuch, R.O., 1983, Hydrology of area 34, Interior Coal Province, Eastern Region, Kentucky and Indiana: U.S. Geological Survey Water-Resources Investigations Open-File Report 82—638, 83 p. Ragsdale. J .O., 1982, Ground-water levels in Wyoming. 1971 through part of 1980: U.S. Geological Survey Open-File Report 82—859, 200 p. Rahn. P.H., 197 5, Ground water in coal strip-mine spoils, Powder River basin, in Proceedings of Fort Union Coal Field Symposium, v. 3, reclamation section: Missoula, University of Montana, Academy of Sciences, p. 348—361. Rainwater, F.H., 1962, Stream composition of the conterminous United States: U.S. Geological Survey Hydrologic Investigations Atlas HA-61, 3 sheets. Randolph, W.J., and Gamble, GR, 1976, Technique for estimating magnitude and frequency of floods in Tennessee: Nashville, Ten- nessee Department of Transportation, 52 p. Rankl. J .G., 1982, An empirical method for determining average soil infiltration rates and runoff, Powder River structural basin, Wyoming: U.S. Geological Survey Water-Resources Investiga- tions Report 81—76, 38 p. 1987, Analysis of sediment production from two small semiarid basins in Wyoming: U.S. Geological Survey Water-Resources In- vestigations Report 85—4314, 27 p. 1989, A point-infiltration model for estimating runoff from rain- fall on small basins in semiarid areas of Wyoming: U.S. Geological Survey Open-File Report 88—337, 60 p. Rankl, J.G., and Barker, D.S., 1977, Rainfall and runoff data from small basins in Wyoming: Cheyenne. Wyoming State Engineer, Wyoming Water Planning Program Report 17, 195 p. Rankl. J .G., and Lowry. M.E., in press, Regional ground-water flow in the Powder River structural basin, Wyoming and Montana: U.S. Geological Survey Water-Resources Investigations Report 85—4229. Raaem, A.C., 1983, Ground-water hydrology before, during, and after coal strip mining of a small watershed in Coshocton County, Ohio: U.S. Geological Survey Water-Resources Investigations Report 88-4155, 36 p. 1984, Ground-water hydrology and quality before and after strip mining of a small watershed in Jefferson County, Ohio: U.S. Geo- logical Survey Water-Resources Investigations Report 83—4215, 39 p. 180 Reed, L.A., 1980, Effects of strip mining the abandoned deep Anna S Mine on the hydrology of Babb Creek, Tioga County, Penn- sylvania: U.S. Geological Survey Water-Resources Investigations Report 80—53, 41 p. [Available only from National Technical In- formation Service, Springfield, Va., as PB—81 121 337.] 1986, Verification of the PRMS sediment-discharge model: Federal Interagency Sedimentation Conference, 4th, Las Vegas, Nev., 1986, Proceedings, v. 2, p. 6-44-6—54. Rehm, B.W., Groenewold, G.H., and Morin, K.A., 1980, Hydraulic properties of coal and related materials, Northern Great Plains: Ground Water, v. 18, no. 6, p. 551—561. Richter, B.D., Kircher, J .E., Remmers, M.A., and Forst, B.A., 1984, Summary of basin and streamflow characteristics for selected basins in western Colorado and adjacent States: U.S. Geological Survey Open-File Report 84-137, 266 p. Ringen, B.H., 1984, Relationships of suspended sediment to stream- flow in the Green River basin, Wyoming: U.S. Geological Survey Water-Resources Investigations Report 84-4026, 14 p. Ringen, B.H., and Daddow, P.B., in press, Hydrology of the Powder River alluvium between Sussex, Wyoming, and Moorehead, Mon- tana: U.S. Geological Survey Water-Resources Investigations Report 89—4002. Ringen, B.H., Shown, L.M., Hadley, R.F., and Hinkley, T.K., 1979, Effect on sediment yield and water quality of a nonrehabilitated surface mine in north-central Wyoming: U.S. Geological Survey Water-Resources Investigations Report 79—47, 23 p. [Available only from National Technical Information Service, Springfield, Va., as PB-299 868.] Robertson, GE, and Smith, DC, 1981, Coal resources and reserves of Missouri: Rolla, Missouri Division of Geology and Land Survey Report of Investigations 66, 49 p. Robson, S.G., Romero, J .C., and Zawestowski, Stanley, 1981, Geologic structure, hydrology, and water quality of the Arapahoe aquifer in the Denver basin, Colorado: U.S. Geological Survey Hydrologic Investigations Atlas HA-647, scale 1:500,000, 3 sheets. Robson, S.G., and Saulnier, G.J., Jr., 1981, Hydrogeochemistry and simulated solute transport, Piceance basin, northwestern Colo- rado: U.S. Geological Survey Professional Paper 1196, 65 p. Roth, D.K., and Cooper, SC, 1985, Hydrology of Area 11, Eastern Coal Province, Ohio, Kentucky, and West Virginia: U.S. Geological Survey Water-Resources Investigations Open-File Report 84—233, 66 p. Roth, D.K., Engelke, M.J., Jr., and others, 1981, Hydrology of Area 4, Eastern Coal Province, Pennsylvania, Ohio, and West Virginia: U.S. Geological Survey Water-Resources Investigations Open-File Report 81-343, 62 p. Roybal, F.E., and others, 1983, Hydrology of Area 60, Northern Great Plains, and Rocky Mountain Coal Provinces, New Mexico, Colo- rado, Utah, and Arizona: U.S. Geological Survey Water-Resources Investigations Open-File Report 83-203, 80 p. Roybal, F...E Wells, J .G., Gold, R.L., and Flager, J .V., 1984, Hydrology of Area 62, Northern Great Plains and Rocky Moun- tain Coal Provinces, New Mexico and Arizona: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—698, 66 p. Roybal, G.H., and Campbell, F.W., 1982, Stratigraphic sequence in drilling data, Fence Lake area: Socorro, New Mexico Bureau of Mines and Mineral Resources Open-File Report 145, 56 p. Rozelle, R.B., Rostock, Robert. and Swain, Thomas, 1976, Studies on the acid-alkaline interaction on the north branch of the Susquehan- na River: Harrisburg, Pennsylvania Science and Engineering Foundation, 47 p. [Unpublished report available from Dr. Rozelle, Department of Chemistry, Wilkes College, Wilkes-Barre, Pa.] Rucker, S.J., IV, and DeLong, LL, 1987, Evaluation of selected surfacewater-quality stations in Wyoming: U.S. Geological Survey Water-Resources Investigations Report 82—4003, 72 p. SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Ruddy, B.C., and Britton, L.J., in press, Traveltime and reaeration of selected streams in the North Platte and Yampa River basins, Colorado: U.S. Geological Survey Water-Resources Investigations Report 88—4205. Runner, G.S., 1981, Technique for estimating magnitude and frequency of floods in West Virginia: U.S. Geological Survey Open-File Report 80—1218, 44 p. Sandberg, G.W., 1979, Hydrologic evaluation of the Alton reclamation- study site, Alton coal field, Utah: U.S. Geological Survey Open- File Report 79—346, 62 p. Sandoval, F.M., Bond, J .J ., Power, J .F., and Willis, W.O., 1973, Lignite mine spoils in the Northern Great Plains—Characteristics and potential for reclamation, in Wali, M.K., ed., Some environmen- tal aspects of strip mining in North Dakota: Grand Forks, North Dakota Geological Survey Educational Series 5, p. 1-24. Sandoval, F.M., and Gould, W.L., 1978, Improvement of saline- and sodium-affected disturbed lands, in Schaller, F.W., and Sutton, Paul, eds., Reclamation of Drastically Disturbed Lands Symposium, Wooster, Ohio, 1978, Proceedings: Madison, Wis., American Society of Agronomy, p. 485-504. Schneider, W.J., and others, 1965, Water resources of the Appalachian Region Pennsylvania to Alabama: U.S. Geological Survey Hydro- logic Investigations Atlas HA—198, various scales, 11 sheets. Scott, A.G., 1984, Analysis of characteristics of simulated flows from small surface-mined and undisturbed Appalachian watersheds in the 'I‘ug Fork basin of Kentucky, Virginia, and West Virginia: U.S. Geological Survey Water-Resources Investigations Report 84—4151, 169 p. Shaw, G.L., 1956, Subsurface stratigraphy of the Permian- Pennsylvanian beds, Raton Basin, Colorado, in Guidebook to the geology of the Raton Basin, Colorado: Denver, Rocky Mountain Association of Geologists, p. 14—18. Shomaker, J .W., Beaumont, EC, and Kottlowski, F.E., 1971, Strip- pable low-sulfur coal resources of the San Juan basin in New Mex- ico and Colorado: Socorro, New Mexico Bureau of Mines and Mineral Resources Memoir 25, 189 p. Shown, L.M., Frickel, D.G., Hadley, R.F., and Miller, R.F., 1981, Methodology for hydrologic evaluation of a potential surface mine—The Tsosie Swale basin, San Juan County, New Mexico: U.S. Geological Survey Water-Resources Investigations Open-File Report 81-74, 63 p. Shown, L.M., Frickel, D.G., Miller, R.F., and Branson, F.A., 1982, Methodology for hydrologic evaluation of a potential surface mine—The Loblolly Branch basin, Tuscaloosa County, Alabama: U.S. Geological Survey Water-Resources Investigations Report 82—50, 101 p. [Available only from National Technical Informa- tion Service, Springfield, Va., as PB—152 223]. Simon, J .L., 1981, The ultimate resource: Princeton, N .J ., Princeton University Press, 415 p. Slack, L.J., 1979, Baseline water quality of Iowa’s coal region: U.S. Geological Survey Open-File Report 79—980, 74 p. 1983, Hydrology of an abandoned coal-mining area near McCur- tain, Haskell County, Oklahoma: U.S. Geological Survey Water- Resources Investigations Report 83—4202, 117 p. Slagle, S.E., Lewis, B.D., and Lee, R.W., 1985, Ground-water resources and potential hydrologic effects of surface coal mining in the north- ern Powder River Basin, southeastern Montana: U.S. Geological Survey Water-Supply Paper 2239, 34 p. Slagle, SE, and others, 1983, Hydrology of area 49, Northern Great Plains and Rocky Mountain Coal Provinces, Montana and Wyo- ming: U.S. Geological Survey Water-Resources Investigations Open-File Report 82—682, 94 p. 1984, Hydrology of Area 45, Northern Great Plains and Rocky Mountain Coal Provinces, Montana and North Dakota: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—527, 90 p. REFERENCES CITED ___1986, Hydrology of area 48, Northern Great Plains and Rocky Mountain Coal Provinces, Montana and Wyoming: U.S. Geological Survey Water-Resources Investigations Open-File Report 84—141, 91 p. Slagle, SE, and Stimson, J .R., 1979, Hydrogeologic data from the northern Powder River Basin, southeastern Montana: U.S. Geo- logical Survey Water-Resources Investigations Report 7 9-1332, 111 p. Smith, R.E., 1976, Field test of a distributed watershed erosion/ sedimentation model, in Soil erosion—Prediction and control: Ankeny, Iowa, Soil Conservation Society of America, p. 201—209. Staubitz, W.W., 1981, Quality of surface water in the coal-mining areas of western Maryland and adjacent areas of Pennsylvania and West Virginia from April 1979 to June 1980: U.S. Geological Survey Open-File Report 81-812, 103 p. Staubitz, W.W., and Sobashinski, J .R., 1983, Hydrology of Area 6, Eastern Coal Province, Maryland, West Virginia, and Penn- sylvania: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—33, 71 p. Steele, T.D., 1978, Assessment techniques for modeling water quali- ty in a river basin affected by coal resource development, in Sym- posium on modeling the water quality of the hydrologic cycle: Baden, Austria, Institute for Applied Systems Analysis, 16 p. __._1979, An overview of river-basin assessment techniques in an energy-impacted region—Yampa River basin, Colorado and Wyoming. Water Supply and Management, v. 3, no. 3, p. 151-171. Steele, T.D., Bauer, D.P., Wentz, D.A., and Warner, J.W., 1976, An environmental assessment of impacts of coal development on the water resources of the Yampa River basin, Colorado and Wyoming—Phased work plan: U.S. Geological Survey Open-File Report 76—367, 17 p. 1979, The Yampa River basin, Colorado and Wyoming—A preview to expanded coal-resource development and impacts on regional water resources: U.S. Geological Survey Water-Resources Investigations Report 78—126, 133 p. Steele, T.D., and Hillier, DE, 1981, Assessment of impacts of pro- posed coal-resource and related economic development on water resources of the Yampa River basin, Colorado and Wyoming-A summary: U.S. Geological Survey Circular 839, 56 p. Stephens, Doyle, 1985, Why Scofield Reservoir is entrophic—Effects of nonpoint-source pollutants on the water-supply mervoir in Utah, in Perspectives on nonpoint-source pollution: Washington, D.C., US Environmental Protection Agency 440/5-85—001, p. 142—147. Stone, W.J., Lyford, F.P., Frenzel, P.F., Mizell, N.H., and Padgett, ET, 1983, Hydrogeology and water resources of the San Juan basin, New Mexico: Socorro, New Mexico Bureau of Mines and Mineral Resource Hydrologic Report 6, 70 p. Stoner, J .D., and Lewis, B.D., 1980, Hydrogeology of the Fort Union coal region, eastern Montana: U.S. Geological Survey Miscellane- ous Investigations Series Map I—1236, scale 1:500,000, 2 sheets. Stumm, Werner, and Morgan, J .J ., 1970, Aquatic chemistry; An in- troduction emphasizing chemical equilibria in natural waters: New York, John Wiley, 583 p. Sullavan, J .N., 1984, Low-flow characteristics of Kentucky streams, 1984: U.S. Geological Survey Open-File Report 84-705, 1 sheet. Sumsion, C.T., 197 9, Selected coal-related ground-water data, Wasatch Plateau-Book Cliffs area, Utah: U.S. Geological Survey Open-File Report 79-915, 25 p. Swenson, F.A., Miller, W.R., and Hodson, W.G., 197 6, Map showing configuration and thickness and potentiometric surface and water quality in the Madison Group, Powder River Basin, Wyoming and Montana: U.S. Geological Survey Miscellaneous Investigations Series Map I—847—C, scale 1:1,000,000, 2 sheets. Thomas, B.E., and Lindskov, KL, 1983, Methods for estimating peak discharge and flood boundaries of streams in Utah: U.S. Geological Survey Water-Resources Investigations Report 83—4129, 77 p. 181 Thomas, R.P., and Gold, R.L., 1982, Techniques for estimating flood- discharges for unregulated streams in New Mexico: U.S. Geological Survey Water-Resources Investigations Report 82—24, 42 p. [Available only from National Technical Information Serv- ice, Springfield, Va., as PB—82 264 953.] Toler, L.G., 1982, Some chemical characteristics of mine drainage in Illinois: U.S. Geological Survey Water-Supply Paper 2078, 47 p. 'Ibth, J., 1963, A theoretical analysis of groundwater flow in small drain- age basins: Journal of Geophysical Research, 68, p. 437 5—4387. Trumbull, James, 1960, Coal fields of the United States: U.S. Geolog- ical Survey Map, scale 1:5,000,000, 2 sheets. Turk, J .T., 1982, Thermodynamic controls on water quality of water from underground coal mines in Colorado: Water Resources Bulle- tin, v. 18, no. 1, p. 75—80. Turk, J .T., and Parker, RS, 1982, Water-quality characteristics of six small, semiarid watersheds in the Green River Coal Region of Colorado: U.S. Geological Survey Water-Resources Investiga- tions Report 81—73, 101 p. [Available only from National Technical Information Service, Springfield, Va., as PB-82 207 390.] Udis, Bernard, Adams, T.H., Hess, R.C., and Orr, D.V., 1977, Coal energy development in Moffat and Routt Counties of the Yampa River basin in Colorado—Projected primary and secondary economic impacts resulting from several coal-development futures: Unpublished Phase 11 Contract Completion Report (U.S. Geolog- ical Survey P.O. 12185), 342 p. [On file in the Colorado District Library of the U.S. Geological Survey, Water Resources Division, Denver, Colo.] U.S. Bureau of Census, 1981, 1980 Census of population, Volume 1—Characteristics of the population, Chapter A—Number of in- habitants, Part 36—North Dakota: U.S. Department of Commerce Report PCBO—l-A36, 37 p. U.S. Bureau of Land Management, 1974, 1975, 1977, North Dakota surface-minerals management quadrangles: Washington, D.C., scale 1:126,720, 3 sheets. 1975a, Resource and potential reclamation evaluation, Hanna basin study site, Hanna coal field, Wyoming: Denver, EMRIA Report 2-1975, 176 p. 1975b, Resource and potential reclamation evaluation, Otter Creek study site, Otter Creek coal field, Montana: Denver, EMRIA Report 1, 234 p. 1976, Resource and potential reclamation evaluation, Bisti West study site, Bisti coal field: Denver, EMRIA Report 5, p. 69—80, F1—F16. 1977a, Resource and potential reclamation evaluation, Bear Creek study area, West Moorhead coal field, Montana: Denver, EMRIA Report 8, 259 p. 1977b, Resource and potential reclamation evaluation, Bisti West study site, Bisti coal field—Summary: Denver, EMRIA Report 5, various pagination. ___1978, Resource and potential reclamation evaluation, Hanging Woman Creek study area [Montana]: Denver, EMRIA Report 12, 309 p. 1981a, Resource and potential reclamation evaluation, Kimbeto study area: Denver, EMRIA Report 17—77, p. J1—J7, L13—L18, Q1—Q19. 1981b, Resource and potential reclamation evaluation, Ojo En- cino study area: Denver, EMRIA Report 19—78, p. Cl7—C20, J1-J9, Q1-Q12. 1982, Resource and potential reclamation evaluation, Pumpkin Creek study area [Montana]: Denver, EMRIA Report 11, 64 p. 1983a, Southern Appalachian coal region draft environmental impact statement II: Washington, D.C., 188 p. 1983b, Uinta—Southwestern Utah round 11 draft environmen- tal impact statement: Salt Lake City, 415 p. 1984, Managing the Nation’s public lands, fiscal year 1983: Washington, D.C., U.S. Government Printing Office, 96 p. 182 1985, A review of the unsuitability criteria in Federal coal leas- ing: Washington, D.C., 105 p. U.S. Congress, 1984, Environmental protection in the Federal coal leasing program: Washington, D.C., Office of Technology Assess- ment, OTA-E—237, 154 p. U.S. Department of the Interior, 1968, Stream pollution by coal-mine drainage, upper Ohio River basin: Federal Water Pollution Con- trol Administration, 110 p. 1975, Final environmental impact statement [on] proposed federal coal leasing program: Washington, D.C., U.S. Government Printing Office, 402 p. 1981, Quality of water, Colorado River Basin: U.S. Department of the Interior Progress Report 10, 190 p. U.S. Energy Information Administration, Office of Coal, Nuclear, Elec- tric, and Alternate Fuels, 1984, Annual energy review 1983: Washington, D.C., U.S. Department of Energy report DOE/ EIA—0384(83), 262 p. U.S. Environmental Protection Agency, 1986a, Maximum contami- nant levels (Subpart B of part 14, national interim primary drinking-water regulations): U.S. Code of Federal Regulations, Title 40, Parts 100 to 149, revised as of July 1, 1986, p. 524—528. 1986b, Quality criteria for water 1986: Washington, D.C., Of- fice of Water Regulations and Standards, EPA 440/5—86—001, unpaginated. 1986c, Secondary maximum contaminant levels (Section 143.3 of part 143, national secondary drinking-water regulations): U.S. Code of Federal Regulations, Title 40, Parts 100 to 149, revised as of July 1, 1986, p. 587—590. U.S. Geological Survey, 1975, Plan of study of the hydrology of the Madison Limestone and associated rocks in parts of Montana, Nebraska, North Dakota, South Dakota, and Wyoming: U.S. Geo- logical Survey Open-File Report 75—631, 37 p. 1979, Plan of study of the Northern Great Plains regional aquifer-system analysis in parts of Montana, Nebraska, North Dakota, South Dakota, and Wyoming: U.S. Geological Survey Water-Resources Investigations Report 79—34, 20 p. _1983a, National water summary—Hydrologic events and issues: U.S. Geological Survey Water-Supply Paper 2250, 243 p. 1983b, Water-Resources Data, New Mexico, water year 1982: U.S. Geological Survey Water-Data Report NM—82-1, 659 p. 1985, National water summary of 1984: U.S. Geological Survey Water-Supply Paper 2275, 467 p. U.S. National Research Council, 1981a, Coal mining and ground-water resources in the United States: Washington, D.C., National Academy Press, 197 p. 1981b, Surface mining—Soil, coal, and society: Washington, D.C., National Academy Press, 233 p. U.S. Office of Surface Mining, 1979, Section 816.57 Hydrologic balance—Stream buffer zones: Federal Register, v. 44, no. 50, Tuesday, March 13, 1979. U.S. Soil Conservation Service, 1972, Water and related land resources, Dolores River basin, Colorado and Utah (a report based on a cooperative study by Colorado Water Conservation Board and U.S. Department of Agriculture): Portland, Ore., 1 v. 1974, Water and related land resources, San Juan River basin, Arizona, Colorado, New Mexico, and Utah (a report based on a cooperative study by Colorado Water Conservation Board and U.S. Department of Agriculture): Portland, Dre, 1 v. 1981a, Little Colorado River basin, Arizona and New Mexico, summary report (a cooperative study between the U.S. Depart- ment of Agriculture Soil Conservation Service, Economic Research Service, Forest Service, and the States of Arizona and New Mex- ico): Washington, D.C., 41 p. 1981b, Little Colorado River basin, Arizona and New Mexico, Water resources, Appendix II (a cooperative study between the U.S. Department of Agriculture Soil Conservation Service, SUMMARY OF THE NATIONAL COAL—HYDROLOGY PROGRAM, 1974—84 Economic Research Service, Forest Service, and the States of Arizona and New Mexico): Washington, D.C., 1 v. U.S. Weather Bureau, 1962, Mean number of thunderstorm days in the United States: U.S. Weather Bureau 'Ibchnical Paper 19, 22 p. Vaill, J .E., and Barks, J .H., 1981, Physical environment and hydro- logic characteristics of coal-mining areas in Missouri: U.S. Geo- logical Survey Water-Resources Investigations Report 80—67, 33 p. [Available only from National Technical Information Serv- ice, Springfield, Va., as PB—81 126 765.] Van Haveren, B.P., and Leavesley, G.H., 1979, Hydrologic modeling of coal lands: U.S. Bureau of Land Management and U.S. Geo- logical Survey, administrative report, 13 p. Van Voast, W.A., 1974, Hydrologic effects of strip coal mining in southeastern Montana—Emphasis, one year of mining near Decker: Helena, Montana Bureau of Mines and Geology Bulletin 93, 24 p. Van Voast, W.A., and Hedges, RB, 1975, Hydrogeologic aspects of existing and proposed strip coal mines near Decker, southeastern Montana: Helena, Montana Bureau of Mines and Geology Bulletin 97, 31 p. Van Voast, W.A., Hedges, R.B., and McDermott, J .J ., 1977, Hydro— geologic conditions and projections related to mining near Colstrip, southeastern Montana: Helena, Montana Bureau of Mines and Geology Bulletin 102, 43 p. 1978, Hydrologic characteristics of coal mine spoils, southeast- ern Montana: Bozeman, Montana University Joint Water Resources Research Center Report 94, 34 p. Waddell, K.M., Darby, D.W., and Theobald, S.M., 1985, Chemical and physical characteristics of water and sediment in Scofield Reser- - voir, Carbon County, Utah: U.S. Geological Survey Water-Supply Paper 2247, 36 p. Waddell, K.M., Dodge, J E Darby, D.W., and Theobald, S.M., 1982, Selected hydrologic data 1978-80, Price River basin, Utah: U.S. Geological Survey Open-File Report 82-916, 72 p. 1986, Hydrology of the Price River basin, Utah, with emphasis on selected coal-field areas: U.S. Geological Survey Water-Supply . Paper 2246, 51 p. Waddell, K.M., Sumsion, C.T., Butler, J .R., and Contratto, P.K., 1981, Hydrologic reconnaissance of the Wasatch Plateau—Book Cliffs coal-fields area, Utah: U.S. Geological Survey Water-Supply Paper 2068, 45 p. Waddell, K.M., Vickers, H.L., Upton, R.T., and Contratto, P.K., 1978, Selected hydrologic data, Wasatch Plateau—Book Cliffs coal-fields area, Utah: U.S. Geological Survey Open-File Report 7 8—121, 33 p. Wangsness, D.J., 1977, Physical, chemical, and biological relations of four ponds in the Hidden Water Creek strip-mine area, Powder River Basin, Wyoming: U.S. Geological Survey Water-Resources Investigations Report 77—72, 48 p. Wangsness, D.J., Crawford, C.G., Wilber, W.G., Miller, R.L., Archood, L.D., and Nutter, L.J., 1981a, Hydrology of area 33, Eastern Region, Interior Coal Province, southwestern Indiana and north- ern Kentucky: U.S. Geological Survey Open-File Report 81—423, 84 p. Wangsness, D.J., Miller, R.L., Bailey, Z.C., and Crawford, C.G., 1981b, Hydrology of area 32, Eastern Region, Interior Coal Province, In- diana: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—498, 76 p. Wangsness, D.J., and others, 1983, Hydrology of area 30, Eastern Region, Interior Coal Province, Indiana and Illinois: U.S. Geo- logical Survey Water-Resources Investigations Open-File Report 82—1005, 82 p. Wangsness, D.J., and Peterson, D.A., 1981, Behavioral and catastrophic drift of invertebrates in two streams in northeastern Wyoming: U.S. Geological Survey Open-File Report 80—1101, 13 p. Ward, J .R., 1976, Preliminary results of preimpoundment water- quality studies in the Tioga River basin, Pennsylvania and New REFERENCES CITED York: U.S. Geological Survey Water-Resources Investigations Report 76-66, 79 p. __1981, Preimpoundment water quality in the Tioga River basin: U.S. Geological Survey Water-Resources Investigations Report 81-1, 142 p. Water, Waste, and Land, Ltd., 1980, Hydrology, geology, and water quality in vicinity of the Maxwell and Allen Mines, Las Animas County, Colorado: Fort Collins, 0010., Final report to CF&I Steel Corp., 1 v. Weber, E.E., and Bartlett, W.P., J r., 1976, Floods in Ohio, magnitude and frequency: U.S. Geological Survey Open-File Report 76-768, 73 p. Wedge, W.K., Bhatia, D.M.S., and Rueff, A.W., 1976, Chemical analysis of selected Missouri coals and some statistical implica- tions: Rolla, Missouri Division of Geology and Land Survey Report of Investigations 60, 36 p. Weeks, J .B., 1978, Digital model of ground-water flow in the Piceance basin, Rio Blanco and Garfield Counties, Colorado: U.S. Geological Survey Water-Resources Investigations Report 78-46, 108 p. [Available only from National Technical Information Service, Springfield, Va., as PB-284 688.] Weeks, J .B., and Welder, EA, 1974, Hydrologic and geophysical data from the Piceance basin, Colorado: Denver, Colorado Water Con- servation Board Basic-Data Release 35, 121 p. Weiss, J .S., and Razem, A.C., 1984, Simulation of ground-water flow in a mined watershed in eastern Ohio: Ground Water, v. 22, no. 5, p. 549-560. Weiss, L.S., 1984, Technique for estimating ground-water drainage to surface coal-mine excavations [abs.]: Annual Midwest Ground- water Conference, 29th, Lawrence, Kans., Proceedings, 1 p. Weiss, L.S., Galloway, D.L., and Ischii, A.L., 1986, Technique for predicting ground-water discharge to surface coal mines and resulting change in head: U.S. Geological Survey WatenResources Investigations Report 86-4156, 232 p. Welder, F.A., and Saulnier, G.J., Jr., 1978, Geohydrologic data from twenty-four test holes drilled in the Piceance Basin, Rio Blanco County, Colorado, 197 5—76: U.S. Geological Survey Open-File Report 78—734, 172 p. Welder, GE, 1968, Ground-water reconnaissance of the Green River basin, southwestern Wyoming: U.S. Geological Survey Hydrologic Investigations Atlas HA-290, scale 1:250,000, 3 sheets. Welder, G.E., and McGreevy, L.J., 1968, Ground-water reconnaissance of the Great Divide and Washakie basins and some adjacent areas, southwestern Wyoming: U.S. Geological Survey Hydrologic In- vestigations Atlas HA-219, scale 1:250,000, 3 sheets. Wells, D.K., 1982, Ground-water data from selected wells in alluvial aquifers, Powder River basin, northeastern Wyoming: U.S. Geological Survey Open-File Report 82-856, 35 p. Wells, D.K., Busby, J .F., and Glover, KC, 1979, Chemical analyses of water from the Minnelusa Formation and equivalents in the Powder River basin and adjacent areas, northeastern Wyoming: Cheyenne, Wyoming State Engineer, Wyoming Water Planning Program Report 18, 27 p. Wentz, D.A., 1974a, Effect of mine drainage on the quality of streams in Colorado, 1971-7 2: Denver, Colorado Water Conservation Board, Colorado Water Resources Circular 21, 117 p. __._1974b, Stream quality in relation to mine drainage in Colorado, in Hadley, R.F., and Snow, D.T., eds., Water resources problems related to mining: Minneapolis, Minn., American Water Resources Association Proceedings Series 18, p. 158-173. Wentz, D.A., and Steele, T.D., 1976, Wyoming-An area of acceler- ated coal development: Pacific Grove, Calif, Conference on Water for Energy Development, 28 p. 1980, Analysis of stream quality in the Yampa River basin, Colorado and Wyoming: U.S. Geological Survey Water-Resources Investigations Report 80—8, 161 p. [Available only from National Technical Information Service, Springfield, Va., as PB—81 108 904.] Wetzel, K.L., and Bettandorff, J .M., 1986, Techniques for estimating streamflow characteristics in the Eastern and Interior Coal 183 Provinces: U.S. Geological Survey Water-Supply Paper 2276, 80 p. Wetzel, K.L., and Hoffman, S.A., in press, Distribution of water- quality characteristics that may indicate the presence of acid mine drainage in the Eastern Coal Province of the United States: U.S. Geological Survey Hydrologic Investigations Atlas HA—705. Wilber, W.G., and Boje, R.R., 1983, Reconnaissance for determining effects of land use and surficial geology on concentrations of selected elements on streambed materials from the coal-mining region, southwestern Indiana, October 1979 to March 1980: U.S. Geological Survey Water-Resources Investigations Report 82—4013, 45 p. Wilber, W.G., Renn, D.E., and Crawford, C.G., 1985, Effects of land use and surficial geology on flow and water quality of streams in the coal-mining region of southwestern Indiana, October 1979 through September 1980: U.S. Geological Survey Water-Resources Investigations Report 85—4234, 49 p. Wilhm, J .L., and Dorris, T.C., 1966, Species diversity of benthic macroinvertebrates in a stream receiving domestic and oil refin- ing effluents: American Midland Naturalist, v. 76, no. 2, p. 427-449. Williams, E.G., Rose, A.W., Parizek, R.R., and Waters, S.A., 1982, Factors controlling the generation of acid mine drainage: Univer- sity Park, Pennsylvania State University, Pennsylvania Mining and Mineral Resources Research Institute, 1 v. Williams, R.S., and Hammond, S.E., 1988, Soil-water hydrology and geochemistry of a coal spoil at a reclaimed surface mine in Routt County, Colorado: U.S. Geological Survey Water-Resources In- vestigations Report 86—4350, 100 p. Winczewski, L.M., 1977, Western North Dakota lignite strip mining processes and resulting subsurface characteristics: Grand Forks, University of North Dakota, M.S. thesis, 433 p. Wood, G.H., Kehn, ’1‘.M., Carter, M.D., and Culbertson, W.C., 1983, Coal resource classification system of the U.S. Geological Survey: U.S. Geological Survey Circular 891, 65 p. Wood, W.A., 1984, Hydrogeologic data for selected test wells drilled in the Fort Union coal region, eastern Montana: U.S. Geological Survey Open-File Report 48—464, 63 p. Woods, RR, 1981, Modeled impacts of surface coal mining on dis- solved solids in the Tongue River, southeastern Montana: U.S. Geological Survey Water-Resources Investigations Report 81—64, 73 p. [Available only from National Technical Information Serv- ice, Springfield, Va., as PB—82 117 771.] Wyoming State Engineer’s Office, 1976, Investigation of recharge to groundwater reservoirs of northeastern Wyoming (Powder River Basin): Cheyenne, Old West Regional Commission report, 11 1 p. Yucel, Oner, 1982, Preliminary feasibility of coal slurry pipelines in Virginia, in Kilpatrick, G. A., and Matchett, Donald, eds., Water and energy, technical and policy issues: New York, American Society of Civil Engineers, p. 43—48. Zimmerman, E.A., and Collier, K.K., 1985, Ground-water data, Green River basin, Wyoming: U.S. Geological Survey Open-File Report 83-943, 511 p. Zuehls, E.E., 1987a, Hydrology of Area 27, Eastern Region, Interior Coal Province, Illinois: U.S. Geological Survey Water-Resources Investigations Open-File Report 84—707, 62 p. 1987b, Hydrology of Area 31, Eastern Region, Interior Coal Province, Illinois and Indiana: U.S. Geological Survey Water- Resources Investigations Open-File Report 85—342, 61 p. Zuehls, E.E., Fitgerald, K.K., and Peters, C.A., 1984, Hydrology of Area 28, Eastern Region, Interior Coal Province, Illinois: U.S. Geological Survey Water-Resources Investigations Open-File Report 83—544, 67 p. Zuehls, E.E., Ryan, G.L., Peart, D.E., and Fitzgerald, K.K., 1981a, Hydrology of Area 25, Eastern Region, Interior Coal Province, Illinois: U.S. Geological Survey Water-Resources Investigations Open-File Report 81—636, 66 p. 1981b, Hydrology of Area 35, Eastern Region, Interior Coal Province, Illinois and Kentucky: U.S. Geological Survey Water- Resources Investigations Open-File Report 81-403, 68 p. 'U.S. GOVERNMENT PRINTING OFFICE: 1990-0-773-976 Qsar Pe PROFESSIONAL PAPER 1464 PLATE 1 Prepared in cooperation with the DEPARTMENT OF THE INTERIOR U. S. BUREAU OF LAND MANAGEMENT . KN U. S. GEOLOGICAL SURVEY 95° 10?, 1 NW 1000 : ’ I. 85° ‘ r ’ , m" fi§§fi?’, FORT UNION . REGION * , -- \ L‘BgSHo. , _ ,A , MOPNTAINS REG .N ‘ ‘ OHIOVd : EGlON ‘ ‘ ‘ WI ‘DgRlVER ’ . \ fl eR QN‘RIVER“? I _ ‘ “DENVER U7“ 7’ - - '- V ”PLAINS 400 LOWER E41 * I fPRO g“ V A V CALIFORNIAN _, , , - I ; INCE _ EXPLANATION 30° COAL-REGION AREA AND NAME agg- COAL-AREA BOUNDARY AND NUMBER PHYSIOGRAPHICPROVINCE BOUNDARY AND NAME—Dashed where boundary is generalized or poorly known COALvPROVINCE NAME 995° 1 0032, Albers Equal Area PrOJection SCALE 117,500,000 200 300 ‘00 MILES KILOMETERS Modified from Fenneman and Johnson, 1946; and Trumbull, 1960 L ' | Survey 1:7,500,000 United States base map ‘ ONS, COAL AREAS, AND PHYSIOGRAPHIC PROVINCES OF MAP SHOWING COAL PROVINCES, SELECTED COAL REGI THE CONTIGUOUS UNITED STATES The Vanadiferous Zone of the Phosphoria Formation in Western Wyoming and Southeastern Idaho By V.E. MCKELVEY, ].D. STROBELL, JR, and AL. SLAUGHTER U.S. GEOLOGICAL SURVEY PROFESSIONAL PAPER 1465 UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON: 1986 DEPARTMENT OF THE INTERIOR DONALD PAUL HODEL, Secretary U.S. GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Cataloging-in-Publication Data McKelvey, V. E. (Vincent Ellis), 1916-— The vanadiferous zone of the Phosphoria Formation in western Wyoming and southeastern Idaho. (US. Geological Survey professional paper ; 1465) Bibliography: p. Supt. of Docs. no.: I. 19.16:1465 1. Vanadium ores—Wyoming. 2. Vanadium ores—Idaho. 3. Phosphoria Formation. 4. Geol~ ogy, Stratigraphic—Permian. I. Strobell, J. D. II. Slaughter, Archibald Logan, 1907- . III. Title. IV. Series: Geological Survey professional paper ; 1465. QE390.2.V36M35 1986 553.4’626’09787 86—607918 For sale by the Books and Open-File Reports Section, US. Geological Survey, Federal Center, Box 25425, Denver, CO 80225 CONTENTS Abstract ................................................................................................................................................... 1 Introduction ................................................................ Regional extent ............................................... Sublette Ridge deposits ........................................................................................... Geology .............................................................................. Vanadiferous zone ...................................... Resources ................................................................................................................ Paris-Bloomington deposits ............................................................................... Geology... .......... Vanadiferous zone ............................................................................................................. Effects of weathering on the vanadiferous zone ................................................................... Variations in thickness and vanadium content of the vandiferous zone .. Resources ......................................................................... Experimental mining and processing . Origin . ...................................... Conclusions ...................... Acknowledgments ....... __ ............. References cited ................................................... ILLUSTRATIONS Page PLATE 1. Profile of the vanadiferous zone outcrop at Sublette Ridge, Wyoming, showing the V205 content and the thickness of the zone at sampled localities .............................................................................................................. In pocket 2. Detailed stratigraphic sections showing the distribution of vanadium within the vanadiferous zone at Sublette Ridge, Wyoming ............................................................................................. In pocket FIGURES 1—3. Maps showing: 1. Approximate areal extent of the vanadiferous zone in the upper part of the Meade Peak Member of the Phosphoria Formation ................. ._ ..................................................................................................................................... 2 2. Selected V205 analyses of the vanadiferous zone ........................................................................................................................... 3 3. Average concentrations of organic C, P, V, and Ag in the Meade Peak Member of the Phosphoria Formation ....................... 5 4. Graphs showing variation of nitrate, phosphate, Cd, Ni, and Zn in seawater with depth in the North Pacific Ocean at 32°41.0’ N., 144°59.5' W ...................................................................... M 23 III IV TABLE D—‘H 13. 14. 15. 16. 17. wewwe HOFWNS" CONTENTS TABLES Page . Partial chemical analyses of the vanadiferous zone at Sublette Ridge, Wyo. .......................................................................................... 6 Chemical analyses of the vanadiferous zone and adjacent limestones at Coal Canyon, Wyo ....... 7 Minor elements in two beds of the vanadiferous zone at Coal Canyon, Wyo ................................................................ 7 Analyses of selected constituents of the vanadiferous‘zone at Coal Canyon, Wyo. .................................................................................. 8 Selected electron microprobe analyses of semifusinite areas in sample VEM 42-47 from the vanadiferous zone of the Meade Peak Member of the Phosphoria Formation at Coal Canyon, Wyo. ................................................................ 9 Average thicknesses and V205 contents of the beds of the vanadiferous zone, based on all available samples 10 Partial composition of the vanadiferous zone and adjacent beds in the Paris-Bloomington area, Idaho ........................... .. 12 Partial chemical analyses of the vanadiferous beds in the Paris-Bloomington area, Idaho (US. Bureau of Mines) .......... 13 Partial chemical analyses of vanadiferous beds in the Paris—Bloomington area, Idaho (Anaconda Copper Mining Co.) .......... 13 . Analyses of samples from the vanadiferous zone at Bloomington Canyon, Idaho ................................................................................ 14 . Estimates of minor-element abundances in the vanadiferous zone at Bloomington Canyon, Idaho, in comparison With their average concentrations in continental crust ........................................................................................................................ 15 . Comparison of the thicknesses and V205 contents of partly leached, enriched, and unenriched rocks of the vanadiferous zone in the Paris-Bloomington area, Idaho .............................................................................................................................................. 15 Comparison of the thicknesses and vanadium contents of the vanadiferous beds on the normal and overturned limbs of the Paris syncline, Idaho .......................................................................................................................................................................... 16 Effect of lenses on the unaltered vanadiferous zone in the Consolidated mine, Paris-Bloomington area, Idaho ........................ 17 Range in thickness and grade of the vanadiferous zone in the Paris-Bloomington area, Idaho ........................................................ 17 Subeconomic vanadium resources in the Paris-Bloomington area, Idaho ..................................... 18 Some trace metals in seawater in comparison with their concentrations in black shales .................................................................. 22 THE VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION IN WESTERN WYOMING AND SOUTHEASTERN IDAHO By V.E. MCKELVEY, J.D. STROBELL, JR., and AL. SLAUGHTER ABSTRACT A black shale in the upper part of the Meade Peak Member of the Permian Phosphoria Formation contains an average of 0.8 to 0.9 percent V205 or more over a large area in western Wyoming and southeastern Idaho. In 1942 and 1943, this vanadiferous zone was sampled at numerous localities in the region by the U.S. Geological Survey, the Wyodak Coal and Manufacturing Co., and the U.S. Bureau of Mines. At Sublette Ridge, 1 mi east of the Idaho State line in Lincoln County, Wyoming, the vanadiferous zone lies about 50 ft below the top of the Meade Peak Member in the nearly vertical eastern limb of a north-striking anticline. The full thickness of the vanadiferous zone in the explored part of Sublette Ridge contains about 1.9 million tons of indicated subeconomic resources above drainage level. The richer upper part contains about 1.35 million tons. In the Paris-Bloomington area of Bear Lake County, Idaho, in the eastern foothills of the Bear River Range about 1 mi west of the towns of Paris and Bloomington, the Phosphoria Formation and associated rocks lie in the Paris syncline, which plunges about 15° N. The western limb of the syncline is overturned; the beds near the trough are nearly vertical, but, high on the limb, they dip as little as 20° W. The vanadiferous zone is about 35 ft below the top of the Meade Peak Member and about 5 ft below the upper phosphate zone. The zone consists of three beds: a shale, a phosphorite, and a siltstone, in ascending order. Estimates of measured, indicated, and inferred subeconomic resources of vanadium in the Paris-Bloomington area have been prepared for the shale bed, the shale and phosphorite beds combined, and the full thickness of the vanadiferous zone. Measured resources for the full zone are estimated to be 600,000 tons averaging 10.8 ft in thickness and 0.93 percent V20 5; indicated resources are 4,000,000 tons averaging 10 ft in thickness and 0.9 percent V205; inferred resources, within the area sampled, are 50 to 75 million tons averaging 10 ft in thickness and 0.6 to 1.0 percent V205. The vanadiferous zone appears to be the product of primary deposition from upwelling water in a reducing marine environment below wave base in a water depth of 100 m or so. These deposits formed on the outer continental shelf on the western side of the North American craton at low latitude, where deep, cold, nutrient- rich seawater came to the surface as a result of divergent upwelling in a tradewind belt. The vanadiferous zone in western Wyoming and southeastern Idaho contains about 41 million tons of indicated subeconomic resources averaging about 0.9 percent V205. Inferred resources are many times larger. Eventually, these inferred resources may prove to be an important source of vanadium and several other metals. INTRODUCTION A black shale containing 0.8 to 0.9 percent V205over a large area was discovered by W.W. Rubey of the U.S. Geological Survey (U SGS) in 1938 in the upper part of the Meade Peak Member of the Permian Phosphoria Formation in the Salt River Range of western Wyoming (Rubey, 1943, 1958). This vanadiferous zone, as it came to be called informally, was sampled by the USGS in 1942 at numerous localities in the adjoining region (McKelvey, 1946); it was sampled in more detail in late 1942 and 1943 by the Wyodak Coal and Manufacturing Co., agent for the Metals Reserve (30., in the Paris-Bloomington area of Idaho (McKelvey and Strobell, 1955) and in the Sublette Ridge area of Wyoming. The U.S. Bureau of Mines (USBM) also explored the zone in Sublette Ridge (Allsman and others, 1949a) as well as in the Salt River Range (Allsman and others, 1949b) and investigated means of recovering the vanadium (Ravitz and others, 1947). Information on the geology, composition, and magni- tude of the vanadiferous deposits in the Salt River Range was summarized by Love (1961). These published data are herein supplemented by additional information acquired in the 1942-1943 investigations of the Sublette Ridge and Paris-Bloom- ington deposits, along with some interpretations of the regional extent, composition, and origin of the zone based in part on subsequent studies (for example, McKelvey and others, 1953, 1959; Gulbrandsen, 1960; Desborough, 1977; E.C.T. Chao, J.A. Minkin, and J.M. Back, written communication, 1986). REGIONAL EXTENT The vanadiferous zone has been found within an area of about 4,500 mi2 in western Wyoming, south- eastern Idaho, and northeastern Utah (figs. 1, 2). In Wyoming and Utah, the zone is a readily identifiable 1 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO 111° / / \ / \ / \ / / \\ / \ \\ \ \ / Aftor: / \\ / \ 2 \ \ \ U5 C «(I ‘3 m / / (92 / I?) / / rn / / ll / i3 / é; / .L D \ / Kemmerer 0,9110“ 0.3 / / WYOMING 0 10 20 30 MILES % I I I I I I I 0 10 20 30 40 KlLOMETERS FIGURE 1.—Approximate areal extent of the vanadiferous zone in the upper part of the Meade Peak Member of the Phosphoria Formation, defined on the basis of the area within which a thickness of 3 ft or more contains 0.3 percent V205 or more (shown by short-dash line). Shown also is the area within which one bed or more contains 0.7 percent V205 or more (long-dash line). Outcrops of the Phosphoria Formation are shown by thin lines, and the location of samples for which analyses are shown on figure 2 are marked by open circles. The areas outlined are (1) Swift Creek-McDougal Pass, (2) Afton, (3) Labarge Creek, (4) Tunp Range, (5) Sublette Ridge, and (6) Paris-Bloomington. REGIONAL EXTENT 111° 6.8’-0.03 | o(2.2'—o.10) I 7.7'—o.12 , 2.2’—1.03 0(1.6’—0.24) 5-8,'°-71 O (2.3 -1.31) 4_7I_0_49 12.1I_o.33 11.5'—o.19 I 2.1'-1.17\O 0(2'21-0'91) 03.02035 (3.8’-0.39) o(2.2'-o.51) o/3.6’—0.71 o o . (1.4’—1.32) 7.5'—O.28 } o (2.0’-0.46) 4.7'—o.45 8.7’—0.48 (1.6’—0.98) O(my—0.59) O7.5'-—0.28 I (2.0’—0.46) 4.4’—O.65 (1.5’—1.46)o 3.0’—0.63 o 1.9'—o.7oO | I 02.8'-0.51 7.4'—o.39 (1.5’-0.76) (3.6’—0.61) O I I 3.4‘—0.80 10.8’—0.93 (’ ,_ (3.6’-1.35) I (2'2 "‘5’ o 3.4'-o.43, : (1.5'—o.7o) O\3.4'—o.74 IDAHO I (2.2'—1.15) 42°—-—— —— —— —— I I I 3.3'—0.58 l O(1.2 —o.90) I 3.9'—o.24 I (2.7'—o.27) I o I UTAH : WYOMING 0 10 20 30 MILES | | | I l l ' J 0 10 20 30 40 KILOMETERS and Paris-Bloomington areas). FIGURE 2.—Se1ected V205 analyses from the vanadiferous zone (see fig. 1 for p Only locations having unfaulted exposures, unaffected by d higher grade part of the zone are shown in parentheses. A based on the average of many samples collected by the USBM (Afton area) and the Wyodak Coal and Manufacturing eep weathering, osition of samples on outcrops of the Phosphoria Formation). are shown; where available, values of samples taken from the nalyses from locations indicated by solid circles rather than open circles are Go. (Sublette Ridge 4 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO stratigraphic unit 2 to 6 ft thick, consisting of non— phosphatic hard black mudstone that weathers to soft black shale containing conspicuous orange coatings along joint and bedding planes. It is 30 to 100 ft below the base of the Rex Chert Member of the Phosphoria Formation and essentially at the base of the upper phosphate zone of the Meade Peak Member. In Bear Lake County, Idaho, the zone is 6 to 15 ft thick and also lies at the base of the upper phosphate zone, 25 to 40 ft below the Rex Chert Member. Unlike the zone in Wyoming, it contains a few to several percent P205. The beds are soft and black in weathered exposure, but their phosphate content gives them an appearance different from that of their Wyoming correlatives. Farther north, in Caribou County, the P205 content increases to'20 to 25 percent, and the zone is not distinguishable lithologically from the upper phosphate zone. The presence of the zone is generally indicated by V205 analyses in the range of about 0.35 to 1.0 percent over a thickness of a few feet. The areal limits of the zone are known only approxi- mately. It is present in the Tunp Range to the east of Sublette Ridge, where it is about 3 ft thick and contains 0.75 to 0.9 percent V205. It is also present in the northern Crawford Mountains of Utah, an occur- rence lithologically quite similar to the one farther north but containing much less vanadium. Southward and eastward, the beds of the vanadiferous zone either tongue out or grade into other lithic units; they are not identifiable on the basis of either lithology or vanadium content in the Wasatch or Uinta Mountains. A similar change in facies occurs to the east and north of the Wyoming Range; the zone is not identifiable in the Wind River Mountains, the Gros Ventre Range, or the Snake River Range. The vanadiferous zone may extend westward for some distance, but its vanadium content west of Bear Lake Valley in southeastern Idaho is so low that the zone is not identifiable as such. A comparison of figures 1 and 3. shows that the vanadiferous zone extends over that part of the Meade Peak Member richest in organic C, P, V, and Ag. However, the zones highest in vanadium lie to the east and south of the carbon-, phosphorous-, and silver-rich zones. The Salt River Range near Afton, Wyo. Sublette Ridge, and the Paris-Bloomington area were selected for further exploration within the area underlain by the vanadiferous zone because of their potential minability, judged on the basis of the vanadium content of the zone and the relatively uncomplicated structural settings. Should exploitation of the van- adiferous zone become economic, however, they would not be the only possibly minable areas. SUBLETTE RIDGE DEPOSITS Sublette Ridge lies in the western part of Lincoln County, Wyoming, about 1 mi east of the Idaho State line. It is a northward—trending ridge of mountainous relief on the eastern side of Thomas Fork Valley. Raymond Canyon, the main point of entry to the deposit, is about 2 mi east of Raymond, Idaho. The Union Pacific Railroad at Border, Wyo., is about 7 mi by road from Raymond Canyon. During the winter of 1942-1943, the Wyodak Coal and Manufacturing Co. and 'the USBM undertook exploration programs. Fifty—four trenches spaced 200 to 1,000 ft apart were dug, and 13 tunnels and a shaft were driven. GEOLOGY Vanadium is present in small amounts in many of the beds of the Meade Peak Member of the Phosphoria Formation, but only the vanadiferous zone, lying about 50 ft below the top of the member, contains amounts of possible economic interest. The Meade Peak Member is underlain by the Grandeur Member of the Park City Formation and overlain by the Rex Chert Member of the Phosphoria Formation. These formations occur on the eastern limb of the Sublette anticline, the axis of which strikes north along the western front of Sublette Ridge. The western limb of the anticline has been downthrown by normal faulting, and most of it is buried beneath Holocene and Tertiary(?) alluvium in Thomas Fork Valley. A part of the western limb is exposed in the northern part of the area east of the fault for a short distance. The Phos- phoria Formation extends over a strike length of 6.5 mi in Sublette Ridge but is concealed by younger rocks for a distance of 1.5 mi north of Petereit Gulch. The beds are nearly vertical along the flank of the anticline but dip as low as 45° in places near the crest. Numerous faults marked by small displacements cut the vana— diferous zone. VANADIFEROUS ZONE The vanadiferous zone in the Sublette Ridge area consists of seven layers of black mudstone overlain by a hard, massive, fossiliferous limestone about 2 ft thick and underlain by an equally hard and massive but locally lenticular limestone. The beds are soft and fissile near the surface, but, about 40 ft beneath the surface, they abruptly become hard and massive; the layers there are separated from one another by barely discernible cracks. Orange coatings, probably the SUBLETTE RIDGE DEPOSITS 11‘2° 110° 1?8° 1 112° 1110“ ”iii \ MONTANA MONTANA [.________.___ L A [______________ | \r~/""—f\i—— \\/‘/~/v-/\| 44° — | 44° — i — IDAHO ' IDAHO | | | \ MONTANA K \ I x \ mm \/ WYOMING UTAH 40° — j 40° _ j i l C D FIGURE 3. —Average concentrations of (A)organic C (in percent) (B) P (in percent). (C )V (in parts per million), and (D) Ag (in parts per million) In the Meade Peake Member of the Phosphoria Formation (after Maughan, 1976, 1980). 6 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO result of pyrite oxidation, are prominent on bedding and joint surfaces of the weathered rocks. Although weathering produced the effects just described, it did not mobilize the vanadium in this area, for the vanadium content is essentially the same in both fresh andweathered parts of the zone. The full vanadiferous zone is 2.5 to 4.3 ft thick and averages about 3.4 ft. Most of the vanadium is con- centrated in the upper 1.5 to 2.5 ft of the zone. The V205 content of the whole zone ranges from 0.45 to 1.05 percent and averages about 0.8 percent. During the early stages of the work, it was assumed that the shaly nature of the rocks would require mining the full width of the zone between the hanging-wall and the footwall limestone beds; consequently, the zone was sampled as a unit in most of the trenches (pl. 1). After hard, unweathered rock was encountered in under- ground workings, attention was given to mining only the upper, more vanadiferous part of the zone, which averages about 1.05 percent V205 over a width of 2.25 ft. The individual layers of the zone are remarkably continuous stratigraphically, and their vanadium contents are similar throughout the area (pl. 2). In fact, essentially the same sequence is found some 40 mi to the north in the Afton area, where the same rich central part of the zone (pl. 1) is 0.5 to 1.0 ft thick and averages 1.6 percent V205 (maximum 2.5 percent) (Love, 1961), about the same as the values found in Sublette Ridge. Although these layers are recognizable in the field on the basis of physical properties, the distinctions between them are slight, and they were not recognized during the 1942 reconnaisance sampling when the zone was generally sampled as a unit. It seems likely that these same layers are also present at other localities in western Wyoming where the zone has been found. The USBM reported partial analyses of several samples, as table 1 shows (Ravitz and others, 1947). The locations and thicknesses represented by these samples were not reported, but, because the V205 content is in the range of 0.52 to 0.92 percent, it is assumed that they represent the full thickness of the zone. The analyses suggest that the main constituents of the rocks are quartz-silicate minerals, organic matter, carbonate minerals, and pyrite. Relatively complete chemical analyses were made by the Tennessee Valley Authority on all of the beds of the Meade Peak Member sampled in 1947 by a USGS field party in Coal Canyon,1 about 1 mi south of Raymond Canyon (McKelvey and others, 1953). Analyses for the beds of the vanadiferous zone and the hanging-wall and footwall limestones are shown in table 2. Gulbrandsen (1960) determined the vana- diferous beds to be quartz-silicate rocks containing grains of quartz, feldspar, and mica in a matrix of illite, organic matter, pyrite, and dolomite. Because of its odor, he believed part of the organic matter to be bituminous. Spectrographic analyses of minor ele- ments in two of the beds of the zone are listed in table 3. Desborough (1977) later collected samples from the vanadiferous zone in Coal Canyon intended to cor— respond to the same beds sampled in 1947 and had them analyzed for the constituents shown in table 4. On the basis of electron microprobe, optical, and chemical studies, he concluded (Desborough, 1977, p. 1Now shown as Rose Canyon on the USGS topographic map of the Geneva 75-min quadrangle. TABLE 1.—Pa)'tz'al chemical analyses (in wt prrrr)1t)oftlze I‘anadiferozmzoneatSubletteRidge, Wyo. (from Raritzand others, 1.94 1‘) [LOL loss on ignition: n.d., not determined] Lot Lot Lot Int [.01 Lot 1-2 1-3 1-4 1-9 1-10 6-5 0.52 0.77 0.53 0.92 0.74 2.6 .9 2.1 .8 1.4 43.8 46.5 42.2 49.0 44.0 4.0 4.2 3.4 3.9 3.6 12.7 12.2 11.6 12.7 10.9 9.9 6.5 10.4 5.2 7.8 4.1 4.1 3.4 4.0 4.0 6.5 5.8 6.2 4.6 4. 7 10.7 10.9 9.2 8.5 4.7 02 .02 .01 .02 .02 .7 .7 .7 n.d. n.d. n.d. n.d. n.d 19.5 20.3 SUBLETTE RIDGE DEPOSITS TABLE 2.—Chemical analyses of the vanadiferous zone and adjacent timestones (in wt percent) at Coat Canyon, Wyo. [U and Se determinations by the U.S. Geological Survey; all others by the Tennessee Valley Authority under the direction of J .H. Walthall (McKelvey and others, 1953; Gulbrandsen, 1960). AI, acid insoluble; LOI, loss on ignition; n.d., not determined; no entry, not analyzed] 1 Vanatliforous zone Bed Bed Bed Bed Bed 72 7:; 74 75 763 4.20 4.20 6.05 8.40 50.22 2.9 2.7 2.3 1.9 1.8 5.6 5 1 5.9 5.0 40.2 .5 .20 .34 2.73 .36 52.58 45.64 41.64 47.18 4.82 9.7 9 6 9.8 8 6 1 1 4.4 4.0 3.9 3.4 5 1.30 .99 .78 1.25 60 2.85 2.78 2.55 2.12 72 .36 .23 18 .52 03 .37 1.45 1 75 .52 06 8.4 10.6 12.5 8.6 44 .32 .52 .83 .51 .13 65.75 56.38 51.2 59.30 5.90 16.6 22.6 24.9 16.7 41.1 .0005—.004 .010 .015 .015 .015 n d 1Beds 71 through 75 are from base to top of vanadiferous zone. Footwall limestone. Hanging—wall limestone. 1) that “V is in, or associated with, organic material and minor amounts of V are in Ti—Fe oxysulfide and an unidentified Fe sulfide; Cr is in a 10A mica. Zn and Cd are in sphalerite; Se is in pyrite; S is in pyrite, TABLE 3.-—Mt'nor elements (in ppm) in two beds of the vanadiferous zone at Coal Canyon, Wyo. [Semiquantitative spectrographic analyses by Harry Bastron, cited by Gulbrandsen (1977). n.d., not detected] Bed Bed Element P-72 P—73 Ag ........................................................ 7 10 B .............................................. 70 100 Ba. 300 300 Be n.d. 1 Co ............... 7 7 Cr .................................. 500 1.000 Cu.... ............................ 70 100 Ga... 10 15 Mn . . 200 150 Mo .................. 150 300 Ni ..................................... 200 700 Pb. ..................... 20 20 Sc . . 15 15 Sr . ..... 150 150 V ............................... 5,000 7,000 Y ......................................... 50 30 Yb ............ 7 20 Zn ..... 700 15,000 Zr .................................................... 300 300 ‘sphalerite, and organic material; Ti is in TiO2 and TizFe(SO302; and M0 is in powellite.” Agreement among these analyses for minor elements, made by different methods on different samples, leaves much to be desired, but they nevertheless support an interesting observation: a sizable group of elements—Ag, Cd, Cr, Cu, Ga, Mo, Ni, and Zn— roughly follow V in their concentrations. As table 3 shows, for example, an increase in V by a factor of 1.4 from bed P-72 to P-73 is paralleled by a 1.43 increase for Ag and Cu, at 1.5 increase for Ga, a 2.0 increase for Cr and Mo, and a 3.5 increase for Ni. (The 21.4 factor of increase shown for zinc may be a gross error; the value of 700 ppm shown for bed P-72 probably should be 7,000 ppm.) In a further attempt to define the site of the vanadium, E.C.T. Chao (written communication, 1985) undertook a study of a doubly polished thin section prepared by Cheryl Edwards from a chip of sample VEM 42-47 from the zone in Coal Canyon. He reported that: The pyritiferous, carbonaceous black shale, VEM 42—47, is a very fine—grained shale with most grains less than 50 micrometers in size. It contains several larger (up to a few tenths of mm) fossil fragments. The dominant minerals are as follows: Quartz (many with less than 10 micrometer size pyrite inclusions), detrital calcite (both single crystal fragments and aggregates), illite (platy micaceous clay), 8 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO TABLE 4. ~A nalyses of selected constituents of the vanadiferous zone at Coal Canyon, Wyo. (from Desborough, (.977) [n.d., not determined] Bed Bed Bed Bed 72-CC 73-CC 74-CC 75~CC Total c,1 in percent ...................................... 9.24 8.13 8.16 14.08 Carbonate C, in percent .134 1.63 1.16 1.28 Organic C,3 in percent ......... 7.9 6.5 7.0 12.8 Total 5,4 in percent ....................... 3.38 3.03 3.38 4.34 V, in percent ........................ .....09 .14 .19 .95 Zn, in percent ._.05 .09 .09 .69 Cd, in percent .n.d. .007 .01 .07 Ag in ppm ...................... 6. 5. 6. 10. Ti,7 in ppm .............................................. 3,000. 2.000. 2,000. 2.000. M0, in ppm 50. 100. 70. 300. 1Combustion, gasometric C02. H P04 digestion. 3D] ference between 1 and 2. 4Combustion. Colorimetric. Atomic absorption. Semiquantitative six-step spectrographic. X-ray fluorescence. 6 7 sphalerite (both single crystal fragments and in fine- grained aggregates), and abundant pyrite both granular and framboidal. In addition, this black shale contains very fine grained (10/1m)0r less in grain size) lithic fragments of clay lenses consisting of quartz and illite in a brownish dark matrix. The black shale also contains fairly abundant thin lenses and fragments of semifusinite (tentatively identified by its optical characteristics), a maceral of the inertinite group of coal. It also contains several opaque minerals, un- identified minerals such as a K-Al-silicate of low bire- fringence, rutile, zircons, plagioclase feldspar, and other accessory minerals. Authigenic secondary overgrowths on quartz have been observed but they are extremely rare. Based on two qualitative analyses, the V is contained in fragments of semifusinite, One grain without sphalerite, the other with inclusions of sphalerite. This suggests that semifusinite (an aromatic hydrocarbon of terrestrial plant origin) is the host of V in this rock. This would support George Desborough’s earlier finding... Phosphorus is present in the fossil fragments... This black shale contains an unusual amount of sphalerite with traces of cadmium, in isolated grains and in fine aggregates, apparently of sedimentary origin. The presence of framboidal pyrite, detrital calcite, phosphatic fossil fragments, and coaly fragments (and its V content) provide clues to the deposi- tional enviroment...(anaerobic enviroment with dominant marine influence). On the basis of its nonfluorescence in ultraviolet illumination, most of the organic matter is judged to be semifusinite. Because of its fine-grained mineral impurities, however, it is not uniform in composition. The electron microprobe analyses shown in table 5 were made by J.A. Minkin and J.M. Back (written communication, 1986) and selected from 25 such analyses of randomly selected semifusinite areas in the doubly polished thin section of VEM 42—47.,The electron beam focuses on an area a few micrometers in width, so that the analysis reflects the composition of any impurities present. Because the electron micro- probe does not detect light elements, such as carbon and hydrogen, the difference between the total shown and 100 percent can be assumed to be largely organic matter. Chao grouped these analyses on the basis of their composition, as table 5 shows. For example, analyses having a low total of inorganic constituents are assumed to be nearly pure semifusinite; potassium is assigned to illite; iron is assigned to pyrite, and excess sulfur is assumed to be organic; phosphorus is assumed to be in apatite. These analyses show that the vanadium in the semifusinite is likely to be in more than one form. Some of it clearly is in organic matter, since the nearly pure semifusinite contains appreciable amounts of vanadium. Because the highest amounts—up to 4.2 percent V203 (5.1 percent V205)—are in the areas composed mainly of illite, however, some vanadium may be attached to that mineral. Some may also be in apatite, although the areas containing apatite also contain the lowest amounts of vanadium. The presence of appreciable amounts of vanadium in association with the unidentified minerals of groups 6 and 7 in table 5 suggest that it may be in or attached to one or more other minerals also. SUBLETTE RIDGE DEPOSITS 9 TABLE 5.——Selected electron microprobe analyses (in percent) of semifusinite areas in sample VEM 42-47from the vanadiferous zone of the Meade Peak Member of the Phosphoria F ormation at Coal Canyon, Wyo. [J.A. Minkin and J.M. Back, analysts; characterization of groups by E.C.T. Chao (written communication, 1986). n.d., no data; --, not present] Group Group 22 Group Group Group Group Group 11 A B C 33 44 55 66 77 38.54 39.4 38.5 27.6 20.46 0.11 64.0 24.8 16.04 16.6 15.7 12.0 11.61 -- 10.8 9.6 .16 .4 .66 .96 .45 14.03 .20 11.6 .17 .11 .14 .11 .01 .40 .11 .212 4.29 3.5 3.9 3.3 1.01 .16 2.3 2.8 .82 1.2 6.5 19.7 3.32 .08 .42 .89 1.81 2.0 3.6 3.9 .56 .27 .7 8.6 .3 .20 1.75 .77 .07 .05 3.7 .27 .02 .03 .10 .12 .05 .04 .04 .08 2.98 4.2 3.4 3.9 1.06 .75 1.56 2.54 3.58 4.1 5.4 15.0 10.59 5.91 1.53 1.30 .09 n.d. n.d. n.d. .09 10.17 n.d. n.d. 68.8 71.74 79.65 87.36 49.28 31.97 85.36 62.7 1Nearly pure semifusinite. Six samples in group show V203 range of 1.23 to 2.22 percent. 2Semifusinite having appreciable illitic clay, quartz, and low to moderate pyrite content. Four samples in group show V203 range of 1.6 to 4.2 percent. 3Semifusinite having moderate illitic clay and high pyrite content. Three samples in group show V203 range of 1.96 to 3.9 percent. 4Semifusinite having high organic sulfur, moderate ash, and moderate pyrite. Four samples in group show V203 range of 1.46 to 2.51 percent. 5Semifusinite containing apatite. Four samples in group show V203 range of 0.65 to 1.08 percent. As yet unidentified mineral containing Si, Al, Ti, K, and Mg. One sample. 7As yet unidentified mineral(s) characterized by moderately high Si, high Ca and Mg, and moderate Al. Two samples show V203 range of 1.05 to 2.54 percent. RESOURCES The Wyodak Coal and Manufacturing Co. estimated that the vanadiferous zone in the explored part of Sublette Ridge contains about 1.9 million tons above drainage level averaging about 0.8 percent V205. Experimental stoping indicated that the upper 2.25 ft of the zone, averaging 1.06 percent V205, could be mined separately by the shrinkage method of mining. This uppermost part of the zone contains about 1.35 million tons. The feasibility of mining these rocks commercially has not been demonstrated. Hence, these quantities can be classed as indicated subeconomic resources. PARIS-BLOOMINGTON DEPOSITS The Paris-Bloomington area of Idaho is in Bear Lake County, in the eastern foothills of the Bear River Range. It is about 1 mi west of the towns of Paris and Bloomington and 10 mi southwest of the Union Pacific Railroad at Montpelier. The hills are low and rounded, local relief being about 500 ft. They are cut by four eastward-flowing streams—from north to south, Hammond Creek, Sleight Creek, Paris Creek, and Bloomington Creek—that head in the mountains to the west. From March to December 1943, the Wyodak Coal and Manufacturing Co., agent for the Metals Reserve Co., sampled the Paris-Bloomington deposits in nearly 2,500 ft of underground workings and in six trenches and short adits. For its part in this project, the USGS prepared a geologic strip map of the area adjacent to the trace of the vanadiferous zone and a geologic map of the area underlain by the zone, both later published at scales of 114800 and 1:12,000, respectively (McKelvey and Strobell, 1955). The average thicknesses and vanadium contents of the beds of the zone in surface and underground workings are shown on these maps, as are the locations of geologic and geographic features mentioned in this report. GEOLOGY Rocks exposed in the area include the Phosphoria Formation (Permian), the overlying Dinwoody, Woodside, and Thaynes Formations (Triassic), and the underlying Grandeur Tongue of the Park City Forma— tion (Permian) and the Wells Formation (Pennsyl- vanian and Permian). Brigham Quartzite and Ute(?) Limestone (Cambrian) are in fault contact with these rocks along the Paris overthrust (Armstrong and Cressman, 1963). Tertiary conglomerate, sandstone, and gravel overlie and conceal these Mesozoic and Paleozoic rocks, especially in the southern part of the area. Alluvium, slope wash, and talus of Quaternary age mantle much of the area. The Phosphoria Formation is divided into the Meade Peak Member, about 200 ft thick, and the overlying Rex Chert Member and cherty shale members, about 250 ft in combined thickness. The top of the vana— diferous zone is about 35 ft below the Rex Chert 10 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO Member and about 5 ft below the upper phosphate zone of the Meade Peak. The Phosphoria Formation and associated forma- tions lie in the asymmetrical Paris syncline, whose axial plane dips about 35° W.; the axis plunges about 15° N. south of Paris Canyon but flattens northward to about 5° or less. The western limb of the syncline is overturned; its beds are nearly vertical near the trough of the syncline, but, high on the limb, they dip as little as 20° W. The beds on the eastern normal limb dip 10° to 35° N.-NNW. The beds on the western limb are tightly squeezed and decrease in thickness with increasing distance above the trough. At vertical distances of 1,000 to 2,500 ft above the trough, large parts of the incompetent formations are cut out by reverse faults, which are nearly parallel to the bedding. Because of the north- ward plunge of the syncline, these bedding faults increase in abundance at the surface and progressively appear in stratigraphically higher beds. Thus, a bedding fault cuts out most of the Meade Peak Member at the surface north of Paris Canyon, but no such faults of any size are known in the member south of it. Four major transverse faults displace the Meade Peak Member as much as 1,000 ft. Minor transverse faults, many of less than 5 ft displacement, and drag folds that cause local thickening or thinning of the beds are abundant on the overturned limb of the syncline. The beds on the normal limb are little disturbed but in places are warped or broken by faults of a few feet displacement. The Paris thrust plate once overlay the entire area but has been eroded back to the west. South of Sleight Canyon, its margin now lies 500 to 2,000 ft west of the Meade Peak Member on the overturned limb. North of Sleight Canyon, however, the Paris thrust plate overlies the entire Phosphoria Formation as well as the Din— woody and Woodside Formations. The Paris fault plane dips westward 45° in Bloomington Canyon and decreases northward to 5° at Hammond Creek. VANADIFEROUS ZONE The vanadiferous zone is composed of three beds: in ascending order, a shale, a phosphorite, and a siltstone (table 6). These beds are stratigraphically continuous throughout the area. Small amounts of vanadium are found in other beds of the Meade Peak Member, mainly in and near the lower and upper phosphate beds and in the beds adjacent to the vanadiferous zone. Although these adjacent beds are locally enriched as a result of weathering, other beds in the Meade Peak Member for the most part contain less than 0.2 percent V205. TABLE 6.—Average thicknesses and V205 contents of the beds of the vanadiferous zone, based on all available samples [All vanadium analyses for the Paris-Bloomington deposits made by W.E. Ryan of the Wyodak Coal and Manufacturing Co.] Average thickness, Average V205 content, Bed in ft in percent Siltstone ...................................... 4.30 0.87 Phosphorite .......... 2.20 .54 Shale ...................... 3.4 1.38 Total .................................... 9.9 .97 The shale bed (the lower part of the zone) is black, medium hard, and fissile. The upper 1 ft contains some thin oolitic phosphorite layers, and the lower 6 to 8 in are in places more coarsely bedded and blocky. In the lower part of a winze in Paris Canyon, about 200 ft below adit level, the whole bed is calcareous, hard, and less fissile. Where the rock is moderately weathered, it is soft and clayey, bedding is indistinct, and its color is dark brown to tan. Lenses of black massive siltstone are found locally in the shale bed, particularly in the upper part. They range from 0.3 to 3 ft thick and from 1 to 25 ft long. Most of these lenses are soft and earthy, but, in places, some are hard and calcareous, and the change from hard to soft may be abrupt. The shale bed in the Paris—Bloomington workings, including the lenses, averages 3.4 ft in thickness and 1.38 percent V205. It also contains about 12 percent P205. Most of the lenses contain less than 0.2 percent V205. The phosphorite bed is black, massive, hard, and coarsely oolitic and has some shaly siltstone inter- bedded in the upper half. Where the bed is weathered, it is soft and dark brown to tan, crumbly in the lower part and clayey in the upper part. A few oolite lenses occur between the massive and shaly parts of the bed; they range from 0.5 to 1 ft thick and are generally less than 10 ft long. The average thickness of the phos- phorite is 2.20 ft, and its average V205 content is 0.54 percent; the P205 content is about 23 percent. Lenses of siltstone 0.5 to 3.0 ft thick and as much as 80 ft long locally occur at the contact of the phosphorite and the siltstone beds. The V205 content is generally low but may be high locally as the result of secondary enrichment. The siltstone bed is black, medium hard, blocky, and well bedded. Individual layers are 2 to 8 in thick. A soft, shaly layer as much as 1 ft thick is locally developed at the top of the bed. In the lower part of the Paris Canyon winze, the bed is massive, hard, and somewhat calcareous. Where it is weathered, it is soft, clayey, and dark to light brown. Its average thickness is 4.3 ft, and its average V205 content is 0.87 percent; it contains about 3 percent P205. PARIS-BLOOMINGTON DEPOSITS 11 The beds adjacent to the vanadiferous zone are mainly siltstone. Unweathered beds stratigraphically below the zone are hard and calcareous and form a good hanging wall for mining where the zone is on the overturned limb of the syncline. Where these beds are weathered, however, they are soft and require much timbering to hold as a hanging wall. The siltstone stratigraphically above the zone is hard, massive, oolitic, and yellowish brown and contains about 16 percent P205. On the normal limb of the syncline, this bed forms the hanging wall and, in most of the workings, will hold for only a day or two without support. Several partial chemical analyses of the vanadiferous beds are shown in tables 7 through 10. The major constituents of the zone are, in order of abundance, silica, organic matter, calcium, phosphate, and alumina. Quartz-silicate minerals and organic matter are the dominant components of the shale and siltstone beds, although carbonate minerals make up 15 to nearly 30 percent of the shale bed in the lower part of the Paris Canyon winze. Organic matter seems to make up about a quarter of the rock, but little is known as to its nature. Some of the rock has an odor of petroleum, and part of the organic matter may be hydrocarbon. Whatever its form. it is readily com- bustible; the stockpiles at the 14 South incline caught fire spontaneously in January 1944 and burned vigorously. Most of the phosphorite bed is composed of carbonate fluorapatite, approximately Ca9.7Mg.1Na.2(PO4)5.6 (C03)_1(CO3F)I3F2, according to Gulbrandsen (Manheim and Gulbrandsen, 1979). This mineral is probably also the site of the lesser amounts of phosphate in the shale and siltstone beds. Minor constituents of the zone include Fe, S, K, F, Na, Ti, and Mg, in approximate order of abundance. Much of the iron is in hydrous oxides in the weathered rocks and in sulfides in the fresh rocks. Some of the sulfur may be in organic matter. Potassium and sodium, as well as some of the iron and magnesium, are probably present in silicate minerals. Most of the magnesium: is probably combined with calcium in dolomite. The fluorine is in the carbonate fluorapatite. Beds P—142 and P—143 (table 10) are from the upper part of the shale beds, and P-148 is from the phos- phorite; the other samples are from the siltstone in the upper part of the vanadiferous zone. Several things in these analyses are worth noting. One is the large amount of organic matter in all the beds. Another is that 49 elements (counting organic matter as carbon) have been detected in the zone. Many of the elements are present in relatively uniform amounts—mostly a few or a few tens of parts per million—and show no marked relation to vanadium in their variation (for example, As, Be, Co, Dy, Er, Eu, Ga, Gd, Hg, Ho, Mn, Pb, Pr, Re, Sc, Sm, Ti, Y, Yb, and Zr), but several other elements do seem to follow vanadium in concentration variations (for example, Ag, Cd, Cr, Mo, Ni, Se, Sn(?), T1, and Zn). The distribution of the highest values does not correspond exactly, however. For example, al- though beds P—142 and P-143 contain the most V and the most Ag, Cu, and Sn as well, bed P-151 contains the most Cd, Mo, Ni, and T1 and by far the most Zn; bed P-154 contains the most Se by far; bed P—156 contains as much Cu as bed P-142 and more Cr than all other beds. Table 11 shows estimates of the average and maximum contents of minor elements concentrated in the vanadiferous zone in amounts that are more than 10 times their averages in continental crust. The average concentration factors of Cd, Mo, Se, Ag, Tl, V, and Zn are greater than 30; Cd, Sc, and Ag show exceptionally high concentration factors of 2,350, 11,000, and 230, respectively. The maximum con- centration factors of Cd and (Se are more than twice these average values, and maximum values for Mo, Tl, V, and Zn are hundreds of times larger than their averages in continental crust. Earth Sciences, Inc., which did some experimental mining in Bloomington Canyon in the 1970’s, reported that the vanadiferous shale bed there averaged 4.1 ft thick and contained 1.0 percent V205, 13.7 percent P205, 0.0115 percent U308, 0.0225 percent Se, 0.05 percent M0, 3.6 percent S, 20.1 percent C, 0.6 percent H, 0.6 percent N, 18.8 percent CaO, 13.8 percent SiOZ, 4.3 percent A1203, 2.0 percent combined H2O, 8.0 percent free H20, 13.3 percent undetermined, and 37.7 percent loss on ignition (De Voto and Stevens, 1979). The uranium content, as well as the phosphate content, is much higher than it is in the vanadiferous zone in western Wyoming. EFFECTS OF WEATHERING ON THE VANADIFEROUS ZONE The effects of Holocene weathering on rocks of the vanadiferous zone have been relatively minor. They involve partial oxidation of organic matter and pyrite, some leaching of carbonate and a consequent enrich- ment of phosphate, a change in color from black to dark brown, formation of orange coatings on joints, a change (in hardness) from hard to soft, and a change from massive to thin bedded or fissile. As we mentioned for the Sublette Ridge deposits, such weathering does not appear to have mobilized the vanadium. Weather— ing in phosphatic beds has caused an increase in phosphate that is proportional to a decrease in car- bonate and organic matter; because much of the 12 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO TABLE 7.—Partial composition (in wt percent) of the vanadiferons zone and adjacent beds in the Paris- -Bloomington area Idaho [Analyses made under the direction of N. Herz Homestake Minin gr. specific gravity; N, none; tr, trace; 1; Co. AI, acid insoluble; LOI. loss on ignition; Sp. -— not present; ? presence uncertain; no entry. not analyzed] g." “m 1111111111111111111. 14 9111111. 111 91111111 259 144 202 l'tmmlnlnled i : i e incline. {1, 1‘1 l'l m1n1-. 1111. 40 ft 71111 90 [1 Portal in down down 1 11-011 1n in in Avenzel Phonehnllc sill-tone V205 0.344 0.14 11.494 11.17 1123 N1 .014 C110 2311“ 30.0 15.44 31.5 25.2 P 05 15.1‘1 10.11 13.114 26.7 16.4 Af 20.94 24.11 51.71 16.0 211.1 L01 20.5 24.6 10.114 11.11 16.7 “A patitse"2 39.34 28.1 33.11“ 68.9 42.5 C|C033 6 9‘ 211.11 4 Sp. gr. 2. 704 2.71 2.524 2.1111 2.611 Siluwne v205 0.116 11. 65 11.5114 0.11:1 11.611 0.9:1 0.711 N1 N‘ tr N4 11‘ cm 5.6 9.0 6.74 :17 5.9 Pfo5 3.4 3.3 .54 :1.:1 2.9 A 65.4 511.1 5611‘ 1111.1 59.4 597 L01 19.9 21.1 20.14 24.9 .6 211.11 22.9 “Apatite” 11.11 8.6 1.34 11.6 111.4 11.11 7.7 C41C03 2.0 11.2 10.744 -- 25 -. SP 3'12-3' 2.2.1 3- 1" 2.111 2.22 2.11 2.19 Phosphor“: v205 0.45 0.51 11.311 11.42 11.49 11.217 11.55 11.511 11.46 N1 . N tr tr C10 382 311.11 32.6 211.7 25.4 211.4 27.11 29.7 1705 29.1 25.11 2116 2:14 24.4 22.1 23.11 23.5 A 14.2 20.0 22.4 21.6 17.11 19.6 20.6 L01 5.7 10.3 13.6 111.3 14.4 111.6 12.6 Apltite" 75.7 67.011 5:16 60.11 6:14 59.11 61.1 0.603 .. 1 :1 -- -- -- -- Sp. .1». 2.94 2.72 2.50 2.66 2.611 2.62 2.77 2.64 Shale v205 1.68 1.06 0.96 1.32 11.117 1.2111 1.37 121 N1 .017 5144 .112 11 .112 C10 15.1 111.5 26.6 16.11 22.11 111.2 22.1 .2 P 05 11.9 12.3 7.7 11.11 17.2 14.5 16.5 12.6 Al 40.2 24.6 17.0 - 1' 21.3 20.3 17.2 21111 L01 17.4 35.3 214.9 27.7 37.2 24.6 32.5 “Apuiu” 30.9 32.11 20.0 44. 7 37.7 42.9 :12 9 C11003 -- 4.1 29.3 .. .. 11 Sp. gr. 2.37 2.13 2.117 2.24 1.91 2.211 2.14 Silulone (cult-luau: Ind phmphllir) V205 0.15 0.15 0.211 0.60 1.67 0. 51 C10 40.0 311.11 29.4 19.1 111.0 26. 4 Pfo5 2.2 4.2 3.11 . 14.9 13.7 116 A . 7.6 6.1 6.4 9.11 111.11 22 .5 111. 5 L01 . 41.3 42.11 4:111 411.11 47.2 :17. :1 414 “Apntiue” 5.6 11.0 . .. 213.0 311.7 35.6 22.3 Cnco3 66.4 45.11 4:1.5 7.7 -- Sp. gr. 2.50 2.211 2.311 1.97 1.91 2.25 2.20 Sn" lens in nhlle bed v 1.20 1011 C1285 17.6 11.7 111.7 s 4 Al05 25.4 19. :1 L0] 34.3 51.15 “Apntile” 27.11 . 21.9 C11603 6.6 .. Sp. gr 2.12 1.34 Hard lens in shale bed v205 0.29 0.29 11.29 Clo 26.6 25.11 26:1 P 05 4.2 2.1 3.4 A 5.6 3.5 4.5 LOl 44.0 44.11 44.4 “Apntite” 10.11 7.1 11.9 CIC03 311.0 :1911 311.11 Sp. gr. 2.41 2.42 2. 4] Son lenx between phosphor“: and uilumne beds v 1199 c.235 24.8 p 7.6 “0.5 111.1 L01 _ _ 32.3 ‘Apntite 19'“ cnco3 26' 4 511 g,_ 2.3 l.—\\'era1ze of analyses shown in this table. 2“Apatite” here is assumed to be llJCaO'3P. ()5'C01'Cal’. calculated by converting all the P2()— to “apa i 1." at ding e (‘a() necessary to satisfy the P205 {and F. and cowerting the remaining ( a() to Ca( Insufficient (‘ a() present to satisfx the proportions called for 1n the “apatite" formula. Sample from only part of the bed. The amounts of “apatite" and (‘aC03 shown are PARIS—BLOOMINGTON DEPOSITS 13 TABLE 8.—-Partial chemical analyses (in wt percent) of the vanadiferous beds in the Paris-Bloomington area, Idaho [Analyses by the Us. Bureau of Mines (Ravitz and others, 1947). LOI, loss on ignition; n.d., not determined] Lot Lot Lot Lot Lot Paris 6—61 6-101 6-112 6-123 6—134 composite5 v205 _______________________________________ 0.91 1.17 0.86 0.49 1.60 0.69 P205 ............. 11.3 10.3 3.0 22.8 12.8 13.1 Si02 ........... 23.8 30.2 45.4 16.8 15.0 23.4 F6203. ........ 2.2 3.0 1.7 2.3 1.4 1.8 A1203 ....... .._.10.1 11.1 0.2 4.3 4.7 6.8 CaO ......................... 13.8 12.6 3.6 30.2 17.4 17.4 S ............................. 3.2 2.0 2.1 1.2 4.6 2.8 002 ................ 3.2 .6 .4 .7 .6 .6 Organic C _. _...17.1 12.6 14.4 8.5 25.7 16.0 M003 .................. .02 .02 .02 <.01 .07 .03 TiO2____ ........ .3 .3 .6 .2 .2 n.d. LOI .......................................... 27.7 24.8 27.7 16.2 41.2 30.0 1Full thickness, Paris Canyon. 2Unweathered siltstone bed, Paris Canyon winze. 3Unweathered phosphorite bed, Paris Canyon winze. 4Unweathered shale bed, Paris Canyon winze. 5Composite of lots 6-11, 6-12, and 6-13. TABLE 9.—Partial chemical analyses (in wt percent) of vanadtferous beds in the Pam's-Bloomington area, Idaho [Analyses by the Anaconda Copper Mining Co. No entry, not analyzed] Bed Bed Bed Bed 11 22 33 44 Si O ........................ 30. 2 11.0 16.6 22.0 1.10 1.14 1.40 1Channel sample from full zone exposed in No.2 Crosscut of Paris Canyon adit. Calcined sample from bed 1. 3Channel sample of shale and phosphorite beds 73 ft from portal in 16 South incline. 4Calcined sample from bed 3. vanadium in such rocks is held in the apatite lattice, it locally shows slight increases (Lotspeich and Mar- quard, 1963). A period of deep chemical weathering preceded the deposition of the Tertiary Wasatch Formation in the Paris-Bloomington area and elsewhere in the region. Near the erosion surface on which the Wasatch sedi- ments were deposited, the rocks of the Meade Peak are light gray, tan, and. pink, in contrast to their dark brown or black color elsewhere, and are similar in appearance to the burned rocks of the 14 South stockpiles. Weathering has removed calcium car- bonate, organic matter, and vanadium, as well as part of the calcium from the apatite. This zone of leaching is exposed only in the trenches above the Consolidated mine in the Paris-Bloomington area and may extend to a depth of 150 to 175 ft below the surface. Enrichment of vanadium is found at or near the base of the zone of weathering. The rocks there are soft to medium hard and dark brown to black, and bedding planes are well defined. Visible occurrences of red hewettite (CaO - 3V205 - 9H20), green sincosite (CaO-V204-P205-5H20), and orange pascoite (2Ca0~3V205-11H2O) are common. Locally, the hewet- tite forms tabular masses as much as a few inches thick and 2 ft in diameter. The enrichment in some places is considerable. The enriched part of the zone in the Consolidated mine contains 1.75 percent V205, in comparison with 0.79 percent V205 for unenriched rock (table 12). The enriched part of the shale bed in the 16 South incline contains 1.66 percent V205 in comparison with 1.38 percent V205 in the shale below the enriched zone. Elsewhere, the obvious secondary concentrations of vanadium are spotty, and the increase in average vanadium content is slight. Large masses of hewettite are present in the 14 South stopes, for example, but they are so irregularly distributed that the average V205 content of the “enriched” rock is only 0.06 percent more than that of the unenriched rock. The limits of the obviously enriched zone are ir- regular, but, in the 14 and 16 South inclines, where it is 14 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO TABLE 10. —Analyses of samples from the vanadiferous zone at Bloomington Canyon, Idaho (from Gullrrandsen, 1975; written communication, 1985) [Organic matter analyzed by Don Krier; Ti. Hg, and N (total) by E. Campbell; Se by J .S. Wahlberg, J .0. Johnson, and R.J. Young; As by E.J. Fennelly; Re for beds P-l43, P-150, P-151, P-154, P-155, and P-156 by Rowe and Steinnes (1976); emission spectrographic analyses for all other elements by J .D. Fletcher. See McKelvey and others (1953) for description of beds. n.d.. not detected] Bed Bed Bed Bed Bed Bed Bed Bed Bed Bed Element P-142 P-l43 P-l46 P-148 P-150 P-151 P—153 P-154 P-155 P-156 Amount. in wt percent 12 12 12 28 35 38 >34 33 30 1.8 2.6 1.8 3.3 3.8 4.1 4.4 3.9 3.6 .61 2.4 .68 2.4 2.5 1.5 1.4 2.1 1.3 .24 .21 .14 .24 .25 .37 .35 .35 .51 7.9 19.5 5.6 4.3 1.1 1.0 1.2 3.2 3.5 >.3 >.32 .09 >.32 >.32 .07 .08 .09 .08 .55 .60 .39 1.2 1.4 1.7 1.4 1.4 1.4 ‘ .11 .13 .08 .22 .26 .28 .28 .30 .24 P 7 0 >68 >68 6.2 4.0 1.4 1.2 1.6 2.9 2.1 Organic matter.. n d 35.8 n.d. n.d. 21.0 22.9 n.d. 24.2 24.4 18.1 N (total) ............ 2.1 ,41 1.0 1.0 1.3 1.1 1.2 1.5 .70 Amount. in parts per million 22 14 16 14 12 14 14 16 16 20 40 25 50 50 20 20 25 12 <10 <10 <10 <10 <10 <10 <10 <10 <10 140 68 25 130 140 190 190 200 270 240 180 340 300 340 380 360 2.9 3.3 2.4 3.7 2.3 2 2.7 3.3 2.9 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 760 420 380 970 980 170 180 170 180 <63 <63 99 130 110 73 69 <63 130 2.3 2.2 1 0 9.5 10 2.6 3.5 2.1 1.4 1.500 1.400 860 850 940 1.200 980 1.200 1,600 420 99 120 180 110 180 250 170 260 14 19 9.8 24 17 <6.8 8.7 14 7.8 9.3 10 7.9 27 11 <4.6 6.6 8.7 <4.6 2.5 1.8 1.9 3.3 3.5 1.9 1.6 2.5 2.4 9.8 5.6 3.9 7.2 8.9 9.2 9.1 8.2 7.9 23 <15 <4.6 18 26 19 23 16 <15 <3.2 <3.2 <3.2 <3.2 <3.2 >3.2 <3.2 <3.2 <3.2 <22 <22 <22 <22 <22 <22 <22 <22 <22 .94 .60 1.3 .80 .78 .86 .80 1.6 1.1 3.7 <3.2 3.4 6.7 <3.2 <3.2 <3.2 3.8 <3.2 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 <4.6 170 140 130 150 100 58 68 96 90 <3.2 <3.2 <3.2 <3.2 <3.2 <3.2 <3.2 <3.2 <3.2 66 120 34 140 270 79 98 91 83 44 33 6 1 140 280 30 22 23 4.6 <10 <10 <10 <10 <10 <10 <10 <10 <10 120 <68 <68 <68 <68 <68 <68 <68 <68 410 73 120 340 540 180 290 150 160 14 11 <68 10 14 19 11 14 13 <32 22 25 33 26 <32 <32 <32 25 3.6 <10 <10 3.9 4.4 <10 4.9 5.8 .2 14 12 10 20 17 17 17 19 16 820 21 430 440 590 480 1.200 940 420 6.7 <4.6 6.3 7 4 7.6 6.8 6.6 6.8 7.4 100 25 32 17 19 32 26 28 78 2.200 1.500 780 460 370 440 470 550 1.900 <10 <10 <10 <10 <10 <10 <10 <10 <10 15 10 7.1 74 110 26 36 36 11 <3.2 <3.2 <3.2 2.9 <3.2 <3.2 <3.2 <3.2 3.3 <150 <150 <150 <150 <150 <150 <150 <150 <15O >4.000 1.700 2.400 2.900 2.900 4.100 4.000 3,000 2.400 <10 <10 <10 <10 <10 <10 <10 <10 <10 140 180 220 370 140 80 95 150 130 7.9 16 7. 14 7.7 4.3 5.6 7.2 21 1,400 620 580 4.700 10,000 490 1.600 380 360 190 410 150 310 280 250 240 430 120 PARIS-BLOOMINGTON DEPOSITS 15 TABLE 11.—Estimates of minor-element abundances in the vanadiferous zone at Bloomington Canyon, Idaho, in comparison with their average concentrations in continental crust (after Gnlbrandsen, 1 97 7) Vanadiferous zone Continental crust No. of Average. Maximum. Average, Concentration Element samples in ppm in ppm factor1 30 50 1.8 17 130 200 10 13 2170 980 .2 2.350 1,200 1,600 100 12 1 1.6 .08 12.5 60 280 1.5 40 560 1,200 .05‘ 11,000 16 20 .07 230 40 110 .45 89 40 100 2 20 4,600 17,000 135 34 2,400 10,000 70 34 1Ratio of vanadiferous zone average to continental crust average. 2Spectrophotometric analysis by E.J. Fennelly. Semiquantitative spectrographic analysis by J. Fletcher. 4Analysis by E. Campbell. 5X-ray fluorescence analysis by J .S. Wahlberg, J .D. Johnson, and R.J. Young. 6Analysis by J. Fletcher and H.H. Lipp. TABLE 12. ——Comparison of the thicknesses and V20 5 contents of partly leached, enriched, and unenriched rocks of the vanadiferous zone in the Paris-Bloomington area, Idaho [No entry, not analyzed] Partly leached rock Enriched rock Unenriched rock Average Thickness. V205, Thickness, V205, Thickness, V205, Thickness, V205, in ft in percent in ft in percent in ft in percent in ft in percent Consolidated mine Siltstone ................................................................ 3.51 1.501 4.74 0.674 4.4 0.94 Phosphorite .................................... 1.41 .741 2.84 .494 2.1 .57 Shale ......................................... 2.91 1.8912 3.44 1.224 3.3 1.45 Total ................................................................ 7 .8 1.753 10.9 .793 9.8 1.033 16 South incline Shale5 .................... 3.36 1.266 5.27 1.667 4.18 1.388 4.1 1.40 14 South incline Shale ..................... 3.4 1.11 3.9 1.37 3.8 1.31 3.6 1.29 1Sampled along 180 ft of drift. 2Includes siltstone bed below shale where enriched 45 to 140 ft north of No. 1 crosscut. 3Weighted average. 4Sampled along 283 ft of drift. 5Includes siltstone below shale 85 to 175 ft from portal. 6Sampled along 85 ft of incline. 7Sampled along 45 ft of incline. 8Sampled along 150 ft of incline. 16 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO defined by the presence of hewettite, the zone extends downdip for 40 to 50 ft. In the northern drift of the Consolidated mine, where the rocks are faulted and broken, ennrichment extends to still greater depths. Differences in permeability of the individual beds also affect the depth of weathering and consequent enrich- ment. The phosphorite bed, for example, is commonly weathered to greater depths than the other beds are. Enrichment in the Bloomington Canyon workings is almost entirely restricted to the shale bed and to the beds below it that are only weakly vanadiferous elsewhere. This restriction suggests that solutions migrating downward along the shale bed locally departed from it and moved downward into the underlying beds along steeply dipping fractures. In the northern drift of the Consolidated mine, Where the rocks dip 50° to 60° W., all of the exposed beds are locally enriched, an indication that solutions there worked downward mainly along bedding planes but locally departed from them along fractures to enter either hanging-wall or footwall beds. Local lenses, particularly those between the phosphorite and silt- stone beds, are strikingly enriched in places. Weather- ing apparently removed the calcite from these lenses and left a spongy mass favorable for the deposition of secondary minerals. The rocks below the zone of enrichment are hard, black, and in places contain pyrite and white coatings on fractures and joints. Sincosite is found locally below the zone of enrichment in the Bloomington Canyon workings but is of no quantitative importance. The lenses and siltstone beds below the shale are calcareous in the Consolidated mine, but the beds of the vana- diferous zone there are only weakly calcareous. All the rocks in the bottom of the Paris Canyon winze, however, are calcareous, and the V205 content of the zone as a whole is only about 0.6 percent, in contrast to the average of 0.9 percent elsewhere in the mine. Because the vanadium content of calcareous rocks from the bottom of the winze is lower than that of rocks elsewhere in the mine, samples from other workings may not be reliable indicators of the grade of the zone in deeper parts of the syncline, since the so-called unenriched rocks may in fact be somewhat enriched, even though they contain no visible signs of enrichment. Should this scenario prove true, the grade of the zone at depth might be more comparable to that near the base of the winze than to the higher values in the other workings. VARIATIONS IN THICKNESS AND VANADIUM CONTENT OF THE VANADIFEROUS ZONE The thicknesses and vanadium contents of the beds of the vanadiferous zone vary considerably in in- dividual samples, although averages of groups of samples generally differ only slightly. However, some lateral variations, due chiefly to weathering, can be observed. As table 13 shows, the beds on the normal limb of the syncline are about 20 percent thicker than those on the overturned limb. The siltstone and phosphorite beds on the normal limb are somewhat less vanadiferous. The vanadiferous zone as a whole contains less than average amounts of vanadium in the apparently un- altered part of the Consolidated mine, owing to the presence of abundant, weakly vanadiferous calcareous lenses. These lenses also add to the thickness of the zone, making it somewhat thicker than it is elsewhere on the overturned limb (compare tables 13 and 14). The most conspicuous differences in grade and thickness are between individual samples, as table 15 shows. Extreme differences in thickness occur on the overturned limb; differences in the vanadium contents of these rocks reflect the effects of leaching and enrichment. TABLE 13.—Compa rison ofthe thicknesses and Pa nadium contents ofthe vanadiferous beds on the normal and overturned limbs of the Paris syncline, Idaho Normal limb1 Overturned limb1 Average V205. Thickness. V205, Thickness, V205, Thickness, in percent in ft in percent in ft in percent in ft Siltstone .................................................... 0.82 4.9 0.89 4.0 0.85 4.5 Phosphorite ...... 43 2.5 .61 2.1 .51 2.3 Shale2 ........................................................ 1.36 3.8 1.35 3.1 1.36 3.4 .92 11.2 .98 9.2 .95 10.2 1Based on samples obtained in underground workings only. 2Includes siltstone bed below shale where it is high grade in the Consolidated mine and 16 South incline. PARIS-BLOOMINGTON DEPOSITS 17 TABLE 14.——Effect of lenses on the unaltered vanadiferous zone in the Consolidated mine, Paris-Bloomington area, Idaho With lenses Without lenses Lenses only Thickness, V205, Thickness, V205, Thickness, V205, Bed in ft in percent in ft in percent in ft in percent Siltstone ...................................................... 4.8 0.67 4.5 0.68 1 1 0.48 Phosphorite .49 2.2 .55 1.4 .26 Shale ............... 1.22 2.7 1.40 2.4 .30 All beds ...................................................... 10 .86 9.4 1.22 4.9 .33 TABLE 15. ——Range in thickness and grade of the vanadiferous zone in the Paris-Bloomington area, Idaho Range of local averages1 Extreme ranges2 Thickness. V205 content, Thickness, V205 content. in ft in percent in ft in percent Siltstone ................................. 3.7-5.2 0.72-0.98 1.1-6.9 0.06-2.43 Phosphorite ..... .. 2.0-3.0 .39—.67 .5~4.7 .08-2.43 Shale ....................................... 2.9-4.1 .84-1.05 .05—7.2 .06-5.733 Full zone ................................ 8.8-11.8 .84-1.05 5.8-16.1 .07-1.80 1Range in local average thickness and grade is the range in average V205 content determined at each of the underground workings; it applies to the full zone as well as to individual beds. Extreme range in thickness and grade is the range in individual samples of the bed. For the full zone, it is the range in complete sections of the three beds. 3Includes Siltstone bed (0.6 ft thick) below the shale, which contains 2300 percent V205 at this point. The distribution of vanadium within each bed is generally uniform. Sections of the vanadiferous zone in which the thinnest recognizable units of each bed were sampled separately show that the vanadium content of the Siltstone is remarkably constant from top to bottom, that the Siltstone layers of the phosphorite bed contain more vanadium than the phosphatic ones, and that the upper phosphatic and lower blocky parts of the overlying shale bed contain slightly less vanadium than the central part of the bed. RESOURCES The tonnages, stratigraphic distributions, and vanadium contents of the vanadiferous beds—classed as measured, indicated, and inferred subeconomic resources—are summarized in table 16. If the tonnage were projected to the northern end of the area mapped, it would be approximately tripled. Estimates were made of the shale bed separately because of its much higher grade; of the shale and the phosphorite, in case it should prove impractical to mine the full zone; and of the full zone. The Wyodak Coal and Manufacturing Co. did not undertake experi— mental mining of the shale bed alone, but it did successfully complete experimental mining of the other two units in the 14 South workings on the normal limb of the syncline by the longwall mining method. Although no experimental mining was attempted on the overturned limb, operations manager J .D. Johnson believed that the top-slicing method of mining would be suitable for that part of the deposit. Although some possible means of vanadium recovery have been identified (Ravitz and others, 1947; De Voto and Stevens, 1979), the commercial feasibility of production from the zone has not been demonstrated, and its resources are considered subeconomic for the present. They are classed as measured where the vanadiferous beds are known to be present, and their thicknesses and vanadium contents have been well established by taking closely spaced samples and measurements. The limits of error in these estimates are judged not to exceed 15 percent. Resources have been measured in the Paris Canyon area, where the zone was sampled in 725 ft of underground workings, four trenches, and one short crosscut; in the Con- solidated mine area, where the zone was sampled in nearly 500 ft of underground workings; and in Bloomington Canyon, between the 12 South adit and 16 South incline, where the zone was sampled in 18 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO TABLE 16.—Subeconomic vanadium resources in the Paris-Bloomington area, Idaho Average Tons thickness, V205 content, (short) in ft in percent Measured Shale bed only1 .......................................... 180,000 3.6 1.35 Shale and phosphorite ............................ 320,000 6.0 1.02 beds. Shale, phosphorite, and .......................... 600,000 10.8 .93 siltstone beds. Indicated Shale bed only ............................................ 1,200,000 3.0 1.4 Shale and phosphorite ....... 2,300,000 5.5 1.0 Shale, phosphorite, .................................. 4,000,000 10.0 .9 and siltstone beds. Inferredz Shale bed only ...................................... 15,000,000- 3.0 1.0-1.4 25,000,000 Shale and phosphorite ...................... 25,000,000- 5.5 .7-1.1 40,000,000 Shale, phosphorite, ............................ 50,000,000- 10.0 .6—1.0 and siltstone beds. 75,000,000 1Siltstone below shale included where it is enriched in Consolidated mine and 16 South incline. 2Inferred resources stated for area south of northernmost intersection of zone in Paris Canyon. nearly 1,200 ft of underground workings. The measured resources lie above a depth of 200 ft below drainage level. In areas between or adjacent to those containing measured resources, the cover is such that the structural continuity of the zone is not well established, and samples and measurements are few and widely spaced. Resources in these areas—estimated partly on the basis of samples and measurements and partly on the basis of geologic projection from adjacent areas— are classed as indicated. Because the hard, calcareous rocks encountered in the bottom of the Paris Canyon winze may be typical of unaltered rock, the beds sampled in most of the workings may be slightly enriched and therefore not representative of the rocks at depth. Indicated re- sources are therefore limited only to the beds between the surface and the depth below the old Tertiary surface equivalent to the bottom of the winze. This elevation in Paris and Bloomington Canyons is about 6,000 ft, but, at the divide between the two, it is assumed to be about 6,250 ft. Because the vanadiferous zone is faulted out at the surface north of trench 4 North (on the northern side of Paris Canyon), the existence of resources on the overturned limb is not certain enough to justify the estimation of any indicated resources there. There are no exposures east and north of the 16 South incline to define the extent of the zone on the normal limb. Since no evidence contradicts the reasonable assumption that the zone extends for some distance in this direction, resources extending 500 ft east of the 16 South incline are arbitrarily classed as indicated. The limits of inferred subeconomic resources are defined by the boundaries of the measured and in- dicated resources in the southern and southwestern parts of the area, by the supposed position of a possible fault along the eastern edge of the Paris syncline, and, arbitrarily, by the northernmost exposure of the zone in Paris Canyon. The Paris syncline is known to extend as far north as Liberty, about 7 mi north of Paris Canyon, and it seems likely that the zone extends over the entire structure. For the purpose of this report, however, it is assumed that its northern limit cor- responds to the northern limit of the area mapped, about 3.5 mi north of Paris Canyon. As table 16 (footnote 2) states, if inferred resources were considered to extend to the latter limit, the estimates would be approximately triple those shown in the table. Most of the inferred resources lie above a depth of 3,500 ft below drainage level. The vanadium content of the greater part of the inferred resources may more nearly approximate that of the zone in the bottom of the Paris Canyon winze (0.6 percent V205) than that of the zone sampled in other workings (0.9 percent V205). Even if this assumption PARIS-BLOOMINGTON DEPOSITS 19 is correct, part of these resources—those on the over- turned limb of the syncline above an elevation of about 6,000 ft and those northeast of the 16 South incline on the normal limb above an elevation of 5,800 to 6,000 ft—should be comparable in grade to those of the measured and indicated resources. To emphasize the uncertainties caused by these factors, the tonnages and grades of the inferred resources are expressed as ranges in table 16. To prepare these estimates, samples were weighted according to their abundance as well as to their area of influence. Thus, in determining the grade and thick- ness in a given block, the average of many samples from an underground working was given more weight _ than samples from individual trenches, even though the area of influence of the latter may have been greater. Because samples in the underground workings were rather uniformly spaced (about 10 ft apart), averages were determined by weighting the samples by thickness but not by distance between them. Samples of part of a bed were not included. The beds differ in density, and, if units had been combined according to weight rather than volume alone, the average grade would differ by 0.02 to 0.03 percent V205. Because the discrepancy thus produced would have been small, weight was disregarded in calculating the average grades of the shale—phosphorite combina- tion and the full zone. The beds range in density from 12.3 to 15.3 ft3/t and 14 ft3/t. The estimates of the full zone and of the shale- phosphorite combination were based on a density of 14 ft3/t; those of the shale bed alone were based on a density of 15.3 ft3/t. EXPERIMENTAL MINING AND PROCESSING The Wyodak Coal and Manufacturing Co. demon- strated in 1943 the feasibility of mining both the shale and phosphorite together and the full zone. The USBM (Ravitz and others, 1947) undertook preliminary studies of recovery processes. In the mid-1970’s, Earth Sciences, Inc., did experimental mining in Blooming- ton Canyon and developed a plan for extracting the shale bed by augur mining. Mine production of 211,000 tons/yr was planned to yield, after roasting and leaching, 3,000,000 lb of V205 as ferrovanadium, 67,500 lb of selenium, 33,900 lb of U308, and 59,100 tons of phosphate concentrate containing 32 percent P205; molybdenum was recoverable, too, but in amounts too small to market (De Vote and Stevens, 1979). Cost analysis projected a 9.5 percent rate of return on investment. Although this rate was thought to be minimally adequate then, development did not follow; given the subsequent decline in commodity prices, it may be assumed that production is still not economically feasible. ORIGIN Many marine black shales contain concentrations of vanadium in the range of 0.2 to 0.35 percent V205 (for example, Davidson and Lakin, 1961; Vine, 1969; Vine and Tourtelot, 1970), and a few contain somewhat larger amounts (Krauskopf, 1955; Coveney and Martin, 1983; Poole and Desborough, 1985). They also com— monly contain suites of trace metals similar to those of the vanadiferous zone of the Phosphoria Formation. None have been reported, however, in the range of 2.0 to 2.5 percent V205 found within the zone in western Wyoming or at the average of nearly 1 percent over an average thickness of about 10 ft found in the Paris- Bloomington area of Idaho. Exceptional also is the total vanadium content of the Meade Peak Member. The member averages 0.02 to 0.1 percent V205 in southeastern Idaho and western Wyoming (fig. 3) (Maughan, 1976, 1980). All of its beds at Coal Canyon except one contain vanadium in amounts greater than the average crustal abundance, and 10 beds outside the vanadiferous zone contain it in the range of 0.2 to 0.42 percent V205 (McKelvey and others, 1953; Gulbrandsen, 1960). The Meade Peak Member in general and its vana- diferous zone in particular, then, have exceptional concentrations of vanadium, although they are other— wise similar in trace-metal content to many other black shales. Their origins very likely are similar, but the higher concentration of V205 in the Meade Peak Member requires explanation. The vanadiferous zone is clearly the product of primary deposition from upwelling water in an anoxic marine environment below wave base in a water depth of 100 to 300 m or so. The deposits formed on a large embayment on the outer continental shelf on the western side of the North American craton at low latitude, where deep, cold, nutrient-rich seawater welled up as a result of di— vergent upwelling in a tradewind belt. The car- bonaceous mudstones, of which the vanadiferous zone is a part, are the outer shelf facies equivalents of nearer shore phosphorite, chert, carbonate rock, red- beds, and evaporites. The carbonaceous, phosphatic, and cherty sediments are direct or indirect manifesta- tions of the high biologic productivity characteristic of the shelf environment. Concentration of the vanadium and associated metals took place at about the same time as deposition or burial of the associated sediment, which accumulated very slowly in an environment of circulating seawater. 20 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO The reasons for these conclusions are straight- forward. The lateral continuity over thousands of square miles of thin beds of very fine grained sediment underlain and overlain by rocks containing marine fossils allows only syngenetic deposition in a marine environment. The abundant organic matter (part of which may be bituminous), pyrite, and other sulfides testify to accumulation under reducing conditions. The geographic setting of the Meade Peak Member, including its relation to upwelling, has been established by regional stratigraphic studies of the Phosphoria Formation (for example, McKelvey and others, 1959; Sheldon, 1964, 1981). The requisite slow deposition and circulating 'seawater are also necessary for phosphorite (McKelvey, 1946). The concentrations of vanadium in the zone are so high (locally, 150 times or more its crustal abundance of 150 g/t) (Rankama and Sahama, 1950) relative to its ordinary occurrence that they could have been attained only in the absence of the dilution that would have resulted from sedimentation at ordinary rates. Furthermore, the vanadium content of seawater (2.5 ppb (Brewer, 1975)) is so low that even the total extraction of the vanadium in a standing column of seawater several times the probable depth of the Meade Peak sea would not begin to equal the amount of vanadium found in the vanadiferous zone.2 Beyond these observations, the origin of the vana- diferous zone can be addressed only with questions and speculations. For example, how are vanadium and other metals held? What led to their precipitation from seawater? What factors are responsible for their exceptionally high concentration in both the vana- diferous zone and in the Meade Peak Member as a whole? What was the specific environment in which the high concentrations were localized? As McKelvey (1946) discussed, the vanadium in the vanadiferous zone may occur as sulfide, as adsorbed ions on clay minerals, or contained in organic matter. Vanadium has been found in each of these forms in some sedimentary rocks (Rankama and Sahama, 1950). Some support for its presence in a clay or other silicate mineral came from R.A. Gulbrandsen (written com- munication, 1985), who tested a sample of the Coal Canyon vanadiferous zone in hydrochloric and hydro— fluoric acids. He found that essentially no vanadium was dissolved by hydrochloric acid but that vanadium was dissolved by hydrofluoric acid. Since hydrofluoric acid dissolves silicate minerals, this result suggests 2For example, a water column 1 m2 and 1,000 m deep—10 times the probable depth of the Meade Peak sea—containing 2.5 pbb V (see table 16) would contain only 2.5 g V. One square meter of the upper part of the vanadiferous zone in Sublette Ridge, averaging 0.686 m thick and 0.594 percent V, contains 8,521 g V, the equivalent of the vanadium in about 3,400 such water columns. that vanadium occurs in one or more silicate minerals. Desborough’s (1977) subsequent electron microprobe studies showed that vanadium is held in organic matter. The recent work of E.C.T. Chao, J.A. Minkin, and J.M. Back (written communication, 1986) con- firmed both observations—that vanadium is within semifusinite but that some of it is attached to illite.3 According to George Breit (written communication, 1986), “...there seems to be a general progression in the form of vanadium in black shales. Soon after deposition the vanadium is tightly associated with organic frag- ments. In older units, especially those past the point of oil generation, the vanadium is tied up in clays and occasionally sulfides.” - In what chemical form is the vanadium held by the organic matter? The richest bed of the vanadiferous zone in western Wyoming has a V205 content of 2 percent or more and an organic matter content of about 25 percent; thus, its V205 content would be 8 percent or more. Carbonaceous matter is known to be a good chemical adsorbent for metals, and bituminous shales that are high in vanadium and some other metals are relatively high in organic carbon (Des- borough, 1977; Maughan, 1980; Holland, 1984). But many shales rich in organic carbon, including some in the Phosphoria Formation, have low trace-metal contents, so the correlation is not necessarily direct. Bader (1937, cited by Rankama and Sahama, 1950, p. 599) believed that the vanadium in vanadiferous crude oils and bituminous shales is held in porphyrin complexes. A Persian crude oil has been reported to contain 2.82 percent V (5.03 percent V205), but the maximum vanadium content in most vanadiferous crudes, including the heavy crudes from which vanadium is recovered commercially, is of the order of 1,000 ppm (Hunt, 1979). Bader’s average vanadium content for asphalt is 5,400 ppm, although Fischer (1973) reported that many asphaltites contain about 1 percent. Recently, a vanadyl deoxophylloerythroetio- porphyrin was extracted from an oil shale at Julia Creek in Queensland, Australia (Ekstrom and others, 1983; Miller and others, 1984). This oil shale, which contains as much as 0.5 percent V, is in a mica- montmorillonite clay; that part of the vanadium in organic matter is in kerogen (Riley and Saxby, 1982). Saoiabi and others (1983) also reported vanadyl and 3Semifusinite, which consists of carbonized woody tissue, is character- istically of terrestrial origin, but, because the environment of deposition of the vanadiferous zone was more than 125 mi from the nearest land, where an arid climate probably prevailed, a terrestrial origin for it here seems dubious. The point of its origin is not important, however. As Chao recognized, its environment of deposition was marine, and it is clear that the living plants could not have acquired such large quantities of vanadium. ORIGIN 21 nickel porphyrins from Moroccan oil shales. The vanadium content of these porphyrins is not known to us, but, if they were the vanadium analogue of the nickel porphyrin abelsonite (C31H32N4Ni) (Milton and others, 1978), they would contain 17.85 percent V205— just about the amount needed if half the organic matter in the highly vanadiferous beds were in that form. Premovic and others (1986) also found V02+ porphyrins in the Triassic Serpiano marl of Switzer— land and the Cretaceous La Luna shaly limestone of Venezuela, which contain about 2,700 and 1,100 ppm V, respectively. About half of the total vanadium in both formations is in kerogen, and the remainder is in the extractable organic, HCL-soluble, and HF-soluble fractions. The V02+ porphyrins are in the extractable organic and kerogen fractions. In view of these occur- rences and others reported by Premovic and others, it seems likely that at least some of the vanadium in the Meade Peak Member is in the form of porphyrins. In addition to the phosphorite in the vanadiferous zone in the Paris—Bloomington area, several other phosphorites in the Meade Peak Member contain vanadium, generally in the range of 0.2 to 0.4 percent V205. In fact, vanadium is currently being recovered as a byproduct of the manufacture of elemental phosphorus in southeastern Idaho. Some of the vanadium in phosphorite probably occurs in carbonate fluorapatite in substitution for phosphorus (Rankama and Sahama, 1950; McConnell, 1953). Some of it, however, may also be held by organic matter (Kraus- kopf, 1955; Gulbrandsen, 1966). Phosphorites also contain moderately high concen- trations of the other metals found in the vanadiferous zone. Southeastern Idaho phosphorites, for example, average 90 ppm Cd, 800 ppm Cr, 85 ppm Cu, 30 ppm M0, 100 ppm Ni, 30 ppm Se, 5 ppm Ag, 3 ppm Th, and 250 ppm Zn (Gulbrandsen, 1977). These concentrations, for the most part, are much higher than those of the average phosphorite reported by Altschuler (1980)—— namely, 18 ppm Cd, 125 ppm Cr, 7 ppm Cu, 9 ppm Mo, 53 ppm Ni, 4.6 ppm Se, 2 ppm Ag, and 195 ppm Zn. No gold determinations have been made for phosphorite in the Sublette Ridge or Paris-Bloomington areas, but 0.4 ppm Au has been found farther north in western Wyoming (Love, 1984). According to Rankama and Sahama (1950), vana- dium in upper oxygenated seawater is in the quin- quevalent state. Reduction in oxygen-deficient bottom waters would lead to its precipitation in some form. Holland (1979) found that the concentrations of Ag, Cu, Ni, V, and Zn at the highest median value reported for 20 sets of black shales by Vine and Tourtelot (1970) represent a surprisingly narrow range of enrichment factors over their concentrations in seawater of 2.5 to 8x105 (table 17). Even at the highest 95th percentile in the same 20 sets of samples, the enrichment factor for these metals ranged only from 5 to 23X105. Noting the reported removal of such metals by precipitation from the anoxic part of the water column in the Black Sea (Brewer and Spencer, 1974), Holland (1979, p. 1679) concluded that, although “...many metals participate actively in biochemical cycles...the available data (see, for instance, Brewer, 1975) suggest that the concen- tration of trace metals in living organisms is not large enough to account for the observed enrichment of trace metals in black shales. Apparently, the concen- tration of metals in organic-rich sediments owes more to chemical precipitation and to reactions with dead organic remains than to their incorporation in living organisms” (see also Holland, 1984). The common occurrence of these. metals in bi- tuminous shales deposited in the upwelling environ- ment led Brongersma-Sanders (1969) to suggest that they are transported from surface to deeper water layers by organisms and concentrated there by resolu- tion after death. Precipitation of the metals may take place if the decay of organic matter leads to deoxy- genation of the bottom waters and the formation of H28. She cited data given by Schutz and Turekian (1965) on Ag, Ni, and Co showing “...high concentra- tions in areas of high organic productivity (upwelling water) accompanied by an increase of concentration with depth” (Brongersma-Sanders, 1969, p. 234). Thus, upwelling water itself may be important as a source of the metals. Bruland (1980) reported a marked increase in the content of Zn, Cd, Ni, and Cu with depth in the central North Pacific; the increase for Zn, Cd, and Ni between the surface and a depth of 1.0 to 1.5 km is severalfold, a pattern almost identical to that long known for phosphate and nitrate (fig. 4). Cutter (1982) found that selenium also increases with depth and that it changes from selenate and selenite forms in the oxic zone to an organic selenide form in the anoxic zone. If, as seems likely, vanadium and the other characteristic trace metals of black shales follow the same pattern, upwelling would move metal-rich materials into the environment of deposition, just as it does for phosphate and the other major nutrient elements.4 Baturin (1982) confirmed the effectiveness 4Hite (1978) also considered a high metal source important, but he envisioned that source as a warm brine, enriched in phosphate as well as metals, refluxed to the sea from an evaporite basin and reacting with cold seawater. Perhaps the most problematic aspect of this novel hypothesis is the idea that the brine (presumably) retains its coherence as a sheet extending a few hundred miles seaward from its source, across bottoms where the fossil record shows sessile organisms lived, to the black shale environment. 22 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO TABLE 17.——Some trace metals in seawater in comparison with their concentrations in black shales (Holland, 1979) Highest median Highest 95th percentile of 20 sets Enrichment of 20 sets Enrichment of black factor over of black factor over Seawater, shale samples. seawater. shale samples. seawater. in ppb in ppm1 x105 in ppm1 X105 10 2.5 20 5 1.000 30 3,000 90 200 8 500 20 300 .3 700 .7 500 8 1,000 16 50 17 100 34 1,000 4 2,000 8 1.500 5 7.000 23 lVine and Tourtelot (1970). zBrewer (1975). 3Boyle and others (1977). 4Sc1ater and others (1976). of upwelling in helping to concentrate the metals when he reported that diatomaceous ooze in the area of strong upwelling off southwestern Africa contains 455 ppm Ni, 337 ppm Zn, 129 ppm Cu, 500 ppm Mo, and 360 ppm V, all fixed mainly in organic matter. Although the metal content of organisms is generally very small, Z.A. Vinogradova and V.V. Koval’skiy (see Calvert, 1976) reported 200 ppm Cu and 2,600 ppm Zn in Black Sea plankton. Remarkable concentra— tions of vanadium occur in the blood of ascidians and holothurians. Rankama and Sahama (1950) reported an astonishing 10.4 percent V (18.6 percent V205) in holothurian blood (dry weight)—again, just about the right concentration needed if roughly half of the organic matter in the richly vanadiferous beds were in such a form. The idea that a significant part of the organic matter in the vanadiferous zone might be derived from holothurian blood is too bizarre to discuss further, but it does seem possible (even likely) that organisms played some part in extracting vanadium and other metals from seawater, similar to the role that has long been postulated for them in concentrating phosphate (see Gulbrandsen’s (1969) review). Addressing the similar assemblage of metals found in phosphorites, Prevot and Lucas (1980) sug- gested that organisms bring these elements together and that the first trap for these elements is biologic. Subsequent steps in their concentration are accom- plished by both biochemical and inorganic chemical processes. As we indicated previously, it is possible that vanadium and some other metals are in the form of a porphyrin. Determination of the crystal structure of the vanadyl porphyrin from Australia is regarded as confirmation of the hypothesis that this porphyrin is a degradation product of chlorophyll (Ekstrom and others, 1983). But, although this hypothesis in turn may be taken as proof of the biologic origin of crude oil, it is also proof that the vanadium, replacing the magnesium of chlorophyll, was not sited in the porphyrin by a primary biologic process. Premovic and others (1986) reached a similar conclusion with respect to the Serpiano marl and the La Luna shaly limestone. Are organisms the common denominator that brings the black shale metals (here taken to be Ag, Cd, Cr, Mo, Ni, Se, Ti, U, V, and Zn) together, as Prevot and Lucas suggested? According to Bowen (1966), only Mo, Se, V, and Zn are essential to the life of some or all organisms. Cr and Ni may have an essential role, and Cd is known to be involved in a special biologic process. N o biologic role has been reported for Ag, Ti, or U. Three other metals that are also essential to life (Co, Cu, and Mn) are present in black shales in low concentrations only. All of the black shale metals are known to be moderately to highly toxic to organisms in more than trace amounts. It thus seems unlikely that biologic necessity brings all of these metals together, particularly in the high concentrations in which they are found in the black shales. However, even though a biologic role is unknown for Ag and is limited or uncertain for Cd, Cr, and Ni, all four elements are known to occur in both marine plants and animals, and all the black shale metals are known to occur in seawater. It is possible, therefore, that, even though some metals do not play a biologic role, they might be carried to the bottom in small concentrations in living or dead organic tissue. Organic matter has been shown to be effective in adsorbing metals from seawater (Krauskopf, 1955). DEPTH, IN KILOMETERS 23 ORIGIN 0 I I I | I I 0 x xI x I I I x x x x x X X xx Xx 1 _ x. 1 _ X _ x - _ x E m U) as E m 2 — x — a 2 — x — 2‘ E 9 a _ Q — x _ >4 — x Z Z ‘ 3 — — ‘ 3 — — E x E ‘L 34 3 o _ x — — x " 4 —— x — 4 — x —— _ x x 5 AI | I I I I Ix 5 B I I I xI I 0 10 20 30 40 0.0 1.0 2.0 3.0 NITRATE PHOSPHATE (NANOMOLES PER KILOGRAM) (NANOMOLES PER KILOGRAM) 0 I I I I I I I I I I x _ x —. x x 1 ~ x _ _ x _ 2 — x _ _ x _ 3 — X _. _ x _ 4 — x _ x 5 CI I | I l I | ¥ I I I 0.0 0.2 0.4 0.6 0.8 1.0 1.2 CADMIUM (NANOMOLES PER KILOGRAM) DEPTH, IN KI LOMETERS I 0 I 1* I I I I I II x — x _. x' x 1— x —— _ x ._ 2— x“ _ x— 3— x— __ x- 4— X“ x D 5 I I I I | I | I l | 0 2 4 6 8 10 12 NICKEL (NANOMOLES PER KILOGRAM) DEPTH, IN KILOMETERS (NANOMOLES PER KILOGRAM) x x x x x __ x _ _ x _ _ x _ _ x _ _ x _ _ x _ _ x ._ El I I I I I I Ix I 0 2 4 6 8 10 ZINC FIGURE 4.——Variation of (A)nitrate, (B) phosphate, (C) Cd, (D) Ni, and (E) Zn in seawater with depth in the North Pacific Ocean at 32°41.0’ N., 144°59.5’ W. (drawn from Bruland, 1980). 24 VANADIF‘EROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO All things considered, the rough correlation between metal content and amount of organic carbon noted by other authors (for example, Maughan, 1976, 1980, 1984; Desborough, 197 7) for metalliferous black shales does not seem to indicate that a biologic process is directly responsible for the deposition of the metals but rather shows the importance of organic matter in progressively concentrating the metals. Organisms withdraw small quantities of them, perhaps using them as nutrients, and these metals are then carried in dead tissue to the bottom, where they may be dissolved to add to their concentrations in bottom or interstitial water. Decaying organic matter creates the reducing conditions necessary for the fixation of vanadium and other metals and forms a site favorable for the fixation of some. As time passes, vanadium may move from organic matter to illite or other minerals (George Breit, written communication, 1986). The importance of an anoxic environment to the precipitation of the metals is worth stressing, since one thing that they all have in common is that their precipitation is favored by reducing conditions. In summary, it seems reasonable to assume that the concatenation of such processes, combined with the slow accumulation of sediment, may account for the occurrence of vanadium and several other metals in the Meade Peak Member and in some other black shales. But the extraordinary concentrations of vanadium in the richest part of the vanadiferous zone (an order of magnitude higher than the ordinary black shale highs) seem to require conditions or processes not yet identified. The increase and subsequent de- crease in the vanadium content from the base to the top of the zone over an area of several hundreds of square miles in western Wyoming (see, for example, pl. 2) imply a gradual buildup and then decline in one or more favorable environmental conditions, which can- not be defined with certainty. One possibility is that the principal variable was sediment accumulation rate, which increased and then decreased the time available for the resolution of vanadium from decaying organic tissue and for its withdrawal from circulating seawater. That variable alone, of course, would not account for the variation in the distribution of high concentrations of other metals within the vanadiferous zone. Although the specifics of the environment in which the high concentrations of vanadium and other metals were deposited cannot yet be identified, the factors controlling the deposition of vanadium are independent of those controlling the deposition of phosphate, uranium, and carbonate, which increase westward from Wyoming to southeastern Idaho without appreci- able change in the vanadium content ofthe zone. Such an increase would be expected if the redox potential (Eh) were the controlling factor in the deposition of vanadium and if hydrogen ion concentration (pH), temperature, and CO2 content controlled the precipita- tion of apatite (which hosts the uranium) or carbonate, for those factors do not necessarily vary in concert. Although the deposition of V, P, U, and carbonate may take place simultaneously, the maximum concentration of vanadium and most other metals is favored by a pH and temperature low enough to prevent or retard accumulation of diluting phosphate or carbonate. Apatite, for example, is unstable below a pH of about 7.0, and carbonate is unstable below a pH of about 7.5 (Gulbrandsen, 1969; Bentor, 1980). Apatite in modern sediments precipitates where the dissolved oxygen in bottom waters is lowest, probably owing to increased preservation of organic phosphorus rather than to an Eh control over the precipitation of apatite (Burnett and others, 1980). Although other metals have been treated here as a group along with vanadium, their distribution within the zone does not conform exactly to that of vanadium, nor does it in other black shales. For example, in a group of 18 samples of black shales from the Penn- sylvanian Mecca Quarry and Logan Quarry of Illinois and Indiana, analyzed by Coveney and Martin (1983), only the high value of zinc (14,300 ppm) and total organic matter (46.5 percent) corresponded with the high vanadium value (10,100 ppm). High U (240 ppm) and Mo (1,600 ppm) values were in another sample, and high values of Pb, Se, Cu, and Ni (500, 400, 500, and 1,300 ppm, respectively) were each in other samples. According to Disnar (1981), the capacity for organic matter to fix U, Cu, Pb, Zn, and Ni increases with increasing pH, but, for Mo and V, it increases with decreasing pH; both groups of metals might be fixed coevally in the intermediate pH range. The divergence in the metal concentrations expected from these relations is not seen in either the Meade Peak or the Mecca-Logan samples, but it seems likely that the differences in, the distribution of the metals do reflect differences in their behavior ‘under varying Eh, pH, temperature, and perhaps other chemical conditions. The exceptionally high concentrations of vanadium and other metals throughout the Meade Peak Member also deserve explanation. Perhaps these concentrations are merely the result of a long continued strong upwelling environment. An enriching factor, however, might have been volcanic ash, the presence of which is shown by the occurrence of buddingtonite (the am- monium feldspar) in most of the Meade Peak mud- stones. According to Gulbrandsen (1974, p. 697), its “...distribution and concentrations...seem to show that great airfalls did not occur, but that small falls were ORIGIN 25 frequent for along period of time....” Buddingtonite is formed by the alteration of volcanic glass in an ammonia—rich environment, probably beneath the sediment-seawater interface. Its alteration conceivably might have released metals, which added to the amounts derived from seawater. Premovic and others (1986) favored volcanic ash as the source of vanadium in the Serpiano and La Luna rocks. CONCLUSIONS The vanadiferous zone of the Meade Peak Member in the Sublette Ridge area of Wyoming and the Paris— Bloomington area of Idaho contains about 5.9 million tons of indicated subeconomic resources averaging about 0.9 percent V205. Indicated resources in the Afton area of Wyoming area above drainage level total about 35 million tons of the same quality (Love, 1961). Inferred resources in western Wyoming and south— eastern Idaho of about the same tenor are many times larger. The zone also contains Cd, Cr, Mo, Ni, Se, Ag, Th, Zn, and (in the Paris-Bloomington area) U in amounts many times their average crustal abundance. Although.the feasibility of mining these deposits has been shown by the Wyodak Coal and Manufacturing Co. and Earth Sciences, Inc., the feasibility of re- covering vanadium commercially remains to be demonstrated. When the need for vanadium is reflected in increased prices, however, processes for its extrac- tion will be explored again, under the justifiable expectation that economic production will be possible one day. The initial output probably will not be large— the production planned by Earth Sciences, Inc., would have supplied only 8.5 percent of national vanadium demand—but Meade Peak resources could eventually be a more important source of vanadium and several other metals. ACKNOWLEDGMENTS We are grateful to E.C.T. Chao, J.A. Minkin, and J .M. Back for their electron microprobe studies of the vanadiferous shale. We also thank George Breit, George Ericksen, Michael Fleischer, R.A. Gulbrand- sen, H.D. Holland, and E.K. Maughan for their helpful suggestions. Gertrude A. Sinnott kindly assisted in obtaining references. We also thank Kathie R. Fraser for her fine work in editing the manuscript. REFERENCES CITED Allsman, P.'I‘., Majors, F.Z., Mahoney, S.R., and Young, W.A., 1949a, Investigation of Sublette Ridge vanadium deposit, Lincoln County, Wyo.: U.S. Bureau of Mines Report of Investigations 4476. 1949b, Investigation of Salt River Range vanadium deposit, Lincoln County, Wyo.: U.S. Bureau of Mines Report of Investi- gations 4503, 18 p. Altschuler, Z.S., 1980, The geochemistry of trace elements in marine phosphorites, pt. I, Characteristic abundance and enrich— ment, in Bentor, Y.K., ed., Marine phosphorites— Geochemistry, occurrence, genesis: Society of Economic Paleontologists and Mineralogists Special Publication 29, p.19-30. Armstrong, F.C., and Cressman, ER, 1963, The Bannock thrust zone, southeastern Idaho: U.S. Geological Survey Professional Paper 374-J, p. J1-J22. Bader, E., 1937, Vanadin in organogenen Sedimenten, I, Die Grfinde der Vanadinanreicherung in organogenen Sedimenten: Zentralblatt for Mineralogie, Geologie, und Palaontologie, Abt. A, p. 164. Baturin, G.N., 1982, Phosphorites on the sea floor: New York, Elsevier, 343 p. Bentor, Y.K., 1980, Phosphorites—The unsolved problems, in Bentor, Y.K. ed., Marine phosphorites—Geochemistry, occurrence, gene- sis: Society of Economic Paleontologists and Mineralogists Special Publication 29, p. 3-18 Bowen, H.S.M., 1966, Trace elements in biochemistry: New York, Academic, 241 p. Boyle, E., Sclater, RR, and Edmond, J.M., 1977, The distribution of dissolved copper in the Pacific: Earth and Planetary Science Letters, v. 37, p. 38-54. Brewer, P.G., 1975, Minor elements in seawater, in Riley, J.P., and Skirrow, G., eds., Chemical oceanography, 2d ed.: New York, Academic, p. 415-496. Brewer, P.G., and Spencer, D.W., 1974, Distribution of some trace elements in the Black Sea and their flux between dissolved and particulate phases: American Association of Petroleum Geolo- gists Memoir 20, p. 137-143. Brongersma-Sanders, M., 1969, Origin of trace metal enrichment in bituminous shales, in Hobson, G.D., and Speers, C.C., eds., Advances in organic geochemistry, Third International Congress Proceedings: Oxford, Pergamon, p. 231-236. Bruland, K.W., 1980, Oceanographic distribution of cadmium, zinc, nickel and copper in the North Pacific: Earth and Planetary Science Letters, v. 47, p. 176-198. Burnett, W.C., Veeh, H.H, and Soutar, A., 1980, U-series, ocean- ographic and sedimentary evidence in support of Recent formation of phosphate nodules off Peru, in Bentor, Y.K., ed., Marine phosphorites—Geochemistry, occurrence, genesis: Society of Economic Paleontologists and Mineralogists Special Publication 29, p. 61-72. Calvert, SE, 1976, The mineralogy and geochemistry of nearshore sediments, in Riley, J.P., and Chester, R., eds., Chemical ocean- ography, v. 6: New York, Academic, p. 187—280. Coveney, R.M., Jr., and Martin, SR, 1983, Molybdenum and other heavy metals of the Mecca Quarry and Logan Quarry shales: Economic Geology, v. 78, p. 132-149. Cutter, G.A., 1982, Selenium in reducing waters: Science, v. 217, p. 829-831. Davidson, D.F., and Lakin, H.W., 1961, Metal contentof some black shales of the Western United States: U.S. Geological Survey Professional Paper 424-C, p. C329-C331. Desborough, GA. 1977, Preliminary report on certain metals of potential economic interest in thin vanadium—rich zones in the Meade Peak Member of the Phosphoria Formation in western Wyoming and eastern Idaho: U.S. Geological Survey Open-File Report 77-341, 27 p. 26 VANADIFEROUS ZONE OF THE PHOSPHORIA FORMATION, WYOMING AND IDAHO De Voto, RH, and Stevens, D.N., 1979, Uraniferous phosphate resources and technology and economics of uranium recovery from phosphate resources, United States and free world: U.S. Department of Energy Report GJBX-79, 3 v. Disnar, J. R., 1981, Etude experimental de la fixation de metaux par un material sedimentaire actuel d’origine algaire, II, Fixation ‘in vitro’ de U022+, Ni2+, Zn2+, Pb2+, 002+, Mn2‘, ainsi que dc V03‘, M0422 Ge032': Geochimica et Cosmochimica Acta, v. 45, p. 363-379. Ekstrom, A., Fookes, C.J.R., Hambley. T., Locke, H.J., Miller, S.A., and Taylor, J.C., 1983, Determination of the crystal structure of a petrophyrin isolated from oil shale: Nature. v. 306, p. 173-174. Fischer, RR, 1973. Vanadium: U.S. Geological Survey Professional Paper 820, p. 679-688. Gulbrandsen, R.A., 1960, Petrology of the Meade Peak phosphatic shale member of the Phosphoria Formation at Coal Canyon, Wyoming: U.S. Geological Survey Bulletin 1111~C, 146 p. 1975, Analytical data on the Phosphoria Formation, Western United States: U.S. Geological Survey Open-File Report 75-554, 45 p. 1966. Chemical composition of phosphorites of the Phosphoria Formation: Geochimica et Cosmochimica Acta, v. 30, p. 769-778. 1969. Physical. and chemical factors in the formation of marine apatite: Economic Geology. v. 64, p. 365-382. 1974. Buddingtonite. ammonium feldspar, in the Phosphoria Formation, southeastern Idaho: U.S. Geological Survey Journal of Research, v. 2. p. 693-697. 1977, Final environmental impact statement. v. 1. Develop- ment of phosphate resources in southeastern Idaho: Washington, D.C., U.S. Department of the Interior and U.S. Department of Agriculture, p. 1—53. Hite, R.J.. 1978. Possible genetic relationships between evaporites, phosphorites, and iron-rich sediments: The Mountain Geologist. v. 14. no. 3. p. 97~107. Holland. H.D.. 1979, Metals in black shales—A reassessment: Economic Geology. v. 74. p. 1676-1680. 1984, The chemical evolution of the atmosphere and the oceans: Princeton. N.J.. Princeton University Press. 582 p. Hunt, J.M.. 1979. Geochemistry and geology of petroleum: San Francisco. Freeman. 619 p. Krauskopf, KB, 1955, Sedimentary deposits of rare metals: Eco- nomic Geology 50th anniversary volume. pt. 1. p. 411-463. ‘Lotspeich, F.O., and Marquard. EL. 1963, Minor elements in bedrock. soil. and vegetation at an outcrop of the Phosphoria Formation on Snowdrift Mountain in southeastern Idaho: U.S. Geological Survey Bulletin 1181-F. 42 p. Love, J.D., 1961, Vanadium and associated elements in the Phos- phoria Formation in the Afton area. western Wyoming: U.S. Geological Survey Professional Paper 424-C, p. C279- C282. 1984, Gold. silver. and other selected trace elements in the Phosphoria Formation of western Wyoming: Wyoming Geological Association Annual Field Conference. 35th, Casper. Wyo.. 1984, Guidebook, p. 383-387 Manheim, F.T.,and Gulbrandsen, RA. 1979, Marine phosphorites. in Burns. R.G., ed., Marine minerals: Mineralogical Society of America Short Course Notes. v. 6, p. 151-170. Maughan. E.K., 1976. Organic carbon and selected element distribu— tion in the phosphatic shale members of the Permian Phosphoria Formation, eastern Idaho and parts of adjacent States: U.S. Geological Survey Open-File Report 76-577, 92 p. 1980, Relation of phosphorite, organic carbon, and hydro- carbons in the Permian Phosphoria Formation, Western United States of America: Bureau de Recherches Geologiques et Minieres Document 24, p. 63-91. 1984, Geological setting and some geochemistry of petroleum source rocks in the Permian Phosphoria Formation, in Wood- ward, J., Meissner, FF and Clayton, J., eds, Hydrocarbon source rocks of the greater Rocky Mountain region: Denver, Colo.. Rocky Mountain Association of Geologists, p. 281-294. McConnell, D., 1953. Radioactivity of phosphatic sediments: Eco— nomic Geology, v. 48. p. 147~148. Mc Kelvey, V.E., 1946, Preliminary report on the stratigraphy of the phosphatic shale member of the Phosphoria Formation in western Wyoming. southeastern Idaho, and northern Utah: U.S. Geological Survey Open-File Report, 138 p. McKelvey, V.E., and Strobell, JD. Jr., 1955. Preliminary geologic maps of the Paris-Bloomington vanadium area. Bear Lake County. Idaho: U.S. Geological Survey Miscellaneous Field Studies Map MF-41. scales 1:12.000 and 1:4.800. McKelvey, V.E., Smith, L.E., Hoppin, R.A.. and Armstrong, F.C., 1953, Stratigraphic sections of the Phosphoria Formation in Wyoming, 1947-1948: U.S. Geological Circular 210, 33 p. McKelvey, V.E., and others, 1959, The Phosphoria, Park City, and Shedhorn Formations in the western phosphate field: U.S. Geological Survey Professional Paper 313-A, 17 p. Miller, S.A., Hambley. T.W., and Taylor, J.C., 1984, Crystal and molecular structure of a natural vanadyl porphyrin: Australian Journal of Chemistry. v. 37, p. 761-766. Milton, C., Dwornik, E.J.. Estep—Barnes, P.A., Finkelman, R.B., Pabst, A., and Palmer, 8., 1978, Abelsonite. nickel porphyrin, a new mineral from the Green River Formation, Utah: American Mineralogist. v. 63, p. 930-937. Poole, F.G., and Desborough, G.A., 1985, Metal concentrations in marine black shales: U.S. Geological Survey Circular 949, p. 43-44. Premovic, P.I.. Pavlovic, M.S.. and Pavlovic, N.Z., 1986, Vanadium in ancient sedimentary rocks of marine origin: Geochimica et Cosmochimica Acta, v. 50, p. 1923-1931. Prevot. L., and Lucas, J., 1980. Behavior of some trace elements in phosphatic sedmentary formations, in Bentor, Y.K., ed., Marine phosphorites—Geochemistry, occurrence, genesis: Society of Economic Paleontologists and Mineralogists Special Publication 29, p. 31—40. Rankama, K., and Sahama. T.G.. 1950. Geochemistry: Chicago. University of Chicago Press, 912 p. Ravitz. S.F., Nicholson. I.W., Chindgren, C.J.. Bauerle. L.G.. Williams. F.P.. and Martinson, M.T., 1947, Treatment of Idaho- Wyoming vanadiferous shales: American Institute of Mining Engineers Technical Publication 2178, 14 p. Riley, K.W.. and Saxby, JD, 1982. Association of organic matter and vanadium in oil shale from the Toolebec Formation of the Erhomang Basin, Australia: Chemical Geology, v. 37. p. 265-275. Rowe, J.J., and Steinnes. E., 1976, Determination of rhenium in sedimentary rocks by instrumental activation analysis using epithermal neutrons: Radiochemistry and Radioanalysis Letters, v. 26. p. 324. Rubey. W.W.. 1943, Vanadiferous shale in the Phosphoria Forma- tion, Wyoming and Idaho [abs]: Economic Geology, v. 38, p. 87. 1958. Bedford, Wyo., geology: U.S. Geological Survey Geologic Quadrangle Map GQ-109, scale 1:62.500. Saoiabi. A.. Ferhat, M., Barbe, J.M.. and Guilard, R.. 1983, Metals, including vanadyl and nickel porphyrins. in the oil shales of Timahdit. Morocco: Fuel. v. 62. p. 963-965. Schutz, DE. and Turekian, KK. 1965. The investigation of the geographical and vertical distribution of several trace elements in sea water using neutron activation analysis: Geochimica et Cosmochimica Acta. v. 29, p. 259-313. REFERENCES CITED 27 'Sclater, F.R., Boyle, E., and Edmond, J.M., 1976. On the marine and Planetary Science, v. 9, p. 251-284. geochemistry of nickel: Earth and Planetary Science Letters, v. Vine, J.D., 1969, Element distribution in some Paleozoic black 31, p. 119—128. shales and associated rocks: US Geological Survey Bulletin Sheldon, RR, 1964, Paleolatitudinal and paleogeographicdistribu- 1214-13, 32 p. tion of phosphorite: US Geological Survey Professional Paper Vine, J.D., and Tourtelot, E.B., 1970, Geochemistry of black shale 501-C, p. C106-C113. deposits—A summary report: Economic Geology, v. 65. p. —1981, Ancient marine phosphorites: Annual Review of Earth 253-272. *U.S. GOVERNMENT PRINTING OFFICE: 1986-181-‘409:40016 DEPARTMENT OF THE INTERIOR U.S. GEOLOGICAL SURVEY PROFESSIONAL PAPER 1465 PLATE 1 6 6 o‘ o‘ o o o o 129 é? /\° ’\‘° l I I II B. o 1-5'@ r (E. Limb) \ TURNER CANYON \ \ B. 1 abandoned \\ (E. Limb) \ B. 1x brecciated l (w. Limb) /' S. .2'9,@ 0.721 4 (W. Limb) ,’ Shaft down 71' (w. Limb) B. 2 1.3’@ I (E. Limb) \ EVANS CANYON \ B. 2x Faulted out \ (W. Limb) \ 15 000 N \ I B. 23 In progress I I 1 \ B. 2b Planned } I I / B. 3 2.8’@ l " ( LEYLAND CANYON \ \\ I i II B. 3a 3.6’@ (—fi 3.3 (a 0.782 EXPLANATION Surface, dashed where unsurveyed Tunnel Trench number by E.-W. coordinates. Trenches dug by the 0.8. Bureau of Mines (B), Wyodak Coal and Manufacturing Co. (W), and the US. Geological Survey (S) Thickness and grade (V205) of vanadif BYOUS shale. Assays and thickness by Wyodak Coal and Manufacturing Co (0.75‘) and by the US. Bureau of Mines (0.752) 0 1000 2000 FEET l,___ __~ 0 1000 M ETE RS o o‘ o‘ o‘ 60% 6’? «90 4’9 r 1 A) s‘ T 37 ”‘3'5 @ 0'80 g PETEREIT GULCH W. T 33 N—Faulted out 5 W. T 30 N—Faulted out / B. 4 T 25 N-3.1’@ 0-851 \ TUNNEL CANYON w. T 21 N-3.1’@ 0.71. 3.3 @ 0.782 = B. 5 T 17 N—3.9’@ 0.8‘ 2 3/ w. T 15 N—3.7’@ 0.92‘,4.0@ 0.592 ; w. T 13 N—3.2’@ 0.751 : /i/TunneI-in 50' W. T 10 N-3.0’@ 0.751 / Tunnlelein 97' |-' 7 ' Tum ”‘ 8 RAYMOND CANYON B. 6 T 5 N—2.8’@ 0.45‘ Tunnel-in 97' w. T 4 N—41’@ 0.45‘ ,3.0@ 0.422 Tunnemn 601 B. 7 T 2 N—4.2'@ 0.6‘ :_ Tunfiemn 33' w. T 0 N—S-3.2’@ 0.61’3.3'@ 0.552 " o-N-S — w. T 2 S—3.7’@ 0.521. 3.5'@ 0.672 ,\ w. T 4 s—3.5'@ 0.65‘,3.5'@ 0.662 g X w. T 6 5—3.1 @ 0.91 ,4.2'@ 0.262 < w. T 11 s-3.7’@ 0.95‘,4.0'@ 0.542 ) w. T 16 5—2.9'@ 023‘ ,3.o'@ 0.852 l ’ w. T 20 34.01;: 0.50’,3.5'@ 0.435z . ~. B. 7a.T 24 8—380; 0.78‘,3.1'@ 0.802 )\ w. T 29 5-3.5’ \ w. T 33 5—3.4'@ 0.92‘.3.5'@ 0.582 } w. T 37 5—200; 0.501 .38. T 36 8—? / w. T 39 s—3.5'@ 0.851 , 2.6’@ 0.6652 / _ r . 1 . _5' \ 0352 Tunnel in 36' w. T 44 s 3.2 64 0.75 3 (a 1 I, COAL CANYON s. T41 s-3.7'@ 0.85 \1unnel-in 99' ,3.1'@ 0.582 ; \ w. T 47 5—3.2' ‘ \\ w. T 49 S—4.3’@ 0.921. 3.5’@ 0.662 ) 5000 s— B. 9. T 54 5—3.0'@ 0.921 / B. 93 T 56 S—Faulted out W. T 59 5—3.0’@ 1.051 , 3.4’@ 0.712 B10. T 515- 3.4'@ 0.57‘ B 11. T 65 S—Faulted out w’ \\ JACKSON CANYON (DEER CANYON) W. T 66 5—3.6’@ 0.71 . 3.2’@ 0.852 W. T 69 5—3.3’@ 0.921 . 3.2’@ 0.052 W. T 74 8—3.4'@ 0.91.3.4’@ 0.7451 W. T 79 S—Faulted out W. T 84 S—3.8’(u; 0.851. 3.8'(03 0.8252 A‘ W. T 89 8—3.5’ B 12 T 99 S—2.5’@ 0.671 B 13 T 105 8—3.5’, 2.7’@ 0.7952 N/N 8.133 Faulted out B 14. T 119 S—2.8’@0.651 B 15 Faulted out /'\ Crosscut-opened up 360’ S Bed not recognized _—— -_'-___ 10 000 5- SPRING CANYON YORK CANYON FRANCIS CANYON 150003 6 8 <0 Transit control by Wyodak Coal and Manufacturing Co. VANADIFEROUS ZONE OUTCROP AT SUBLETTE RIDGE, WYOMING, SHOWING THE V205 CONTENT AND THE THICKNESS OF THE ZONE AT SAMPLED LOCALITIES DEPARTMENT OF THE INTERIOR U.S. GEOLOGICAL SURVEY a‘. 7’ JAN 27 1987 11 (a y PROFESSIONAL PAPER 1465 PLATE 2 6400 LEVEL ADIT R 7 N 119 FEET IN HANGING 6400 LEVEL ADIT R 7 S 1245 FEET IN 9% WALL 6400 LEVEL ADIT R 7 S 132 FEET IN LIMESTONE 6900 LEVEL ADIT 38 FEET IN F0 0 FEET v205% 3T" 0.25'@ 0310 ii. 0.45'@ 0.575 025001.90 0.45’@ 1.15 0 55% 0.87% a '03, SJU 0.". ('31- 0.55’@ 0.415 0.55’@ 0.210 ... 0.50'@ 0.078 T FEET fl; ———>— -—~——> 0.30’@ 1.83 030(4) 1.98 1 2.6’@ 0.72% 4%— 155091.121. 0.30’@ 1.13 f 0.650;: 0.395 .‘ 1.05’@ 0.125 3'. EXPLANATION Limestone Siltstone Conglomerate PERCENT V205 0 1.0 2.0 VANADIFEROUS FTOWT 7.5 FEET LA (‘ 0.51% (, 0.91% 0.3563 0.255 0.60’@ 0.670 113060 1.870 0.20’@ 1.920 0.30’(1L‘ 1.050 0.65’@ 0.305 0.4514? 0.175 0.25’GL‘ 0.105 37’@ 0.62% 2.65’@ 0.83% SILTSTONE 4440 FEET— 0.75’@ 0.595 _ 0.3'@ 1.78 _ 0.60’@ 1.45 ‘ 0.40’@ 0.415 0.60’@ 0.300 0.40’@ 0.100 0.65’@ 0.075 6840 LEVEL ADIT 4 105 FEET IN J I N.D. 0.10’@ 2.04 0.50’@ 2.02 0.50’@ 1.09 0.50’@ 0.490 0.50’@ 0.310 0.50’@ 0.115 0.40’@ 0.071 FEET 4i Assays performed by the US. BUREAU of MINES DISTRIBUTION OF VANADIUM WITHIN THE VANADIFEROUS ZONE AT SUBLETTE RIDGE, WYOMING