0137’;— Wo 7D 9“ 11.959 AYS Amificial Recharge to a Freshwater-Sensitive Brackish-Water Sand Aquifer, Norfolk, Virginia GEOLOGICAL SURVEY PROFESSIONAL PAPER 939 Prepared in cooperation with the Department of Utilities, Norfolk, Virginia writ”??? \yfl‘ man “.1 END “ *sfi’v“ 23mm §?\ 600,9 ‘ \y _‘ \‘w M \ I ’EB 22 1977 Artificial Recharge to a Fresmater-Sensflzive Brackish-Water Sand Aquifer, Norfolk,Virginia By DONALD L. BROWN and WILLIAM D. SILVEY GEOLOGICAL SURVEY PROFESSIONAL PAPER 939 Prepared in cooperation with the Department of Utilities, Norfolk, Virginia UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON: 1977 UNITED STATES DEPARTMENT OF THE INTERIOR THOMAS S. KLEPPE, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director Library of Congress Cataloging in Publication Data Brown, Donald L, 1938- Artificial recharge to a freshwater—sensitive brackish-water sand aquifer, Norfolk, Virginia. (Geologica: Survey Professional Paper 939) Bibliography: p Supt: of Docs. no.: I 19.16:939 1. Water, Underground—Virginia-Norfolk region-Artificial recharge. l. Silvey. ”William 0., joint author. II. United States Geological Survey. III. Norfolk, Va. Dept. of Utilities. IV. Title. V. Series; United States Geological Survey Professional Paper 939. T040437 552.4’9 76-20538 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-02920-1 CONTENTS Page English-metric equivalents ........................................................... V Deterioration of aquifer hydraulic properties ................ . ............. Abstract ........................... 1 Head buildup Introduction. 1 Excess head buildup due to temperature and viscosity Location of area ........... 1 Excess head buildup due to hydraulic conductivity Previous investigations ............................. 1 deterioration ....................................................................... Acknowledgments ................................ 2 Current-meter traverses . ............................................... Well field .......................... .. 3 Specific capacity ............ Injection well 1 (IW—l) ............................................................. 4 Conductivity surveys ................................................................. Injection well 2 (IW-2) ............................................................. 4 Aquifer heterogeneity ............................................................ Observation wells ............... 4 Water quality ................................................................................ Annular-space well (ASW) .............................. 4 Comparison of the chemistry of city and formation water ...... Observation well 2 (OW42) ............................. 4 Water sampling .......................................................... , .............. Observation well 3 (OW-3)... 4 Analytical results ..................................................... Test well 1 (TW—l) ................................ 4 Changes in the concentrations of calcium, magnesium, and Geology and hydrology of the injection sand 4 sodium ................................................................................ Aquifer tests .................................................................................. 9 Effect of cation exchange on formation clay during Injection system ............................................................................ 11 injection ............................................................................. Pre—injection calibrations 12 Laboratory determination of hydraulic conductivity Current meter .................. 12 deterioration ....................................................................... Transducers 14 Injection test 4 ......................... Turbidity ...................... 14 Injection specific capacity Water-quality monitor.... 14 Current-meter traverses ............................................................. Injection tests ................... l4 Cause of clogging of IW— 2 during injection phase of test 4 Injection test 1 ........................................................................... 14 Chemical effects observed during withdrawal phase of test 4 Injection test 2 ........................................................................... 14 Analysis of project ........................................................ -. ............... Injection test 3 ........................................................................... 15 References cited ............................................................................. ILLUSTRATIONS FIGURE 1. Map showing location of test site .................................................................................................................................................... 2. Sketch showing Norfolk injection project well field, Moore’s Bridges Filter Plant .. ‘ 3. Diagrammatic sketch of injection well IW—2 .............................................................. 4. Geophysical logs and stratigraphy of test well TW-l ..................................................................................................................... 5. Graph showing composite cumulative curve of particle-size distribution of the injection sand ................................................... 6. Geophysical logs of injection zone, test well TW—l ....................................................................... 7. Photograph showing thin section of a part of the injection sand (955 to 975 ft or 291 to 297 m)... . 8. Hydrograph of injection well IW—2 showing specific capacities derived from May 1970 step-drawdown test 9. Graph showing current-meter traverses in injection well IW‘—2 prior to injection of freshwater ................................ , .................. 10. Graphs showing theoretical and measured head data for injection well IW-2 and observation well OW—S, test 1 ..... _ ................ ll. Graph showing pre-injection and post-injection hydraulic gradients .......................................................................... ;. l2. Graph of current-meter traverses in injection well IW=2 of the pre-injection flow and the flow after 5 hours of injection during test 1 .................................................................................................................................................................................. 15. Graph showing current-meter traverses in injection well IW—‘2 of the pre-injection flow and the withdrawal flow during test 1 ............................................................................................................................................................................................... 14. Graph showing current-meter traverses in injection well lW—2 of the pre~injection flow, test 2, and the flow after 24 hours of injection during test 2 .............................................................................................................................................................. l5. Graph showing current- -meter traverses in injection well IW—2 during the injection phase, test 2 .............................................. l6. Graph showing current-meter traverses in injection well IW—2 of the pre-injection flow and the withdrawal flow, test 2 ............................................................................................................................................................................................... l7. Graph showing current-meter traverses in injection well IW—2 of the pre-injection flow, test 2, and the flow after 75 hours of withdrawal pumping during test 2 ................................................................................................................................ Page 15 16 16 17 19 28 28 31 32 32 32 33 35 36 36 39 40 42 43 46 49 53 23 24 25 26 IV FIGURE 18. TABLE 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39, . Specific capacity of injection well IW—2 during the injection phase of tests 1, 2, and 3 .......... . Concentration of major constituents from Moore’s Bridges Filter Plant city water and formation water .................................... . Variations in water chemistry of mixed freshwater and formation water during withdrawal, injection tests 1 and 2 ................. . Sodium to chloride ratio and associated concentrations of calcium in samples collected during injection tests 1 and 2 ............ . Effect of water chemistry on laboratory hydraulic conductivity for core samples from injection zone of observation well CONTENTS Graph showing current-meter traverses in injection well IW—2 of the pre-injection, injection, and withdrawal flows, test 3 ............................................................................................................................................................................................... Graph showing changes in injection rate with time during tests 1, 2, and 3 ................................................................................ Graph showing pre-injection withdrawal specific capacity of injection well IW—‘2 and injection specific capacity during tests 1, 2, and 3 ................................................................................................. . ..... Graph showing conductivity profiles in observation well OW—3 during injection phase, test 2 ......................... Graph showing conductivity profiles in observation well OW—3 and current-meter traverse in injection well IW—2, test 2 ......................................................................................................................................................................... Graph of conductivity profiles in observation well OW—3, test 4.... Graph of conductivity profiles in observation well OW—2, test 4 ................................................................................................... Lithologic section showing injection zone in injection well IW—2 and zones of detection of freshwater in observation wells in OW—2 and OW—3 ............................................................................................................................................................ Graph showing chloride, sodium, and dissolved solids versus specific conductance of recovered injected water from tests 1, 2, and 3 .............................................................................................................................................................................. Graph showing calcium concentration and sodium to chloride ratio versus specific conductance during the recovery of injected water, tests 1 and 2 .......................................................................................................................................................... Graph showing effects on the hydraulic conductivity caused by injecting city water containing various chemicals into a core saturated with formation water .......................................................................................................................................... Graph showing variations in specific capacity of injection well IW—2 during injection tests .................................... Graph showing specific capacity of injection well IW—2 prior to any injection of freshwater and variation of specific capacity during test 4 .................................................................................................................................................................... Graph showing variation of specific capacity in injection well IW—2 for injection phases 5, 6, and 7, test 4.... Graph showing variation of specific capacity in injection well IW—2 for injection phases, test 4 ................... Graph showing current-meter traverses in injection well IW—2 of pre-injection and injection flow, test 4 ..... Graph showing current—meter traverses in injection well IW—2 of pre-injection, injection, and withdrawal flow, test 4.. Graph showing current-meter traverses in injection well IW—2 of withdrawal flow, test 4 .......................................................... Graph showing changes in injection and withdrawal flow and conductivity profile in injection well IW—2 after 8 Mgal (30,300 m3) withdrawn, test 4 ........................................................................................................................................................ Graph showing changes in injection and withdrawal flow and conductivity profile in injection well lW—2 after 15.1 Mgal (57,150 m3) withdrawn, test 4 ........................................................................................................................................................ Graph showing changes in chloride concentration with volume of water withdrawn, test 4 ....................................................... Sketch map showing possible well locations for a five-well injection field at Moore’s Bridges Filter Plant ............................... TABLES . Heavy minerals identified from cores from injection zone of observation well OW—3 .................................................................. . Statistiml parameters of the injection sand ....................................................................... . X-ray diffraction analysis of core samples from injection zone, observation well OW—3 .............................................................. Thickness and sand-shale ratios for the injection zone within the well field ................................................................................ . Specific capacities at various pumping rates after 1 hour for the injection sand in test well TW—l and injection well IW—2 .............................................................................................................................................................................................. OW—3 ............................................................................................................................................................................................. 35 37 38 40 41 42 43 44 45 47 48 49 52 CONTENTS ENGLISH-METRIC EQUIVALENTS [Although the conversion factors are shown [0 four significant figures, the metric equivalents in the text oi this paper are shown only to the number of significant figures consistent with the values of the English units] English unit Metric equivalent inch (in) foot (it) gallon (gal) gallon (gal) million gallons (Mgal) gallon per minute (gal/min) gallon per minute per foot (gal min—1ft“) foot per day (ft/d) cubic foot per day per foot (ft3d‘1ft-') pound per square inch (lb/in?) horsepower 25.4 .3048 3.785 .003785 3785 .06309 .207 .3048 .0929 6.8948 .7457 millimetres (mm) metre (m) litres (1) cubic metre (ms) cubic metres (m5) litre per second (l/s) litre per second per metre (I s-‘m-l) metre per day (m/d) cubic metre per day per metre (m3d-1m—1) kilopascals (kPa) kilowatt (kW) ARTIFICIAL RECHARGE TO A FRESHWATER-SENSITIVE BRACKISH-WATER SAND AQUIFER, NORFOLK, VIRGINIA By DONALD L. BROWN and WILLIAM D. SILVEY ABSTRACT During late 1971 and early 1972, three injection and withdrawal tests were made at the Norfolk, Va., injection site. In test 1, freshwater was injected at the rate of 400 gal/min (25 VS). The specific capacity of the well decreased from 15.4 to 9.3 gal min—'ft—l or 3.2 to 1.915‘1m—l ofdraw- down at the end of 260 minutes of injection. In test2, the initial injection rate of 400 gal/min (25 l/s) decreased to 215 gal/min (l4 l/s) after 7,900 minutes of injection. The specific capacity dropped from 14.2 to 3.7 gal min-lft—l or 2.9 to 0.77 ls-lm-l during the test. At the start of test 3, the aquifer accepted water at a maximum rate of 290 gal/ min (18 1/ s), but the injection rate decreased to 100 gal/min (63 VS) within 150 minutes and continued to decrease to a low of 70 gal/ min (4.4 l/s) after approximately 1,300 minutes. The specific capacity decreased from 3.7 to 0.93 gal min—1h-l or 0.77 to 0.19 ls-‘mrl. Attempts at redevelopment of the injec- tion well failed to improve the specific capacity. Current-meter surveys made during injection and withdrawal pumping indicate that the reduction in flow and specific capacity were due to a uniform reduction in the hydraulic conductivity of all contri- buting zones in the aquifer and not to a complete shutoff of flow from selected parts of the aquifer. The hydraulic and chemical data indicate that the uniform loss of specific capacity of the contributing zones was due to dispersion of interstitial clay and that this clay would readily respond to chemical treatment for the purposes of decreasing or eliminating dispersion. Subsequently, a pre—flush of 3,000 gal (11 m3) of 0.2N calcium chloride solution was injected in front of the freshwater prior to injection test 4. The initial specific capacity was 4.3 gal min—1ftx or 0.891s‘lm" and, by redevelopment pumping during injection, the specific capacity was improved to 5.3 gal min—’ft—l or 1.1 ls“m'1. After injecting 4 Mgal (million gallons) (15,100 m3) of freshwater, an additional 3,000 gal (11 m3) of 0.4N calcium chloride solution was added to the formation. A total of 20,146,100 gal (76,300 m3) of freshwater was injected during test 4. The specific capacity remained fairly constant throughout the injection of the first 16 Mgal (60,600 m5), indicating the stabilization of interstitial clay in the aquifer was accomplished. After 16 Mgal (60,600 m3) had been injected, particulate clogging began occurring, and the specific capacity fell to less than 3 gal min-1ft”l or 0.62 ls‘lm-l. Current-meter traverses made during test 4 injection showed that because of the deterioration of aquifer properties caused by tests 1,2, and 3, the calcium chloride preferentially treated the most permeable part of the aquifer. As a result, a combination of dispersion and particulate clogging caused the lower 40 ft ( 12 m) of the aquifer to be plugged so that the freshwater selectively injected into the upper part of the aquifer. As withdrawal pumping began, the lower part of the aquifer became unclogged resulting in the brackish formation water mixing with the freshwater. Only 20 percent of the volume injected during test 4 was recovered as potable freshwater. Tests 1 and 2 showed that if clogging of the screen in the injection well can be prevented, as much as 85 percent of the injected water can be recovered and will remain within the drinking- water standards of the U.S. Public Health Service (1962). Treatment of the injection well with a clay stabilize: prior to the if‘jt‘t‘.‘ tion of any freshwater will minimize clogging and increase recovery of freshwater. If plugging of the screen can be prevented so that injection and withdrawal flow patterns remain similar, the storage of, freshwater in a brackish-water sand aquifer is feasible. INTRODUCTION The water supply for the city of 1‘21: Milk ”21., came: from surface impoundments 1n the independent cities of Nansemond, Norfolk, and Virginia Beach. During the winter months, when water demand is 10w and reseroirs are full, water must be diverted. from the reservoirs and allowed to escape to the ocean W"? ”a the potential use of t; a; water unfulfilled. It has been estimated (Schweitzer, 1: ' commun., 1968) that as much as 2. 3 133:3! é‘biliion game (9, 460, 000 m3) of water per winter (111319“ 2011] , . .. . able for use if sufficient storage areas were availahie The U.S. Geological Survey and the city £13“ orfolk entered into a cooperative program to def: frame if it would be possible to utilize the water j herttiy {lowing to waste by processing it in the treatment giants and storing it underground 1n aquifers containirr. saiine water. The freshwater would then be retri e reef. during the summer months when peak water demands and tow water levels in reservoirs place strains on the ,‘sivesent water system. LOCATECN OF AREA The Norfolk injection site is at Moore's Bridges Filter Plant, Norfolk. Va. (fig. 1). Excess water from the filter plant supplies the injection project. The water is taken from the treatment system after it has been chlorinated, settled, and filtered, but prior to the final chlorination and liming. PREVIOUS INVESTIGATIONS The first attempt at artificial recharge of freshwater into brackish-water aquifers in the Coastal Plain of Virginia was conducted by D. J. Cederstrom (Cederstrom, 1957) in 1946 at Camp Peary. During that experiment, water was injected into a brackish-water aquifer over a period of 85 days. The well into which the water was injected had screens in the intervals of 430 to 440 ft (131 to 134 m) and 450 to 475 ft (137 to 145 m) below land surface. The well ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 1000 2000 3000 FEET I—l—f—u' 300 600 900 METRES EXPLANATION. .IW-Z Injection well and number \ .TW—l l; \\ Test well and number ’ ”I WASHINGTON, .ow—3 ’/ g D.C. Observation well and number ’ v n " I e VIRGINIA , / I I / § I ’71 ( i) Q ‘3 v [/1 Maj E ad’ / ’ ’I *‘s 0 83° r/ \’ \/ \N\ "g; g S , \ 4— K o o 100 MILES 2 37 if l__‘._|_,_r|_'_|_| ( — ,/,, pressure monitoring pipe (/ / l\7 \ //// ’1 18-inch MED-millimetre) fiberglass casing sfiafi "” °- . .. . . .93 33' ~83 ‘5', __ a», 32-mch (8 10-mIl/Imetre) pit casing 00 E (“‘3 E r" a O) s a -. ’ 100 l— 1Q l x K :9 L: . — 50 200 l— Cement grout (I) 300 _ Lu 5 8-inch (ZOO-millimetre) fiberglass casing _ 100 E LU Lu “- 2 Z Z — 400 _ —_ 8 Lu < c t z 0 LL / emen grou E [I a 150 g D 500 — (D Z 3 3 < - Wooden block and stainless-steel straps for fixing —’ ; IZ-inch (38-mi/Iimetre) pressure monitoring ; 9 600 — pipe to well easing O L“ d m (I1 I — 200 I '— l— 0- o. D o 800 _ 8-inch l200-mi/limetre), hea vy-duty, wire-wrapped, _ 250 40-5/0 t, stainless-s teel screen I Vz-inch (38-mi/limetrel, heavy-duty, wire-wrapped, ' 40-3/0 t,stainless-steel screen 900 — . . 10 feet (3 metres) of blank stainless—steel pipe Grave, paCkN ’gg°° Gravel pack 00: 800; o — 300 1000 — Total Depth 990 feet (302 metres) FIGURE 3.—Diagrammatic sketch of injection well IW-2. barrel was employed to core the injection zone in OW—3. mination of clay type, grain size, porosity, hydraulic The tool is designed to take core in unconsolidated conductivity, and identification of mineral content of the material by encasing the core in a rubber sleeve as the core sand. is cut. Selected intervals of the core were submitted to the Table 1 lists the accessory minerals identified by Geological Survey laboratory in Denver, Colo, for deter— William Lockwood of the laboratory. Figure 5 is a DEPTH BELOW MEAN SEA LEVEL, IN FEET SPONTANEOUS 387 625 682 792 1538 2111 POTENTIAL MV ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA RESISTIVITY, IN OHMS PER METRE GAMMA RAY SQUARED PER METRE 1.0 AN D CALI PER 100 2.0 ROCKS OF \ late MIOCENE AGE 10 ROCKS OF middle MIOCENE AGE , KS OF late EOCENE A ROCKS OF CENOMANIAN AGE ROCKS OF NOMANIAN—ALBI AGE ROCKS OF ALBIAN—APTIAN AGE ROCKS OF APTIAN-N EOCOMIAN AGE FIGURE 4.—Geophysical logs and stratigraphy of test well TW—l. 2,5 30 15 POROSITY, IN PERCENT BULK DENSITY. IN GRAMS/Cc 3.0 —15 118 190 208 241 469 643 DEPTH BELOW MEAN SEA LEVEL, IN METRES GEOLOGY AND HYDROLOGY OF THE INJECTION SAND 7 w100 '3 (I) I E 90 < 9 g 80 z 5‘: / t— 70 j— I (5 E 60 i E n: 50 LL! Z L: w 40 / Lu _1 Q E 30 SE / u. 0 20 I § /// 8 -” a: m 0- 0 v— m o o O Q ~— to O o O m 0 O. O O ". L‘? O O. 0' Lu 0 O O O o o .—' LD v— 5 Lu PARTICLE-SIZE DIAMETER, IN MILLIMETRES — N E <7) SAND E CLAY SILT GRAVEL 8 >0.004 0004—00625 Very fine Fine Medium Coarse Very Coarse 5 2 00625—0125 0125—025 025—050 0.5—1.0 1.0—2.0 Lu 9 o o E _z_ 5.0 12.2 9.8 39.4 33.0 0.8 0.2 O E L g 0 Lu 0. FIGURE 5.—Composite cumulative curve of particle-size distribution of the injection sand. cumulative curve showing the distribution of grain sizes by percent of the total weight. The curve is a composite of particle-size analyses of four cores taken at selected depth intervals in the injection zone. Using figure 5 and the statistical parameters shown in table 2, a description of the injection sand can be obtained. The injection zone is bounded above and below by laterally persistent silty-clay to clayey-silt confining beds (fig. 6). The clay in the confining beds was identified by X- ray analysis as a multi-layered mixture of illite and mont- morillonite, plus mon tmorillonite and minor amounts of kaolinite (table 3). Because the montmorillonite and mixed-layer clays are swelling clays, the amount of these clays present in the injection sand was of concern. If the clay percentage was significant and swelling occurred when the clay was sub- jected to freshwater, the hydraulic conductivity of the injection zone would be decreased. To determine the spacial relationship of the clay and sand, several thin sections were prepared by the laboratory from the OW—3 core. A red thermosetting plastic was injected into the core under a vacuum while the core was still in the rubber sleeve. The “original” fabric, pore pattern, effective porosity, and interstitial relationship of the clay was thus preserved (fig. 7). Figure 7 shows slight cross-bedding and a minor amount of interstitial clay. Even though table 3 shows that the clay content in the injection sand is as high as 25 percent, the interstitial clay content (fig. 7), which caused the clogging problems, is as low as 5 percent of the total DEPTH BELOW LAND SURFACE, IN FEET 850 T 900 950 SPONTANEOUS 7 20 MV Ier Sample 2 a E n: U) L o .. m 3 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA DUAL INDUCTION LOG RESISTIVITY, lN OHMS PER RED PER METRE 10 POTENTIAL Injection Zone METRE SQUA 1.0 , FORMATION DENSITY LOG GAMMA GAMMA BULK DENSITY, IN GRAMS/CC 20f ,, 2.25 " 259 '0 C (V a) IN U T N__ EDIUM INDUCTION _ > > E 9 (no 7 274 U E 290 w ULK DENSITY MPENSATION ROSITY -———- 304 > > :9 (/30 'U C (U U) 5 30 POROSITY, IN PERCENT FIGURE 6.—Geophysical logs of injection zone, test well TW—l. DEPTH BELOW LAND SURFACE, IN METRES AQUIFER TESTS 9 TABLE 1,—Heavy minerals identified from cores of injection zone of observation well OW—S Sample No. 69Va-1 69Va-2 69Va-3 69Va-4 69Va-5 69Va-6 69Va-7 69Va-8 Interval (ft) .................... 892— 912- 975- 892— 912— 935— 955— 955— 912 932 995 912 932 955 975 975 Lithology ...................... clay clay clay sand sand sand sand sand Percent of sample Total heavy minerals 0.8 0.5 0.3 1.3 4.0 0,6 0.9 1.3 Percent of total heavy minerals Apatite ................ < 1 ............... Amphibole 1 .......... Actinolite <1 2 2 2 1 Tremolite < l .......... l Biotite... <1 ............... 2 l Chlorite.... 4 12 1 ..... 1 Diopside .. <1 <1 1 .. . l 2 ..... Epidote . <1 <1 <1 5 4 l9 9 5 Gamet.. ..... 5 <1 10 10 15 5 15 Homblende .. ..... <1 1 .......... 4 Hypersthene. .......... <1 ......................... Kyanite. ..... <1 <1 1 l 1 Magneti . ..... 11 24 23 20 24 27 29 Muscovite. l 4 8 2 9 l l 2 Pyrite.... 85 4 9 7 ..... 3 10 ..... Rutile ..... <1 <1 ............... l Staurolite.. ..... 14 5 14 22 12 25 19 Titanite .......... 3 5 2 l ..... Tourmaline. ..... <1 <1 2 <1 1 l ..... Zircon ............ <1 1 ..... 1 ..... Unidentifiable.. 9 35 45 28 30 18 21 14 TABLE 2.—Statistical parameters of the injection sand Value Phi parameter (phi units) Remarks Mean ..................................... 2.3 Fine sand Standard deviation ............... 1.3 moderately sorted Skewness coefficient ............. 0.46 Slightly skewed (slightly poorer sorting in fine sizes) Kurtosis coefficient .............. 2.6 Moderately peaked (better sorted in the center than on the two ends) clay content. The relatively small amounts of interstitial clay present in the thin sections appeared to negate a clogging problem. However, it is now apparent that a very small percent of clay can significantly reduce hydraulic conductivity. Moreover, the injection well (IW—2) is in an area of facies change that has an increase in the percentage of silt and clay in the injection zone (table 4). The effect of the clay on the hydraulic conductivity will be discussed in detail later in the paper. The formation water in the injection zone is brackish having a dissolved-solids concentration of 3,010 mg/l (milligrams per litre). It is a sodium chloride bicarbonate type water having a chloride concentration of 1,360 mg/ 1. A detailed description of the chemical quality of the native ground water is presented later in the paper. TABLE 4.—Thiekness and sand-shale ratios for the injection zone within the well field Depth to top of Sand-shale ratio Well injection sand Thickness Total sand upper lower (ft below sea level) (ft) (percent) half half TW—l 885 104 83.0 81.0 85.0 OW—2 889 86 84.0 84.0 84.0 OW—3 891 82 83.0 76.0 90.0 lW—l 891 88 63.0 77.0 48.0 IW—2 888 84 59.5 79.0 40.5 AQUIFER TESTS Constant-rate and step-drawdown aquifer tests were made on IW—2 and TW—l. Consistent transmissivity values were not obtained from interpretations by the Theis method of analysis or by the “leaky aquifer” analysis. Slope changes indicating recharge boundary conditions TABLE 3.—X—ray diffraction analysis of core samples from injection zone, observation well 0W—3 [Results in percent] Depth1 of Mixed-layer Mon tmori 1- sample ([1) Lithology Quartz Feldspar Kaolinite Illitic-mica clays lonite Chlorite Total 892—9122 Clay 32 16 5 17 30 .......... 100 975—9952 Sand 71 6 4 6 13 .......... 100 975—9952 Sand 78 5 3 4 10 ..... 100 892—9123 Clay 20 12 13 27 ..... 18 ..... 90 892—9125 Sand 52 24 3 3 1 18 ..... 101 935—9553 Sand 67 23 1 l ..... 2 1 95 955-9753 Sand 52 19 7 7 ..... 2 2 89 ICore was recovered in 20th lengths and the lithology varies considerably within a cored interval, 2Core analysis by Wayne Hower of Halliburton Company who reported that dye staining of consolidated fragments of the core showed the kaolinite was limited to isolated “clumps" whereas the mixed-layer clays were present in the waterways, In his opinion, the core would be “quite sensitive to freshwater" (Hower, oral commun,, 1972). 3Oore analysis by Barbara Anderson, U.S. Geological Survey, Denver, Colo. 10 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA § x. Q . . it I ‘ I r: ‘6 ‘ \s‘fgl', v‘ x 9’3»! “fin-9"?” ”93 ‘ f I}, I? i’ W «WWWWigngsaw 3,51%»: é’;§s.‘~i§.v @3523'512w u” ”1’" ., “a. ,A . f’ig.‘ Ex 'l_: q? '\:"‘flffi‘“'h‘€:«::1‘“3“;‘$§V‘x’g‘ . ,9. ‘, . :IO‘Q’fi 5'3; '3’}; :5! " l‘ “I?“ .g : . . I ’ I 'V ’ q. .5 ‘ ,u 3‘) ,b “ #94. I 1938 Vaqéh‘l): ‘ua‘ifié’x: ,’ \> 3“.“ 3y 'J’gf in. ;‘r\~ wag"; 1.‘.‘\‘* 'l a». ‘;’:!} Yfiléa n13“; gig/3‘53» 1‘.;'7%2;$ g. ‘i 1. ' “ i t .1. 9' ..\» ‘ .1 ‘ 4. ,' V I, ‘ ”a :‘i ‘3‘ ,I1 *ftfih" a” ,, ‘ ‘ $ ’ wk“! 3 - “I ~ . w 3px;»: 33 W55“? 'fisfirw ' in ‘ ‘4 "~ ! " fl: .‘$j. _‘t.*. ”l? ‘1 ‘ bfim' ‘ w ‘3 U _ 1&6 3‘3] ,, is“ “I“ "t 4 i! 0.. W 3 y“ '5- '4‘ ‘3' “I “353 fim‘ ‘2? {‘i .1” ' , 1 I: y 4: “b as A <4? 0f; 9‘ 4,— M o 1" "i q' d“ “1" . I». , ~ % c 4 ‘ u i" ‘3 ,, :‘J’y’iu ‘35»? i! i ' .1 W' , i I. 1”. 1€$n~ 'Afi Qafi‘t‘"?! i-fi‘fi' >3, 3'3“? {‘9 ‘fii‘rytac‘ ': ifi I“ ‘ “3% ‘3‘ 1‘ J‘ , . ' a‘.' v; *5‘ x "i; ‘ . . ' _ }:§&!,£J »u‘¥;‘§.§;§;, sf"..- 33 aria; 3kg; 3“ 3:33, :4“ - Nb 1 3|“ ‘ . lg ‘ «t ’5-g*”v,.;r v “(I ‘ " '. ~31 u gfif:%hg¥gayr a %€\%.Zi~§?’$:ffifgf eggfié‘zxtigwgglfi : K .Q I; ‘3’» u a“ . “ .a‘éfii‘k’i‘Va-e" 356%.“ N ”fig“ 914” “ § "‘1‘?“ ‘ I? @133? M. s‘ V “$".~;’i£$’ fi'tifi'é, ‘ "I‘ .1 ""‘1w’ “I?" qw“ ; l :1 7 , ,1‘ ”5,9“: ‘3 kaQflufifrik 41‘ .5 ¢.. vi“ .-. $55 21:: ~- .a {W . . , . “'é';.$r ‘*» {3" £5 ‘ ‘. r“ ,1’ V ”a - i: fi"£&g.¢€3;u‘3‘9 { ‘ 1’ ‘ , g Q\ 'E‘V""‘i’ f» iii. ,4. § . , ‘4 ‘ . ‘2’ 31‘24533’9‘3 firs: , . iii?" a}! v- :‘rfifi “'5 ‘ ' ' . >.c‘ , We 5 . 3 35.4» Mg? 33., >' i ‘0’ i ' fi‘V '1 1 a. V, I ‘ It‘wr’x‘Q‘fi‘m ,.. 3;; .sz v‘fi“ ‘3‘.» ' ‘ i =u * 4n ‘3 t, 9- " f g'gififldffi'§‘~ ‘2‘; , .i ‘ 33w“ « > ‘ I . :$ T‘ *‘.‘J ‘ i. i d {#3:}! y. "u:. a ‘1 6 . J” ’?§‘¢’ .4» hi; «5‘ r. r". ‘ I 33:3‘" " .11? fig? 3». flu‘ §, IV“ Y “Q"; h s i “‘3 _. ’3‘ fig?! «“iv-{fiur FIGURE 7.—Thin section of part of the injection sand (955 to 975 ft or 291 to 297 m). INJECTION SYSTEM 11 begin affecting the aquifer-test data after 10 minutes of pumping and continue to affect the logarithmic plots up to about 60 minutes. Apparent recharge occurs in the plot of both drawdown and recovery data and has been inter- preted as indicating thickening of the producing aquifer or increased permeability in the aquifer away from the well field. On the basis of the aquifer-test data and an interpretation of the geology, a geologic model was made of the aquifer. This interpretation theorized a gently arcuate, concave westward, sand buildup striking north- northeast lying to the west of the well field. The eastern limit of the buildup passes close to the western edge of the well field. Using the geologic framework and the aquifer-test data, 0. J. Cosner of the Geological Survey constructed a mathe- ‘ matical model of the hydrology of the well field. Values of transmissivity ranged from as low as 5,360 ftE’d—lft—l or 498 m31d—1m'l in the injection well to 16,600 ft3d—1ft-1 or 1,540 m-"d—lm—l in the area of maximum sand buildup. In areas other than the sand buildup or the actual well field, a regional transmissivity value of 8,300 ft3d—1ft—1 or 770 de—lm—1 was used. The digital model included a 20-ft (6- m) layer of clay above the injection sand that had a vertical hydraulic conductivity of 1.41x10—5 ft/d (4.3x10—6 m/d). The specific storage of the confining clay was assumed to be 4x10‘5. The storage coefficient of the aquifer, deter- mined from tests, was 1.5x10“4. The model was verified by simulating drawdown and recovery curves obtained from a 48-hour aquifer test in which IW—2 was pumped at 800 gal/min (50 US) and TW—l was observed. The model shows that after pumping or injecting on a long-term basis (4 days or more), water levels would tend to stabilize because of leakage through the confining layers (Cosner, oral commun., 1970). No long-term aquifer tests were made to establish the validity of the model-derived hydraulic conductivity of the confining bed. Several constant-rate pumping tests and a 3-hour step- drawdown test were made of IW—2 to determine specific capacities prior to any injection (table 5). Step-drawdown TABLE 5.—Specific‘ capacities at various pumping rates after I hour for the injection sand on test well TW—I and injection well IW—Z Well Dale Screened interval Type of test Pumping rate 22:3:5 (depth In ft) (gal/min) (gal min'lfi-l) TW—l 3- 7-69 900—960 step drawdown 100 17.7 TW—l 3- 7-69 900-960 step drawdown 150 13.7 TW—l 3- 7-69 900—960 step drawdown 250 12.1 TW—l 3-11-69 900—960 constant rate 200 13.0 IW—2 5-20-70 896—976 step drawdown 250 16.2 IW—2 5-20-70 896—976 step drawdown 500 15.8 IW'—2 5-20-70 896—976 step drawdown 1000 13.4 IW—2 5-21-70 896—976 constant rate 800 14.5 tests were made to determine the well loss to be expected when withdrawing water at a rate double that of injection. Figure 8 is a hydrograph of the stepadrawdown test made of IW—2 showing how the specific capacity values were derived. The well was pumped with a 10-in (250-mm) turbine pump, and the discharge measured by 10— by 6-in and 10- by 5-in (250- by 150-mm and 250- by 130—mm) orifice plates. Water-level measurements were made during pumping and recovery in the pumped well IW—2 and in observation wells OW—2, OW’—3, and TW—l. Comparison of the specific capacities obtained during withdrawal and injection tests will be presented later in the paper. INJECTION SYSTEM All pipes and valves are constructed of schedule 80- polyvinyl chloride. All waterways in the pumps are either rubber lined or epoxy coated so at no place in the system is there iron in contact with the injection or withdrawal water, except as stainless steel. Chemically stable materials were used to prevent extraneous sources of iron masking chemical reactions occurring within the aquifer. The city water is injected by a 10-horsepower (7.46 kW) centrifugal pump rated at 500 gal/min (31.5 US) against a total dynamic head of 30 ft (9 m). It operates at 800 revolutions per minute. The pump will deliver 800 gal/ min (50.5 l/s) to free discharge at the well house which is located 230 ft (70 m) from the centrifugal-pump vault. The pump vault is situated adjacent to a 60-in (1,520-mm) concrete line from which the injection water is supplied. The centrifugal pump has a rubber liner so that the city water is not in contact with any ferrous-metal surfaces prior to injection. The injected water is withdrawn by a 30-horsepower (22 kW), 10-in (250-mm), vertical hollow-shaft turbine pump rated at 800 gal/ min (50.5 1/ 5) against a total dynamic head of 105 ft (32 m). It has 120 ft (37 m) of 8-in (200-mm) epoxy- coated pump column and a four-stage stainless-steel pump-bowl assembly. All waterways are epoxy coated. The injection flow system is so arranged that water flows in the same direction through the metering system during either injection or withdrawal. The metering system consists of a 4.8-in (l22—mm) stainless-steel orifice plate and a McCrometerl Model MC 01400 saddle meter. One-inch (25-min) diameter, thin wall, polyvinyl chloride pipes in lO-in (250-mm) lengths are cemented in the 6-in (150-mm) flow line both upstream and downstream from the orifice to act as straightening vanes and eliminate turbulent flow through the orifice. Pressure changes between the upstream and downstream sides of the orifice are recorded by a differential pressure transducer. The orifice was calibrated in the laboratory prior to installation. IThe use of trade names does not imply endorsement by the US. Geological Survey but is provided for complete description of the system. 12 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 0 I l I | I I I I I l l I l l I | I I ‘I l I I I I I I 10 — Step 1 fl —5 20 _ 0 = 250 gallons per minute PRE- PUMPING WATER LEVEL _ (15.8 litres per second) 5 a) A = 1 .48 f t .7 t “J E 30 _ Specific capacity = 16.2 gallons per minute per foot SI 5 ee (4 2 me res) __ E .2. (3.35 litres per second per metre) -10 g uj ——— —————— — a 2 40 h Step 2 T _ “I EL: 0 = 500 gallons per minute = 16.27 feet (4.96 metres) (<1) 8 (31.5 litres per second) E 50 — _-15 O D . . . w < 2WD > where SwTFSwfiexcess head buildup, in feet, due to colder water (head buildup in IW—2 for injection water of temperature T, at time t, minus head buildup in the well for the formation water at time t); K1=horizontal hydraulic conductivity of the formation to formation water, in cubic feet per minute per square foot; K1 C=horizontal hydraulic conductivity of the formation to freshwater at the injection temperature, in cubic feet per minute per square foot. chis defined as K](u/uc), where up is the viscosity of cold freshwater at injec— tion temperature and u is the viscosity The following assumptions are used in making the cal- culations: The water moves into the formation in a hor- izontal, radial pattern; it remains confined in the 80-ft (24-m) zone in which the well screen is placed; accumula- tion of water in storage within the radius r,- (t) is negligible for any time t during the test; and a definite interface exists between the colder denser injection water and the forma- tion water. If the average transmissivity based on pumping- test data between well OW—3 and IW—2 is taken as 7,075 ft~"’d—1ft‘l (657m3d—1m‘1), the following substitutions can be made in equations 3 and 5: K1=0.06l ft3min—1ft‘2; K 150041 ftsmin'lft—2, taking the water temperature of the injection water as 10°C, its viscosity as 1.3 cP (centipoises) and the viscosity of the forma~ tion water as 0.87 cP; Q=400 gal/min or 53.5 ftS/min; D=80 ft; rw=0.33 ft; 0=0.30 (the porosity of the injection sand was ob- tained from core analysis and interpretation of compensated gamma-gamma density logs. Both sources indicated an effective porosity of 35 to 40 percent. A value of 30 percent is used here, as a part of the effective porosity is always oc— cupied by essentially static water along the pore walls). After 1 minute of injecting 10°C water, the radius of the injection water from the well would be: (53.5)(1) 71,“): ———-}—-———- (3.1a)(80)(0.3) ri(t)=0.905 ft (0.28 m). Solving equation 5 gives: SW41“: 1 _ 1 _(2.3)(53.5)_10g0.905 0.0410061 (2)(3.14)(80) 0.33 Sth —Swt—0.87 ft (0.26m) of excess head buildup after 1 minute due to colder water. + (0.33)2, The negative of the drawdown curve from a pre- injection aquifer test was used as the head buildup to be expected in the wells. The calculations for excess head were added to the inverse drawdown plot in order to approximate head buildup changes with time and tem- perature. The equations were solved for excess head buildup changes with time and temperature. The equa- tions were solved for excess head buildup at various tem- peratures and times so that predicted head buildup could be approximated prior to any injection test. The dif- ference in temperature between the city water and forma- tion water may vary as much as 20°C depending upon the season. Figure 10 shows plots for IW—2 and OW—3 of the theo- retical head buildup divided by discharge and actual head buildup divided by discharge recorded during the first 270 minutes of injection test 1. The figure shows that the theoretical data approach the empirical data in the observation well, but the head build- up divided by discharge in the injection well is far in excess of the predicted values. Plots of the values for tests 2 and 3 gave similar results, but the excess head buildup in IW—2 and ASW became greater with sucCessive tests. EXCESS HEAD BUILDUP DUE TO HYDRAULIC CON DUCTIVITY DETERIORATION The field data indicate that only a part of the excess head buildup can be explained by temperature and dens- ity differences. The balance must be associated either with hydraulic conductivity deterioration in the aquifer or with entrance losses due to clogging of the screen and gravel pack. As a first approach to estimating hydraulic conductivity alteration due to clay dispersion, equation 5 may be solved for K16, the hydraulic condUctivity of the aquifer to the injected water, and observed head dif- ferences may be inserted in place of the expression 18 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 0190 I l I I I II l| I I I r I II I] l I I I I II I 092 _ 0—0—0 Head measured during Injection phase of test 1. We” IW—Q — D—D—U Predicted head from negative of drawdown, measured _ 0_150_ during a pre—injection pumping test _072 LI.I — e—h-e Correction for excess head build—up due to temperature — D ’5 ~— and viscosity differences between formation water and — Z 3 _ city water, assuming a discharge of 400 gallons per minute. _ 8 2 _ (25 litres per second) _ a E 0.100— e 0.48 E o. _ a. Z _ E 9 — ‘ I: 2‘ — MM, — .4 o _ _ n: (I E LIJ 0.050— — 0.24 m 0— | I | l I | | | I I I I | I | l I I | I l I | l I | g E I— w E u. 20-050 I llllTll I I Illllll I I IIIIII0-24g ‘- E _ d g _ Well ow 3 A < LU UJ I — — I Z 0010— — 0048.2. _ _ _ IJJ uJ o (D _ _ Z Z <( < _ _ I I U 0 — _ 0.005 0 D A Data points _0'024 _I 0001 I IIIIIII| I I I IIIIII I |lll|l0_0048 1 10 100 1000 TIME SINCE PUMPING BEGAN.|N MINUTES FIGURE 10.—Theoretical and measured head data for injection well IW—2 and observation well OW—3, test 1. (Sw Tt'Swt ). That is, the difference between the head build— up actually measured during an injection test and the theoretical head buildup taken from the drawdown measured during pumping at an equal rate may be substi- tuted for (Sth—Swt). This yields the equation: 239 Iogflfl (K1) 2WD 7w Kit: (5) 2.3Q K1 (Sim 'SPIJ'QWD 10g where S inf =the head buildup measured after a time (t) ~ of injection at the rate (Q); Sp =the drawdown measured after the same time during pumping at a rate (Q), prior to any injection; and the remaining terms are as previously defined. If no alteration has occurred in the aquifer and if entrance losses are negligible, equation 6 should yield a hydraulic conductivity equal to that calculated for the injection water from the relation K15=K1(u/uc). Deterioration of the aquifer’s hydraulic properties should be indicated by a lower hydraulic conductivity value. This approach assumes all excess head to be due to hydraulic conductiv- 7N) 7'w itv deterioration rather than to clogging of the screen. Application of equation 6 at several different time values during injection test 1 yielded a hydraulic con- ductivity of about half that calculated from the relation K1€=Kl (u/uc ). Application of the equation in later tests yielded even lower hydraulic conductivity values, but it is believed that these later results may reflect the results of screen clogging as well as hydraulic conductivity deterioration. Figure 11 shows distance-drawdown plots from two aquifer tests—one prior to any injection, and one following test 3, after removal of all injection water and extensive redevelopment to remove screen—clogging particles. Both tests were made at a discharge rate of 400 gal/ min (25 1/5). Calculation by the distance-drawdown method using the gradient between ASW and OW—S shows that the average lateral hydraulic conductivity following the three injection tests was about 50 percent of the original value. The value agrees with the estimates obtained from test 1 using equation 6. The fact that the redevelopment pumping did not restore the hydraulic conductivity to its original value illustrates the irreversible nature of the deterioration. DETERIORATION OF AQUIFER HYDRAULIC PROPERTIES 19 DISTANCE FROM PUMPING WELL, IN METRES 0 0.3 3.0 30.5 , 305 0 I lIIIlll] I IIIIIIII llllllll} IIIIIII 10— — _5 E 20— _ m “J E "" +- 3 s w” z Il—J vi 3 30— “ E Z 3 2 -10 E O 2 o o v 8 '— < 40—- — LE 2 2 g GRADIENT PER LOG CYCLE ; o WELLS . . . . . . o D Pre-Injectlon Post-Injection Q E IW—2-ASW 11.3 feet 3.4 metres 60.3 feet 18.4 metres 3 a: ASW-OW—3 8.3 feet 2.5 metres 18.3 feet 5.6 metres H5 3 D 50— OW—3—OW—3 4.4 feet 1.3 metres 7.2 feet 2.2 metres _ OW—2—TWr1 2.6 feet 0.8 metre 2.6 feet 0.8 metre 60— — —20 7O 1 IIIIIIII 1 IlllllII I lllllllI l IIIlIII 0.1 1.0 10 100 1000 DISTANCE FROM PUMPING WELL.IN FEET FIGURE ll.—-Pre-injection and post-injection hydraulic gradients. Figure 11 shows that after injection of freshwater, the hydraulic gradient steepened greatly near the pumping well and has, in fact, been modified as far as 50 ft (15 m) from the pumping well. The hydraulic gradient steepened between OW—3 and OW—2 but not between OW—2 and TW—l. This is because freshwater was injected slightly beyond the radius of OW—3 but not as far as OW—2. Every- where that the freshwater displaced the brackish water, deterioration of the aquifer hydraulic conductivity occurred. It can be shown, by plotting head buildup versus distance for each injection test, that the deterioration becomes more severe with each test and does not improve significantly with development pumping. The greatest amount of deterioration, as would be expected, is close to the injection well. CURRENT-METER TRAVERSES Figure 12 shows the flow pattern of water entering the screen during test 1 and the flow pattern during pumping before any injection tests. Minor clogging occurred in two zones; 917 to 923 ft (280 to 281 m) and 951 to 961 ft (290 to 293 m) below sea level. A decrease in flow of 14 percent in the upper zone and 15 percent in the lower zone is believed to have resulted from clogging of the injection sand by particulate matter. Agitation of the drilling mud, silt, and sand in the fill-up pipe or movement of drilling mud not removed from the filter pack by development may have been the source of the material. During the withdrawal phase, the clogged zones became productive again, and the flow pattern reverted to the pre-injection condition (fig. 13). 20 IN FEET DEPTH BELOW MEAN SEA LEVEL, 890 900 9103 920 930; 9405 950 I 960 970 ARTIFICIAL RECHARGE TO A BRACKISI-I-WATER AQUIFER, VIRGINIA I I Top of screen —280 ZONE CLOGGED ZONE CLOGGED | Bottom of schen l J 20 4O 60 80 100 PERCENTAGE OF TOTAL FLOW DEPTH BELOW MEAN SEA LEVEL, IN METRES FIGURE 12.—Current-meter traverses in injection well IW—2 of the pre-injection flow and the flow after 5 hours of injection during test 1. 890 900 910: 940 DEPTH BELOW MEAN SEA LEVEL, IN FEET 950 960 970 DETERIORATION OF AQUIFER HYDRAULIC PROPERTIES ITop of screenl\ I 275 ‘0 gallons per minut — 290 lBottom of scrleen/ l 0 20 FIGURE 13.—Current-meter traverses in injection well IW—‘2 of the pre-injection flow and the withdrawal flow 40 60 PERCENTAGE OF TOTAL FLOW during test 1. 80 100 DEPTH BELOW MEAN SEA LEVEL, lN METRES 21 22 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA During injection test 2, the flow pattern deviated from that determined by pre-injection tests (fig. 14). The change was not so much in the minor zones that clogged, but rather in a new zone that developed and began taking water. The sand from 898 to 906 ft (274 to 276 m) below sea level took nearly 20 percent of the total flow by the end of 24 hours of injecting. In all the pre-injection test 1 traverses, the sand zone from 897 to 917 ft (273 to 280 m) below sea level contributed virtually no water from the aquifer. In tests 1, 2, and 3, but especially in test 2, as pressure became greater in the injec— tion well and sand zones began clogging, the 897- to 917-ft (273- to 280-m) zone began taking water. In test 2, this zone began taking water after 35 minutes of injection and continued taking water in the lower part for the remainder of the test. The lower part of the sand zone contributed water during the withdrawal phase of test 2 and remained open throughout test 3. Examination of the drill cuttings and geophysical logs indicates that the sand zone from 897 to 917 ft (273 to 280 m) is clean, moderately sorted, porous, and permeable and should have contributed water since the initial develop- ment of the well. Probable causes for the lack of contri- bution from this zone are either clogging by drilling mud that was not removed from the gravel pack during development or by cement implaced during the construction of the well. Evidence that drilling mud remained in the gravel pack consists of glauconite grains and Miocene microfossils recovered in the turbid water pumped during the with- drawal phase of test 2. Both items are foreign to the injection zone. The pH data observed prior to any injection tests present strong evidence supporting the theory of cement influence of water in IW—2. Injection well 2 was completed in May 1970 and was not pumped between May 1970 and July 1971. Water samples gathered by a thief sampler in December 1970 had a pH of 11.7 compared with a normal formation water pH of 7.8. During subsequent test pumpings over a period of several months, the pH was never above 7.8 even when the well was idle for as long as a month. These facts indicate that the cement reaction had gone to completion before injection test 1 began. Nine current-meter traverses were made in IW—2 during the injection phase of test2 (fig. 15). It is significant to note that the plot of traverses 1 through 9 nearly parallel each other even though the injection rate decreased from a high of 400 gal/ min (25 l/s) during traverse 1 to 240 gal/ min (15 VS) during traverse 9. If the reduction in the ability of the aquifer to accept the water was caused by physical clogging with particulate matter, the flow pattern could be expected to change with time. Clogging would be most severe in the zones initially accepting the most water; deterioration of these zones would then force flow into the less permeable zones. Figure 15 supports the alternate theory that the reduction in acceptance of water is due to rather uniform reduction in hydraulic conductivity of the entire screened interval. Figures 16 and 17 illustrate the flow patterns during the withdrawal phase of test 2. The percentage of total flow from individual sections of the aquifer did not change appreciably with time, even though the flow increased from 575 to 670 gal/min (36 to 42 VS). Traverse 6 (fig. 17) more nearly matches the pre-injection flow pattern than traverse 1 (fig. 16), probably because of the increased percentage of formation water being pumped. Base exchange again occurred as the brackish formation water replaced the injected water in the formation and increased the hydraulic conductivity, although not to its original value. During injection in test 3, the flow pattern shows a drastic variation from those established during injection in tests 1 and 2 (fig. 18). The entire basal section of the aquifer, from 945 to 972 ft (288 to 296 m) below sea level, took less than 10 percent of the total recharge. In the pre- injection traverse, this same zone contributed about 32 percent of the total flow. Moreover, the zone from 923 to 945 ft (281 to 288 m) below sea level, which in the pre- injection traverse contributed only 10 percent of the total flow, took 41 percent of the flow. The flow at the time of the traverse was 70 gal/min (4.4 l/s). This plugging of selected zones and development of less permeable zones during injection conforms to the expected pattern of changes due to clogging by particulate matter.’ During test 3, the post-injection withdrawal flow pattern roughly parallels the pre-injection flow pattern. The reason the traverses do not match is that the lower section of the aquifer had failed to develop during the period of withdrawal pumping prior to making the screen traverse. The injection current-meter traverse (fig. 18) could not be made until the injection rate stabilized. The initial 100 min of injection accounted for nearly all the decrease in the rate of injection (fig. 19). By the time the injection traverse was made, the acceptance rate of the formation had nearly stabilized; particle clogging had effectively reduced the input into the more permeable zones, forcing water to enter less permeable zones, and causing cor- respondingly higher injection heads. The origin of this particulate clogging may have been fine material that accumulated at the bottom of the well during the extensive redevelopment pumping following test 2. This material may have been agitated into suspension at the start of injection in test 3 and lodged in the screen and gravel pack. Another explanation might be that material from horizons above the screen sifted down- ward through breaches in the gravel pack around the screen, musing some choking of the upper part of the screen. 890 DETERIORATION OF AQUIFER HYDRAULIC PROPERTIES Top of Scree'n\ I 900 ZONE DEVELOPED ’275 910 CLOGGED ‘280 IN FEET 920 930 e-injection test 2 withdrawal,0 406'gailohé' erminute (25. litres. .PPFSGCF’M 940 DEPTH BELOW MEAN SEA LEVEL, 950 ZONE DEVELOPED 960 ZONE CLOGGED 97o; I Bottom of s¢reen/ I l FIGURE l4.—Currem-meter traverses in injection well IW—2 of the pre-injection flow, test 2, and the flow after 24 20 4O 60 80 100 PERCENTAGE OF TOTAL FLOW hours of injection during test 2. DEPTH BELOW MEAN SEA LEVEL, |N METRES 23 DEPTH BELOW MEAN SEA LEVEL IN FEET ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 890 Top Iof screen\ I 900 910 920 Data point; 930 940 950 Traverse Time, in Discharge, lecharge, 3 . . In galJons In litres number minutes per minute per second - 1 35 400 25 400 25 375 24 350 22 960 300 19 290 18 260 16 240 15 970 3 2:1( Bottom of screen{ l T I O 40 80 1 20 1 60 200 REVOLUTIONS PER MINUTE FIGURE 15.——Current—meter traverses in injection well IW—2 during the injection phase, test 2. DEPTH BELOW MEAN SEA LEVEL, IN METRES IN FEET DEPTH BELOW MEAN SEA LEVEL. 890 900 910 920 930 940 950 960 970 DETERIORATION OF AQUIFER HYDRAULIC PROPERTIES I T0 of sore n Injection test 2 wt draw =770 gallons per minute 49 litres per second traverse Ithdrawal, 0=575 gallons per 36 litres per second | Bottom of screen/ O 20 4O 60 80 100 PERCENTAGE OF TOTAL FLOW 275 280 285 290 295 DEPTH BELOW MEAN SEA LEVEL, IN METRES FIGURE l6.—Current—meter traverses in injection well IW—2 of the pre-injection flow, and the withdrawal flow, test 2. 25 26 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 890 i Top of screen'\ I i 900 910 re—InJectlon test Withdrawa =770 gallons per minute 49 litres er second) 920 930 940 DEPTH BELOW MEAN SEA LEVEL, IN FEET 950 42 litres per second): 960 970 I Bottom of sqreen/ I i O 20 40 60 80 PERCENTAGE OF TOTAL FLOW 1OO DEPTH BELOW MEAN SEA LEVEL, IN METRES FIGURE l7.—Currem-meter traverses in injection well IW-2 of the pre-injection flow, test 2, and the flow after 75 hours of withdrawal pumping during test 2. DETERIORATION OF AQUIFER HYDRAULIC PROPERTIES 890 i I I I Top of screen\ 900 910 l M CD 0 920 ....... DEVELOPED 9 (4:4- litres per second) " (0 (:0 O ZONE DEVELOPED re-Injectlon test 3 withdrawal, O=4OO gallons per minute 25 litres per second) : CO A O ZONE DEVELOPED DEPTH BELOW MEAN SEA LEVEL, IN FEET L 950 ZONE CL OGGED -290 960 97o " I Bottom of sqreen/ [ I O 20 4O 60 80 PERCENTAGE OF TOTAL FLOW FIGURE l8.—Current-meter traverses in injection well IW-2 of the pre-injeclion, injection, and withdrawal flows, test 3. 100 DEPTH BELOW MEAN SEA LEVEL, IN METRES 27 28 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 700 I |I|||I| 650 600 —40 550 0 D A Data points 500 Test 1 450 400'— 350 Test 2 IN GALLONS PER MINUTE 300 250 Test 3 200 -10 INJECTION RATE, 150 INJECTION RATE. IN LITRES PER SECOND 100 \ 50 0 IIIIIII I IIIIIIII IIIIIIIO 10 100 1000 10.000 TIME SINCE PUMPING BEGAN, IN MINUTES FIGURE 19.—Changes in injection rate with time during tests 1, 2, and 3. SPECIFIC CAPACITY Figure 20 shows the change of injection specific capacity with time (see table 6). It illustrates, as did the decline in injection rate, the progressive deterioration of the aquifer hydraulic conductivity with successive tests. Neither development nor withdrawal pumping returned the specific capacity of the well to pre-test levels, indicating permanent deterioration of the aquifer properties. Moreover, after the first few minutes, decline in specific capacity during injection continued at nearly the same rate as established in the previous injection test. Figure 20 shows that a definite departure from the pre- injection specific capacity pattern occurred in all three injection tests. The test 2 line falls between test 1 and the pre-injection line rather than between the lines for tests 1 and 3 because the head values for the annular—space well (ASW) had to be used rather than those of the injection well due to failure of the transducers in IW—2. The slope of the line is valid, however. Figure 20 suggests that if addi- tional injection tests were made under the same condi- tions as the previous tests, further deterioration of the specific capacity of the well could be expected. CONDUCTIVITY SURVEYS During the injection phase of test 1 and the injection and withdrawal phases of test 2, conductivity traverses were made in the screen of OW—S using a downhole probe to detect the arrival and movement of freshwater. The background conductivity along the traverse in OW—3 was 4,800 micromhos on a scale of 0 to 6,000. Changes of 50 micromhos were considered significant. No apparent freshening of the water in OW—S occurred during test 1. The first definite detection of freshwater in OW—3 was during test 2 after 5,433 min (3.77 days) of injection (fig. 21). Approximately 1.8 Mgal (6,800 m3) of freshwater had been injected at that time. The freshwater first appeared at the depth interval of 895 to 899 ft (273 to 274 m) below sea level, near the top of the injection zone. Figure 21 indicates that there may have been some internal flow within OW—3 during the injection test caused by small vertical head differences in the aquifer; uncertainty regarding such internal movement complicates the interpretation of the breakthrough curves. Nevertheless, there appear to be two additional zones in which breakthrough of freshwater occurred as the DETERIORATION OF AQUIFER HYDRAULIC PROPERTIES 29 20 IIIIIII |I|II|| IIIIII 03 Pre-injection test 1 withdrawal (temperature corrected) 0) CI; ‘L A 0 D 0 A Data points .s M \w 0 Test 17\\ \=tl%’ where V=volume of water that has been injected up to time t, in cubic feet; m=thickness of aquifer, in feet; 9=porosity. (7) Equation 7 is equivalent to equation 3 except that rw is considered to be negligible, V is used in place of 30 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 860 870 — 265 f: 5433 minutes, (3.77 days) ‘- 1.81 million gallons, i=6525 minutes, (4.53 days) (I) . Lu 5 880’ (6850 CW? d 2.08 million gallons, (7870 _ . ( o: E metres) lnlecte cubic metres) injected t—6936 minutes, 4.82days) —270L—" 2 / 2.20 million gallons. (8330 E —890 cubic metres) injected —— Z —‘ 2‘26740 minutes,(4.68 days) T. DJ _J a fi 2:2: 2.12 million gallons. L; _l 900 (8020 cubic metres) UJ < lnjected ‘2754 Lu < w a Z 910 < 2 Lu CE 2 _ 920 2802 g a 9 0 Lu _J C0 930 LEE E 851 o. ‘2 E L5 940 DJ 0 950 —290 960 —> DECREASING CONDUCTIVITY FIGURE 21.—Conductivity profiles in observation well OW—3 during injection phase, test 2. TABLE 6,—Specific capacity of injection well IW-2 during the injection TABLE 6.—Specific capacity of injection well IW—2 during the injection phase of tests I, 2, and 3 phase of tests I, 2, and 3—Continued Injection rate Time since Water-lard Specific capacity Injection rate Time since Wralter-let/el Specific capacity (gal/min) injection began c 3:13:21" (gal min=lft=‘) (gal/min) injection began C 31:21" (gal min='It=‘) (min) (min) (‘0 (fl) Injection Test 1 Injection Test 3 400 6 26.0 15.4 245 20 66.7 3.7 400 10 29.5 13.6 240 32 69.4 3.5 400 20 33.1 12.1 215 40 75.6 2.8 400 30 33.9 11.8 185 50 72.9 2.5 400 40 34.5 11.6 175 60 75.8 2.3 400 50 34.8 11.5 155 67 75.8 2.0 400 60 35.5 11.3 135 73 75.8 1.8 400 90 37.0 10.8 125 77 75.8 1.6 400 120 38.7 10.3 115 90 75.9 1.5 400 160 40.0 10.0 105 120 76.2 1.4 400 180 40.0 10.0 90 250 84.0 1.1 400 220 41.0 9.8 70 900 72.9 0.96 ‘ 400 260 43.0 9.3 70 1020 75.9 0.92 ‘ 70 1100 75.4 0.93 Injection Test 21 i 400 35 28.1 14.2 ‘Change in water level measured in ASW rather than in IW-2. 400 145 32.0 12.5 3;? 1338 22'; 3'2 Q(t-tc ), and it is recognized that the thickness of aquifer, 330 1440 44:5 724 m, accepting flow may differ from the screen length, D. At the time of detection of freshwater in OW—3 during 295 2790 49.7 5.9 . . 285 3030 50.6 5.6 test 2, a total of 1,813,000 gal (6,860 m3) had been injeCted. Egg 8532(1) gill) 33 Using equation 7 and this volume, expressed in cubic feet, 215 7896 5718 3:7 and using m=80 ft (24 In), 0=0.30, 7i (t) is calculated as 56.7 DETERIORATION OF AQUIFER HYDRAULIC PROPERTIES 31 860 I I I W — 2 OW — 3 870 — 265 50 feet, (15 metres) 880 Percentage of total flow entering screen _ _ . . (0 after 4290 minutes (3.04days) of in- Conductuvuty profile after 6936 mlnutes, _27OuJ E 890 jection test 2. O=270 gallons per E (432951“) 0f Injection test 2'. 7 A E iii minute, (17 litres per second). 2'2 m'H'P'? gallons, (8330 mm": L; Z 1.48 million gallons, (5600 cubic metres) Injected. Z — 900 metres) injected. ~— — . —275 j g 32 percent [ii LU 910 7 E ’1 25 percent _J E < _ Lu (0 920 280 (0 <2): 5 percent 2 LU <1: 2 UJ 930 2 i —285 E a 6 percent 9 m 940 Lu I [D E 12 percent E ”J 950 a O 1 percent T290 3 960 16 percent 970 — 295 —> DECREASING CONDUCTIVITY 980 FIGURE 22,—Conductivity profiles in observation well OW—3 and current—meter traverse in injection well IW—2, test 2. ft (17.3 m). However, the current-meter traverses show that only about 60 ft (18 m) of the injection-well screen was taking water; if m is taken as 60 ft (18 m) rather than 80 ft (24 m), r, (t) is calculated as 65.5 ft (20 m). Again, the current-meter traverses show that 40 percent of the total inflow occurred in the sand between 900 and 920 ft (274 to 280 m) below sea level, which correlates with the intervals of early breakthrough in OW—3. If equation 7 is solved using 40 percent of the injected volume and using m=20 ft (6 m), r, (t) is calculated to be about 72 ft (22 m). Thus, the injected water should have reached OW—3, at a distance of 50 ft (15 m) from the injection well, from 1 day to more than 2 days prior to actual detection, depending upon the assumptions used. The calculations indicate that the injection front probably did not move out equally in all directions and form a cylinder. Subsequent calculations relating to the arrival time in OW—2, 75 ft (23 m) from the injection well, confirm this interpretation. A possible alternative is that the front may have had an elliptical form, as should be expected if the aquifer were homogeneous but anisotropic. However, comparison of the arrival times in OW—3 and OW—2 during test 4 rules out this possibility. It does seem clear, however, that the injection front was elongate in a direction roughly normal to the line through OW—2, OW—3, and the injection well. Trial calculations show that this elongation could not be due to superposition of a radial flow on the original hydraulic gradient in the aquifer, as this original gradient was very small. AQUIFER HETEROGENEITY A reasonable explanation of the arrival time data can be offered on the basis of the geology. The zone of greatest hydraulic conductivity in the upper part of the aquifer is probably a channel-fill or a shoestring sand. A deposit of this type would have the coarsest material along the center of the channel, with transition to progressively finer material along the sides. The average hydraulic conductivity across the channel would accordingly be lower than that along the channel axis, and injected water would tend to follow the channel axis. The channel would have a meandering orientation, but presumably its overall linea tion would be at some angle to a line through OW—2, OW—3, and the injection well (IW—2). This interpretation 32 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA 860 870 —265 880 r:10,130 minutes.(7.03 days) #211,700 minutes, (8.12 days) _ 1.84 million gallons, (6960 202 million gallons.(7650 '— _ iii "9935 minutes, (690 daVS) cubic metres) Injected cubic metres) injected ,2708 u. 1.82 million gallons, (6890 / / E Z 890Tcubic metres) inject? Lu _ 2 .J z E 900 7 w _275_. _, LL] :5 E _l a) 910 < 3 a LIJ 920 —2802 E < ; LU 9 E g 930 g I —285d E 940 cc LU I 0 E Lu 950 —ZQOD 960 ~295 970 —> DECREASING SPECIFIC CONDUCTANCE FIGURE 23.—Conductivity profiles in observation well OW—3, test 4. is supported by aquifer-test data, which indicate that the specific capacities of wells IW—2 and TW—l are higher than those of OW—3 and OW—2. WATER QUALITY COMPARISON OF THE CHEMISTRY OF CITY AND FORMATION WATER The city water is a calcium sulfate chloride type, and the formation water is a sodium chloride bicarbonate type. The major constituents are listed in table 7. The dissolved- solids concentration of the city water varied between 110 and 190 mg/l, and the formation water contained about 3,000 mg/ 1. WATER SAMPLING Prior to injecting any freshwater into the host forma- tion, consideration was given to the possibility of chemical reactions which might interfere with the injec- tion process. For example, the freshwater to be injected was generally saturated with dissolved oxygen while the formation water contained none. Thus, if high concen- trations of iron and manganese were present in the formation water, precipitation of iron or manganese TABLE 7.—Concentration of major constituents from Moore’s Bridges Filte1 Plant city water and formation water. [Values given in mg/l unless otherwise noted] Constituents Formation water City water Silica (SiOz) .................................................. 13 3.8 Calcium (Ca) ................................................ 14 17 Magnesium (Mg) .......................................... 8.7 2.6 Sodium (Na) .......... 1,140 9.5 Potassium (K) ........ 25 1.6 Bicarbonate (HCOS). 624 9.0 Sulfate (804) ................................................. 136 36 Chloride (Cl) ................................................ 1,360 21 Nitrate (N03) ........ .1 1.2 Phosphate (P04) .......................................... .28 .00 Boron (B) ...................................................... 3.4 .04 Fluoride (F).... 1.4 .l Dissolved solids.. 3,010 111 pH ................................................... 7.9 5.8 Specific conductance (micromhos) .............. 5,000 190 hydroxides could occur within the formation when the freshwater was injected, resulting in loss of hydraulic conductivity of the aquifer. Chemical analysis, however, WATER QUALITY 33 870 880 -270 890— 1:31 days _~ 7:40 days ~7_ t=41 days ___, r245 days __ 6.67 million gallons, (25.250 cubic metres) 900* injected —I 8.77 million gallons. (33.190 cubic metres) injected 910 920 930 940 950 DEPTH BELOW MEAN SEA LEVEL, IN FEET 960 9.06 million galions. (34,290 cubic metres) injected] 10 2.3 million gallons (38720 cubic metres) injected —275 I N (D 0 Ii) 93 DEPTH BELOW MEAN SEA LEVEL. IN METRES I M (D O -295 970 980 —> DECREASING SPECIFIC CONDUCTANCE FIGURE 24.——Conductivity profiles in observation well OW-2, test 4. indicated that the iron and manganese concentration in the formation water is too small (less than 0.1 mg/l to account for significant loss in hydraulic conductivity because of precipitation. Another possible chemical reaction was precipitation of calcium carbonate. However, studies carried out by Ivan Barnes (written commun., 1968) indicated that the concentration of calcium in both the freshwater and formation water was below equilibrium (saturation) values, and no precipita- tion would occur when the two waters mixed. After examining the data, it was believed that when injection occurred, the major chemical effect would be simple dilution; in which case, three chemical zones would be formed around the wellbore. These include, in order of decreasing distance from the wellbore: (1) Undiluted formation water; (2) mixed formation and freshwater; and (3) freshwater. In order to chemically define these zones during the freshwater withdrawal phase, water samples were collected for chemical analysis whenever a change in specific conductance occurred (increase in specific conductance of 200 to 500 micromhos. The range in specific conductance was approximately 200 micromhos in freshwater to 5,000 micromhos in formation water. It was also believed that in collecting samples with respect to changes in specific conductance, the resulting chemical data would indicate not only when unsuspected reactions occurred, but also which constituents were involved. Unfortunately, as will be discussed later, some reactions took place when there was little or no change in specific conductance, and the with- drawn water was still virtually fresh. ANALYTICAL RESULTS It was previously stated that the formation water is predominately a sodium chloride type containing approx- imately 1,100 and 1,400 mg/I \sodium and chloride, respectively. The freshwater contained 9.5 and 2.1 mg/l of sodium and chloride, respectively. It was assumed that as these two waters mixed a dilution would occur, and the analytical data as shown in table 8 tended to support this assumption. However, when these data concerning variations in the concentrations of sodium, chloride, and dissolved solids were plotted against specific conductance (fig. 26), the assumptions weakened. For example, the chloride values produced almost a straight-line relation- ship with respect to specific conductance, but sodium and DEPTH BELOW MEAN SEA LEVEL, IN FEET ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA DISTANCE, IN METRES 10 5 0 2'5 210 115 I I I l 110 1.5 20 830 I l l I I I I I I I I I I I I 840_ 0W~2 lW-1 Iw-2 ow—3 :255 850— — —260 860— — 870— :265 U) E 880— — E —270 E 890— — E 900 — ii —275 > L3 910— — < % —280 920— — Z < LU 930— _ 2 —285 % UJ III 950* _—290 I I— E 960 — ‘ Q —295 970 — — 980— — —300 990— — 1000— V i ——305 J I I I I | I I I I I | I I | 80 70 60 50 40 30 20 10 0 1O 20 30 40 50 60 EXPI-ANAT'ON DISTANCE, IN FEET Sand I I I I Clay Zone taking 90 percent of flow during injection test 4, (based on current-meter traverse). Zone where injected water appeared in observation wells during injection test 4 (based on conductivity profiles). FIGURE 25,—Lithologic section showing injection zone in injection well IW—2 and zones of detection of freshwater in observation wells OW—2 and OW—3. WATER QUALITY 35 SODIUM CONCENTRATION, IN MILLIEQUIVALENTS PER LITRE 0.1 0,5 1,0 5.0 10 50 10000 I | IIIII] T 5000_ 0 Data point — CHLORIDE DISSOLVED SOLIDS 1000— 500 — SPECIFIC CONDUCTANCE IN MICROMHOS AT 25 DEGREES CELSIUS IIIII I |II||TI| I IIII SODIUM —- IIIIIIII IIIIIIII 1O 50 100 CHLORIDE AND DISSOLVED SOLIDS CONCENTRATION. IN MILLIGRAMS PER LITRE 500 1000 5000 10.000 FIGURE 26,—Chloride, sodium, and dissolved solids versus specific conductance of recovered injected water from tests 1, 2, and 3. the dissolved solids did not. Sodium and dissolved-solids concentrations tended to increase at a greater rate than changes in specific conductance or chloride, which indicated that sodium ions were being evolved by chemical reaction. The dilution concentration of other constituents, such as calcium and magnesium, indicated that there was little or no relationship between variations in concentration with respect to changes in specific conductance. CHANGES IN THE CONCENTRATIONS OF CALCIUM, MAGNESIUM, AND SODIUM The concentrations of calcium and magnesium in city water were 17 and 2.6 mg/l, respectively, and in formation water averaged 14 and 8.7 mg/ 1, respectively. The calcium and magnesium values in table 8 indicate that the concen- trations of these two constituents in the mixed water were at times much lower or higher than their concentrations in either city or formation water. The water containing the low concentrations, however, was still virtually freshwater. During the later periods of withdrawal, when formation water became dominant in the mixture, the calcium and magnesium concentrations began to increase and were greater than could occur in a simple mixture. Near the end of withdrawal in test 1, the concentration of calcium was almost twice that in either city water or formation water. There appeared to be little doubt that calcium and magnesium were involved in some form of reaction as the freshwater entered the formation and that the reaction was reversible (fig. 27). If calcium and magnesium concentrations were not decreasing as a result of calcite precipitation, then the decrease must have been due to simple cation exchange with the sodium-saturated clay within the formation. When calcium in the freshwater was exchanged onto the clay, exchangeable sodium from the clay should be released into the water. Thus, there should be an excess of sodium in the water when calcium concentrations approach minimum values. Figure 27 shows that calcium from the water was being lost (sorbed onto the clays) when the water was still fresh. As the percentage of formation water increased (based upon specific conductance), the calcium was exchanged from the clay into the water. Because of the small amount of sodium exchanged 36 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA TABLE 8.—Variations in water chemistry of mixed freshwater and formation water during withdrawal, injection tests I and 2 Specific Formation conductance water Calc1um Magnesrum.Sod1um Potass1um B1carbonate Sulfate Chloride (micromhos) (pexcem)(mg/ll (mg/1) (mg/l) (mg/1) (mg/l) (mg/1) (mg/l) Freshwater 180 0.0 17 2.6 9.5 1.6 9 36 21 Injection test 1 290 1.0 6.2 48 8.4 60 34 35 370 2.0 4.0 72 8.8 73 35 49 840 11 8 0 2000 35 23' 3400 61 31 1 4600 93 23 1 400 26 272 75 500 710 35 422 100 860 2 0 2.1 2.3 160 13 129 45 170 8 0 2 l 1000 40 582 130 Formation water 5000 100 14 8.6 1200 40 618 150 1400 Freshwater 180 0.0 17 2.6 9.5 1.6 9 36 21 Injection test 2 190 0.3 10 3.2 17 2.9 8.0 36 24 245 .5 20 3 .8 20 3.7 31 42 27 360 1.8 5.6 1.3 63 8.6 67 40 46 460 3.3 5.3 1.1 85 8 9 81 40 67 1400 14 10 2.5 220 13 256 52 220 1800 28 16 4.8 360 15 228 48 410 2300 36 16 5 .5 460 17 274 66 520 2600 43 16 6.0 530 19 312 71 620 2900 51 17 6.4 590 21 352 74 720 3200 58 19 7 .2 680 23 400 69 820 3800 68 24 8.8 800 25 456 81 960 4900 96 22 1 0 1000 29 620 63 1200 Formation water 5000 100 15 8.8 1 100 29 608 69 1300 compared to the total in the water, the sodium data do not show clearly that there was excess sodium during the periods when calcium was being lost from the water. There is a slight change in the slope of the sodium curve (fig. 26) at low concentrations, indicating that the concen- tration of sodium was increasing at a greater rate than the specific conductance of the mixed waters. An indirect technique to show that cation exchange was occurring was attempted based on the assumption that (1) chloride ions do not enter into any reactions during either the injection or withdrawal of the fresh water (fig. 26) and (2) sodium was involved in only the cation-exchange reaction. The ratio of sodium to chloride in city water is 0.71 and in formation water is 1.30. The ratios obtained from all the analytical data indicated that the sodium to chloride ratio was never less than 0.71, but was frequently more than 1.30. Furthermore, when the ratio was more than 1.30, the calcium concentration was much less than that of either the city or formation water, as shown in table 9. If a comparison is made between the calcium concen- tration data and the sodium to chloride ratios shown in figure 27, it can be seen that, with respect to specific conductance, excess sodium concentrations occurred TABLE 9.—Sodiurn to chloride ratio and associated concentra- tions of calcium in samples collected during injection tests I and 2. Specific Sodium-chloride ratio Calcium conductance (m x1) (rn /l) (microthS) eq/ g Injection test I 290 2.11 6-2 370 2.27 4'0 840 1.45 8-0 2000 1.23 23 3400 1-27 31 4600 1-19 23 5000 1.30 14 Injection test 2 190 1.09 10 245 1.14 14 350 2.11 5-6 460 1.96 5-3 1380 1.54 10 1800 1.35 15 2300 1.36 16 2600 1.32 ‘6 2900 1.26 17 3200 1.11 19 3800 129 24 4900 1.29 22 5000 1-30 15 during the withdrawal periods when calcium concen- trations were approaching minimum values. EFFECT OF CATION EXCHANGE ON FORMATION CLAY DURING INJECTION It is unlikely that cation exchange during the injection phases had any effect except to slightly lower the zeta potential (tendency to disperse) of the clays. If the sodium montmorillonite-illite clay were going to disperse when subjected to freshwater, the exchange of calcium for sodium would only slightly lower this tendency. The chemical data were actually showing a solution to the problem rather than a cause of clogging. The fact that even low concentrations of calcium would exchange for sodium on the clay indicated that the clay would respond to chemical treatment. LABORATORY DETERMINATION OF HYDRAULIC CONDUCTIVITY DETERIORATION In order to substantiate the hypothesis that clay dispersion caused hydraulic conductivity deterioration in the aquifer, core samples of the injection sand taken during the drilling of OW—3 were sent to the hydrologic laboratory to determine if the aquifer was “water sensitive.” Testing procedures were similar to the techniques described by Hewitt (1963). The cores were saturated with formation water for 24 hours prior to testing. Hydraulic conductivity was determined by running formation water through the core until the values DETERMINATION OF CONDUCTIVITY DETERIORATION 37 ‘°'°°°_ I I I IIIII __ I I _ ‘3 _ End withdrawal — _ a 5000— /\ — _ 0 — Concentration of calcium ~— CoglcentratIfon 0ft" _ ‘0 In formation water SO mm In orma Ion “J water LL] —— _ — _ II (9 Lu 0 L0 — — — -—< (\l I'— <2: 8 :I: TEST1 5 (I 1000— _ :_ * 9 — _ 2 ._ _ _ _ E — Concentration_ .— LL,’ — of calcium in— — — 0 Z 500_ ,. fresh water __ _ E ‘DEFICIEN o _ _ _ _ D o z o _ _ __ _ o 9 'i- 8 _ _ __ _ % Begin WIthdrawal Concentration of sodium} in fresh water 0 D Data points I | I I I I I I l I I 1 000.1 0.5 1.0 2 o 1 2 CALCIUM CONCENTRATION, IN MILLIEOUIVALENTS PER LITRE SODIUM TO CHLORIDE FIATIO, IN MILLIEOUIVALENTS PER LITRE FIGURE 27.—Calcium concentration and sodium to chloride ratio versus specific conductance during the recovery of injected water, tests 1 and 2. stabilized. City water was then introduced into the core, displacing the formation water. Hydraulic conductivity was measured until the value stabilized (table 10). Laboratory results matched the field tests in that the hydraulic conductivity of the sand was irreversibly reduced. Reductions in hydraulic conductivity ranged from 50 percent to over 70 percent/The aquifer, on the basis of Hewitt’s 1963 classification (water permeability- Klinkenberg (gas) permeability is less than 0.3), would be classified as strongly sensitive to freshwater. Various chemicals were added to the city water to prepare a pre-flush prior to injection in an attempt to overcome the dispersion problem (table 10 and fig. 28). Sodium hydroxide and sodium carbonate were added to adjust the pH of the city water to values similar to, or greater than, that of the formation water. Deterioration of the hydraulic conductivity was not prevented by this treat- ment. Sodium hexametaphosphate, a compound used to clean wells that have had excessive invasion of drilling mud, was introduced into the core as a mixture in the city water. The compound did not prevent reduction in the hydraulic conductivity, and, because it acts as a dispersant, probably magnified the damage. The fact that a turbid effluent resulted when either untreated or chemically treated city water was introduced into the core saturated with formation water indicates dispersion was occurring. When the city water wastreated with calcium chloride, the double layer around the clay particles—and consequently the zeta potential—was reduced. N 0 reduction in hydraulic conductivity occurred, and the effluent was clear, indicating dispersion and migartion of clay did not occur in any significant amount. When a calcium chloride pre-flush was used (fig. 28, sample 73Va9), followed by untreated city water, a reduction in hydraulic conductivity of only 11 percent occurred. Reed (oral commun., 1973) has found that a 11- 38 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA SAMPLE 73Va5 SAMPLE 73Va7b I I I I Formation water. 1400 milligrams i per litre CI’1 Formation water 0 Z ; 8 I— City water, (turbid) City water + NaOH, (turbid) L O o: LU E Formation water Formation water 0 ...... ............. .. 0.0 0.05 0.01 0.15 020 0,25 0.30 0.0 0.10 0.20 0.80 0.40 0.50 0.60 0.70 HYDRAULIC CONDUCTlVlTY. lN METRES PER DAY SAMPLE 73Va9 Formation water City water + CaCi .(clear) SAMPLE 73V86 Formation water City water (clear) 0 E if) E City water + (NaP03)n. (turbid) Formation water LL O 95 3 Formation water City water, (turbid) O 0.0 0.05 0.10 0.15 0.20 0.25 0.30 0.0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 HYDRAULIC CONDUCTIVITY, IN METRES PER DAY FIGURE 28,—Effects on the hydraulic conductivity caused by injecting city water containing various chemicals into a core saturated with formation water. to 12-percent reduction in hydraulic conductivity occurs in wells treated with polymeric-hydroxy aluminum in water-flood projects and should be expected if either the calcium or aluminum treatment is used to stabilize clay. The core was resaturated with formation water, then injected with untreated city water. Clogging did occur, and the effluent was slightly turbid, indicating the calcium for sodium cation exchange is a reversible reaction. The reduction in hydraulic conductivity was not so severe as in previous tests, suggesting that the cation exchange may not be as complete when exchanging sodium for calcium at the concentrations used in this experiment. This is to be expected, as the calcium ion is held more tightly than the sodium ion. INJECTION TEST 4 39 TABLE 10.——-Effecl of water chemistry on laboratory hydraulic conductivity for core samples from infection zone of observation well OW—3 Laboratory Klinkenberg Hydraulic sample Depth permeability Water Water. Input conductivity Effluent number (ft) (millidarcy) (m/d) type modification pH (m/d) condition 73Va2a 892—912 1050 8.7x10—1 formation none 8.5 4.2x10“ clear 2a .......... city none 6.5 2.15x10-1 turbid 2a .......... formation none 8.5 2.42 x10“ clear 2b 892—912 .......... formation none 8.?) 4.98xw" Clear 2b .......... city NaOHl 8.35 2.25 x10“1 turbid 2c 892—9 1 2 .......... formation none 8.3 1.34 clear 2C .......... city N32C032 10.2 4.52 x10‘1 turbid 73 Va5 955—975 1320 l 1 formation none 7.5 2.43 x 10—1 clear 5 .......... city none 6.5 6.49 ><10‘2 turbid 5 .......... formation none 7.5 7.2x10—2 clear 73Va6 955—975 2150 l 8 formation none 7.5 2.56x10“ clear 6 .......... city none 6.5 7.76 x10-2 turbid 6 .......... city Na(P03)n3 ..... 7.48x10—2 turbid 6 .......... formation none 7.5 7.77x10‘2 clear 73Va7b 955—975 .......... formation none 7.0 6.2x10—1 clear 7b ..... city NaOH“ 7.3 2.7x10“1 turbid 7b ..... formation none 70 2.5x10“ clear 73Va9 955—975 .......... formation none ..... 2.08 xlO—l clear 9 .......... city CaC125 ..... 2.08 x10“ clear 9 ..... city none ..... 1.86 x10” clear 9 .......... formation none ..... 2.20 ><10—1 clear 9 .......... city none ..... 1.74x10—1 slightly turbid ‘Enough added to bring pH equal to or greater than 8.3. 240 mg/l added. 5100 mg/l added. INJECTION TEST 4 A fourth injection test was made to determine the effectiveness of chemically treating the clay to prevent dispersion under field conditions. Because the damage to the aquifer from clay dispersion during the first three tests severely reduced the capacity of the well to accept water, it was decided to use an inexpensive, nonpermanent, calcium chloride pre-flush treatment to stabilize the clay. As the pre—flush moves away from the well screen, the calcium ions in the solution are exchanged onto the clay replacing the sodium ions. The pre-flush becomes a sodium chloride solution with time due to addition of sodium ions exchanged by the clay. Once the calcium ions in the pre-flush solution are reduced and reach the concentration of the formation water, the effectiveness of the pre-flush is negated, and dispersion, migration, and particle plugging occur in the aquifer. Thus, a decrease in injection rate and increase in injection head buildup will occur when the freshwater enters the untreated part of the aquifer away from the well. It is neither practical nor necessary to treat the entire area that will come into contactwith the injected water. In any problem of flow toward a well, the cross-sectional area of flow decreases sharply as the well is approached, and the greatest head losses occur close to the well. This can be shown by a simple application of Darcy’s law, which states that as the cross—sectional area of flow decreases, the ‘Enough added to modify pH to between 7.0 and 8.0. 51.375 grams per litre (014 percent solution) added. hydraulic gradient must increase, other factors remaining equal. In the present problem, this leads to the conclusion that if the area immediately around the well can be treated, most of the increased head losses, due to hydraulic conductivity deterioration, can be avoided. The exact distance from the well to which the treatment should extend is controversial; however, discussions with personnel from oil field service companies indicate that the preferred radius of treatment lies within the limits of 3 to 10 ft (0.9 to 3 m) from the borehole. Injection test 4 was begun Nov. 24, 1972. A pre-flush of 3,000 gal (11 m3) of 0.2N calcium chloride was injected in front of the city water. Based on current-meter data, this volume would theoretically treat the aquifer to a radius of 8 ft (2 m) in the most permeable zones. The injection rate stabilized at 185 gal/min (12 VS) after 10 minutes and was maintained at that rate for 115 min- utes. The hydraulic gradient declined throughout this time indicating that the treatment was working effectively. After 115 minutes, the injection rate began to decline slowly, and the injection head pressure began to increase slowly. At this point, over 20,000 gal (76 m3) had been injected, and the freshwater was beyond the area of treatment. It was suspected that redevelopment pumping would increase the specific capacity of the well, but redevelop- ment pumping could not be attempted until a sufficient 40 quantity of freshwater was injected to ensure that formation water was not brought back into the vicinity of the wellbore. Because the calcium for sodium base exchange is reversible, if formation water were brought into contact with the “desensitized clay,” it would return them to a water-sensitive condition. Continuous injection of 398,000 gal (1,510 m3) was made over a 2,580-min (1.79-day) period before any redevelopment pumping was attempted. Water was then injected for periods of 11,380 min (7.9 days), 10,025 min (6.96 days), 2,495 min (1.73 days), 2,695 min (1.87 days), and 20,450 min (14.2 days) between redevelopment pumpings. After the 20,450-min injection phase, redevelopment pumping was done on a daily basis. Thirty-nine injection phases were used over a period of 95 days in order to inject a total of 20,146,100 gal (76,250 m3) of freshwater into the brackish-water aquifer. INJECTION SPECIFIC CAPACITY Figure 29 shows the injection specific capacity of IW—2 during the early part of test 4 as compared to the specific capacity measured during pre-injection and injection tests 1, 2, and 3. In the first 1,000 min, the 20 ARTIFICIAL RECHARGE TO A BRACKISI-I-WATER AQUIFER, VIRGINIA decrease in specific capacity was 51 percent in test 2, 75 percent in test 3, and only 32 percent in the initial phase of test 4. A 40-percent decrease in specific capacity occurred during the initial 260 minutes of test 1. The expected decrease in specific capacity with time, based on pre- injection aquifer test data, is about 15 percent. After the initial injection phase of test 4, the decrease in specific capacity during the first 1,000 min of injection for each new injection period following redevelopment ranged from 3 to 20 percent. The variation in percentage of decrease reflects the effectiveness of redevelopment pumping. A decrease of 11 to 12 percent per initial 1,000 min of injection is an average value for test 4. This value agrees with the decrease Reed (oral commun., 1973) found to occur in treated water-flood wells and with the decrease in laboratory hydraulic conductivity when freshwater was injected into core treated with a calcium chloride pre-flush (table 10). Figure 30 shows that each redevelopment pumping period increased the specific capacity of the injection well. After discharging the water standing in the well casing (about 3,000 gal or 11 m3), the water first pumped from the 18 IIIIIIl IIIIIIII (Temperature corrected) PRE-INJECTION TEST 1 WITHDRAWAL 0 Data point 16 [— 8 E LL L3 35 2 “L E jJ—J D. O D 0\ Z 14 W E E \ 8 I (I) E 12 E g TEST2 CL 0 TEST 1 g} jIO ‘o\o E < ~2_ O .1 Z Z _ 8 _ >’ >‘ I: l: O O E 6 E <1 <1: 0 MST 4 Phase1 _10 o 9 L: 4 T‘ M N E — O 8 TESTS Lu 1 E}; w 2 R 0 IIIIIIII I IIIIIIII 1 IIIIIIIO 10 50 100 500 1000 5000 10,000 TIME SINCE INJECTION BEGAN, IN MINUTES FIGURE 29.—Variations in specific capacity of injection well IW-2 during injection tests. INJECTION TEST 4 A: p-i 165 I |||l|l|] I II 93 o ||||l| ll Illl I 5'" A Illlll I III \ | .09 w 0 Data point I S” N I S9 _. 9° 0 N o I N 00 I .N 0) | _‘ '_I \/ PHASE 1 - 0.4 million gallons.(1510 cubic metres) injected including .003 million gallons. I "o 5 015.5 LL PRE—INJECTION WITHDRAWAL c: (Temperature corrected) quSO 0. Lu \\ l—l4.5 3 X Z 214.0 0: \R UJ 0—13.5 (I) 2 913.0 "\0 _1 {512.5 ||l||||lI lllLJllll IIIIIIII| I lllllll 35-5 lllllllll |||l|l| I lllllll] IIIIIIII >_. WPHASE 3- I376 million gallons (14 230cubIc metres) injected I: 5.0 d 2 n. < O 9 ‘i- 0 Lu 0. (0 IL \I SPECIFIC CAPACITY. lN LITRES PER SECOND PER METRE 4‘5 (11.4 CUblC metres) Of 1.4 percent CaCIz *—_09 W pre—flush ‘ 4.0 >//o/ 1\ “M '08 345 m ‘D _ //’\PHASE 2- 201 million gallons\ 0'7 3 O (7610 cubIc metres) ' Injected “‘0 “mac e06 2.5 I IILJIIII I IIIIIIIl I IIIIIIII I |l||lll I O 100 1000 10.000 100.000 TIME SINCE INJECTION BEGAN, IN MINUTES FIGURE 30.—Specific capacity of injection well IW—2 prior to any injection of freshwater and variation of specific capacity during test 4. aquifer contained sand, clay, and mica in excess of 360 mg/l. The water was heavily laden with sediment for 5 to 6 minutes during the first redevelopment pumping, but the large sediment concentration had decreased until it existed for less than a 1-minute duration after the fourth redevelopment pumping. The sediment coming from the well contained microfossils and glauconite particles that are foreign to the aquifer. This material probably represented drilling-mud invasion into the gravel pack and formation during construction of the well. The clay particles coming from the well were flocculated and probably represented material lossened during treatment by the calcium chloride. Sediment was noticed in the redevelopment discharge from IW—2 following injection test 2. Development pumping, by alternately surging and injecting water, produced large concentrations of sand prior to and after test 3. The sediment recovered during that development attempt was identical to that recovered during test 4, with the exception that the clay was in a dispersed state prior to test 4. The specific capacity of IW—2 improved with redevelop- ment pumping, and the amount of sediment discharged decreased during the redevelopment pumping of injection test 4. It was believed that the well yield could be further improved if sediment movement could be prevented completely. Coppel (oral commun., 1973) stated that treat— ment to desensitize water-sensitive aquifers has reduced or prevented sand production in wells that had a history of sand production. Reed (oral commun., 1973) suggested that the 0.2N solution of calcium chloride pre-flush was not the most efficient concentration to stabilize the clay and recommended the injection of another pre-flush using a 0.4N solution. After 4.04 Mgal (15,290 m3) had been in- jected, 3,000 gal (11 m3) of a 0.4N calcium chloride solu- tion was injected in front of untreated city water. Figure 31 shows the injection specific capacity for the calcium chloride water, and the specific capacity recorded after two subsequent redevelopment periods. The specific capacity during injection phase 5 showed a marked improvement for about 60 minutes following the treat- ment with the calcium chloride solution but then declined at a greater rate (27 percent between 60 and 1,000 min) than the 11- to 12-percent average decrease. The specific capacity for the injection phases 6 and 7 ranged from 3.3 to 5.8 gal/min—Ift“l (0.68 to 1.2 ls‘lm—l) and decreased at a 42 rate similar to that of phase 4 (fig. 30). Thus, the aquifer characteristics did improve significantly with the addi- tional treatmentbut clogged again sometime after 120 min of injection. The sand flow during the redevelopment periods diminished but did not completely stop. Figures 30 and 31 show that an optimum specific capacity value can be maintained if the injection period is limited to about 1,440 min (1 day). In IW—2, considering the altered condition of the aquifer and the sediment discharge, the most efficient method was to inject for about 1,440 min, withdraw for 30 min to clear the screen of sediment, wait 1 hour to allow the water level to approach static conditions, and then begin the next injection. Figure 32 illustrates that the injection specific capacity did not vary significantly from the time 9.05 Mgal (34,250 m3) had been injected through the time 15.87 Mgal (60,070 m3) had been injected. After injecting 16.35 Mgal (61,890 m3), the specific capacity deteriorated and redevelopment pumping could not restore it. ° CURRENT-METER TRAVERSES Flgure 33 shows the pre-injection test 4 flow pattern and the flow patterns observed during the injection phase of test 4. The injection flow pattern shows that essentially no ARTIFICIAL RECHARGE TO A BRACKISH—WATER AQUIFER, VIRGINIA water is entering the aquifer deeper than 930 ft (283 m) below sea level. This interval, based on the pre-injection test-4 traverse, should have been taking about 45 percent of the total flow. The flow pattern suggests particulate clogging in the screen and gravel pack rather than the uniform reduction in flow percentages that occurred when clay dispersion affected the hydraulic conductivity of the aquifer as shown in tests 1 and 2. The withdrawal current- meter traverses (fig. 34) show that the flow pattern returns to the pre-injection test 4 pattern, indicating that the sediment lodged in the screen was removed during the initial surge of withdrawal pumping. Within the zones taking water, little clogging occurred during the injection of the first 15 Mgal (56,800 m3) of freshwater (fig. 33). The current-meter traverses show that in the zone taking the highest percentage of the water at the start of the test (903 to 915 ft or 275 to 279 m below sea level) only 3 ft (0.9 m) at 912 to 915 ft (278 to 279 m) below sea level had clogged after nearly 63 days of injecting. During the first 63 days of injecting, the specific capacity of the well had remained rather uniform. When the specific capacity of IW—2 began declining rapidly (after 16 Mgal or 60,600 m5 had been injected), it was suspected that 8 I I I I I | | | I I I I I | I I | | I l I I I | I I I I I | I | I | I —1.6 0 Data point 5 —1.5 g 9. 7 A E tr PHASE 5 — 4.65 million gallons, (17,600 cubic metres) —1.4 2 E injected including 3000 gallons, (11 cubic 0: Lu metres) of 2.8 percent CaCI2 pre-flush —1.3 E *— 3 6 / e g PHASE 6—520 million gallons, "1'20 (19,680 cubic metres) 8 E injected ~1.1 “3 o. [I (I) 5 E Z FLO 9 e _l ~0.9 C: < l: (D 4 __l E —0.8 2 >’ PHASE 7-8.76 million gallons, ; 5 (33,160 cubic metres) ”07 t injected 0 < < 3: 3 ~0.6 it 0 0 9 —o.5 9 E E 8 2 —o.4 8 a. u. U) (I) ‘0.3 1 | lllllllI l lllllllI l lllllllI lllllll 10 100 1000 10,000 100,000 TIME SINCE INJECTION BEGAN, lN MINUTES FIGURE 31,—Variation of specific capacity in injection well IW—2 for injection phases 5, 6, and 7, test 4. INJECTION TEST 4 43 \l O) ,4 PH 3 l PHAS 20 01 A (A) 0 Data point SPECIFIC CAPACITY, IN GALLONS PER MINUTE PER FOOT l\) 10 50 100 TIME SINCE INJECTION BEGAN, IN MINUTE 2:) WATER PHASE Million gaHons 8 9.05 9.71 20 13.76 23 15.87 27 16.85 28 1 7.32 39 20,15 INJECTED Cubic metres —\ 01 34,250 36,750 52,080 60,070 61 .880 65,560 76,270 9 .o .o .o .o e f e f c 01 O) \1 CO to O —‘ N (A) J> SPECIFIC CAPACITY, IN LITRES PER SECOND PER METRE .0 J; .0 w 500 1000 5000 10.000 FIGURE 32.—Variation of specific capacity in injection well IW—2 for injection phases, test 4. clogging was occurring within the aquifer above the depth of 930 ft (283 m) below sea level. The current meter was inoperative at this time, however, and traverses could not be made to confirm the suspicion. Prior to treatment of the aquifer with the calcium chloride, the zone from 904 to 922 ft (276 to 281 m) below sea level accepted a maximum of 50 percent of the water injected. After treatment with a calcium chloride solution, the zone took between 80 and 90 percent of the water injected. This fact suggests that the treatment may have operated preferentially in this zone. It was the most permeable zone and would accept most of the calcium chloride solution and show the greatest improvement in hydraulic conductivity. The preferential treatment of this zone may have taken so much of tahe calcium chloride solution that only a token amount reached the aquifer below 922 ft (281 m). Therefore, when the pre-flush was followed by freshwater, dispersion and clogging by sand and clay particles occurred below 922 ft (281 m). CAUSE or CLOGGING or 1w-2 DURING INJECTION PHASE OF TEST 4 The quantity of sediment produced during the redevelopment phases up through 16.35 Mgal (61,880 m3) injected in test 4 usually diminished after about 1 or 2 minutes of redevelopment pumping. After injecting 16.35 Mgal (61,880 m3) of freshwater, the heavy discharge of sediment obtained during the daily redevelopment lasted for 8 to 10 minutes. Not only had the quantity of sediment increased, but the character of the sediment also changed. It consisted predominately of clay granules with some colloidal clay and silt to fine—grained particles of quartz and mica. The granules consist of flocculated clay and silt. The colloidal clay remained in suspension for several hours in undisturbed water but not for days, as it had before treatment. A possible cause for the large amount of sediment discharge and the subsequent decrease in specific capacity may have been due to disturbance of the gravel pack, so that its effectiveness as a filter was decreased. During the redevelopment cycle, a combination of injection and with- drawal pumping was employed; this practice tends to agitate the gravel pack of the well. It was after this vigorous redevelopment pumping that the sediment dis- charge during withdrawal became noticeably larger and lasted up to 15 minutes before diminishing. The appearance of some dispersed clay and the overall DEPTH BELOW MEAN SEA LEVEL, lN FEET 890 900 920 930 950 960 970 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA Top of screeii | I ZONE DEVELOPED 4.8 million gallons Injected 0=‘171 gallons per minut ‘ ('11 litres per second ZONE CLOGGED 2.01 million gallons injected 0=223 gallons per minute (14 litres per second) - 280 - 285 Pre-injection withdrawal, test 4 O=560 gallons per minute (35 litres per second) ZONE CLOGGED - 290 0 D A Data points *295 Bottom of Screen/ | 20 4O 60 80 PERCENTAGE OF TOTAL FLOW 100 FIGURE 33.—Current-meter traverses in injection well IW—2 of pre-injection and injection flow, test 4. DEPTH BELOW MEAN SEA LEVEL, lN METRES DEPTH BELOW MEAN SEA LEVEL, IN FEET 890 900 910 920 930 940 950 960 970 INJECTION TEST 4 1 Top of screen\ N on O N oo 01 I Bottom of sclreen/ l 2o ' 4o 60 80 100 PERCENTAGE OF TOTAL FLOW DEPTH BELOW MEAN SEA LEVEL, IN METRES 290 295 FIGURE 34.—Current-meter traverses in injection well IW—2 of pre-injection, injection, and withdrawal flow, test 4. 45 46 ARTIFICIAL RECHARGE TO A BRACKISH—WATER AQUIFER, VIRGINIA decrease in size of the sediment indicated that the sediment may have been coming from the aquifer below 922 ft (281 m) below sea level, which did not receive effective calcium chloride treatment. During redevelopment pumping, the higher pumping rates produced appreciable water from the lower part of the aquifer with enough velocity to move clay and silt into the borehole. As pumping continued and the rate fell off, the velocity of the water coming from the lower section was not sufficient to move the sediment into the borehole. The sand trap below the screen may have been the source of most of the sediment that caused particulate clogging. As injection began, after each phase of redevelopment pumping in test 4, the surge of water into the wellbore and screen could agitate the material into suspension, so that injection of sand and clay particles into the screen and gravel pack could take place. During redevelopment, the currerxt meter indicated water movement in the bottom 40 ft (12 m) of the well, but during injection it indicated no water movement even though the pumping rates were comparable. The phenomenon may occur because the screen is designed for withdrawal pumping. The openings of the screen are V-shaped with the open end of the V facing inward. During normal withdrawal pumping, particles passing through the smallest opening enter the well and are removed, provided the well is pumping at sufficient rates to keep the sand in suspension. However, if the sediment clogging the well is coming from agitation of an internal source (such as the fill-up pipe) while injection is occurring, the sand would lodge in the screen and block the flow of water. Upon withdrawal pumping, the shape of the openings would allow easy removal of the material and open the screen to flow from the aquifer. Figure 35 shows that below 950 ft (290 m) below sea level, the screen in IW—2 contributed a larger percentage of the total flow as withdrawal pumping progressed, indi- cating that development occurred within the zone. The flow from 946 to 960 ft (288 to 293 m) increased from 23 percent to 36 percent of the total flow after 26.3 Mgal (99,550 m3) had been withdrawn. At the same time, the percentage of total discharge from 910 to 922 ft (277 to 281 m) below sea level decreased from 41 to 13 percent. Both the chemical and hydraulic data suggest that the clay in the gravel pack in the 910- to 922-ft (277- to 281- m) zone was exposed to the brackish water, which resulted in repacking, which, in turn, reduced the flow from that zone. CHEMICAL EFFECTS OBSERVED DURING WITHDRAWAL PHASE OF TEST 4 Chemically, the following sequence of events were predicted during the withdrawal phase of test 4: (1) The specific conductance of the first water withdrawn would be about equal to the input value; (2) the specific conduc- tance would remain below 1,000 micromhos until approx- imately 17 Mgal (64,300 m3) was withdrawn; and (3) the calcium concentration in the repumped water would remain at 15 to 17 mg/l, assuming that most of the calcium in the calcium chloride solution was exchanged by the clay. There would be some increase in the concentration of sodium at the outer edges of the freshwater zone, corresponding to periods in which the calcium chloride was injected (calcium for sodium exchange). The calcium concentration in the repumped water would not increase above 17 mg/l until the concentration of sodium in the mixed freshwater and formation water rose to between 600 and 700 mg/ 1. At that point, a reverse exchange would occur, which would indicate the return to a high percentage mixture of native water, including some exchanged sodium. During this later period of pumping, the calcium concentration would remain relatively high until most of the exchanged calcium had been replaced by sodium from the formation water. These events were predicted on the basis of data observed during withdrawal phases of tests 1 and 2. The authors were aware that most of the water was injected selectively in the interval 910 to 930 ft (277 to 283 m) below sea level. It was expected that the stabilization of the clay in that zone, combined with the clogging in the bottom of the aquifer, would cause the withdrawal flow pattern to be the same as the injection flow pattern. However, this did not occur during test 4 withdrawal, and the flow pattern reverted to the pre-injection flow pattern (fig. 35), causing many of the predicted chemical reactions to be masked. After 8 Mgal (30,300 m3) of water had been withdrawn, the zone that had taken 80 percent of the water during injection (904 to 915 ft (276 to 279 m) below sea level) was yielding less than 10 percent (fig. 36). Further, it can be seen that zones that took little or no freshwater began to develop and were yielding formation water. As pumping continued and after 17 Mgal (64,300 m3) of water had been withdrawn, more than 50 percent of the water coming out of the well was formation water (fig. 37). Based on the chloride concentration, less than 20 percent of the water recovered was potable (US. Department of Health, Education, and Welfare, 1962) (fig. 38). The net result of this differential yield between injected and withdrawn freshwater was that predicted chemical observations were obscured by the preponderance of formation water. The specific conductance was above 1,000 micromhos after only 4.2 Mgal (15,900 m3) was with- drawn instead of 17 Mgal (64,300 m3), as was predicted. The calcium content did show some deficiency in the first water withdrawn, indicating some base exchange still occurred in test 4. The deficiency of calcium in test 4 was minor compared to the calcium deficiency recorded in the first three tests, and was further evidence that treatment of DEPTH BELOW MEAN SEA LEVEL, IN FEET 890 900 910 920 930 . 960 970 INJECTION TEST 4 '0 560 gallons per minute M on O M CD 01 DEPTH BELOW MEAN SEA LEVEL, IN METRES 290 l Bottom of sqreen/w ‘ 2O 4O 60 8O PERCENTAGE OF TOTAL FLOW FIGURE 35.—Currem-meter traverses in injection well IW—2 of withdrawal flow, test 4. 100 47 48 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA Injection current meter traverse in lW—2. lW-2. Conductivity profile in Withdrawal current meter traverse in lW-2, 14.8 million gallons, (56.000 8.0 million gallons, (30,300 Water 8.4 million gallons, (31,800 890cubic metres) injected cubic metres) withdrawn samples cubic metres) withdrawn I l I T l l l ':'§-E .| (LG ._ ., . LU 900 _ i —275 E ”" uJ “' 2 3 910-, 7 3 LT]; _i E — — 0 —280 g _, 920 _ “j E 3: . < (j) :v:v. .- Lu 2 930 — a) g _ __ —285 DECREASING CONDUCTIVITY 0 20 4O 60 80 PERCENTAGE OF TOTAL FLOW WATER SAMPLES COLLECTED BY THlEF SAMPLER 5—30-73 Deptgeseleo/vgl mean Chloride Specific. conductance Feet Metres (mg/l) (Mlcromhos) 9'02 275 880 3800 909 277 880 3700 932 284 1200 4900 962 293 1400 5800 FIGURE 37.—Changes in injection and withdrawal flow and conductivity profile in injection well IW—2 after 15.] Mgal (57,150 m3) withdrawn, test 4. of dissolved oxygen is an immediate reaction during injection but has little effect on the potability of the water. Although plugging by gas bubbles is a possibility, satura- tion and head data suggest that the hydraulic properties of the aquifer were not noticeably affected by the loss of dissolved oxygen. ANALYSIS OF PROJECT The Norfolk injection project has demonstrated that the sand aquifers containing saline water in the Norfolk area, and quite likely throughout the Coastal Plain, are water sensitive and must be treated as such if they are to be recharged with freshwater. Moreover, if physical clogging of the screen of the injection well can be prevented during injection of freshwater, the percentage of recoverable potable water is sufficient to make the proposal of under- ground storage and retrieval of freshwater from a brackish aquifer feasible. During injection tests 1 and 2, 65 percent of the water recovered was within Public Health Service standards, and as much as 85 percent of the mixed water recovered could be used if necessary. During these two tests, the clogging of the aquifer, due to clay dispersion, caused a uniform reduction in aquifer hydraulic conductivity. The first three injection tests were conducted prior to identifying the water-sensitive nature of the aquifer. Con- sequently, deterioration of aquifer properties as a result of injecting freshwater was irreversible, and original condi- 50 ARTIFICIAL RECHARGE TO A BRACKISI-I-WATER AQUIFER, VIRGINIA VOLUME WITHDRAWN, IN CUBIC METRES 378.5 3785 378,500 1400 IllIIIII I I 1800— 1200— 1100-— 1000—- 900— 800— 0 Data point 700 600 500 400 300 200 CHLORIDE CONCENTRATION, IN MILLIGRAMS PER LITRE 100 IIIIIII II IIII ||I||ILL 0,1 VOLUME WITHDRAWN. 1 IN MILLIONS OF GALLONS 10 100 FIGURE 38.—Changes in chloride concentration with volume of water withdrawn, test 4. tions could not be restored. Treatment of the clay to prevent further deterioration of the aquifer hydraulic properties was possible, as evidenced by the injection phase of test 4. However, the aquifer deterioration had progressed to the point that physical clogging of the screen and the gravel pack was the dominant factor in the injection process. Current-meter traverses made during injection and withdrawal in tests 1 and 2 showed that zones taking water during injection gave up water during withdrawal in approximately the same percentage of the total discharge. That is, a zone that took 10 percent of the water injected also yielded 10 percent of the water during withdrawal. This demonstrates that the injection front was pre- dictable, provided that clogging did not occur and cause variations in flow patterns between injection and withdrawal. The importance of preventing clogging can be seen from the low recovery percentages of potable water during tests 3 and 4. Because of the water sensitivity factor, the deterioration of the aquifer had progressed to the extent that clogging by sand and clay particles became the dominant factor in the injection process. Current-meter traverses showed that the injection flow patterns had changed drastically from those of tests 1 and 2. Zones that had taken water during injection tests 1 and 2 clogged completely during tests 3 and 4, resulting in localized rather than uniform injection of freshwater. Current-meter traverses made during withdrawal pumping are consistent and similar to those made prior to injection of any freshwater. Zones that clog during injection become productive during withdrawal and produce nearly the same percentage of the total discharge from the well as was produced during pre—injection conditions. As a result, when withdrawal pumping commenced in tests 3 and 4, flow came not only from the zones that took freshwater but from the previously clogged zones that contained formation water. Consequently, only 20 percent of the water recovered in test 4 was potable. Sand production during withdrawal and development pumping in IW-2 after test 2 was coincident with the particulate clogging on the screen and in the gravel pack. ANALYSIS OF PROJECT 51 It is unclear, at this time, whether clay dispersion and repacking within the aquifer allowed the gravel pack to settle below the top of the screen, thereby allowing move- ment of aquifer sand into the screen. The sand produced during redevelopment pumping contains glauconite and Miocene fossil material foreign to the aquifer. There is some evidence that the high injection heads during test 3 may have caused a channel along the boundary of the gravel pack and well screen so that sand and clay from an overlying formation could move downward through the channel into the screen. The younger material is probably best explained by incomplete development pumping during the construction of the well, resulting in drilling mud being left in the gravel pack. The majority of sand production, on the basis of examination of the screen by down-hole television during test 4 withdrawal, appears to be coming from a breach in the gravel pack below a depth of 972 ft (296 m) below sea level. The obvious question is whether the sand production and resulting clogging could have been prevented if the formation had been treated with clay stabilizers prior to injection of any freshwater. The question cannot be answered definitely unless a new well is drilled. Based on published data and observations made through long exposure to the problems of the injection project, we believe that physical clogging can be minimized provided proper well construction and formation treatment are utilized. X—ray and thin sections of core samples indicate that the consolidation factor is very low and that the bonding agent is essentially the clay surrounding the quartz grains. If dispersion of the clay is prevented prior to injection of freshwater by treatment with trivalent cations, such as aluminum, clogging by sand should be a minor problem in a properly developed well. The chemical quality of water recovered indicated that the injection and retrieval of freshwater from an aquifer containing brackish water is entirely feasible. During this study, only 15 percent of the injected water was considered not potable according to US. Public Health Service standards (chloride more than 250 mg/l). Further, adverse chemical reactions (deterioration in water quality or large scale chemical precipitation) did not occur as a result of injection and withdrawal of the freshwater from the brackish-water sand. The freshwater injected contained no coliform bacteria, and biological contamination was not found in the water withdrawn. Chemical measurements, such as those for pH and dissolved oxygen, indicated that reactions other than simple silution were occurring when the water was injected. The pH effect was eliminated by pre-treatment of the formation clay with calcium chloride; however, the cause of the loss of dissolved oxygen is unclear at this time. Laboratory experiments (Kimbler, Kazmann, and Whitehead, 1973) have suggested that the density differ- ence between freshwater and saltwater would cause verti- cal movement, which could affect the recovery of the injected freshwater. The conductivity profiles and current- meter traverses made in this investigation suggest that under injection conditions the flow was essentially hori- zontal. Natural stratification within the aquifer may pre- vent upward migration of freshwater, but not enough evidence is available, especially regarding long-term storage under static conditions, to warrant any conclu- sions. Also, the effect of injection on the natural move- ment of water could not be fully evaluated. The injected water was not left in residence long enough and the observation-well network was not dense enough to observe the effect, if any,.injection has on the movement of formation water. An injection well field at Moore’s Bridges Filter Plant, ideally, would have at least 4 wells; each capable of injecting 1,000 gal/min (63 Us) and withdrawing 2,000 gal/min (126 l/s). Injection would be continuous until a sufficient quantity had been injected so that withdrawal demands would not remove more than 60 percent of the injected freshwater. Invasion of drilling fluids into the sediments usually occurs during hydraulic rotary drilling. In a withdrawal well this is normally not a problem, but in an injection well, where clay dispersion is a possibility and prevention of clogging is essential, invasion of drilling fluids can jeopardize the life of the well. Therefore, to minimize the invasion of drilling mud into the aquifer, a proposed injection zone could be drilled using a hydraulic reverse rotary method. The Norfolk study has shown that clogging of the formation by iron precipitation is not sufficient to prevent the use of steel casing, although a stainless-steel screen would probably be required because of the brackish water. If a different aquifer or well field location were chosen, then new calculations of the iron reactions would have to be made. The advantages or disadvantages of a gravel pack have been widely published for withdrawal wells, but, because of the serious problems that can be caused by clogging in the gravel pack of an injection well, natural completion (no gravel pack) using a wire-wrapped screen may help to minimize clogging problems. A screen with a round or square wire, rather than a V—shape, may reduce internal clogging. To effectively inject 1,000 gal/min (63 VS) and with- draw 2,000 gal/min (126 l/s), the specific capacity of an injection well would be in the range of 20 to 30 gal min—1h—l or 4.1 to 6.2ls‘1m". In the Norfolk area, in order to obtain a specific capacity within that range at the proposed pumping rates, more screened aquifer is required than was used in IW—2. Test drilling in the vicinity of IW—2 has shown that the sand from approximately 750 to 1,000 ft (229 to 305 m) below land surface and the chemical quality of the water in this interval are rather uniform. Using electric-10g and core- sample data to determine screen placement, this entire 52 ARTIFICIAL RECHARGE TO A BRACKISH-WATER AQUIFER, VIRGINIA section could be selectively screened in order to obtain the desired specific capacity. Immediately after setting the screen, development pumping exceeding maximum operational rate would continue until sand pumping ceased. At this time treat- ment of the formation to prevent clay dispersion would begin—before freshwater is introduced into the well. The detailed method of treatment is available from several oil field service companies. The design of an injection well would be more efficient than that of IW—2 if the same size casing extended from the surface to the top of the screen. This would allow the installation of additional pump column and provide for greater drawdown. With such construction, the screen may be adjusted or even removed and replaced if necessary. The design also could provide access for a current meter in order that changes in the injection flow pattern from that of the pre—injection pattern could be detected. Withdrawal pumps on the first injection well would be capable of pumping at least 2,000 gal/min (126 l/s) against a total dynamic head of 300 ft (91 m). Aquifer-test data could then be used to refine pump requirements for subsequent wells. The only special construction require- ment for a withdrawal pump would be to have stainless- steel line-shaft, epoxy—coated column pipe, and an all- bronze bowl assembly. Each well would be equipped with its own injection pump located at the freshwater collection point. The injection pump would be capable of pumping more than 1,000 gal/ min (63 1/ 5) against a total dynamic head of 120 ft (37 m). The water injected, ideally, would contain the least possible amount of particulate matter. Spacing in a well field is normally based on hydraulic characteristics of the aquifer in that the wells are spaced far enough apart to prevent serious hydraulic interference when several wells are pumping. In an injection well field, Storage Tanks (26 metres} 50 feet TW—l W 5 i O - Possible well location Pumping M Station ii I FIGURE 39.—Possible well locations for a five-well injection field at Moore’s Bridges Filter Plant. REFERENCES CITED 53 however, two additional considerations are of equal importance: (1) The distance from the source for injection water, and (2) the shape of the injection front. Injection tests 2 and 4 demonstrated that the spread of water from the injection well is non-radial and is elongated roughly in a northwest direction. Welis placed perpendicular to the long axis of the injection front in the shape of an expanded W, with a minimum of 600 ft (183 m) between the wells, would make efficient use of space available (at the Filter Plant for instance) and the hydraulic properties of the aquifer could be achieved (fig. 39). This spacing and well configuration would only apply to the area under investigation. If a well field were to be located in a different area, an analysis of the hydraulic properties of the aquifer at that location would be necessary. Using this spacing, and assuming an aquifer thickness of 150 ft (46 m), approximately 100 Mgal (378,000 m3) of freshwater could be stored by each well in the area of the Filter Plant before well interference became serious. REFERENCES CITED Baptist, 0. C., and Sweeney, 8. A., 1955, Effects of clays on the permea- bility of reservoir sands to various saline waters, Wyoming: Bureau of Mines Rept. Inv. 5180, 23 p. 1957, Physical properties and behavior of the Newcastle oil- reservoir sand, Weston County, Wyo: Bureau of Mines Rept. Inv. 5331, 43 p. Brown, D. L, 1971, Techniques for quality of water interpretations from calibrated geophysical logs, Atlantic Coastal area: Ground Water, July»August, v. 9, no. 4, 13 p. Brown, D. L., and Silvey, W. D., 1973. Underground storage and re— trieval of fresh water from a brackish-water aquifer: in Second International Symposium on Underground Waste Management and Artificial Recharge, New Orleans, La. Reprints, vol. 1, p. 379~419, Am. Assoc. Petroleum Geologists. Cederstrom, D. J., 1957, Geology and ground-water resources of the York-James peninsula, Virginia: U.S. Geo]. Survey Water—Supply Paper'1361, 237 p. Coppel, C. P., Jennings, H. Y., and Reed, M. G., 1972, Field results from wells treated with hydroxy-aluminum: SPE-AIME, San Antonio, Texas. ‘ Gray, D. H., and Rex, R. W., 1966, Formation damage in sandstones caused by clay dispersion and migration: Clays and Clay Minerals, 14th Nat. Cont, Pergamon Press, London, 10 p. Hewitt, C. H., 1963, Analytical techniques for recognizing water sensitive reservoir rocks: Jour. Petrol. Tech., August, p. 813—818. Jones, T. 0., 1964, Influence of chemical composition of water on clay blocking of permeability: Jour. Petrol. Tech., No. 16, p. 441—446. Kimbler, O. 14., Kazmann, R. G.. and Whitehead, W. R., 1973, Saline aquifers—future storage reservoirs for fresh water?: Underground Waste Management and Artificial Recharge, 2nd International Symposium, vol. 2, AAPG, New Orleans, La. Land, C. S., and Baptist, 0. C., 1965, Effect of hydration of mont- morillonite on the permeability to gas of water-sensitive reservoir rocks: Petrol. Transactions, Oct., p. 1213—1218. Meade, R. H., 1964, Removal of water and rearrangement of particles during the compaction of clayey sediments—Review: in Mechan- ics of Aquifer Systems, U.S. Geo]. Survey Prof. Paper 497—3, p. B2—B23. Reed, M. G., 1972, Stabilization of formation clays with hydroxy- aluminum solutions: Jour. Petrol. Tech. July. U.S. Department of Health, Education, and Welfare, 1962, Drinking water standards, 1962: Public Health Service Pub. 956, 61 p. White, E. J., Baptist, 0. C., and Land, C. S., 1962, Physical properties and clay mineral contents affecting susceptibility of oil sands to water damage, Powder River Basin, Wyoming: Bureau of Mines Rept. Inv. 6093, 20 p. 1964, Formation damage estimated from water sensitive tests, Patrick Draw Area, Wyoming: Bureau of Mines Rept. Inv. 6520, 20 p. 7 DAY Mineral Resource Perspectives 1975 GEOLOGICAL SURVEY PROFESSIONAL PAPER 940 _ @1273’ ~ 7% v. ~7st v Mineral Resource Perspectives 1975 GEOLOGICAL SURVEY PROFESSIONAL PAPER“ 940 UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1975 UNITED STATES DEPARTMENT OF THE INTERIOR ROGERS C. B. MORTON, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director Library of Congress Cataloging in Publication Data United States. Geological Survey. Mineral resource perspectives 1975. (Geological Survey professional paper; 940) Bibliography: p. Supt. of Doc. no.: I 19.161940. 1. Mines and mineral resources--United States. I. Title. II. Series: United States. Geological Survey. paper; 940. TN23.U74 1975 553’.0973 75-619150 Professional For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-02673-4 F. PREFACE The Mining and Minerals Policy Act of 1970, the Club of Rome’s report “The Limits of Growth” (Meadows and others, 1972), and the U.S. National Commission on Materials Policy (1973) focused public at— tention on the Nation’s dependence on mineral resources, and reports such as “United States Mineral Resources” (U.S. Geological Survey Profession- al Paper 820) and “Mineral Resources and the environment” (National Academy of Sciences, 1975) have broadened that concern. The problems of access to raw materials, though often minimized, are complex issues re- quiring factual data and careful analysis. This report, which supplements the annual report of the Secretary of the Interior under the Mining and Minerals Policy Act of 1970 (PL. 91—631), provides current Geological Survey perspectives on the status of mineral resources research on other than energy materials, which will be considered elsewhere, and discusses progress toward a better understanding of the problems of minerals availability. The U.S. Bureau of Mines (1974b) has prepared a companion appraisal of mineral reserves, “Commodity Data Summaries 1974.” The task of evaluating mineral resources is a continuing one. Al- though mineral resources are not renewable—that is, they cannot be re- generated or replaced at rates comparable with their extraction—mineral reserves can be created from subeconomic resources or new discoveries, but the creation process is possible only through a concerted, well-planned, time-consuming effort by industry, academia, and Government. This re- port considers some of these problems and describes some current re- search by the U.S. Geological Survey that is directed toward understand- ing the problems of access to mineral resources and the methods of their appraisal and expiration. U. {h lw (aflwj/ Director U.S. Geological Survey .- Preface ______________________________________ Introduction __________________________________ What is our minerals problem? __________________ Classifying mineral resources _________________ __ Appraising mineral resources ___________________ Methods __________________________________ Organizing data ___________________________ The resource outlook ___________________________ Exploration for mineral resources _______________ Techniques for mineral exploration __________ Lead time ________________________________ Chances of success ________________________ Recent exploration activity _________________ Recent research on mining and extraction ____ Roles of industry and government in research ____ Mineral resources research programs in the U.S. Geological Survey ___________________ Regional appraisal _________________________ Areal geologic studies ______________________ CONTENTS oomeoooqqosmmmNHI—A n—u—t 13 14 Mineral resources research programs—Continued Detailed studies in mining districts __________ Geochemical exploration ____________________ Geophysical exploration ____________________ Resource analysis __________________________ Mineral resource studies by specialists ______ State reports requested by the U.S. Senate _-__ Mineral appraisal of Federal lands __________ Conclusions: Some pressing problems ____________ Determining criticalness of mineral commodities Depending on imported minerals ____________ State of the mining industry ________________ Reporting reserve data _____________________ Recovering potential mineral byproducts ______ Considering subsea mineral resources ________ Staffing for mineral appraisal and exploration _ Increasing public awareness ________________ References cited _______________________________ ILLUSTRATIONS FIGURE 1. Classification of mineral resources ______________________________________________________________ 2. Organizations in the Geological Survey having program responsibilities for mineral resources research TABLE 1. 2. 3. 4. 5. 6. ~ 7. 8. TABLES United States’ dependence on foreign sources for some of its minerals ____________________________ General outlook for domestic reserves and resources through 2000 AD. ____________________________ Foreign exploration—development data reported in 1973 ____________________________________________ Commodities in the mineral-resource—specialists program __________________________________________ Federally owned land in 11 western States and Alaska as of June 30, 1971 ________________________ Major landholding agencies in the Federal Government ___________________________________________ Three levels of resource appraisal used by the U.S. Geological Survey _____________________________ Mineral resource appraisal of Federal lands completed to September 1974 __________________________ Page 12 Page 11 17 17 17 19 MINERAL RESOURCE PERSPECTIVES 1975 INTRODUCTION Energy materials are so much in the limelight to- day that little attention has been focused on the equally important but less obvious problem of avail- ability of other minerals for an expanding world in- dustrial complex, or indeed for the increased produc- tion of energy. This report summarizes the status of mineral resources and mineral exploration in the United States and mineral resources research in the US. Geological Survey in 1974. It is intended for con- cerned citizens, scientists, planners, and policy- makers, whether in industry, Government, or private life, who want to know more about current and po- tential mineral resource problems and what steps are being taken, or might be taken, to solve them. Bold-face type is used in the text to emphasize ideas which the authors consider to be especially signifi- cant. References are cited in parentheses by the au- thor’s name and the date of publication; full cita- tions are given at the end of the report. WHAT IS OUR MINERALS PROBLEM? Our society is dependent on minerals, and without a steady supply of them it could not survive. Con- sider a few of the mineral products in a typical American home: Building materials: sand, gravel, stone, brick (clay), cement, steel, aluminum, asphalt, glass. Plumbing and wiring materials: iron and steel, copper, brass, lead, cement, asbestos, glass, tile, plastic. Insulating materials: rock wool, fiberglass, gypsum (plaster and wallboard). Paint and wallpaper: mineral pigments (such as iron, zinc, and titanium) and fillers (such as talc and asbestos). Plastic floor tiles, other plastics: mineral fillers and pigments, petroleum products. Appliances: iron, copper, and many rare metals. Furniture: synthetic fibers made from minerals (principally coal and petroleum products) ; steel springs; wood finished with rotten-stone polish and mineral varnish. Clothing: natural fibers grown with mineral fertil- izers; synthetic fibers made from minerals (principally coal and petroleum products). Food: grown with mineral fertilizers; processed and packaged by machines made of metals. Drugs and cosmetics: mineral chemicals. Other items, such as windows, screens, light bulbs, porcelain fixtures, china, utensils, jewelry: all made from mineral products. In our homes, our oflices, our industries—in al- most every facet of our daily life—we use minerals without knowing it, so hidden are many of their uses among the technical complexities of modern indus- trial processes and products. The entire US. economy is based on minerals. In 1972, the last full year prior to the pinch of the oil embargo, domestic raw materials valued at $32 bil- lion were converted into energy and processed ma- terials, the value of which exceeded $150 billion and formed the basis of the Gross National Product of $1.1 trillion (Mining and Minerals Policy, 1973). Our problem is simply that the United States does not have an adequate known domestic supply of all the minerals needed to maintain our society for the foreseeable future. We never have had all we needed, but in the past we could easily obtain materials from abroad. Today we meet a smaller percentage of our needs from domestic supplies (table 1), and minerals from overseas are increasingly costly and, in some cases, of uncertain availability. Nationalization of mines in some countries discourages participation by American mining companies; cartel agreements among the major producing nations can suddenly and dramatically raise prices or even halt supply, as has happened recently with petroleum; and “develop- ing” nations are now competing in the world market for the purchase of mineral raw materials. The many facets of our problem become apparent when we consider the ways we might try to alleviate it: By reducing the demand for scarce minerals, through substitution of others, reduction of waste, or elimination of some uses. 2 MINERAL RESOURCE PERSPECTIVES 1975 TABLE 1.——Um'ted States’ dependence on foreign sources for some of its minerals [Mining and Minerals Policy, 1973] a. Less than half imported from foreign sources: Copper Tellurium Iron Stone Titanium (ilmenite) Cement Lead Salt Silicon Gypsum Magnesium Barite Molybdenum Rare earths Vanadium Pumice Antimony b. One-half to three-fourths imported from foreign sources: Zinc Nickel Gold Cadmium Silver Selenium Tungsten Potassium c. More than three-fourths imported from foreign sources: Aluminum Tantalum *Manganese Bismuth Platinum Fluorine Tin *Strontium *Cobalt Asbestos *Chromium *Sheet mica *Titanium (rutile) Mercury *Niobium *Commodities more than 90 percent imported. By supplementing the raw mineral supply, through recovery and recycling of scrap and used mate- rials. Recycling of scrap (from junk) already provides significant amounts of some metals— over half of our annual needs for antimony and one-quarter to one-third of that for iron, lead, nickel, mercury, silver, and gold (Mining and Minerals Policy, 1973, p. 20). Research on re- cycling of domestic and industrial waste is de- veloping new processes for use by industry. By increasing our domestic supply, through discov- ery of new mineral deposits and through de- velopment of technology for the feasible re- covery of low—grade deposits. From the perspective of the 1970’s, increasing our domestic supply of minerals seems imperative. A widespread misconception, however, allows that this is simply a matter of economics and technology— that the Earth’s crust is an infinite storehouse that can readily be tapped for new supplies of all kinds of mineral raw materials by either raising the price or developing new technology (which in turn may raise the price). The economic and technologic factors gov- erning mineral supply cannot be ignored, but neither can another more fundamental factor: geologic availability. A mineral resource is a concentration of elements or compounds in certain rocks, or in cer- tain geologic environments, in a form that can yield a useable mineral commodity, but such concentra- tions are geologic rarities, and even low-grade de- posits are anomalous concentrations. If a mineral is not geologically available, favorable economics and technology are not pertinent. Nor is “raising the price” as simple as it sounds. Raising the price may actually reduce the supply of a mineral commodity brought from existing facili- ties to market per unit of time, because the higher price enables the company to mine a lower grade ore, thereby reducing the amount of concentrate de- rived from the same mill per unit of time. The min- ing and milling capacity of a property is almost never increased to accommodate what are believed to be short-term price fluctuations. The use of increasingly lower grades of ore poses two other problems. First, more energy must be spent, both to dig the ore from the ground and to ex- tract the mineral from it. Second, the environmental impact is greater, not only as a result of mining larger volumes of rock and discarding larger volumes of waste, but also because of the increased energy and water requirements. Realistic appraisals of the quantities of mineral resources remaining to be developed and yet to be discovered are a matter of concern for many policy- makers in both industry and Government, but miner- al resources cannot be inventoried like cans on a shelf. Reserves, or mineral resources that have been found, sampled, and measured and that can be legally and profitably mined under present conditions, can be inventoried (although on a national scale the in- ventory is only accurate to the extent that mining companies are Willing to release data). But in addi- tion to these reserves, there are: known, low-grade deposits not profitable to mine now; new deposits of reserve quality that can be logically inferred to exist but are as yet undiscovered; and even new types of deposits not yet recognized. These are all mineral resources, and appraising them accurately is per- haps as difficult as appraising the 1985 wheat crop. Resources can be estimated only by successive ap- proximations, subject to constant change as known deposits are worked out, as new deposits are found, and as new techniques of exploration, mining, and processing are developed, literally creating reserves from material that was previously unuseable. And so an important facet of the minerals problem is the need to improve methods for appraising resources and to have the successive estimates come closer and closer to being coincident with the amount of ma- terial actually occurring. CLASSIFYING MINERAL RESOURCES Through the years geologists, mining engineers, and economists have used many terms to describe he CLASSIFYING MINERAL RESOURCES 3 and classify mineral reserves and resources. Some of these terms have gained wide use and acceptance, although their use by different authors commonly has been marked by a vagueness that makes precise comparison of data difficult. More uniform use of reserve and resource terminology is critical to better communication on this vital subject. Modern attempts to establish a standard ter- minology of mineral resources began during World War II when the U.S. Bureau of Mines and the U.S. Geological Survey (1947) were assessing the Na- tion’s mineral position. That terminology was re- viewed and revised later by a committee of the So- ciety of Economic Geologists (Blondel and Lasky, 1956). More recently McKelvey (1972) and Brobst and Pratt (1973) have discussed concepts of mineral resources and their terminology and emphasized the need for understanding the broader implications of the difference between reserves and several cate- gories of mineral resources. In anticipation of increased activity in the assess- ment of mineral resources during the coming years, the staffs of the U.S. Bureau of Mines and the U.S. Geological Survey have collaborated in a joint state- ment on the classification of mineral resources. The classification (summarized here) is intended to be useful for all mineral commodities, including the energy materials, although supplementary essays dealing with classification problems peculiar to cer- tain commodities may be required. The distinction between resources and reserves is based on current geologic and economic factors. A resource is a concentration of naturally occurring solid, liquid, or gaseous materials in or on the Earth’s crust in such form that economic extraction of a commodity is currently or potentially feasible. A reserve is that portion of the identified resource from which a useable mineral or energy commodity can be economically and legally extracted at the. time of determination. The term are is used for re- serves of some minerals. Resources thus include, in addition to reserves, other mineral deposits that may eventually become available—known deposits that cannot be profitably mined at present, because of eco- nomics, technology, or legal restraints, and also un- known deposits, rich and lean, that may be inferred to exist on the basis of geological evaluation but have not yet been discovered. An analogy from the field of personal finance may help to clarify the dis- tinction: reserves are represented by the funds in one’s bank account and by other liquid assets; re- sources include, in addition, all other assets and all anticipated future income, from whatever source. Public attention usually is focused on reserves of mineral or energy materials. Long-term public and commercial planning, however, must be based on the probability of geologic discovery of new deposits and technologic development of economic extraction proc- esses for currently unworkable deposits. Thus, all components of total resources must be continuously reassessed in the light of new geologic knowledge, of progress in technology, and of shifts in economic and political conditions. Another requirement of long-term planning is the weighing of multicommodity or total resource avail‘ ability against a particular need. The general classi- fication system must therefore be uniformly appli- cable to all commodities so that data for alternate or substitute commodities can be compared. To serve these planning purposes, total resources are classified in terms of both economic feasibility and the degree of geologic assurance of their occur- rence (fig. 1). As shown in figure 1, total resources are divided into two major fields, identified and undiscovered resources; they in turn are subdivided. The resource terms that will be most useful to readers are: Identified resources: Specific bodies of mineral- bearing material, the location, quality, and quantity of which are known from geologic evidence and, if they are in the demonstrated category, are supported by engineering meas- urements. Reserves (already defined). Identified subeconomic resources: Resources that may become reserves as a result of changes in economic, technologic, and legal conditions. Undiscovered resources: Bodies of mineral-bearing material surmised to exist on the basis of broad geologic knowledge and theory. Explor- ation that confirms their existence and re- veals quantity and quality will permit their reclassification as reserves or as identified subeconomic resources. Hypothetical resources: Undiscovered resources that may reasonably be expected to exist in a known mining district under known geologic conditions. Speculative resources: Undiscovered resources that may exist either as knewn types of deposits in a favorable geologic setting where no discoveries have been made or as unknown types of deposits that remain to be recognized. 4 MINERAL RESOURCE PERSPECTIVES 1975 mu "MRS 103W!!!” UNDISCOVSRED SPKUMTNZ {In until-covered dint-ices) mun end HYPO‘HIHICAL (In Known dilute-cl) natured Indie: tad Inf-2nd RESERVES aconanzc _._) Paramatglnal z n In 0 v: z n n m Increasing degree of economic feasibility + + .1. + Summiml 6 Increasing degree of geologic assurance FIGURE 1.——Classification of mineral resources. Measured, indicated, and inferred are terms ap- plicable to both reserves and identified subeconomic resources. Measured—Identified resources for which ton- nage is computed from dimensions revealed in out- crops, trenches, workings, and drill holes and for which grade is computed from the results of de- tailed sampling. The sites for inspection, sampling, and measurement are spaced so closely and the geo- logic character is so well defined that size, shape, and mineral content are well established. The com- puted tonnage and grade are judged to be accurate within limits which are stated, and no such limit is judged to be different from the computed tonnage or grade by more than 20 percent. Indicated—Identified resources for which ton- nage and grade are computed partly from specific measurements, samples, or production data and part- ly from projection for a reasonable distance on the basis of geologic evidence. The sites available for in- spection, measurement, and sampling are too wide- ly or otherwise inappropriately spaced‘to permit the mineral bodies to be outlined completely or the grade to be established throughout. Inferred—Identified resources for which quan- titative estimates are based largely on broad knowl- edge of the geologic character of the deposit and for which there are few, if any, samples or measure- ments. Continuity or repetition is assumed on the basis of geologic evidence, which may include com- parison with deposits of similar type. Bodies that are completely concealed may be included if there is specific geologic evidence of their presence. Esti- mates of inferred reserves or resources should in- clude a statement of the specific limits within which the inferred material may lie. The terms proved, probable, and possible (used by industry for economic evaluations of ore in specific deposits or districts) commonly have been used loosely and interchangeably with the terms meas- ured, indicated, or inferred (used by the Depart- ment of the Interior mainly for regional or national estimates). “Proved” and “measured” are essentially synonymous. “Probable” and “possible,” however, are not synonymous with “indicated” and “inferred.” “Probable” and “possible” describe estimates of part- ly sampled deposits—in some definitions, for ex- ample, “probable” is used to describe deposits sam- pled on two or three sides, and “possible” for de- posits sampled only on one side; in the Bureau/ Survey definitions, both would be described by the term “indicated.” Long-range estimates of resources are little af- fected by increases of reserves brought about by new discoveries or new technologic or economic develop- ments. What such developments do, in effect, is sim- ply add ton-s or ounces to the reserve category by taking them away from one of the resource cate~ gories shown in figure 1. Such developments are the life blood of the mineral industry because they re- plenish the dwindling reserves. But they do not alter the total resource picture substantially. A signifi- cant increase in the total resources is brought about only by a major scientific or technologic break- through. Consider, for example, recent estimates of gold re sources. In 1967, when the price of gold was $35 per ounce, the US. Bureau of Mines estimated 244 mil- lion ounces of gold as identified subeconomic re- sources, potentially recoverable at prices as high as $145 per ounce. With the price in the $160—190 range, as it was late in 1974, those formerly subeconomic resources can be regarded as reserves—but the total resources remain unchanged. On the other hand, gold in seawater is not now considered a resource because of its low concentration—only as much as 0.05 parts per billion, or about 1 ounce of gold per million cubic metres of seawater. Discovery of a method for recov- ering this gold economically on a commercial scale would add spectacularly to the world’s total resources of gold, on the order of 1,000 ounces per cubic kilo- metre of seawater. In the 1960’s, recognition of a new type of resource (the so-called disseminated gold, or Carlin type) was a scientific milestone that added millions of ounces of gold to the categories of identi- fied, hypothetical, and speculative resources. CLASSIFYING MINERAL RESOURCES 5 APPRAISING MINERAL RESOURCES Mineral reserves and resources are dynamic quan- tities and must be constantly appraised as known deposits are worked out, new deposits are found, new extractive technologies and uses are developed, and new geologic knowledge indicates new areas and en- vironments favorable for exploration. A final, once- and-for-all “inventory” of any mineral resource is nonsense. Indeed, appraising the national resources of any mineral commodity is at this time an emerging science, and a universally accepted appraisal method has yet to be found. METHODS Resource appraisals made in the past have differed greatly because they were made by different methods to fill different needs. Many are of limited use as appraisals of total resource potential because of limi- tations imposed by the purpose for which they were made. In general, two types of approaches have been taken. One utilizes extrapolations based on produc- tion and economic information; the other utilizes extrapolations based on information about geologic occurrence. More recently, geomathematical tech- niques have been applied to help quantify both types. Approaches that extrapolate from production and economic data consider such variables as improve- ments in discovery and recovery techniques and changes in the economics of exploration, production, and marketing (for example, Hubbert, 1962, 1967, 1969; Moore, 1966, 1970). Some incorporate the de- clining efficiency in exploration, the growth of res serves by additions, and (or) the effects of price changes on resource potential (Arps and others, 1970; Brinck, 1972; Bieniewski and others, 1971). Such approaches tend to allow neither for possible breakthroughs that might make subeconomic occur- rences recoverable, nor for possible new discoveries in unexplored areas, and so they have limitations for appraising total resource potential. Approaches that emphasize geologic occurrence extrapolate data about mineral deposits from ex- plored to unexplored areas. Such geologic evaluation has been used to estimate the number of undiscov- ered porphyry copper deposits in the Southwestern United States, British Columbia, Chile, «and Peru (Lowell, 1970) and to estimate the quantity of sev- eral mineral commodities remaining to be discovered in the area north of 60° latitude in Canada (Derry, 1973). This concept of basing mineral resource ap- praisals on geological extrapolations has been ex- tended by the use of such geomathematical tech- niques as statistical treatment of opinion polls (Harris and others, 1970) and construction of deci- sion models (Allais, 1957, based on Nolan, 1950; Slichter and others, 1962). Ideally, resource appraisals should be made by dividing the total unmined resources of each com- modity into specific geologic, technologic, and eco- nomic subsets, each with its own estimate and each explicitly defined so that its validity can be tested. The procedure by which these estimates are made should be flexible so that new information can be readily incorporated. The precision of an estimate should be stated as explicitly as possible and ex- pressed within prescribed limits of confidence; in practice, however, such confidence intervals are difl‘i— cult to construct because the assumptions used in making geologic extrapolations about occurrences of mineral commodities are likely to vary, and variabil- ity of this sort is not easily reduced to numbers. The assumptions used to construct an estimate should therefore also be stated explicitly. Bias should be avoided; the analyst should not alter an estimate in order to be conservative or optimistic. The most recent comprehensive mineral resource appraisal by commodity for the United States (and, for some commodities, for the world) was published by the US. Geological Survey (Brobst and Pratt, 1973). The methods used to appraise 65 commodities are diverse, but most. of them are variants of geo- logic extrapolation. The appraisals span a broad range, from qualitative to quantitative, and none of the authors would claim that his evaluation is a definitive resource appraisal. In “Project Appalachia” the Geological Survey of Canada currently is combining expertise in regional geology, mineral-deposit geology, and mathematical processing to provide three approaches to evalua- tion—(1) a traditional metallogenic (geologic ex- trapolation) appraisal, (2) a metallogenic-based geo- mathematical evaluation, and (3) an opinion-poll appraisal. The Canadian Appalachian region was se- lected for this pilot study because its size is con- venient, its geology is relatively well known, and it contains a variety of mineral deposits. Results of the three appraisals will be compared and will be fol- lowed up with an opportunity for a second round of opinions based on comments from the first. On July 1, 1974, the US. Geological Survey began a new multidisciplinary program of field and labora- tory studies designed to assess the metal and se- lected nonmetal resources of Alaska. Because mineral appraisals of such large, remote, and little-known regions as Alaska are still largely experimental, a two-year program called the Prototype Alaskan Min- 6 MINERAL RESOURCE PERSPECTIVES 1975 eral Resource Assessment Program (PAMRAP) will develop guidelines, techniques, and products as a model for a future statewide mineral assessment program. For each 1:250,000-sca1e (1°-by-3°) quad- rangle studied, geologic, geochemical, and geophysi- cal maps, and computer-enhanced data from satel- lites (multispectral ERTS imagery) will be inter- preted to produce a mineral-potential map. This will be accompanied by a tabular summary of known mines, prospects, and occurrences, and by resource- assessment diagrams showing estimates of identi- fied, hypothetical, and speculative resources. ORGANIZING DATA Increasing attention to both geologic and eco- nomic evaluation of mineral resources will inevitably generate great masses of numerical data and infor- mation which will require skillful organization if it is to be effectively utilized. Because the subject of mineral resources is broad in scope and complexity, involving principally the fields of geology, mining engineering, and economics but also many other closely related, highly technical subjects, and be- cause of the great variety of basic raw materials and the Wide range of needs and use patterns, data on mineral resources are difi‘icult to organize, gain access to, and use effectively. Computerized process- ing now offers the best available means of improving the usefulness of data. The computer can separate a whole file into its parts, perform operations on parts, and then reorganize the file. Moreover, a com- puterized file can be constantly updated, thus elim- inating a major problem with the old manual sys- tems. A computer can accomplish in seconds many operations of data which would never even be at- tempted by manual methods; thus all important relationships of available data can be determined. Computerized files of resource data now in use within the U.S. Geological Survey and U.S. Bureau of Mines include: 1. CRIB—Computerized Resources Information Bank (for mineral resource data), U.S. Geo- logical Survey. 2. MAS—Minerals Availability System, U.S. Bu- reau of Mines. 3. Oil and gas file of the Oflice of Oil and Gas and U.S. Geological Survey. 4. Production/Consumption surveys of the U.S. Bureau of Mines. THE RESOURCE OUTLOOK To try to summarize the “status of resources” in a simple numerical table, showing tons of this or that commodity, would be like trying to sketch a sunset in black and white. Describing the full spec- trum of resources, like painting a sunset, requires the skillful blending of a full palette of hues—the geologic, technologic, economic, and legal factors that make the appraisal of resources so complex. To disregard these factors and to tabulate resource estimates as raw numbers removed from their con- text inevitably lead to misinterpretation. The essentials of the resource situation can be expressed by grouping the important commodities according to the general domestic outlook for re- serves and resources for the remainder of the twen- tieth century. These groupings are listed in table 2. An indication of our dependence on foreign sources for some commodities was given in table 1. TABLE 2.—-General outlook for domestic reserves and resources through 2000 A.D. [Within each group, commodities are listed in order of relative importance as determined by dollar value of U.S. primary demand in 1971. An asterisk marks those commodities which may be in much greater demand than is now projected because of known or potential new applications in the production of energy] Group 1: RESERVES in quantities adequate to fulfill pro- jected needs well beyond 25 years. Coal Phosphorus Construction stone Silicon Sand and gravel Molybdenum Nitrogen Gypsum Chlorine Bromine Hydrogen Boron Titanium (except Argon rutile) Diatomite Soda *Barite Calcium Lightweight aggregates Clays Helium Potash Peat Magnesium *Rare earths Oxygen *Lithium Group 2: IDENTIFIED SUBECONOMIC RESOURCES in quantities adequate to fulfill projected needs beyond 25 years and in quantities significantly or slightly greater than estimated UNDISCOVERED RESOURCES. Aluminum Vanadium * Nickel *Zircon Uranium Thorium Manganese Group 3: Estimated UNDISCOVERED (hypothetical and speculative) RESOURCES in quantities adequate to ful— fill projected needs beyond 25 years and in quantities sig- nificantly greater than IDENTIFIED SUBECONOMIC RESOURCES; research efforts for these commodities should concentrate on geologic theory and exploration methods aimed at discovering new resources. Iron Platinum *Copper Tungsten *Zinc * Beryllium Gold *Cobalt *Lead * Cadmium Sulfur *Bismuth * Silver Selenium *Fluorine * Niobium Group 4: IDENTIFIED-SUBECONOMIC and UNDISCOV- ERED RESOURCES together in quantities probably not adequate to fulfill projected needs beyond the end of the century; research on possible new exploration targets, new types of deposits, and substitutes is necessary to relieve ultimate dependence on imports. Tin *Antimony Asbestos * Mercury Chromium *Tantalum a THE RESOURCE OUTLOOK 7 Known U.S. reserves of many minerals represent only a few years’ supply. The outlook for resources is somewhat better, but to bring them into the cate- gory of available reserves will require enormous and costly efl’orts of exploration and research. Details on the technologic and geologic problems of utilizing resources of specific commodities are in US. Bureau of Mines (1970, Bulletin 650); Brobst and Pratt (1973, US. Geological Survey Professional Paper 820); and a shorter report by Pratt and Brobst (1974) that summarizes the principal findings of Professional Paper 820 with regard to the resources of 27 major mineral commodities. EXPLORATION FOR MINERAL RESOURCES According to a US. Geological Survey compilation from the literature, the world’s mineral-reserve wealth was increased in 1973 by nearly 100 million tons of copper; 40 million tons of nickel; several bil- lion tons of iron ore and bauxite; about 38 million tons of zinc; and about 21 million tons of lead. These developments and increased productive capacities are the result of many years of concerted effort by gov- ernments and industry. To maintain the forward thrust of such activities, extensive research and increased exploration activi- ties for new deposits in unexplored but geologically favorable areas are essential. New or improved ex- ploration techniques and geological concept-s must be devised. The technology to develop mineral deposits and to increase the recovery of metal from low-grade ores must be improved. New policies and guidelines must be established so that long-range programs of research, discovery, and development may be con- ducted and investment capital attracted to support the mineral industry. Closer cooperation between government agencies, universities, and private indus- try is essential for a better understanding of geo- logical and technical problems. TECHNIQUES FOR MINERAL EXPLORATION Most of the mineral deposits discovered before the twentieth century were found because they were exposed at the Earth’s surface and contained high concentrations of the minerals sought. Commonly these minerals were fairly obvious; they occurred as brightly colored stains or as shiny grains on out- crops or in stream sediments. In many parts of the world today, most of these obvious mineral deposits have already been discov- ered. The emphasis in the search for new resources is therefore on finding low-grade deposits that would not have been profitable to mine in the past, deposits that are poorly exposed or not exposed at all, and deposits of previously unknown types that exhibit no obvious visual surface expression. Mineral exploration today is a highly sophisti- cated, expensive and involved science that includes regional appraisal, reconnaissance geological studies, and physical exploration. The role of the traditional mineral prospector who searched for and sampled outcrops continues to decline in importance as the most obviously mineralized areas become thoroughly explored. The equipment used today is more sophisti- cated and the techniques are costlier than the early prospector could have imagined. At the same time, domestic lands available for prospecting are becom- ing more restricted with the establishment of primi- tive and wilderness areas, parks, monuments, and wildlife refuges. In some areas, urban expansion and environmental-protection measures limit access and restrict operations. Modern exploration activities include geologic, geochemical, and geophysical investigations; three- dimensional sampling by core drilling or other methods; laboratory analyses including ore treat- ment, concentration, and recovery tests; economic appraisal; and evaluation of transportation, water, and energy requirements. Favorable results must be obtained from these studies before a property or a mineral deposit can be considered for development. Geology is the principal discipline used in mineral exploration; a thorough understanding of the physi- cal and chemical characteristics of mineral deposits is essential. In regional geologic appraisals an ex- ploration geologist may use such concepts as global tectonics, metallogenic provinces; metallotects (geo- logic features believed to have influenced the locali- zation or concentration of elements in the Earth’s crust) ; the relationships between geochemical abun- dance and mineral resources; the relationship be- tween crustal abundance and the size of resources; and multivariate geostatistical analysis. Geochemical exploration uses the systematic measurement of one or more chemical properties of a naturally occurring material to discover and delin- eate abnormal chemical patterns that may be related to potentially economic mineral deposits. The chemi- cal property most commonly measured is the con- centration of an element, or of a group of elements, in rock, soil, or streambed sediments, in vegetation, in well, stream, lake, or ocean water, in glacial debris, or in airborne volatile materials. Exploration geophysics applies the principles of physics to the search for mineral deposits that occur in the Earth’s subsurface. Most geophysical work is 8 MINERAL RESOURCE PERSPECTIVES 1975 done with sophisticated electronic equipment that can detect subtle contrasts in such physical prop- erties as specific gravity, electrical conductivity, heat conductivity, seismic velocity, and magnetic suscepti- bility. The common techniques used, either singly or in combination, are gravity, magnetic, electrical, electromagnetic, seismic, and radioactivity methods. Measurements may be made from aircraft, at the Earth’s surface, or in boreholes. In recent years, a variety of “telegeologic” or “re- mote sensing” techniques—measurements of various geologic or related properties from aircraft or satel- lites—have provided insights into complex structures of some regions where much of the bedrock is con- cealed. Side—looking radar imagery or photography has provided useful base maps in areas where con- ventional photographic methods had failed because of adverse weather and atmospheric conditions. LANDSAT, formerly Earth Resources Technology Satellite, imagery of many areas provides informa- tion on potential mineral and mineral-fuel deposits and is proving to be valuable for a better understand- ing of ground-water conditions and water-manage- ment problems. Preliminary results in Alaska and elsewhere suggest that LANDSAT data may have significant value in monitoring environmental effects of mineral and fuel production activities. Modern laboratory techniques are applied in ex- ploration projects. Electron-probe X-ray micro- analysis, new rapid chemical analysis, Visual color- comparison techniques, and neutron-activation an- alysis have been found particularly useful. In addition to exploration on land and in the air, government and industry are engaged in designing and testing exploration and mining equipment for undersea exploration operations. LEAD TIME Following a preliminary regional appraisal and assuming that jurisdictional and land acquisition problems in selected areas are resolved, reconnais- sance geological mapping and studies, detailed geo- logical mapping, geochemical and geophysical in- vestigations, physical exploration, bulk sampling, ore testing, reserve calculations, and economic eval- uations commonly require many years of concerted effort and large amounts of capital investment before a mineral deposit may be considered for development. Lead time of 20 years or more may be necessary from the beginning of the exploration project to the delineation of an economic mineral deposit. Mine development and plant construction activities require several additional years and more capital expenditure before a mineral commodity is pro- duced and marketed. For example: In 1929, officals of the Copper Range Company recog- nized that a large deposit of copper ore existed in the White Pine areas of the Michigan Upper Peninsula, but it was not until 1955 that pro- duction started at the mine following completion of exploration and development projects at a cost of 61.7 million dollars. For a period of 10 years, the Bear Creek Mining Com- pany has been exploring a copper-silver prospect in the Spar Lake district of western Sander County, Montana. The American Smelting and Refining Company, which has entered into an agreement to develop the property, indicates that the feasibility study and mine development will require 13 to 18 years. The Amax’s Henderson molybdenum mine in Colo- rado, which is expected to start production in 1975, required 8 years of development at an esti- mated cost of about 250 million dollars. At the Granduc mine, British Columbia, geological investigations, surface exploration drilling, and ore testing required nearly 12 years, and 5 addi- tional years were needed for development work before production began in 1970. Total cost of these activities was nearly 115 million dollars. The Michiquillay copper ore deposit in Peru was investigated initially in the 1950’s, but produc- tion is not scheduled to start before 1980 or later. None of these examples takes into account the time and expenses involved in the initial stages of regional appraisal. Because the mineral industry is so competitive, little information, published or un- published, is made available in the early stages of regional appraisal and land-acquisition activities. Factors such as location, size of area under study, availability of base and geological maps, aerial photo- graphs, and geologic reports all affect the time needed and expenses involved at this early stage. In the period 1955—69, nearly one billion dollars was spent in exploration ventures in the Western United States. In 1972, more than 32 million dollars was spent in worldwide geophysical and geochemical studies, 4.3 million of which was spent in the United States. CHANCES 0F SUCCESS An exploration project that leads to discovery, development, and production is exceptional, and the odds against making a significant discovery are high: Under the Strategic Mineral Development program in the period 1939-1949, about 10,000 prospects j—Q . _.__‘ EXPLORATION FOR MINERAL RESOURCES 9 were examined; of these 1,342 deposits were investigated in detail, but only 1,053 contained enough tonnage to be of interest. The outstand- ing development resulting from this program was the San Manuel copper mine in Arizona. The Defense Minerals Exploration unit (established by the DME Act) in the period 1951—1958 re- ceived 3,888 applications for Federal financial assistance to explore deposits of strategic and critical minerals. Of 1,159 contracts granted, 399 resulted in the certification of discovery of some valuable minerals, and by 1959, 45 of the deposits were in production. The outstanding discoveries of this program included zinc ore deposits in Tennessee and lead deposits in the Viburnum district of Missouri. RECENT EXPLORATION ACTIVITY In recent years, worldwide exploration for nonfuel minerals has focused principally on copper, iron, bauxite, lead-zinc, nickel-copper, uranium, and fer- tilizer materials. Improved prices for gold and silver have stimulated search for precious metals. Offshore exploration is receiving more attention than hereto- fore. Exploration activities are gaining momentum ‘in the United States and in Brazil, Europe, South Africa, and the Southwest Pacific. Governmental constraints, however, and new investment and min- ing policies in Australia, British Columbia, Zambia, and Peru have discouraged exploration efiorts in these areas. Major exploration targets in the United States have been disseminated copper, gold, molybdenum, and uranium deposits, stratiform copper deposits, stratabound lead-zinc deposits, and mineral deposits in volcanic sedimentary sequences. The Precambrian Belt series (extending from Libby, Montana, to the Coeur d’Alene district of Idaho) ; certain districts in Wisconsin, Minnesota, and Tennessee; and some un- explored areas in Alaska have received more than usual attention. Exploration for copper porphyries massive sulfides, and Mississippi Valley-type lead- zinc deposits has continued in the Appalachian re- gion. Interim results were reported on nickel-copper investigations in Minnesota Where, Within a large, low-grade mineralized area, some deposits are esti- mated to contain as much as 100 million tons of rock that is 0.8 percent combined copper-nickel. Other projects include exploration for zinc in central Ten- nessee; for molybdenum in Idaho and Colorado; for copper in Arizona, Montana, Utah, Nevada, Wyom- ing, and Washington; and for platinum in Montana. Also in the United States, known base-metal ore bodies have been extended in the Coeur d’Alene area of Idaho, and large deposits have been developed at Buick, Mo., and Gainsboro, Tenn. Preliminary reports have been made of exploration, development, and reactivation of old precious metal mining districts in Nome, Alaska; Silverton and Cripple Creek, 0010.; Pinson, Nev. ; Mocassin, Mont. ; and Wenatchee, Wash. The Homestake Mining Company announced the expenditure of several million dollars at the Homestake gold mine, South Dakota, to develop about 13 million tons of ore hitherto considered sub- economic. Exploration for uranium deposits has con- tinued principally in Wyoming, Utah, Colorado, and New Mexico. No significant discoveries have been announced, but some earlier discoveries are being developed. Recent developments in Canada include lead-zinc- silver deposits of the Strathcona mine in Baffin Islands, the Arvik mine in the Northwest Territories, and the extension of the Kidd Creek ore body, On- tario, Bathhurst district in New Brunswick and the Gays River area, Nova Scotia. Significant discoveries of copper were also reported in Sonora, Mexico. Elsewhere in the world, exploration for aluminum ore (bauxite) is focused principally on Australia, Brazil, and West Africa; for iron, on Brazil, West Africa, Australia, North Korea, and India; for cop- per, on Panama, Peru, Chile, Argentina, Australia, Philippines, Indonesia, Pakistan, Iran, India, Yugo- slavia, Sweden, U.S.S.R., Zaire, Zambia, Ethiopia, and Southwest Africa; and for nickel, on Africa, Australia, Indonesia, Philippines, India, and Brazil. Trade journals regularly report new developments in the mineral industry. A comprehensive survey, for example, of international mining activities was pub- lished in the British journal Mining Magazine (1974). Results of exploration efforts outside the United States, measured by additions to the world’s re- serves of the major metals, are summarized by con- tinent in table 3. Impressive as these data are for tons of reserves reported, explored and (or) de- veloped, and expenditures noted, the reader must keep in mind at least three facts: 1. Lead time to reach production levels may be 15 or more years for many of the individual Projects. 2. The products from the foreign developments will for the most part be available to the world market and are not automatically destined for United States consumption. 3. Expenditures yet to be made may be many times those already reported. 10 MINERAL RESOURCE PERSPECTIVES 1975 TABLE 3.—Foreign exploration-development data reported in 1973 Reported Ore in . Indicated reserve investment development Location and (ore in millions M (millions of (millions mineral of tons) From To dollars) of tons) Africa Bauxite ____________ 1,490 42 55 470 900 Copper _____________ 250 0 8 5 5 200 230 Iron ore ____________ 8,880 35 68 600 7,000 Lead-zinc ___________ 8 8 18 ________________________________________ Nickel ______________ 225 0 7 1.5 291 50 Gold 2 ______________ 130 0 5 0.9 150 ______________________ Total _______________________________________________________________________ 1,711 ______________________ Asia Bauxite ____________ 95 45 59 100 55 Copper _____________ 970 0 9 1 5 400 475 Iron ore _____________ 2,420 _________________________________________________________________________ Lead-zinc __________ 95 17.7 _________________________________________________________ Nickel ______________ 35 0.8 2 0 ________________________________________ Total _______________________________________________________________________ 500+ ____________________ Australia Bauxite ____________ 3,735 45 60 900 1,020 Copper _____________ 230 0 7 3.0 145 205 Iron ore ____________ 1,900 60 64 ________________________________________ Lead-zinc ___________ 245 24.7 ________________________________________ Nickel ______________ 230 1.3 3.3 471 70 Gold 3 ______________ 6 0.2 0 4 ________________________________________ Total _______________________________________________________________________ 1,516+ ____________________ Canada Copper _____________ 1,600 0 4 1 7 90 190 Iron ore ____________ 3,300 23 50 327 700 Lead-zinc ‘ __________ 255 7 20 96 10 Nickel ______________ 540 0.33 1 6 ________________________________________ Total _______________________________________________________________________ 513 ______________________ Europe Bauxite ____________ 60 54 60 ________________________________________ Copper _____________ 775 0 7 1.6 175 610 Iron ore ____________ 800 50 60 ________________________________________ Lead-zinc ___________ 150 9 11.6 18 7 Nickel ______________ 820 1 2 1.3 72 110 Total _______________________________________________________________________ 265+ ____________________ Latin America Bauxite ____________ 4,000 High grade 1,200 ______________________ Copper _____________ 8,690 0 4 1 0 1,842 3,300 Iron ore ____________ 24,475 46 67 1,230 24,000 Nickel ______________ 190 1.0 2 6 430 100 Gold 5 ______________ 50 0.15 ________________________________________________________ Total _______________________________________________________________________ 4,702+ ____________________ Oceania Bauxite ____________ 80 50 __________________________________________________________ Copper _____________ 2,470 0.45 2 5 1,397 1,870 Lead-zinc ___________ 5 5.2 _________________________________________________________ Nickel _____________ 1,590 1.2 1 65 1,398 1,065 Total _______________________________________________________________________ 2,795+ ____________________ 1For bauxite, percent A1203; for gold, oz/ton. 2Some low-grade gold ores contain uranium. 3 Also contains silver values. 4 Includes Greenland. 5Includee complex ores that contain silver, lead. zinc, and copper. RECENT RESEARCH ON MINING AND EXTRACTION mining in some areas. This technology has made pos. Research into various aspects of solution mining, .sible the recovery of copper from some deposits or leaching, has advanced from laboratory and field hitherto considered uneconomic, and it is being used tests to commercial operations, and in the future successfully in treating gold ore. A principal dis- solution mining will bean alternative to conventional advantage of the method is that it does not recover L1 i—V ROLES OF INDUSTRY AND GOVERNMENT IN RESEARCH 11 such important byproducts from copper ores as gold, silver, selenium, and tellurium. The technical and economic feasibility of deriving aluminumfrom large domestic alunite deposits in the Western States and from high-alumina clays in Georgia is under serious study. Domestic aluminous laterites from deposits in Oregon have been used successfully instead of bauxite in a conventional alumina refinery. Offshore exploration for deposits of manganese nodules, which also contain appreciable quantities of copper, nickel, and cobalt, has attracted world- wide attention, particularly by interests in the Uni-ted States, Germany, Japan, and the U.S.S.R. These nodules have great potential value, and the technology to gather them from the deep seabed (4,000—6,000 metres) is rapidly being developed. If a satisfactory legal regime can be forged that will protect the miner’s investment, commercial man- ganese nodule production will likely be operable before 1980. Other significant results of scientific and engineer— ing investigations in developing technology for ex- tracting, processing, and recycling mineral-s, metals, and fossil fuels are summarized in U.S. Bureau of Mines (1974a). ROLES OF INDUSTRY AND GOVERNMENT IN RESEARCH For many reasons the ultimate objectives of indus- try’s and Government’s research in mineral resources lie at the ends of diverging paths. The mineral in- dustry is composed of specialized units focused on individual or small groups of commodities, each of which has different search and development prob- lems. The industry is so highly competitive and in- dependent in operation, even for the same commodi- ties, that no national focus has been possible beyond those imposed by supply and demand. The primary purpose of a private company is to find ore and pro- vide a marketable product at a profit. A large part of industry’s exploration activity is therefore di- rected toward developing reserves, generally from identified subeconomic resources and only to a much more limited extent from hypothetical and specula- tive resources. A company is understandably most concerned with its position in the short-term market- place, whereas Government is deeply concerned with developing knowledge about the Nation’s long-term total resource potential and its options for access to resources in the short term. The overall balance in treatment and understanding needed for public- policy decisions cannot be achieved by industry alone, as industry concentrates its efforts on specific com- modities and possibly overlooks others. In recent years the cost in time and money of ex- ploring for ore and developing reserves has increased markedly; at the same time the rate of discovery has decreased markedly. These factors have contributed materially to a reduction in domestic supplies and to price fluctuations which have resulted in overall re duction in funds available to industry for activities in mineral exploration. Some of these difficulties stem from Government actions. Even if industry funds were available, the costs are becoming too great to tolerate continued duplication of uncoordi- nated effort. Thus Government has the option to nurture the supply system through incentives to industry as a Whole for increasing exploration activ- ity and efliciency through mineral research and development. A responsibility of Government is to collect, syn- thesize, and analyze basic data about mineral re- sources and to make the information available to those who need and want such information for mak- ing both public and private decisions. A broad base of data is needed by large and small mining com- panies for development of new target areas for exploration and new exploration and production methods. Planners also require it for determining quantitative analysis of the geologic and economic availability of national and international mineral supplies. In the years ahead, it will be increasingly necessary for Government to have a reservoir of information for planning purposes and to be able to monitor supplies and sources of supply, as well as uses of minerals at home and abroad, in order to help the Nation and its mineral industry across the in- evitable economic highs and lows. MINERAL RESOURCES RESEARCH PROGRAMS IN THE U.S. GEOLOGICAL SURVEY The U.S. Geological Survey conducts mineral re- sources research programs in several of the organi- zational units shown in figure 2. The focus of this report is on those nonenergy programs that include the development and testing of resource theory and its application to improved appraisal and discovery technology; the development of resource information systems; and the analysis of these data to locate new resource targets and to provide forecasting and mon- itoring capabilities. These activities are supported by analytical laboratory facilities and by programs in geophysical and geochemical laboratory and field research, mineralogic and petrologic research on ore- forming environments, and development of new tech- 12 MINERAL RESOURCE PERSPECTIVES 1975 niques for geochemical and geophysical exploration. Other very important resource programs which are not discussed within the scope of this report include: (1) those for energy materials such as oil and gas, coal, uranium, thorium, and oil shale; (2) studies directed specifically toward environmental impacts of mining; (3) detailed mining engineering studies of leasable mineral commodities on Federal lands; and (4) 'hydrologic studies that support resource appraisal and development. US. Geological Survey mineral resource research programs collect and interpret Earth-science data in order to define the origin and occurrence of useful materials and to appraise the current and future availability of mineral resources in the United States. These programs lead to the definition of regions that seem likely to contain ore deposits, but not to their exploration or development as ore bodies which is the responsibility of the private sector. Incentives, however, to increase domestic exploration for se- lected minerals are provided to private companies through an exploration loan program and through research to improve exploration technology. Knowledge about the geologic availability of do— mestic and overseas mineral resources is the first link of a chain leading to discovery of ore deposits; this knowledge is basic to the development of sound Government policy decisions, and it fosters timely production of minerals by industry. US. Geological Survey programs aim to provide this kind of knowl- edge through four concurrent and continuing types of research: Investigations of the geologic environment of mineral occurrence—These studies are, aimed at obtaining basic information about known mineral occurrences that will help in identifying new ex- ploration targets and accelerating the conversion of undiscovered mineral resources into reserves. They include (a) geologic, geochemical, and geophysical field investigations in known mining districts as well as in broad regions and mineral belts, and (b) geo- chemical laboratory research on the ore-forming environment and the processes of ore deposition. Clues to the location of a deposit can be established from the characteristics of occurrence and mode of formation of similar deposits. Geologic research in- volves documenting type examples of metal-mining districts and the detailed examination of mineralized rocks and the processes that form them. Research that synthesizes geochemical, mineralogic, and petro- DIRECTOR l r 1 Land Information and Analysis Office Conservation Division Water Resources Division 1 I Office of Environmental Office of Energy Resou/rce/s/ Geology / / / / / ,oééeC/{////// G°°°"°/"}/‘}"/}/°/7/}/% /ANCILLARY OFFICES /Office of // /International Geology / FIGURE 2.—Organizations in the Geological Survey having program responsibilities for mineral resources research. Pri- mary responsibility for programs described in the report are indicated by stipple patterns, shared and supporting program responsibility indicated by line pattern. .m‘ts Y MINERAL RESOURCES RESEARCH PROGRAMS IN THE U.S. GEOLOGICAL SURVEY 13 logic field and laboratory work aims at evaluating sources of ore-forming solutions, how the solutions move, and establishing the conditions favorable for ore deposition. Geologic studies in potentially mineralized areas.— These studies are aimed at applying basic resource information and multidisciplinary field techniques to the appraisal and discovery of resources in potential- ly mineralized areas. The objectives are to develop greater competence in rapid and accurate resource appraisal, to provide information needed for de- cisions about using potentially mineralized land, to identify new geologic targets to explore for conven- tion-a1 as well as new types of ore deposits, and to make quantitative studies of the resources of criti- cally short commodities. Reports on this work are increasingly sought where Federal, State, and pri- vate lands are available for exploration by industry or involve Government management of leases; they are also needed when land-use and environmental impact decisions are being made prior to urban de- velopment of areas that may contain minerals of potential commercial value. Programs of this sort are also conducted by the U.S. Geological Survey in other countries under the auspices of USAID or con- tracts negotiated by the Department of State with other national governments, and have three benefits: (a) They assist foreign governments in on-the—job training of geologists and in development of mineral exploration programs, thereby speeding discovery and appraisal of the country’s resources; (b) they add to basic knowledge of mineral deposits that can then be applied to the search for similar deposits in the United States; and (c) they provide opportuni- ties for the evaluation of foreign resources and their potential impact on domestic supply. Research in mineral exploration technology.— These programs integrate geologic, geochemical, and geophysical concepts and data gathered both in the laboratory and in the field to develop and field-test new methods for assessing mineral resource poten- tial as well as those for finding minerals, especially in new deposits and districts. Computer technology is used to process and aid in interpreting field and laboratory data. Resource analysis—Programs of analysis using computerized data systems permit storage and cate- gorized retrieval of information obtained from Gov- ernment and industry. Analysis includes testing and evaluating computer programs for specific applica- tion to improving resource appraisals; to designing, developing, and testing predictive model-s of resource occurrence for use in economic studies of the con- version of resources into reserves; and to continual updating of resource-data banks. Results of all these programs are published as re- ports and maps that provide the public with a con- tinuously expanding base of resource-related infor- mation. The principal publications of the U.S. Geo— logical Survey are Professional Papers, Bulletins, Circulars, and various series of maps. Information is also made available in open-file reports or pub- lished in professional scientific journals. Reports and maps released by the Government are described in a monthly listing, New Publications of the Geo- logical Survey, which is free upon application to the U.S. Geological Survey, National Center 329, Reston, Va. 22092. In addition to the four types of continuing re- search, and the extensive technical support required by that research, the U.S. Geological Survey is committed to certain prescribed objectives based on statutory acts (such as giving advice to the Govern- ment and response to Congressional mandates) and on contractual agreements between the various parts of the Survey itself or with outside agencies (such as projects financed jointly with the National Sci- ence Foundation or with State agencies). The pre- scribed objectives are annually becoming a larger part of the Survey’s mineral resource research pro- gram, and their fulfillment is possible only because of a substantial past investment in continuing scien- tific research backed by effective technical support. Work at the operating level is organized into indi- vidual projects that may be concerned with one or more of the four major types of research. Some pro- jects are regional in scope; others are more localized, intensive topical studies. These projects are des- cribed individually in the Science Information Ex- change of the Smithsonian Institution, Washington, D. C. The current status or noteworthy achieve- ments of projects are reported in Geological Survey Research 1971;, U.S. Geological Survey Professional Paper 900. A brief summary of current mineral re- source programs follows. REGIONAL APPRAISAL Regional resource mapping programs in geology, geophysics, and geochemistry are conducted in the United States as well as overseas. Increasingly, the successful and productive programs are multidisci- plinary efforts. Characterization of resource poten- tial depends on a geologic map framework that in turn is based on factors such as isotopic geo- chronology, geochemical distribution, and aeromag- netic map features. These programs are designed 14 MINERAL RESOURCE PERSPECTIVES 1975 to broaden the base of resource-appraisal theory and to provide answers about resource potential of large areas. Regional resource appraisal is receiving new em- phasis now because of the Wilderness Program on Forest Service lands and studies of Indian and Bureau of Land Management Interior lands (see p. 17). The Alaska Native Claims Act of 1974, for example, requires resource availability decisions for 80 million acres to be made within a 5-year period. In the Tucson area of Arizona, where min- ing and urban development interests are competing for the same lands, an assessment of potential cop- per resources has been made. Several mineral belts or regions in Nevada, Oregon, Colorado, and New Mexico are being examined at the 1:250,000 scale (2° quadrangle). The potential for developing new exploration targets of critical commodities such as platinum, nickel, copper, and chromium associated in ultramafic terranes is the basis for regional 2°- quadrangle appraisal mapping in California, Oregon, Minnesota, and Michigan. Gold- and silver-bearing volcanic rocks of Nevada are also being examined in detail. Aeromagnetic and gravity methods have long been used singly or in combination to analyze the regional framework and to identify major rock types that may contain mineral resources. Under the national aeromagnetic map program, the U.S. Geological Survey, along with cooperating State agencies, ad- ded 125,000 square miles of magnetic coverage in 1973. The work was done mainly by private con- tractors. Aeromagnetic maps are now publicly avail- able, or are in press, for 40 percent of the Nation, including complete coverage of ten states—Arizona, Colorado, Connecticut, Indiana, Iowa, Maryland, Massachusetts, Michigan, Minnesota, and Virginia. The flight-line spacing on these maps ranges from reconnaissance of 5 miles (about 8 kilometres) down to 1 mile (1.6 kilometres), which is well suited for helping to define mineral prospects. The truck-borne magnetometer has been effective as a rapid and economical means of examining the magnetic “sig- nature” of an area, which is used in planning aero- magnetic work, or for detailed evaluation of specific anomalies shown on existing aeromagnetic maps. AREAL GEOLOGIC STUDIES Areal geologic studies are aimed at defining broad mineral belts that contain conventional types of de- posits and new types of resource possibilities. Pre- cambrian shield areas, for example, are sources of many mineral resources elsewhere in the world, but they have not been extensively explored in the United States except for iron; therefore, studies of stratigraphy, structure, and chemical facies of Pre- cambrian iron formations and greenstone terranes in the Great Lakes shield and elsewhere are under- way. Similarly, areas containing post-Precambrian crystalline volcanic rocks are being examined for possible massive sulfide deposits; carbonate plat— form rocks are being examined as stratigraphic traps for base metal deposits of the Mississippi Valley type; coastal plain sedimentary rocks and local placer deposits are being examined for high-alumina clay and titanium deposits; areas containing porphyry-type disseminations are being examined for copper and uranium; and new or extensions of old mineral belts containing hydro- thermal vein and manto-type deposits are being ex- amined for base and precious metals. DETAILED STUDIES IN MINING DISTRICTS Detailed studies in mining districts have long been a cornerstone of resource research to define ore- forming environments and they continue to be a vital part of the program. Data on the geologic character- istics of a new type of low—grade disseminated copper-silver deposits at Spar Lake, Montana, in Precambrian quartzites will provide the basis for future exploration as the need for such deposits de- velops. Studies of low-grade disseminated gold de- posits of the Carlin type continue; discovery of these deposits opened new opportunities for precious- metals exploration in similar geologic terranes else- where. Porphyry deposits are the major source of copper in the United States and are producers of important quantities of molybdenum, gold, and some base metals. Extensive studies of this type of deposit in both production and exploration phases are in progress at Ray, Arizona; in Puerto Rico; and in the Cathcart Mountain area, Maine. Research on natural processes that produce ore de- posits continues in the east Tennessee zinc district and the San Juan Mountains of Colorado. Both studies are concerned with understanding spatial re- lationships, the chemical and thermal nature of solu- tions, their plumbing systems, and the sources of metals that were concentrated. Investigations of the geochemistry of ore deposits provide new insights into the processes responsible for anomalous concentrations of minerals. Continu- ing laboratory studies of mineral-phase equilibria have yielded information regarding physical and chemical conditions at the time of ore deposition. Field and laboratory studies are directed toward c—n ”(—7 MINERAL RESOURCES RESEARCH PROGRAMS IN THE U.S. GEOLOGICAL SURVEY 15 understanding the source of metals, the mode of transportation, and the reasons for localization of the metals. For example, mineralogic and sulfur isotopic studies at Homestake, S. Dak., have shown that gold can be related to original sedimentary ac- cumulations, thereby enlarging the area of search for additional deposits. GEOCHEMICAL EXPLORATION Geochemical exploration programs are based on the systematic measurement of one or more chemical properties of naturally occurring materials, and the discovery and delineation of chemical patterns that may be related to potential mineral deposits. One approach to a better understanding of how geo- chemical exploration can be more useful in the search for mineral deposits is to evaluate and com- pare data from various kinds of samples (rock, soil, stream sediment, gases) collected in known mineral districts or from other localities thought to have a high potential for new deposits. Such investigations are under way in a variety of climatic environments (humid-tropical, humid—temperate, arid, and al- pine). These studies will help to establish the best type of sample to collect in order to find a particular type of mineral deposit in a particular climatic en- vironment. Other geochemical investigations are oriented to- ward model studies of individual mineral deposits or mineral districts. These studies characterize the pat- terns of major, minor, and trace elements found in rocks in and around known mineralized areas. The characteristic chemical dispersion patterns can then be searched for in other areas to locate new deposits, especially those that are deeply buried under rock, soil, water, and ice. Other studies deal with the chemical effects of weathering processes on mineral deposits and how these changes can be used to locate new deposits. Studies of the abundances of various trace elements in vegetation, in surface and ground waters, and in specialized types of samples such as chemically pre- cipitated coatings on rock surfaces and gases in the atmosphere and soils should provide a better under- standing of how various chemical elements move under natural conditions and, therefore, how geo- chemists can best use this information in the search for new mineral resources. Studies are also being conducted in the applica- tion of computer technology to geochemical data. Many surveys now involve the analysis of thousands of samples for as many as 30 or 40 separate chemi- cal elements. Such a large set of data can be evalu- ated adequately only by computer. New statistical and graphic concepts are also being applied to geo- chemical data sets to evaluate with greater con- fidence the mineral resource potential of selected areas in the United States. New analytical methods being tested are needed to fulfill the new search specifications. The emphasis is on methods that are cheaper, faster, and more accurate. Many chemical elements that in the past were not routinely sought in samples can now be determined, thereby increas- ing the possibilities of finding new resources of these elements. GEOPHYSICAL EXPLORATION Geophysical exploration is aimed at the detection of subsurface mineral deposits where drilling and excavation are prohibitively expensive but where rapid appraisal of the possibilities for the occur- rence of mineral resources is required. The capa- bilities ‘of several geophysical methods are currently being tested and evaluated. Magnetic anomalies may denote unmapped or blind intrusive bodies that have associated mineral- ization, such as the porphyry copper deposits at Ruth, Nev., or they may directly reflect iron ore de- posits, as at Marmora, Ontario. Negative anomalies also may be significant where the magnetite content of the host rock has been greatly reduced by hydro- thermal alteration associated with ore emplacement, as at Cripple Creek, Colo. Electrical resistivity and electromagnetic surveys provide additional information about target areas delineated with gravity and magnetic surveys, par- ticularly in the identification of buried conductors which are indicative of massive sulfide deposits or in locating buried channels that may control uran- ium deposits. Radioactivity surveys are used not only to detect radioactive minerals directly but also for correlation of rock units in geologic mapping. Gamma-ray spec— trometers can be used to discriminate between radia- tion from uranium, thorium, and potassium sources. For example, radioactive potassium associated with hydrothermal alteration that occurred during min- eralization can be studied. Several methods for measurement of natural earth currents are proving to be useful and economical reconnaissance techniques in exploring for conduc- tive anomalies that may be associated with geother- mal systems, which are in turn associated with large self-potential anomalies. These methods of- fer promise for the development of improved min- eral resource exploration techniques for the future. 16 MINERAL RESOURCE PERSPECTIVES 1975 Remote-sensing experiments using a multispectral 4-band camera and thermal infrared scanner are underway over areas of known base-metal and uranium mineralization to study resolution of subtle surface-alteration patterns. In studies of the Gold- field area in Nevada, a combination of digital com- puter processing and color compositing of LAND- SAT multispectral scanner images were used to enhance spectral reflectance differences; hydrother- mally altered areas associated with mineral deposits were detected and mapped (Rowan and others, 1974). The technique of total-field resistivity mapping promises to be a rapid yet good-resolution recon- naissance approach for deep resistivity studies. An improved electromagnetic technique designated ELF (extra low frequency) is being developed which uses a ground transmitter as a multiple-frequency source and an aircraft to carry the receiver. In addition to providing a reliable signal along with rapid, sys- tematic coverage, the depth of penetration for a large target is about 500 metres in areas of resistive country rock. The application of paleomagnetism to the study of time and sequence of uranium deposition is being investigated. Samples of the ore zones, alteration halos, and host rocks are analyzed to detect signifi- cant variations in magnetization relative to varying intensities and directions of magnetic fields in past geologic history. Progress is being made in interpretation of geo- physical data using electronic computers. Highly ef- fective computer programs for automatic interpre- tation of two- and three-dimensional models from both d.c.-resistivity and electromagnetic data are being developed, such as techniques for making auto- matic depth determinations and for calculating the pseudogravity field from magnetic data. More rigor- ous mathematical treatment of filtering is being ap- plied to magnetic data in an attempt to separate superimposed magnetic anomalies, the sources of which are at different depths below the surface. This should aid in detecting buried stocks that may be mineralized in areas where volcanic rocks are at the surface. RESOURCE ANALYSIS Research in resource analysis is a relatively new component of the mineral resource appraisal pro- gram. The primary emphasis is on improving data storage and retrieval capability and developing sys- tems for resource estimation and prediction. The development and expansion of the computerized re- source information bank (CRIB), which is the pri- mary storage and retrieval system for mineral re- source information in the US. Geological Survey (Calkins and others, 1973) is focused on insuring CRIB’s compatibility with data systems in other State as well as Federal agencies. Application of computer graphics to geologic-data solutions and displays is also being examined. Geostatistics and computer theory are applied to seeking mathematical models of resource occurrences, the economics of the exploration process, and resource convertability to reserves. MINERAL RESOURCE STUDIES BY SPECIALISTS An integral part of the mineral resource appraisal program is an intensified emphasis on long-term studies of specific critical commodities—an approach to resource studies that has served the Nation ef- fectively in the past. The comprehensive assessment of resources of each mineral commodity on a nation- wide scale, presented in Professional Paper 820 (Brobst and Pratt, 1973), could not have been pre- pared without the specialized knowledge of several dozen commodity experts, each of Whom had de- voted a major part of his professional career to field, laboratory, and library research on the econ- omic geology of one or a few mineral commodities. Understanding of the US. resource base now re- quires global resource knowledge. In addition to es- tablishing priorities for national commodity-study emphasis, the specialist program provides oppor- tunities for field studies at home and abroad, lab- oratory support, and technical services to broaden the experience and competence of individual special- ists. Mineral resource specialists assigned to all the commodities of national significance must maintain up-to-date mineral information for program plan- ning as well as for advisory purposes. Commodities currently covered by the mineral-resources—specialist program are listed in table 4. STATE REPORTS REQUESTED BY THE US. SENATE The first of the modern Federal documents on the mineral and water resources of various States, prepared at the request of the Interior and Insular Affairs Committee of the US. Senate, was the US. Geological Survey report on Wyoming submitted to Senator Gale McGee in 1960. Between 1963 and 1974 the Geological Survey, at Senatorial request, provided reports on Montana, Alaska, Colorado, Idaho, Nevada, South Dakota, Eastern Montana, 5:; ‘_A__.« MINERAL RESOURCES RESEARCH PROGRAMS IN THE U.S. GEOLOGICAL SURVEY 17 TABLE 4.—Commodities in the mineral-resources-speeialists program Aluminum Niobium Barite Nitrates Beryllium Peat Boron Phosphate Bromine Platinum-group Chromium Potash Clays Rare earths Copper Rhenium Feldspar Salt Fluorine Scandium Gold Silver Graphite Stone, construction Iron ore Sulfur Kyanite Talc Lead Tantalum Lightweight aggregates Tin Limestone and dolomite Titanium Lithium Tungsten Mercury Vermiculite Mica Zeolites Molybdenum Zinc Nickel New Mexico, Washington, California, Missouri, Ore- gon, Arizona, Utah, and North Dakota. With the exception of the original Wyoming report, these were prepared by the U.S. Geological Survey in col- laboration or cooperation with the respective State geological or mining agencies, and in some cases with State universities or other Federal or State agencies, such as the U.S. Bureau of Mines and Bureau of Reclamation. All were published orig- inally as U.S. Senate Documents or Committee Prints, but some were reprinted separately as re- ports of the various State agencies. Because of limi- tations of staff and the short time available for preparation (generally only a few months), these reports have not entailed new field surveys but have relied on the collective experience of numerous professionals in the various agencies. Early in 1974, the U.S. Geological Survey was requested to revise and update the report on South Dakota, originally published in 1964. These reports are a long-term obligation that has been met as Survey staff, funding, and scientific support from respective State geologic agencies became available. MINERAL APPRAISAL 0F FEDERAL LANDS Appraisal of the mineral resources of Federal lands is the responsibility of the Government and was a cardinal point in the legislation that created the U.S. Geological Survey. In the past, legislation of two types has directed such appraisal. The Min- eral Leasing Acts of 1920 and 1947 provided for the appraisal of identified leasable minerals—energy minerals, potash, sodium, phosphate, and sulfur—on Federal lands. The Wilderness Act of 1964 provided for the appraisal of mineral resources of all types on primitive and wilderness-type lands prior to their withdrawal from future mineral entry. Currently, the anticipation of increased monetary returns from mineral leasing under the proposed act to replace the Mining Law of 1872 emphasizes the need for sys- tematic appraisal of mineral resources of all Federal lands. Federally owned or federally managed lands com- prise about 762 million acres, or about one-third of the Nation; Almost half of the federally-owned lands (353 million acres) is in Alaska; almost 90 percent of the remainder (358 million acres) is in 11 western States (table 5), and about 43 million acres of that lie in States east of the Rocky Mountains. The prin- cipal agencies responsible for management of Fed- eral lands are shown in table 6. The Bureau of Indian Affairs manages about 50.5 million acres of Indian lands, which are not federally owned. TABLE 5.———Federally owned land in 11 western States and Alaska as of June 30, 1971 [From U.S. Bureau of Land Management, 1972] Federal lands Percent of (millions of total area State acres) of State Arizona _________________________ 31.9 44.0 44.9 44.8 23.9 36.0 33.8 63.8 27.6 29.6 60.8 86.5 26.0 33.5 32.2 52.3 34.8 66.0 12.6 29.6 Wyoming ______________________ 30.0 48.1 Total in 11 Western States- 358.5 Alaska __________________________ 353.5 96.7 TABLE 6.——Major landholding agencies in the Federal Govern- ment Agency Millions of acres1 Department of the Interior Bureau of Land Management ____________ 474. Bureau of Sport Fisheries and Wildlife ____ 27.9 National Park Service ___________________ 24.5 Bureau of Reclamation __________________ 7.6 Bureau of Indian Affairs ________________ ’ 2.2 Department of Agriculture Forest Service _________________________ 186.8 Department of Defense ______________________ 30.3 1As of June 30, 1971 (from U.S. Bureau of Land Management, 1972). ‘As of June 30, 1973 (from Counselman, 1973): does not include 50.5 million acres of Indian-owned lands managed by B.I.A. Minerals from the Federal lands have contributed markedly to the industrial and economic development of the Nation and can be expected to continue to do so. According to the U.S. National Commission on Materials Policy (1973, p. 79), “In 1965 the western Public Lands produced over 90 percent of the Na- tion’s domestic copper, 95 percent of the mercury 18 MINERAL RESOURCE PERSPECTIVES 1975 and silver, and 100 percent of the nickel, molyb- denum, and potash, and about 50 percent of the lead.” Oil and gas production from onshore Federal and Indian lands has supported 5 to 6 percent of the Nation’s total domestic energy production in the recent past. During fiscal year 1974, receipts from all energy and mineral leasing activities conducted by the Bureau of Land Management, including rentals, royalties, and bonuses from lands and from the Outer Continental Shelf, totalled about 7 billion dollars. Not all Federal lands are open to mineral explora- tion development. As of June 30, 1967, more than 111 million acres had been withdrawn from mining, and more than 103 million acres from mineral leas- ing. Most of the withdrawn lands are in National Parks, Monuments, Fish and Wildlife Game Refuges, Indian Reservation-s, reclamation projects, military reservations, and scientific testing areas. The Wil— derness Act of 1964 provides that areas designated as Wilderness will be withdrawn from mining laws and mineral leasing activities by December 31, 1983. Other land-s not subject to leasing are the Naval Petroleum and Oil Shale Reserves, the National Park System, and a one-mile buffer zone around Naval Petroleum, oil shale, and helium reserves. In total, the Federal lands subject to mineral develop- ment aggregate about 822 million acres, including public lands not withdrawn, acquired lands, and lands which have been patented with minerals re- served to the United States. Submerged lands of the Outer Continental Shelf are also subject to mineral development. Once resources are identified on Federal land, the Bureau of Land Management and the Forest Service have the main responsibility for administering the laws concerning leasable and locatable minerals. Leasable minerals include energy minerals and po- tash, sodium, phosphate, and sulfur. Locatable min- eral-s include those metallic and nonmetallic mineral-s that may be located under the Mining Law of 1972. Mineral materials, largely sand, stone, gravel, pu- mice, and clay, are obtainable by competitive sale by contract. The US. Geological Survey is responsible for classification of public lands for their leasable mineral resources for retention-disposal, exchange, and multiple-use purposes. The US. Geological Sur- vey is also responsible for prelease resource evalua- tion and postlease administration of leasable min- erals. In its statutory role to assess mineral resources on Federal lands the US. Geological Survey cur- rently provides three levels of resource appraisal of specific tracts of Federal lands, depending on the time, personnel, and funds available (table 7). Level I is a library and record analysis that provides enough data to delineate those areas that are known to have promising mineral potential but is inade- quate to form the basis for decisions about the dis- position of lands. Level II is a more extensive in- vestigation, including field reconnaissance, which evaluates the mineral potential of areas and desig- nates potential targets for detailed studies or ex- ploration for possible future development. These re- sults are adequate for making land-use decisions with regard to mineral potential but not for leas- ing purposes. Level III consists of detailed studies required for resource evaluation and lease manage- ment decisions and mineral resource research on mining districts. Research at this level not only in- creases our understanding of the genesis, distribu- tion, and economic potential of metallic and non- metallic resources, but also contributes to the de- velopment of modification of exploration tools and techniques. The Cost of these levels of appraisal varies depending on the degree to which an area has already been developed, the amount of available geo— science data, the difliculty and expense of obtaining such data, and the length of time available to make the study. Much remains to be done before the mineral re- sources on all Federal lands have been appraised. Table 8 shows the magnitude of the task ahead; much is known, however, about the geology and mineral resources in many as yet unappraised areas because of US. Geological Survey, State, industry, university, and other studies. When the status of knowledge on these areas is determined, it will be possible to revise table 8; nevertheless, the task will remain a formidable one. In addition, appraisals will eventually be needed for 50.5 million acres of Indian land. Finally, future planners must also bear in mind that much of the 1,600 million acres of pri- vately owned lands, mostly in the eastern part of the United States, has never been adequately ap- praised. Following are summaries of the principal current US. Geologic-a1 Survey programs directed wholly or in part toward mineral resource appraisal of Fed- eral lands, exclusive of the continuing mineral classi- fication program. Wilderness Program—The US. Geological Sur- vey and the US. Bureau of Mines jointly have con- ducted mineral surveys on wilderness and proposed wilderness areas in National Forests under the Wilderness Act of 1964. The Survey evaluates the A A__.__¢'_I‘_-._‘_ A‘_. i 4.4. __ _‘ ._‘_._l MINERAL RESOURCES RESEARCH PROGRAMS IN THE U.S. GEOLOGICAL SURVEY 19 TABLE 7.—Three levels of resource appraisal used by the U.S. Geological Survey Objective Method of study Advantages Limitations I-__- General inventory of past production and resource actiVlty. Identify areas needing more detailed study. II--- Uniform reconnaissance- level appraisal to estab- lish base for total re- source estimate. III -- Sufficiently detailed geo- science data to determine reserves and to make management decisions regarding leases and en- vironmental consequences. Library and records survey; search of all sources for unpublished data on known districts; computer storage of data. Level I plus reconnaissance geologic, geochemical, and geophysical mapping; re- mote sensing; sampling of broad areas that are promising. Computer stor- age, retrieval. and inter- pretation of data. Detailed geologic mapping, geochemical sampling and assaying, and geophysical surveys in small areas of known potential. Computer storage, retrieval, and interpretation of data. Superficial; most data will be spotty; inadequate for determination of total re- source; often biased; un- developed areas will be overlooked. Status report: assessment of all known resource in- formation. Important first step for any assessment. Not sufficient to define re- serves or to be used by management. Sufficient detail to present resource evaluation for use of decision makers. New areas identified for classification and develop- ment. Deeply buried deposits missed. Development of geologic theory leading to identifi- cation of new types of de— posits; basis for stockpile decisions. Time consuming, expensive. TABLE 8.—-Mineral resource appraisal of Federal lands com- pleted to September 1974 Federal land area $111113]; Total Federal land areas --------------------- 762 Federal lands appraised at Level I Alaska (Native Claims areas) ____________ 80 Desert lands (Bureau of Land Management) - 1 Urban areas,1 San Francisco, California --__ 1 Federal lands appraised at Level II Wilderness lands (Forest Service) ---------- 15.4 Urban areas,1 Tucson, Arizona ------------ 5 Quadrangle appraisal mapping 1 ____________ _ Federal lands appraised at Level III 2 undetermined 1Includes some Indian and(or) private lands. ' 2The status of extensive U.S. Geological Survey mineral classrfication programs of the past 70 years, which routinely provided data about the potential mineral resources of Federal lands to Federal land-administering agencies, is not tabulated in this compilation. mineral potential of the areas on the basis of recon- naissance geologic, geochemical, and geophysical ex- aminations. The Bureau evaluates all previous and existing exploration and mining in the area, includ- ing examination of all mining claims. In addition, a minability appraisal is made where appropriate. As of September 1974, 447 primitive, wilderness, and proposed wilderness areas totaled about 38.7 million acres; 129 areas, about 15.4 million acres, had been evaluated. Studies of the remaining areas are scheduled to be completed by January 1, 1983. In January 1975, 16 new wilderness areas, totalling 212,618 acres of the Eastern United States, were established, and 17 proposed wilderness areas were designated for study. Desert land studies—An evaluation of the geology of the Randsburg-Searles Lake area, California, is the first of several studies that will provide infor- mation on the known mineral resources and mining activities of the area, which information will be used by the Bureau of Land Management as a basis for classifying these Federal lands for protection of the desert environment. Less than 1 million acres have been appraised in this new program. Urban studies—Reconnaissance geologic mapping and compilation of previous geologic work is being conducted in the Tucson area, Arizona, much of which is composed of Federal or Indian lands. Data on the mineral potential are based on past produc- tion and unpublished material contributed by the mining industry. The resulting maps will provide State and local planners with factual data for land- use decisions. Southeastern Arizona is one of the world’s richest copper provinces, and it currently faces the problems of industrial growth, urbaniza- tion, and protection of environmental values. Ap- proximately 5 million acres have been examined in this program. Quadrangle appraisal mapping—Reconnaissance mapping and local detailed mapping of selected 2° quadrangles, large segments of which are Federal lands, include extensive geophysical surveys and some geochemical sampling. The studies are de- signed to test the potential of areas that have known promise as mineral belts and to determine parts of those areas that warrant more detailed studies, as well as to evaluate other geologic environments which elsewhere in the world have produced mineral resources. Indian land studies—A new program to sum- marize existing information on the geology and mineral resources of Indian lands was begun in 1975 in cooperation with the Bureau of Mines and 20 MINERAL RESOURCE PERSPECTIVES 1975 Bureau of Indian Afi‘airs. Compilation of resources summaries of nearly 10 million acres of Indian lands is in progress. Some of these lands contain large resources of fuels and minerals, and their orderly de- velopment is essential to the welfare of Indian peo- ple and the Nation’s economy. ' Alaska land appraisal.—Alaska is almost equal in size to one-fifth of the total area of the conterminous United States. Although a large part of Alaska (353.5 of its 365.5 million acres) was Federally owned land as of June 30, 1971, the land ownership pattern is changing significantly. By 1978, as re- quired by the Alaska Statehood and Alaska Native Claims Settlement Acts, the State will own 103 mil— lion acres and Natives will own 40 million acres. Proposals for adding about 80 million acres to the National Park, Wildlife Refuge, Forest, and Wild and Scenic River Systems have been submitted for Congressional consideration under provisions of the Alaska Native Claims Settlement Act. Much of the land included in these proposals may be withdrawn from mining and mineral entry even though some of it includes parts of potential mineral belts. Because of these pending changes, the US. Geo- logical Survey’s Alaska mineral resource appraisal program is attempting to increase the rate of ap- praisal as well as its quality and scope. Started in July 1974, the program describes and evaluates energy, metal, and nonmetal mineral resources and their potential for discovery. CONCLUSIONS: SOME PRESSING PROBLEMS Because there are so many minerals, the related problems are even more complex than those for energy. The national and international flow of min- erals, for example, and the possibilities for substi- tution of one mineral for another, are very complex phenomena that are not well understood. Public awareness of the need for many minerals has not yet been aroused because vital mineral uses are not obvious in a complex society, but sudden shortages would affect our personal lives in unforeseen ways. Prior planning and critical evaluation are as neces- sary to the development and utilization of nonfuel mineral resources as they are for the energy minerals, if potentially serious problems are to be avoided or at least minimized. DETERMINING CRITICALNESS 0F MINERAL COMMODITIES Objective means of evaluating the overall na- tional and international importance of each com- modity must be developed so that priorities for re- search may be assigned to various commodities and to the most productive lines of research for each. Factors to be considered include: 1. Importance of the commodity to industry and commerce, 2. Defense or other needs of overriding importance, 3. Extent to which the commodity can be replaced by others, 4. Extent of domestic reserves, 5. Extent of economic, technologic, and legal use of domestic resources, 6. Dependence on foreign imports, and 7. Political stability (internal or international) of foreign import sources. Every min-era] commodity being mined today is essential to some industrial process or product, but to assume that either government or industry can fully pursue the research needed on all mineral commodities would be far from realistic. DEPENDING 0N IMPORTED MINERALS Patterns in world mineral supply and demand which had persisted for some four decades began to change radically after World War II. In 1940, the United States produced and consumed nearly 50 per- cent of world mineral supplies, including fossil fuel, but in 1971 the United States’ share had dropped to about 27 percent, and it continues to diminish. Between 1945 and 1972, world consumption of 18 basic mineral commodities increased about six times, Whereas U.S. consumption less than doubled. Start- ing from a far lower per capita base, both the de- veloped and the more advanced developing countries are catching up in industrial production. Further- more, world population is multiplying rapidly, plac- ing increasing demands on limited mineral supply. The United States continues to be ever more de- pendent on foreign sources of supply for essential mineral raw materials. In 1972 and 1973, our over- all dependence on imports for 15 critical industrial materials other than energy was about 60 percent of our consumption (Council on International Eco- nomic Policy, 1974, figs. 2 and 15) ; the situation for individual materials ranged from total dependence to total independence. In 1974, the Nation was more than 90 percent dependent on imports of primary materials for four commodities, 75 to 90 percent dependent for 8 additional commodities, and 50 to 75 percent dependent for 8 commodities. Seventeen other major commodities are imported. Forecasts for the year 2000 indicate that we shall then be com- pletely dependent on imports for 12 commodities, more than 75 percent dependent for 19 commodities, ——§_ — ' I—'l_ —F—‘ 1 l V ' CONCLUSIONS: SOME and more than 50 percent dependent for 26 com- modities. Among causes of this increase are (a) the depletion of known domestic ore deposits, and (b) the “free market” economy, whereby it is cheaper to use imported supplies of some raw ma- terials than to process and use our own, even though we may have ample supplies. The increas- ing dependence on imports seems inescapable but by no means approaches the degree of dependency of any other major industrialized country with the ex- ception of the USSR, which is relatively self-suffi- cient in most essential commodities and which has, in the last five decades, made enormous investment in geologic research and prospecting in carefully planned programs. Rapidly evolving complex changes in the attitude of people and nations foreshadow fundamental changes in economic patterns throughout the world. Sparked by rising national consciousness and eco- nomic pressures, the industrially underdeveloped and developing nations are increasingly less willing to export only raw materials and would prefer to ex- port their mineral materials in at least the semi- finished stage. These nations possess many of the world’s major mineral deposits, upon which the in- dustrial world will depend during the next quarter century, and they are forming national companies to prospect for and exploit their ore deposits. Many highly industrialized nations, including France, Ja- pan, and Germany, are actively supporting these nation-a1 companies of the developing nations to in- sure continued raw material supplies for their own economies. Some industrialized nations, including those already mentioned, negotiate through their own government-supported industrial firms, thereby increasing government-to-government negotiations in this field. In contrast, United States industry has not entered into comparable joint ventures with the Federal Gov- ernment. The United States is thereby put into a less competitive position in negotiations abroad for access to available foreign supplies. And at the same time, environmental constraints on the activities of the mining and manufacturing industries in the United States are forcing a trend toward additional domestic dependence on foreign supplies of both raw and partly processed mineral products. Some options available to the United States to insure adequate supplies of mineral raw materials from abroad include: Diversification of overseas sources of supply. Government activities to assure access to foreign sources of supply, combined with steps by in- PRESSING PROBLEMS 21 dustry to increase overseas processing of raw materials. A major economic stockpiling program by industry or Government for those commodities, including energy materials, supplies of which are de- pendent on a few foreign sources. Restraint in consumption of the scarcer raw ma- terial supplies. Products and processes can be redesigned to use abundant rather than scarce materials. National and worldwide programs for recycling of metals. Bilateral technical assistance programs in mineral fields, financed by the United States. Placement of staff trained in mineral deposit and mineral industry development in United States Missions overseas to insure accurate, up-to-date information on present and potential mineral supplies. Fostering of close cooperation between Government and industry. The natural short-range View of industry, working toward an impressive balance sheet, must be melded with the natural long- range view of Government, working toward long continued economic and social progress and stability. STATE OF THE MINING INDUSTRY Keeping a finger on the pulse of domestic mining activities, an implied responsibility under provisions of the Mining and Minerals Policy Act of 1970, is very difficult because of the complexity of the mining industry. One possible measure of the state of health of the industry is the amount of resource explora- tion activity, and present exploration activity can be considered a barometer of the future mining activ- ity. A systematic quantified method to obtain this information should be developed jointly with the industry. Improved liaison will lead to the develop- ment of a more accurate resource data base. REPORTING RESERVE DATA The difficulty of making reliable estimates of mineral resources has been compounded by the long- standing, but declining, practice of many mining companies to maintain secrecy about the magnitude of their reserves in the ground. Public release of such data obviously could be disadvantageous to corporations in the competitive business atmosphere. Reliable reserve and resource estimates, however, are increasingly needed for use in making policy decisions both in corporate board rooms and Govern- 22 MINERAL RESOURCE PERSPECTIVES 1975 ment conference rooms. Adequate reporting of such information can be done in strictest confidence so as not to prejudice private interests. RECOVERING POTENTIAL MINERAL BYPRODUCTS Potential mineral byproducts or coproducts are being lost while their associated principal mineral products are being mined and recovered (Brobst and Pratt, 1973, p. 7—8). A few of these minerals are lost because of selective mining—the gold, silver, selenium, and tellurium that remain behind during in—place leaching of copper deposits. Most of them are lost in selective processing—vanadium in mag- netite deposits; fluorine, vanadium, uranium, scan- dium, and rare earths in marine phosphorites ; cad- mium, bismuth, cobalt, and mercury in lead ores; and several metals in coal ash. Industry or public research and development are needed to identify these escaping resources and determine ways of re- covering them. Other commodities that have poten- tial for recovery as byproducts include aluminum and soda ash from oil shale; vanadium and nickel from crude oils; gold and tungsten from sand and gravel operations; trace metals from brines and mine waters; and several metals from seawater and sea bed deposits. CONSIDERING SUBSEA MINERAL RESOURCES As offshore drilling for oil becomes more common, the potential for other mineral resources under the sea will receive more public attention. Neither the continental margins nor the deep seabed resource potentials can be ignored in future appraisals of the world’s mineral resources. The continental margins, being geologically part of the continents, contain in general the same kinds of resources as the contin- ents and have contributed some mineral production, whereas the deep seabed, being composed largely of basaltic igneous rock covered by clays and other fine-grained sediments, contains a much smaller variety of minerals—chiefly metals occurring in manganese nodules, pavements, and crusts, and de- posits that are genetically related to mafic igneous rocks. Minerals that have been produced from con- tinental margin rocks mainly by means of shafts ex- tending from on-land to beneath the seabed include coal, iron ore, nickel, copper, tin, gold, mercury, barite, and limestone. In recent years only coal, iron ore, and barite were being produced in significant amounts by this method. Barite, although a bedrock deposit, was mined from an open-pit operation be- neath the sea. Surficial deposits on the world’s continental shelves include sand and gravel, calcium carbonate in the form of shells, precious coral, iron sands, placer deposits containing tin, diamonds, titanium sands, zircon, monazite, and gold and other heavy minerals, and phosphorite nodules. Sand and gravel is by far the most important by volume. No minerals have yet been produced commercially from the deep seabed, but commercial manganese- nodule production will likely begin before 1980. Metalliferous brines and muds also have some po- tential in the deep seabed, but little is known about them, and any possible commercial recovery seems several decades in the future. STAFFING FOR MINERAL APPRAISAL AND EXPLORATION Continued appraisal of and exploration for min- eral resources require a continuing supply of geol- ogists, geochemists, and geophysicists as well as mineral economists. Student enrollments in these disciplines have decreased steadily in recent years, and at the same time employment demands have increased sharply. Since the National Science Foundation’s Register of Scientific and Technical Personnel was discon- tinued in 1970, no system exists for monitoring em- ployment demands and predicting trends. Informal observations made to the American Geological Insti- tute by university department heads, who are prob- ably in the best position to observe where hiring emphasis is being placed from year to year, indicate a recent large upsurge in demand for Earth scien- tists, particularly at the Batchelor’s and Master’s levels, although the number of Master’s candidates decreased by 4 percent and the number of Doctoral candidates decreased by 5 percent in 1974 (Hender- son, 1974). Timely availability of adequate numbers of university graduates at various academic levels requires that employment needs be anticipated 2 to 7 years in advance, but the mineral industry generally has not been able to anticipate its need for Earth scientists by more than a few years. The level of exploration activity seems to be the index by which to measure the need for Earth scien- tists. Long-term increases in exploration activity can be anticipated reasonably because of the vast in- crease in mineral production that must occur during the next decade or so. Over the past 7 years, student enrollment evidently has conformed quite closely to demand or kept ahead of demand in a market- place in which no mineral crises and no sudden large upsurges in mineral exploration activity have oc- JA8_ ___4. ‘ 4 4_«__ ¢_‘____..__‘§_1___. _¢_u_ i... ‘ .fijeg'w:fi '1 I ~v . :3 iii-"#7 v.1. Of ,. CONCLUSIONS: SOME PRESSING PROBLEMS 23 curred. Now, with the existing crisis in energy and impending crises for several metals, the educational process may not be able to keep abreast of demand. It seems likely that anticipated increases in mineral- exploration activity may result in a near-term short- age of geologists and a chronic shortage of geo- physicists and geochemists. The identification of new employment requirements and support for short- term retraining programs are probably areas in which Government, alone or jointly with industry, can assist in meeting changing needs. INCREASING PUBLIC AWARENESS Finally, the popular misconception that a steady supply of minerals from the crust of the Earth is simply a matter of favorable economics and tech- nology has induced widespread public complacency. This notion ignores that fundamental factor govern- ing mineral supply: geologic availability. Neither technologic magic nor astronomical dollar value can make it possible to extract iron, aluminum, gold, sulfur, or phosphorus from rocks in which they are not present. Our total resources in 1975 are vast, but they cannot be mined, much less used, until they have been identified, appraised, and finally moved into the category of reserves. We must begin now, in both industry and Government, to inform the public about the real nature of our minerals problem, and to stimulate the research that will make our mineral resources available. REFERENCES CITED Allais, M., 1957, Method of appraising economic prospects of mining exploration over large territories; Algerian Sahara Case Study: Management Sci., v. 3, no. 4, p. 285—347. Arps, J. J., Mortada, M., and Smith, A. E., 1970, Relationship between proved reserves and exploratory effort: Soc. Petroleum Engineers, SPE paper 2995. Bieniewski, C. L., Persse, F. H., and Brauch, E. F., 1971, Availability of uranium at various prices from resources in the United States: U.S. Bur. Mines Inf. Circ. 8501, 92 p. Blondel, F. A. J ., and Lasky, S. G., 1956, Mineral reserves and mineral resources: Econ. Geology, v. 51, no. 7, p. 686—697. Brinck, J. W., 1972, The prediction of mineral resources and long-term price trends in the non-ferrous metal mining industry: Internat. Geol. Cong., 24th, Montreal, 1972, Proc. Sec. 4—Mineral deposits, p. 3—15. Brobst, D. A., and Pratt, W. P., 1973, eds., United States min- eral resources: U.S. Geol. Survey Prof. Paper 820, 722 p. Calkins, J. A., Kays, Olaf, and Keefer, E. K., 1973, CRIB—the mineral resources data bank of the U.S. Geological Survey: U.S. Geol. Survey Circ. 681, 39 p. Council on International Economic Policy, 1974, Special Re- port—Critical Imported Materials: Washington, D.C., U.S. Govt. Printing Office, 49 p. Counselman, B. J., compiler, 1973, Annual report on Indian Lands: Department of the Interior, Bureau of Indian Affairs, 66 p. Derry, D. R., 1973, Potential ore reserves—an experimental approach: Western Miner, v. 46, no. 10, p. 115—122. Harris, D. P., Freyman, A. J., and Barry, G. S., 1970, The methodology employed to estimate potential mineral supply of the Canadian Northwest—an analysis based upon geo- logic opinion and systems simulation: Canada Dept. En- ergy, Mines and Resources, Mineral Resources Branch, Mineral Rept. MR—105, 56 p. Henderson, B. C., 1974, Student enrollment rises somewhat: Geotimes, v. 19, no. 11, p. 20—21. Hubbert, M. K., 1962, Energy resources: Natl. Acad. Sci.— Natl. Research Council Pub. 1000—D, 141 p. Hubbert, M. K., 1967, Degree of advancement of petroleum exploration in United States: Am. Assoc. Petroleum Geol- ogists Bull., v. 51, no. 11, p. 2207—2227. Hubbert, M. K., 1969, Energy resources, in Resources and Man: San Francisco, W. H. Freeman and 00., p. 157—242. Lowell, J. D., 1970, Copper resources in 1970: Mining Eng., v. 22, no. 4, p. 67—73. McKelvey, V. E., 1972, Mineral resource estimates and public policy: Am. Scientist, v. 60, no. 1, p. 32—40; reprinted in Brobst and Pratt, 1973, p. 9—19. Meadows, D. H., Meadows, D. L., Randers, J¢rgen, and Behrens, W. W., III, 1972, The limits to growth: New York, Universe Books, 205 p. Mining and Minerals Policy, 1973, Report of the Secretary of the Interior under the Mining and Minerals Policy Act of 1970 (PL. 91—631), 2nd Ann.: Washington, D.C., U.S. Govt. Printing Office, 73 p. plus pt. 2—Appendices. Mining Magazine, 1974, International mining survey: Mining Mag. v. 131, no. 3, p. 191—219. Moore, C. L., 1966, Projections of U.S. petroleum supply to 1980: Washington, D.C., U.S. Dept. Interior, Oflice of Oil and Gas, 13 p. Moore, C. L., 1970, Analysis and projection of historic patterns of U.S. crude oil and natural gas, in Future petroleum provinces of the United States—A summary: Washington, D.C., Natl. Petroleum Council, p. 133~138. National Academy of Sciences, and National Research Council, 1975, Mineral resources and the environment: Washington, D.C., Natl. Acad. Sci., 348 p. Nolan, T. B., 1950, The search for new mining districts: Econ. Geology, v. 45, p. 601—608. Pratt, W. P., and Brobst, D. A., 1974, Mineral resources— potentials and problems: U.S. Geol. Survey Circ. 698, 20 p. Rowan, L. C., and others, 1974, Discrimination of rock types and detection of hydrothermally altered areas in south- central Nevada by the use of computer-enhanced ERTS images: U.S. Geol. Survey Prof. Paper 883, 35 p. Slichter, L. B., Dixon, W. J., and Myer, G. H., 1962, Statistics as a guide to prospecting, in Computer short course and symposium on mathematical techniques and computer applications in mining and exploration: College of Mines, Univ. Arizona, p. F11—F27. U.S. Bureau of Land Management, 1972, Public Land Statis- tics: Washington, D.C., U.S. Govt. Printing Office, 101 p. U.S. Bureau of Mines, 1970, Mineral facts and problems: U.S. Bur. Mines Bull. 650, 1291 p. 24 MINERAL RESOURCE PERSPECTIVES 1975 U.S. Bureau of Mines, 1974a, Bureau of Mines research 1973— A summary of significant results in mining, metallurgy, and energy: Washington, D.C., U.S. Govt. Printing Oflice, 107 p. U.S. Bureau of Mines, 1974b, Commodity data summaries 1974—Appendix I to mining and minerals policy—Third annual report of the Secretary of the Interior under the Mining and Minerals Policy Act of 1970: 193 p. U.S. Bureau of Mines, and U.S. Geological Survey, 1947, Mineral position of the United States, Appendix to Hear- ings before a subcommittee of the Committee on Public Lands, United States Senate, 80th Congress, first session, on investigation of the factors affecting minerals, fuels, forestry, and reclamation projects, May 15, 16, and 20, 1947: Washington, D.C., U.S. Govt. Printing Office, p. 167—338. U.S. Geological Survey, 1975, Geological Survey Research 1974: U.S. Geol. Survey Prof. Paper 900, 349 p. U.S. National Commission on Materials Policy, 1973, Material needs and the environment today and tomorrow: Wash- ington, D.C., U.S. Govt. Printing Oflice, variously paged. *U.S. GOVERNMENT PRINTING OFFICE: 1975 0‘585—475/136 3 8 t‘_-£l._m #4- AA’.__ 1‘4A -‘#_~ Q “A L; "IA. .-#Q .44; Q‘ A- REGION GEOLOGICAL SURVEY PROFESSIONAL PAPER 9417A ‘ . done in roonernfion wifh I] S nannernnO n‘ Hnndnn nnd llrhnn nnvnlnnmnnfi nflirn n; Pnlirv nnvnlnnmnno nnrl DACAflrrl" COVER PHOTOGRAPH of San Francisco Bay Region taken April 14, 1972, at altitude of 65,000 feet from U—2 aircraft. Courtesy National Aeronautics and Space Administration (Ames Research Center, Moffett Field, Calif.) Front shows city of San Francisco and Golden Gate at bottom, San Francisco Bay and city of Oakland in middle, Sacramento-San Joaquin Delta and crest of Sierra Nevada at top. Back shows Bolinas lagoon and trace of San Andreas fault at bottom, San Pablo Bay in middle, Sacramento valley and crest of Sierra Nevada at top. Studies for Seismic Zonation of the San Francisco Bay Region Edited by R. D. BORCHERDT BASIS FOR REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION, CALIFORNIA GEOLOGICAL SURVEY PROFESSIONAL PAPER 941—A A series of closely related earth science studies that define the nature and severity of earthquake hazards associated with geologic conditions jointly supported by the US. Geological Survey and the Department of Housing 69’ Urban Development, Office of Policy Development and Research, as a part of a program to develop and apply earth—science information in support of land—use planning and decisionmahing. UNITED STATES GOVERNMENT PRINTING OFFICE,WASHINGKTON‘:1975 UNITED STATES DEPARTMENT OF THE INTERIOR ROGERS C. B. MORTON, Secretary GEOLOGICAL SURVEY V. E. McKelvey, Director Library of Congress catalog-card No. 75—15239 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 Stock Number 024-001-02676-7 FOREWORD Earthquake hazard reduction program This report represents a milestone in the evolution of methods for reducing the hazards of earthquakes. The great Alaska earthquake of 1964 triggered an aware- ness among public officials of the seriousness of the earthquake hazard to many of the Nation’s major cities. If the effects of the Alaska earthquake are used as a gage, it is clear that when a major earthquake hits California cities such as Los Angeles or San Francisco, casualties could be in the tens of thousands and damage could be in the tens of billions of dollars. After the Alaska earthquake, the US. Geological Survey began to focus its diverse earth-science capabilities more specifically toward the goal of reducing earthquake hazards. The possible effective- ness of land-use planning to avoid the most serious hazards began to be recognized as a supplement to the common practice of incorporating earthquake—resistant designs into structures. For decades geologists had known, for example, that structures built astride the San Andreas fault were in jeopardy, but only in a few places had the fault been delineated in sufficient detail to serve as a guide to community officials and developers. Even if the fault data had been available, standard procedures were inadequate for translating the data into land-use plans or actions. Indeed, land-use planning was, and still is, in an early phase of evolution in the United States. No national land-use policy has been adopted. In order to satisfy some of the most urgent needs for basic data, several projects were started after the Alaska earthquake. The entire 1,400-km (868-mi) length of the San Andreas fault was mapped for the first time on the best available topographic base maps. Nets of closely spaced seismic instruments were installed in an experimental field laboratory along the San Andreas fault to study the basic mechanisms of earthquakes and the patterns of energy radiation and attenuation as earthquake waves pass through different types of rocks and soil. New laboratory studies were initiated to explore the physical principles of earthquakes. Re— search demonstrated the feasibility of earthquake prediction and of earthquake control and modification. The science of earthquakes is complex, requiring data and research in seismology, geology, soil mechanics, geophysics, hydrology, and engineering. Nevertheless, if earthquake hazards are to be reduced, earth-science data must be translated from scientific and technical language into a form that can be used effectively in the decisionmaking process. The San Francisco Bay Region Environment and Resources Planning Study Out of this recognition of the need to use earth-science information in regional planning and decisionmaking came an experimental program—the San Francisco Bay Region Environment and Resources Planning Study. The study, begun in January. 1970, is jointly supported by the US. Geological Survey, Department of the Interior, and the Office of Policy Development and Research, Department of Housing and Urban Develop- ment. The Association of Bay Area Governments participates in the study and provides a liaison and communication link with other regional planning agencies and with county and local governments. ' Although the study focuses on the nine-county 7,400-mi2 San Francisco Bay region, it bears on a difficult issue that is of national concern—how best to accommodate orderly development and growth while conserving our natural resource base, insuring public health and safety, and minimizing degradation of our natural and manmade environment. The complexity, however, can be greatly reduced if we understand the natural characteristics of the land, the processes that shape it, its resource potential, and its natural hazards. These subjects are chiefly within the domain of the earth sciences: geology, geophysics, hydrology, and the soil sciences. Appropriate earth-science information, if available, can be rationally applied in guiding growth and development, but the existence of the information does not assure its effective use in the day-to-day decisions that shape development. Planners, elected officials, and the public rarely have the training or experience needed to recognize the significance of basic earth-science information, and many of the con- ventional methods of communicating earth-science in- formation are ill suited to their needs. The study is intended to aid the planning and decisionmaking community by (1) identifying impor- tant problemsthat are rooted in the earth sciences and related to growth and development in the bay region, (2) providing the earth-science information that is needed to solve these problems, (3) interpreting and publishing findings in forms understandable to and usable by nonscientists, (4) establishing new avenues of com— III 1'V FOREWORD munication between scientists and users, and (5) exploring alternate ways of applying earth-science information in planning and decisionmaking. Since the study was started in 1970, it has produced more than 70 reports and maps. These cover a wide range of topics: reduction of flood and earthquake hazards, unstable slopes, engineering characteristics of hillside and lowland areas, mineral and water resources management, solid and liquid waste disposal, erosion and sedimentation problems, bay water circulation patterns, and others. The methods used in the study and the results it has produced have elicited broad interest and a wide range of applications from planners, government officials, industry, universities, and the general public. Studies for seismic zonation of the San Francisco Bay region This report brings together the results of a number of earth-science studies that provide a basis for reducing earthquake hazards. The enormous amounts of energy released during large, or even moderate, earthquakes produce a complex chain of effects, most of which are potentially hazardous to man and his works. Many of these effects, such as fault displacement and ground shaking, are direct results of the earthquake. Others, like landslid— ing and liquefaction, result from the action of ground motion generated by earthquakes on unstable geologic units or structures. Still other effects result from the reaction of manmade structures to earthquake forces. Moreover, many of these effects are complexly interre- lated in ways that make analysis difficult; ground shaking, for example, may be amplified or reduced by local geologic conditions, and the level of ground shaking may determine whether or not landslides are triggered or liquefaction induced. The diversity and complexity of earthquake effects make the reduction of earthquake hazards extraor- dinarily difficult and require adoption of a coordinated and disciplined plan of attack. Such a plan must incorporate the efforts of many individuals and the skills and techniques of several professions. Earth scientists, structural and civil engineers, professional planners, elected officials, and private citizens are among those who are essential participants in the effort. This report is designed to provide the earth-science basis for such a comprehensive approach to reducing earthquake hazards. It brings together and correlates significant results from several fields of geology and from seismology and engineering seismology. Because these results are derived from the natural processes brought into play by a damaging earthquake, they are a logical starting point for an attack on the problem. The method outlined here for seismic zonation is applicable, with modifications, throughout the San Francisco Bay region and elsewhere in regions of high earthquake hazard. Its effectiveness, however, depends on the degree to which these results are used or applied. Although many of the research findings can be applied directly in hazard-reduction programs now underway, others suggest the need for continuing communication among participants in the hazard-reduction process and for the conduct of related research. Thus, while few readers will be prepared to apply all that is presented here, the contents should assist them in determining where additional expertise is needed. CONTENTS Page Page Foreword ________________________________________________ III Borcherdt, W. B. Joyner, R. E. Warrick, and J. F. Gibbs___-A52 Abstract __________________________________________________ A1 Liquefaction potential, by T. L. Youd, D. R. Nichols, E. J. Helley, Introduction _____________________________________________ 2 and K. R. Lajoie ________________________________________ 68 Faults and future earthquakes, by R. L. Wesson, E. J. Helley, Landslides, by T. H. Nilsen and E. E. Brabb J. ______________ 75 K. R. Lajoie, and C. M. Wentworth ______________________ 5 Predicted geologic effects of a postulated earthquake, by R. D. Estimation of bedrock motion at the ground surface, by R. A. Borcherdt, E. E. Brabb, W. B.Joyner,E.J. Helley,K. R. Lajoie, Page, D. M. Boore, and J. H. Dieterich ____________________ 31 R. A. Page, R. L. Wesson, and T. L. Youd ________________ 88 Differentiation of sedimentary deposits for purposes of seismic General conclusions ________________________________________ 95 zonation, by K. R. Lajoie and E. J. Helley ________________ 39 1906 intensity scale for San Francisco ______________________ 96 Response of local geologic units to ground shaking, by R. D. References cited ___________________________________________ 97 ILLUSTRATIONS Page FIGURE 1. Map showing apparent intensity of the 1906 earthquake in San Francisco _________________________________________ A3 2. Generalized geologic map of San Francisco _____________________________________________________________________ 4 3. Maps showing faults that may cause earthquakes or surface displacement in the San Francisco Bay region and sources of data on recent faulting _________________________________________________________________________________ 7 4. Diagrams showing four types of fault movement _________________________________________________________________ 12 5. Map showing zones of surface fault displacement associated with earthquakes during historic times, San Francisco Bay region _______________________________________________________________________________________________ 14 6. Photographs showing examples of surface displacement along the San Andreas fault, 1906 earthquake __________ 15 7. Photograph showing curb offset by creep on Calaveras fault _____________________________________________________ 15 8. Map showing distribution of documented fault creep in the San Francisco Bay region _____________________________ 16 9. Map showing epicenters of earthquakes of magnitude greater than or equal to 1, San Francisco Bay region, 1969—72 and block diagram showing relation between fault surface, earthquake epicenter, and earthquake focus ________ 18 10. Sketch showing vertical deformation of young sedimentary deposits in trench across Coyote Creek fault, southern California _________________________________________________________________________________________________ 20 11. Block diagram showing landforms developed along recently active strike- slip faults _______________________________ 21 12. Aerial photograph and map showing topographic features along a segment of the San Andreas fault near the Carrizo Plain, southern California _________________________________________________________________________________ 22 13. Photograph and sketch of faulted wave-cut platform exposed in sea cliff near Point Afio Nuevo ____ _________________ 24 14. Graph showing length of surface fault rupture in relation to earthquake magnitude _______________________________ 24 15. Map showing surface rupture associated with the 1906 earthquake along part of the San Andreas fault near Fort Ross, Calif. _______________________________________________________________________________________________ 26 16. Map showing surface ruptures associated with the 1971 San Fernando earthquake _________________________________ 27 17—24. Graphs showing: 17. Horizontal ground motion recorded at Pacoima damsite, 1971 San Fernando earthquake __ _______________ 33 18. Comparison of spectra for horizontal ground motion at Pacoima damsite, 1971 San Fernando earthquake ____ 34 19. Peak horizontal ground acceleration in relation to shortest distance to slipped fault as a function of earth- quake magnitude _________________________________________________________________________________ 34 20. Peak horizontal ground acceleration 1n relation to shortest distance to slipped fault for earthquakes of magni- tude 5.0 to 5. 9 as a function of site geology ___________________________________________________________ 35 21. Peak horizontal ground acceleration 1n relation to shortest distance to slipped fault for earthquakes of magni- tude 6.0 to 6.9 as a function of site geology ___________________________________________________________ 35 22. Peak horizontal ground velocity in relation to shortest distance to slipped fault as a function of earthquake magnitude and site geology _________________________________________________________________________ 36 23. Peak horizontal ground displacement in relation to shortest distance to slipped fault as a function of earth- quake magnitude and site geology ____________________________________________________________________ 36 24. Comparison of attenuation curves for four representative finite- element models of strike- s-lip faulting with data on peak acceleration from figure 19 _____________________________________________________________ 38 25—29. Maps showing. 25. Location of detailed study area ________________________________________________________ . ________________ 40 26. Topography of Mountain View—Sunnyvale area ___________________________________________________________ 41 V VI CONTENTS FIGURES 25—29. Maps showing—Continued page 27. Soil units in Mountain View—Sunnyvale area ___________________________________________________________ 42 28. Alluvial deposits in Mountain View—Sunnyvale area ______________________________________________________ 44 29. Geology of Mountain View—Sunnyvale area ____________________________________________________________ 46 30. Example of unpublished engineering data from which thickness and physical properties of alluvial deposits were partly derived ___________________________________________________________________________________________________ 48 31. Schematic cross section of southern San Francisco Bay region and description of certain physical properties of the generalized geologic units _________________________________________________________________________________ 49 32. Correlation diagram showing groupings of geologic units for evaluating ground response and liquefaction potential ,1 50 33. Map of San Francisco Bay region showing distribution of generalized geologic units and locations where ground mo- tions generated by nuclear explosions were recorded _________________________________________________________ 53 34. Recordings of horizontal ground motion generated by two nuclear explosions _______________________________________ 54 35. Graphs showing horizontal spectral amplifications computed for 13 recording sites in San Francisco _________________ 57 36. Histograms of average horizontal spectral amplification values computed for sites underlain by bay mud, alluvium and either granitic rocks or rocks of Franciscan Formation ___________________________________________________ 58 37. Graphs showing horizontal and vertical spectral amplification curves computed from recordings of a nuclear explosion and recordings of the San Francisco earthquake of March 22, 1957 ___________________________________________ 59 38. Graph showing observed intensity for sites underlain by bedrock as a function of distance to zone of surface rupture for 1906 earthquake _________________________________________________________________________________________ 60 39. Graph showing increments in 1906 intensity as a function of average horizontal spectral amplification computed at corresponding sites from recordings of nuclear explosions _____________________________________________________ 61 40. Map showing maximum earthquake intensities predicted for San Francisco _______________________________________ 62 41. Schematic geologic section at site of down-hole seismometer array near margin of San Francisco Bay __________________ 63 42. Recordings of horizontal ground motion from a down-hole seismometer array _____________________________________ 64 43. Records showing comparison of observed and computed surface ground motions determined from motions observed in bedrock at a depth of 186 m _______________________________________________________________________________ 65 44—47. Graphs showing: 44. Computed and observed spectral amplification of ground motion at site of down—hole seismo meter array "H 67 45. Comparison of high- and low—strain spectral amplification for site underlain by 220 m of alluvium (input scaled to maximum velocity of 31 cm/s) _____________________________________________________________ 67 46. Comparison of high- and low-strain spectral amplification for site underlain by 11 m of bay mud and 173 m of alluvium (input scaled to maximum velocity of 19 cm/s) ___________________________________________ 67 47. Comparison of high- and low-strain spectral amplification for site underlain by 11 In of bay mud and 173 m of alluvium (input scaled to maximum velocity of 12 cm/s) ___________________________________________ 67 48. Graph showing estimated stress ratio required to produce liquefaction under field conditions during 30 cycles of seis- mic loading ______________________________________________________________________________ , ________________ 70 49. Graph showing criteria used for estimating liquefaction potential in the field _____________________________________ 71 50. Preliminary map showing liquefaction potential for the southern San Francisco Bay region _________________________ 73 51—59. Photographs showing: 51. Earthquake-generated landslide, Hebgen Lake, Mont. ___________________________________________________ 75 52. Aerial view of Turnagain Heights landslide, Anchorage, Alaska ___________________________________________ 75 53. Aerial view of debris avalanche that destroyed towns of Yungay and Ranrahirca, western Peru ____________ 76 54. Aerial view of landslides in Lopez Canyon area north of San Fernando, Calif. _____________________________ 76 55. Transverse cracks that formed in upper part of old landslide near Paicines, Calif., after a moderate earthquake _______________________________________________________________________________________ 76 56. Earthflow-type landslides triggered by earthquake of April 18, 1906, near Half Moon Bay, Calif __________________ 77 57. Roadbed of Ocean Shore railroad, south of San Francisco _________________________________________________ 77 58. Landslide that occurred about 6 km north of Bolinas Lagoon, Marin County, Calif, in March 1907 ____________ 77 59. Landslides in artificial fill along the shore of Lake Merced, near San Francisco, Calf, triggered by the San Francisco earthquake of March 22, 1957 ______________________________________________________________ 78 60. Block diagram showing nomenclature of parts of a landslide ,‘ ____________________________________________________ 78 61. Block diagrams showing common types of landslides in the San Francisco Bay region _____________________________ 79 62. Map showing distribution of landslides in San Mateo County, Calif ________________________________________________ 82 63—66. Maps of part of northeastern Contra Costa County, Calif, showing: 63. Distribution of landslide deposits _______________________________________________________________________ 83 64. Generalized slope ,11____,,____,____-11--__,1___,-___-,,1___,,1___,,__,VU__-,1__,H_-____V,,n____,,i__ 84 65. Distribution of geologic units susceptible to landsliding ____________________________________________________ 85 66. Relative slope stability _________________________________________________________________________________ 86 67. Map of part of San Francisco Bay region showing location of demonstration profile and estimated length of surface rupture along the San Andreas fault for a postulated earthquake of magnitude 6.5 _____________________________ 89 68. Diagrammatic section showing predicted geologic effects of a postulated earthquake of magnitude 6.5 ____________ 91 69. Diagrammatic section showing Fourier amplitude spectra of ground shaking estimated for an earthquake of mag- nitude 6.5 _________________________________________________________________________________________________ 92 TABLE CONTENTS VII TABLES Page Faults with Quaternary displacement _________________________________________________________________________ A10 Historic surface fault displacements associated with earthquakes in the San Francisco Bay region _________________ 13 . Width of the main rupture along the Coyote Creek fault resulting from the 1968 Borrego Mountain earthquake ______ 28 . Statistics for samples of low-strain amplifications and intensity increments with respect to the Franciscan Formation for various geologic units ___________________________________________________________________________________ 61 Summary of the analysis of liquefaction potential using standard penetration data and criteria plotted in figure 49 "e , 72 Parameters for estimating amplitude spectra for shaking of bedrock _______________________________________________ 93 BASIS FOR REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION, CALIFORNIA STUDIES FOR SEISMIC ZONATION OF THE SAN FRANCISCO BAY REGION ABSTRACT Studies by 15 researchers in various earth-science and engineering disciplines suggest that seismic zonation of the San Francisco Bay region is feasible using existing geologic and geophysical knowledge. Seismic zonation is defined as the delineation of geographical areas with different potentials for surface faulting, ground shaking, flooding, liquefaction, and landsliding during future earthquakes of specific size and location. Seismic zonation, as defined, is a necessary foundation for the development of regional land-use policies to minimize future losses during earthquakes. The need for seismic zonation was clearly demonstrated by the large variations in damage resulting from the great California earthquake of April 18, 1906. In some areas of the San Francisco Bay region, the losses of life and property were catastrophic, whereas in other areas the losses were minor. In an integrated sequence of papers, data required for seismic zonation are compiled and analyzed. Methodologies are emphasized for constructing the necessary tools from data currently available on a regional scale. Basic tools derived for seismic zonation are (1) a map showing active faults, (2) data on attenuation of shaking in bedrock, (3) geologic data, (4) a map showing qualitative estimates of ground response, (5) a map showing areas of potential inundation by tsunamis, (6) a map showing liquefaction potential, and (7) a map showing landslide susceptibility. The map showing active faults delineates areas of potential faulting of the ground surface and the location of potential sources of strong ground shaking. Categorizing the faults according to geologic and geophysical evidence for recency of movement and fault length permits crude estimates of earthquake potential, maximum earth- quake magnitude, and characteristics of future ground deformation. The San Andreas, Hayward, and Calaveras faults are considered to have the highest potential for large earthquakes, with estimates of maximum magnitude of 8.5, 8.2, and 6.5, respectively. The data on attenuation of shaking with distance from the source suggest that duration and peak-amplitude parameters of ground motion can be predicted from empirical relations for sites on bedrock and firm alluvium at distances greater than 10, 20, and 40 km (6, 12, and 25 mi) for earthquakes of magnitude 5.0 to 5.9, 6.0 to 6.9, and 7.0 to 7.9, respectively. Data are not available for smaller distances and larger magnitudes; hence extrapolation based on numerical models of faulting is required. Geologic data provide the basis for extrapolating results of local site studies to larger areas for purposes of seismic zonation. The unconsolidated sedimentary deposits of the San Francisco Bay region are differentiated into five geologic-genetic units based on geomorphic relations, soils, physical properties, fossils, and radiocarbon dates. Regrouping these units according to common sets of physical parameters such as thickness, density, sorting, data from standard penetration tests, and shear-wave velocities shows that certain groups of units are important for studying ground response, others for studing liquefaction, and still others for studying slope stability. The map showing qualitative estimates of ground response delineates on a regional basis those areas for which the effects of ground shaking, as amplified by surficial deposits, are expected to be least, intermediate, and greatest. These expectations are based on analysis of the observed 1906 California earthquake intensities, accelerograms recorded from the 1957 San Francisco earthquake, amplifications of ground shaking measured at 99 sites, numerical models of ground response, and geologic data. Combining the data on the potential location and magnitude of future earthquakes, on the attenuation of strong shaking on bedrock, and on the estimated high-strain response of surficial deposits permits quantitative predictions of ground shaking at specific sites. Such predictions suggest that certain geologic units substantially amplify frequencies of ground shaking near the fundamental mode of the unit and that the peak amplification in some instances can be at least as high at high-strain levels as that at low-strain levels. The map showing areas of potential inundation by tsunamis was prepared earlier by Ritter and Dupré (1972) and is mentioned here for completeness. The map delineates coastal areas and areas along the margins of San Francisco Bay likely to be inundated by an earth- quake-generated wave of 6 m (20 ft) at Golden Gate Bridge. The map showing landslide susceptibility delineates general areas where landsliding is considered likely. The map is based on the present distribution of landslide deposits, bedrock geology, and degree of slope. The parts of the San Francisco Bay region having the greatest susceptibility «to landsliding are hilly areas underlain by weak bedrock units of slope greater than 15 percent. Application of these seven basic tools along a demonstration profile for a postulated magnitude 6.5 earthquake on the San Andreas fault illustrates a methodology for seismic zonation of the San Francisco Bay region at the current state of the art. Pending completion of this suggested seismic zonation, a map showing maximum intensities delineates areas with potential earthquake problems, and the seven basic tools help identify the problems and their possible severity. A1 A2 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION INTRODUCTION The San Francisco Bay region (nine bay area counties) includes three major active faults and seven minor faults. Historically, 4 violent earthquakes and 33 smaller, but damaging, earthquakes have occurred on these faults. As a result, the San Francisco Bay region is considered one of the most earthquake-prone urbanized areas in the United States. The most devastating earthquake to hit the San Francisco Bay region was that of April 18, 1906. Damage to. property from the earthquake and the ensuing fires was estimated at $400 million (1906 value). Approximately 700 people lost their lives. The amount of damage from the earthquake was strongly dependent on the geologic character of the ground. For example, in the Telegraph Hill area of. San Francisco, where rock is exposed at the surface, the effects of the earthquake were "weak,” with “occasional fall of chim— neys and damage to plaster, partitions, plumbing, and the like.” But at a distance of less than one-quarter mile, in an area underlain by artificial fill and water- saturated mud, the effects of the earthquake were “vio- lent,” with “fairly general collapse of brick and frame structures when not unusually strong” (Wood, 1908). Comparison of the 1906 distribution of intensity (fig. 1) and the geologic map of San Francisco (fig. 2) de- monstrates the need for zoning the region to account for variations in earthquake hazards originating from var- iations in geologic conditions. The principal hazards to life and property from ear- thquakes in the San Francisco Bay region are potential failures of manmade structures, such as buildings, dams, waterlines, and bridges. Current urbanization implies that the effects of another 1906- type earthquake would be catastrophic. A recent study (Algermissen, 1972) predicts loss of life ranging from 2,300 to more than 100,000 people depending on time of day and the number of dam failures. The potential loss of property is estimated to be billions of dollars; the loss in productivity and earnings, substantially larger. Such potential losses can be minimized if manmade struc- tures are adequately engineered to withstand future earthquakes. Construction of earthquake-resistant structures re- quires the prudent and conscientious application of advanced engineering design techniques that consider the geologic setting of the surrounding region. In the past, the geologic setting has been considered princi- pally on a site by site basis and only for major structures (such as high-rise buildings and nuclear power plants). For most structures, no consideration has been given to the influence of the geologic setting on potential earthquake damage. This situation is due partly to a past lack of knowledge, partly to the expense of assessing the geologic setting on a site by site basis, and partly to a lack of appropriate public policy to incorporate and enforce developments in the earth sciences and engineering. The geologic setting of a region influences earthquake damage by controlling (1) the potential location and size of damaging earthquakes, (2) the potential for rupture of the ground surface by faulting, both slow creep and sudden movement, (3) the potential for damaging levels of ground shaking on different geologic units at various distances from the source of the earthquake, (4) the potential for flooding from dam failures, tsunamis, seiches, and tectonic changes of land level, and (5) the potential for shaking-induced ground failures such as landslides and those related to liquefaction. This study summarizes the state-of-the-art for asses— sing these potential earthquake effects on a regional scale for purposes of seismic zonation. Such an evaluation of the geologic setting on a regional scale provides the necessary foundation for developing policies that will minimize future earthquake losses. It also provides a basis for incorporating geologic factors into codes for the routine design of earthquake-resistant structures. The different sections of this report examine (1) faults and their earthquake potential, (2) estimation of bedrock motion at the ground surface, (3) geologic parameters for seismic zonation, (4) response of various geologic units to shaking, (5) liquefaction potential, and (6) landslide susceptibility. Seismic zonation requires the composite application of results from these studies for potential earthquakes of specific size and location. Such an application is demonstrated in terms of a profile showing effects predicted for a postulated earthquake (magnitude 136.5) on the San Andreas fault. This example illustrates the extent to which such effects as surface faulting, ground shaking, flooding, liquefaction, and landsliding can be predicted quantitatively on a regional scale for purposes of seismic zonation. The nature of the seismic zonation problem suggests a broad audience ranging from the research earth scientist and engineer to the practicing land-use planner, engineer, and politician. Accordingly, the authors have directed much of their discussion toward a broadaudience, with many results spanning a wide range of interests. The compilations and analyses presented in the first six papers are of most interest to earth scientists and engineers involved in research. The methodology presented in the seventh paper is of most interest to those involved in developing and implement- ing appropriate land-use and construction practices. STUDIES FOR SEISMIC ZONATION 122° 30’ | 122°22’ 37°48’ — 'U {b O N ’11 N O O 0 P1 :b 2 37°42' — - Very violent - Violent L EXPLANATION Very strong Weak Strong 0 1 2 3 MILES l; I I l l I l o 1 2 3 KILOMETRES Yerba Buena |. oaspumj W’S‘ ling FIGURE 1.—Distribution of apparent intensity of the 1906 earthquake in San Francisco, Calif. (after Wood, 1908). Detailed description of 1906 intensity scale for San Francisco is presented at end of report. Compare with figure 2, which shows distribution of geologic report units. A3 A4 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 12°30 12°22 l l Yerba 37°48’ — Buena I. (I) g 3 w ’11 > a i Q § 0 N '1‘“ :::::“:::~ on ~ é Q ...... O . t- O to lb 2 37°42, — ‘ _ . _ E X P L A N A T I O N Bay mud (in places covered by artificial fill as of 1906) Alluvium (>30m (100 ft) thick) 0 1 2 3 MILES if I I l l I J Alluvium (<30m (100 ft) thick) 0 1 2 3 KILOMETRES Bedrock I FIGURE 2.—Generalized geologic map, San Francisco, Calif. (compiled by K. R. Lajoie from data of Schlocker and others, 1958). FAULTS AND FUTURE EARTHQUAKES By R. L. WESSON, E. j. HELLEY, K. R. LAJOIE, and C. M. WENTWORTH INTRODUCTION The San Francisco Bay region is located within a broad complex of faults associated with the San Andreas fault system. Movement along these faults and the associated geologic deformation have produced the mountains and valleys of the California Coast Ranges that help make the bay region a scenic place to live. But this tectonic deformation is not entirely an asset. It continues today, producing small earthquakes that wake sleeping residents and earth movements that crack sidewalks and buildings. More importantly, continued deformation holds potential for generating catastrophic earthquakes, which result from the sudden movement of blocks of the earth’s crust along faults. All faults—and they are numerous in the bay region—have been surfaces of movement at least once in the geologic past. But which faults are likely to sustain future movement? And what will be the characteristics of this movement? The answers to these questions are critical to seismic zonation. If the answers can be found, land-use regulation and design and construction practices can be instituted to minimize the consequences of future movement. Our present methods are primarily empirical because physical laws govern- ing earthquake behavior are still inadequately under- stood. Our judgment of whether or not a fault is likely to move in the near future is based on whether or not it is moving today or has moved in the recent geologic past. This determination is complicated because the geologic record of past movement is incomplete. It is not now possible to determine with certainty if a fault will sustain movement in the future. We must assume that if a fault has been active over a considera- ble length of time (millions of years) and has been his- torically active or shows evidence of movement in the geologically recent past, it will most likely sustain movement in the future. Evidence from historically ac- tive faults both here and abroad indicates that this assumption is generally valid. Extensive studies of the geologic and tectonic settings of historic earthquakes, particularly those in California such as the great California earthquake of 1906, the Kern County ear- thquake of 1952, the Parkfield earthquake of 1966, and the San Fernando earthquake of 1971, reveal that the faults responsible for these earthquakes were charac— terized by at least one or more of the following features: (1) historic earthquakes with or without surface fault displacement, (2) ephemeral physiographic features such as sag ponds, offset streams, and linear ridges that suggest recent fault displacement, and (3) offset Holocene and Pleistocene deposits and geomorphic fea- tures. The presence of these characteristics provides a basis for determining which faults in the San Francisco Bay region are likely to sustain future movement. Faults along which these characteristics are developed are commonly termed “active faults.” Faults in the San Francisco Bay region that display features like those mentioned above are shown in figure 3A. They include the well—known San Andreas, Hay- ward, and Calaveras faults and many other less well—known faults potentially capable of causing sub- stantial damage. The faults are classified according to the evidence available for recent or current movement. This evidence is summarized in table 1 and the sources of data are identified on figure 33. All the faults in the San Francisco Bay region are part of What is broadly termed the San Andreas fault system. Most of these faults trend northwestward, and most display a similar sense of movement. This movement shifts the rock mass on the southwest side of each fault relatively tOWard the northwest. Fault displacements occur suddenly during earthquakes or very slowly by a process called fault creep, and for most northwest- trending bay region faults, present-day movement is almost entirely horizontal. This kind of fault movement occurs on the San Andreas, Hayward, and Calaveras faults. It is technically described as strike slip, and faults with predominantly horizontal movement are termed “strike-slip” faults. Most strike-slip faults in the San Andreas system, including those just named, exhibit a right-lateral sense of movement; that is, to an observer looking along the fault zone, the rock mass to the right of the fault moves toward him (fig. 4). For left-lateral strike-slip faults, the horizontal movement is opposite in sense—the right-hand rock mass moves away from the observer. A different sense of movement characterizes dip-slip or vertical-slip faults. Movement on them is predomin- antly vertical, and the rock mass on one side of a fault A5 A6 surface is elevated relative to the opposite mass. Depending on the geometry of the fault surface and the sense of movement, these faults are termed "normal” or “reverse” (or thrust) faults (fig. 4). Although less common than strike slip, some dip-slip faults are recognized in the bay region. Many bay region faults are not yet well enough known to identify the sense of movement positively and unequivocally. The faults in the San Francisco Bay region are grouped geographically and numbered in figure BB and table 1. The groups include the following: San Andreas fault—The San Andreas fault (1) trends through the Santa Cruz Mountains in the southern part of the bay region and along the coastal margin in the northern part. The 1906 earthquake made this one of the best known active faults in the world. Within the last few years, both local and State governments, through land—use regulation, have recognized the potential danger of this fault. Hayward and related faults—The Hayward fault (2) trends northwestward along the base of the hills behind the East Bay cities from Fremont northwest to Richmond. North of San Pablo Bay, the Rodgers Creek and Heaidsburg faults (3) continue along much the same trend. Farther north, three faults in the northeast of Alexander Valley (4, 5, and 6) continue the same trend. Segments of this fault system were responsible for damaging earthquakes in 1836, 1868, and 1969. Calaveras and related faults—The Calaveras fault (7) diverges northward from the San Andreas fault south of Hollister and continues northward along the eastern margin of the Santa Clara Valley and into the Diablo Range. Related faults, some of which may be connected with the Calaveras, include the Pleasanton fault (8) near Pleasanton, the Concord fault (9) through Concord, the Green Valley fault (10) north of Suisun Bay, faults on the west side of Napa Valley (11), and the Silver Creek fault (12) southeast of San Jose. Faults west of the San Andreas—Faults west of the San Andreas include the Zayante (13), San Gregorio (14), and the Seal Cove (15) faults. The Zayante lies west of the San Andreas but trends more westerly. The San Gregorio trends northward across the mouth of Mon- terey Bay and along the coast of San Mateo County. The Seal Cove and associated faults (15) north of Half Moon Bay may represent a northward continuation of this zone. The Pilarcitos fault (16) branches westward from the San Andreas fault on the San Francisco peninsula and may join the San Andreas beneath the Pacific Ocean south of San Francisco. Faults along the east margin of the Santa Cruz Mountains—This group of faults is poorly exposed; it includes the Sargent (18), Black Mountain (19), Berrocal (20), Serra (21), and Vasona (22) and occurs in an irregular band roughly parallel to, and about 5 km (3 mi) east of, the San Andreas fault. Several of these are REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION probably thrust faults with southwest-dipping fault surfaces. Faults along the west margin of the Great Valley- Faults along the west margin of the Great Valley include the Rio Vista fault (24) and the Antioch fault (25), which trends through the town of Antioch. The Antioch fault may extend northwestward to the Montezuma Hills fault (26). Faults in the Livermore Valley—Known faults in the Livermore Valley include the Livermore, Tesla, and Grenville faults (23). Faults between San Andreas and Healdsburg— Rodgers Creek faults—Faults with trends more wes- terly than the San Andreas fault occur in the San Anselmo, Petaluma, and Santa Rosa areas (27, 28, 29, and 30). Our knowledge of young fault movement in the San Francisco Bay region is still incomplete. Some of these faults may have moved more recently than is recog- nized, and other faults with as much potential for causing damaging earthquakes may be as yet unrecog- nized. This chapter is therefore a progress report, not a final definitive statement. EVIDENCE SUGGESTING FUTURE MOVEMENT ALONG FAULTS The process that leads earth scientists to believe that movement may occur along particular faults is one of determining which identifying characteristics result from historic fault movement and using these charac- teristics to evaluate the possibility of movement on those faults. What characteristics of the San Andreas fault were known prior to the 1906 earthquake? First, the fault was known to have produced surface displace- ment in earlier earthquakes in 1838 and 1890 (Lawson and others, 1908; Louderback, 1947). Second, the trace of the fault was characterized by such physiographic features as linear ridges and depressions, sag ponds, and scarps (Schuyler, 1898; Anderson, 1899). Third, the fault was known to offset geologic deposits of Pleis- tocene age (Lawson, 1893, 1895), formed within the last 3 million years. These and related characteristics are found worldwide along faults with historic movement. They are now accepted as evidence that a fault is likely to sustain future movement. This evidence can be divided on the basis of its age into three categories: (1) historic fault displacement—at the surface, either suddenly in association with earth- quakes or gradually as fault creep, or at depth, inferred from earthquakes that can be reliably attributed to the fault; (2) displacement during Holocene time (last 10,000 years)—faulted Holocene deposits and fault- produced topography of Holocene age; and (3) displace- ment during Quaternary time (last 3 million years)— faulted Pleistocene deposits and fault-produced topog- STUDIES FOR SEISMIC ZONATION A7 , Y % n; W "'SACIRAMEN'ROF“3.,w“; 5% a . EXPLANATION Fault with known historic movement Historic seismicity, surface rupture, or creep Fault with known Holocene movement (past 10,000 years) Well-defined fault topography or patterns of alluvial deposits incompatible with surficiai processes Fault with known Quaternary movement (past 3 million years) Ofi‘set Quaternary strata; bedrock faults associated with Quaternary faults — 37° 0 10 20 30 MILES L J 1 I l l l l I I I l I l I I 1 0 10 20 30 40 KILOMETRES A FIGURE 3.—A, Faults that may cause damaging earthquakes or surface displacement in the San Francisco Bay region. The most recent displacement on these faults is known to have occurred during historic time (past 150 years), during Holocene time (0—10,000 years before present), or during Quaternary time (0—3 million years before present). B, Sources of data, with explanation on facing page. Numbers on boxes are keyed to table 1, which summarizes the evidence for recent displacement. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION kw , \ 0 10 20 30 MILES l Illlll | |""| l l | 0 10 20 30 40 KILOMETRES B FIGURE 3.—Continued. STUDIES FOR SEISMIC ZONATION raphy of Pleistocene age. (The Holocene Epoch and preceding Pleistocene Epoch constitute the Quaternary Period.) These categories form a framework that encompasses geophysical data, data on prehistoric and historic events, and geologic data. This framework can be used to distinguish recency of faulting on the basis of Map Fault name No. orlocality 1. San Andreas A. Sources of Data Sarna-Wojcicki, A., and Pampeyan, E. H. (unpub. data) . Brown (1972) Brown and Wolfe (1972) . Radbruch (1967) .Dibblec (1972b, c, d) . Brown (1970a); Gealey (1951); Weaver (1949 .Ilelley, E. J. (unpub. data) Radbruch—Hall, D. H. (unpub.data); Wentworth, C. W., and Frizzell, V. A., Jr. (unpub. data) 5. Alexander Valley Helley, E. J. (unpub. data) 6. Big Sulphur Creek McLaughlin, R. J. (unpub. data) 7. Calaveras A. Brown (1970a) B. Dibblee (1972a, b, d; 1973a, b, c, e) Radbruch (1968a) Sharp (1973) Brown (1970a); Brown, R. D., Jr., and Frizzell, V. A., Jr.(unpub.data) Fox and others (1973) 2. Hayward >w>ow 3. Healdsburg— Rodgers Creek C" 4. Northeast of Alexander 8. Pleasanton 9. Concord 10. Green Valley 11. West side Napa Valley 12. Silver Creek Dibblee (1972b; 1973c) 13. Zayante Hall. N. T. (unpub. data) 14. San Gregorio A. Brown (1972) B. Weber and Lajoie (1974) Lajoie, K. L., and Brown, R. D., Jr. (unpub. data) Smith (1960); Lajoie, K. L.(unpub. data) Greene and others (1973) McLaughlin (1973) McLaughlin and others (1971 ); McLaughlin, R. J.(unpub. data) McLaughlin and others (1971); McLaughlin, R. J. (unpub. data) 21. Serra Bonilla (1965) 22. Vasona McLaughlin, R. J. (unpub. data) 23. Livermore A. California Department of Water 15. Seal Cove 16. Pilarcitos 17. Monterey Bay 18. Sargent 19. Black Mountain 20. Berrocal Valley Resources (1966) B. Burke, D. B. (unpub. data) 24. Rio Vista Reiche (1950) 25. Antioch Burke and Helley (1973) 26. Montezuma Hills A. Burke, D. B. (unpub. data) ' B. Sims and others (1973) Helley, E. J., and Fox, K. F., Jr. (unpub. data) Wentworth, C. W. (unpub. data) Wright, R. H., and Sorg, D. (unpub. data) Wentworth, C. W. (unpub. data) 27. Southeast of Santa Rosa 28. Burdell Mountain 29. San Geronimo Valley 30. Tolay FIGURE 3.—Continued. A9 stratigraphic evidence, in the fashion of Wentworth, Ziony, and Buchanan (1970) and Ziony, Wentworth, and Buchanan (1973). The time span of each category overlaps younger ones, emphasizing that displacement more recent than the youngest identified may have occurred. Full use of this framework is possible only where much information is available, as in the San Francisco Bay region. In many places, information may be limited largely to historic data, topographic features, and sparse stratigraphic data. EVIDENCE FOR HISTORIC DISPLACEMENT Evidence for historic displacement along faults in the bay region comes from the historic record of surface faulting and studies of recent seismicity. These studies are based partly on data collected from networks of closely spaced seismograph stations and are providing precise locations and focal-mechanism solutions for small earthquakes as well as accurate measurements of fault creep. SURFACE FAULT DISPLACEMENT DURING EARTHQUAKES Six earthquakes in the San Francisco Bay area have been accompanied by documented surface fault dis— placement (table 2). These earthquakes occurred on the San Andreas, Hayward, and Calaveras faults (fig. 5). By far the most extensive and best described ground rupture occurred with the 1906 earthquake along the San Andreas fault (Lawson and others, 1908). During this earthquake, surface displacement across the fault was as much as 5 m (16 ft)1 (fig. 6A). The pattern of disturbed ground along the fault, typical of surface displacement during earthquakes on strike-slip faults, is frequently described as a giant mole track (fig. 68). Ground rupture along the Hayward fault accompanied the earthquake of 1868 (Lawson and others, 1908; Radbruch, 1967), and surface displacement due to fault slip accompanied four other earthquakes in 1836, 1838, 1861, and 1890 (table 2). At least three additional earthquakes in the San Francisco Bay area (1800, 1865, 1911) were large enough to have caused surface displacement, although none was described. FAULT CREEP Tectonic fault creep consists of gradual relative movement across a fault at rates as large as a few centimetres (an inch or two) per year. It is less 1A larger displacement, 6 m (20 ft), is commonly cited as the maximum. This figure appears to be based on observations by G. K. Gilbert (in Lawson and others, 1908) of fault displacement of a road near the south end of Tomales Bay. Gilbert’s field notes and his published description ofthis locality (Lawson and others, 1908, p. 71) suggest that some of this displacement was nontectonic and resulted from the shitting of road fill resting on marshy ground. Four unequivocal fault offsets located about 2 km (1 1/2 mi) south of the road locality averaged about 4.70 m (15.25 ft) according to Gilbert (Lawson and others, 1908, p. 71). A10 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION TABLE 1.—Faults with Evidence for displacement Map number Fault or ‘ (fig. SB) 3:72:11? Historic displacement Holocene displacement Surface fault 7 displacement associated Fault creep Small earthquakes Offset Holocene GerOTPth with earthquakes depOSIts features 1 San Andreas 7 Yes (see Lawson and Yes (see Steinbrugge and Yes (see Brown and Yes . 7 7 7 7 Yes (see Brown (1970a.b); Hayward and related faults: 2 Hayward 7 3 Healdsburg—- Rodgers Creek. 4 Northeast of Alexander Valley. Yes (see None known Livermore Valley (several faults in the eastern Live ground-water anomalies in young alluvium, creep, others (1908); Louderback, (1947); Brown and others (1967)). Lawson and others (1908); Louderback (1947); Radbruch (1967)). None known V None known 7 77 7 7 None known 7 Yes (see Radbruch (1968a)). ‘ None observed 7 7 None known 7 None known 777 7 7 . None known 77 7 777 None known 7 7 77 77 None known 7 V 777 None known None known 7 77 77 Possibly (see Lawson and others (1908)). None known77 7 7777777 None known7777 7 7 7 7 None known77 7 7 V 7 7. None known . 7 7 None known 777777777777 None known 77777777777777 Zacher (1960); Tocher (1960)). Yes (see Radbruch and others (1966); Cluff and Steinbrugge (1966). None known 7 Probably (D. H. Radbruch— Possibly (see Wesson and None observed 7 .. 77777 Hall (unpub. data». None known None known 77 7 7 7 7 7 Yes (see Radbruch (1968a))7‘ Possibly (see Radbruch (1968a); Gibson and Wollenberg (1968)), Yes (see Sharp (1973)) Possibly (M. G. Bonilla, R. D. Brown, and C. M. Wentworth (unpub. data); V. A. Frizzell, Jr., and R, V. Sharp (unpub. data); Dooley (1973)). None known 7 7 7 77 7 Possibly (Lowney/ Kaldveer Assoc. (unpub7 report, 1971)). None known 777 777 None known None known V 7 777 None known None known 7 7 None known77 7777777 None known 77 7V 7 _ 7 None known V .7 . None known 77 77 7 7 None known777777 77.777 Lee (1971)). Brown and Wolfe (1972); U .8. Geological Survey (unpub. data». o Yes (see Brown and Yes (see Helley, Lajoie, Yes (see Radbruch Lee (1971)). and Burke (1972; (1968a)), , unpub. data». Yes (R. D. Brown Yes (see Unger and unpub. data». Eaton (1970); McEvilly (1970)). Yes (E. J. Helley 7 77 7 7 (unpub. data». Yes (C. M. Wentworth and others (1972a, b; V. A. Frizzell, Jr. (unpub. 1973a, b)). data». . None known 77 777 7 7 77 None observed 77777 7 Yes (E. J. Helley (unpub. data». . 7 777777 V 777 Yes7777 7 7 777 Yes 777777777777777777777 Yes (see Yes 7777 777 7 7 7 77 Yes (see Radbruch (1968a)). Brown and Lee (1971)). Possibly (see Lee and others (1971)). Yes (see Brown (1970a): Gibson and Wollenberg (1968)). Possibly (see Gibson and Wollenberg (1968)). Possibly (see Sharp (1973); Lee and others (1972a,b,c,); Wesson and others, (1972a,b', 1973)). Possibly (see Lee and others (1972a,b,c); Wesson and others (1972a,b,c; 1973)). Yes (E. J. Helley (unpub. data»- Yes (see Sharp (1973)). 1 Possibly (M. G. Bonilla, R. D. Brown , and C. M. Wentworth (unpub. data». Yes (see Brown (19708); Dooley (1973); V. A. Frizzell, Jr. (unpub. data». Yes (see Fox and others (1973); E. J. Helley (unpub. data». Yes (see Fox and others (1973); E. J. Helley (unpub. data». Possibly, but not within resolution of data (see Lee and others (1972a,b,c); Wesson and others (1972a,b, c». Possibly (see Wesson and others (197Za,b: 1973); Lee and others (1972a,b,c)). Possibly (Lowney/ Yes (see Brown (1970a))7 Kaldveer Assoc. (unpub. report, 1971)). Possibly (see Hall and others (in press, 1975)) Yes (see Greene and others (1973); G, Weber (unpub. data». Possibly (see Wesson and Yes 7 others (1972a,b; 1973)). Yes (see Greene and others (1973); G. Weber, J.Tinsley, and K. R. Lajoi» (unpub. data». Yes (see Glen, (1959); Jack (1968); K. R. Lajoie, J.Tinsley, and G. Weber (unpub. data». Yes (see Smith (1960); K. R. Lajoie (unpub. data». Possibly (see Greene and Yes (see Greene and Yes (see Greene and and others (1973); others (1973)). others (1973)). Griggs (1973)). 1 Yes (see Greene and others (1973); Griggs (1973)). None known on main strand; yes on associated strand (K. R. Lajoie, J. Tinsley, and G, Weber (unpub. data». None known 77 77 7 None known 7 7 7. . None‘ known 777777777777 Yes (see Allen (1946); Brown (1970a); McLaughlin (1973 )). Yes (see McLaughlin (1973 ))7 Yes (see Brown and Lee (1971)). 0 Probably (see Lee and others (1972a,b,c); W. H. K. Lee and P. Bauer (unpub. data». Possibly (see Lee and others (1972a,b,c); Wesson and others (1972a,b,c; 1973)). None known7777 7 7777 7 None observed 7 V7 7 7 777 None observed 777 7777777 None observed 7777777777 Yes (R. J. McLaughlin (unpub. data». None observed 77777777 7 None observed 777 77 7V ..7 7 Possibly 77777777 .777777 7 None observed 77777777 . 7 None observed 777 rmore Valley show various types of evidence for Quaternary displacement, including and small earthquakes. At present, however, these faults are poorly delineated. (See Bernreuter and Tokarz (1972); John Blume and Associates (unpub. mitt); California Department ofWater Resources (1966); Hansen (1964 ); o M. G. Bonilla and J. E. Schoellhamer (unpub. data); Gibson and None known. 77.77 77777 5 Alexander Valley 7 6 Big Sulfur Creek 7 7 Calaveras and related (?) faults: 7 Calaveras 7 7 7 8 Pleasanton 9 Concord 77 7 10 Green Valley 11 Western side of Napa Valley. 12 Silver Creek 7 Faults west ofthe San Andreas: 13 Zayante 7 7 7 7 14 San Gregorio 77 7 7 15 Seal Cove 7 77 V 16 Pilarcitos 77 7 77 7 77 17 Faults in Monterey Bay. Faults alon the eastern margin 0 Santa Cruz Mountains: 18 Sargent 7777777 7 7 19 Black Mountain 7 7 7 7 20 Berrocal77 7 , 7 77777 21 Serra 77 7.77777777777 22 Vasona 77 77777777777 23 Faults along the western margin of Great Valley: 24 Rio Vista 7 7.77777 25 Antioch 77777777777 7 Footnotes are at end of table. None known77 77. 7 None known 7777777777777 Probably (see Burke and Helley (1973)). lenberg (1968). Yes (see Reiche (1950)). Probably (see McEvilly and Yes (see Burke and Casaday (1967)). Helley (1973)). None known 777777777777 Yes (see Reiche (1950)). Yes (see Burke and Helley (1973)). Quaternary displacement STUDIES FOR SEISMIC ZONATION A11 Additional factors for assessing earthquake potential Quaternary displacement Offset Quaternary Estimated recurrence interval (in years) for maximum earthquake, inferred from geologic Total known fault length (in kilometres) (estimate of maximum magnitude earthquake Magnitude of largest historic earthquake Present ability to. predict pattern of surface faulting Comments dep051ts slip rate" in parentheses“) Yes (see Cummings 10¢],000 8.3 (see Lawson and 1,200 (Bl/2)9 Generally good, locally Right-lateral strikeslip (1968)). (for magnitude 7~8+I others (1908))?(3 very good. fault, maximum displacement m 1906, 6 m (20 ft). Yes . 7 . 7 [0—100 7:‘/2 (see Slemmons 72 (7.0) Yes (R. D. Brown and E. J. Helley (unpub. data)). None observed ,. , None observed , ,. . Yes (R. McLaughlin (unpub. data)). Yes . _ . Yes (see Gibson and Wollenberg (1968)). None observed,. e , , Yes (M, G. Bonilla and C. M. Wentworth (unpub. data); Dooley (1973)). Yes (see Fox and others (1973); E. J, Helley (unpub datall. Yes (see Dibblee (1972a,b,c)). Yes,.,, ,., Yes (see Greene and (for magnitude 6—7) 10—-100 (1967))?!" 5.7 (see McEVilly11970)).‘ 72 (7.0) . Geyservxlle to Milpitas, 163 (7.5) 34 (see Lee and 35 (6.6) others (1972a,b,c); Wesson and others (1972a,b: 1973)). None known2 ,,,,, 13 (7) None known'2 . 1,." 6.4 115 (Hollister to San Ramon) (7.3) ) 4.3 (see Lee and 9 (7 others (1971)).‘ 5.4 (see Sharp (1973); 18 (?) Murphy and Cloud (1957)).5 22 (includes extension across Carquinez straits) (6.3) 2—3 (see Lee and others (1972a.b,c): Wesson and others (1972a,b:1973)).5 4—5, on possible northward estension (R. L, Wesson (unpub. dated).s 2—3 (see Lee and others (1972a,b,c); Wesson and others (1972a.b:1973)).5 38 (6.6) 17 (7) 3.5 (R. L. Wesson and others (unpub. data)),‘ 20 (minimum estimate) (62) Not known 82 (7.4) 200 (includes 61 (see Richter (1958)).5v10 135 (7.4) possible north- Generally good. locally very goo . Generally good, locally very 800d Right-lateral strike-slip faults. Locally very good, abundant evidence, l'ault not well mapped. Locally very good. Fault not well mapped: Locally very good. Generally fair, locally very good to very poor. 777777777777 Ri ht-lateral strike~slip aults. Northward extension of Green Valley Locally very good. probable. Locally very good. Poor. Poor. Northward extension toward San Jose not well known. Poor. Locally very good. Right-lateral strike-slip others (1973); Brabb (1970)). ward extension to San Andreas fault, connect- fault. Southward extension. Yes (see Jack (1968); Cooper (1971); K. R. Lajoie, ' J. Tinsley, and G. Weber (unpub. data)). Yes (see Cummings (1968 )). Yes (see Greene and others (1973)). Yes (see McLaughlin (1973 l I. Yes (see Dibblee (1966 ); Pampeyan (1970); R. McLaughlin (unpub. data)). Yes (R, J, McLaughlin and W D. Sorg (unpub. data)). Yes (see Bonilla (1965)). Possibly", . . ,..,,,. Yes (see Reiche (1950)). Yes (see Burke and Helley (1973)). None known” (76)11 None known” 43 (6.7) 6.1 (see Richter (1958)).5:1° 42 (across entire bay) (6.7) 5.0 (see McEvilly (1966)),4 95 (Portola Valley to Hollisber) (7.4) 33 (Mount Madonna to Hollister) (6.7) 55 (Lake Elsman to Hollister) (6.9) 31 (Portola Valley to Los Gatos) (6.7) 3.6 (see Lee and others (1972)).4 4.5? (R. L. Wesson (unpub. data)).5 33 (Los Gatos to Mount Madonna) (6.7) None known12 ,,,,,,,, 4 (’2) 13 (see Lee and 14 (?) ' others (1972a,b,c); Wesson and others (197Za,b; 1973)).5 None known12 .. ,,,,,,. 5 (?) 4.9 (see McEvilly and 14 (?) Casaday (1967); 37 (including en echelon 1899 earthquake“; northward extension) Tocher (1959)).‘ (6.6) ing at Bolinas) Locally very good. Southward extension on shore probable. Steep southvVest-dippin'g right-lateral fault, up on southwest side, dip decreases to northwest. Locally good. Poor. Westward-dipping thrust fault. Poor. Westward-dipping thrust fault. Very good. Westward-dipping thrust fault, Poor. Poor, No longer exposed, buried by dredged material. RightJateral strike-slip fault. Locally very good. A12 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION TABLE 1 .—Faults with Evidence for displacement Map number P 3“" or ’ (fig. 38) ggfifigi Historic displacement Holocene displacement Surface fault . t - displacement associated Fault creep Small earthquakes Offset Holocene G€.°m°rph‘c with earthquakes dePOSItS features 26 Montezuma Hills. ,, , None known,,.. . . ,,,,, None known. ..,,,, Yes (see Wesson and Possibly . H, ,, Yes (D. B. Burke (unpub. others (1972a,b,c; 1973)). data)). Faults between San Andreas and Healdsburngodgers Creek faults: 27 Southeast of None known,,, , , , .,,,, None known . , ,, , ,, None known , , , ,,, None observed . , , Yes (E. J. Helley (unpub. Santa Rosa. data». 28 Burdell Mountain . .. , None known , H, . ..,, None known , , H H ,. None known H, H, , , None observed , , Yes (C. M, Wentworth (unpub. data)). 29 San Geronimo None known ,,,,,,,,,,,,, None known ,,,,,,,,,,,, None known H H, ,H ,, , None observed ,,,,,,,,,,, Yes (R. H. Wright (unpub, Valley. data); Berkland (1969)). 30 Tolay 7777777777777777 None known . .7, 7. None known , ,,,, , None known H, , None observed . ,,, Yes (C. M, Wentworth (unpub, data l I. ‘Cited paper contains numerous additional references. 2Magnitude uncertain or estimated. 3Ground rupture along surface trace of fault associated with earthquake, ‘Instrumental location of epicenter and focal mechanism suggest occurrence on named fault. 5Instrumental location of epicenter compatible with, but not compelling evidence for, location on named fault. sLocation of epicenter based on data from felt earthquakes and is compatible with occurrence on named fault 7Recurrence intervals for maximum magnitudes (from Wallace, 1970). 8Maximum magnitude estimated assuming (1) total length of fault is known and (2) half the total length would break in maximum earthquake. The maximum magnitude is calculated Left-lateral strike slip Right-lateral strike slip Dip slip (normal) Dip slip (reverse) FIGURE 4.-—Four types of fault movement, characterized by the sense of movement relative to the fault and to the horizontal. Most faults in the bay region show right-lateral strike slip, the characteristic sense of movement for the San Andreas fault. Movement on oblique-slip faults has both strike- and dip-slip components. spectacular than the sudden fault movements that accompany earthquakes. Fault creep is most obvious where it breaks or offsets streets, curbs, sidewalks, and other structures (fig. 7). Such breaks and offsets, where mapped through populated areas, define linear trends that approximate those of underlying bedrock faults. Fault creep was first measured in central California in 1956 along the San Andreas fault south of Hollister (Steinbrugge and Zacher, 1960; Tocher, 1960), although its existence was predicted many years before (Louder- back, 1942). Fault creep has now been documented for long sections of the San Andreas fault (Brown and Wallace, 1968; Nason, 1971; Savage and Burford, 1973) and the Hayward fault (Cluff and Steinbrugge, 1966; Bonilla, 1966; Blanchard and Laverty, 1966; Radbruch and Lennert, 1966; Bolt and Marion, 1966; Radbruch, 1968b; Nason, 1971) and at several localities along the Calaveras fault (Rogers and Nason, 1971). Evidence for fault creep has been described along the Concord fault (Sharp, 1973) and along the Antioch fault (Burke and Helley, 1973). Fault creep is now recognized as a common and widespread characteristic of active faults in the San Francisco Bay region (fig. 8). SMALL EARTHQUAKES Small earthquakes provide convincing evidence of contemporary fault movement where (1) many are alined along a fault, (2) the sense of movement inferred from them is systematic, and (3) that sense of movement agrees with the sense of movement derived from geologic data. An abundance of small earthquakes does not, by itself, demonstrate that a fault has the potential to generate a large earthquake, but in California several recent earthquakes in the magnitude range from 5.0 to 7.7 were preceded by many small earthquakes clustered near the larger one (Wesson and Ellsworth, 1973). To demonstrate that small earthquakes are related to a given fault requires both relatively large numbers of earthquakes and the ability to locate them accurately. Accurate location depends primarily on the number and distribution of seismograph stations and also on the ability to correct for the complexities of wave propaga- tion in the earth’s crust (Wesson and others, 1973b). Most of the accurately located earthquakes in the San Francisco Bay region can be assigned to faults with historic or geologic evidence for recent movement (fig. 9). Quaternary displacement STUDIES FOR SEISMIC ZONATION A13 Additional factors for assessing earthquake potential Quaternary displacement Offset Quaternary deposits Estimated recurrence interval (in years) for maXimum earthquake. inferred from geologic slip rate7 Magnitude of largest historic earthquake Total known fault length (in kilometres) (estimate of maximum magnitude earthquake in parentheses") Present ability to. predict pattern of surface faulting Comments Yes (D. B. Burke None known ,, , 16 m (unpub. datall. Yes (E. J. Helley ,, , . , . None known‘2 . 27 (6.4) (unpub. data)). Yes (C. M. Wentworth and . .,,,. ., , None known12 _ ,. . 19 ('f) E. J. Helley (unpub. data)). None observed ,,,, ,,,,,,,, None known12 ,,,,,,,, 15 ('2) None observed , . , . .W. .W. N, None known” . ”H...” ll (1’) - Locally good. Poor. Primarily right-lateral Locally good. strike-slip faults, but also have significant reverse displacement. Northeast side up along more westerly reaches. Locally good. Fault not well mapped. Fault not well mapped. by taking the arithmetic average of three estimates using the empirical relations of Tocher (1958) M:O.9 logic (L) +5.6, Iida (1965) M:0.76 loge (L) +6.07, and either Bonilla and Buchanan (1970) M:2.57 logm (L) +2.79 for strikeslip faults or Bonilla and Buchanan (1970)M:2.96 logic (L) +1.85 for other faults, where M is magnitude and L is one-half the fault length in kilometres. Both the assumptions and the empirical magnitudefault length relations are to some degree uncertain; therefore, the estimated maximum magnitude must be regarded as no more than a crude estimate. Only faults longer than 20 km (12 mi ) were considered. 9Maximum magnitude for the San Andreas fault is assumed to be equal to the magnitude of a historic earthquake in 1906. InUncertainty in location of epicenter permits assignment of earthquake (October 22, 1926) to either the Monterey Bay or San Gregorio fault zone. “Greene and others (1973), using a variety of magnitude—fault length relations, estimated maximum magnitudes between 7.2 and 7.9 for this fault zone. 12No significant earthquake can be assigned reliably to this fault on basis of present data. TABLE 2.—Historic surface fault displacements associated with earthquakes in the San Francisco Bay region Date Fault Rupture length Locality References June 10, 1836 ........... Hayward ,,,,,,,,,,,, Unknown ,,,,,,,,,,,,,,,,, ,Hayward ___________________ Louderback (1947). Late June, 1838 ________ San Andreas ,,,,,,,, Unknown ,,,,,,,,,,,,,,,,,, Woodside _________________ Louderback (1947). July 3, 1861 .. _.._,_..,c_1,-Calaveras __________ Unknown ,,,,,,,,,,,,,,,,,,,, 29 km (18 mi) northwest Brewer (1930); of Calaveras Reservoir, Trask (1964); west side San Ramon Witney (1865); Valley, Dublin. Lawson and others (1908); Radbruch (1968b). October 22, 1868 .......... Hayward _. .-_.m .._1,>30 km (20 mi) __________ Warm Springs northward Lawson and others (1908). to San Leandro, possibly as far north as Berkeley. April 24, 1890 ___________ San Andreas ......... >10 km? (6 mi) ____________ San Juan Bautista to ‘ Lawson and others (1908). Pajaro Gap?. April 18, 1906 __________ San Andreas "-11 ._.,,>43O km (270 mi) .......... San Juan Bautista north- Lawson and others (1908). ward to Shelter Cove or Point Delgada. EVIDENCE _F OR HOLOCENE DISPLACEMENT Holocene deposits are those formed during the last 10,000 years, during which climatic and sea level conditions have been similar to those now prevailing. These deposits are common where depositional proces- ses are still active, in such places as stream flood plains and terraces, alluvial fan surfaces, low coastal terraces, marshes, and beaches. The age of these deposits is determined in various ways, but the most reliable age assignments depend ultimately on radiocarbon dating. Where Holocene deposits are displaced or offset by faults, the fault surface—and the movement that produced it—must be younger than the deposits. Although the age of landforms such as hills, terraces, valleys, and stream channels is more difficult to determine, some can be shown to have formed during Holocene time. Where landforms of known Holocene age are cut or displaced by a fault, they too provide evidence that establishes the time of fault movement. Faults that cut or displace Holocene deposits or Holocene landforms must have moved within the last 10,000 years, along time by conventional calendars but a brief and very recent episode in geologic time. Major geologic processes like faulting are long lived, lasting millions or tens of millions of years. Thus it is prudent to consider a fault that has moved within the past 10,000 years as still active and as a factor to be weighed carefully in planning for the future. OFFSET HOLOCENE DEPOSITS Holocene deposits in the San Francisco Bay region consist primarily of stream and marine terrace deposits, alluvial fan deposits, the muds deposited in San Francisco Bay, beach deposits, and slope wash or colluvium. These deposits can be identified and dated by their surface morphology and the type of soil profiles developed upon them, by the presence of shells or bones from modern species, by the presence of aboriginal A14 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION '~;~' \.~‘ . ‘ LE z if ‘ .c’ “ ”SACRAMENTQED ”my" ‘ ,f EXPLANATION < 1838 Fault with historic surface displacement Triangle next to date indicates termination of known surface rupture Quaternary fault Displacement occurred during past 3 million years —37° ( , >1! " if \I \ o 10 20 3OM|LES \\\\\\ \ \W ’ -' Illllll|lll l I II I I \\\\\ t ’\ \ » lo 10 20 30 4o KILOMETRES \ I \\\ FIGURE 5.—Zones of surface fault displacement associated with earthquakes in the San Francisco Bay region during historic times (see table 2). See figure 3A for explanation of symbols used for Quaternary faults. STUDIES FOR SEISMIC ZONATION FIGURE 6.—Examples of surface displacement that accompanied the 1906 earthquake. A, Fence offset 21/2 m (8 ft) by right-lateral displacement on San Andreas fault. Trace of fault approximately perpendicular to fence. One kilometre (1/‘2 mi) northwest of Woodville, Marin County. Camera was alined with straight part of fence at right (in middle ground) to illustrate the zone of flexure beyond the abrupt offset. Total offset including flexure was about 3.5 m (11 ft) (from Lawson and others, 1908, plate 49A). B, "Mole track” produced by right-lateral displacement on San Andreas fault 1 1/2 km (1 mi) northwest of Olema, Marin County (from Lawson and others, 1908, plate 40A). View northwest. artifacts or skeletal remains, or by radiocarbon dating of enclosed organic materials (Helley and Brabb, 1971; FIGURE 7.—North curb on Sixth Street in Hollister, Calif, offset by right-lateral fault creep on the Calaveras fault. General trend of fault trace indicated by dashed line. Date of street construction is 1925i2 years. Fault trend and date of street construction from Rogers and Nason (1971, figure 7). View east. Helley and others, 1972; Wright, 1971). Holocene faulting of these deposits is expressed as linear scarps on modern flood plains, anomalously straight contacts of fluvial deposits, and the disruption of surface and subsurface hydrologic processes along relatively straight lines. The study of trenches dug across suspected fault zones is one of the most useful techniques for determining the recency of fault move- ment; however, it is not infallible and does require careful attention in locating the trench, preparing the trench walls, logging the exposed stratigraphic and structural relations, and dating any amenable mate- rial. Most important, a trench must expose geologic relations that provide unequivocal evidence of the relative age of fault movement. A sketch of a trench wall (fig. 10) through the fault break of the 1968 Borrego Mountain earthquake in southern California provides an example of this technique. The relative movement along this fault zone is primarily strike slip, with a small dip-slip component. This dip-slip movement offsets the deposits vertically across the fault. The amount of this offset was measured at four stratigraphic levels in the trench. The progres- A16 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION _ 38a # fPoint ""3 Reyes O o O ‘“ SAN FRANCISCO #9 '7 ,. if» is ‘1 ~ _"..:: tge EXPLANATION ...x Fault on which creep has occurred during historic time Fault with known Quaternary movement . 0.5 cm /yr (27 yr) 0. 5 cm /yr Average annual creep rate (27 yr) Length of record — 37D 0 10 20 30 MILES \ \‘ \\\\ l J l | 1 l I l \ \ \“ \ I—FFI I l l | | ‘\ lo 10 20 30 4o KILOMETRES \ I \\\ FIGURE 8.—-Distribution of documented fault creep along faults in the San Francisco Bay region. Creep rate on the Hayward, Calaveras, and San Andreas faults from Nason (1971). SPC, offset curb in San Pablo; CC, offset curb in Concord; A, Antioch fault; BT, offset tunnel in Berkeley; HAC, offset curb in Hayward; IB, offset building in Irvington; SS, deformed survey array in Sunol; ABR, offset bridge at Anderson reservoir; HC, offset curb in Hollister; and SJB—N, offset fence north of San Juan Bautista. Creep rate on Concord fault from Sharp (1973). No data available for creep rate on Antioch fault (Burke and Helley, 1973). See figure 3A for explanation of symbols used for Quaternary faults. STUDIES FOR SEISMIC ZONATION sively older beds show greater offset, and a graph of the offsets against radiometric age indicates a rate of vertical deformation. Analysis of this data not only provided abundant evidence for pre-1968 Holocene movement, but also permitted a quantitative estimate of the rate of deformation. PHYSIOGRAPHIC FEATURES Faults that are surfaces of contemporary or youthful movement can also be identified by how they interact with other geologic processes. A fault undergoing tectonic creep or one with recent episodes of abrupt displacement causes subtle but distinctive changes in the terrain it crosses. For a series of similar repetitive events, these changes are additive. Anomalous and very distinctive patterns are produced where active faults cross streams, landslides, basins that are concurrently undergoing deposition, and other ongoing geologic processes. The alined features that produce such patterns include scarps, trenches, notches, ridges, stream offsets, sag ponds, and lines of springs or vegetation (figs. 11, 12). Some of these features are direct results of fault movememt, but some have more complex origins. The absence of identifiable fault topography in places along the fault zone does not necessarily imply lack of recent displacement. The preservation of such features depends on the local rates of erosion and deposition, which vary greatly from place to place. Landslides and downslope movement of soil, as well as many other geologic processes, can effectively bury or erase the physiographic evidence of fault displacement within a few years. Although topographic features caused by fault movement seldom can be dated precisely, many can be dated approximately by geologic interpretation. In the bay region, many certainly were formed during Holocene time and therefore indicate the youngest category of prehistoric fault movement. Recently, the dating of sediments in some sag ponds along the San Andreas fault in the San Francisco Bay region (Andrei Sarna-Wojcicki, written commun., 1973) has confirmed their previously tentative age assignment as Holocene. EVIDENCE FOR QUATERNARY DISPLACEMENT Displaced deposits of Pleistocene age and some fault-produced physiographic features record fault movements that took place between 10,000 and 3 million years ago. Faults that display such evidence are here considered Quaternary faults (fig. 3A)——-they clearly have moved in Pleistocene time and may have moved during Holocene time. Pleistocene deposits in the San Francisco Bay region consist of alluvial fan and stream terrace deposits 10,000 years old and older, marine terrace deposits 70,000 to 1 million(?) years old, older San Francisco Bay A17 mud, and semi—indurated, structurally deformed conti- nental sand and gravel deposits. The Pleistocene deposits are differentiated and dated approximately by their relation to present drainage, degree of erosional dissection, relative development of soil profile, degree of induration and weathering, fossil assemblages and radiometric ages. Many faults that cut these deposits are expressed as subdued linear surface features or as near-surface ground-water barriers. Emergent wave-cut marine terraces and their as- sociated deposits record Quaternary faulting and associated warping along the coast. In some places approximate rates of deformation can be inferred from the relations shown by sequences of deformed terraces (fig. 13). ESTIMATING THE RATE OF FUTURE MOVEMENT ALONG'FAULTS AND THE CHARACTERISTICS OF ASSOCIATED EARTHQUAKES When the evidence suggests that future fault move- ment is likely, questions arise about the frequency and characteristics of the expected earthquakes. What proportion of the future movement will be associated with large earthquakes, and what proportion will occur as creep? What will be the characteristics of surface fault displacement? How much displacement may occur in a given event? But even more important are questions regarding the size of future earthquakes and the intensity of shaking. Until earthquake processes are better understood, the past must be used as the primary guide to the future in considering these questions. Present estimates of the maximum size and frequency of earthquakes along a given fault are based on (1) the geologically determined rate of slip and historic records of ground deformation, (2) the seismic history of the fault and the surrounding tectonic regime, (3) geologic evaluation of the tectonic setting, and (4) the empiri- cally derived relation between magnitude of earth— quakes and fault length or other parameters. GEOLOGICALLY DETERMINED SLIP RATES The offset of distinctive rock units establishes the rate of fault movement only within fairly wide bounds. Commonly these offsets average the rate of movement over millions of years and cannot be used to distinguish between sudden slip and creep. Data for the San Andreas fault suggest an average slip rate of 1—2 cm/yr (0.4—0.8 in./yr) over the last 20 million years (Dickinson and Grantz, 1968). But to predict movements in the immediate future, the most recent few hundreds to thousands of years are the most important. The history and rate of fault movement have been obtained within this brief time period in a few special circumstances in southern California using absolute age dating techniques. Clark, Grantz, and Rubin (1972) REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION fol C W1 SACRAMENTQ‘E‘: ‘7‘,“ s’ ‘r g EXPLANATION Quaternary fault + Earthquake epicenter — 37° + ' v ,, o 10 20 30 MILES \ ; ., | I 1 l l I l l \+\\‘\ \ , ~“ ’t a :5 A l_'_'"l : I I \\ _, .1; \wfi" _ lo 10 20 30 4o KILOMETRES \ .‘H + \\\ fl , 1 ,_ y A STUDIES FOR SEISMIC ZONATION were able to establish closely the slip rate along the Coyote Creek fault in southern California by radiomet- rically dating sediments of Holocene Lake Cahuilla (fig. 10). The 1971 San Fernando earthquake led to similar investigations (Bonilla, 1974) that provide evidence bearing on recurrence intervals on another fault. Such investigations can allow joint estimates of the mag- nitude and frequency of prehistoric earthquakes. For example, Clark, Grantz, and Rubin (1972) estimated that an earthquake with a similar amount and type of displacement, and presumably a similar magnitude, to that of the 1968 earthquake would be required along each segment of the Coyote Creek fault about every 200 years to account for the movement accumulated over the past 3,000 years. Comparable observations are not yet available in the San Francisco Bay region but are likely as more data are obtained. A major problem in estimating magnitudes and recurrence intervals is to allow for fault movement due to creep or to other aseismic processes. Wallace (1970) estimated magnitudes and recurrence intervals for maximum expectable earthquakes along parts of the San Andreas fault system by relating to a particular tectonic model the long-term slip rate inferred from geologic observations and the measured offsets from earthquakes and creep. This model assumes that the presently observed creep rates represent the longer term rates and predicts that the higher the rate of A19 tectonic creep, the longer the recurrence interval for an earthquake of a given magnitude. The recurrence intervals and maximum magnitudes obtained in this way for the San Andreas (100—1,000 years for mag- nitude 8+) and Calaveras (10—100 years for magnitude 6—7) faults are no more than a general guide to the frequency of potentially hazardous earthquakes; they represent only one of several possible interpretations of the creep-earthquake relation, but they are a step toward answering the questions of expected magnitude and frequency. Eventually it may be possible to combine the geologic data on past rates of movement and fault creep with geodetic measurements of contemporary movements (for example, Savage and Burford, 1973) to obtain more accurate estimates of recurrence intervals. SEISMIC: HISTORY The historic record of seismicity helps in estimating characteristics of future earthquakes by providing a measure of the magnitude of earthquakes that can occur along a given fault. In California the 200 years of written history and 40 years of reasonably good instrumental records are a small sample of the events on all active faults, but the usefulness of the available record can be expanded by assuming that similar faults are capable of sustaining similar earthquakes. Thus one can assume that an earthquake with a magnitude comparable to the 1868 earthquake along the Hayward 8 FIGURE 9.——A, Epicenters of earthquakes in San Francisco Bay region, 1969—72, with magnitudes 21.0 (Lee and others, 1972 (three references); Wesson and others, 1972a and b, 1973a, 1974). The pattern of epicenters is incomplete because distribution of seismo- graph stations is uneven and has changed with time. Most epi centers lying in elongate concentrations parallel to major faults were most likely associated with earthquakes on these faults.The location of these epicenters off the fault traces stems from the necessarily simplified crustal model used in the location procedure. See figure 3A for explanation of Quaternary fault symbols. B, Relation between fault surface, earthquake epicenter, and earthquake focus. F, earthquake focus or hypocenter (that point, generally at some depth within the earth’s crust, from which the seismic energy appears to radiate); E, earthquake epicenter (a point on the earth’s surface that is vertically above the focus); f, fault surface (shaded) (a fracture surface along which failure and accompanying dislocation have displaced adjacent blocks of the earth’s crust). In both diagrams arrows indicate a right-lateral strike-slip sense of movement, as on the San Andreas fault. A20 DEPTH, IN METRES EAST REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION /FauIt zone / 560 mm (22 in.) . 740 mm (29.2 in.) IlIlIIIllIlIlIIIIIIIIIl llIllll IIIIIl|l|l|l|l|lll|l|II|l|IlIIIIIHIIIIIIIIIIII 1.7m (67in.) o 0 0’0” 0 n 0 00 o 0 I) ‘ 40002 a r 4 _ HORIZONTAL DISTANCE, IN FEET 2 4 6 8 10 12 14 16 18 20 22 24 26 28 3O 5 I I I I I I I I I I I 1 I I I l I I | T | I 0 2 3 4 5 6 7 8 HORIZONTAL DISTANCE, IN METRES 2000 — — 80 > 1600 — _ g - 60 E ‘- I- 233 _ <2: SE 1200 — m8 0*; 25 LI. 83 ‘ 4° 3; _l_ _Jz 52 800— _ 5- I:E I: E E > _ 20 > 400 ~ *19680ffset l l l' I l I O O 400 1200 2000 2800 3600 4400 Recurrence YEARS BEFORE PRESENT Interval 205 yea rs 10 12 14 DEPTH, IN FEET STUDIES FOR SEISMIC ZONATION Offset drainage channel Linear ridge Bench Scarp Sag pond Linear valley Scarp Spring A21 Offset drainage channel Shutter ridge _ Linear valley FIGURE 11.—Block diagram showing landforms developed along recently active strike-slip faults. fault could occur along the San Gregorio fault because of similarities in tectonic regime, fault geometry, surface expression, and seismic characteristics (Greene and others, 1973). This kind of assumption is given added credibility by historical records of earthquakes in other seismically FIGURE 10. —Simplified sketch of trench wall showing vertical de- formation of initially flat-lying sediments and sedimentary contacts associated with predominantly horizontal movement on the Coyote Creek fault, southern California (from Clark and others, 1972, fig. 83). The trench, dug shortly after the 1968 Borrego Mountain earthquake, crosses a branching break of the fault zone along which about 50 mm (2 in.) of vertical displacement and about the same amount of horizontal displacement took place during the earth- quake. Deposits at points A, B, and C were datedradiometrically. The vertical displacement of the sedimentary contacts plotted against the age of the corresponding deposits yields an average rate of vertical defamation of about 0.5 mm/yr (0.02 in./yr) for the past 3,000 years. This suggests a recurrence interval for earthquakes the size of the 1968 event of about 200 years (Clark and others, 1972). active parts of the world (Iran, Japan, Turkey) where historic data cover a much greater time span than in California. EMPIRICAL MAGNITUDE—FAULT LENGTH RELATIONS Several empirical relations between earthquake magnitude and the length of associated surface rup- tures along faults have been derived (Bonilla, 1970; Bonilla and Buchanan, 1970; Albee and Smith, 1967; Iida, 1965; Tocher, 1958; Bolt, 1973). These magnitude- fault length relations may be used as crude estimates of the maximum-magnitude earthquake that might be expected from a particular fault, if the length of the fault is well known (fig. 14). The broad scatter shown by the data is partly due to (l) the variety of field conditions in the areas where the observations of faulting were made, (2) the inadequacy of the length of surface fault rupture as a measure of the length of faulting at depth, and (3) theoretical consider- ations (Dieterich, 1973; Thatcher and Hanks, 1973). A22 Another difficult problem is the inability to determine, for faults lacking historic ruptures, the proportion of a given fault zone that might be involved in any one earthquake (Wentworth and others, 1973). Despite these serious limitations, magnitude-fault length rela- tions are widely used to estimate the size of the maximum expectable earthquake on known faults. Estimates of maximum magnitude for various San REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION Francisco Bay region faults are shown in table 1. These estimates are based on the assumption that half the total fault length can break in any one earthquake; however, some of the faults present special problems. For example, lengths of 135 km (84 mi) or 200 km (125 mi) can be assumed for the San Gregorio fault (No. 14, fig. 3), depending on whether the Seal Cove fault and its offshore extension (No. 15, fig. 3) are considered as part A FIGURE 12.—-Physiographic features of recent faulting along a segment ofthe San Andreas fault near the Carrizo Plain, southern California. A, Aerial view west-southwestward. Low sun angle enhances physiographic features. SeeB for location. B, Part of strip map of San Andreas fault showing topographic features such as offset drain age channels, linear ridges and valleys, and linear scarps associated with recent faulting, View inA shown by stippled area (from Vedder and Wallace, 1970). A23 STUDIES FOR SEISMIC ZONATION 1 MILE l I 1 KILOMETRE a . a; , ./'View in l 1‘ figure 12 ~_~ ../" I I I ( l 230 | i < : / Offset ’/ drainage TRUE NORTH .- \ .4 f . o i . . . ‘/ a t - t, ’ -‘ :4, . z . o -. ‘ V b 2 APPROXIMATE MEAN DECUNATIQN,1975 EXPLANATION Fault traces [A Q I Linear Solid line, field or photogeologic evidence Qf recent movement shown ' ' | valle by scarp lines, trenches, sag ponds, spring lines, and vegetation 3500730,, PANORAMA HILLS QUADRANGLE/ [I y contrasts 1 19°40’ Dashed line, less obvious evidence of recent movement, but very prob- ably a fault break Brief notes along fault breaks indicate locations where features men~ tioned are especially clear. Visible fault features are not limited to the heatimts noted, but are present to some degree all along the mapped fault lines B FIGURE 12.-——Continued. A24 '1. “ '-\\ \, . X . .‘(\‘ x \ 18‘. FIGURE 13.»—A, Faulted wave—cut platform of probable Sangamon age (70,000—120,000 years before present) near Point Afio Nuevo, San Mateo County, Calif. B, Sketch showing geologic relations. Movement along strand of San Gregorio fault (No.14, fig. 3 and table 1) has displaced wave-cut platform (B) about 5.2 m (17 ft) and thrust Miocene siliceous shale (C); over unconsolidated Sangamon marine terrace deposits (D) and highly sheared Miocene shale (E). Terrace deposits overlying the steeply dipping siliceous shales in the upthrown block (C) have been removed by erosion. Neither the age of the most recent displacement nor the amount of lateral displacement associated with the thrust displacement on this fault is known. The youngest features displaced are the wave-cut platform and overlying terrace deposits, which are probably 70,000-120,000 years old. Therefore, the displacement is younger than that age. Displacement along a related subparallel fault strand 1.2 km (0.8 mi) east of this site has also offset this same wave-cut platform and deformed alluvial deposits dated at 9,510:140 years before present (Weber and Lajoie, 1974). of the San Gregorio. These alternate assumptions give estimates of about magnitude 7.4 and 7.7, respectively (Greene and others, 1973). In summary, accurate estimates of the maximum- magnitude earthquake for any given fault zone cannot be made at present. Consideration of the available data in the manner described above yields reasonable approximations. Maximum magnitude is not the whole REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 1000 |||II I 100 II||l| I 10 ||III| I LENGTH OF SURFACE RUPTURE OF MAIN FAULT, IN KILOMETRES LENGTH OF SURFACE RUPTURE 0F MAIN FAULT. IN MILES I 2 3 4 5 6 7 8 9 EARTHQUAKE MAGNITUDE FIGURE 14.—Length of observed surface rupture in relation to earthquake magnitude (Bonilla and Buchanan, 1970). Observa- tions of rupture length often underestimate the actual source dimension of the earthquake because (1) the rupture expressed at the surface may represent only a small part of the total rupture 0r (2) the surface rupture may be obscured by vegetation or water. Shown for comparison are four suggested magnitude- fault length relations: Bonilla and Buchanan (1970), best fit to all the data plotted (world); Bonilla and Buchanan (1970), best fit to all the data from strike-slip faults; and Iida (1965) and Tocher (1958), best fit to subsets of the data. story, because damaging earthquakes of lesser mag- nitude commonly occur with greater frequency. In fact, statistics indicate that over the last 40 years the number of earthquakes smaller than a given magnitude increases by a factor of about 10 for each unit decrease in magnitude. Thus, the frequency of occurrence of various sizes of earthquakes is also required to appraise the potential hazards of various faults. ESTIMATING THE CHARACTERISTICS OF GROUND DEFORMATION ASSOCIATED WITH FUTURE MOVEMENT ALONG FAULTS At the time of the great California earthquake in 1906, the San Andreas fault broke almost instantane- ously along at least 430 km (270 mi) of its length and offset the ground surface as much as 5 m (16 ft). The main line of rupture followed preexisting fault- controlled topography and was accompanied by sub- sidiary faulting and tectonic distortion of the ground STUDIES FOR SEISMIC ZONATION that at least locally extended as far as 1 km to several kilometres (thousands of feet to several miles) from the main fault (Lawson and others, 1908). Data from this and numerous other historic faulting events in the world give a basis for estimating the location, character, and maximum amount of ground deformation along many of the faults in the San Francisco Bay region. Of principal concern in fault-related ground deforma- tion are (1) detailed prediction of the pattern of surface faulting, especially the width of the zone, (2) the amount of displacement across the surface traces of faults, and (3) tectonic distortion of the ground, including uplift, subsidence, and horizontal distortion. PATTERN OF SURFACE FAULTING The pattern of surface faulting, especially along the strike-slip faults, involves a main fault zone of varying but generally narrow width along which the principal offsets occur and lesser branch and secondary faults that extend to, or occur at, considerable distance from the main zone (figs. 12, 15). Reverse faults commonly produce more complex rupture zones, and the zones typically are broader and less regular in plan (fig. 16). Major displacements can be expected along linea- ments defined by recognizable fault-caused topographic features (figs. 12, 15). Studies of several surface faulting events indicate that historic ground ruptures closely follow mappable geomorphic features that delineate preexisting fault traces (1857 Fort Tejon—Wallace, 1968; 1906 San Francisco—Lawson and others, 1908, Wallace, 1969; 1966 Parkfield—Brown and Vedder, 1967; 1968 Borrego Mountain—Clark and others, 1972, Clark, 1972; 1971 San Fernando—Yerkes and others, 1974; 1973 Managua— Brown and others, 1973); these observations suggest that patterns of surface faulting are predictable. Clark (1972) estimated, for example, that along about 50 percent of the length of the surface rupture from the Borrego Mountain earthquake of 1968, the position of the main surface fractures could have been predicted to within about 100 m (300 ft) before the earthquake. In the San Francisco Bay region, the San Andreas, Hayward, Concord, Antioch, and a few other faults are mapped in sufficient detail to accurately show the location of fault traces and of the expected future displacements (Brown and Wolfe, 1972; Brown, 1972; Radbruch, 1968a; McLaughlin, 1971; Sharp, 1973; Burke and Helley, 1973). Much of this map‘ information is adequate to influence decisions on structural design and land use. The confidence with which surface traces can be mapped at a scale of 1:24,000 (l cm=240 m; 1 inch=2,000 feet) varies considerably depending on fre- quency and amount of Quarternary displacement, the style of fault movement, and rate of destruction of A25 geomorphic features (controlled largely by climate and local geology and topography). Recent fault traces along strike—slip faults such as the San Andreas can be map- ped more confidently than those on faults with dip-slip movement. Consequently, dip-slip faults with youthful movement are only now being recognized in regions where active strike-slip faults have long been known. ZONE WIDTH Although the most obvious fault displacement tends to be localized along recognizable and mappable fault lineaments, some permanent ground deformation from fault movement extends outward from the main fault trace. This deformation, manifested as fractures, relatively small surface faults, and local warping, defines an irregular zone that parallels and includes the more obvious and more continuous traces of the main fault. The width of this zone of surface deformation varies with the type of faulting, earthquake magnitude, the local geologic setting, and perhaps other factors. An example of this variation for strike-slip faulting associated with the Borrego Mountain earthquake is shown in table 3 (Clark, 1972). Because the zone width is so variable and because it seldom can be well defined by surface morphology prior to a major fault event, detailed site studies are usually required for accurate delineationof the zone. Such detailed site information is not yet widely available. In its absence, estimated zone widths are often based on comparison with known patterns of deformation associated with well- documented modern fault geometry accompanying major earthquakes. Data on zone widths for North American earthquakes in the magnitude range from 5.5 to about 8.5 were analyzed by Bonilla (1970). The data are sparse because only a few events are well documented, but they indicate the general range in width of zones that can be anticipated. For strike-slip faults, the maximum half- width of the zone, from the centerline of the main fault zone to the outer edge of the deformation zone, is about 92 m (300 ft). For dip-slip faults the zone is as much as 900 m (3,000 ft). These values are probably conservative estimates except for very large earthquakes. They have been suggested as the basis for some kinds of planning decisions (Brown, 1972; Hall and others, 1974), but they should be used cautiously and where possible should be supplemented by site investigations. Some evidence from studies of worldwide data suggests that maximum zone width for strike-slip faults may be significantly greater than that cited above and that deformation zones of strike-slip faults may be as wide as those associated with dip-slip faults (US. Geological Survey, 1971b, p. A169). A26 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION The designation of deformation zones by these rupture zone as branch faults and secondary ruptures criteria is relatively simple where a single fault trace on faults that are not visibly connected to the main zone can be used as the center of the zone. Where multiple, and at considerable distance from it. Large branch and parallel, or overlapping fault traces are recognized, the secondary faults probably will be recognized as problem zone width is established by measuring out a half-width faults in their own right on the basis of geologic from the two outermost fault traces. evidence, but smaller ones or those with less frequent Surface offset may also extend well beyond the main displacement may not. Secondary and branch faults EXPLANATION - o Houses Stream ilk-kn! Marsh Fault traces Furrow or crack W Scarp O 1000 FEET l—I—;I—'—TJ O 300 METRES CONTOUR INTERVAL 25 FEET 123° 1330” + 38°30’ FIGURE 15.—Surface rupture associated with the 1906 earthquake along part of the San Andreas fault near Fort Ross, Calif, showing primary and secondary fault segments, Map simplified from P. Matthes (Lawson and others, 1908). STUDIES FOR SEISMIC ZONATION 1% ' .Sth “’18.“; " gylmarz} filer \ lg c /" e} Boundaries of zone containing main tectonic rup- tures in Sylmar segment of San Fernando fault zone 0. 300 Magnitude of shortening across tectonic ruptures or other compressional features, in metres 1 18° 25” EXPLANATION 15° 0.80L 0.04R ___L___ ...:.. . Left Right Tectonic rupture, showmg dlp Magnitude of lateral component of movement, Queried where tectonic origin is WW; dotted where in metres surface expression w pow L m‘ R with/out number signifies ofiset less than 0.01 m 22::::::::::::::::::: __,é_ Direction of shortening at tectonic ruptures or component of shortening on features other than surface ruptures 0.5 _D_T__ U . Vertical component of relatlve movement Dot or D (m downthrtrwn side qf rupture, U on upthrown side. Magnitude of vertical camponent of movement given in metres FIGURE 16.——Part of detailed map of surface ruptures associated with the 1971 San Fernando earthquake (US. Geological Survey, 1971a, fig. 2). The greater complexity of this pattern relative to that of the 1906 earthquake (fig. 15) reflects the difference in type of faulting (reverse slip for 1971, strike slip for 1906) and the properties of the near-surface materials. A28 TABLE 3.——Width ofthe main rupture zone along the Coyote Creek fault resulting from the 1.968 Borrego Mountain earthquake Percentage of total C t ' 1 th umula We eng length of rupture Width of rupture m ft km mi 050 (L 150 19.7 12.2 64 50» 100 15a 300 2.7 1.7 9 100— 500 300L600 5.4 3,4 17 50(L1.000 1.6003300 10 0.6 3 “1,000 “3,300 2.2 1.4 7 historically have shown measurable displacements at distances of as much as 10 km (6 mi) from the main fault zone. AMOUNT OF DISPLACEMENT Approximate limits on the amount of surface fault displacement associated with future earthquakes can be estimated from data on past events and estimates of the maximum earthquake magnitude of which a fault is capable. Maximum offsets in historic faulting events in the San Francisco Bay region range from a centimetre (1/2 in.) or so in some creep events to 5 m (16 ft) on the San Andreas fault in 1906 (Lawson and others, 1908). Wallace (1968) suggested that offset on the San Andreas fault in southern California during the 1857 Fort Tejon earthquake was about 10 m (30 ft), and so 5 In (16 ft) cannot be assumed as the maximum possible for magnitude 8 earthquakes in the San Francisco region. Other documented evidence in the bay region is limited to the 0.9 m (3 ft) of horizontal and 0.3 m (1 ft) of vertical offset that occurred on the Hayward fault in 1868. None of the reverse or thrust faults in the region are known to have undergone historic surface offset that can be used as a guide to future events. These data together with other historic data compar- ing earthquake magnitude and maximum fault offset (Bonilla and Buchanan, 1970; Clark, 1972; US. Geological Survey, 1971b) suggest upper bounds for maximum displacements on strike-slip faults of 10 m (30 ft) for magnitude 8, 6 m (20 ft) for magnitude 7, 2 m (6 ft) for magnitude 6, and 0.5 m (2 ft) for magnitude 5. The available historic data for the strike-slip faults in the San Francisco Bay region (Bonilla, 1970) suggest that the vertical displacement along these faults may be at most about one-third of these values but commonly is less than one-tenth. TECTONIC GROUND DISTORTION In addition to discrete rupture along fault traces, large areas of the ground surface can be more subtly, but also permanently, affected by vertical and horizontal distortions of the earth’s crust, including gross uplift and subsidence. These effects are related geometrically to the main fault offset but may extend for tens of REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION kilometres (or tens of miles) from the fault. These distortions include bending, warping, and changes in elevation (Bonilla, 1970; Sharp and Clark, 1972). The 1964 Alaska earthquake (magnitude 8.5) caused crustal deformation over an area of perhaps 285,000 km2 (1 1,000 mi2), producing a maximum uplift of 12 m (38 ft) and a maximum downwarp of 21/2 m (71/2 ft) (Plafker, 1969). This may be an extreme example, and the amount of distortion may reflect the fact that it was related to thrust faulting; however, similar distortions of smaller magnitude and extent have been observed with several other earthquakes. Studies of the wide- spread surface distortion associated with the 1971 San Fernando earthquake suggest that its general charac- ter could have been predicted from the Quaternary geologic history of the area (Yerkes and others, 1974). Commonly the maximum distortion is adjacent to the fault, and it gradually decreases with distance away from the fault. Such distortions must be expected to accompany any large earthquake in the San Francisco Bay region. Because the most hazardous faults in the bay region are characterized by predominantly horizontal movement, the amount of tectonic uplift or subsidence probably will be less than a metre (about 3 ft). However, significant changes in elevation or slope can be caused by dip-slip faulting or by other mechanisms of surface distortion such as landslides. IMPLICATIONS FOR REDUCING EARTHQUAKE HAZARDS Which faults are capable of generating damaging earthquakes? What will be the frequency and character of the earthquakes? What will be the pattern and character of surface fault displacement associated with earthquakes? These questions are central to the practical aspects of reducing the hazards associated with earthquakes. The answers are still incomplete, but the information available now can substantially reduce loss of life and property in future large earthquakes. Nevertheless, the San Francisco Bay region faults discussed are all believed capable of future movement. Future loss can be reduced through planning that recognizes (1) that earthquakes on these faults will be the sources of locally severe and widespread shaking and (2) that surface fault displacement and associated deformation will be localized along these faults. Current knowledge provides an ample basis for this kind of land-use planning, the formulation of local building codes, and similar governmental decisions. Because the most severe ground deformation is typi- cally restricted to a zone several tens to hundreds of metres wide along the fault trace, limitations in land use or requirements for special engineering design to reduce loss need affect only a small area, compared with STUDIES FOR SEISMIC ZONATION the area for which precautions against severe seismic shaking should be taken. The response that follows recognition of a particular fault hazard can vary greatly. One sort of response may be appropriate for facilities such as nuclear reactors, hospitals, and schools, quite another for parks, recrea— tion areas, or other less potentially hazardous uses. For potentially hazardous land uses, prudence demands a conservative response even though the evidence for current and youthful fault movement is relatively weak. But any fault showing evidence for late Quater- nary or more recent movement must be regarded as capable of future movement and should be carefully evaluated in any kind of land-use or planning decision. Faults showing evidence for older Quaternary dis- placement must be regarded with extreme caution, especially if they are near sites under consideration for potentially hazardous facilities. The likelihood that earthquakes of various sizes will occur may be taken into account by considering several levels of performance of structures during an earth— quake. One level might be represented by survival of the structures without impairment of function or safety, and another level might permit substantial structural damage but not so much as to represent catastrophic failure. In the siting and design of nuclear power reactors, for example, an expectable “operating basis earthquake” and a maximum expectable “safe shut- down earthquake” are considered (Atomic Energy Commission, 1973) so that capital investment and generating capacity are protected at reasonable cost but so that catastrophe can be avoided in the event that full potentials of the fault are realized during the life of the structure. Similar procedures can be applied to other structures with less extreme impact on public safety. The immediate implications of fault creep for any engineering structure astride the trace of a creeping fault are clear; many structures in the bay region, particularly where the Hayward fault passes through the densely populated east side of the bay, exhibit damage from fault creep. But fault creep also may or may not imply a high level of stress in the earth’s crust. What this means in terms of the potential for a large earthquake is not yet known and is currently the subject of vigorous debate (see, for example, Savage and Burford, 1973; Nason, 1971). Whether tectonic creep sufficiently relieves stress to inhibit the occurrence of a large earthquake, whether creep is a precursor of a large earthquake, or whether the actual situation is some combination of these is uncertain at present. But fault creep and large earthquakes can occur on the same segment of a fault; the sections of the Hayward fault responsible for large damaging earthquakes in 1836 and 1868 have been creeping at an average rate of about A29 1 cm (1%; in.) per year for at least the last 54 years (Nason, 1971). Planning and land-use decisions .at site, local, and regional scales should take into account the broader zone of deformation as well as fault traces themselves. Until proved otherwise by geologic site investigations, prudence suggests zone widths of 184 m (600 ft) for the largest strike-slip faults and 1,800 m (6,000 ft) for the largest dip-slip faults. In the San Francisco Bay region, most dip-slip faults are relatively short (less than 16 km or 10 mi), and for these, narrower zone widths are appropriate. Hall, Sarna-Wojcicki, and Dupré (1974) have assigned zone widths of 850 m (2,800 ft) to the Zayante fault, an oblique-slip fault in Santa Cruz County. The level of potential hazard from branch and secondary faults is less than that associated with the main fault, but for some facilities it may be essential to perform geologic site investigations to insure that branch and secondary faults are not a problem. Tentative upper bounds on the maximum surface fault displacement for an earthquake of a given magnitude are 10 m for magnitude 8, 6 m for magnitude 7, 2 m for magnitude 6, and 0.5 m for magnitude 5. In the San Francisco Bay region, the impact of permanent tectonic distortion of the earth’s surface may be small in comparison with the damage caused by shaking, landslides, and other effects. Vertical move- ments present problems along shorelines and on canals and pipelines. Horizontal distortions may cause prob- lems for pipelines as well but may also cause havoc in the definitions of land ownership. SUMMARY A well-designed program to reduce hazards from earthquakes requires concerted and coordinated effort both by the scientific and engineering disciplines and by those public agents having social and political respon- sibilities. It also requires an information base that can be used as input to engineering analysis and to policy decisions. The most fundamental segment of that information base deals with (1) the location of faults capable of generating damaging earthquakes, (2) the magnitude of earthquakes anticipated on these faults, (3) the amount of fault displacement anticipated, (4) the nature and areal distribution of deformation accom- panying earthquakes or fault movement, and (5) the frequency of recurrence of earthquakes on a known fault. The evaluation presented in this chapter reflects our current assessment of these topics for the San Francisco Bay region, and it suggests an approach that can be applied to other areas with high earthquake risk. The results are encouraging for those who wish to imple- A30 ment a program of earthquake hazard reduction, but they also indicate a need for further effort and for attention to new discoveries. About 30 faults in the bay region are potentially capable of producing damaging earthquakes. Most of these can be accurately located, and those that are the largest and potentially most destructive can be very well located. Detailed maps, suitable for most planning and decisionmaking purposes, are available for many of these faults. Magnitudes of historic earthquakes are known for more than half of the recognized faults. These data indicate that at least eight moderate or large- magnitude events have occurred on known bay region faults and that one very large earthquake (magnitude 8.3) was located on the San Andreas fault. Current methods of estimating maximum magnitude in the absence of historic data are still crude, but they provide an approximate measure of the size of earthquake that can be expected on faults that have no historic record of damaging earthquakes. Fault displacement of as much as 5 m (16 ft) was recorded after the 1906 earthquake on the San Andreas, and maximum horizontal displacement of as much as 10 m (30 ft) is judged possible with a magnitude 8 earthquake on a strike-slip fault. Estimated upper bounds for displacement (horizontal) accompanying smaller earthquakes on strike-slip faults in the bay region are 6 m (20 ft) for magnitude 7, 2 m (6 ft) for magnitude 6, and 0.5 m (2 ft) for magnitude 5. Vertical displacements associated with earthquakes on strike- slip faults are likely to be less than one-third of the horizontal displacement. Displacement associated with dip-slip faults is more difficult to evaluate, but these evidently are fewer and shorter in the bay region than strike-slip faults. The nature and areal distribution of deformation related to (fault movement includes (1) permanent ground deformation localized as a zone along the fault and (2) systematic deformation of the earth’s surface on a regional or subregional scale. Accurate delineation of the width of the zone of deformation along the fault is best accomplished through careful geologic site studies including, where necessary, trenching, excavation, or other subsurface investigations. Where such data are not available, zone width can be crudely estimated by analogy with measured zones of deformation that have accompanied historic faulting. This method suggests that, for strike-slip faults, permanent ground deformation may be expected to extend for 92 m (300 ft) on either side of a recognizable strike—slip fault trace and 425 m (1400 ft) on either side of a recognizable dip-slip fault trace (Hall and others, 1974). Designation of deformation zones on this basis is admittedly a stop-gap measure and should REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION be re-evaluated with the availability of new geologic data at the site. Regional or subregional deformation of the earth’s crust commonly accompanies major earthquakes. It is manifested predominantly as upwarping or subsidence for dip-slip faults and predominantly as horizontal distortion for strike-slip faults. In the bay region horizontal distortion appears to be the predominant process, although some local vertical warping (about 0.5 m or 1.5 ft) accompanied the 1906 earthquake on the San Andreas. The magnitude of this process and its potential hazard for the bay region are not completely known, but it appears to be less important in evaluation of earthquake hazards than other earthquake effects. The frequency of recurrence of earthquakes is perhaps the most difficult to assess of all these topics. Until more geologic data are available, recurrence estimates are tentative at best and depend heavily on our knowledge of recurrence of historic earthquakes. The historic record in the bay region is little more than 150 years old, a woefully inadequate sample for faults that have been active for millions or tens of millions of years. But even that record shows a crude pattern of damaging earthquakes on major bay region faults. Attempts to determine recurrence intervals for bay region earthquakes are further complicated by the unresolved relation between fault creep and damaging earthquakes because several bay region faults exhibit fault creep along parts of their length. Despite the need for more accurate data on frequency of recurrence, the phenomenon of recurrence is well established. Many important questions are still unanswered, but enough is known now to move positively toward reducing the hazard from future earthquakes. Some steps in this direction are obvious. All residents would agree that schools and hospitals should not be located astride the traces of major faults; most would accept requirements for geologic site studies in the deforma- tion zones along major faults; and many would agree on siting restrictions that would locate major highway interchanges, dams, or power plants away from faults that may generate earthquakes. These kinds of actions are ultimately a product of the democratic process, and they depend as much on social and economic values as on our scientific knowledge. Other steps toward reducing earthquake hazards cannot be taken without more information than is given here. Building codes, for example, are an important mechanism for protecting life and property from earthquakes. But such codes require specific informa- tion on the nature of seismic shaking, possible modes of structural response, and other factors that go far beyond the initial geologic process that causes the earthquake. These and other problems relating to hazard reduction are treated in subsequent sections of this report. ESTIMATION OF BEDROCK MOTION AT THE GROUND SURFACE By R. A. PAGE, D. M. BOORE, and J. H. DIETERICH INTRODUCTION In terms of human and economic losses, seismic shak- ing is the most significant factor contributing to the overall earthquake hazard. Shaking contributes to losses not only directly through vibratory damage to manmade structures but also indirectly through trig- gering of secondary effects such as landslides or other modes of ground failure. Thus, an important element in seismic zonation on a regional basis is the geographical assessment of potential ground shaking. The intensity and character of ground shaking de- pends upon earthquake source parameters such as magnitude, driving stress causing the fault to slip, and dimensions of the slip surface, as well as upon distance from the fault. In addition, experience shows that surfi- cial geologic materials may influence the level and na- ture of ground motion. Hence, the problem of evaluating the potential of seismic shaking divides naturally into two parts: estimation of bedrock motion at the ground surface, which is the subject of this paper, and estima- tion of the response of surficial geologic units to bedrock motion, which is discussed by Borcherdt, Joyner, and others (this report). FACTORS INFLUENCING DAMAGE POTENTIAL OF GROUND MOTION Three factors, amplitude, frequency content, and duration, govern the damage potential of ground motion and thus must be included in any scheme to characterize ground motion for purposes of design or hazard assessment. Damage tends to increase with the amplitude of bedrock motion; however, the relation between damage and amplitude is generally complex because of the response of surficial geologic deposits and manmade structures to large ground motions. Fre- quency content is a critical factor because structures, and in some cases surficial deposits, may respond in a resonant manner depending upon the frequency content of the ground motion. Relatively large deformations and stresses can occur in a structure or unconsolidated surficial deposit if the shaking includes significant amounts of energy at frequencies close to the natural resonant frequencies of the system. Duration of shak- ing, which is perhaps the least widely recognized factor influencing damage from shaking, is important because failure mechanisms in structures and uncdnsolidated surficial deposits commonly are dependent upon the cumulative number of induced stress cycles as well as the amplitudes of the stress. For example, had the duration of strong shaking during the San Fernando, Calif, earthquake of 1971 been longer, the earthquake damage and loss of life caused by shaking would have been greater than it was, as several critical structures were very near failure (Housner and Jennings, 1972). CHARACTERIZATION OF BEDROCK MOTION Ground motion can be characterized in a number of ways. The most complete characterization is a time history of ground movement, that is, a specification of ground motion in three independent spatial coordinates for every instant of time in terms of acceleration, velocity, or displacement. As basic data for the dynamic analysis of structural designs, time histories are a valuable tool to the seismic engineer. For purposes of seismic zonation, however, a more compact characteri- zation of the ground motion is desirable. For sites underlain by bedrock, one such characterization employs four physical parameters scaled or computed from standard strong-motion seismograms: maximum ground acceleration, maximum ground velocity, maximum ground displacement, and duration of shak- ing above some threshold amplitude. On a typical strong-motion recording of an earthquake at distances within which damage is sustained (see fig. 17), the peak ground acceleration commonly occurs at frequencies in the range 2 to 10 Hz, whereas the dominant frequencies of velocity and displacement are in the ranges 0.5 to 2 Hz and 0.06 to 0.5 Hz, respectively. Thus, specification of the peak amplitudes for acceleration, velocity, and displacement also conveys information about the frequency content of the motion. The strategy for predicting ground motion (Bor- cherdt, Brabb, and others, this report) involves two A31 A32 steps: first, estimation of bedrock motion at the ground surface and, second, modification of that motion to ac- count for the dynamic response of surficial geologic deposits. For use as input to the second step, we want to express ground motion in terms of the Fourier amplitude spectrum of ground acceleration. We cannot estimate the Fourier spectrum directly from peak ground-motion parameters and duration of ground motion; however, an upper bound to the spectrum is given by the zero-damped velocity response spectrum (Hudson, 1962), for which a smooth estimate can be obtained from the peak ground-motion parameters using an empirical technique described by Newmark and Hall (1969). The velocity response spectrum is a common tool for seismic design. It is defined by the maximum response velocities of a suite of linear, damped, single-degree-of- freedom oscillators subjected to a specified time history of motion. A velocity response spectrum for a given level of damping is thus a plot of maximum velocity as a function of oscillator period or frequency (see fig. 18). The usefulness of the response spectrum for design purposes stems from the ability to model structures by comparable oscillators and to estimate stresses induced by the ground motion from knowledge of the response spectrum and of the equivalent natural frequencies and damping of the structure. The relation between the Fourier and zero-damped velocity response spectra is illustrated (fig. 18) for one horizontal component of the Pacoima damsite recording of the 1971 San Fernando earthquake. Figure 18 also illustrates the graphical method of Newmark and Hall (1969) for constructing smooth response spectra from peak ground-motion parameters. The first step is to plot lines of constant acceleration, velocity, and displace- ment (dot-dashed lines, fig. 18) equal to the maximum values of ground motion (1.25 g, 115 cm/s (43 in./s), and 43 cm (17.8 in.), respectively). To obtain the smooth response spectrum, these lines are then shifted upward (dashed lines, fig. 18) on the plot by multiplying the ground-motion values by factors (4.8, 3.0, and 1.7, respectively) that reflect the dynamic amplification of the ground motion by the oscillator and the duration of ground motion. The amplification factors are dependent on the level of damping of the oscillator and, at low levels of damping, on the duration of shaking. In this example, the amplification factors are chosen to produce the illustrated fit between the smooth tripartite response spectrum (dashed line) and the computed zero-damped response spectrum (heavy solid line). These factors are derived for a magnitude 6.6 earth- quake; larger factors would be appropriate for a larger magnitude, and hence longer duration, earthquake. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION SCALING OF GROUND-MOTION PARAMETERS WITH DISTANCE AND EARTHQUAKE MAGNITUDE Since 1966, the increase in the amount and quality of strong-motion data obtained within 50 km (31 mi) of the causative fault during moderate-sized earthquakes (magnitude 5.0 to 6.9) has made it possible to predict on a statistical basis peak ground-motion values for earth- quakes smaller than magnitude 7.0 at distances greater than 10 to 20 km (6 to 12 mi) for sites on competent geologic materials ranging from bedrock to firm alluvium. The increase in the quantity of strong-motion data comes largely from increased numbers of strong- motion recorders placed in operation in recent years by the Seismological Field Survey of the National Oceanic and Atmospheric Administration (now the Seismic En- gineering Branch of the US. Geological Survey). Even with the greater number of recorders, however, no strong-motion seismograms have been obtained within 40 km (25 mi) of a magnitude 7 earthquake and within more than 100 km (62 mi) of a magnitude 8 shock. The quality of the strong-motion data also has improved in the sense that detailed aftershock investigations and field studies after recent moderate-sized earthquakes have delineated the inferred slip surface for the main earthquake and have made possible more accurate es- timates of distances of strong-motion sites from the causative fault. To determine the scaling factors for ground-motion parameters with distance, we must know the distance from a recording site to the source of the seismic energy responsible for the peak recorded motion. Lacking such information for most earthquakes, we approximate that quantity by the shortest distance to the slipped fault surface. Close to a fault the approximation may be poor because the source of peak motions may not be the point on the fault closest to the recording site (Lindh and Boore, 1973; Boore and Zoback, 1974; Hanks, 1974; Trifunac, 1974). For this reason, there is considerable uncertainty in empirical relations between ground- motion parameters and distance close to the fault (at distances less than the width or depth of the fault), even for earthquakes for which some near-fault data exist (shocks smaller than magnitude 7). PEAK GROUND ACCELERATION A plot of peak horizontal ground acceleration against shortest distance to the slipped fault (fig. 19) reveals that acceleration increases with magnitude at all distances for which data exist and that the rate of attenuation with distance is similar for all magnitudes at distances beyond 10 to 40 km (6 to 25 mi). Although STUDIES FOR SEISMIC ZONATION A33 S ‘ Ground acceleration _ 400 8 1000 — O 2 Lu _ - O 3 a “J ‘ _ a) Z“- _ E5 in _ in. z 98 ‘ Se '— Ecl‘nJ 0 0 go LLJD: LL18 ADJ — _ _'(D ”JD. » LIJ 0 on: 00) ‘ om <1“ — _~u> — — — Lu 88 81 .105 4m Lul- mm L” > >2 3: — o '2 ‘ Z LIJ o 100 — 50 50 — — 20 Ground displacement w LIJ 5 an E I—' BE 0 - O E 52 z o.|-IJ Lu 2° 0 O; 5 a. Q o l | 50 20 0 5 10 15 TIME, IN SECONDS FIGURE 17 .—Records of N. 14° E. component of horizontal ground motion at Pacoima damsite for San Fernando, Calif., earthquake of February 9, 1971 (after Trifunac and Hudson, 1971). Velocity (center) and displacement (bottom) records are obtained by integrating acceleration record (top) once and twice, respectively. there is considerable scatter in the data for a particular range of distance and magnitude, the systematic trends are obvious. Distances are known to an accuracy of about 2 to 3 km (1 to 2 mi) for most events of magnitude 5.0—5.9 and 6.0—6.9, and to within 5 km (3 mi) for events of magnitude 7.0—7.9, with one exception for which a 7 minimum value, is plotted. If peak-acceleration data taken from the literature are plotted without regard to the measure of distance (that is, whether the distance is to the epicenter, hypocenter, or closest point on the slipped surface) or without regard to accuracy of the distance, the scatter at distances less than 30 km (20 mi) A34 1000 IIIII 100 RESPONSE VELOCITY, IN CENTIMETRES PER SECOND H O 2 I I | I I 0.1 1.0 FREQUENCY, IN HERTZ FIGURE 18.—Comparison of spectra for N. 14° E. component of ground motion at Pacoima damsite for San Fernando, Calif ., earthquake of February 9, 1971. Amplitude at a given frequency may be read in terms of velocity (vertical axis), acceleration (bottom left to top right axis), or displacement (bottom right to top left axis). Zero—damped velocity response spectrum (heavy solid line) envelopes Fourier spectrum of ground acceleration (light solid line). Derivation of smooth tripartite response spectrum (dashed lines) from peak ground-motion values (dot-dashed lines) is described in text. 10.0 is at least twice that observed in figure 19 (compare Page and others, 1972, fig. 4) and tends to obscure the marked attenuation of acceleration within this distance range. The site conditions corresponding to the data shown in figure 19 range from crystalline rock to thick sections of firm alluvium. To estimate the effect of surficial conditions on peak ground acceleration, the data are replotted separately for the magnitude intervals 5.0—5.9 and 6.0-6.9 (figs. 20, 21) with a twofold classification of site geology. A site underlain by less than 5 m (16 ft) of alluvium is classified as a rock site, whereas a site underlain by a greater thickness is labeled an alluvium site. Within the range of surficial materials and levels of ground motion sampled (figs. 20, 21), site geology appears to contribute no more to the observed variation in peak acceleration than other factors that are yet to be thoroughly investigated, such as variations in driving stress and dimensions of the fault surface for earth- quakes of equivalent magnitudes and variations in the propagation of seismic waves. At sites close to the fault, REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION DISTANCE. IN MILES 10 100 | l | I l | I II I l I l l l I | I I o 1.0 _— —_ (—11 ‘ ‘ _ _ x A O OO _ _ l—> _ A x 00 _ ‘ o I _ Q9 00 O A x o o ' A o 39 0.1 :— ‘ o .0 —_ E i ‘ ‘ ‘ 0 I 2 lg: — A 00 - o — 33 - 0 CD 0 O — d — o 8 00 — 8 _ A A _ < A 0o o I A - o o 0'01 T A A8 000- i _ o o _ _ A A (,0 — _ 0 o .. _ A A A - ‘ MAGNITUDE 0 - ' ‘ A 5.0—5.9 ‘ “ °° ‘ _ o 6.0—6.9 _ I 7.0—7.9 0001 I | I l I I ll 1 I l I I I I I l I I 10 100 DISTANCE, IN KILOMETRES FIGURE 19.——Peak horizontal ground acceleration in relation to shortest distance to slipped fault as a function of magnitude. Acceleration expressed in terms of gravitational acceleration, g. Data from free-field sites, including buildings up to two stories high. Most data from Page, Boore, Joyner, and Coulter (1972, table 4). Arrow pointing to left signifies distance to fault is actually smaller than shown; arrow to right signifies distance is greater. Crosses are interpolated values for postulated earthquakes (see Borcherdt, Brabb, and others, this report). where ground shaking is sufficient to cause significant damage to ordinary structures, surficial deposits may substantially influence peak accelerations. As more near-fault strong-motion records of damaging levels of shaking are obtained, we anticipate that the effects of site geology upon peak acceleration will become clear. For example, on weak foundation materials such as the bay mud and Holocene alluvium in the San Francisco Bay region (see Lajoie and Helley, and Borcherdt, Joyner, and others, this report), we expect that for some sites peak levels of acceleration will be limited by the ability of the geologic materials to transmit the intense motion from the bedrock to the ground surface. Although the peak accelerations, which are generally a measure of the ground-motion amplitude at the higher frequencies, may be reduced for such sites from those STUDIES FOR SEISMIC ZONATION DISTANCE, IN MILES 10 100 l I I | I I I I I I I I I I l I I I I MAGNITUDE 5.0—5.9 1-0 .— 0 Rock —_ I o Alluvium : .4 O O _ __ O _ _ o a _ . . _ 0 30 0.1 :— . —: Z _ O o _. <2 — . _ *2 ~ 0 - n: - _ Lu .4 _ _ '6‘ o _ o o - <1: 0 O o 0.01 E— . o #2 : O O — _ O O . _ _ g .9 _ 0001 l I I I I I II I I | l I I I II I l J 10 100 DISTANCE, IN KILOM ETRES FIGURE 20.—Peak horizontal ground acceleration in relation to shortest distance to slipped fault for earthquakes of magnitude 5.0 to 5.9 as a function of site geology. Data from figure 19. Rock site is underlain by less than 5 m (16 ft) of alluvium; alluvium site by more than 5 m (16 ft). Arrow as in figure 19. expected on bedrock, the overall damage potential is likely to be greater on the soft ground than on rock, because of possible ground failure (see Youd and others, this report), possible extended duration of shaking, and possible ground-motion amplification at the lower frequencies (see Borcherdt, Joyner, and others, this report). From the data presented, it is clear that the depen- dence of peak acceleration (1 upon distance r can be represented by an inverse power law, a = kr‘B, where k is a factor depending on magnitude. Within the uncer- tainty of the data, the exponent B is independent of magnitude and ranges from 1.4 to 1.7. This relation applies outside the immediate vicinity of the fault, that is, at distances greater than one fault depth or width. Within the immediate vicinity of the fault, the rate of attenuation must be less. Although there is considera- ble scatter in the peak-acceleration data (figs. 19, 20, 21), it is clear that useful statistical predictions of peak A35 DISTANCE, IN MILES 10 100 I I | I I I I I I I I | I I I I | I I o 10 _ MAGNITUDE _ ' I 6.0—6.9 _ Z 0 Rock I _ O Alluvium — _ O .. _ o". 0 fl 0 O o 30 0.1 :— 0 o . —: é : 0 I E — o. o _ O: _ o (D. g _ LIJ fl - o 0 co — 8 < — 0. fi . — o 0.01 :— 9 ' — _ 0°“. : _. g) _ _ o O — _. . _ _ oo — 0.001 I I | J I I I I I I I I I I I II I I I 10 100 DISTANCE, IN KILOM ETRES FIGURE 21.—Peak horizontal ground acceleration in relation to shortest distance to slipped fault for earthquakes of magnitude 6.0 to 6.9 as a function of site geology. Data from figure 19. Rock Site is underlain by less than 5 m (16 ft) of alluvium; alluvium site by more than 5 m (16 ft). acceleration can be made now for sites on rock and competent alluvium at distances greater than 10, 20, and 40 km (6, 12, and 25 mi) for magnitude 5.0—5.9, 6.0—6.9, and 7.0—7 .9 earthquakes, respectively. PEAK GROUND VELOCITY Standard strong-motion seismographs produce ac- celerograms, that is, graphical records of ground acceleration as a function of time. Ground velocity and displacement are calculated by integrating accelero- grams once or twice, respectively. Because of the data processing involved, both velocity and displacement data are scarce compared with acceleration data, which are scaled directly from accelerograms. Recently estab- lished programs for routine computer processing of digitized accelerograms (Hudson and others, 1971) are beginning to provide the velocity and displacement data required to establish the dependence of peak ground velocity and displacement upon magnitude, distance, A36 DISTANCE. IN MILES 10 100 I I I I | I I | l ‘I I I I | I I I I I — MAGNITUDE SITE GEOLOGY _ 5.0-5.9 A Rock A Alluvium 3 100 6.0—6.9 o o : . _ g 100 r 7.0—7.9 C) _ o _ o 8 +4: 5 w _ o 0: Lu _ (I) 33 A XO 0 o: a — A O O Ell-J E ‘ 0° —_ 10 w E x. pd} a : ‘33 _ o O — o ’2 A x ~ 2 :3 1° :— 4 @ ‘ E. E : A 000 ‘ C E ’ ‘ ° 0.0 “ C9,) 0 - o w a _ > > 0CD — I. O O 1 —- o I— I I I I I I l I I I I I l I II I I I 10 100 DISTANCE, IN KILOMETRES FIGURE 22,—Peak horizontal ground velocity in relation to shortest distance to slipped fault as a function of magnitude and site geology. Data from free—field sites, including buildings up to two stories high. Solid symbols for rock sites (underlain by less than 5 m (16 ft) of alluvium); open symbols for alluvium sites (more than 5 m (16 ft) of alluvium). Large symbols for velocities from integrated accelero- grams; small symbols for velocities estimated by approximate integration techniques typically accurate to 10 to 20 percent. Distances accurate to 5 km (3 mi) or less with one exception, for which the minimum value is plotted. Crosses are interpolated values for postulated earthquake (see Borcherdt, Brabb, and others, this report). Arrow as in figure 19. and site conditions. In many respects the behavior of peak horizontal ground velocity parallels that of peak acceleration. Peak velocity increases with magnitude at all distances for which data are available and attenuates with distance from the slipped fault (fig. 22) following an inverse power law. The rate of attenuation, however, is slightly less than that for peak acceleration. The scatter in the velocity data is comparable to that observed in the acceleration data (compare fig. 19). Within the data shown, which includes sites on rock and firm alluvium only and corresponds only to low-strain levels of ground motion, there is no clear dependence of peak velocity upon site geology. PEAK GROUND DISPLACEMENT The displacement parameter discussed in this chap- ter is the peak horizontal dynamic ground displacement REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION DISTANCE, IN MILES , 10 100 100 _ I | I I I I I I l I I I I I I I I I Z MAGNITUDE SITE GEOLOGY _ (I) : . 5.0-5.9 A Rock A Alluvium “:4 E. _ 6.0—6.9 O o 2 EU) <—<> 7.0—7.9 D .2 10 E Lu “J ' O I p; o“: 2 <5 x 0 _ Lu ._J o _ (LE 10 E: w _ _ _ 5% I A A XA . D _ 2:) 0 — A 0 D 'J 5; : X D - §> < —— o E ‘ l 2 O E . “— 1 2 ~ >- o o 10 I I I l I I I I I l I I I I I I I 10 100 DISTANCE, IN KILOM ETRES FIGURE 23.—Peak horizontal ground displacement in relation to shortest distance to slipped fault as a function of magnitude and site geology. Data from free-field sites, including buildings up to two stories high. Solid symbols for rock sites (underlain by less than 5 m (16 ft) of alluvium), open symbols for soil sites (more than 5 m (16 ft) of alluvium). Large symbols for displacements from twice-integrated accelerograms; small symbols for values from 10-second displacement meters. Distances accurate to 5 km (3 mi) or less. Crosses are interpolated values for postulated earthquake (see Borcherdt, Brabb, and others, this report). Arrow as in figure 19. obtained from double integration of accelerograms or recorded directly by strong-motion displacement meters with natural periods of 10 5 (seconds) and does not contain spectral components longer than about 10—15 s. Because the resonant periods of structures are usually less than a few seconds, even for very large structures, no significant information is lost by using this measure of displacement rather than the actual ground dis- placement. A plot of peak dynamic displacement versus distance as a function of magnitude (fig. 23) shows that peak dynamic displacement increases with magnitude and attenuates with increasing distance from the fault at a rate probably less than those observed for the accelera- tion and velocity data. The data are scanty, and no inference is made regarding the effects of site geology. DURATION OF SHAKING Within the seismic engineering literature, no single measure of duration of shaking is in common usage and, in fact, discussion of duration is often not based on a quantitative definition. A crude but useful measure of duration is the time interval between the first and last peaks equal to or greater than 0.05 g on the accelero- STUDIES FOR SEISMIC ZONATION gram. This measure roughly corresponds to the "in- tense” or “strong” phase of shaking witnessed close to the fault during moderate-sized earthquakes and defines the time interval during which significant damage results from shaking. Durations of shaking for several earthquakes in the magnitude range 5.0—7 .9 increase with magnitude and with decreasing distance to the fault (Page and others, 1972). The dependence of duration on magnitude reflects the increase in fault length with magnitude and the finite velocity at which rupture propagates along the fault. EXTRAPOLATION BEYOND EMPIRICAL DATA BASE The existing strong-motion data provide an empirical basis for predicting surface bedrock motion from earthquakes in the magnitude range 5.0—6.9 at dis- tances greater than 10 to 20 km (6 to 12 mi) and at distances beyond 40 km (25 mi) for magnitude 7.0—7.9 shocks. There is, however, little observational data from which to predict motion in the immediate vicinity of the causative fault for earthquakes in the magnitude range 5.0—6.9 and no observational basis for predicting motion within 40 km (25 mi) of a magnitude 7.0—7.9 earthquake or within more than 100 km (60 mi) of a magnitude 8 shock. Thus for purposes of design or zonation, it is necessary to extrapolate from the existing base of data to small distances and to larger magnitudes. We briefly discuss some of the theoretical and numerical studies of the fault process and of the attenuation of ground motion with distance that can be used to guide extrapolations of recorded data. Extrapolation of the apparent power-law attenuation of peak acceleration, velocity, and displacement with distance (figs. 19 through 23) to within a kilometre of the fault surface suggests unrealistically large ground motions; hence the attenuation curves must flatten close to the fault to reflect finite limits of motion at the fault surface. Values for such limits have been obtained for various simplified models of the fault mechanism (Housner, 1965; Ambraseys, 1969; Brune, 1970; Dieterich, 1973; Ida, 1973) and are shown in normalized form in figure 24. Peak velocity depends on the density and rigidity of the material surrounding the fault and on the driving stress (also referred to as stress drop) causing the fault’to slip. Peak acceleration depends on these parameters and also on the high-frequency limit or cutoff in the frequency content of the motion. The high-frequency cutoff may arise from the mechanics of rupture (Ida, 1973) and the inelastic absorption of energy in highly sheared rock present in major fault zones (Boore, 1973). Estimates of the driving stress operating during earthquakes are as much as a few hundred bars and according to various fault models (for example, Brune, 1970) suggest peak velocities at the A37 fault surface in excess of 100 cm/s (40 in./s) and peak accelerations in excess of 2 g as recorded on a standard strong-motion accelerometer with a natural frequency of 16 hertz. Whether these peak values of ground motion will occur at a particular site depends on the shear strength of the underlying geologic material. On competent rock such values are expectable, whereas on unconsolidated alluvium the strength may be in- sufficient to transmit such intense motion to the surface (Ambraseys, 1973). The attenuation of these ground motions with distance close to the fault can be studied with numerical simulations of the faulting process using finite-element models. The results of such a study (Dieterich, 1973) have been used to derive scaling laws that relate peak ground-motion parameters to the stresses actirig at the fault surface and to the dimensions of the fault. The acceleration data (fig. 19), normalized accordingly, are plotted together with attenuation curves computed from several different finite-element models (fig. 24). The model results are in reasonable agreement with the data. The more rapid rate of attenuation in the empirical data at distances greater than a few minimum fault dimensions reflects inelastic absorption of energy, a process that is not included in the model. The use of this numerical model for the prediction of ground motion from an assumed earthquake requires a knowledge of both the stresses acting at the fault surface and the dimensions of the fault surface. At present, stress estimates for a given earthquake are . uncertain by factors as large as five; as knowledge of the mechanics of faulting increases, the uncertainty in stress estimates should decrease. Although the inten— sity of ground motion is physically dependent upon the stresses acting at the fault surface, the current uncertainties in stress estimates are such that mag- nitude is currently a more satisfactory parameter than stress for scaling ground motion for purposes of seismic design. SUMMARY Strong—motion recordings of earthquakes currently provide a suitable basis for predicting peak parameters and duration of ground motion at sites on rock and firm alluvium at distances greater than 10, 20, and 40 km (6, 12, and 25 mi) for earthquakes of magnitudes 5.0—5.9, 6.0—6.9, and 7 0—79, respectively. There are still, however, very few recordings of ground motion from the critical region close to the fault, where earthquake damage is intense, and there are no recordings from within 100 km.(60 mi) of an earthquake larger than magnitude 8.0. Simplified physical models of faulting provide theoretical estimates of ground motion close to the fault inside the distance range of existing observa- A38 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 0-1 __ I I I I l I I I I I I I I l I I | | I I I | I | I I | | l I | I I | I | | I I I_ Z A : 0.01 _— __ m " _ n: - _ < — a co 3 a " . . - E 8 ‘ n: a: _ _ a 0 ° 5 g o. o o m on a. A ‘ o 0.001 — A : .. _ 3 Numerical model 0 00 00 ° I 7 *—-x-—x- “ g o d _ A _ H Earthquake .‘ I‘ ‘0 _ ' magnitude i ' ._ i u . . _ A 5.0—5.9 ‘ . - O 6.0—6.9 ‘ - I 7.0-7.9 . 0.0001 | l I L I I I l l I l I l i I | I l I I l I l l | | l l L I I l - l l l l | I I 0.01 0.1 1 10 100 DISTANCE FROM FAULT MINIMUM FAULT DIMENSION (ARBITRARY UNITS) FIGURE 24.—Comparison of attenuation curves for four representative finite-element models of strike-slip faulting with peak acceleration data of figure 19. The data are normalized by dividing ground acceleration by stress drop and distance to the fault by minimum dimension of faulting (Lmin) as suggested by the scaling relation of Dieterich (1973). Attenuation curves give peak accelerations along lines oriented perpendicular to fault and originating at point of initial surface rupture and assume a 10-hertz high-frequency cutoff of the ground motion. Locations of grid points at which accelerations were computed are shown as X. At distances less than 0.2 Lmin all curves extend uniformly to fault. Differences between attenuation curves reflect differences in source geometry and point of origin-for the four models. tional data; however, the reliability of such estimates is determiried by the accuracy of both the model and its input parameters and can be ultimately tested only by comparison with observational data. There are two indications that the intensity of ground motion close to the fault in the zone of destructive shaking is significantly greater than that which was widely assumed for seismic design prior to 1971. One is the sparse but growing number of accelerograms recorded close to the fault, which have caused sig- nificant upward revision of acceleration-distance rela- tions (compare Seed and others, 1969; Schnabel and Seed, 1973). The other indication is the extensive damage caused by recent moderate-sized earthquakes occurring in or on the edge of urban areas. For example, the 1969 Santa Rosa, Calif, earthquakes (magnitudes 5.6 and 5.7) caused $6 million damage to buildings (Steinbrugge and others, 1970), and the San Fernando, Calif, earthquake (magnitude 6.6) resulted in $500 million damage (Housner and Jennings, 1972). Another sobering example of extensive damage in an urban area is the Managua, Nicaragua, earthquake of 1972 (Brown, Ward, and Plafker, 1973). In the next several years estimates of surface bedrock motion near to the fault where damage is intense will become more reliable as strong-motion recordings are collected at a growing rate and as more refined and complete theoretical and numerical models of the faulting processes are developed. Advances in techniques of processing strong-motion records will provide many more reliable data on velocity, displace- ment, and duration. DIFFERENTIATION OF SEDIMENTARY DEPOSITS FOR PURPOSES OF SEISMIC ZONATION By K. R. LAJOIE and E. J. HELLEY INTRODUCTION Geologic data are the basis for special-purpose interpretive maps such as ground—response maps, liquefaction-potential maps, and slope-stability maps. However, most standard geologic maps do not, in themselves, contain sufficient data for these purposes, particularly in areas underlain by unconsolidated sedimentary deposits. For example, most geologic maps differentiate bedrock units in considerable detail but only crudely differentiate young, unconsolidated de- posits. Yet, large historic earthquakes (for example, those that occurred near San Francisco, in Mexico City, and in Anchorage) demonstrate that some of the greatest structural damage and resultant loss of life due to high amplitudes of ground shaking and extensive ground failure occur in areas underlain by unconsoli- dated sedimentary deposits. The potential for such damage is greatest in flat lowlands because these areas generally are underlain by thick unconsolidated sedimentary deposits are are often highly developed and densely populated; parts of the San Francisco Bay region are examples of such areas. The present distribution and physical properties of the various unconsolidated sedimentary deposits are controlled by their age and depositional environment. Geologic units defined and mapped on the basis of temporal, genetic, and physical criteria, therefore, can be used to outline regions of potential earthquake- induced hazards such as liquefaction, seismic amplifica- tion, and ground failure. The primary physical proper- ties used to differentiate, map, and then regroup the unconsolidated deposits into broader units with similar seismic behavior include thickness, bedding, density, induration, texture (grain size), and porosity. These primary parameters control secondary parameters, such as seismic velocities and penetrometer resistance, which are useful to predict general behavior during earthquakes. Because none of these parameters were measured systematically in the present reconnaissance mapping project and some were estimated, the range of physical properties within each geologic unit is not precisely known. Therefore, the derivative maps based on these units, for example, the liquefaction-potential map (fig. 50) shows only in a general manner those areas where a particular seismic hazard most likely exists. Presently, the main applications of such maps are to call attention to areas where land users and planners should consider certain problems and to provide a base for future, more rigorous studies of seismic behavior. This paper describes the reconnaissance techniques used for rapidly differentiating and mapping unconsoli- dated sedimentary deposits in the San Francisco Bay region and briefly discusses the physical parameters used to regroup these deposits for delineating areas where liquefaction and ground-motion amplification might occur. GEOLOGIC MAPPING TECHNIQUES One of the main efforts in the current geologic study of the San Francisco Bay region has been to differentiate into geologically distinct and seismically significant units the alluvial deposits underlying the gently sloping sedimentary plain between the bay and the surrounding hills. Shortcuts and specialized mapping techniques have been used because of the large area (approximately 19,300 km2 (7,450 mi2)) and short time (3 years) involved. The alluvial units are defined by various combinations of geologic and genetic criteria such as depositional environment, geomorphic expres- sion, soil-profile development, age, induration, compac- tion, and texture. The distribution of the units is determined primarily from topographic maps, pub— lished soil series maps, and aerial photographs. The evolution of such a geologic map for the area of detailed study shown in figure 25 is illustrated in figures 26, 27, 28, and 29. The contour lines on the topographic map (fig. 26) clearly reveal the major geomorphic features such as the hilly uplands, the flat marshlands adjacent to the modern bay, and the broad alluvial plain sloping gently from the hills to the bay. The contour lines also reveal smaller geomorphic features such as distinct alluvial fans, stream levees, and flood plains, all of which were formed by separate but closely related alluvial proces- ses and which reveal the distribution of genetically related deposits. On the basis of relative soil-profile development, the 18 alluvial soil series described in the Soil Conservation Service report on the region (fig. 27) fall into two distinct groups that reflect some basic difference in the deposits A39 Pillar Point 0 10 MILES O 10 KlLOMETRES l 122“ 30’ figures 26—29 FIGURE 25.—Location of detailed study area. on which they are developed. The soil units having strongly developed weathering profiles constitute one group, and those having weakly to moderately de- veloped weathering profiles constitute the other. Soil profiles are developed by physical and chemical weath- ering processes at the surface of the earth; therefore, well-developed soil profiles generally indicate that the materials on which they are formed have been exposed to either intense weathering conditions or moderate weathering conditions for a considerable length of time. Because the weathering conditions in the bay region are moderate and relatively uniform, the alluvial deposits having strongly developed weathering profiles were inferred to be significantly older than the deposits having weakly to moderately developed profiles. This inferred age relation was used to differentiate younger and older alluvial deposits (fig. 28). The younger deposits make up the alluvial fans being formed under the existing hydrologic regime. The REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION streams forming these young fans are graded to present sea level. These younger deposits and the bay mud into which they grade are informally referred to as Holocene deposits (fig. 29). The older alluvial deposits, now partly covered by the Holocene deposits, make up alluvial fans formed by these same streams when they were graded to lower stands of sea level during the late Pleistocene (prior to about 10,000 years before present). These older deposits are informally referred to as late Pleistocene alluvium (fig. 29). The Holocene alluvium is differentiated further into (1) depositional facies (fig. 29) on the basis of textural characteristics (that is, gravel, sand, silt, and clay) derived primarily from published soil reports and unpublished engineering foundation reports and (2) depositional environment (that is, stream levees and flood basins) determined from geomorphic expression as revealed on topographic maps and aerial photographs. These facies grade from coarse-grained gravel and sand deposits, which form prominent stream levees at the highest parts of the alluvial fans, into medium-grained sand and silt deposits, which form broad flood plains and subdued levees along the lower margins of the alluvial fans. These stream deposits grade into and interfinger with fine-grained silt and clay deposits that form the flat floors of flood basins between stream levees on the outer margins of the alluvial fans directly adjacent to the bay marshlands. These fine-grained basin deposits and some of the medium—grained levee deposits interfinger with and grade into the bay mud, the carbonaceous silty clay deposited in the marshes and on the mudflats of San Francisco Bay during Holocene time (approximately the past 10,000 years). This gradation from coarse—grained to fine-grained sediment in the Holocene alluvium is a natural consequence of very recent stream erosion, transporta- tion, and deposition. The coarsest rock debris eroded from the bedrock uplands is deposited near the base of the hills where the rapidly flowing streams enter the broad, gently sloping alluvial plain. Only the finer grained debris is carried by the ever-slackening water to the lower parts of the alluvial fans and eventually into the bay itself, where it is deposited as bay mud. The landward extent of the saturated plastic bay mud underlying the former marshes and tidal mudflats of San Francisco Bay (fig. 26) was inferred from early (ca. 1850) US. Coast and Geodetic Survey hydrographic charts (Nichols and Wright, 1971) rather than from direct field observation because cultural activity over the past 50 years has obscured its original distribution. Long exposure to erosion and weathering processes has altered the original geomorphic expression and physical character of the late Pleistocene alluvium, thus it has not been separated into depositional facies STUDIES FOR SEISMIC ZONATION A41 NATURAL ]FLOOD BASIN '7’ TIDAL MUDFLATS ‘ Position of shoreline about; 1850 STREAM , ‘ LEyEES FORMER 2' TIDAL MARSH ‘ i ‘ NATURALV FLOOD BASIN 37°25’ 0 lMlLE 1K|LOMETRE 122°00’ FIGURE 26.—Topographic map of the Mountain View—Sunnyvale area, Mountain View 71/2-minute quadrangle. The enhanced contour lines show the irregular topography of the bedrock uplands and fluvial geomorphic features on the broad alluvial plain such as alluvial fans, stream levees, and flood basins. Distribution of former tidal marshland from Nichols and Wright (1971). A42 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 37°25’ AvD2’ \ 122°00' lMlLE 1 KILOMETRES 0 O STUDIES FOR SEISMIC ZONATION similar to those of the Holocene alluvial deposits. In the areas of the bay region where detailed soils data are not available, the alluvial deposits have been differentiated using geomorphic and genetic criteria EXPLANATION Soil Series in the Mountain View-Sunnyvale area" Alluvial soil series with weakly to moderately developed weathering profiles: An Alviso clay Ba Bayshore clay loam Ca Campbell silty clay loam Cc Campbell silty clay loam, clay substrate Cf Castro clay Ch Clear Lake clay, drained Cra Cropley clay, 0- to 2—percent slopes GbB Garretson gravelly loam, 0- to 5—percent slopes Pf Pacheco loams, clay substrate Sv Sunnyvale silty clay, drained YaA Yolo loam, 0- to 2—percent slopes YeA Yolo silty clay loam, 0- to 2—percent slopes sz Zamora clay loam, O- to 2-percent slopes Alluvial soil series with strongly developed weathering profiles: PoA Pleasanton loam, 0- to Zpercent slopes PoC Cropley clay loam, O- to 2-percent slopes PpA Pleasanton gravelly loam, 0- to 2-percent slopes PpC Pleasanton gravelly clay loam, 2- to 9—percent slopes SdA San Ysidro loam, 0- 2-percent slopes Upland soil series (nonalluvial soils): AsE Ayer clay, 15- to 30-percent slopes AuG Azule clay loam, 30- to 75-percent slopes AVE Azule silty clay loam, 15- to 30-peroent slopes AvD2 Azule silty clay loam, 9- to 15-percent slopes AvE2 Azule silty clay loam, 15- to 30-percent slopes DaD Diablo clay, 9- ’oo 15-percent slopes DaE Diablo clay, 15- to 30-percent slopes F‘bG Felton-Ben Lomond complex, 50- to 70-percent slopes LGE Los Gatos clay loam LGE2 Los Gatos clay loam, 15- to 30percent slopes LKG3 Los Gatos and Maymen, 50- to 75-percent slopes MEFZ Maymen fine sandy loam, 15- to 30-percent slopes PhG3 Permanente stony loam, 50- to 75-percent slopes PRC Positas-Saratoga loam, 2— to 9—percent slopes PRD Positas-Saratoga loam, 9- to 15-percent slopes SgC Saratoga-Positas loam, 2- to 9-percent slopes ShE2 Soper gravelly loam, 15- to 30-percent slopes ShF Soper gravelly loam, 30- to 50-percent slopes Miscellaneous map symbols: KfB Kitchen middens, archeological site LfF Landslides Ma Made land Tf Tidal flats *From US. Soil Conservation Service (1968) FIGURE 27.—Soil units in Mountain View—Sunnyvale area. Soil units are defined primarily on the basis of profile development, texture, and slope. Soil-profile development is controlled by many factors including time, climate, parent material, slope, and biological activ- ity. In an area such as this, where weathering conditions and parent material are relatively uniform, the time factor is clearly expressed by relative development of soil profiles, which can be used as a means of differentiating alluvial deposits on the basis of relative age (fig. 28). A43 derived from aerial photographs and extrapolated from areas where detailed soils data provide the main and most reliable means of recognition. Fossils, archeologi- cal remains, and radiometric ages corroborate the relative ages and correlations based on these limited data. The upper part of the late Pleistocene alluvium contains a Rancholabrean fossil vertebrate fauna containing mainly extinct species (for example, camel, bison, mammoth, and ground sloth), whereas the Holocene alluvial deposits contain a fossil fauna completely modern in aspect (for example, deer and elk). The upper part of the late Pleistocene alluvium contains fossil wood and fresh-water molluscan shells that yield radiocarbon ages of about 22,000 years before present. Holocene alluvial deposits contain fpssil wood, shells, and archeological remains that yield radiocar- bon ages of about 5,000 years before present and younger. Peat and shell deposits in the bay mud yield radiocarbon ages that range from about 9,600 years before present at the base of the unit in the lowest parts of the basin to modern at the top of the unit. These ages from the bay mud date the latest marine transgression into the basin, which agrees very well with the post-Pleistocene rise in sea level established from worldwide data (Milliman and Emery, 1968). In general, the partial flooding of the basin caused by this rise in sea level raised the base level of the local streams and began the present depositional cycle represented by the Holocene alluvial deposits. The fact that the Holocene alluvial deposits around the margins of the bay are graded to the present sea level is used as one means of identifying these deposits and adds credence to the relative ages and correlations based on other criteria. , Most of the age control used for establishing an absolute chronology and for correlating sedimentary deposits in the bay region was gleaned from published and unpublished geologic, archeologic, and engineering reports. These diverse and readily available sources provided the data for constructing an initial geologic model and determined where subsequent dating control was needed. Many of the samples dated specifically for this study were obtained from the numerous drill cores collected over a period of 25 years for engineering foundation studies of various proposed or existing transbay bridges. PHYSICAL PARAMETERS OF GEOLOGIC UNITS Physical parameters of the various geologic units such as texture, thickness, bulk density, induration, and seismic velocity are needed to assess their A44 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION ‘ w l r BAY MUD N'N‘Cx" M1- 37°25’ YOU‘NGER'ALLUVIAL‘ FAN DEPOSITS! .., a ’7 FAN VDEPOSITS $ ROCK 122°05’ 0 1 MILE 1 KILOMETRE STUDIES FOR SEISMIC ZONATION seismic-hazard potential. These parameters were com- piled and interpreted from readily available sources (fig. 30) that were augmented by direct measurement where no data existed or where verification of existing data was needed. This information also was used to differentiate units where soil series data were ambigu- ous or did not exist. Very few data are available on the total thickness of the unconsolidated and semiconsolidated deposits filling the bay basin. A few deep drill holes and several seismic profiles (Hazelwood, 1974) indicate that these sediments are probably more than 600 m (1,970 ft) thick in San Jose and thin irregularly northward from 60 to 90 m (200—300 ft) near San Francisco (see generalized cross section in figure 68 for data on sediment thickness in the Palo Alto—Coyote Hills area). The thickness of the younger units is fairly well known from numerous shallow bore holes described in engineering reports (fig. 30) and from shallow seismic surveys where bore-hole data are lacking. The Holocene alluvium (fig. 29) generally ranges in thickness from about 15 m (50 ft) near the heads of alluvial fans to about 3 m (10 ft) near the margins of the bay. The Holocene bay mud ranges in thickness from O to as much as 37 m (120 ft). The thickness of the late Pleistocene alluvium (fig. 29) is not precisely known because its base is not well defined in the thick sedimentary section beneath the bay and the surrounding alluvial plain. Where the base of the late Pleistocene alluvium can be seen on stream terraces in narrow valleys, these sediments are about 3 m (10 ft) thick. They are probably as much as 46 m (150 ft) thick beneath the bay where they overlie old estuarine mud, as identified by saltwater fossils brought up from that depth in a drill sample. Still older FIGURE 28,—Distribution of younger and older alluvial deposits in Mountain View—Sunnyvale area as determined primarily from re- lative development of soil profiles, based on soil series as mapped by Soil Conservation Service (fig. 27). Alluvial deposits on which weak to moderate weathering profiles are developed were initially infer- red to be younger than alluvial deposits on which strong weathering profiles are developed. Radiocarbon and fossil data have confirmed this relative age classification. The younger alluvial deposits con- tain modern vertebrate and invertebrate fossils and organic re- mains that yield radiocarbon ages of about 5,000 years before pre- sent and younger. Therefore, these deposits and the bay mud with which they interfinger are informally referred to as Holocene de- posits (see figure 29). The older alluvial deposits locally contain extinct late Pleistocene vertebrate fossils such as camel, sloth, bi- son, and mastodon and organic remains that yield radiocarbon ages of about 20,000 years before present. Therefore, these older deposits are informally referred to as late Pleistocene alluvium (see figure 29). A45 Pleistocene alluvial deposits probably underlie these deeply buried estuarine muds, but their total thickness is not known. In the southern bay area these deposits may grade downward into Pliocene and early Pleis- tocene alluvial deposits of the Santa Clara Formation or may lie unconformably on them. Published soils maps and unpublished engineering reports (fig. 30) are the main sources of data on the physical properties of the various geologic units. Each unit generally has a distinctive range of values for properties such as grain size, sorting, bulk density, compaction, induration, and moisture content. These properties have been used in defining and delineating some of the units, in particular the late Pleistocene alluvium and the three facies of Holocene alluvium (see figs. 29, 30). The primary physical properties such as grain size and sorting are controlled by the depositional environ- ment, whereas the secondary properties such as induration, compaction, and bulk density are related to (and generally increase with) the age of the geologic units. This variation of the secondary physical proper- ties with age is reflected in the resistance to penetra- tion, which increases from low values for the Holocene deposits to high values for the various bedrock units (col. 4, fig. 31). The very low resistance to penetration of the bay mud is attributable to its extremely high water content and loose packing, which reflect its youthful age. Older deeply buried estuarine muds would have higher resistances to penetration owing to compaction. Pene- trometer resistance can be used to estimate relative densities that, with data on grain size, sorting, and moisture content, can be used to evaluate liquefaction potential in shallow, unconsolidated deposits (see Youd and others, this report). The velocities of seismic waves in geologic materials are determined by various primary and secondary phys- ical properties and therefore can be used as a rough index of these properties and as a means of evaluating seismic behavior. The seismic wave velocities in the sedimentary units and some bedrock units are listed in figure 31. The compressional, or P—wave, velocities (Vp) were obtained primarily from shallow seismic refrac- tion surveys conducted to identify or determine strati- graphic thicknesses of the younger sedimentary units and from deep seismic refraction surveys conducted to determine the total thickness of sedimentary material in the bay basin (Hazelwood, 1974). The shear, or S -wave, velocities (Vs) of the bay mud and alluvial units were derived from limited surface surveys by the au- thors and down-hole experiments by Warrick (1974; A46 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 7 37°25' L 122°05’ 7 12°00 o 1MILE 0 1 KILOMETRE FIGURE 29.—Geologic map of the Mountain View—Sunnyvale area. STUDIES FOR SEISMIC ZONATION oral commun., 1974). The S -wave velocities of the bed- rock units were not measured directly but were esti- EXPLANATION DESCRIPTION OF MAP UNITS Holocene deposits (less than 10,000 years old): Holocene estuarine deposits (0—9,000 years old): thm Bay mud. Water-saturated estuarine mud; predominantly clay and silty clay underlying marshlands and tidal mud- flats of San Francisco Bay. Occasional lenses of well-sorted fine sand and silt; occasional shelly and peaty layers. Inter- fingers with and grades into fine-grained and medium— grained alluvium; generally overlies early Holocene alluvium or late Pleistocene alluvium 0—40 m (0-120 ft) thick Holocene alluvial deposits (0—5,000 years old): Qhaf Fine-grained alluvium. Plastic, poorly sorted carbonaceous clay and silty clay in poorly drained interfluvial basins mar- ginal to bay marshlands. Locally contains thin beds of well- sorted silt, sand, and fine gravel; contains modern vertebrate fossils and freshwater gastropod and pelecypod shells. In- terfingers with and grades into bay mud and medium- grained alluvium; overlies late Pleistocene alluvium. Gen- erally less than 5 m (15 ft) thick Qham Medium-grained alluvium. Loose, moderately drained, mod- erately sorted sand forming alluvial plains and stream levees. Locally contains beds of well-sorted clay, silt, and gravel; contains modern vertebrate fossils and fresh water gastropod and pelecypod shells. Intermediate in character and lateral extent between fine-grained and coarse-grained alluvium with which it interfingers; generally overlies late Pleistocene alluvium. Generally less than 7 m (21 ft) thick. Qhac Coarse-grained alluvium. Loose, well-drained, moderately sorted, permeable sand and gravel forming stream levees and flood plains on higher parts of alluvial fans; gravel be— comes dominant toward fan heads. Locally contains beds of well-sorted silt, sand, and gravel; contains modern verte- brate fossils and fresh water pelecypod and gastropod shells. Thickness ranges from as much as 15 m (50 ft) at fan heads to 6 m (20 ft) Where these deposits interfinger with and grade into medium-grained alluvium; overlies late Pleisto- cene alluvium and bedrock Pleistocene deposits (10,000—3,000,000 years old): Qpa Late Pleistocene alluvium (10,000—70,000? years old). Weath— ered, slightly consolidated and indurated alluvial fan de- posits consisting primarily of gravel and sand with some silt. Less permeable than Holocene alluvium. Locally con- tains fresh water pelecypod and gastropod shells and extinct late Pleistocene vertebrate fossils. Overlain by Holocene deposits on lower parts of alluvial plain; incised by channels that are partly filled with Holocene alluvium on higher parts of alluvial plain. Maximum thickness unknown but at least 45 m (150 ft) near margins of present bay where these deposits overlie deeply buried Pleistocene estuarine deposits Bedrock: QTa Pliocene and early Pleistocene alluvium. Tectonically de- formed alluvial fan deposits with local minor amounts of shallow-water marine deposits. Weakly to moderately in- durated gravel, sand, and silt with subordinate amounts of lacustrine silt and clay; local thin tufl" beds; contains late Pliocene and early Pleistocene vertebrate fossils. Underlies late Pleistocene alluvium; overlies or is in fault contact with Franciscan Formation. Consists of the Santa Clara Forma- tion in southwestern part of bay area sz Mesozoic Franciscan Formation. Well-indurated sandstone, chert, and altered volcanic rocks. In map area underlies or is in fault contact with Pliocene and early Pleistocene allu- v1um FIGURE 29.—Continued. A47 mated using the relation Vs sz/2. The P-wave velocity is low in loosely packed mate- rials such as the Holocene alluvial deposits, which can be easily compressed, and high in hard materials such as the Franciscan Formation, which cannot be easily compressed. Within the Holocene alluvium the P-wave velocities are strongly dependent on water content. Above the water table the P-wave velocity is lower than that of water, whereas below the water table it is virtu- ally that of water. The S -wave velocity is low in mate- rials with low shear strengths, such as the saturated bay mud, and high in materials with high shear strengths, such as the well-indurated Franciscan rocks. In general, the P-wave andS -wave velocities systemati- cally increase with induration and compaction of the geologic materials, and because these two properties generally increase with age of the geologic‘ units, the seismic velocities increase in 'a similar manner. Also, within an individual geologic unit both the P-wave and S -wave velocities tend to increase with depth owing to increasing compaction and induration. DELINEATING AREAS OF SEISMIC HAZARDS Once the unconsolidated deposits have been differen- tiated into map units whose distribution and general physical properties are known or can be reasonably inferred, these units can be combined in various ways on the basis of particular similarities for specific seismic zonation purposes. Figure 32 shows how the sedimen- tary units delineated in this chapter have been recom- bined to reflect liquefaction potential and possible rela- tive ground response. LIQUEFACTION POTENTIAL Liquefaction is defined as the transformation of a granular material from a solid state to a liquefied state as a consequence of increased pore-water pressure (Youd and others, this report). If shear stresses result- ing from sloping terrain or nonuniform loading are pre- sent, the liquefied sediment may fiow, generating ground failures that could result in serious damage to manmade structures. One cause of liquefaction is ground shaking during earthquakes. Seismic shaking tends to compact granular sediments, which causes a transfer of load from intergranular contacts to the in- terstitial pore water, thereby increasing the pore-water pressure. Seismically induced liquefaction is most likely to oc- cur in beds of loose, water-saturated, well-sorted silt and sand within 30 m (100 ft) of the ground surface. Geologic and engineering data indicate that these con- ditions exist to varying degrees in all five Holocene and A48 late Pleistocene sedimentary units (fig. 29). These five units were recombined to delineate three zones of differ- ent liquefaction potential in the southern bay region (fig. 50). Beds of loose well-sorted silt and sand within and directly beneath the bay mud have the highest liquefac- tion potential because they lie below sea level and are permanently saturated. The distribution of bay mud therefore defines a zone of moderate to high liquefaction potential (zone 1, fig. 50). Similar beds occur in the Holocene alluvium but are not permanently saturated i REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION because of the fluctuating ground water table. The dis- tribution of the Holocene alluvium therefore defines a zone of moderate liquefaction potential (zone 2, fig. 50) that is divided into two subzones (2a and 2b, fig. 50) on the basis of depth to the water table. In subzone 2a the depth to the water table is less than 3 m (10 ft). These areas are underlain by the fine-grained Holocene al- luvium and medium—grained alluvium that form the low, poorly drained parts of the Holocene alluvial fans. In subzone 2b the depth to the water table is greater than 3 m (10 ft). These areas are underlain by the o ~ a B E E + 9 + 'U y. m ‘. to 3’0 0: L“. 5 L”. OD ‘ (U _ ‘ N _ 10 — 3 $ 5 B 3 a a B 4 30 El. 298 1. El. 295 Soft dark ra silt cla 4‘- ? g y y y [Em / J Soft to stiff tan silt c 4' 20 1.4 ‘-—-'- 5 _ GWS. 3- 15‘2 Dense brown silty sand— {:5 7.31.55 some pebbles 1A .5:- 5;“ 1o <7 Soft brown to blue-gray my sand and clayey silt— '.'. some pebbles and roots MEAN O — O -:}'l‘. SEA LEVEL -::r E | /l l T 10 “J 8 60 40 20 0 Dense brown sand silty sand If; u— c: x E 5 __ TONS and small gravel -,:\:o‘ E f 20 Graphic driving rate Stiff tan t0 gray Clayey // of penetromete, Sllt —some pebbles EH 3’ .._-———f Stiff and slightly compact '24 30 10 _ l i l i ' gray clayey fine sand \ 400 300 200 100 o 2 SECONDS PER FOOT Stiff tan and gray mottled 1-4 ', clayey silt } 40 E X P L A N A T | O N g z E Slightly compact sandy silt 15 — —1 B-Number _ with thin sand leases 52‘, S'ze of sarn ler (inches) Top hOIe elevation . “N 50 ' " Description Hard blue-gray clayey Sllt ”Tm? Blows per foot 140 pound , hammer with 30" drop -m l 4 7,, . Unconfined compressive I ' 7 _ GWS Elevation strength (T/ftZ) : 60 20 2 7‘ Unit weight (1b/ft3) Shear strength (lb/ft ) m VS 3163 m 3: Percent Vane shear ‘_ Moisture FIGURE 30i—An example of unpublished engineering data from which thickness and physical properties of alluvial deposits were partly derived. The abrupt decrease in driving rate of the penetrometer at about 6 m (20 ft) below sea level is interpreted to reflect the stratigraphic contact between the Holocene alluvium and the late Pleistocene alluvium. STUDIES FOR SEISMIC ZONATION coarse-grained Holocene alluvium and some of the medium-grained Holocene alluvium that together form the high, well-drained parts of the Holocene alluvial fans. These deposits are only seasonally saturated and are not subject to liquefaction during much of the year. Therefore, the beds of well-sorted silt and clay that may exist in subzone 2a are water saturated for a greater part of the year and have a slightly higher liquefaction potential than similar beds that may exist in subzone 2b. Most beds of well-sorted silt and sand in the late Pleistocene alluvium are slightly compacted or chemi- cally altered and are not as likely to liquefy as similar deposits in the loose Holocene alluvium. The surface distribution of Pleistocene alluvium therefore defines a zone (zone 3, fig. 50) of low liquefaction potential. The bedrock units are generally too well indurated to liquefy, and so their distribution delineates a zone of negligible liquefaction potential. WEST A49 The liquefaction-potential map based on the above criteria does not outline areas where liquefaction, with or without resultant ground failure, will occur. It merely outlines those areas where units occur that may contain potentially liquefiable materials and where liquefaction and resultant ground failure may be ex- pected during moderate or large earthquakes. There- fore, this map highlights those areas where liquefaction potential exists and should be evaluated prior to various types of land use. GROUND RESPONSE Medvedev (1965) showed that seismic impedance (defined as the product of S -wave velocity and bulk density, VS p) can be used to make rough estimates of relative ground response. The potential for seismic am- plification increases as the impedance contrast between an overlying and an underlying unit increases if other parameters such as stratigraphic thickness are con— EAST LOW HIGH \\T$\ IMPEDANCE CONTRAST . Thickness Relative bulk Penetration P-wave S-wave Impedance, Unit (m) densrty, p resrstance1 velocrty,Vp velocrty,Vs V p (g/cm3) (blows/ft) (m/sec) (m/sec) 5 thm 0-36 1.3—1.7 O 1400 4 90—130 2 117—153 Qha 0—15 1.9 20—80 300—600 3 200—300 380—570 Qpa 10—45 2.1 100 1500—2100 200—400 5 420—630 QTa 0—250? 2.0 loo-refusal 2500 1200 2400 Ts 0—300 2.4 Refusal 1500—3300 500-1400 1200—3360 sz ------ 2.7 Refusal 2800-4000 1400—2000 3780—5400 1Test used 140-Ib hammer dropped 30 in. 2From Warrick (1974) 3Above water table. Below water table Vp =1500—1700 4Figures in italics are estimated values 5Warrick (oral commun., 1974) FIGURE 31.—Schematic cross section of southern San Francisco Bay region and description of certain physical properties of the generalized geologic units. thm, bay mud; Qha, Holocene alluvium; Qpa, late Pleistocene alluvium; QTa, early Pleistocene and Pliocene alluvium; Ts, Tertiary sandstone; sz, Franciscan Formation. A50 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION Geolo ic unit Lajoie and Helley, BorcahneddcithJeZnen YOUd and others, g this report . ’ this report this report Liquefaction potential Holocene zone Bay mud thm bay mud J. thm Bay mud lthm Bay mud l 1 Alluvium Qal Holocene alluvium Qha _, Fine grained Qhaf Holocene Qhaf Holocene 2a Medium grained Qham alluvium Qham alluvium Coarse grained Qhac __ Qhac Alluvium Qal 2b _ . Qal . Late Pleistocene alluvium Qpa Late Pleistocene Qpa Late Pleistocene I 3 alluvnum __ alluvnum Bedrock Tle b __ Pliocene-Pleistocene alluvium QTa QTa Merced Formation QTm Santa 1 Santa Clara Formation QTsc Clara Formation Tertiary rock Ts Bedrock Ts . Bedrock Pre'Tertiary Pre-Tertiary rock Mz b and Tertiary Granite Mz g bedrock Great Valley sequence Mz gv . l Franciscan Formation Mz f -- M“ FranCIscan Formation FIGURE 32.—Correlation diagram showing groupings of geologic units for evaluating ground response (Borcherdt, Joyner, and others, this report) and liquefaction potential (Youd and others, this report). stant. The table in figure 31 lists the generalized geologic units in the southwestern bay area in increas- ing age, which is roughly proportional to density and S -wave velocity and therefore to impedance (column 7). The Holocene bay mud has the lowest impedance owing to both its low density and its low S -wave velocity. The three facies of the loose Holocene alluvium (fig. 29) have very similar physical properties and therefore are com- bined into one geologic unit that has only slightly lower impedance values than the weakly consolidated late Pleistocene alluvium. The variability of physical prop- erties within and between bedrock units is reflected in their relatively wide range of moderate to high imped- ance values. The schematic cross section (fig. 31) shows the generalized stratigraphic relations in the southwestern bay region with the high contrasts in impedance be- tweenunits represented by slightly heavier contact lines. Considering only the impedance contrasts, am- plification of bedrock motion is expected to be highest where bay mud overlies late Pleistocene alluvium or where Holocene alluvium overlies Pliocene and early Pleistocene alluvium. The impedance data suggest that the highest levels of amplification would occur where thick deposits of bay mud directly overlie Franciscan bedrock. Because seismic amplification is dependent on fre- quency, and therefore controlled by other factors such as stratigraphic thickness, predicted amplification poten— tial using only impedance data is neither very precise nor directly applicable to engineering design. These crude predictions are consistent, however, with com- parative low-strain ground-motion measurements (Borcherdt, Joyner, and others, this report) that show that the highest amplifications occur on bay mud sites. It is probably significant that four of the generalized geologic units with distinct low-strain amplifications (table 5) roughly correspond to the four groups of geologic units with similar impedance values (groups separated by heavy lines in column 7 of fig. 31; correla- tion shown in fig. 32). Combining the geologic units into groups with simi- lar impedance values provides a useful means of evaluating data on earthquake intensity and low- strain-level response. For example, if deposits with similar impedance values behave differently in an earthquake, other parameters such as variation in stratigraphic thickness might be investigated as the causative factor. SUMMARY The alluvial deposits in the San Francisco Bay region, which in the past were usually treated as one geologic STUDIES FOR SEISMIC ZONATION unit, are differentiated into two main units, Holocene alluvium and late Pleistocene alluvium, primarily on the basis of soil profile development. The Holocene al- luvium is further differentiated into three textural units, coarse-, medium-, and fine-grained alluvium, on the basis of depositional environment as determined from aerial photographs, soils reports, and engineering data. The fine-grained alluvium interfingers with and grades into the Holocene bay mud. Physical properties such as relative bulk density, compaction, and indura- tion as expressed by penetrometer resistance and seis- mic velocities are generally lowest in the bay mud, in- termediate in the Holocene alluvium, and highest in the late Pleistocene alluvium. These sedimentary units may be regrouped in various ways for purposes of seismic zonation on the basis of similar physical properties. For example, all the uncon- solidated sediments contain some potentially liquefi- able beds of loose well-sorted fine sand and silt. There- fore, zones of different liquefaction potential based on the moisture content, relative compaction, and distribu- tion of the geologic units can be delineated. The Holocene bay mud delineates a zone of high liquefaction potential because the beds of loose sand and silt within these estuarine sediments are permanently saturated. The beds of loose silt and sand in the Holocene alluvium are only seasonally saturated, and these three geologic units are grouped together. Their distribution de- lineates a zone of moderate liquefaction potential. This zone is subdivided into two subzones; one where the ground water table is less than 3 m (10 it) deep has slightly higher liquefaction potential than the other. The beds of well-sorted silt and fine sand within the late A51 Pleistocene alluvium are slightly compacted and indu- rated, and so they are not as susceptible to liquefaction as similar beds in the Holocene deposits. The distribu- tion of the late Pleistocene alluvium therefore de- lineates a zone of low liquefaction potential (Youd and others, this report). For purposes of comparing ground amplification, the sedimentary deposits and bedrock units can be grouped into four units according to similarities in seismic im- pedance (Vs p). The bay mud forms a unit with the lowest impedance, and the late Pleistocene and Holocene alluvium are combined to form a unit of low to moderate impedance. The bedrock units in the south bay region fall into two groups with higher seismic impedances than the sedimentary deposits. The early Pleistocene and Tertiary sandstones form a unit with moderate to high impedance, and the Franciscan For- mation forms a unit with the highest impedance (fig. 31). Medvedev (1965) pointed out that the highest levels of seismic amplification can be expected where the im- pedance contrast between overlying and underlying geologic units is greatest. This relation suggests that the highest seismic amplifications in the bay region would be expected where thick deposits of bay mud directly overlie Franciscan bedrock. In the southern bay region the highest seismic amplifications would be ex- pected where bay mud overlies the late Pleistocene a1- luvium or where deposits of late Pleistocene and Holocene alluvium overlie bedrock. This appraisal does not consider other factors, such as stratigraphic thick- ness, that will greatly affect seismic amplification. Data on contrasts in impedance do, however, provide means of evaluating these other parameters. RESPONSE OF LOCAL GEOLOGIC UNITS TO GROUND SHAKING By R. D. BORCHERDT, W. B.]0YNER, R. E. WARRICK and J. F. GIBBS INTRODUCTION The most widespread earthquake damage to man- made structures is generally a direct result of ground shaking. Local geologic conditions can change the characteristics of earthquake ground shaking. In par- ticular, the intensity of shaking in certain frequency bands can be amplified by thick deposits of unconsoli- dated materials. Such materials exist over a large pro- portion of the San Francisco Bay region, and after the 1906 earthquake, effects of exaggerated ground shak- ing were documented at sites underlain by these mate- rials (Lawson, 1908). For example, violent effects were observed on the muds near San Francisco Bay and on the thick alluvial deposits underlying San Jose and Santa Rosa. Lawson (1908, p. 160—253) reported evi- dence for exaggerated shaking on alluvial deposits in 18 other communities. This phenomenon is not considered in present building codes, partly because observations of damage are not sufficiently quantitative to be incor- porated easily into designs for earthquake—resistant structures and partly because data on the seismic re- sponse of different geologic units are limited. However, recent increases in the number of comparative seismic recordings and advances in numerical models for the dynamic response of surficial geologic deposits are yielding improved quantitative data. This chapter summarizes these data for the San Francisco Bay region and examines their usefulness for purposes of seismic zonation. OBSERVED GROUND-MOTION AMPLIFICATIONS FROM NUCLEAR EXPLOSIONS IN NEVADA Comparative ground-motion measurements of a single seismic event provide quantitative estimates for the effects of various geologic units on ground shaking. Comparative measurements of ground motion gener- ated by distant nuclear explosions in Nevada have been made at 99 sites in the San Francisco Bay region (fig. 33). Nuclear explosions in Nevada are especially useful for such studies in the bay area, since at these distances (approximately 530 km (330 mi)) source characteristics and travel paths are nearly the same for each recording site. In addition, most of the ground-motion energy is in A52 the frequency band for which the effects of the local geologic units are greatest. This coincidence of fre- quency bands causes the amplification effects of the geologic units to be readily apparent on the analog re- cordings of ground velocity. Detailed analyses of these ground-motion data were presented by Borcherdt (1970) and Gibbs and Borcherdt (1974). Results of these analyses are summarized here. A brief summary of the geology aids in understanding these results. The numerous geologic units in the region can be grouped into three general categories on the basis of gross physical properties (see Lajoie and others, this report, for a discussion of methods for differentiat- ing sedimentary deposits and a discussion of the physi— cal parameters used in regrouping them). The three general categories determined to have distinctly differ- ent seismic properties (Borcherdt, 1970) are as follows: 1. Bay mud (equivalent to the younger bay mud unit of Borcherdt, 1970) consists mostly of recently depos- ited soft plastic carbonaceous clay, silt, and minor sand containing more than 50 weight percent wa— ter; thickness as much as 40 m (130 ft); shear velocities 90 to 130 m/s (290 to 430 ft/s). 2. Alluvium (equivalent to the older bay sediment unit of Borcherdt (1970) and the late Pleistocene and Holocene alluvium of Lajoie and Helley (Lajoie and Helley, this report) consists mostly of silty sandy clay, silty clayey sand, and sand and gravel with less than 40 weight percent water; thickness as much as 600 m (2,000 ft); shear velocity approx- imately 200 m/s (660 ft/s) at the surface, increasing with depth. 3. Bedrock consists of Pliocene and early Pleistocene alluvium of Lajoie and Helley (this report), which includes the Santa Clara Formation, consisting of semiindurated and indurated sandstone, siltstone, and mudstone; Tertiary rocks, consisting of marine sandstone and shale of Eocene, Miocene, and Pliocene age; the Page Mill Basalt, lava flows and pyroclastic rocks of Miocene age; and pre- Tertiary rocks, which include the Franciscan Formation, consisting mostly of sandstone, shale, radiolarian chert, and greenstone (volcanic rocks), minor amounts of granitic rocks, and the Great Valley sequence, consisting of indurated STUDIES FOR SEISMIC ZONATION A53 122°22’30” 121°52’30” 37° . ‘1. . “ '::.::. . 2:12 .. 45, . . . : ' :Calaizerds: ’ I : . f Reservoir a: .j.:. 370 O 22’ — O 30" {A V ‘1- EXPLANATION o 5 10 MILES - 0 5 10 KILOMETRES Bay mud Bedrock 0 Recording location Alluvium FIGURE 33.—Distribution of generalized geologic units and locations where three components of ground motions generated by nuclear explosions were recorded. sandstone and siltstone; thicknesses vary; shear Examples of. horizontal ground-velocity recordings velocities estimated to range from 500 to 2,000 m/s obtained on each of these three units are shown on (1,600—6,600 ft/s). figure 34A. All the recordings were made at sites within REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION A54 .3 56398 .8205: .«o mmnnfioom» vfimummsvm .m .oomsqgm cum 5 mmfim an N. 586298 .3205: we 356.808 338 .va £136 2: an fin: ofiofiomm «o 093 m5 8 @5283 @0365 98 mmEEoow» 93 4306393 .8225 95 NE gangmcom :oEoE “259% Rumours: mo mmEEoowml.vm $5on 4% ~ _ _/ _ _ _ _ _ _ _ _ _ . _ _ E _ T1||_ mucoomm 0H éggfééigéiIl D. i§§§§§§§§§il H n_ __ m 0 ééégggééééthSoilll w W B m XQOmOmm % A O §E§§§§§§<<§§I .H. 55:32 E My Z 9 9 AI M 3 no 0 W 3 l a 3 S M Rn. no 3 3 O 0 N G n... _ _ _ _ _ _ _ — DDS— >, INTENSITY lNCREMENT (1906 SAN FRANCISCO SCALE) 0 I ._. lJlJll l 20 30 llllJl l I] 0.5 1 5 10 AVERAGE HORIZONTAL SPECTRAL AMPLIFICATION __2 I I 0.2 FIGURE 39.—Increments in 1906 intensities as a function of average horizontal spectral amplification computed at corresponding sites from recordings of nuclear explosions. Both the intensity increment values (81) and the average spectral amplification values (AHSA) were computed with respect to the corresponding average value determined for sites underlain by rocks of the Franciscan Forma- tion. The empirical relation (81 = 0.19+2.97 log(AHSA)) is based on the data from all sites for which there was an observed 1906 intensity value (circles and dots). The empirical relation (BI = 0.27+2.70 log(AHSA)) is based on only the data from sites in the city of San Francisco (dots) for which there was unequivocal evi- dence for the ascribed degree of intensity. (The second empirical relation is preferred, although intensity increments predicted from either relation differ by less than two—tenths of an intensity incre- ment.) TABLE 4.—-Statistics for samples of low-strain amplifications and intensity increments with respect to Franciscan Formation for various geologic units Average horizontal spectral amplification Intensity increment GEOIOEIC unit (1906 San Francisco scale) Standard Standard Mean . . Mean . . devxation dev1ation Granite ,,,,,,,,,,,, 0.63 0.11 -0.29 0.21 Franciscan Formation iiiiii 1.00 0.38 0.19 0.47 Great Valley sequence ,,,,,,,,, 1.42 0.45 0.64 0.34 Santa Clara Formation ....... 1.70 0.64 0.82 0.48 Alluvium __________ 2,76 1,16 1.34 0.58 Bay mud ,,,,,,,,,,, 7.06 3.78 2.43 0.58 with respect to the bedrock, spectral amplifications were predicted with the numerical model. This predic- tion is compared with the amplifications observed from the recordings of the distant San Fernando earthquake (fig. 44). The fundamental frequency and the frequency A61 of the first two higher modes for the computed response agree to within about 10 percent of those observed. The amount of amplification is also in reasonable agree- ment. The maximum computed amplification is about 35 percent less than the maximum observed. Such com- parisons of low-strain amplifications (max. strain z 1045) are useful for understanding the amplification phenomena and for the development of numerical mod- els. The corresponding low-strain parameters for the unconsolidated deposits can be determined satisfactor- ily both in the laboratory and in situ. However, esti- mates of high-strain parameters at present must result largely from laboratory measurement of soil paramet- ers. Using laboratory data on the dynamic behavior of bay mud and alluvium, a quasilinear procedure was applied to estimate the high-strain response of the surficial geologic deposits. The procedure is based on the as- sumption that the linear low-strain model is applicable, provided the model parameters are chosen in accord with high-strain data from laboratory studies. This as- sumption has been applied previously by Idriss and Seed (1968). The high-strain parameters were deter- mined from curves presented by Hardin and Dernevich (1972) and from data on the properties of the unconsoli- dated deposits compiled by Harold Olsen (written com- mun., 1972). Comparisons of the high-strain spectral response and the low-strain response for three sites are shown in figures 45, 46, and 47. The higher strain responses were computed using the bedrock motion from the San Fer- nando earthquake of February 9, 1971 at Pacoima Dam. The accelerogram was scaled according to the perpen- dicular distance of the sites from the San Andreas fault. The model calculations suggest that the principal ef- fect of high strain is to shift the frequencies of the higher modes to lower values and to reduce the amplification corresponding to these modes. For frequencies in the vicinity of the fundamental mode, there is only a slight change in the amount of amplification. The high-strain responses of the two bay' mud sites (figs. 46, 47) are distinctly different, even though the low—strain seismic models used for these two sites were identical. This difference is due to the difference in the input strain level at the two sites. The higher strain level (fig. 46) has two effects: (1) to increase the impe- dance contrasts at the bedrock-alluvium and alluvium-mud interfaces and (2) to increase the damp- ing of the seismic waves. The first effect tends to in- crease the amount of amplification, and the second tends to decrease it. However, since the attenuation increases with frequency, the higher level of input strain results in less amplification at the higher modes but increased amplification for the fundamental mode. A62 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 122°30’ 122°22’30” Verba B Mme was): EXPLANATION PREDICTED MAXIMUM EARTHQUAKE INTENSITY (1906 SAN FRANCISCO SCALE) Very violent 37°45’ Violent San Andreas fault ' A ' .7 R Reservoir 37°37’30” 2 3 MILES 2 3 KILOMETRES , ; \2- .\ FIGURE 40,—Maximum earthquake intensities predicted for San Francisco. Each value is the maximum of those predicted assuming a large earthquake on the San Andreas fault or the Hayward fault. The intensity values are predicted from the empirical relations (figs. 38, 39) based on only the good intensity data for the 1906 earthquake together with a generalized geologic map compiled by K. R. Lajoie (written commun., 1974). Letters A—E indicate grades of the San Francisco intensity scale (see section "1906 Intensity Scale for San Francisco”). STUDIES FOR SEISMIC ZONATION Surface 1 Bay mud Cl 11 m (36 ft) 12 m (39 ft) 40 m (131 ft) Alluvium 184 m (604 ft) FIGURE/ 4/l.—Sc/he/m/atic/ ge/o/lo/gic// section at //t/he site of the d//own/- h/ole seismometer array. In brief, these calculations show that an increase in level of input strain does not necessarily mean a de- crease in amplification for all frequencies. The model predicts that for the higher frequencies (frequencies more than about three times the funda- mental) the corresponding levels of ground shaking can be substantially reduced. Part of this reduction in amplitude may be an artifact of the quasilinear model because the high-strain parameters are chosen for the dominant frequency in the strain history. This possibil- ity has been raised by other researchers (for example, Dobry and others, 1971). Recent work with nonlinear models (W. B. Joyner, unpub. data) suggests that ground motions calculated using the quasilinear model are adequate for frequencies near the fundamental fre- quency, but at the higher frequencies the ground mo- tions are underestimated by the quasilinear model. Further investigation of the problem is needed. With the assumption that the high-strain modeling procedure is valid, it becomes possible to predict surface ground motions from postulated future earth- quakes. Such predictions have been made at four differ- ent sites along a profile perpendicular to the San An- A63 dreas fault and are described in Borcherdt, Brabb, and others (this report). SUMMARY The amount of damage in San Francisco from the 1906 earthquake was observed to depend strongly on the geologic character of the ground. This dependence suggests the need for zonation maps to reduce losses from future earthquakes. Comparative measurements of ground shaking gen- erated by nuclear explosions and the 1957 earthquake show that there is a significant and consistent difference in the response to shaking of different geologic units in the San Francisco Bay region. Comparison of the mea— sured amplifications with the 1906 intensities show that an increase in amplification corresponds to an in- crease in intensity. This correlation suggests that equidistant sites with large observed amplifications may also be sites of relatively high intensity in future earthquakes. These data together with available geologic information were used to predict the maximum intensity that sites in the San Francisco Bay region might sustain in another 1906—type earthquake on either the San Andreas fault or the Hayward fault (fig. 40; Borcherdt and Gibbs, 1975). Such a zonation, prop- erly interpreted, is a useful first step toward predicting the earthquake hazard associated with various geologic conditions. Such a zonation does not provide quantita- tive estimates of ground shaking, nor does it necessarily define the nature of the problems in the various areas, such as surface faulting, liquefaction, or landsliding. It does delineate many potentially hazardous areas and provides a basis for development of general landwuse policies to reduce the hazards of future earthquakes. The data from the 1957 earthquake suggest that the response to shaking of the various geologic units, as based on data from nuclear explosions, can be extrapo- lated quantitatively to maximum strain levels of ap- proximately 10‘4. For higher strain levels, the results of the numerical model suggest that there can be large amplification effects near the fundamental frequency of a site. However, further analysis of numerical proce- dures and more field data on the dynamic behavior of surficial geologic units at high-strain levels are needed before these preliminary model predictions can be used quantitatively with confidence in the design of earthquake-resistant structures. In summary, certain qualitative conclusions can be drawn regarding expected intensities of shaking on a regional scale from future earthquakes. The-data cur- rently available for the San Francisco Bay region suggest that the level of ground shaking will vary sub- stantially depending on the type of underlying geologic deposit. For sites equidistant from the fault, excluding A64 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION LOCAL EARTHQUAKE Surface AW 40 m (131 ft) HORIZONTAL MOTION I WmmMWWW 186 m (610 ft) I I I I I 10.0 15.0 20.0 25.0 TIME, IN SECONDS DISTANT EARTHQUAKE MWWWWAWWWGMWW 12 m 6%) WWW 40m (131 ft) HORIZONTAL MOTION W 186 m (610 ft) J I I J I 0 10.0 20.0 30.0 40.0 TIME, IN SECONDS 50.0 FIGURE 42.—Recordings of horizontal ground motion (N. 40° W. component) from a down-hole seismometer array near the margin of San Francisco Bay. Vertical scale within each set of records is the same to illustrate amplification effects of the soil profileA, Local earth- quake, magnitude 3.5, 79 km (49 mi) from recording site. B, Distant earthquake, magnitude 6.3, 480 km (298 mi) from site (San Fernando earthquake of February 9, 1971). those in the immediate fault zone, the effects of am- plified ground shaking are expected to be least for those sites underlain by bedrock, intermediate for those sites underlain by alluvium, and greatest for those sites un- derlain by artificial fill and bay mud. This qualitative classification of geologic units, together with the geologic maps of the San Francisco Bay region described in Lajoie and Helley (this report) provides a qualitative ground-response map of the San Francisco Bay region. The general areas for which the effects of amplified ground shaking are expected to be the greatest are also those areas that are generally considered to be most susceptible to liquefaction (see Youd and others, this report). A65 STUDIES FOR SEISMIC ZONATION o .om .Ava .mme axasvfihwm ownwfiom cam 23me 9.: N3 83.8an 983 2832: @35on 23. .N .mev .3: wstwfigww 303 m up wmugwnum 96? E832: @3330 9? JV dc 39 E wwfi mo flag a an M3968 E @33on :oEoE 59a rod—“Shane“. $859: 959% oomm‘am 135958 was @33on mo :oflhfifioOlfiV 559m o.mN 0 .ON V mDZOOmm z_ .m:>=._. 0.3 0.2 o.m o 9 28 E 9: B 53% m E :2on .3853; 8238 montam S 228:. .9358; 83320 825m E .559: 5:058: umSQEoo REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION A66 0.0m ohm odN .vmdcsnooldv 359% m woZOOmm Z_ .m_>__._. 0A.: 0.2 o.m _ _ 928 E 9: B Show N a 8:9: .3855; 8338 83:5 “a £030.: fiEoNCo; twimmno momtzm Hm cozoE 5:05.62 .083an0 STUDIES FOR SEISMIC ZONATION 15 I I I I I [ T 10 IIIIIIIIIIIII IIIIJIIIIIII 20 15 Observed SPECTRAL AMPLIFICATION 10 IIIIIIIII'IIIIII‘IWI I I I I I I l I | 2 3 4 FREQUENCY, IN HERTZ mIllllJll_l_IJ_L|Illlll O b—l FIGURE 44.-—Computed and observed spectral amplification curves for the site at the down-hole seismometer array. The curves were determined using the horizontal motion recorded in bedrock at a depth of 186 m (610 ft) from the San Fernando earthquake (fig. 423). 1° I I I I METRES FEEI 0 m 1 Bay mud SPECTRAL AMPLIFICATION 0 I I I O 1 2 3 FREQUENCY, IN HERTZ FIGURE 46.—Comparison of high— and low-strain spectral am- plification for a site underlain by 11 m (36 ft) of bay mud and 173 m (567 ft) of alluvium. The high-strain spectral amplification was computed by using as input the bedrock motion recorded at Pacoima Dam, (San Fernando earthquake of February 9, 1971) scaled to a maximum velocity of 19 cm/s (8 in./s). A67 METRES FEET SPECTRAL AMPLIFICATION FREQUENCY, IN HERTZ FIGURE 45,—Comparison of high- and low-strain spectral am- plification for a site underlain by 220 m (722 ft) of alluvium. The high-strain spectral amplification was computed by using as input the bedrock motion recorded at Pacoima Dam, (San Fer- nando earthquake of February 9, 1971) scaled to a maximum velocity of 31 cm/s (12 in./s). 10 T I I I METRES FEET 0 :Bay mud 9 — _ 200 8 _ 100 Alluvium _ 400 A 7 — ’ — 200 High strain ’ ‘rLow strain SPECTRAL AMPLIFICATION U'I \ ‘\ 3 1 2 _ 1 .— 0 I I I I 0 1 2 3 FREQUENCY, IN HERTZ FIGURE 47.—Comparison of high- and low-strain spectral am- plification for a site underlain by 11 m (36 ft) of bay mud and 173 m (567 ft) of alluvium. The high-strain spectral amplification was computed by using as input the bedrock motion recorded at the Pacoima Dam, (San Fernando earthquake of February 9, 1971) scaled to a maximum velocity of 12 cm/s (5 in./s). LIQUEFACTION POTENTIAL By T. L. YOUD, D. R. NICHOLS, E. J. HELLEY, and K. R. LAJOIE INTRODUCTION Liquefaction of clay-free granular layers has pro- duced abundant and sometimes catastrophic ground failures during earthquakes and hence must be consid- ered in assessing seismic risk or hazard. Conditions requisite for seismically induced liquefaction—— saturated unconsolidated deposits and high seismicity—are widespread in the San Francisco Bay region. Evaluation of the liquefaction potential of these deposits thus forms an important element in mapping seismic hazards of the area. This paper describes how a preliminary liquefaction-potential map of part of the San Francisco Bay region was made and describes types of ground failure commonly associated with liquefaction that might be expected to occur in that region. Map zones are based on detailed geologic studies of the unconsolidated sediments (Helley and Brabb, 1971; Helley and others, 1972; Lajoie and others, 1974; Nichols and Wright, 1971). Liquefaction potential is estimated from an analysis of maximum horizontal surface accelerations, duration of ground motion, depth of water table, and depth and standard penetration resistance of clay-free granular sediments. The analysis is based on the “sim- plified procedure for evaluating liquefaction potential” developed by Seed and Idriss (1971). The results were statistically averaged to provide an estimate of liquefaction potential for each zone. Liquefaction is defined here as the transformation of a granular material from a solid state into a liquefied state as a consequence of increased pore-water pres- sures (Youd, 1973). This definition distinguishes liquefaction as a transformation process rather than liquefied flow or a type of ground failure. Hence, a poten- tial for liquefaction does not necessarily indicate a simi- lar potential for ground failure. However, ground fail- ures are common consequences of liquefaction and hence can be expected to occur in areas susceptible to liquefaction. GEOLOGY AND SEISMICITY OF STUDY AREA The mapped area is bounded by the East Bay hills on A68 the east, the Santa Cruz Mountains on the west, and by the cities of Oakland on the northeast, San Francisco on the northwest, and San Jose on the south. The area contains the broad alluvial plain surrounding the southern part of San Francisco Bay. This plain is under- lain by late Cenozoic sediments that vary greatly in density and degree of consolidation. The sediments are subdivided into units with generally similar geotechni- cal properties, chosen for mapping liquefaction poten- tial (see Lajoie and Helley, this report). Those sediments whose grain-size distribution (clay-free sand and silt) and degree of lithification (completely uncemented) make them potentially liquefiable occur within two units; the older of these was deposited during late Pleis- tocene time, and the younger, during Holocene time. Environments of deposition during both the late Pleis— tocene and Holocene were similar to those of today ex- cept that marine and estuarine conditions were absent during parts of the late Pleistocene. The older deposits are denser and more consolidated and tend to be coarser grained. Because they have been long exposed to weathering processes and changing climatic regimes, they commonly contain well— developed soil profiles. Where exposed, the older de- posits are expressed geomorphically as slightly dissec- ted alluvial fans and aprons generally lying at higher altitudes near the margins of the plains, where they gradually merge into the surrounding foothills." Be- cause these fans are in the highest part of the plain, ground-water levels are generally deep but may be tem- porarily high during wet seasons. The younger alluvial deposits, which are much looser, wetter, and less consolidated than the older fan deposits on which they rest, grade into the modern sediments of San Francisco Bay. The interfingering of alluvial and estuarine (bay) sediments in these younger deposits reflects the post-Wisconsin marine transgression into the basin of San Francisco Bay. The San Francisco Bay region is very active seismi- cally, having been subjected to large historic earth- quakes originating nearby on both the San Andreas and Hayward faults. For example, the 1906 earthquake (magnitude 8.2) was accompanied by a continuous 306 STUDIES FOR SEISMIC ZONATION km (190 mi) surface rupture on the San Andreas fault. The 1868 (and possibly the 1836) earthquake on the Hayward fault also produced significant surface rup- tures. DESIGN EARTHQUAKES Sediments are classified by liquefaction potential as follows: (1) Sediments likely to liquefy in the event of a moderate earthquake (magnitude 6.5) originating nearby on the San Andreas, Hayward, or other local fault are considered to have a high liquefaction poten- tial; (2) sediments unlikely to liquefy even in the event of a major earthquake (magnitude 8.0) nearby on the San Andreas fault, are considered to have a low liquefac- tion potential; and (3) sediments between these two extremes are considered to have moderate liquefaction potential dependent on earthquake size and duration and sediment properties such as grain size and degree of sorting. A moderate-size event would be characterized by approximately 10 significant strong-motion cycles (Seed and Idriss, 1971) with maximum horizontal sur- face accelerations of 0.2 g or greater (Page and others, 1972) over much of the area, and a large event by as many as 30 significant strong-motion cycles (Seed and Idriss, 1971) with maximum horizontal surface acceler- ations of 0.5 g or greater (Page and others, 1972). These parameters are used in the following analyses. METHOD OF EVALUATING LIQUEFACTION POTENTIAL The method used to estimate liquefaction potential is based on the “simplified procedure for evaluating liquefaction potential,” which was developed for mate- rials that underlie relatively level surfaces (Seed and Idriss, 1971, p. 1249) and have relative densities (Dr)1 less than about 80 percent (p. 1256). Because slopes on the alluvial plain surrounding San Francisco Bay are small, the method can be applied over most of the area. However, because the large design earthquake could possibly produce liquefaction in sediments with relative densities greater than 80 percent, the procedure of Seed and Idriss was extended to permit evaluation of liquefaction potential for these extreme conditions. The simplified procedure is based on two basic rela- tions. First, the average cyclic shear stress (Tav), de- veloped during a given earthquake at a depth (h), be- neath a level surface is estimated from the equation Tav z 0.65rd'yh (amax/g), (1) 1Relative density (Dr), in percent, is defined as emax ‘ 2 Dr: emax ' e'min where emax and 9min are void ratios of a given granular material in its loosest and densest states, respectively, and e is the void ratio of the material at the density in question. (100), A69 where rd is an empirically determined stressreduction coefficient, 'y is the unit weight of the soil, amax is the maximum horizontal surface acceleration, and g is the acceleration of gravity. (Equation 1 is equation 4 of Seed and Idriss, 1971, p. 1256.) Second, the ratio of in situ cyclic shear stress (1') required to produce liquefaction in a given number of cycles (1) on laboratory samples molded at the in situ relative density (Dr) to the effective overburden pressure (00’) is related to results of laboratory cyclical triaxial compression tests as (#00310, z 0, (ode/20a)l50 (D, /50)for D, <80 percent,(2) where Cr is a correction coefficient applied to triaxial compression test results, “do is the cyclic deviator stress producing liquefaction in l cycles on a remolded sample of the in situ or similar material at a relative density of 50 percent, and o-a is the initial effective confining pres- sure. (Equation 2 is equation 6 of Seed and Idriss, 1971, p. 1258.) Empirical curves have been constructed by Seed and Idriss (1971) for estimating C r and (Ode/20a)50 from density state and gradational properties of the soil and the number of significant strong-motion cycles. Thus, by comparing the average cyclic shear stress (Tav) developed at any given depth (equation 1) with the cyclic shear stress (7) required to produce liquefaction of the materials at that depth (equation 2), a criterion is established for assessing liquefaction potential. For the moderate-size design earthquake, the follow- ing parameters were used in equations 1 and 2 to esti- mate limiting values of relative density at which liquefaction would be likely to occur: “max = 0-2 g, l = 10 cycles, 0.880 percent, (3) where the value of M is taken directly from the recon— structed curve. Next, T in equation 3 was equated with raw in equation 1 to solve for limitingM values at which liquefaction could occur: ‘ M 20.65 rd(yh/ao’) (amax/g). (4) For the large design earthquake, the following parametric values were used: amax=0.5g, [=30 cycles, and other values as given for the moderate-size design earthquake. Substitution of reasonable combinations of these values into equation 4 yields limiting M values between 0.3 and 1.0. This corresponds to relative den- sities ranging from 87 to 92 percent. (The range is nar- row because of the steep slope of the reconstructed curve, figure 48.) Thus, a limiting density of 90 percent was selected as the maximum at which liquefaction REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 0.6 I l I I I I I 0.5 STRESS RATIO :0 O .o N w J> .0 ._. 4O 50 6O 7O 8O RELATIVE DENSlTY, IN PERCENT 90 100 FIGURE 48.—Estimated stress ratio, (T/(To ’)lD,, required to produce liquefaction under field conditions during 30 cycles of seismic load- ing. Curve A, plot of equation 2 (equation 6 of Seed and Idriss, 1971). Curve B, reconstructed curve plotted from data given in Seed and Peacock (1971). might be expected to occur during a large earthquake in the San Francisco Bay region. The liquefaction-potential criteria can now be sum- marized as follows: Saturated clay-free granular sedi— ments with relative densities less than 65 percent are considered to have high liquefaction potential, even in a moderate earthquake; clay-free granular sediments with relative densities greater than 90 percent are con- sidered to have low liquefaction potential, even in a major earthquake; and saturated clay-free granular sediments with relative densities between 65 and 90 percent have moderate liquefaction potential that de- pends on intensity and duration of ground shaking and textural properties of the sediments. To facilitate application of the liquefaction criteria to field observations, relative densities were estimated from data on standard penetration tests using relations developed by Gibbs and Holtz (1957). This procedure also was used by Seed and Idriss (1971). Standard penetration versus depth curves taken from the rela- tions of Gibbs and Holtz are plotted in figure 49 for relative densities of 65 and 90 percent. Assumed parameters used in constructing these curves include a water-table depth of 3 m (10 ft) at the time of drilling and a dry density of 1.6 g/cm3 (dry unit weight=100 lb/ft3). MAPPING OF LIQUEFACTION POTENTIAL Boring logs from throughout the study area were col- lected from numerous private consultants and gov- ernmental agencies. Standard penetration data from clay-free granular deposits within 15 m (50 ft) of the STUDIES FOR SEISMIC ZONATION 0 I 3 _ _ (10) P u 3 ‘8 6 3; 3. _ o o _ (20) g, g Liquefaction (30) _ potential High Moderate Low DEPTH, IN METRES (FEET) Lo 12 (40) 15 _ (50) 18 I | (60) 0 20 40 6O STANDARD PENETRATION RESISTANCE, IN BLOWS PER 0.3 METRE (1.0 FOOT) FIGURE 49. Criteria used for estimating liquefaction potential in the field. Based on a correlation between standard penetration resistance and relative density developed by Gibbs and Holtz (1957). Curves are based on the assumption that the water table was 3 m (10 ft) deep when penetration data were obtained from exploratory borings. surface were compiled and statistically analyzed in each of the generalized map zones shown in figure 50. These zones were derived from geologic maps of unconsoli- dated sediments (Helley and Brabb, 1971; Helley and others, 1972; Nichols and Wright, 1971; Lajoie and others, 1974.) Zone 1 is an area of bay mud (9,000 years and younger) overlying Holocene and late Pleistocene alluvium. Clay-free granular layers are not generally A71 present in the bay mud but do occur locally near present and former stream channels. Clean granular layers are rather common, however, in the Holocene alluvium and the late Pleistocene alluvium beneath the bay mud. The surficial deposits in zone 2 are Holocene alluvium (less than 10,000 years old) overlying late Pleistocene al- luvium. This zone is subdivided on the basis of water- table depth. Areas in which the water table is normally at 3 m (10 ft) or less are labeled 2a. Areas with a deeper water table are labeled 2b. The Holocene alluvium is generally greater than 3 m (10 ft) thick in subzone 2a and thus extends below the water table. Hence, the Holocene alluvium in this zone has a continual poten- tial for liquefaction. In subzone 2b, the Holocene al- luvium is generally less than 3 m (10 ft) thick and thus is normally above the water table. It therefore has at most only a seasonal potential for liquefaction. Zone 3 is an area of late Pleistocene alluvium consisting of over- consolidated alluvial-fan deposits that extend to a con- siderable depth. Table 5 summarizes the percentages of available standard penetration test data plotting in each category of liquefaction potential for each geologic unit. The data show that granular layers within the bay mud (zone 1) have a generally high potential for liquefaction. Seventy-three percent of the penetration data from these layers indicate high potential for liquefaction, and an additional 21 percent indicate moderate potential. Only 6 percent of the penetration data indicate low potential. The granular layers beneath the bay mud (zone 1) show a much lower, but still significant, potential for liquefaction. Thirty—three percent of the data from these layers indicate high potential for liquefaction, and an additional 28 percent suggest moderate potential. The overall liquefaction potential of this unit is classed as moderate. Virtually all the bay mud and the deposits underlying the bay mud lie below the normal ground- water table; thus, they present a persistent potential for liquefaction. Clay-free granular layers within the Holocene al- luvium in zone 2 show on the average less potential for liquefaction than those within or beneath the bay mud because of both their greater density and the greater depth of the water table. Their potential, however, is still classed as moderate. Twenty-two percent of the penetration values for these sediments indicate high potential for liquefaction, an additional 33 percent indi- cate moderate liquefaction potential, and 45 percent indicate low potential. The Holocene alluvium of sub- zone 2b would be, at most, seasonally or intermittently liquefiable because it is normally above the water table. Although not specifically shown in figure 50 or table 5, granular layers in the relatively recent channel and A72 overbank deposits along present drainage ways are generally characterized by lower penetration resistance and, consequently, higher liquefaction potential than deposits in the adjacent alluvial plains. Liquefaction potential of granular layers in the late Pleistocene alluvium (zone 3) is generally low. Only 11 percent of the penetration values from these sediments indicate high potential for liquefaction, whereas 60 per- cent indicate low potential. Liquefaction potential in this zone is further diminished by a relatively high topographic position and hence a deep water table. TABLE 5.—Summary of the analysis of liquefaction potential using standard penetration data and criteria plotted in figure 49 ['IVvo probable local earthquakes are considered; (1) a moderate event (magnitude =65) and (2) a large event (magnitude 8.0). Sediments likely to liquefy during a moderate event are classified as having high liquefaction potential; those unlikely to liquefy during a large event are classified as having low liquefaction potential; and those between these two categories are classified as having moderate liquefaction potential] Standard penetration test data Percent indicating Number Zone Sedimentary unit of tests Dr<65 percent 65 percent90 percent (high <90 percent (low liquefaction (moderate lique- liquefaction potential) faction potential) potential) 1 Deposits within bay mud ,,,,,,,,, , 73 21 6 53 ] Deposits under- lying bay mud ,1 n 33 28 39 155 2a, Holocene 2b alluvium ,,,,,,,, 22 33 45 708 3 Late Pleistocene alluvium ,,,,,,,, 11 29 60 357 GROUND FAILURES ASSOCIATED WITH LIQUEFACTION Three types of ground failure are commonly associated with liquefaction (Seed, 1968; Youd, 1973). (1) Flow landslides are failures that generally occur on moderate to steep slopes underlain by loose granular deposits. In this case, once liquefaction has occurred, flow deforma- tion commences and continues unabated until the driv- ing shear forces are reduced (as by slope reduction) to a value less than the viscous shear resistance of the liquefied soil. When that state is reached, the material stops flowing and solidifies, usually far from the point of origin. Loose granular deposits on moderately to steeply sloping hillsides in the San Francisco Bay region could be susceptible to this type of failure if they were satu- rated. Such failures occurred on San Bruno Mountain near Colma and near Half Moon Bay during the 1906 San Francisco earthquake (Crandall, in Lawson, 1908, p. 249; Anderson, in Lawson, 1908, p. 395). Because of the generally small slopes, it is unlikely that this type of failure would occur on the broad alluvial plain sur- rounding San Francisco Bay. (2) Lateral-spreading landslides are failures that oc- REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION cur most commonly on gentle to nearly horizontal slopes underlain by loose to moderately dense granular de- posits or layers. In this type of failure, liquefaction oc- curs and flow commences; however, after a finite dis- placement, flow is arrested by a drop in pore-water pres- sure resulting from the tendency for all but very loose granular sediments to dilate during shear. Continued shaking may cause reliquefaction (provided the shak- ing causes reversals in shear stress (Seed and Lee, 1969; Youd, 1973)), and a second episode of flow displacement may occur followed by restabilization. This sequence may continue as long as strong shaking continues. Dis- placements ranging from nearly zero to tens of metres have been produced by these kinds of failures (Varnes, 1958; Youd, 1973). Factors that contribute to greater displacement include greater duration of shaking, loose sediments, and optimal slope conditions. (Slopes that are too flat inhibit movement, and slopes that are too steep inhibit reversals in shear stress necessary for the generation of repeated episodes of liquefaction (Youd, 1973).) Cracks, fissures, and differential settlement are common on, and especially at the margins of, lateral- spreading failures. Although these features and accom- panying slide movements may appear rather inconse- quential in open terrain, they have proved to be very damaging and disruptive to structures and utilities con- structed across, on, or within the slide mass. Lateral spreading probably would be the most perva- sive type of ground failure associated with liquefaction on the broad alluvial plain surrounding San Francisco Bay. Sediments containing granular layers, especially the bay mud and recent channel and overbank deposits in the Holocene alluvium, probably would be the mate- rials most susceptible to this type of failure because of their greater potential for liquefaction and generally uncompacted state, which would permit greater slide movement after liquefaction. Least susceptible to this type of failure would be the late Pleistocene alluvium (zone 1) because of its low potential for liquefaction and generally dense state, which in turn would prevent sig- nificant displacements from occurring even if liquefac— tion should develop. Evidence of lateral spreading was reported at several places within the study area during the 1906 San Fran- cisco earthquake. Most of these slides occurred in the susceptible areas listed above. For example, lateral ground movements, some as large as 2 m (6 ft), occurred in several areas of San Francisco Where artificial fill is underlain by bay mud (zone 1) (Wood, in Lawson, 1908, p. 220—245). In addition, lateral displacement of flood- plain deposits toward the depressions of Alameda and Coyote Creeks was mentioned specifically (Lawson, 1908, p. 400). Many lateral-spreading landslides gener- ated by the 1906 earthquake may not have been re- STUDIES FOR SEISMIC ZONATION A73 122°30’ 15’ 37°45’ ‘ 30’ EXPLANATION a», LIQUEFACTION POTENTIAL Vii» ZONE GEOLOGIC UNIT OF CLAY-FREE GRANULAFI LAYERS Bay mud and Generally moderate, locally - underlying high where clean granular sediments layers are present In bay mud Holocene Moderate, water table normally alluvium above 3 metres (10 feet) Holocene Moderate, water table normally alluvium below 3 metres (10 feet) """ Late Pleistocene : : : . : : alluvium Generally low 37° 15’ 0 2 4 6 MILES Isa—Is—rL—I 0 2 4 6KILOMETRES 122°00’ FIGURE 50,—Preliminary map showing liquefaction potential for the southern San Francisco Bay region. The map shows generalized lique- faction potential of granular layers in each map zone but does not delineate locations of these layers. Hence, the map is useful for desig- nating zones where special consideration should be given to the possibility of liquefaction but is not valid for assessing the liquefaction potential of a given site. corded, especially in undeveloped areas near the bay that were not thoroughly investigated. (3) Quick-condition failures have occurred histori- cally most often in flat areas with high water tables and loose to moderately dense granular sediments extend- ing from near the surface to substantial depths. In this situation liquefaction may lead to a quick condition and often to the loss of bearing capacity with the result that structures, embankments, or other loads founded on the surface sink into the liquefied sediments. At the same time buried tanks or other vessels may rise buoyantly. Other than subsidence of several roadway fills in San A74 Francisco, which may or may not have been due to liquefaction, the authors have found no reports that this type of failure occurred in the study area during the 1906 San Francisco earthquake. SUMMARY A preliminary map of liquefaction potential has been compiled for the southern San Francisco Bay region. This map delineates zones in which existent clay-free granular layers are estimated to have low, moderate, or high liquefaction potential. The map zones are derived from detailed geologic studies of the unconsolidated sed- iments in the region. Liquefaction potential of the granular layers is estimated from an analysis of lithologic, water table, standard penetration, and seis- mic data. Areas underlain by bay mud containing clay—free granular layers have a generally high potential for liquefaction. Granular layers underlying the bay mud have a significant but lower potential, which is classed herein as moderate. Granular layers within the Holocene alluvium have an even lower, but still moder- ate, potential; furthermore, most of this unit is normally above the water table and thus, at most, is only season- ally or intermittently susceptible to liquefaction. Clay— free granular layers within channel and recent over- bank deposits of the Holocene alluvial unit generally are characterized by a greater liquefaction potential than adjacent deposits in the alluvial plain. Granular REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION layers in the late Pleistocene alluvium generally have low potential for liquefaction. Zones delineated in this study as having significant liquefaction potential indicate areas in which the liquefaction process may occur in existing clay-free granular layers; unfortunately, insufficient data are presently available to plot the actual locations of these layers. The data also give no indication of type or amount of ground failure, if any, that might follow liquefaction. However, lateral-spreading landslides are a common consequence of liquefaction beneath gentle slopes. Hence, in the event of a major earthquake, this type of failure is likely to be a result of liquefaction beneath the alluvial plain surrounding San Francisco Bay. Reports from the 1906 San Francisco earthquake verify this conclusion. The criteria used for evaluating liquefaction poten— tial are based on empirical procedures formulated by Seed and Idriss (1971) and approximate estimates of ground-motion parameters and geotechnical properties. In addition, the estimated potential of each zone is based on somewhat limited data generalized to include the entire map unit. Thus, the map of liquefaction potential must be considered preliminary and approximate and not valid for direct determination of liquefaction poten- tial at any specific site. However, despite its limitations, the map should serve the intended purpose of generally delineating areas where the probability that liquefac- tion will occur during a major earthquake is greatest and hence areas where special attention is required. LANDSLIDES By T. H. NILSEN and E. E. BRABB INTRODUCTION Landslides are characteristically abundant in areas of high seismicity and steep slopes. Landslides asso- ciated with earthquakes may cause as many or more fatalities as the initial fault rupture and shaking of the ground. They may also occur long after an earthquake, having been caused or aided by the loosening, shaking, and disruption of the deposits on slopes during the earthquake. As a result, the landslide hazards related to earthquakes may persist long after the ground has stopped shaking and therefore can be a long-term prob— lem. Some of the major earthquakes that have occurred during the past 15 years have vividly demonstrated the hazards of seismically triggered landslides. The Hebgen Lake, Mont., earthquake of 1959 triggered a very large landslide (fig. 51) that killed and injured many people, formed a temporary lake, and blocked travel in the area (Hadley, 1964, fig. 54). The Anchorage, Alaska, earth- quake of 1964 triggered both extensive subaerial (fig. 52) and submarine landslides; tsunamis (seismic sea waves) generated by the submarine landslides caused extensive damage and many fatalities in coastal areas (Hansen and others, 1966). The earthquake in western A». . 3“}; u FIGURE 51.——-Earthquake-generated landslide at Hebgen Lake, Mont. This landslide, one of the largest ever recorded in the United States, was responsible for 26 fatalities and many injuries during the Au- gust 17, 1959, earthquake (magnitude 7.1). Photograph by US. Forest Service. Peru in 1970 triggered a massive debris avalanche (fig. 53) that destroyed the cities of Yungay and Ranrahirca; it caused probably about one-half of the 38,000 fatalities attributed to the earthquake (Plafker and others, 1971). The San Fernando, Calif, earthquake of February 9, 1971, triggered more than 6,000 individual landslides in the surrounding upland areas, most of them of small size (fig. 54); however, only a few damaged manmade structures because residential and industrial develop- ment had been restricted almost wholly to the relatively flat floor of the San Fernando Valley (Morton, 1971; oral commun., September 1973). Each of the major earthquakes described above had magnitudes greater than 6.5. Although smaller earth- quakes may cause less damage (or none) to manmade structures by ground shaking, they are capable of trig- gering slope failures in hillside areas, especially re- newed movements of old, marginally stable landslide deposits. For example, an earthquake in February 1972, with magnitude 5.0 and epicenter located 37 km (23 mi) south of Hollister, Calif, triggered considerable downslope movement on a large old landslide deposit on Halls Ranch, near Paicines (fig. 55; Rogers, 1972). Had FIGURE 52.—Oblique aerial view of Turnagain Heights landslide in Anchorage, Alaska, triggered by earthquake of March 27, 1964 (mag- nitude 8.4). The landslide destroyed more than 75 homes. Photograph by US Army. A75 . A76 Avalanche 1 source Lgnas. Lianganuco FIGURE 53.———Oblique aerial view of debris avalanche that destroyed the towns of Yungay and Ranrahirca in western Peru during the earthquake of May 31, 1970. The avalanche originated at an al- titude between 5,500 and 6,400 m (18,000 and 21,000 ft) and moved 2.4 km (1.5 mi) downslope to the Rio Santa at a speed in excess of 280 km/hr (175 mi/hr) (Plafker and others, 1971, p. 550v558). More than 18,000 people were killed by the avalanche, and an additional 20,000 persons were killed by other effects of the earthquake. the hillside area been developed for residential purposes at the time of the earthquake, considerable damage might have resulted. The most extensive records of seismically triggered landslides in the San Francisco Bay region are descrip- tions of the California earthquake of April 18, 1906, and the San Francisco earthquake of March 22, 1957. The 1906 earthquake triggered many landslides throughout the bay region (Lawson and others, 1908, p. 389—399). Photographic evidence and eyewitness accounts indi- cate that the earthquake triggered many different types of landslides, resulting in several fatalities and major damage to nearby manmade structures. Fortunately, in 1906 most of the residential and industrial development was concentrated along the gently sloping margins of REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION FIGURE 54.—0blique aerial view of landslides (white patches) in Lopez Canyon area, north of San Fernando, Calif, triggered by the earthquake of February 9, 1971 (magnitude 66). Photograph by D. M. Morton. FIGURE 55.—-Transverse cracks formed in upper part of old landslide deposit on Halls Ranch near Paicines, central California, by re newed movement during a moderate earthquake (magnitude 5.0) in February 1972. Photograph by T. H. Nilsen. San Francisco Bay and in some interior valleys, and so the total amount of damage caused by the numerous landslides was not catastrophic. Some landslides that occurred in 1906 provide useful background data for predicting slope failures in future earthquakes. Earthflow-type landslides were triggered locally on very gently sloping surfaces where the ground moisture content was high (fig. 56A); under normal conditions, very few landslides in the San Francisco Bay region develop on slopes of less than 15 percent (Bonilla, 1960a; Brabb and others, 1972; Nilsen and others, 1974). Other earthflow-type landslides developed on steeper slopes, and some were observed to have moved during short periods of time (figs. 563, C). Because the STUDIES FOR SEISMIC ZONATION C FIGURE 56.—Earthflow-type landslides triggered by the earthquake of April 18, 1906 (magnitude 8.3), near Half Moon Bay, Calif. A, Landslide on very gentle slope in water-saturated ground. The scarp is about 3 m (10 ft) high. Photograph 66 of J. C. Branner collection. B, Landslide on steeper slope in same general area asA. Photograph 69 of J. C. Branner collection. C, Landslide at Nunez Ranch. The flow originated about 150 m (500 ft) above the valley floor and took about half an hour to reach the base of the hill (Lawson and others, 1908, p. 397). Photograph 64 of J. C. Branner collection. Note: This photograph and others from the Branner and Gilbert collections are available for inspection at the US. Geologi- cal Survey library, Menlo Park, Calif. A77 1906 earthquake occurred in April, at the end of the rainy season, the ground probably was particularly sus- ceptible to failure by earthflow processes because of high moisture content. Other landslides triggered in 1906 damaged man- made structures such as railroads located in coastal areas (fig. 57). Some landslides that occurred a year or more after the earthquake were probably initiated by the earthquake. In one such example from the northern San Francisco Bay region (fig. 58), open fractures in the ground appeared immediately after the earthquake, but massive failure did not occur until the next rainy sea— son. The 1957 earthquake occurred near Daly City. Al- though the earthquake was of only moderate magnitude (5.3), it triggered landslides that caused considerable damage to manmade structures (fig. 59; Bonilla, 1959). Landsliding is a severe and continual problem in the FIGURE 57.—Roadbed of the coastal Ocean Shore railroad, south of San Francisco, damaged by landslides triggered by the 1906 earth- quake. Photograph 78 of J. C. Branner collection. FIGURE 58,—Landslide located about 6 km (4 mi) north of Bolinas Lagoon, Marin County, Calif, generated in March 1907. Cracks that formed during the earthquake of April 1906 contributed to the triggering of this landslide. Photograph 34 of the G. K. Gilbert collection. A78 FIGURE 59.—Landslides in artificial fill along the shore of Lake Merced, near San Francisco, triggered by the San Francisco earth- quake of March 22, 1957 (magnitude 5.3). Photograph by ~M. G. Bonilla. San Francisco Bay region, as indicated by the high annual costs of landslide damage. Taylor and Brabb (1972) estimated an overall public and private cost in the bay region of more than $25 million from landslides generated during the rainy season of 1968—69. In another study Nilsen and Brabb (1972) determined that a single landslide, which has been active over a period of at least 10 years, cost the city of San Jose more than $750,000. This brief review of the effects and history of landslid- ing in the San Francisco Bay region and other areas points out the severity of the problem and the mag- nitude of the potential hazard. However, we are only beginning to understand many aspects of the problem and are only in the early phases of developing methods for predicting the location, distribution, and types of landslides that would be triggered by a major earth- quake in the bay region. Prediction of an. individual slope failure for a given earthquake requires an under- standing of all the factors that contribute to the slope- failure process and detailed investigations of specific site characteristics. Such detailed information is not available on a regional scale in the San Francisco Bay region. Our work has included (1) historical studies of landsliding during past earthquakes, (2) mapping of old landslide deposits, (3) study of bedrock units that are susceptible to landsliding, (4) studies of recent landslid- ing and its relation to slope, bedrock geology, rainfall, and areas underlain by old landslide deposits, (5) the effects of development and construction activities on landsliding, and (6) some of the economic costs as- sociated with landsliding. This report briefly describes landslide processes, dis- cusses basic data pertinent to analyzing the landslide problem on a regional scale, and utilizes these data to delineate areas within the San Francisco Bay region REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION with various levels of landslide susceptibility during earthquakes. PREVIOUS WORK Most previous geologic studies in the bay region have focused on bedrock units and their structural and stratigraphic relations rather than on surficial deposits and their slope-stability characteristics. Consulting en- gineering geologists have examined the slope-stability characteristics of many small parcels of land in detail, but little of this work has been published; moreover, few regional studies have been undertaken by either con- sultants or governmental agencies. Some of the earlier studies concerned with landslides, slope-stability characteristics, and engineering properties of bedrock and surficial units provided data and valuable contribu- tions to our studies. These include studies by Schlocker, Bonilla, and Radbruch (1958), Bonilla (1960a, 1960b, 1971), Radbruch (1957, 1969), Radbruch and Weiler (1963), Kojan, Foggin, and Rice (1968), Harding (1969), Clague (1969), Pampeyan (1970), Rogers (1971), Waltz (1971), Huffman (1972a, b), Rice and Strand (1972), Burnett (1972), Radbruch and Wentworth (1971), Taylor and Brabb (1972), Nilsen and Turner (1974), and Nilsen, Taylor, and Brabb (1974). These and current studies have yielded important data for analyzing the landslide problem on a regional scale. LANDSLIDE PROCESSES Landslides are the downward and outward movement of slope-forming materials composed of natural rock, soils, artificial fill, or combinations thereof (Eckel, 1958, pl. 1). They move along surfaces of separation by falling, sliding, and flowing, giving rise to many characteristic features (fig. 60). Landslide deposits range in appearance from clearly discernible, largely unweathered and uneroded topo- graphic features to indistinct, highly weathered and eroded features recognizable only by their characteris- TRANSVERSE 5 CRACKS / TRANSVE RSE RIDGES RADIAL CRACKS FIGURE 60,—Nomenclature of parts of a landslide (from Eckel, 1958). STUDIES FOR SEISMIC ZONATION tic topographic configurations. Topographic and as- sociated features useful in recognizing landslide de- posits include (1) small isolated ponds, lakes, and other closed depressions, (2) abundant natural springs, (3) abrupt and irregular changes in slope and drainage pattern, (4) hummocky irregular surfaces and flat or backtilted areas, (5) smaller landslide deposits that are commonly younger and form within older and larger landslide deposits, (6) steep, arcuate scarps at the upper edge of the deposit, (7) irregular soil and vegetation patterns, and (8) disturbed vegetation. Landslides commonly are classified by the type of material underlying the slope before it moved, by the DEBRIS SLIDE Incoherent or broken masses of rock and other debris that move downslope by sliding on a sur- face that underlies the deposit WW 2/ 75 [WWW%% i I l/MI/IWII/Ili/i’ y . WWII/Wm mm ” SLUMP Coherent or intact masses that move downslope by rotational slip on surfaces that underlie as well as penetrate the landslide deposit A79 type of movement, and by the amount of water in the material. Four common types of landslides found in the San Francisco Bay region are shown in figure 61. The formation of landslides under national conditions is affected by (1) type of earth materials— unconsolidated, soft sediments or surficial deposits will move downslope easier than consolidated, hard bedrock; (2) structural properties of earth materials—the orien— tation of the layering of some rocks and sediments rela- tive to slope directions, as well as the extent and type of fracturing and crushing of the materials, will affect landslide potential; (3) steepness of slopes—landslides occur more readily on steeper slopes; (4) water—— EARTHFLOW Colluvial materials that move downslope in a manner similar to a viscous fluid ”I M? /////é// \ / / ROCKFALL Rock that has moved primarily by falling through the air FIGURE 61.-—Common types of landslides in the San Francisco Bay region. A80 landsliding is generally more frequent in areas of sea— sonally high rainfall because the addition of water to earth materials commonly decreases their resistance to sliding; water decreases cohesive forces that bind clay minerals together, lubricates surfaces along which slip- page may occur, adds weight to surficial deposits and bedrock, reacts with some clay minerals, causing volume changes in the material, and mixes with fine- grained unconsolidated materials to produce wet, unst- able slurries; (5) type of vegetation—trees with deep penetrating roots tend to hold bedrock and surficial de- posits together, thereby increasing ground stability; (6) proximity to areas undergoing active erosion ——rapid undercutting and downcutting along stream courses and shorelines makes slopes in these areas particularly susceptible to landsliding; (7) earthquake-generated ground shaking—strong ground shaking can trigger failures at the time of the earthquake and can jar and loosen hillside materials leading to failure at some later time. These are some of the many complex interrelated factors that may contribute to the formation of land- slides, with earthquake-generated ground shaking being only one of several possible triggering mechan- ismsi The prediction of a slope failure at a specific site from‘ a specified level of ground shaking requires an analysis of all such factors, as well as very detailed and expensive onsite investigations. Because such infor- mation is not available on a regional scale and we are only beginning to understand the landslide problem, we have approached the study of earthquake—induced land- slides on a regional scale by trying to delineate relative slope stability on the basis of physical properties. Be- cause landslides triggered by the 1906 earthquake oc- curred throughout the San Francisco Bay region, de- lineation of unstable areas according to their physical characteristics provides a guide to those areas most susceptible to earthquake-induced landsliding. SLOPE STABILITY ANALYSIS ON A REGIONAL SCALE Relative slope stability maps can be prepared in vari- ous ways and from various types of information. No generally applicable formula or technique has been de- veloped that covers all situations and all areas. Differ- ent techniques have been used to prepare relative slope stability maps for different areas, at different scales, for different purposes, and from different types of informa— tion. Many examples of the widely divergent form and style of such maps have been published in recent years for California (Blanc and Cleveland, 1968; Johnson and Ellen, 1965; Johnson and Lobo-Guerrero, 1968; Rogers, 1971; Brabb and others, 1972; Rice and Strand, 1972; REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION Huffman, 1972a, 1972b, 1974; Burnett, 1972; Frame, 1973; Radbruch and Wentworth, 1971; Saul, 1972) and for other parts of the United States (Bailey, 1971; Van Horn, 1972a, b, c; Williams, 1972; Scott, 1972; Maberry, 1972a, 1972b; Simpson, 1973a, 1973b). Complexly interrelated factors contribute to the gen— eration of landslides, and engineering geologists com- monly spend months preparing sophisticated analyses of soil and rock-strength parameters, precipitation re- cords, slope geometry, and other factors to determine the causes of individual landslides. To cover an area as large as the San Francisco Bay region (19,300 km?) (7,450 mi2), no such detailed analysis could be made owing to limitations of time, personnel, and available data. Instead we utilized data that (1) were available at the present, (2) were available throughout the entire map area, (3) could be incorporated easily into the slope- stability analysis, and (4) yielded information about some of the most important factors that control slope stability. The only data that met all these criteria were landslide distribution, slope, and bedrock geology. LANDSLIDE DISTRIBUTION Numerous studies in the bay region and elsewhere have shown that most landslides in a particular year occur in areas of previous landsliding (Nilsen and Turner, 1974; Kojan, 1973; Bailey, 1971). Commonly the new landsliding consists of renewed movements of old landslides triggered by earthquakes, unusually in- tense rainfall, and (or) man’s activities. Hence, the prep- aration of maps showing the distributuion of present landslide deposits is a first step toward delineating areas likely to fail in the future. In addition, such maps are useful for identifying major factors contributing to the formation of landslides. Maps showing the distribution of landslides in most of the San Francisco Bay region have been prepared for ' this purpose at a scale of 1:62.500, primarily by photoin- terpretation with a minimum of field checking. This technique is necessary because of the large size of the area, the inaccessibility of much of it, and time limita- tions. The photointerpretive techniques depend upon the recognition of scarps, anomalous bulges and lumps, hummocky topography, ridge-top depressions and trenches, terraced or backtilted slopes, abrupt changes in slope, altered stream courses, discontinuous drainage patterns, closed depressions, springs, and anomalous color, texture, shade, vegetation, and bedrock patterns. A number of maps have been published, and many others are in preparation (Brabb and Pampeyan, 1972; Nilsen, 1971, 1972a, b, c, d, 1973a, b, c; Sims and Nilsen, 1972; Burnett, 1972; Rice and Strand, 1972). An example of the landslide distribution in San STUDIES FOR SEISMIC ZONATION Mateo County as mapped by Brabb and Pampeyan (1972) is shown in figure 62. The inventory shows the presence of more than a thousand landslide deposits in the region and indicates that landsliding is one of the major erosional processes. Recently, a map of the dis- tribution of landslide deposits in the nine-county San Francisco Bay region was compiled at a scale of 1:125,000 for a slope-stability analysis of the entire region by Nilsen and Wright (unpub. data). An example from the northeastern bay region is shown in figure 63. Although the type of movement, date of most recent activity, and nature of landslide materials were not determined, the maps by themselves can be used as a general guide to areas where landslides may be a prob- lem; they provide a regional picture of the past history of landsliding, and they are useful to planners and en- gineering geologists in making preliminary appraisals of building sites. In addition to the landslide-distribution maps, an isopleth map has been prepared for the southern San Francisco Bay region (Wright and Nilsen, 1974; Wright and others, 1974). This map shows contours depicting variations in geographic density of landslide deposits and permits rapid, quantitative evaluations of the abundance of landslide deposits in different areas. Landslide deposits are abundant along some active faults or parts of active faults and uncommon along others. Where abundant, they may be related to either seismicity along the fault zone, the weak, crushed rocks found in the fault zone, or both. Abundant landslide deposits have been mapped along the Calaveras fault and along some parts of the San Andreas and Hayward faults. The maps of landslide deposits permit correla- tions and comparisons of the distribution of landslide deposits with other factors, but so far they have not in themselves permitted the recognition of those landslide deposits that were originally triggered by earthquakes. SLOPE Degree ofslope is an important parameter controlling the stability of hillside materials. A slope map of the San Francisco Bay region has been prepared by the US. Geological Survey (1972) at a scale of 1:125,000. Studies by Bonilla (1960a), Brabb, Pampeyan, and Bonilla (1972), and Nilsen, Taylor, and Brabb (1974) showed that in the bay region most landslides occur on slopes greater than 15 percent, with very few on slopes of 5—15 percent and virtually none on slopes less than 5 percent. As part of the current study, Nilsen and Wright (unpub. data) have prepared a generalized slope map of the San Francisco Bay region at a scale of 1:125,000 showing the slope intervals 0—5 percent, 5—15 percent, and greater than 15 percent. (An insert from their map is shown in fig. 64.) The generalized slope map shows only A81 areas wider than about 300 m (1,000 ft) and eliminates the thousands of very small, discontinuous areas shown on the original slope map. The generalized map provides an important element for evaluating slope stability on a regional scale. BEDROCK GEOLOGY Certain bedrock units are more susceptible to land- sliding than others because of their physical and chemi- cal characteristics, as well as the types and thicknesses of soils that tend to develop over these rock types. Thus, two adjacent areas that may appear to be similar in most resepcts may differ greatly in landslide suscepti- bility because of the type of bedrock underlying them (Radbruch and Weiler, 1963; Radbruch and Case, 1967; Brabb and others, 1971). For purposes of evaluating slope stability, a map showing the distribution of geologic units in the San Francisco Bay region was prepared at a scale of 1:125,000. On the basis of discussions with geologists at the US. Geological Survey responsible for bedrock mapping and research into the physical properties of hillside materials, certain of these units were judged to be especially susceptible to landsliding; these units are shown in figure 65 for an area in the northeastern bay region. RELATIVE SLOPE STABILITY OF THE SAN FRANCISCO BAY REGION Nilsen and Wright (unpub. data) prepared a slope stability map of the nine-county San Francisco Bay region at a scale of 1:125,000. An example for a part of the northeastern bay region is shown in figure 66. The map delineates areas according to five categories of relative slope stability ranging from generally highly stable to generally unstable. The five categories are defined and tabulated (fig. 66). The map incorporates the most recent data on mapped landslide deposits, slope, and bedrock geology (refer to figs. 63, 64, 65, respectively) and delineates areas in more detail than an earlier map showing relative abun- dance of landslides at a scale of 1:500,000 (Radbruch and Wentworth, 1971); it does not incorporate as much detailed data, especially from field investigations, as earlier slope-stability maps of smaller areas prepared at larger scales (for example, Brabb, Pampeyan, and Bonil- la, 1972). The map can be used only for regional scale investigations and not for slope-stability analyses of in- dividual lots or small subdivision areas. The map provides a generalized regional representa- tion of relative slope stability in the San Francisco Bay region. Although it does not predict that particular slopes will fail during future earthquakes of specific size A82 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 122°30’ 122° 15’ .122 San Francisco international Airport 37°30’ — EXPLANATION . Landslide deposit 37°l5’ — 0 4 MILES }‘_|_L|_|J—T—L—‘ o 4 KILOMETRES _—_)_ 2 l FIGURE 62.-—Landslides in San Mateo County, Calif. More than 1,000 landslides have been mapped in this area. From Brabb and Pampeyan (1972). STUDIES FOR SEISMIC ZONATION A83 and location, the map is useful for delineating areas unstable areas are concentrated along the ridges paral- according to their potential for earthquake-induced leling the San Andreas and Hayward faults, these areas landslides. It defines the relative slope stability of gen- must be considered most susceptibleto landsliding dur- eral areas on the basis of existing physical characteris- ing moderate to large earthquakes. Delineation of such tics. Because the 1906 earthquake generated landslides areas is a basic tool for the development of land-use throughout the region and because most of the more policies for reducing the hazards associated with Kim”. ‘ 1 ‘ / EXPLANATION Landslide depOSlt larger than 150 m (500 ft) in longest dimension Landslide deposit smaller than 150 m (500 ft) in longest dimension 3 4 MILES 0 L l l l l I l I l l l I 0 1 2 3 4 KILOMETRES FIGURE 63.—Distribution of landslide deposits in part of northeastern Contra Costa County, Calif. After Sims and Nilsen (1972) and Nilsen (1971). A84 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION earthquake-induced landslides. Because the relative- SUMMARY slope-stability map is based on bedrock geology, slope, and the distribution of past landslide deposits, the map Landsliding is continually a major hazard to life and also defines general areas susceptible to landsliding property in the San Francisco Bay region. Studies of triggered by other mechanisms such as rainfall and landslide cost showed that at least $25 million was man’s activities. spent in 1968—69 on landslides triggered by processes EXPLANATION Area of 0- to 5-percent slope Area of 5- to 15-percent slope Area of slope greater than 15 percent SCALE 1:125 000 2 o 3 4 MILES I l I I I I I ' I I I I o 1 2 3 4 KILOMETRES FIGURE 64.——Generalized slope map of part of northeastern Contra Costa County, Calif. Modified from US. Geol. Survey (1972) by T. H. Nilsen and R. H. Wright STUDIES FOR SEISMIC ZONATION A85 such as rainfall and man’s construction activities. The number of landslides generated by the 1906 earthquake suggests that in the event of another such earthquake thousands of additional landslides would be triggered, costing possibly billions of dollars and untold loss of life because of extensive development of the hillside areas. Evaluation of the landslide problem on a regional scale requires more generalized data and permits less specific conclusions than evaluation of the problem at specific sites. At present it is not possible to predict particular slope failures on a regional scale from spec- ified levels of earthquake—induced ground shaking. Nevertheless, sufficient data exist in the San Francisco Bay region on a regional scale to provide a basis for EXPLANATION Area underlain by unstable bedrock units Area underlain by stable bedrock units or unconsolidated sediments at low angles of slope fig‘kfi’fi fa‘kin SCALE 1:125000 2 l 0 L l I I I I I ' I I I I 0 1 2 3 4 KILOMETRES 3 4 MILES FIGURE 65.—-Distribution of geologic units susceptible to landsliding in part of northeastern Contra Costa County, Calif. A86 substantially reducing the hazards associated with landsliding, both that occurring on a continuing basis and that triggered by earthquakes. The relative-slope-stability map compiled at a scale of 1:125,000 of the San Francisco Bay region delineates areas according to five categories of estimated relative slope stability (Nilsen and Wright, unpub. data; fig. 66). REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION The map indicates the stability of areas on the basis of landslide distribution, slope, and bedrock geology. The least stable areas are located mainly on the steep slopes of the ridges roughly paralleling the major San Andreas and Hayward faults. The map is preliminary and repre- sents an initial attempt at analyzing the landslide prob- lem of the entire region on a scale useful for land-use "(I N 4Q ”hip," '1 a , K.\/ ) \\C\,\ \r: \p 1/0 SCALE 1:125 000 0 1 2 3 I I I I | I . ' l | I I 0 3 4 KILOMETRES xx'dd‘vEVLNT I 4 MILES EXPLANATION LI Category 1 Generally highly stable. Areas of 0- to 5-per- cent slope, generally not underlain by land- slide deposits § Category 1A Subject to liquefaction. Areas of 0- to 5-percent slope that are underlain by soft, unconsoli- dated bay mud that may be subject to lique- faction and lateral spreading (see Youd and others, this report) Category 2 Generally stable. Areas of 5- to 15-percent slope, generally not underlain by landslide deposits Category 3 Marginally stable or generally stable to locally unstable. Areas of greater than 15-percent slope, generally not underlain by landslide deposits or bedrock units susceptible to land— sliding Category 4 Potentially unstable areas. Areas of greater than 15-percent slope that are underlain by bedrock units highly susceptible to land- sliding Category 5 Generally unstable areas. Areas underlain by or immediately adjacent to landslide deposits and ranging in slope from 0 to 90° Small landslide deposit Generally less than 150 m (500 ft) but greater than 60 to 90 m (200 to 300fl) in longest dimension; only shown where solitary or isolated and more than 300 m (1 000/1) from nearest adjacent landslide deposit FIGURE 66,—Relative slope stability of part of northeastern Contra Costa County, Calif. STUDIES FOR SEISMIC ZONATION A87 planning. In addition to those regional factors incorpo— rated into the map, numerous other factors are known to influence slope stability, This preliminary map provides a basis for analyzing the importance of these factors and for developing future maps which incorporate improved data and techniques. The present map, even with its limitations, is a first step toward delineating .hose gen- eral areas most susceptible to landsliding during future earthquakes and, as such, provides an essential tool for seismic zonation of the San Francisco Bay region. PREDICTED GEOLOGIC EFFECTS OF A POSTULATED EARTHQUAKE By R. D. BORCHERDT, E. E. BRABB, W. B. JOYNER, E. J. HELLEY, K. R. LAJOIE, R. A. PAGE, R. L. WESSON, and T. L. YOUD INTRODUCTION The analyses presented in the preceding six papers show that the geologic setting of the San Francisco Bay region has a dominant influence on potential earth- quake hazards. The strong correlation between geologic conditions and the amount of earthquake damage in 1906 emphasizes the importance of this influence and demonstrates the need for seismic zonation. Seismic zonation can provide the logical basis for preparation of special—purpose land-use maps that, with appropriate public policy, would be a significant step toward reduc- ing the currently expected catastrophic effects of another great earthquake (Algermissen, 1972). Seismic zonation requires a set of integrated mul- tidisciplinary predictions about the geologic effects of potential earthquakes. To illustrate a strategy for mak- ing such predictions, a demonstration profile has been chosen perpendicular to a segment of the San Andreas fault, along which a magnitude 6.5 earthquake has been postulated. This profile includes a wide variety of geologic conditions and provides a means for application of the analyses presented in the preceding six papers. Earthquake hazards to life and property originate from (1) surface faulting, (2) ground shaking, (3) flood- ing, (4) liquefaction, and (5) landsliding. The extent to which each of these geologic effects can be predicted for an earthquake of a given magnitude and location de- pends on the current state of the art. To illustrate techniques and data currently available, each effect is considered for the postulated earthquake. This paper does not provide final estimates for the total earthquake hazard along the demonstration profile. Such an objective would require analysis of in- dividual manmade structures as well as consideration of earthquakes of other sizes and locations. Instead, this paper illustrates the extent to which the above effects can be predicted for an earthquake of this given size and location. Such an analysis demonstrates a methodology for seismic zonation based on available data. A POSTULATED EARTHQUAKE For illustrative purposes, an earthquake of mag- nitude 6.5 is assumed to occur on the San Andreas fault. A88 The location and estimated length of fault rupture are shown in figure 67. Previous large earthquakes as- sociated with rupture of the ground surface along this section of the fault occurred in 1838 (magnitude >65) and in 1906 (magnitude 8.3). The moderate 1957 earth- quake (magnitude 5.3) occurred approximately 25 km (15 mi) north of the postulated surface rupture, but with no associated surface faulting on land. The assumed 6.5 magnitude is moderate in compari- son with the 8.3 magnitude of the 1906 earthquake. However, the damage resulting from the recent San Fernando and Managua earthquakes suggests that damage from a moderate earthquake can be very in- tense but of smaller areal extent that that of a great earthquake. GEOLOGY ALONG DEMONSTRATION PROFILE The demonstration profile extends northeasterly from the community of Sky Londa, across the San An- dreas fault zone, through the city of Menlo Park, and across San Francisco Bay to the southern tip of Coyote Hills (fig. 67). A brief summary of the geology along the profile will aid in understanding the potential geologic effects of the postulated earthquake. The demonstration profile includes seven geologic units that are grouped into five units on the basis of physical properties. In order of increasing age, these un— its are as follows: 1. Bay mud; mostly recently deposited soft clay, silt, and minor sand; contains more than 50 weight percent water; 2. Holocene alluvium; poOrly consolidated clayey silt, sand, and gravel; contains less than 40 weight per- cent water; 3. Late Pleistocene alluvium; primarily same material composition as Holocene alluvium, but contains less water and is more consolidated; in some places overconsolidated (soil-engineering sense); 4. Pliocene and early Pleistocene deposits; primarily continental Santa Clara and marine Merced For- mations consisting of semiconsolidated and con- solidated sandstOne, siltstone, and mudstone; and 5. Pre—Tertiary and Tertiary bedrock; includes Fran- ciscan Formation, consisting mostly of sandstone STUDIES FOR SEISMIC ZONATION A89 122°45’ 30’ 15’ 122°00’ 37°45’ —- PACIFIC OCEAN 30! .—. EXPLANATION Estimated surface rupture for postulated earthquake (magnitude 6.5) 15' — \\ Fault trace Demonstration profile A Site for which ground shaking was calculated for postulated earthquake (magnitude 6.5) 0 5 10 MILES 0 5 10 KILOMETRES 37000! _ l I FIGURE 67.—Location of demonstration profile and estimated length of surface rupture associated with a postulated earthquake of magnitude ~65 on the San Andreas fault, southwestern San Francisco Bay region. A90 and shale with lesser amounts of radiolarian chert, greenstone, limestone, and serpentine; marine sandstone and shale of Eocene, Miocene, and Pliocene age; and Page Mill Basalt, consisting of lava flows and pyroclastic rocks of Miocene age. The general stratigraphic relations of the five geologic units are illustrated in figure 68 (more detailed descrip- tions are given in Lajoie and Helley, this report). PREDICTED GEOLOGIC EFFECTS The potential effects of the postulated earthquake are dependent on the distribution of the five geologic units with respect to the San Andreas fault. A generalized prediction for each geologic effect along the profile is presented in figure 68 and discussed in the following sections. SURFACE FAULTING Historically, the larger earthquakes of the San Fran- cisco Bay region have been associated with surface rup- tures localized along the main surface traces of strike— slip faults. The resulting displacements of the ground surface along the faults have been mainly horizontal, with only minor vertical displacements. They are typi- cally expressed as an en echelon pattern of ground frac- tures that trend obliquely to the overall fault trace. The en echelon fractures have exhibited displacements ranging from a few centimetres (a few inches) to a few metres (several feet) and have defined a surface fault zone ranging in width from a few metres (several feet) to several tens of metres (about 200 ft). On the basis of past observation, the postulated mag- nitude 6.5 earthquake probably would be associated with right-lateral surface displacement along the San Andreas fault that may be as great as 1 m (3 ft) (fig. 67). The length of estimated surface rupture is 40 km (25 mi) plus or minus about 10 km (6 mi). This displacement is likely to be predominantly horizontal with the land west of the fault shifting toward the northwest relative to the land east of the fault. The main zone of surface rupture will range in width from a few metres (several feet) to several tens of metres (about 200 ft), but small fractures and permanent ground distortion may extend to much greater distances. Locally, branch and sub- sidiary faults, such as the Black Mountain fault, the Cupertino fault, and the Canada fault, may also move, but movements on such lesser faults are much more difficult to predict. If sympathetic surface movements do occur along these lesser faults, they are expected to be less than those on the main fault rupture. During the 1906 earthquake, horizontal displace- ments as large as 2.6 In (8.5 ft) were observed along this segment of the San Andreas fault. Branner (1908) re- ported for the 1906 earthquake possible evidence for STUDIES FOR SEISMIC ZONATION slight sympathetic movement on the Black Mountain fault and prevalent ground cracking extending several tens of metres (a few hundred feet) from the main trace of the San Andreas fault. Most of the ground cracking was associated with small local ground failures. The maximum reported fault displacement on the San An- dreas of 6.4 m (21 ft) occurred farther north in Marin County. GROUND SHAKING The characteristics of the ground shaking expected from the postulated earthquake depend on many fac- tors. They depend on (1) characteristics of the earth- quake source (for example, type of offset, magnitude, location, stress drop, and size of associated rupture sur- face), (2) distance from associated rupture surface, and (3) characteristics of the local geologic materials. Exist— ing data on these factors are presented in the first four pages of this report. To illustrate a strategy for predict- ing ground shaking on the basis of these data, four sites along the demonstration profile (fig. 68) were selected to illustrate responses of the several geologic units to ground shaking. The first step in predicting ground shaking for the postulated earthquake is to estimate the bedrock shak- ing at each site. The amplitude spectra for bedrock shaking (fig. 69) were approximated by using a technique proposed by Newmark and Hall (1969). (For discussion of technique see Page and others, this re- port.) The acceleration, velocity, and displacement val— ues used in constructing the spectra (see table 6) were taken from a preliminary set of data (see Page and others, this report) and probably are more representa- tive of a magnitude 6.0 earthquake than one of mag- nitude 6.5. These values are included to illustrate techniques, and it was not considered necessary to re- peat the calculations for the larger values more appro— priate for a magnitude 6.5 earthquake. The second step is to estimate the response to shaking of the unconsolidated deposits at each site. By using the techniques discussed in Borcherdt, Joyner, and others (this report) these responses were estimated from the Pacoima Dam accelerogram (San Fernando earth- quake, Feb. 9, 1971), scaled to the peak velocity values used in construction of the bedrock spectra. The final amplitude spectra for ground shaking at the surface were obtained at each site (fig. 69) by multiplying the amplitude response spectra of surficial deposits, com- puted at the appropriate high-strain levels, by the ap- proximated bedrock spectra. The estimates of ground shaking for the postulated earthquake (fig. 69) are tentative; however, they do suggest some general conclusions for the sites consid- ered: 1. The bay mud and alluvium deposits substantially 1 9 A STUDIES FOR SEISMIC ZONATION 03380 .3239: £m£oom 953395 .8 09$ 9: no 38%“. ~395an 335$ axmzastao 350%th ma... me 323% 9; 638 3:2 83m €2.50 2.: ma woafifiaoa 235 05 3 vwafifiwdv 93 955.695 mo 32535 %n Ewflumpzmsv @9839: mm Sofia wxmaafiumo n 5mg; 3383mm 3% «ESQ nofimbwcofiwv mo .8382 ha nu .ma @me :sfi mmmuccaw :mm may no 36R wmmHm—Eodx «N 0N ma OH m mud—2 ma OH m 00 lEIEL-I NI ‘NOILVAEIE co m m l L g mZON .535. w18 was ongozm B SHHLBW NI ‘NOIiVAE'IB 533:? 208365 Sad E 55.55% wnmoofiem c2: mam I ZOPwm 29.: $9732 =mEm 26“. 23.: $9323 =mEm _Sm>om oz Efimozj I mzcfion 26 lllll meI Emmhmw IcImw In PInwoxm .35 on o {{{{ wmwmwmwmn Cflxw E95 5.25% :36 80:5 32:23 :91 _ I. A .u n u x _ u v _ y 2 20.5550: , I 2838 5:355 Eran. gamma 62 02500: IIIIII NI mAV mflocwzcoc .2 cosmoEEEm UmwmmbE NI mAV $6538: L8 5385322 I Am\ .E mHIm ”3 E0 OMIOC wcimcm x0068 8825:. B mucflwfi 5:5 wEmmEomU GEE omlom ”mite mNTmb mcimzm x8fi$ Mcobm OZ§_ ma OH m o # 2:: 3980 2 $9M eumw§axk :am 133d NI ‘NOIlVAC-I‘IEI SEHLEIIN NI ‘NOIlVAI-HEI EWMVMH' | l mZON ._.._D02m30mmu >02u30mmm >02m30mmm xoofiwm xoohumm xoohuwm 8 $968 cow E:_>:__m \ :58 ham GNOSEIS HEIcI SEIHONI § NI IEIGHLI‘IcIWV {II-H8003 E35: E2>==m . __< \ 3:: ham 8 m CINOOEIS 83d SEIEILEIWILNEIO NI 'HGnll'IdWV HEIIHFIOfl STUDIES FOR SEISMIC ZONATION TABLE 6.——Parameters for estimating amplitude spectra for shaking of bedrock ‘ A. Time history parameters . Distance Acceleration Velocity Displacement Site . . . (km) (m1) (g) (cm/s) (in./s) (cm) (in) 1 ....... 9 5.6 0.40 31 12.2 14 5.5 2 ________ 14 8.7 0.22 19 7.5 9 3,5 3, 4 ______ 21 13.1 0.12 12 4.7 6 2.4 B. Estimated spectral parameters . Velocity Displacement Site Acceleration (g) (factor 3.0) (factor 1.7) (factor 4'8) (cm/s) (in./s) (cm) (in.) 1 __________ 1.92 93 36.6 24 9,4 2 ,,,,,,,,,,,, 1.06 57 22.4 15 5.9 3, 4 ,,,,,,,,,, 0.58 36 14.2 10 3.9 amplify particular frequencies of bedrock shaking and attenuate others; 2. The amplification effect of the unconsolidated de- posits is strongly dependent on frequency, which is partly due to input-strain level (compare the re— sponses for the two sites on bay mud that have identical low-strain seismic models); 3. The amount of amplification near the fundamental frequency of the site in some places actually in— creases with increasing strain level; and 4. The amplification effect of the unconsolidated de- posits in certain frequency bands is greater than the attenuation caused by increasing distance. In general, strong shaking (50—125 cm/s (20—50 in./s)) could be expected from the postulated earthquake for all surface bedrock sites along the profile west of the bay plain (fig. 68). The model calculations suggest that a substantial amplification of bedrock shaking in the fre- quency range below 1.5 hertz could be expected for all parts of the demonstration profile underlain by alluvial deposits, with increased amplifications for the parts un- derlain by bay mud. The predicted amplifications are large enough to suggest that ground shaking for fre- quencies below 1.5 hertz may be stronger at the sites underlain by bay mud and alluvium than at sites under- lain by bedropk much closer to the fault. Manmade structures with natural periods coinciding with those of the underlying unconsolidated geologic deposits are particularly susceptible to damage. FLOODING Areas of potential flooding have not been delineated in the San Francisco Bay region, except for the areas likely to be inundated by a tsunami-generated runup of 6.1 m (20 ft) (Ritter and Dupre’, 1972). Analysis of the problem is not presented in the first six papers in this report; however, for completeness in considering those A93 earthquake effects influenced by geologic conditions, a brief description of the problem is presented for the demonstration profile. For the postulated earthquake, the most probable cause of inundation by water is the failure of dams or dikes. Flood water from such failures could originate from either San Francisco Bay or upland reservoirs. Southern San Francisco Bay is surrounded by a large number of dikes constructed mostly of fine-grained sed- iments dredged from the bottom of the bay. The con- tinuous repair work necessary on many of the dikes suggests their vulnerability to failure, and their loca- tion on bay mud increases their vulnerability. Bay mud has both a high potential for amplifying particular fre- quencies of ground shaking (Borcherdt, Joyner, and others, this report) and a high potential for ground fail- ure due to liquefaction (Youd and others, this report). The area likely to be inundated because of possible dike failures depends on the tidal level at the time of the earthquake. The tidal range at the Dumbarton Bridge extends as much as 3.4 m (11 ft) above mean lower low water. For the postulated earthquake, partial inunda- tion appears likely up to the 1850 shoreline (Nichols and Wright, 1971) and even farther in certain areas where the ground has subsided in recent years owing to the withdrawal of ground water. At the southern tip of the bay, for example, flooding could extend as much as 2 km (1.2 mi) beyond the 1850 shoreline. Possible upland sources of flooding during the post— ulated earthquake are the Upper and Lower Crystal Springs Reservoirs and San Andreas Lake. Each of these bodies of water is in the San Andreas rift valley, which is drained by San Mateo Creek about 20 km (12 mi) north of the demonstration profile. Without a de- tailed engineering analysis of the associated dams, which is beyond the scope of this study, it is not possible to assess the resistance of the dams to the postulated earthquake. However, the areas of possible inundation can be outlined on a topographic map. None of the areas of possible inundation intersect the demonstration profile, even though they are of considerable size. For the postulated moderate earthquake, the likeli- hood of a large vertical offset of the sea floor or large submarine landslide necessary to generate a tsunami seems remote. Ritter and Dupré (1972) reported that 19 tsunamis were recorded by the tide gage at the Golden Gate during the past 100 years. The maximum recorded wave height from a tsunami was only 2.3 m (7.5 ft). Tidal records within San Francisco Bay show that waves attenuate rapidly to less than 50 percent of their original height by the time they reach the area of the demonstration profile. For flooding to result from the unlikely possibility of a tsunami in San Francisco Bay, the corresponding earthquake would have to occur dur- \\ A94 ing high tide when the tops of some of the dikes are less than 1 m (3 ft) above water level. The postulated earthquake of magnitude 6.5 probably is not large enough to generate a seiche in San Francisco Bay. However, seiches could be generated in reservoirs close to the postulated surface faulting; thus, seiches might be'generated in the Upper and Lower Crystal Springs Reservoirs. The tectonic setting of the San Francisco Bay region suggests that large tectonic changes of land level and associated significant changes in sea level (such as oc— curred during the 1964 Alaska earthquake) are very unlikely to accompany the postulated earthquake. Only minor local changes in vertical elevation of about 0.3 m (1 ft) were produced by the 1906 earthquake. LIQUEFACTION By using the procedures discussed in Youd and others (this report), a cross section showing liquefaction poten- tial was constructed along the demonstration profile for the postulated earthquake (fig. 68). Sediments with the greatest potential for liquefaction are the clay-free granular layers within the bay mud unit. Holocene al- luvium has a generally moderate potential with locally high potential in some recent channel and overbank deposits. Late Pleistocene alluvium is generally dense and has a low potential for liquefaction. In addition, much of the Holocene and the late Pleistocene alluvium is normally above water table and thus has, at most, a seasonal potential for liquefaction. The most common type of ground failure expected to result from liquefaction along the profile is that of lateral-spreading landslides (Youd and others, this re- port). In brief, this type of failure consists of movement of a soil mass down a mild slope with resulting cracks, fissures, and differential settlements within and near the margins of the slide mass. Relative displacements as large as tens of metres have been observed for such ground failures. Areas of the profile with the highest potential for this type of ground failure from the post- ulated earthquake are underlain by bay mud along the western margin of San Francisco Bay (fig. 68). Evidence for lateral spreading during the 1906 earth- quake was reported at several locations. For example, lateral ground movements as large as 2 m (6 ft) occurred in virtually every arm of bay mud extending beneath the city of San Francisco. Many more such ground fail- ures may have been generated near the margins of i southern San Francisco Bay but went unreported. Much Lof this area was marshland in 1906. LANDSLIDING The postulated magnitude 6.5 earthquake could be REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION expected to generate several landslides along the dem- onstration profile. By using the techniques presented in Nilsen and Brabb (this report), a cross section showing general landslide susceptibility along the demon- stration profile has been prepared (fig. 68). These ap- proximate landslide susceptibilities are not specifically dependent on the postulated magnitude of 6.5 and would be equally applicable to any moderate earth- quake in the same general location. If the postulated earthquake were to take place dur- ing a wet season and high ground-water levels, many landslides could be expected along the profile. A few large (more than 150 m (500 ft) in maximum dimension) landslides and several small (10—150 m (30—500 ft) in maximum dimension) landslides are likely on the steep slopes between Sky Londa and the San Andreas fault. In this area, existing landslide deposits could be reacti— vated. Several small landslides could be generated in the area between the San Andreas fault and the western margin of the bay plain. Lateral-spreading landslides associated with liquefaction could be expected near the margins of San Francisco Bay. If the postulated earthquake were to occur during a dry season and low ground—water levels, the amount of landsliding is expected to be much less. Some landslid- ing still could be expected between Sky Londa and the San Andreas fault, a few small landslides probably would occur in the other hilly areas along the profile, and lateral-spreading landslides associated with liquefaction still could be expected near the margins of San Francisco Bay. The 1906 earthquake apparently generated hundreds of landslides throughout the hilly regions along the western margin of the bay plain. Numerous landslides were reported in the hilly regions between the San An- dreas fault and Sky Londa. An exceptionally large mass (0.8 km (0.5 mi) across) near Black Mountain moved in 1906. The epicenter for the 1957 earthquake (magnitude 5.3) was located approximately 25 km (15 mi) north of the location for the postulated earthquake. It generated about 15 small landslides along the steep coastal bluffs in the Daly City—San Francisco area (Bonilla, 1960a). SUMMARY Earthquake hazards to life and property originate from surface faulting, ground shaking, flooding, liquefaction, and landsliding. The extent to which the currently available data permit quantitative prediction of these effects on a regional scale, for an earthquake of a given size and location, is illustrated in figure 68. Along the demonstration profile, the predicted severity of the effects varies substantially depending on local geologic conditions. STUDIES FOR SEISMIC ZONATION In brief, and as shown in figure 68, the principal losses from the postulated earthquake could be expected to originate from the following factors: 1. Surface faulting, ground shaking, and landsliding in the distance interval 0 to 10 m (0 to 30 ft), under- lain by pre-Tertiary and Tertiary bedrock and Pliocene and early Pleistocene deposits im- mediately adjacent to the San Andreas fault; 2. Ground shaking and landsliding in the distance in- terval 10 m to 5.2 km (30 ft to 3.2 mi), underlain by pre-Tertiary and Tertiary bedrock and Pliocene and early Pleistocene deposits; 3. Ground shaking in the distance interval 5.2 to 9.3 km (3.2 to 5.8 mi), underlain by Pliocene and early Pleistocene deposits and late Pleistocene al- luvium; 4. Ground shaking and landsliding of the lateral- spreading type associated with liquefaction in the distance interval 9.3 to 12.0 km (5.8 to 7.4 mi), underlain by Holocene alluvium; A95 5. Ground shaking, flooding, and landsliding of the lateral-spreading type associated with liquefac- tion in the distance interval 12.0 to 21.0 km (7.4 to 13 mi), underlain by bay mud; 6. Landsliding in the distance interval 21.6 to 22.1 km (13.4 to 13.7 mi), underlain by pre-Tertiary and Tertiary bedrock; and 7. Ground shaking, flooding, and landsliding of the lateral-spreading type associated with liquefac- tion in the distance interval 22.1 to 22.8 km (13.7 to 14.2 mi), underlain by bay mud. The strong dependence of the predicted effects on geology, together with the availability of extensive geologic and geophysical data, suggests the feasibility of extending this analysis to the entire San Francisco Bay region for other potential earthquakes. Such a re- gional analysis would provide a preliminary seismic zonation of the region from which special-purpose land-use maps could be constructed and then used to reduce earthquake hazards. GENERAL CONCLUSIONS The great California earthquake of April 18, 1906, demonstrated large variations in earthquake vulnera- bility for various areas of the San Francisco Bayregion and provided basic data for the study of earthquakes and their effects. Extensive research facilities have been developed within the last decade to study the basic physical processes associated with earthquakes. These facilities have yielded important new geological and geophysical data, which whenincorporated with data from the 1906 earthquake, provide the basis for de- velopment of guidelines to reduce future earthquake losses. The analyses presented in this report show that the geologic setting of the San Francisco Bay region has a dominant influence on potential earthquake hazards. The geologic setting is shown to control the potential severity of the various earthquake effects from which losses of life and property originate; namely, surface faulting, ground shaking, flooding, liquefaction, and landsliding. The first step required to reduce earth- quake hazard is seismic zonation, which requires prediction of t e potential severity of these various geologic effects on a regional scale for future earth- quakes of specific size and location. An example of such prediction‘ as illustrated by a demonstration profile for a postulated earthquake (Borcherdt, Brabb, and others, this report) suggests that seismic zonation of the San Francisco Bay region is feasible using existing geological and geophysical knowledge. This example illustrates a methodology for seismic zonation at the current state of the art and demonstrates the extent to which the various effects can be predicted quantita- tively on a regional scale using existing data. Tools derived and discussed as a basis for seismic zonation are (1) a map showing active faults (fig. 3A), (2) data on attenuation of shaking in bedrock, (3) geologic data, (4) a map showing ground response, (5) a map showing liquefaction potential (fig. 50), (6) a map show- ing landslide susceptibility (fig. 66), and (7) a map show- ing areas that might be inundated by tsunamis (Ritter and Dupré, 1972). (This last map represents only part of the analyses necessary for a map of potential flooding.) Application of these tools to the problem of seismic zonation shows that predictions on a regional scale for earthquakes of specific size and location are less quan- titative than those that can be made at specific sites where additional data are available. Nevertheless, such predictions are useful for the development of regional land-use policies to reduce losses from future earth- quakes. For example, although it is not possible to predict on a regional scale all those sites that will incur landsliding during the next earthquake of specific size and location, it is possible to delineate those general areas most susceptible to landsliding and hence those areas where special additional studies may be required to evaluate the landslide hazard for specific types of structures. Until more detailed treatments of seismic zonation are avilable, it is hoped that these basic tools and analyses will be useful to a wide variety of users con- A96 cerned with reducing earthquake losses in the bay re- gion. In general, the predicted maximum-intensity (fig. 40) map serves to delineate areas of potential earth- quake problems, and the basic tools mentioned above help identify the nature and possible severity of the problems in each area. In particular and as an informal summary, the analyses presented in the various papers are useful for the following purposes: 1. Estimating the maximum size and location of future damaging earthquakes and delineating areas of potential surface faulting and the locations of po— tential sources of strong ground shaking (Wesson and others). 2. Estimating the degree of bedrock shaking at various distances from earthquake sources of various sizes (Page and others). 3. Interpreting basic geologic data pertinent to ex- trapolating results of specific site studies to a re— gional scale (Lajoie and Helley). 4. Estimating the shaking response of various geologic units at specific sites and delineating those geologic units (defined by Lajoie and Helley) for which the effects of amplified ground shaking at equidistant sites are likely to be least, inter- REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION mediate, and greatest (Borcherdt, Joyner, and others). 5. Estimating the general liquefaction potential of var- ious geologic units and delineating those geologic units (defined by Lajoie and Helley) for which the liquefaction potential of existent clay-free granu- lar layers is low, moderate, and high (Youd and others). 6. Evaluating on a regional scale the vulnerability of various areas to landsliding and delineating those areas with various degrees of landslide suscepti- bility (Nilsen and Brabb). 7. Applying the various basic tools to the problem of predicting the geologic effects of potential earth- quakes of specific size and location for purposes of seismic zonation (Borcherdt, Brabb, and others). These analyses are preliminary and represent a first attempt at the required multidisciplinary analysis of the seismic-zonation problem. There are still many problems associated with seismic zonation, and the state of the art is rapidly changing. Nevertheless, it is hoped that this effort will serve as a useful first step toward the reduction of earthquake hazards on a re- gional scale. 1906 INTENSITY SCALE FOR SAN FRANCISCO The following grades of apparent intensity were used by Wood (1908, pp. 224—225) in the city of San Francisco after the California earthquake of April 18, 1906: Grade A. Very Violent. Comprises the rending and shearing of rock masses, earth, turf, and all structures along the line of faulting; the fall of rock from mountainsides; numerous landslips of great magnitude; consistent, deep, and extended fissuring in natural earth; some structures totally destroyed. Grade B. Violent. Comprises fairly general collapse of brick and frame buildings when not un- usually strong; serious cracking of brickwork and masonry in excellent struc- tures; the formation of fissures, step faults, sharp compression anticlines, and broad, wavelike folds in paved and asphalt-coated streets, accompanied by the ragged fissur- ing of asphalt; the destruction of founda- tion walls and underpinning structures by the undulation of the ground; the breaking of sewers and water mains; the lateral dis- placement of streets; and the compression, distension, and lateral waving or dis- placement of well-ballasted streetcar tracks. Grade C. Very strong. Comprises brickwork and masonry badly cracked, with occasional collapse; some brick and masonry gables thrown down; frame buildings lurched or listed on fair or weak underpinning struc— tures, with occasional falling from under- pinning or collapse; general destruction of chimneys and of masonry, brick, or cement veneers; considerable cracking or crushing of foundation walls. Grade D. Strong. Comprises general but not universal fall of chimneys; cracks in masonry and brickwork; cracks in foundation walls, re- taining walls, and curbing; a few isolated cases of lurching or listing of frame build- ings built upon weak underpinning struc- tures. Grade E. Weak. Comprises occasional fall of chimneys and damage to plaster, partitions, plumb- ing, and the like. STUDIES FOR SEISMIC ZONATION REFERENCES CITED Albee, A. L., and S ith, J. L., 1967, Geologic criteria for nuclear power plant location: Soc. Mining Engineers Trans, v. 238, p. 430—434. Algermissen, S. T. (principal investigator), 1972, A study of the earth— quake losses in the San Francisco Bay area: Natl. Oceanic At— mospheric Admin., US. Dept. Commerce, 220 p. Allen, J. E., 1946, Geology of the San Juan Bautista quadrangle, California: California Div. Mines Bull. 133, 112 p., map scale 1:62,500. Ambraseys, N. N, 1969, Maximum intensity of ground movements caused by faulting: World Conference on Earthquake Engineer- ing, 4th, Santiago, Chile, Proc., v. 1, p. 154—171. 1973, Dynamics and response of foundation materials in epi- central regions of strong earthquakes: World Conference on Earthquake Engineering, 5th, Rome, Italy, Proc., 23 p. Anderson, F. M., 1899, The geology of the Point Reyes peninsula: California Univ. Pubs. Geol. Sci., v. 2, p. 119—153. Atomic Energy Commission, 1973, Nuclear power plants, seismic and geologic siting criteria: US. Federal Register, v. 38, p. 31279—— 31285. Bailey, R. G., 1971, Landslide hazards related to land-use planning in Teton National Forest, northwest Wyoming: US. Dept. Agricul- ture, Forest Service, Intermountain region, Ogden, Utah, 131 p. Berkland, J. 0., 1969, Geology of the Novato quadrangle, Marin County, California: San Jose State College, unpub. M. S. thesis, 146 p. Bernreuter, D. L., and Tokarz, F. J., 1972, Design basis earthquakes for the Lawrence Livermore Laboratory site: California Univ. Lawrence Liv,ermore Lab. Pub. UCRL—51193. Blanc, R. P., and Cleveland, G. B., 1968, Natural slope stability as related to geology, San Clemente area, Orange and San Diego Counties, California: California Div. Mines and Geology Spec. Rept. 98, 19 p. Blanchard, F. B.,. and Laverty, G. L., 1966, Displacements in the Claremont Water Tunnel at the intersection with the Hayward fault: Seismol. Soc. America Bull., v. 56, p. 291—293. Bolt, B. A., 1973, Duration of strong ground motion: World Conference on Earthquake Engineering, 5th, Rome, Italy, Proc., p. 1304—— 1313. Bolt, B. A., and Marion, W. C., 1966, Instrumental measurement of slippage on the Hayward fault: Seismol. Soc. America Bull., v. 56, p. 305—316. Bonilla, M. G., 1959, Geologic observations in the epicentral area of the San Francisco earthquake of March 22, 1957: California Div. Mines and Geology Spec. Rept. 57, p. 25—38. 1960a, Landslides in the San Francisco South quadrangle, California: US. Geol. Survey open—file rept., 44 p. 1960b, A sample of California Coast Range landslides, in Geological Survey research 1960: US. Geol. Survey Prof. Paper 400—B, p. B149. 1965, Geo] gic map of the San Francisco South quadrangle, California: US. Geol. Survey open-file map, scale 1:24,000. 1966, Deformation of railroad tracks by slippage on the Hay- ward fault in‘the Niles district of Fremont, California: Seismol. Soc. America Bull., v. 56, p. 281—289. 1970, Surface faulting and related effects, in R. L. Wiegel, ed., Earthquake engineering: Englewood Cliffs, N.J., Prentice-Hall, p. 47—74. 1971, Preliminary geologic map of the San Francisco South and part of the Hunters Point quadrangles: U.S. Geol. Survey Misc. Field Studies‘ Map MF—311, scale 1:24,000. 1975, Trench exposures of parts of the February 9, 1971, surface A97 faults in the San Fernando fault zone, in Geological and Geophys- ical Studies, v. 3, of San Fernando, California, earthquake of Feb. 9, 1971: Natl. Oceanic Atmospheric Admin, US. Dept. Com- merce (in press). Bonilla, M. G., and Buchanan, J. M., 1970, Interim report on world wide historic surface faulting: U.S. Geol. Survey open-file rept., 32 p. Boore, D. M., 1973, Empirical and theoretical study of near-fault wave propagation: World Conference on Earthquake Engineering, 5th, Rome, Italy, Proc., p. 2397—2408. Boore, D. M., and Zoback, M. D., 1975, Two-dimensional kinematic fault modeling of the Pacoima Dam strong-motion recordings of the February 9, 1971, San Fernando earthquake: Seismol. Soc. America Bull. (in press). Borcherdt, R. D., 1970, Effects of local geology on ground motion near San Francisco Bay: Seismol. Soc. America Bull., v. 60, p. 29—61. Borcherdt, R. D., and Gibbs, J. F., 1975, Prediction of maximum earthquake intensities for the San Francisco Bay region: U.S. Geol. Survey Misc. Field Studies Map (in press). Brabb, E. E., 1970, Preliminary geologic map of the central Santa Cruz Mountains, California: US. Geol. Survey open-file map, 3 sheets, scale 1:62,500. Brabb, E. E., and Pampeyan, E. H., 1972, Preliminary maps of land-. slides in San Mateo County, California: U. S. Geol. Survey Misc. Field Studies Map MF—344, scale 1:62,500. Brabb, E. E., and Pampeyan, E. H., and Bonilla, M. G., 1972, Land- slide susceptibility in San Mateo County, California: US. Geol. Survey Misc. Field Studies Map MF—360, scale 1:62,500. Branner, J. C., 1908, The earth movement on the fault of April 18, 1906; Crystal Springs Lake to Congress Springs: Carnegie Inst. Washington Pub. 87, 104 p. Brewer, W. H., 1930, Up and down California in 1860—64; thejournal of William H. Brewer (F. P. Farquahar, ed.): New Haven, Yale Univ. Press, 601 p. Brown, R. B., Jr., 1970a, Faults that are historically active or that shows evidence of geologically young surface displacements, San Francisco Bay region; A progress report—October 1970: US. Geol. Survey Misc. Field Studies Map MF—331, 2 sheets, scale 1:250,000. 1970b, Map showing recently active breaks along the San An- dreas and related faults between the northern Gabilan Range and Cholame Valley, California: US. Geol. Survey Misc. Geol. Inv. Map I—575, scale 1:62,500. 1972, Active faults, probable active faults, and associated frac- ture zones, San Mateo County, California: US. Geol. Survey Misc. Field Studies Map MF—355, scale 1:62,500. , Brown, R. B., Jr., and Lee, W. H. K., 1971, Active faults and prelimi- nary earthquake epicenters in the southern part of the San Fran- cisco Bay region: U.S. Geol. Survey Misc. Field Studies Map MF—307, scale 1:250,000. Brown, R. D., J r., and Vedder, J. G., 1967, Surface tectonic fractures along the San Andreas fault, in The Parkfield-Cholame, earth- quakes of June-August 1966: US. Geol. Survey Prof. Paper 579, p. 2—22. Brown, R. B., Jr., Vedder, J. G., Wallace, R. E., Roth, E. F. Yerkes, R. F., Castle, R. O. Waananen, A. 0., Page, R. W., and Eaton, J. P., 1967, The Parkfield-Cholame, California, earthquakes of J une- August 1966—Surface geologic effects, water-resources aspects, and preliminary seismic data: U.S. Geol. Survey Prof. Paper 579, 66 p. Brown, R. D., Jr., and Wallace, R. E., 1968, Current and historic fault movement along the San Andreas fault between Paicines and Camp Dix, California, in Dickinson, W. R., and Grantz, Arthur, A98 eds., Proceedings of conference on geologic problems of San An- dreas fault system: Stanford Univ. Pubs. Geol. Sci., v. 11, p. 22—41. Brown, R. B., Jr., Ward, P. L., and Platker, George, 1973, Geologic and seismologic aspects of the Managua, Nicaragua, earthquakes of December 23, 1972: US. Geol. Survey Prof. Paper 838, 34 p. Brown, R. D., J r., and Wolfe, E. W., 1972, Map showing recently active breaks along the San Andreas fault between Pt. Delgada and Bolinas Bay, California: US. Geol. Survey Misc. Geol. Inv. Map 1—692, scale 1224,000. Brune, J. N., 1970, Tectonic stress and the spectra of seismic shear waves from earthquakes: Jour. Geophys. Research, v. 75, p. 4997—5009. Burke, D. B., and Helley, E. J ., 197 3, Map showing evidence for recent fault activity in the vicinity of Antioch, Contra Costa County, California: US. Geol. Survey Misc. Field Studies Map MF—533, scale 1224,000. ‘ Burnett, J. L., 1972, Geologic and slope stability maps of the Hayward hills, California: Hayward, California Planning Dept., Hayward Hills Study Area, scale 1:24,000. California Department of Water Resources, 1966, Evaluation of ground water resources, Livermore and Sunol Valleys, Appendix A, Geology: California Dept. Water Resources Bull. 118—2, 153 p. Chae, Y. S., 1968, Viscoelastic properties of snow and sand: Am. Soc. Civil Engineers Proc., Jour. Eng. Mechanics Div., v. 94, p. 1379— 1394. Clague, J. J ., 1969, Landslides of southern Point Reyes National Seashore: California Div. Mines and Geology Mineral Inf. Ser- vice, v. 22, p. 107—110, 116—118. Clark, M. M., 1972, Surface rupture along the Coyote Creek fault, in The Borrego Mountain earthquake of April 9, 1968: US. Geol. Survey Prof. Paper 787, p. 55—86. Clark, M. M., Grantz, Arthur, and Rubin, Meyer, 1972, Holocene activity of the Coyote Creek fault as recorded in the sediments of Lake Cahuilla, in The Borrego Mountain earthquake of April 9, 1968: US. Geol. Survey Prof. Paper 787, p. 1112—1130. Cluff, L. S., and Steinbrugge, K. V., 1966, Hayward fault slippage in the Irvington-Niles district of Fremont, California: Seismol. Soc. America Bull., v. 56, p. 257—279. Cooper, Allan, 1971, Structure of the continental shelf west of San Francisco, California: California State Univ. San Jose, unpub. M.S. thesis. Cummings, J. C., 1968, The Santa Clara formation and possible post- Pliocene slip on the San Andreas fault in central California, in Dickinson, W. R., and Grantz, Arthur, eds., Proceedings of confer- ence on geologic problems of San Andreas fault system: Stanford Univ. Pubs. Geol. Sci., v. 11, p. 191—207. Dibblee, T. W., Jr., 1966, Geology of Palo Alto quadrangle, Santa Clara and San Mateo Counties, California: California Div. Mines and Geology, Map Sheet 8, scale 1:62,500. 1972a, Preliminary geologic map of the Lick Observatory quadrangle, Santa Clara County, California: US. Geol. Survey open-file map, scale 1:24,000. 1972b, Preliminary geologic map of the San Jose East quad- rangle, Santa Clara County, California: US. Geol. Survey open- file map, scale 1:24,000. 1972c, Preliminary geologic map of the Milpitas quadrangle, Alameda and Santa Clara Counties, California: US. Geol. Sur- vey open-file map, scale 1:24,000. 1972d, Preliminary geologic map of the Calaveras Reservoir, Santa Clara County, California: US. Geol. Survey open-file map, scale 1:24,000. 1973a, Preliminary geologic map of the Gilroy quadrangle, Santa Clara County, California: US. Geol. Survey open-file map, scale 1:24,000. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 1973b, Preliminary geologic map of the Gilroy Hot Springs quadrangle, Santa Clara County, California: US. Geol. Survey open-file map, scale 1224,000. 1973c, Preliminary geologic map of the Morgan Hill quadran- gle, Santa Clara County, California: US. Geol. Survey open-file map, scale 1:24,000. 1973d, Preliminary geologic map of the Mt. Madonna quad- rangle, Santa Clara and Santa Cruz Counties, California: US. Geol. Survey open-file map, scale 1:24,000. 1973e, Preliminary geologic map of the Mt. Sizer quadrangle, Santa Clara County, California: US. Geol. Survey open-file map, scale 1224,000. Dickinson, W. R., and Grantz, Arthur, 1968, Proceedings of conference on geologic problems of San Andreas fault system: Stanford Univ. Pubs. Geol. Sci., v. 11, p. 117—120. Dieterich, J. H., 197 3, A deterministic near-field source model: World Conference on Earthquake Engineering, 5th, Rome, Italy, Proc., p. 2385—2396. Dobry, R. Whittman, R. V., and Roeset, J. M., 1971, Soil properties and the one-dimensional theory of earthquake amplification: Mas-. sachusettsInst.Tech.Research Rept.R71—18, Soils Pub. 275,347 p. Dooley, R. L., 1973, Geology and land use considerations in the vicin- ity of the Green Valley fault, Solano County, California: Califor— nia Univ. Davis, unpub. M.S. thesis, 47 p. Eckel, E. B., ed., 1958, Landslides and engineering practice, in Na- tional Research Council, Highway Research Board Spec. Rept. 29: Dept. of Soils, Geology and Foundations, NAS—NRC Pub. 544, 232 p. Fox, K. F., Jr., Sims, J. D., Bartow, J. A., and Helley, E. J., 1973, Preliminary geologic map of eastern Sonoma County and western Napa County, California: US. Geol. Survey, Misc. Field Studies Map MF-483, 4 sheets, scale 1:62,500. Frame, P. A., 1974, Landslides in the Mt. Sizer area, Santa Clara County, California: California State Univ. San Jose, MS. thesis, 67 p. Gealey, W. K., 1951, Geology of the Healdsburg quadrangle, Califor- nia: California Div. Mines and Geology Bull. 161, 76 p. Gibbs, J. F., and Borcherdt, R. D., 1974, Effects of local geology on ground motion in the San Francisco Bay region—a continuing study: U.S. Geol. Survey open-file rept., 162 p. Gibbs, H. J ., and Holtz, W. G., 1957, Research on determining the density of sand by spoon penetration test: Internat. Conf. Soil Mechanics and Foundations Eng., 4th, London, Proc., v. 1, p. 35—39. Gibson, W. M., and Wollenberg, H. A., 1968, Investigations for ground stability in the vicinity of the Calaveras fault, Livermore and Amador Valleys, Alameda County, California: Geol. Soc. America Bull., v. 79, p. 627-638. Glen, William, 1959, Pliocene and lower Pleistocene of the western part of the San Francisco peninsula: California Univ. Pubs. Geol. Sci., v. 36, p. 147—198. Greene, H. G., Lee, W. H. K., McCulloch, D. S., and Brabb, E. E., 1973, Faults and earthquakes in the Monterey Bay region, California: US. Geol. Survey Misc. Field Studies Map MF—518, 4 sheets, scale 1:250,000. Griggs, G. B., 1973, Earthquake activity between Monterey and Half Moon Bay, California: California Geology, v. 26, p. 10&110. Hadley, J. B., 1964, Landslides and related phenomena accompanying the Hebgen Lake earthquake of August 17, 1959: US. Geol.’ Survey Prof. Paper 435, p. 107—138. Hall, Nelson T., Sarna-Wojcicki, Andre, and Dupré, William, 1974, Faults and their potential hazard in Santa Cruz County, Califor- nia: U.S. Geol. Survey Misc. Field Inv. MF—626, scale 1:62,500. Hanks, T. C., 1975, The faulting mechanism of the San Fernando earthquake: Jour. Geophys. Research (in press). STUDIES FOR SEISMIC ZONATION Hansen, W. R., 1964, Sewage effluent disposal and its effects on ground water beneath Livermore Valley: Assoc. Eng. Geologists Bull., v. 1, p. 22—36. Hansen, W. R., Eckel, E. B., Schaem, W. E., Lyle, R. E., George, Warren, and Chance, Genie, 1966, The Alaska earthquake, March 27, 1964: Field investigations and reconstruction effort: U.S. Geol. Survey Prof. Paper 541, 111 p. Hardin, B. 0., and Dernevich, V. D., 1972, Shear modulus and damp- ing in soils; design equations and curves: Am. Soc. Civil En- gineers Proc. Soil Mechanics and Found Div., p. 667— 692. Harding, R. C., 1969, Landslides—A continuing problem for Bay area development, in Danehy, E. A. (ed.), Urban environmental geol- ogy in San Francisco Bay region: Assoc. Eng. Geologists, San Francisco Section, Spec. Pub., p. 65—74. Haskell, N. A., 1953, The disperson of surface waves on multilayered media: Seismol. Soc. America Bull., v. 43, p. 17—34. Hazelwood, R. M., 1974, Preliminary report of seismic refraction sur- vey along the east side of San Francisco Bay, Alameda County, California: US. Geological Survey open-file rept., 10 p. Helley, E. J., and Brabb, E. E., 1971, Geologic map of late Cenozoic deposits, Santa Clara County, California: US. Geol. Survey Misc. Field Studies Map MF—335, 3 sheets, scale 1:62.500. Helley, E. J., Lajoie, K. R., and Burke, D. B., 1972, Geologic map oflate Cenozoic deposits, Alameda County, California: US. Geol. Sur- vey Misc. Field Studies Map MF—429, scale 1:62,500. Hudson, D. E., Brady, A. G., Trifunac, M. D., and Vijayaraghavan, A., 1971, Strong-motion earthquake accelerograms, corrected ac- celerograms and integrated ground velocity and displacement curves: California Inst. Tech., Earthquake Eng. Research Lab., EERL 71—51, v. 2, 321 p. Huffman, M. E., 1972a, Geology for planning on the Sonoma County coast between the Russian and Gualala Rivers: California Div. Mines and Geology Prelim. Rept. 16, 38 p., scale 1:24,000. 1972b, Geology for planning in the Sonoma Mountain and Mark-Riebli Road areas, Sonoma County, California: California Div. Mines and Geology open-file rept. 72—25, scale 1:24,000. 1974, Geology for planning on the Sonoma County coast bet- ween the Russian River and Estero Americana: California Div. Mines and Geology Prelim. Rept. 20, 36 p. Housner, G. W., 1965, Intensity of earthquake ground shaking near the causative fault: World Conference on Earthquake Engineer- ing, 3rd, Wellington, New Zealand, Proc., v. 1, p. III—94—III—109. Housner, G. W., and Jennings, P. C., 1972, The San Fernando, California earthquake: Earthquake Eng. and Structural Dynamics 1, p‘. 5—31. Hudson, D. E., 1962, Some problems in the application of spectrum techniques to strong-motion earthquake analysis: Seismol. Soc. America Bull., v. 52, p. 417—430. Ida, Y., 1973, The maximum acceleration of seismic ground motion: Seismol. Soc. America Bull., v. 63, p. 959-968. Idriss, I. M., and Seed, H. B., 1968, Seismic response of horizontal soil layers: Am. Soc. Civil Engineers Proc., Soil Mechanics and Found. Div., v. 94, p. 1003—1031. Iida, Numizi, 1965, Earthquake magnitude, earthquake fault and source dimensions: Nagoya Univ. Jour. Earth Sci., v. 13, p. 115— 132. Jack, Robert, 1968, Quaternary sediments of the Montara Mountain area, San Mateo County: California Univ., Berkeley, unpub. M.A. thesis. Jahns, R. H., and Hamilton, D. H., 1971, Geology of the Mendocino Power Plant, in Mendocino Power Plant Preliminary Safety Analysis Report: US. Atomic Energy Comm., v. 1, appendix 2.5A, 61 p. Johnson, A. M., and Ellen, S. D., 1965, Preliminary evaluation of the interaction between engineering development and natural A99 geologic processes on the Bovet property, town of Portola Valley, California—with a section on the San Andreas fault by W. R. Dickinson: Report prepared for City of Portola Valley, California, scale 122,400. Johnson, A. M., and Lobo-Guerrero, A. U., 1968, Preliminary evalua- tion of the relative stability of ground, Marianni property, town of Portola Valley, California—with a section on the San Andreas fault by W. R. Dickinson: Report prepared for the City of Portola Valley, California, 22 p. Kanai, K., 1952, Relation between the nature of surface layer and the amplitudes of earthquake motions: Tokyo Univ. Earthquake Re- search Inst. Bull., v. 30, p. 31—37. Kojan, E., Foggin, G. T., III, and Rice, R. M., 1972, Prediction and analysis of debris slide incidence by photogrammetry, Santa Ynez—San Rafael Mountains, California: Internatl. Geol. Con- gress, 24th, sec. 13, p. 124—131. Lajoie, K. R., Helley, E. J., Nichols, D. R., and Burke, D. B., 1974, Unconsolidated and semi-unconsolidated sedimentary deposits, San Mateo County, California: US. Geological Survey Misc. Field Studies Map MF—575, scale 1162,500. Lawson, A. C., 1893, The post-Pliocene diastrophism of the coast of southern California: California Univ. Pubs. Geol. Sci., v. 1, p. 115—160. 1895, Sketch of the geology of the San Francisco peninsula, Extract from the Fifteenth Annual Report of the US. Geol. Sur- vey, 1893—94, Washington, Govt. Printing Office, p. 401—476. Lawson, A. C., and others, 1908, The California earthquake of April 18, 1906—Report of the State Earthquake Investigation Com- mission: Carnegie Inst. Washington Pub. 87, 2 vols. Lee, W. H. K., Eaton, M. S., and Brabb, E. E., 1971, The earthquake sequence near Danville, California, 1970: Seismol. Soc. America Bull., v. 61, p. 1771-1794. Lee, W. H. K., Meagher, K. L., Bennett, R. E., and Matamoros, E. E., 1972, Catalog of earthquakes along the San Andreas fault system in central California for the year 1971: US. Geol. Survey open-file rept., 67 p. Lee, W. H. K., Roller, J. C., Bauer, P. G., and Johnson, J. D., 1972, Catalog of earthquakes along the San Andreas fault system in central California for the year 1969: US. Geol. Survey open-file rept., 48 p. Lee, W. H. K., Roller, J. C., Meagher, K. L., and Bennett, R. E., 1972, Catalog of earthquakes along the San Andreas fault system in central California for the year 1970: US. Geol. Survey open-file rept., 73 p. Lindh, A. G., and Boore, D. M., 1973, Another look at the Parkfleld earthquake using strong-motion instruments as a seismic array (abs): Seismol. Soc. America, Earthquake Notes 44, p. 24. Louderback, G. D., 1942, Faults and earthquakes: Seismol. Soc. America Bull., v. 32, p. 305—330. 1947, Central California earthquakes of the 1830’s: Seismol. Soc. America Bull., v. 37, p. 33—74. Maberry, J. 0., 19723, Map showing landslide deposits and areas of potential landsliding in the Parker quadrangle, Arapahoe and Douglas Counties, Colorado: US. Geol. Survey Misc. Geol. Inv. Map I—770- E, scale 1: 24 0.00 1972b, Slope map of the Parker quadrangle, Arapahoe and Douglas Counties, Colorado. U. S. Geol. Survey Misc. Geol. Inv. Map I—770—F, scale 1:24,000. McEvilly, T. V., 1966, The earthquake sequence of November 1964 near Corralitos, California: Seismol. Soc. America Bull., v. 56, p. 755—7 73. 1970, Magnitudes, epicenters, and fault plane solutions, in The Santa Rosa earthquakes of October 1969: Mineral Inf. Service, v. 23, p. 44—48. McEvilly, T. V., and Casaday, K. B., 1967, The earthquake sequence of A100 September 1965 near Antioch, California: Seismol. Soc. America Bull., V. 57, p. 113—124. McLaughlin, R. J ., 1971, Geologic map of the Sargent fault zone in the vicinity of Mount Madonna, Santa Clara County, California: US. Geol. Survey open-file map, scale 1:12,000. 1973, Geology of the Sargent fault zone in the vicinity of Mount Madonna, Santa Clara and Santa Cruz Counties, California: California State Univ. San Jose, unpub. M. S. thesis, 131 p. McLaughlin, R. J., Simoni, T. R., Osbun, E. D., and Bauer, P. G., 1971, Preliminary geologic map of the Loma Prieta—Mount Madonna area, Santa Clara and Santa Cruz Counties, California: US. Geol. Survey open—file map, scale 1224,000. Medvedev, S. V., 1965, Engineering seismology: National Technical Inf. Service, NTIS No. TT65—50011, 260 p. Milliman, J. D., and Emery, K. 0., 1968, Sea levels during the past 35,000 years: Science, v. 162, p. 1121. Morton, D. M., 1971, Seismically triggered landslides in the area above the San Fernando Valley, in The San Fernando, California, earthquake of February 9, 1971: US. Geol. Survey Prof. Paper 733, p. 99—104. Murphy, L. M., and Cloud, W. K., 1957, United States earthquakes, 1955: US. Coast and Geodetic Survey, 83 p. Nason, R. D., 1971, Measurements and theory of fault creep slippage in central California: Royal Soc. New Zealand Trans, v. 9, p. 181—187. Newmark, N. M., and Hall, W. J., 1969, Seismic design criteria for nuclear reactor facilities: World Conference Earthquake En- gineering, 4th, Santiago, Chile, Proc., v. 2, p. 37—50. Nichols, D. R., and Wright, N. A., 1971, Preliminary map of historic margins of marshlands, San Francisco Bay, California: US. Geol. Survey open-file map, scale 1:125,000. Nilsen, T. H., 1971, Preliminary photointerpretation map of landslide and other surficial deposits of the Mount Diablo area, Contra Costa and Alameda Counties, California: US. Geol. Survey Misc. Field Studies Map MF—310, scale 1:62,500. 1972a, Preliminary photointerpretation map of landslide and other surficial deposits of parts of the Altamont and Carbona 15-minute quadrangles, Alameda County, California: US. Geol. Survey Misc. Field Studies Map MF—321, scale 1262,500. 1972b, Preliminary photointerpretation map of landslide and other surflcial deposits of the Byron area, Contra Costa and Alameda Counties, California: US. Geol. Survey Misc. Field Studies Map MF—338, scale 1262,500. 1972c, Preliminary photointerpretation map of landslide and other surficial deposits of the Mount Hamilton quadrangle and parts of the Mount Boardman and San Jose quadrangles, Alameda and Santa Clara Counties, California: US. Geol. Sur— vey Misc. Field Studies Map MF—339, scale 1262,500. 1972d, Preliminary photointerpretation map of landslides and other surficial deposits of parts of the Los Gatos, Morgan Hill, Gilroy Hot Springs, Pacheco Pass, Quien Sabe, and Hollister 15-minute quadrangles, Santa Clara County, California: US. Geol. Survey Misc. Field Studies Map MF—416, scale 1:62,500. 1973a, Current slope stability studies by the US. Geological Survey in the San Francisco Bay region: Landslide, The Slope Stability Review, v. 1, no. 1, p. 2—10. 1973b, Preliminary photointerpretation map of landslide and other surficial deposits of the Livermore and part of the Hayward 15-minute quadrangles, Alameda and Contra Costa Counties, California: US. Geol. Survey Misc. Field Studies Map MF—519, scale 1:62,500. 1973c, Preliminary photointerpretation map of landslide and other surficial deposits of the Concord 15-minute quadrangle and the Oakland West, Richmond, and part of the San Quentin 7.5- minute quadrangles, Contra Costa and Alameda Counties, REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION California: US. Geol. Survey Misc. Field Studies Map MF—493, scale 1:62,500. Nilsen, T. H., and Brabb, E. E., 1972, Preliminary photointerpretation and damage maps of landslide and other surficial deposits in northeastern San Jose, California: US. Geol. Survey Misc. Field Studies Map MF—361, scales 1212,000 and 1:24,000. Nilsen, T. H., Taylor, F. A., and Brabb, E. E., 1975, Structurally damaging landslides in Alameda County, California, from 1940 to 1971—economic analysis and influence of ancient landslide deposits: U.S. Geol. Survey Bull. 1398 (in press). Nilsen, T. H., and Turner, B. L., 1974, The influence of rainfall and ancient landslide deposits on recent landslides (1950—1971) in urban areas of Contra Costa County, California: US. Geol. Sur— vey Bull. 1388. Oakeshott, G. B., ed., 1959, San Francisco earthquake of March 1957: California Div. Mines Spec. Rept., V. 57, p. 73—123. Page, R. A., Boore, D. M., Joyner, W. B., and Coulter, H. W., 1972, Ground motion values for use in the seismic design of the trans- Alaska pipeline system: US. Geol. Survey Circ. 672, 23 p. Pampeyan, E. H., 1970, Geologic map of the Palo Alto 7 .5-minute quadrangle, San Mateo and Santa Clara Counties, California: US. Geol. Survey open-file rept., scale 1224,000. Peacock, W. H., and Seed, H. B., 1968, Sand liquefaction under cyclic loading simple shear conditions: Am. Soc. Civil Engineers Proc., Jour. Soil Mechanics and Found. Div., v. 94, no. SM3, p. 689—708. Plafker, George, 1969, Tectonics of the March 27, 1964 Alaska earthquake: U.S. Geol. Survey Prof. Paper 543.1, _74 p. Plafker, G., Erickson, G. E., and Concha, F. J., 1971, Geological aspects of the May 31, 1970, Peru earthquake: Seismol. Soc. America Bull., v. 61, p. 543—578. Radbruch, D. H., 1957, Areal and engineering geology of the Oakland West quadrangle, California: US. Geol. Survey Misc. Geol. Inv. Map I—239, scale 1:24,000. 1967, Approximate location of fault traces and historic surface ruptures within the Hayward fault zone between San Pablo and Warm Springs, California: US. Geol. Survey Misc. Geol. Inv. Map I—522, scale 1262,500. 1968a, Map showing recently active breaks along the Hayward fault zone and southern part of the Calaveras fault zone, Califor- nia: U.S. Geol. Survey open-file map, 2 sheets, scale 1224,000. 1968b, New evidence of historic fault activity in Alameda, Contra Costa, and Santa Clara Counties, California, in Dickin- son, W. R., and Grantz, Arthur, eds., Proceedings of conference on geologic problems of San Andreas fault system: Stanford Univ. Pubs. Geol. Sci., v. 11, p. 46—54. 1969, Areal and engineering geology of the Oakland East quad- rangle, California: US. Geol. Survey Geol. Quad. Map GQ—769, scale 1:24,000. Radbruch, D. H., Bonilla, M. G., and others, 1966, Tectonic creep in the Hayward fault zone, California: US. Geol. Survey Circ. 525, 13 p. Radbruch, D. H., and Case, J. E., 1967, Preliminary geologic map and engineering geologic information, Oakland and vicinity, Califor- nia: U.S. Geol. Survey open-file map, scale 124,000. Radbruch, D. H., and Lennert, B. J ., 1966, Damage to culvert under Memorial Stadium University of California, Berkeley, caused by slippage in the Hayward fault zone: Seismol. Soc. America Bull., v. 56, p. 295—305. Radbruch, D. H., and Weiler, L. M., 1963, Preliminary report on landslides in a part of the Orinda Formation, Contra Costa Coun- ty, California: US. Geol. Survey open-file rept., 49 p. Radbruch, D. H., and Wentworth, C. M., 1971, Estimated relative abundance of landslides in the San Francisco Bay region, Califor- nia: U.S. Geol. Survey open-file rept., scale 1:500,000. Rice, S. J ., and Strand, R. G., 1972, Geologic and slope stability maps of the Tennessee Valley, Lucas Valley and north coastal areas, STUDIES FOR SEISMIC ZONATION Marin County, California: California Div. Mines and Geology open-file rept. 72—22, scale 1:12,000. Richter, C. F., 1958, Elementary seismology: San Francisco, W. H. Freeman, 768 p. ’ Reiche, Parry, 1950, Rio Vista, California, fault scarp (abs): Geol. Soc. America Bull., V". 61, p. 1529—1530. Ritter, J. R., and Dupré, W. R., 1972, Map showing areas of potential inundation by tsunamis in the San Francisco Bay Region, California: U.S. Geol. Survey Misc. Field Inv. Map MF—480, scale 1:125,000. Rogers, T. H., 1971, Environmental geologic analysis of the Santa Cruz Mountain1 study area, Santa Clara County, California: California Div. Mines and Geology, San Jose, Calif, Santa Clara County Planning Dept., 64 p. 1972, Bear Valley—Melendy Ranch earthquakes of 24—27 Feb- ruary 1972: California Geology, v. 25, p. 138. Rogers, T. H., and Nason, R. D., 1971, Active displacement on the Calaveras fault zone at Hollister, California: Seismol. Soc. America Bull., v. 61, p. 399—416. Saul, R. B., 1972, Geology and slope stability of the southwest quarter of the Walnut Creek quadrangle, Contra Costa County, Califor— nia: California Div. Mines and Geology Map Sheet 16, scale 1:12,000. Savage, J. C., and Burford, R. 0., 1973, Geodetic determination of relative plate motion in central California: Jour. Geophy. Re- search, v. 78, p. 832—845. Schlocker, J., Bonilla, M. G., and Radbruch, D. H., 1958, Geology of the San Francisco north quadrangle, California: U.S. Geol. Sur- vey Misc. Geol. Inv. Map I—272, scale 1:24,000. Schnabel, P. B., and Seed, H. B., 1973, Accelerations in rock for earthquakes in the western United States: Seismol. Soc. America Bull., v. 63, p. 501—516. Schuyler, J. D., 1898, Reservoirs for irrigation, in Eighteenth Annual Report of the U.S. Geol. Survey for 1896—1897, Part IV: Washington, D. C., p. 711—713. Scott, G. R., 1972, Map showing landslides and areas susceptible to landsliding in the Morrison quadrangle, Jefferson County, Col- orado: U.S. Geol. Survey Misc. Geol. Inv. Map I—790—B, scale 1:24,000. “ Seed, H. B., 1968, Landslides during earthquakes due to soil liquefac- tion: Am. Soc. Civil Engineers Proc., Jour. Soil Mechanics and Found. Div., v. 94, p. 1053—1122. Seed, H. B., and IdI‘lSS, I. M., 1971, Simplified procedure for evaluating soil liquefaction potential: Am. Soc. Civil Engineers Proc., Jour. Soil Mechanics and Found. Div., v. 97, p. 1249—1273. Seed, H. B., Idriss, I. M., and Kiefer, F. W., 1969, Characteristics of rock motions during earthquakes: Am. Soc. Civil Engineers Proc., Jour. Soil Mechanics and Found. Div., v. 95, p. 1199—1218. Seed, H. B., and Lee, K. L., 1969, Pore-water pressures in earth slopes under seismic loading conditions: World Conf. on Earthquake Engineering, 4th, Santiago, Chile, Proc. V. 3, p. A—5—1 to A—5—11. Seed, H. B., and Peacock, W. H., 1971, Test procedures for measuring soil liquefaction characteristics: Am. Soc. Civil Engineers Proc., Jour. Soil Mechanics and Found. Div., v. 97, p. 1099—1119. Sharp, R. V., 1973,.Map showing recent tectonic movement on the Concord fault, Contra Costa and Solano Counties, California: U.S. Geol. Survey Misc. Field Studies Map MF—505, scale 1:24,000. Sharp, R. V., and Clark, M. M., 1972, Geologic evidence of previous faulting near the 1968 rupture on the Coyote Creek fault, in The Borrego Mountain earthquake of April 9, 1968: U.S. Geol. Survey Prof. Paper 787, p. 131—140. Simpson, H. E., 1973a, Map showing landslides in the Golden quad- rangle, Jefferson County, Colorado: U.S. Geol. Survey Misc. Geol. Inv. Map I—761—B, scale 1:24,000. 1973b, Map showing areas of potential rockfalls in the Golden A101 quadrangle, Jefferson County, Colorado: U.S. Geol. Survey Misc. Geol. Inv. Map I—761—C, scale 1:24,000. Sims, J. D., Fox, K. F., Jr., Bartow, J. A., and Helley, E: J., 1973, Preliminary geologic map of Solano County and parts of N apa, Contra Costa, Marin, and Y010 Counties, California: U. S. Geol. Survey Misc. Field Studies Map MF—484, 5 sheets, scale 1:62,500. Sims, J. D., and N ilsen, T. H., 1972, Preliminary photointerpretation map of landslides and other surficial deposits of parts of the Pittsburg and Rio Vista 15-minute quadrangles, Contra Costa and Solano Counties, California: U.S. Geol. Survey Misc. Field Studies Map MF—322, scale 1:62,500. Slemmons, D. B., 1967, Pliocene, and Quaternary crustal movements of the Basin-and-Range Province, USA: Osaka City Univ. Jour. Geosciences, vol. 10, p. 1—11. Smith, D. D., 1960, Geomorphology of the San Francisco peninsula: Stanford Univ., unpub. Ph.D. thesis, 433 p. Steinbrugge, K. V., Cloud, W. K., and Scott, N. H., 1970, The Santa Rosa, California earthquakes of October 1, 1969: Washington, U.S. Govt. Printing Office, 99 p. Steinbrugge, K. V., and Zacher, E. G., 1960, Fault creep and property damage, in 'Creep on the San Andreas fault: Seismol. Soc. America Bull., v. 50, p. 389-396. Taylor, F. A., and Brabb, E. E., 1972, Map showing distribution and cost by counties of structurally damaging landslides in the San Francisco Bay region, California, winter of 1968—69: U.S. Geol. Survey Misc. Field Studies Map MF—327, scale 1:1,000,000. Thatcher, W., and Hanks, T. C., 1973, Source parameters of southern California earthquakes: J our. Geophys. Research, v. 78, p. 8547—- 8576. Tocher, Don, 1958, Earthquake energy and ground breakage: Seis- mol. Soc. America Bull., v. 48, p. 147—153. 1959, Seismic history of the San Francisco Bay region, in San Francisco earthquakes of March 1957: California Div. Mines Spec. Rept. 57, p. 39—48. 1960, Creep on the San- Andreas fault—creep rate and related measurements at Vineyard, California, in Creep on the San An- dreas fault: Seismol. Soc. America Bull., v. 50, p. 396—404. Trask, J. B., 1964, Earthquakes in California from 1800 to 1864: California Acad. Sci. Proc., v. 3, p. 130—153. Trifunac, M. D., 1975, A three-dimensional dislocation model for the San Fernando, California, earthquake of February 9, 1971: Seis- mol. Soc. America Bull. (in press). Trifunac, M. D., and Hudson, D. E., 1971, Analysis of the Pacoima dam accelerogram—San Fernando, California, earthquake of 1971: Seismol. Soc. America Bull. v. 61, p. 1393—1411. Unger, J. D., and Eaton, J. P., 1970, Aftershocks of the October 1, 1969, Santa Rosa, California, earthquakes (abs): Geol. Soc. America Abs. with Programs, v. 2, p. 155. U.S. Geological Survey, 1971a, Surface faulting, in The San Fernando California earthquake of February 9, 1971: U.S. Geol. Survey Prof. Paper 733, p. 55—76. 1971b, Worldwide correlation of surface displacement and Richter magnitude, in Geological Survey research 1971; U.S. Geol. Survey Prof. Paper 750—A, p. 168. 1972, Slope map of the San Francisco Bay region: U.S. Govt. Printing Office, 3 sheets, scale 1:125,000. U.S. Soil Conservation Service, 1968, Soils of Santa Clara County, California: Washington, D. C., U.S. Govt. Printing Office, 227 p. Van Horn, Richard, 1972a, Slope map of the Sugar House quadrangle, Salt Lake County, Utah: U.S. Geol. Survey Misc. Geol. Inv. Map I—766—C, scale 1:24,000. 1972b, Landslide and associated deposit map of the Sugar House quadrangle, Salt Lake County, Utah: U.S. Geol. Survey Misc. Geol. Inv. Map I—766—D, scale 1:24,000. 1972c, Relative slope stability of the Sugar House quadrangle, Salt Lake County, Utah: U.S. Geol. Survey Misc. Geol. Inv. Map A102 I—766—E, scale 1:24,000. Varnes, D. J ., 1958, Landslide types and processes, chap. 3, in Eckel, E. B., ed., Landslides and engineering practice: Natl. Research Council, Highway Research Board Spec. Rept. 29, NAS—NRC Pub. 544, p. 20—47. Vedder, J. G., and Wallace, R. E., 1970, Map showing recently active breaks along San Andreas fault between Cholame Valley and Tejon Pass, California: US. Geol. Survey Misc. Inv. Map 1—574, scale 1:24,000. Wallace, R. E., 1968, Notes on stream channels offset by the San Andreas fault, southern Coast Ranges, California, in Dickinson, W. R., and Grantz, Arthur, eds., Proceedings of conference on geologic problems of San Andreas fault system: Stanford Univ. Pubs. Geol. Sci., v. 11, p. 6—21. 1969, Planning along active strands of the San Andreas fault: Assoc. Eng. Geologists, National Meetings, San Francisco, 1969, Field Trips (Guidebook), p. E5—E6. 1970, Earthquake recurrence intervals on the San Andreas fault: Geol. Soc. America Bull., v. 81, p. 2875—2890. Waltz, J. P., 1971, An analysis of selected landslides in Alameda and Contra Costa Counties, California: Assoc. Eng. Geologists Bull., v. 8, p. 153—163. Warrick, R. E., 1974, Seismic investigation of a San Francisco Bay mud site: Seismol. Soc. America Bull., v. 64, p. Weaver, C. E., 1949, Geology of the Coast Ranges immediately north of the San Francisco Bay region, California: Geol. Soc. America Mem. 33, 242 p. Weber, G. E., and Lajoie, K. R., 1974, Holocene movement on the San Gregorio fault zone near Afio Nuevo, San Mateo County, Califor- nia (abs): Geol. Soc. America Abs. with Programs, v. 6, p. 273. Wentworth, C. M., Bonilla, M. C., and Buchanan, J . M., 1973, Seismic environment of the Burro Flats site, Ventura County, California: US. Geol. Survey open-file rept., 35 p. Wentworth, C. M., Ziony, J. I., and Buchanan, J. M., 1970, Prelimi- nary geologic environmental map of the greater Los Angeles area, California: US. Atomic Energy Comm., TID—25363, 41 p., scale 1:250,000. Wesson, R. L., Bennett, R. E., and Lester, F. W., 1972a, Catalog of earthquakes along the San Andreas fault system in central California, April—June, 1972: US. Geol. Survey open-file rept. 42 p. Wesson, R. L., Bennett, R. E., and Meagher, K. L., 1972b, Catalog of earthquakes along the San Andreas fault system in central California, January —March, 1972: US Geol. Survey open-file REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION rept. 60 p. Wesson, R. L., and Ellsworth, W. L., 1973, Seismicity preceding mod- erate earthquakes in California: Jour. Geophys. Research, v. 78, p. 8527—8546. Wesson, R. L., Lester, F. W., and Meagher, K. L., 1974, Catalog of earthquakes along the San Andreas fault system in central California, October—December 1972: US. Geol. Survey open-file rept., 46 p. Wesson, R. L., Meagher, K. L., and Lester, F. W., 1973a, Catalog of earthquakes along the San Andreas fault system in central California, July—September, 1972: US. Geol. Survey open-file rept., 49 p. Wesson, R. L., Roller, J. C., and Lee, W. H. K., 1973b, Time-term analysis and geological interpretation of seismic travel-time data from the Coast Ranges of central California: Seismol. Soc. America Bull., v. 63, p. 1447—1471. Whitney, J. D., 1865, Geology of California, vol. 1, part 1. Geology of the Coast Ranges: California Geol. Survey, p. 1—197. Williams, P. L., 1972, Map showing landslides and areas of potential landsliding in the Salina quadrangle, Utah: US. Geol. Survey Misc. Geol. Inv. Map I—591—L, scale 1:250,000. Wood, H. 0., 1908, Distribution of apparent intensity in San Francis- co, in The California earthquake of April 18, 1906, Report of the State Earthquake Investigation Commission: Carnegie Inst. Washington Pub. 87, p. 220—245. Wright, R. H., 1971, Map showing locations of samples dated by radiocarbon methods in the San Francisco Bay region: US. Geol. Survey Misc. Field Studies Map MF—317, scale 1:500,000. Wright, R. H., Campbell, R. H., and Nilsen, T. H., 1974, The prepara- tion and use of isopleth maps of landslide deposits: Geology, Oct. 1974, p. 483—485. Wright, R. H., and Nilsen, T. H., 1974, Isopleth map of landslide deposits, southern San Francisco Bay region, California: US. Geol. Survey Misc. Field Studies Map MF—550, scale 1:125,000. Yerkes, R. F., Bonilla, M. G., Youd, T. L., and Sims, J. D., 1974, Geologic environment of the Van Norman reservoirs area, north- ern San Fernando Valley, California: US. Geol. Survey Circ., 691, pt. 1., 35 p. Youd, T. L., 197 3, Liquefaction, flow and associated ground failure: U.S. Geol. Survey Circ. 688, 12 p. Ziony, J. I., Wentworth, C. M., and Buchanan, J. M., 1973, Recency of faulting; A widely applicable criterion for assessing the activity of faults: World Conference on Earthquake Engineering, 5th, Rome, Italy, Proc., p. 1680—1683. x$11.5. GOVERNMENT FRDITING OFFICE: 1975-0-690-036/7 elected "xampi’é‘smm the fénciscofl Bay Region, California WORK DONE IN COOPERATION WITH U.S. DEPARTMENT OF HOUSING AND URBAN DEVELOPMENT. OFFICE OF POLICY DEVELOPMENT AND RESEARCH GEOLOGICAL SURVEY PROFESSIONAL PAPER 94l-B COVER PHOTOGRAPH of San Francisco Bay Region taken April 14, 1972, at altitude of 65,000 feet from U-2 aircraft. Courtesy National Aeronautics and Space Administration (Ames Research Center, Moffett Field, Calif.) Front shows city of San Francisco and Golden Gate at bottom, San Francisco Bay and city of Oakland in middle, Sacramento-San Joaquin Delta and crest of Sierra Nevada at top. Back shows Bolinas Lagoon and trace of San Andreas fault at bottom, San Pablo Bay in middle, Sacramento valley and crest of Sierra Nevada at top. Seismic Safety and Land-use Planning— Selected Examples from California By M. L. BLAIR and W. E. SPANGLE, WILLIAM SPANGLE and ASSOCIATES BASIS FOR REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION , CALIFORNIA GEOLOGICAL SURVEY PROFESSIONAL PAPER 941—B Mellzodxfor using selsmie zonalion and hazard mapping in land—use planning and regulation jointly supported by the U .S. Geologieal Survey and the Department of Homing and U roan Development, Office of Polity Development and Research, as part of a program to develop and apply earth—science lnfin‘mation in support of land—use planning and decisionmaking UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1979 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Searetary GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Cataloging in Publication Data Blair, M. L. Seismic safety and land-use planning—selected examples from California. (Basis for reduction of earthquake hazards, San Francisco Bay region, California) (Geological Survey professional paper ;941-B) Bibliography: p. B80-B82. Supt. of Docs. no.: I 19.16z941-B 1. Seismology—California. 2. Land use—Planning~California. I. Spangle, W. E., joint author. 11. Title. III. Series. IV. Series: United States. Geological Survey. Professional paper ; 941-B. QE535.2U6B56 363.3’4 79-20292 For sale by the Superintendent of Documents, US. Government Priming Office Washington, DC. 20402 Stock Number 024-001-03231-7 FOREWORD This report is a product of the San Francisco Bay Region Environment and Resources Planning Study, an experimental program that was designed to facilitate the use of earth- science information in regional planning and decisionmaking. The study, conducted from 1970 to 1976, was jointly supported by the US. Geological Survey, Department of the Inter- ior, and the Office of Policy Development and Research, Department of Housing and Urban Development. The Association of Bay Area Governments actively participated in the study and also provided liaison with other regional agencies and with local governments. Although the study was focused on the nine—county 74,4(l0-square—mile San Francisco Bay region, it explored a problem common to all communities: how best to plan for orderly development and growth and yet conserve our natural resource base, insure public health and safety, and minimize degradation of our natural and manmade environment. Such planning requires that we understand the natural characteristics of the land, the processes that shape it, its resource potential, and its natural hazards. These subjects are chiefly within the domain of the earth sciences—geology, geophysics, hydrology, and the soil sciences—and information from these sciences can help guide growth and development. But the mere existence of‘ information does not assure its effective use. Relatively few planners, elected officials, or citizens have the training or experience needed to recognize the significance of basic earth— science information, and many of the conventional methods of presenting earth—science in- formation are ill-suited to their needs. The San Francisco Bay Region study has aided planners and decisionmakers by: identifying important geologic and hydrologic problems that are related to growth and development; providing the earth-science information that is needed to solve these problems; interpreting and publishing findings in forms understandable to and usable by nonscientists; establishing avenues of communication between scientists and users; and exploring different ways of applying earth-science information in planning and decisionmaking. More than 100 reports and maps have been produced. These cover a wide range of topics, such as flood and earth- quake hazards, unstable slopes, engineering characteristics of hillside and lowland areas, mineral and water resources, solid and liquid waste disposal, erosion and sedimentation, and bay-water circulation patterns. Seismic safety and land-use planning—selected examplesfmm Calz'fiflnia is one of the final reports in the San Francisco Bay Region study. The authors are city and regional planners and have participated in the seismic-safety planning of several California communities. In this report they discuss the earth-science data needed for effective planning and the methods that can be used by local and regional government to reduce earthquake risk to acceptable levels. Much of the discussion is illustrated with examples drawn from experience in California where seismic-safety planning is mandated by State law. Although public attitudes, procedures, and legal requirements differ in other states, the basic earth-science needs and the planning strategies discussed are relevant wherever earthquake hazards are an issue in decisions related to public policy. This report is the companion volume of an earlier publication, Studies/221‘Seismic Zonation of [he San Francisco Bay Region (U.S. (leol. Survey Professional Paper 94l—A), which dealt chiefly with the geologic and bydrologic causes of earthquake damage. For conformity with the earlier report, and because the scientific literature on earthquakes customarily uses metric values, measurements in this report are expressed in metric terms followed by their English equivalents. % )0 Z . / Robert I). Brown, ]r. Project Director San Francisco Bay Region Study III CONTENTS Page Foreword __________________________________________________ BIII Abstract __________________________________________________ 1 Introduction ______________________________________________ 1 Purpose and scope __________________________________ 2 Overview of seismic hazards __________________________ 2 Seismic hazards defined __________________________ 4 Surface faulting ______________________________ 5 Ground shaking ______________________________ 7 Ground failure ________________________________ 7 Tsunamis and seiches ________________________ 7 Earthquake damage ______________________________ 7 Damage to buildings __________________________ 8 Damage to dams ____________________________ 8 Damage to utilities __________________________ 10 Damage to transportation systems ____________ 10 Fire ________________________________________ 10 Earthquake magnitude and intensity ______________ 11 Professions involved in seismic risk reduction __________ 14 Land-use planning ________________________________ 18 Geology, engineering geology ______________________ 18 Civil engineering ________________________________ 19 Soils engineering ________________________________ 19 Structural engineering ____________________________ 19 Architecture ____________________________________ 19 Building inspection ______________________________ 19 Governmental framework for reducing seismic risk ________ 19 Federal programs ____________________________________ 20 Research and technical information ________________ 21 US. Geological Survey ________________________ 21 National Oceanic and Atmospheric Administration ____________________________ 21 National Science Foundation __________________ 21 Emergency preparedness and disaster relief ________ 21 Civil Defense Preparedness Agency ____________ 22 Federal Disaster Assistance Administration ____ 22 Program Administration __________________________ 22 A—95 review ________________________________ 22 Environmental impact assessment ____________ 22 Programs of the Department of Housing and Urban Development ________________________ 23 Insurance ________________________________________ 23 State role—California ________________________________ 24 Legislation and advice ____________________________ 24 California legislature ________________________ 24 Governor’s Earthquake Council ________________ 25 Joint Committee on Seismic Safety ____________ 25 Seismic Safety Commission ____________________ 26 Research and information ________________________ 26 Structural standards ______________________________ 27 Critical facilities ________________________________ 28 Department of Water Resources ______________ 28 California Department of Transportation ______ 28 Emergency preparedness and disaster relief ________ 28 Land-use planning and regulation ________________ 28 Office of Planning and Research ______________ 29 California Environmental Quality Act ________ 29 Page Governmental framework for reducing seismic risk—Continued Area-wide planning—San Francisco Bay region ________ B29 Association of Bay Area Governments ____________ 30 San Francisco Bay Conservation and Development Commission ____________________________________ 31 California Coastal Zone Conservation Commission __ 32 Metropolitan Transportation Commission __________ 32 Evaluating seismic risk __________________________________ 33 Identifying seismic hazards __________________________ 33 Surface rupture (A6—A12; A25—A30) ______________ 33 Ground shaking (A52—A57) ______________________ 34 Liquefaction (A68—A74) __________________________ 34 Landsliding (A75-A87) __________________________ 34 Flooding (A93—A94) ______________________________ 34 Selecting the design earthquake ______________________ 35 Predicting geologic effects ____________________________ 36 Ground shaking __________________________________ 36 Liquefaction and lateral spreading ________________ 40 Landsliding ______________________________________ 40 Flooding ________________________________________ 40 Seismic zonation ________________________________ 40 Inventorying cultural features ________________________ 46 Current land use ________________________________ 48 Structures with high and involuntary occupancy ___- 48 Hazardous structures ____________________________ 48 Lifelines ________________________________________ 48 Facilities for emergency response __________________ 49 Other critical facilities ____________________________ 49 Assessing seismic risk ________________________________ 49 Dollar loss ______________________________________ 49 Deaths and injuries ______________________________ 49 Population at risk ________________________________ 50 Relative risk ____________________________________ 50 Scenarios ________________________________________ 51 Determining acceptable risks __________________________ 52 Land-use planning and seismic safety ______________________ 53 The planning process __________________________________ 54 Identify issues and define objectives ________________ 55 Collect and interpret data ________________________ 55 Formulate plans __________________________________ 56 Evaluate impacts ________________________________ 56 Review and adopt a plan __________________________ 56 Implement the plan ______________________________ 56 Planning examples ____________________________________ 57 Flaming to reduce risk from ground shaking ______ 57 Site investigation and design requirements ,-__ 57 Abating structural hazards ____________________ 58 Planning to reduce risk from ground failure ________ 58 Landsliding ___________________________________ 58 Liquefaction ________________________________ 61 Planning to reduce risk from surface rupture ______ 62 Surface rupture—undeveloped areas __________ 64 Surface rupture—developed areas ____________ 67 Planning to reduce risk from flooding ______________ 68 Tsunamis ____________________________________ 69 Dam and dike failure ________________________ 71 CONTENTS VI . Page Page Land-use planning and seismic safety—Continued Land-use planning and seismic safety—Continued Putting it all together ________________________________ B69 Postearthquake reconstruction ________________________ B76 Land-capability analysis __________________________ 71 Earthquake prediction ________________________________ 78 Land-use policy and regulation ____________________ 73 Conclusions ______________________________________________ 78 Project review ____________________________________ 76 References ______________________________________________ 80 ILLUSTRATIONS Page FIGURE 1— 4. Maps showing: 1. Plate boundaries and epicenters of 30,000 earthquakes recorded between 1961 and 1967 ____________________ B3 2. Approximate location of San Andreas fault in California __________________________________________________ 4 3. The shallow, intermediate, and deep-focus earthquakes in the circum-Pacific earthquake belt ______________ 5 4. Seismic risk zones of the United States based on damages associated with historic earthquakes _____________ 6 5. Diagram of four types of fault movement characterized by the sense of movement relative to the fault and to the horizontal __________________________________________________________________________________ 7 6. Faults in southern California and epicenters associated with San Fernando earthquake ____________________ 9 7—11. Photographs showing: 7. Damage to small buildings caused by the San Fernando earthquake ______________________________________ 10 8. Damage to hospitals caused by the San Fernando earthquake ____________________________________________ 12 9. Slide damage to Van Norman Dam ______________________________________________________________________ 14 10. Damage to Sylmar converter station ____________________________________________________________________ 15 11. Damage to highway interchange ________________________________________________________________________ 15 12. Graph showing relationship between magnitude and energy ______________________________________________ 16 13. Map showing intensity distribution of San Fernando earthquake of February 9, 1971 ______________________ 17 14. Diagram showing steps in evaluating seismic risk ________________________________________________________ 33 15. Map of California showing behavior of different segments of San Andreas fault __________________________ 36 16. Map showing location and length of surface rupture associated with a postulated earthquake of magnitude 6.5 on the San Andreas fault __________________________________________________________________________ 37 17. Diagram showing predicted geologic effects of postulated earthquake of magnitude 6.5 on the San Andreas fault ________________________________________________________________________________ 38 18. Generalized geologic map of area crossed by demonstration profile ________________________________________ 39 19. Map showing relative amplification of bedrock motion ____________________________________________________ 41 20—25. Maps showing: 20. Liquefaction and lateral spreading ______________________________________________________________________ 42 21. Landsliding __________________________________________________________________________________________ 43 22. Flooding ______________________________________________________________________________________________ 44 23. Seismic hazard zones __________________________________________________________________________________ 45 24. Maximum earthquake intensity predicted on a regional scale ____________________________________________ 47 25. Day and night populations, areas of high occupancy, and risk zones, Palo Alto, Calif ________________________ 53 26. Diagram of the planning process ________________________________________________________ * 1; ______________ 55 27—33. Maps showing: ' , 27. Estimated building damage levels for a 1906-type earthquake, San Francisco, Calif ________________________ 59 28. Density of precode, Type C buildings, San Francisco, Calif ________________________________________________ 60 29. Special geologic study areas, San Francisco, Calif ________________________________________________________ 61 30. Risk zones for land-use planning, Santa Clara County Baylands __________________________________________ 63 31. Active and probably active faults and fracture zones for a portion of San Mateo County, Calif ______________ 65 32. Special Studies Zone for a portion of San Mateo County, Calif ____________________________________________ 66 33. Known and inferred fault trace locations and setback lines, Portola Valley, Calif __________________________ 67 34. Aerial photograph showing Hayward fault traces in urban and urbanizing area ____________________________ 68 35. Map showing Hayward redevelopment plan ______________________________________________________________ 70 36. Land-capability map for multi-family residential use ____________________________________________________ 74 37. Natural hazards map, City of San Jose "a": ________________ . ___________________________________________ 77 38. Seismic safety zones, Santa Clara County __-_,_____,____-,,___; _________________ 1 ________________________ 79 TABLE HHD-‘r—‘D-‘D—‘b-lr-‘D-l {DmQGSOIACADNJr—I ,_. otooogaspvgnpomp-i CONTENTS v11 TABLES Page . Magnitude/size comparisons between any two earthquakes ______________________________________________________ B16 . Modified Mercalli intensity scale ______________________________________________________________________________ 16 High priority recommendations of the Task Force on Earthquake Hazard Reduction ______________________________ 20 Cost of earthquake insurance for residential buildings __________________________________________________________ 24 Recommendations of the Governor’s Earthquake Council ________________________________________________________ 25 . Summary of recommendations of the Joint Committee on Seismic Safety __________________________________________ 26 San Francisco Intensity Scale for 1906 earthquake ______________________________________________________________ 46 . Projected total loss from earthquake shaking, 1970—2000 ________________________________________________________ 50 . Deaths and hospitalized injuries ________________________________________________________________________________ 50 . Life loss from dam failure ____________________________________________________________________________________ 51 . U.S. population-at-risk by seismic risk zone and state ____________________________________________________________ 52 . Earthquake ratings for common building types __________________________________________________________________ 54 . Scale of risks for various building uses ________________________________________________________________________ 54 . Risk analysis of a motor inn __________________________________________________________________________________ 54 . Criteria for permissible land use in Portola Valley ______________________________________________________________ 62 . Risk zones for settlement and ground failure, Santa Clara'County ________________________________________________ 62 . Land and building uses appropriate for various risk zones, Santa Clara County ____________________________________ 64 . Costs associated with ground shaking resulting from events on the San Andreas, Hayward, or Calaveras faults ______ 72 . Summary of costs for multifamily residential use ________________________________________________________________ 73 BASIS FOR REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION, CALIFORNIA SEISMIC SAFETY AND LAND-USE PLANNING—SELECTED EXAMPLES FROM CALIFORNIA By M. L. BLAIR and W. E. SPANGLE, WILLIAM SPANGLE and ASSOCIATES ABSTRACT Earthquakes are inevitable, but their damaging effects can be greatly reduced. Land-use planning and management based on maps showing hazards and seismic zones can be particularly effective in reducing loss of life, as well as injury and property damage from earthquakes. Many Federal, State, and areawide programs encourage and sup- port city and county actions to reduce seismic risk. The Federal gov- ernment provides funds for seismic research, directs emergency pre- paredness activities, provides disaster relief, considers seismic hazards in program administration, and subsidizes insurance in tsunamiprone areas. Among the States, California’s program for re- ducing seismic risk is the most comprehensive. Through various State agencies, California provides data on seismic hazards, estab- lishes structural standards, regulates construction and operation of certain critical facilities, maintains an emergency preparedness plan, and requires cities and counties to prepare seismic safety plans. Although areawide agencies can reduce seismic risk in many ways, those in the San Francisco Bay region do so primarily by establishing procedures and criteria for project review to ensure that seismic hazards are considered. Effective local planning to reduce seismic risk is based on an evaluation of the nature and degree of risk—risk being defined as a function of the nature, severity, and frequency of seismic hazards and of the exposures of persons and property to those hazards. Asses- sing risk starts with recognizing the overall seismicity of the area and identifying the potential for ground shaking, landsliding, liquefaction, surface rupture, and flooding. A "design earthquake" is then selected as a basis for predicting, as precisely as possible, the location and severity of the various seismic effects. Cultural features are inventoried and mapped, special attention being given to facilities such as large dams whose failure could be catastrophic, facilities necessary for disaster response, and high-occupancy struc- tures. Using this information, the degree of seismic risk can be estab- lished and expressed in terms of potential dollar loss, deaths, and injuries, population exposure, relative risk, or scenarios describing the probable effects of a design earthquake. Plans and regulations can then be formulated to reduce risk to a level the public is willing to accept. Local land-use planning and regulation can be used to reduce seismic risk, particularly in undeveloped or sparsely developed areas. Many cities and counties in California are successfully inte- grating plans to reduce seismic risk into their general planning pro- grams. Methods include considering seismic hazards in analyzing land capability, developing land-use policy and regulations consis- tent with seismic risk, and establishing project review procedures to ensure consideration of seismic hazards in land-use decisions and land-development practices. Plans and programs to reduce seismic risk can also be formulated to respond to scientifically valid earth- quake predictions and to direct postearthquake reconstruction. INTRODUCTION Then the thunder crashed and rolled, and lightning flashed; and there was a great earthquake of a magnitude unprecedented in human history. The great city of “Babylon” Split into three sections, and cities around the world fell in heaps of mbble*** And islands vanished and mountains flattened out*** (Revelations 16: 18—20) Earthquakes have inspired fear and awe throughout man’s time on earth. Often attributed to the “wrath of God” or vengeful spirits, earthquakes dramatically confront man with the insignificance of his power be— fore the forces of nature. As Charles Darwin (as quoted in Elders, 1974, p. 20) observed in 1835 when the ground convulsed beneath him during an earthquake in Chile: A bad earthquake at once destroys our oldest associations; the earth, the very emblem of solidity, has moved beneath our feet like a thin crust over a fluidrvone second of time has created in the mind a strange idea of insecurity, which hours of reflection would not have produced. Indeed, individuals are virtually helpless during the course of an earthquake. They must “ride it out” wher- ever they happen to be at the time the earthquake strikes. But helplessness is_ confined to those seconds when the ground is shaking; man has the knowledge and ability to avert many of the damaging effects of earthquakes. ' The basic premise of this report is that actions can and should be taken to lessen the impact of earth- quakes in seismically active areas. Based on geologic and seismologic data, land uses and design and occu- pancy of structures can be adjusted to reduce significantly the loss of life, injury, and damage from earthquakes. Failure to make these adjustments will result in needlessly high costs in human suffering and Bl B2 property damage when the as yet unpredictable, but inevitable, earthquakes occur. PURPOSE AND SCOPE A major aim of the San Francisco Bay Region Envi- ronment and Resources Planning Study has been to gather information on the complex effects of earth- quakes and to present such information in a form di- rectly applicable to land-use planning and decision- making. A report edited by Borcherdt (1975) presents the earth-science phase of this study. It consists of a series of scientific articles defining earthquake hazards and methods of predicting their relative severity throughout a planning area. Seismic zones delineating the different effects of earthquakes, ranked in terms of relative severity, are identified in the Borcherdt report. Such seismic zona- tion provides .the geological and seismological basis for relating land-use plans and regulations, structural de- sign criteria, construction practices, and emergency re- sponse plans to recognized earthquake hazards. A map or set of maps may be prepared showing zones having a similar degree of hazard. In conjunction with maps of cultural features, these hazard maps indicate areas where the same kinds of risk-reduction methods may be successfully applied. The seismic zones indicate the wide geographic variation in the effects and local geologic conditions. If areas subject to severe effects can be accurately identified, land use and development decisions can reflect potential risk; and significant loss of life, injury, and damage can be avoided. This report is a companion volume to the Borcherdt report (Professional Paper 941—A) and shows how in- formation on seismic hazards can be effectively incor- porated into land-use planning and decisionmaking to reduce seismic risk. The first report (Professional Paper 941—A), was written by earth scientists to pre- sent the state-of-the-art for seismic zonation of the San Francisco Bay region. This report (Professional Paper 941—B), is written by planners, and it outlines possible applications of seismic hazard information with em- phasis on land-use planning and regulation. Success in increasing seismic safety requires an in- terdisciplinary effort including earth scientists, engineers, and planners. To be useful, scientific and engineering data must be translated to a form under- standable to planners and public policy makers. In some cases, the engineer can serve as an intermediary between scientist and planner. Engineering interpre- tations relate seismic information directly to issues of project feasibility, structural design, and cost; thus they provide an important input to planning decisions. However, in dealing with broader relationships of land-use patterns and intensities of seismic hazards, REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION the planner must draw more directly from basic geologic and seismologic information. A communica- tion bridge between scientist and planner must be es- tablished. This report is an attempt to establish such a bridge. In this report, the seismologic information needed by planners is summarized as simply as possi- ble consistent with accuracy. Although written by planners, the report was carefully reviewed by earth scientists at the US. Geological Survey, ensuring that the geologic information is presented and interpreted correctly. Seismic hazards are briefly defined to provide the back- ground for a discussion of ways of improving seismic safety. Federal, State, and areawide roles in response to seismic hazards are summarized, and examples are given of specific State and regional programs drawn from California and the San Francisco Bay region. A systematic approach to assessing seismic risk is set forth, and examples of various techniques are summarized. Typical planning responses of local gov- ernment to seismic risk are shown to relate to the na— ture of the hazard and the level of development in areas identified as hazardous. Examples of local plan- ning actions are presented wherever possible. The geologic information upon which this report is based pertains to the San Francisco Bay region, thus, most of the planning examples are drawn from this area. The methods of developing and interpreting in- formation and the ways to improve seismic safety used by bay region governments are, however, directly ap— plicable to other areas of the country with similar seismic hazards and governmental organizations. Planners and decisionmakers in most earthquake- prone areas of the United States may find that this report provides a useful framework on which to base their ongoing planning activities. OVERVIEW OF SEISMIC HAZARDS Knowledge of the nature and cause of earthquakes and their effects has increased tremendously in the last decade. Seismology has both contributed to, and ben- efited from, a revolutionary new concept—the theory of plate tectonics. According to this theory, the earth’s lithosphere or outer shell is formed of a mosaic of a dozen or more rigid plates in constant motion relative to each other (Dewey, 1972, p. 56). Most of the world’s large-scale active geologic processes—vulcanism, mountain building, formation of oceanic trenches, and earthquakes—are concentrated at or near plate boundaries (Press, 1975, p. 15). Records of seismic activity have been important in the development of the plate tectonic theory. As shown in figure 1, recorded earthquake epicenters, which tend B3 SEISMIC SAFETY AND LAND-USE PLANNING .GH .n .32 “$ng anwfigofi mo ”858:“. 30% mBohm SE. .818an 53m mo umwoo amok Q5 mac? 98m 550535 a mm m «in £de 3395. cam 23 \3 3.58 $8953 Eammsmb m mm N Mama 053nm umwm wfi MEESM bag—son mum“ 03E 353838 w. E H ”33.562 Joaw mo oimem SN #56 .2385: SE. 6me «Eu 8% 5953 3282 mwstvfihwo ooodm «a 235.2% ES gravy—don Bflmla 55on Bin 240E? B4 to cluster along plate boundaries, have been used to locate these boundaries. The concept of plate tectonics in turn provides a long-missing explanation of the un- derlying cause of earthquakes occurring along plate boundaries according to Press, (1975, p. 15), stresses build up where the relative motion of the plates is resisted by frictional forces. When the stress increases to the point where it exceeds the strength of the rocks of the lithosphere or overcomes the frictional forces at the boundary of a plate, fracturing occurs and an earthquake results. Plate boundaries are of three types, each with dis- tinctive geologic and seismologic characteristics (Dewey, 1972, p. 57—59). (1). Mid-oceanic ridge axes where plate boundaries are diverging. In these areas hot basaltic material is welling up from the earth’s interior and cooling to form new crustal material. Here the focus of earthquakes is usually shallow (less than 70 kilometers, 43 miles). (2). Transforms where the plates are sliding past each other. Lithospheric materials are not created or consumed along these boundaries; volcanic activity is limited, and earthquakes are shallow. (3). Subduction zones where plates are converging, one plate diving under the other to be eventually con- sumed in the asthenosphere—the molten or semimol- ten layer of the earth’s mantle beneath the solid litho- sphere. This zone is associated with deep oceanic trenches and volcanic island arcs having shallow, in- termediate (70—300 km, 43—186 mi), and deep (300— 700 km, 160—434 mi) earthquakes; and with continen- tal areas having primarily shallow earthquakes and high mountain ranges created by the compressive force of the converging plates. Examples of each type of plate boundary are shown by the numbers 1, 2, and 3 in figure 1. The San Andreas fault system marks the boundary between the Pacific and North American plates. The boundary is a transform—the plates move past each other. The por- tion of California, including Los Angeles and part of the bay region west of the San Andreas fault is on the Pacific Plate which is moving northwest on an average of a few centimeters a year (fig. 2). At this rate it would take Los Angeles about 10 million years to come abreast of San Francisco Bay (Yanev, 1974, p. 26). Plate movement is not uniform, however; portions of the plate boundary including that passing through the bay region have been locked for many years. It is considered highly probable that these locked portions will eventually give, resulting in earthquakes. The San Andreas fault system is part of the Circum-Pacific Earthquake Belt, sometimes called the “ring of fire”. This belt, shown in figure 3, outlining the Pacific, Cocos, and Nascan plates, is where intense vol- canic and seismic activity takes place. Nearly 80 per- cent of the world’s earthquakes occur along this belt, REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION i J _—>Z I | I k \ 0\ San '7 < \\ ‘4 Francrsco ’ \ ,6 O 1906 break 0 100 200 MILES i—e—r—H 0 100 200 300 KILOMETERS FIGURE 2.—Approximate location of San Andreas fault in California. accounting for the relatively high seismicity of the west coast of North America. Not all earthquakes can presently be explained by plate movements. Midplate earthquakes, hundreds of miles from known plate boundaries, do occur, and sci- entists are not yet agreed on their causes. In North America the most widely felt earthquakes ever re- corded were centered in New Madrid, Missouri. The largest of these shocks was felt nearly everywhere in the United States east of the Rocky Mountains (US. Geological Survey, 1971, p. 4). Figure 4 shows a map of relative seismicity in the United States by S. T. Algermissen (1969). This seis- mic risk map of the United States is based on the known distribution of damage from earthquakes. It has been incorporated into the 1973 Uniform Building Code as the basis for recommending differing struc- tural standards for risk zones 1, 2, and 3. The map indicates that most of the country can expect at least minor damage from earthquakes. SEISMIC HAZARDS DEFINED An understanding of plate tectonics is useful in ex- plaining why and where earthquakes are likely to oc— cur. And information based on experience from past earthquakes, as summarized in the seismic risk zone map of the United States, gives an overall picture of the relative seismicity of different parts of the country. SEISMIC SAFETY AND LAND-USE PLANNING Aleutian ls. ‘ e 0 ”099.9 Solomon Is. in ' ‘ /New Hebndes ~ 3 '° Samoa Is Fiji 15...: . 0 New : Tonga ls. Caledonia . o] :8 , 0°} New Zealand SEISMIC BELTS \ -—- Betts of shallow- focus earthquakes o o o o Intermediate-focus earthquakes on. o Deep-focus earthquakes Hawaiian Is. Equator / / Ii FIGURE 3.—The circum-Pacific Earthquake Belt showing shallow, intermediate, and deep focus earthquakes (figure adapted from Strahler, 1971, p. 443). However, this is only the tip of the iceberg in defining earthquake hazards. An earthquake unleashes a complex chain of natural events often catastrophic, and difficult to predict. The major geologic effects of earthquakes include surface faulting (ground rupture), ground shaking, ground failure, and flooding from tsunamis and seiches. These earthquake hazards are defined below. SURFACE FAULTING Faults are “planes or surfaces in earth materials along which failure has occurred and materials on op- posite sides have moved relative to one another in re- sponse to the accumulation of stress” (Nichols and Buchanan-Banks, 1974, p. 2). Fault movement does not always extend to the surface of the earth, but when it REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION B6 .8me domflnfiwwzv moxmsvsfimw oCSmE firs woe—308?. mwmmfimw do woman mbam GSED o5 mo mason gm: EEmBml.v 550$ can one ono as. one 90: an: can can / _ / _ w \ .32 S ..8m 3382. 2852 3692. 3:3 26:2. 2: 2 maria. azimfima :. 3.3.38 .0: mm; 0:2 sum» 5 moxmacztoo uEmnEnu .0 8:25qu .o 55:03. 22305 2...— Aatéum 93:65.3 5:: $3593: on 8 35:3 33:63 «En 35.93: .5203 SEE .0 3:23.33 a...» ”83.2 52: .o 8:330 M32655» on»... 5.: 35533 3:58.... .is. 0.: «Eu «9.2.35.3 «2363 3 5.3959... .565. 05 co 033 a as: «.5 loom n.2,: 95 .0 5:2; van :5 922:. 2 3:83:00 Hommfimu in: - m uzo~ .23 ..2.2 2: .o __> 3.33:. 2 3:83:00 63:5“. «.2032 u N uzoN .038 .22 05 .0 _> can > 3:55.... o. mucoamotggmucouon o.— / :5: .235 32.3 .25Eauca. 5;, 35.93: 2 WW/ 35:3 8:3 35 maxmzaztno €33. ”oumEnc 3:5 - u uzo~ dunes. oz - 0 ~23 I mmp. LN Ni‘lltJ‘IflN-u . . \. QCSHNO .Nu m. .m w v .w omflfl cam . N . FAN”. FOZHEMMEDZ FDOIF_>> FOO r .v a . m. Fwn. ©3698;an_m>m<3mmwu . 36 v 92.3 «EESMO gnaw /. 0mm maxim m - .8250 ahwmoQ O . W a 5 5W5 w. :3“ a— who I; own . . .m w on m. ow: Fiona : . . ._x. o .N v >onE< >.w=m> 0:535 @- n m w. .w_> .m m > / c n o n n .. If“ r w a“ .m m. 3M8 m m .m .m m. 33. own .9 W O 3/ m . Bag—Em N. w x .. EzmoEfiO « m. .m 4 v . _ «£2.30 . . . .v _ w. m m . T. c.3532 . I can . . . .m y n . 2.33 v/ . v. . . m >_ . _ neunfifld / {liL .. . _ . mz . 35m _ . . m . . . _ nuancofiou // .. NU. _ x .538on _ / «G . . . . . own _ .. euoaquam . . ‘ ‘ on: ow: om: cm: on: ‘w: om: comp 19 0mm, B18 has issued a Guide to Interprofessional Collaboration in Environmental Design to aid its members who are engaged in interdisciplinary projects (ICED, 1974). Land-use planning is a young profession and has roots in the “design professions.” Many land-use plan- ners come into planning with design training or expe- rience. Land-use planners, as a group, thus have had closer ties to engineering, architecture, and landscape architecture than to the earth sciences. However, in recent years because of increasing costs associated with failure to recognize geologic conditions in land development, planners and other design professionals are working more closely with earth scientists and are using more earth—science information. Earth scientists have become increasingly aware of the value of their information and expertise to land—use planning. For example, AEG (Association of Engineering Geologists), through meetings, conferences, and publications has provided a substantial body of information that contributes to a better understanding of the role of the engineering geologist (and earth scientists generally) in relation to others concerned with reducing seismic risk. The publication of the proceedings from their an- nual meeting in 1973 includes papers by authors from many fields. The integration of physical and economic planning was addressed in a paper (Lakshmanan, 1972) commis- sioned by the U. N. Centre for Housing, Building and Planning. This paper provides useful insights to some of the problems planners face in working with profes- sionals from other fields and gives some general guides to interprofessional collaboration. Lakshmanan (1972 writes: It is fair to suggest, that in the current state—of-the-art, no such overarching framework for integrative economic and physical plan- ning is available. The complexity of issues involved does not hold out much promise for such integrative models in the near future. A policy of learning by doing and willingness to adapt action with experience is warranted. This is a social learning process, learning while doing, with a willingness to experiment without excessive commitment. The approach is to focus an interdisciplinary team of physical and economic planning skills on a broad issue. This will require new planning roles and skills in social, economic, and physi- cal design, a good appreciation of implementation processes, and knitting together the separate planning systems into a broader de- velopmental planning system. LAND—USE PLANNING The land-use planner employed by a public agency has responsibility for developing plans, regulations, and procedures to guide and control the physical devel- opment pattern of a planning area. In responding to seismic risk through land-use planning, the planner can be a key coordinator drawing information from sci- entists and engineers, developing recommendations for public policy, interacting directly with elected de- REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION cisionmakers and the public, and reviewing and evaluating land-development proposals. All of these functions require an ability to draw information from other disciplines, apply it appropriately to the plan- ning problems at hand, and aid the earth scientists in communicating their scientific and technical knowl- edge to the public. In these tasks, the land-use planner will need to interact directly or indirectly with many professions. In a simple geologic setting with low-intensity de- velopment, a team consisting of a land-use planner and an engineering geologist frequently is adequate to ad- dress the problems related to seismic risk. As intensity of development increases or the seismic hazards be- come more complex, additional, more specialized pro- fessionals will need to be involved in addressing the problems. The scope and role of key professionals deal- ing directly with, or making direct application of, earth-science information in the discharge of their own professional responsibilities are discussed below. GEOLOGY, ENGINEERING GEOLOGY The geologist or engineering geologist frequently works directly with the land-use planner in simple situations and, for more complex problems, advises on programs, data requirements, and areas of expertise needed on the interdisciplinary team. Geologistsand engineering geologists have the same basic education and frequently have somewhat similar experience. En- gineering geology is a specialization within geology, emphazing application of geologic informatiOn to en- gineering problems. Many other specializations are recognized within the geologic profession, and some are particularly important in analyzing specific seismic risk problems. The geologist, as researcher and interpreter, may be a primary member of a team working on seismic risk reduction through land-use planning, or he may pro- vide the information used by others. In either case his work is basic in formulating public policy, developing land-use regulations and project review procedures, and reviewing development proposals. Some local agencies in California have added geologists to their staffs or contracted for geologic services as needed to assist the land-use planner in identifying and respond- ing to geologic problems. The engineering geologist brings to the land-use planning team training and experience in the interpre— tation of geologic conditions affecting safety and econ- omy of engineering works. He provides a bridge be- tween earth-science researchers, land-use planners, engineers, and architects concerned with seismic safety policy and its application. Geologists are responsible for compiling maps of SEISMIC SAFETY AND LAND-USE PLANNING seismic hazard zones and collaborating with land-use planners and engineers in relating levels of risk expo- sure in hazard zones to land uses and critical facilities. They also collaborate in assessing probable damage to existing development in relation to seismic hazards. Information developed by seismologists dealing with the forces, lines of direction, periodicity, and other characteristics of earthquakes is fundamental to evaluation of the seriousness of seismic problems in any given area. This information is the starting point for land-use planning to reduce seismic risk, However, the planner, and others on the land-use planning team, usually depend on published information with geologist, engineering geologist, or structural engineer serving as interpreter. CIVIL ENGINEERING Civil engineering is a broad profession with several well-recognized areas of specialization, such as soils, structural, foundation, sanitary, and transportation. However, many civil engineers are engaged in general practice and have experience with a wide range of civil engineering problems. With this breadth of experience the civil engineer can, with collaboration from mem- bers of the planning team, provide perspective on the general nature of the engineering problems that are likely to result from an earthquake. The civil engineer can also help identify the engineering specialty needed to address seismic risk problems in particular situa- tions. SOILS ENGINEERING Soils engineering, a branch of civil engineering, deals with the mechanical properties of soil and their effects on structures. The soils engineer draws infor- mation from soils science, geology, and hydrology and applies it to specific engineering problems. He carries out side investigations and recommends design and construction solutions to soil problems such as expan- siveness, erodibility, and soil creep. Soils engineers prepare soils reports on specific development proposals; they may also be employed by public agencies to review and evaluate soils reports prepared by others. Soils en- gineers work closely with engineering geologists and civil, foundation, and structural engineers on particu- lar design problems. The soils engineer can provide valuable assistance in identifying the limits particular soils may place on aseismic construction. STRUCTURAL ENGINEERING Structural engineering is another specialization within civil engineering. This branch of engineering is responsible for designing structures. The structural engineer collaborates in the design of major public B19 facilities and assesses the need for, and costs of, struc- tural measures to mitigate problems associated with particular sites. ARCHITECTURE Architecture is the science and art of designing buildings to blend form and function with safety. The architect charged with responsibility for preparing plans and specifications and providing on-site supervi- sion during construction must work within design pa- rameters recommended by engineering geologists and soils and structural engineers to produce a safe build- ing. On any particular project, his interaction with the land-use planner is primarily as a member of a design team, frequently as the lead professional with respon- sibility for coordinating the team effort. BUILDING INSPECTION All the efforts of the professionals can be seriously compromised unless building inspection is carefully and expertly carried out. The building inspector is the public official responsible for seeing that building code provisions are adhered to. He reviews final construc- tion plans and inspects construction to insure that local code requirements are met. In carrying out his duties, the building inspector, who often is not an engineer, relies on the engineering or public-works department personnel or consulting structural engineers for evalu- ation of seismic safety and other aspects. The land-use planner is often in the position of coor- dinating the work of these professionals to assist in developing plans and policies, framing regulations, es- tablishing review procedures, and reviewing develop- ment proposals. Timing may be critical. Basic geologic information is needed early to identify and evaluate seismic hazards and to assist in developing appropriate policies and regulations. The contribution of the en- gineers is needed primarily at the time of project de- sign and review. A structural engineer, however, is also needed during the formulation of seismic safety plans to give a general evaluation of the safety of exist- ing and proposed structures in the area. GOVERNMENTAL FRAMEWORK FOR REDUCING SEISMIC RISK Governmental agencies at all levels have a part in reducing seismic risk. Under the present system, risk reduction through land-use planning is carried out primarily by local governments. However, the opera- tions of local agencies affect and are affected by the planning and decisionmaking of government agencies at Federal, State, and regional levels. These govern- ment agencies often preempt or influence local de- B20 cisionmaking by imposing requirements for funds, criteria for programs, shared responsibility for specific functions such as transportation, and regulations such as those concerning environmental quality or the con- tent of local plans. To an increasing extent, local governments are de- pendent on Federal and State funds to carry out their responsibilities. This means that plans and programs developed at the local level are often framed with an eye not only to locally expressed objectives and con- cerns, but also to Federal and State funding require- ments. Thus, individual governmental decisions be- come part of a network of decisions made by other agencies, at different jurisdictional levels over a period of time. Because of increasing political, economic, and legal interdependency, effective planning by local gov- ernment often depends on complementary decisions of other local, Federal, State, and regional agencies. The following sections outline the major Federal, State, and regional programs and activities that pro- vide the context for local seismic-safety planning in the San Francisco Bay region. Programs directly or indi- rectly influencing local land-use planning are em- phasized. The description of State programs is limited to California because California has gone further than other states in enacting programs to reduce seismic risk. The description of area-wide activities is similarly limited to agencies in the San Francisco Bay region. Such focusing on California and the bay region allows discussion of actual plans and programs illustrating various methods of reducing seismic risk consistent with the powers and responsibilities of typical state and regional governmental agencies. FEDERAL PROGRAMS The Federal government has a broad constitutional mandate to protect the health, safety, and welfare of the residents of the United States. Direct Federal ef- forts to reduce risk from seismic hazards have evolved from the Federal commitment to provide disaster relief to states and localities devastated by earthquakes. With increasing urbanization of seismically active areas, particularly the west coast, the potential cost of disaster and recovery assistance, borne primarily by the Federal government, is awesome. The moderate San Fernando earthquake of 1971 caused property damage estimated at more than a half billion dollars. Federal aid including grants and loans exceeded $450,000,000. Approximately $135,000,000 was allo- cated from the President’s Disaster Fund—an amount larger than for any natural disastr since the fund’s establishment (US. Senate, 1971, p. 71; US. Office of Emergency Preparedness, 1971 and 1972, p. 179). REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION Expenditures of this magnitude have focused atten- tion on the need to reduce future damages. In 197 0, a Task Force on Earthquake Hazard Reduction was es- tablished by the Office of Science and Technology (since disbanded) to "develop an appropriate national action program for the reduction of the human suffer- ing and property damage attendant upon an earth- quakem” (US. Office of Science and Technology, 1970, p. 1). Table 3 lists the Task Force’s high-priority rec- 0mmendations. These recommendations provide a framework for Federal involvement in seismic-hazard reduction. Not all of the recommendations are currently being carried out in a consistent, coordinated program, but many have been incorporated into the activities and pro- TABLE 3_High—priority recommendations of the Task Force on Earthquake Hazard Reduction [US Office of Science and Technology. 1970, p. 5] A. Significant benefits probably beginning to accrue in the short term (less than 5 years after eginnmg of recommended act- ion): A—l: Engineered earthquake resistance for new governmental facilities. A—2: Engineered earthquake resistance for new nongovernmental facilities. A—3: Seismicity (or risk, or probability) maps. A—4: Earthquake geologic hazards maps. A—5: Urban planning to minimize seismic hazard. A—6: Earthquake hazards abatement in older facilities. A—7: Cost—benefit studies. A—8: State and local government role in geologic hazards reduction. A—9: Federal total plan for immediate response. A— 10: Federal responsibility in reconstruction. A—11: Federal responsibility in earthquake insurance. A—12: Strong motion equipment and analyses. A—13: Full-scale testing. B. Significant benefits probably beginning to accrue in the inter- mediate term (5—10 years after beginning of recommended act- ion): B—l: Applied research on seismic design criteria. B—2: Postearthquake analyses. B—3: Fault mapping, dating, and specialized geologic mapping. B—4: Local seismic networks. B—5: State responsibility in earthquake hazards reduction. B—6: Newly discovered hazards and older construction. B-7: Taxes and tax reform. C. Si nificant benefits probably beginning to accrue mainly in the anger term (10 years or more after beginning of recommended action): C—1‘ Basic research in earthquake engineering. C—2: Earthquake prediction research. C-3: Earthquake control research. C~42 Geodetic research. C—5: Worldwide seismic network continuation. C—6 Tsunami hazard research. C—7 Basic research in seismology. C—8 Basic research on causes and mechanisms of crustal failure. SEISMIC SAFETY AND LAND-USE PLANNING grams of several Federal agencies. At present the Fed- eral Government has four major functions in reducing risks from future earthquakes: conducting or funding research and providing technical information, en- couraging emergency preparedness and providing dis- aster relief, considering seismic hazards in program administration, and requireing insurance. Each of these Federal functions is discussed below. RESEARCH AND TECHNICAL INFORMATION Federal agencies which sponsor research and provide technical information concerning seismic hazards in- clude the U.S. Geological Survey, the National Oceanic and Atmospheric Administration, and the National Science Foundation. U.S. GEOLOGICAL SURVEY The USGS (U.S. Geological Survey) is responsible for carrying out the seismic research and hazard mapping recommended by the Task Force on Earthquake Hazard Reduction. The USGS publishes maps of faults, evaluates their degree of activity, and compiles and maintains records of historical and recent seismic ac- tivity. Geologists and seismologists investigate the re- lationship of geologic structure to seismic wave amplification and ground shaking, liquefaction poten- tial, and other seismic hazards. Post earthquake analysis is carried out in the field by USGS profession- als in cooperation with investigators from universities, professional organizations, and other groups. The USGS also investigates the feasibility of earth- quake prediction and control and is responsible for is- suing official earthquake predictions. The USGS ad- ministers EROS (Earth Resources Observation Sys- tem) to evaluate the application of data obtained from remote sensing. A few previously unknown faults have been identified from remote sensing imagery from high-altitude aircraft and satellites, but most known faults have been recognized and mapped by geologists in the field. On May 22, 1974, Congress enacted Public Law 93—288 (88 Stat. 143), which is known as the “Disaster Relief Act of 1974” to provide an orderly and continuing means of assistance by the Federal Gov- ernment to State and local governments in carrying out their respon- sibilities to alleviate the suffering and damage which results from***disaster**" By subsequent redelegations of authority, the Director of the Geological Survey was empowered to exercise the authority, functions, and powers granted by Section 202 of the Disaster Relief Act of 1974 with respect to disaster warnings for an earthquake, volcanic eruption, landslide, mudslide, or other geological catastrophe. (Federal Register, vol. 42, no. 70, p. 19292, Tuesday, April 12, 1977). Section 202 (a) of the Act states that B21 “The President shall insure that all appropriate Fed- eral agencies are prepared to issue warnings to State and local officials.” In addition, Section 202 (b) states that The President shall direct appropriate Federal agencies to provide technical assistance to State and local governments to insure that timely and effective disaster warning is provided. The Federal Register statement cited (p. 19292) de- scribes the Survey’s capabilities and limitations for advance recognition and warning of various kinds of geologic-related hazards and the procedures pro- posed to carry out the responsibilities delegated under the Act. The USGS has undertaken several pilot studies in urban areas, such as the San Francisco Bay Region Environment and Resources Planning Study, to pro- vide geologic and seismologic information for use in land-use planning and decisionmaking. Under the Earth Sciences Applications Program, several projects, similar to the San Francisco Bay Region Study, have been undertaken to provide earth-science information for use in land-use planning. Studies have been com- pleted of the Greater Pittsburgh area, Connecticut River Valley, Washington-Baltimore region, and the Phoenix-Tucson area. Still underway (in 1977) are projects in the Puget Sound area, in the Front Range Corridor in Colorado, and in Fairfax County, Virginia. NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION NOAA National Oceanic and Atmospheric Adminis- tration) is responsible for setting up and maintaining a tsunami warning system for coastal areas of the United States. The agency’s previous functions with respect to solid-earth geophysics have been assumed by the U.S. Geological Survey, but N 0AA still maintains and provides data related to seismology. NATIONAL SCIENCE FOUNDATION The NSF (National Science Foundation) provides funds for research by colleges and universities, non- profit research organizations, and other groups in all the scientific disciplines. Under its program, Research Applied to National Needs, NSF has funded research to investigate the social and engineering aspects of seismic-hazard reduction. Such projects have included studies of the impacts of hazard-mitigation measures and the potential impact of improved capability to pre- dict earthquakes. EMERGENCY PREPAREDNESS AND DISASTER RELIEF The Civil Defense Preparedness Agency and the Fed- eral Disaster Assistance Administration are the agen- cies most directly responsible for emergency prepared- ness and disaster relief. B22 Civil Defense Preparedness Agency. The main objec- tive of the CDPA (Civil Defense Preparedness Agency), established within the Department of Defense in 1972, is to improve prospects for survival of the population in the event of a nuclear war. A secondary objective “is to improve the readiness of State and local governments to respond to peacetime emergencies” (US. Civil De- fense Preparedness Agency, 1974, p. 6). In this connec— tion, the agency provides planning assistance to state and local governments to develop their natural disas- ter preparedness plans and capabilities and plays a key role in coordinating Federal state, areawide, and local emergency response plans. Among other activities, the CDPA funds emergency planning efforts of other gov- ernment agencies, makes surplus Federal property and equipment available for emergency response, and op- erates emergency warning and communications sys— tems. Federal Disaster Assistance Administration. The FDAA (Federal Disaster Assistance Administration), in the Department of Housing and Urban Develop- ment, is responsible for “programs concerning disaster research, preparedness, readiness evaluation, disaster relief, and recovery, and coordination of other agency disaster assistance activities (US. Office of the Federal Register, 1973, p. 253). The agency publishes the Fed- eral Earthquake Response Plan which outlines the Federal Government’s role in response to a major earthquake. FDAA administers the Disaster Relief Act of 1974 enacted to help State and local governments alleviate the suffering and damage caused by floods, tsunamis, earthquakes, mudslides, and other emergencies and major disasters. The Disaster Relief Act of 1974 provides for financial and technical assis- tance to the states to develop plans, programs, and reg- ulations for hazard reduction, disaster preparedness, and disaster relief. The act requires that property to be replaced, repaired, or restored with the assistance of Federal relief funds be insured, if insurance is avail- able, against future losses. To receive any disaster loan or grant, a state or local government must agree to evaluate natural hazards in the disaster area and take actions to mitigate the hazards, such as control of land-use and construction practices (US. Congress, 1974, See. 406). PROGRAM ADMINISTRATION Federal officials who administer Federal and Feder— ally funded programs have an opportunity to influence decisions related to seismic safety. The extent to which this is done often depends on how legislation and regu- lations are interpreted by administrators. Two Federally mandated procedures provide the basic framework for consideration of seismic risk in REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION administering Federal and Federally funded programs: (1) review procedures set forth by the US. Office of Management and Budget (A—95), and (2) environmen- tal impact assessment required by the National En- vironmental Policy Act of 1969. A—95 REVIEW A—95 review procedures are designed to implement the Intergovernmental Cooperation Act of 1968 by in- suring that Federally funded projects are consistent with State, areawide, and local planning objectives. Under these procedures applicants for Federal funds for a wide variety of projects must notify designated State and regional clearinghouse agencies which re- view proposed projects for consistency with State, areawide, and local plans and programs. The clearing- house agency forwards the project description to any affected agencies for their review and comment. The comments are only advisory, but a Federal agency must defend in writing any decision to fund a project which has received a negative review. This procedure at least assures that the review comments of affected public agencies are considered. Comments may be made concerning the natural characteristics of a proposed project site. In seismically active areas, the nature and extent of seismic hazards are an appropriate and necessary subject of comment in the A—95 review process. ENVIRONMENTAL IMPACT ASSESSMENT The NEPA (National Environmental Policy Act) of 1969 requires that an EIS (environmental impact statement) be prepared for proposed legislation and for other Federal actions that may significantly aifect the quality of the human environment. The statement must describe the environmental impact of the pro- posed action, identify adverse and unavoidable en- vironmental effects, list alternatives to the proposed action, describe how local short-term uses of man’s en- vironment are related to maintaining and enhancing long-term productivity, and identify any irreversible and irretrievable commitments of resources. Guidelines and procedures for preparing environ- mental impact statements are issued by the CEQ (Council on Environmental Quality) and administered by the EPA (Environmental Protection Agency). The current guidelines (Council on Environmental Quality, 1973) help Federal agencies prepare environmental impact statements and require that environmental fac- tors, such as seismic hazards, be explicitly considered before most Federal actions. To the extent possible, en- vironmental impact assessment and A—95 review are coordinated. SEISMIC SAFETY AND LAND-USE PLANNING PROGRAMS OF THE DEPARTMENT OF HOUSING AND URBAN DEVELOPMENT Although most Federal activities are subject to A— 95 review and environmental impact assessment re- quirements, each agency develops its own operating procedures to carry out the intent of the legislation and regulations. The programs of HUD (Housing and Urban Development) support planning, community development, and housing activities of many state, areawide and local governments throughout the coun- try. The way these programs are administered can af- fect seismic safety in HUD-assisted activities and proj- ects. The Housing and Community Development Act of 1974 (US. Dept. of Housing and Urban Development, 1975b) combined several HUD programs covering urban renewal, neighborhood development, and public facilities into a single program funded with Commu- nity Development Block Grants. Local community de- velopment programs are subject to A—95 review, and individual projects funded with Community Develop- ment Block Grants are subject to environmental im- pact assessment. In reviewing applications, HUD may challenge statements of fact and program decisions re- lated to seismic safety, and it may require additional information from the applicant. Implications of the community development program and proposed proj- ects should be evaluated in the review of applications for Community Development Block Grants. Under the Comprehensive Planning Assistance Pro- gram (Section 701 of the Housing and Community De- velopment Act of 1974), HUD provides grants to cities and counties, and metropolitan, regional, and State planning agencies. Grants may be used for planning to mitigate and reduce hazards, among other activities. All agencies applying for grants must adopt a land-use element by August 23, 1977 (US. Department of Hous- ing and Urban Development, 1975a, p. 36862). The land-use element must identify areas where growth should and should not take place giving comprehensive consideration to environmental factors. Planning ac- tivities supported by these grants must be conducted in accord with the National Environmental Policy Act of 1969 (Public Law 91—190). HUD (US. Department of Housing and Urban Development, 1975a, 36860) specifies that each agency shall: (1) Identify salient elements of the natural and the man-made envi- ronments, their interrelationships, and major problems and/or opportunities they present for community development; (2) Assess those environmental factors which will: (i) Minimize or prevent undue damage, unwise use, or unwar- ranted pre-empting of natural resources and opportunities; (ii) Recognize and make prudent allowance for major latent environmental dangers or risks (e.g., floods, mud slides, earthquakes, air and water pollution); and B23 (iii) Foster the human benefits obtainable from use of the nat- ural environment by wise use of the opportunities availa- ble (e. g., use of natural drainage systems for park and recreational areas, HUD has also issued Minimum Property Standards which define minimum levels of acceptable design and construction for Federally subsidized housing and for housing approved for Federally insured mortgages. Where earthquakes are a recognized hazard, the stan- dards require a comprehensive soil investigation, spe- cial foundation design, and structural design to with- stand lateral forces in accord with the latest Uniform Building Code (US. Dept. of Housing and Urban De- velopment, 1973). INSURANCE The Task Force on Earthquake Hazard Reduction recommended studies of the feasibility of offering earthquake insurance at actuarial rates (US. Office of Science and Technology, 1970, p. 26). Private earth- quake insurance has been available since the early 1900’s, but relatively few property owners have pur- chased it. For example, of the more than half a billion dollars in property damage caused by the San Fer- nando Earthquake, only about $46,000,000 was cov-. ered by insurance (Baker, 1971, p. 31). The potential for expanding private insurance cov- erage is limited, because insurance companies are re- quired to maintain reserves sufficient to meet obliga- tions in the event of a major earthquake. Property owners have also been reluctant to assume the extra cost of earthquake insurance, although it is relatively low. Table 4 lists typical costs of earthquake insurance in California in 1974 for residential buildings accord- ing to risk classes which are based on structural type. A deductible amounting to 5 percent of the building value usually applies. Several proposals for Federal action to increase the extent of insurance coverage for damage from earth- quakes and other natural disasters have been made, but, to date, the only direct Federal response has been enactment of the National Flood Insurance Program in 1968 (Public Law 90—448) as amended in 1973 (Public Law 93—234). Under this program, an individual must purchase flood insurance to be eligible for any kind of Federal financial assistance, including loans from Fed- erally insured or regulated institutions, for acquisition of property or construction in identified special flood- hazard areas. The insurance is available at rates pres- ently subsidized by the Federal Government in partici- pating communities which have adopted and enforced land use and development controls to reduce the flood hazard. Areas subject to flooding from tsunamis are included in the program. When the Federal Insurance Adminis- B24 TABLE 4.—-—Cost of earthquake insurance for residential buildings [Yanev, 1974, p. 234] Class Cost per of $1,000 risk Building description coverage I Small wood-frame and frame $1.50 stucco buildings (less than four stories). II Steel-frame and reinforced $2.50 poured-concrete buildings. III Reinforced concrete-frame (col- $3.00 umned) buildings and other reinforced shear-wall masonry buildings. IV Wood-frame buildings with masonry veneers or other liabilities. V Steel-frame and reinforced concrete-frame buildings with segmented mansonry walls (brick, concrete block, etc.). VI Single-family dwellings (less than three stories) 0 reinforced masonry; other reinforced masonry buildings with liabilities. VII Unreinforced brick and larger reinforced masonry buildings without steel or concrete framing. Unreinforced masonry, masonry-veneer, or adobe buildlngs. $3.50 $3.50 $4.00 $7.50 V111 $25.00 tration of the Department of Housing and Urban De- velopment (US. Federal Insurance Administration, 1975, p. 13429) has identified such “coastal high hazard areas,” the community: Must provide that all new construction or substantial improvements within the designated coastal high hazard area be located landward of the reach of the mean high tide; Must provide that all new con- struction and substantial improvements within the designated coast- al high hazard area be elevated on adequately anchored piles or columns to a lowest floor level (including basement) at or above the 100-year flood level and securely anchored to such piles or columns: Must provide that all new construction and substantial im- provements within the designated coastal high hazard area have the space below the lowest floor free of obstructions or are constructed with “breakaway walls” intended to collapse under stress without jeopardizing the structural support of the building so that the impact on the building of abnormally high tides or wind-driven water is minimized. Such temporarily enclosed space shall not be used for human habitation; Must prohibit, within the designated coastal high hazard area the use of fill for structural support; Must prohibit, within the designated coastal high hazard area, the location of any portion of a new mobile home park, expansion to an existing mobile home park, and any new mobile home not in a mobile home park. These regulations are designed to encourage land- use decisions that minimize losses caused by flooding from the sea. Federally insured or regulated institutions such as banks or savings and loan companies are not permitted to provide funds to substantially modify or purchase existing structures within coastal high-hazard areas REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION unless the. property is insured under the program. When more accurate maps of hazard areas are re- leased, the insurance will be available at actuarial rates which will reflect the flood risk. STATE ROLE—CALIFORNIA Officials in nearly every agency of California State government need to consider seismic safety in carrying out their duties. The location and construction of public facilities, management of State lands, provision of services, and delegation of powers and responsibilities to local governments should all reflect an awareness that damaging earthquakes are inevitable. Few, if any, populated areas of the State are free from the risk of a major earthquake, and most large population centers are less than 80 kilometers (50 miles) from faults that can generate earthquakes (Jennings, 1975). Although the historical record and geologic knowl- edge are insufficient to precisely establish the fre- quency of future earthquakes, one estimate is that California can expect a great earthquake (Richter magnitude greater than 7.7) every 60—100 years, a major earthquake (Richter magnitude 7.0—7.7) every 20 years, and a moderate earthquake (Richter mag- nitude 6.0—6.9) every 8‘10 years (California State Legislature Joint Committee on Seismic Safety, 1974, p. 200). LEGISLATION AND ADVICE The State legislature and, more recently, various advisory bodies, have led in defining and coordinating California’s role in reducing seismic risk. CALIFORNIA LEGISLATURE Legislative actions to reduce seismic risk have al- most always followed damaging earthquakes. The most far-reaching actions followed the earthquakes of 1933 in Long Beach, 1964 in Anchorage, Alaska, and 1971 in San Fernando. Below is a brief chronology of legislative actions linked to each of these earthquakes. Details of key legislation are provided in relevant sec- tions. After the 1933 Long Beach earthquake, the legisla- ture adopted the Field Act establishing seismic stand- ards for the construction of school buildings and the Riley Act setting forth lateral force requirements for certain other buildings. After the 1964 Anchorage, Alaska, earthquake and in response to two conference reports on this major earthquake, the legislature in 1969 established the Joint Committee on Seismic Safety to advise it on earthquake hazards. SEISMIC SAFETY AND LAND-USE PLANNING After the 1971 San Fernando earthquake the Gov— ernor appointed the Governor’s Earthquake Council, and the Legislature intensified support for the Joint Committee on Seismic Safety and passed several seismic-safety bills. The major acts include the Alquist-Priolo Special Studies Zones Act to reduce risk from fault rupture; the Hospital Safety Act to strengthen construction standards for hospitals; the Dam Safety Act to require evaluation of the safety of dams, mapping of potential inundation areas, and preparation of evacuation plans; and an amendment to the Government Code requiring each county and city to adopt a seismic safety element as part of its general plan. GOVERNOR’S EARTHQUAKE COUNCIL In January 1972 the GEC (Governor’s Earthquake Council) was established to recommend measures to reduce future earthquake losses. The recommen- dations presented in the Council’s report of November 21, 1972 closely parallel those of the Federal Task Force on Earthquake Hazard Reduction and emphasize measures which can be undertaken administratively by the executive branch of State government. Table 5 summarizes the major recommendations and responsi- ble agencies (California Governor’s Earthquake Coun- cil, 1972, p. 5-15). JOINT COMMITTEE ON SEISMIC SAFETY The Joint Committee on Seismic Safety was estab- lished by the State Legislature in 1969 as an out- growth of studies following the 1964 Anchorage, Alaska, earthquake. The committee’s purpose was to “develop seismic safety plans and policies and recom- mend to the Legislature any legislation needed to minimize the catastrophic effects upon people, prop- erty, and operation of our economy should a major earthquake strike any portion of California” (Califor- nia State Senate, 1969, Senate Concurrent Resolution No. 128 of 1969, Resolution Chapter 378). The commit- tee, composed of four senators and four assemblymen, relied on the technical and professional expertise of more than 70 persons who served on five advisory groups: engineering considerations and earthquake sciences; disaster preparedness; postearthquake recov- ery and redevelopment; land-use planning; and gov- ernmental organization and performance. The San Fernando earthquake also spurred the Joint Committee’s efforts. A special subcommittee was formed to conduct a postearthquake investigation, and the resulting report provided a basis for many of the Committee’s legislative recommendations (California State Legislature, Joint Committee on Seismic Safety, 1972). The reports and recommendations of the advi- B25 sory groups and the committee’s basic legislative rec- ommendations are contained in its final report, Meet- ing the Earthquake Challenge, published in January 1974 (California State Legislature, Joint Committee on Seismic Safety, 1974). The Committee’s recommen- dations, which emphasize legislative action, are sum- marized in table 6. Perhaps the most important recommendation of both the Joint Committee on Seismic Safety and the Gov- ernor’s Earthquake Council was to create a permanent organization within State government to coordinate State efforts to improve seismic safety. SEISMIC SAFETY COMMISSION In July 1975, the California Seismic Safety Commis- sion was established by the Legislature. The 17- member commission includes geologists; structural, civil, mechanical, and soils engineers; architects; plan- ners; representatives of local government; and State legislators. The Commission’s responsibilities as stated in the California Government Code, (1974, Sec. 8897) are: (a) Setting goals and priorities in the public and private sectors; (b) Requesting appropriate state agencies to devise criteria to pro- mote seismic safety; (0) Recommending program changes to state agencies, local agen- cies, and the private sector where such changes would reduce the earthquake hazards; (d) Reviewing reconstruction efl‘orts afier damaging earthquakes; TABLE 5.—Recommendations of the Governor’s Earthquake Council [Adapted and summarized from California Governor's Earthquake Council, 1972, pp. 5—15] Subject State Agencies Area Recommended Actions Responsible Research, Support basic seismological research. OES (Ofiice of ‘ PYOVISIOIE of Support research in earthquake Emergency SeVioes), information engineering. UCB (UniverSity of Fund expanded seismo aphic network, California at Develop procedures and? provide funds for Berkeley): ‘ postearthquake studies. CDMG (Califorrua Prepare earthquake geologic hazard maps, Div1sion of seismicity ma 5. Mines & Geology). Disseminate e quake-related EERJ (Earthquake information. Engineering Research Institute).‘ \ DGS (Department of General Services). Critical Assess safety of public utility systems and PUC (Public Utilities structures dams. Commission). DWR (De ment of Water sources). Emergency Mandate local disaster plans including OES (Office of pre aredness an response evacuation rocedures. Coordinate F local disaster plans. Require disaster training in schools. Assess emergency operations including medical and communication era], State,'areawide, and Emergency Services). capabilities. Land-use Provide incentives and technical guidance Department of planning for preparation of seismic safety Conservation. Insurance ...... element. Consider local earthquake risk in public improvement projects. Require geologic reports on private and public projects in seismically active areas. Mandate inclusion of disaster coverage in standard fire insurance policies. Encoura 9 insurance industry to advise policy elders of disaster coverage. Office of Planning and Research. Department of Insurance. 'EERI is a private, nonprofit organization. B26 TABLE 6,—Summary of recommendations of the Joint Committee on Seismic Safety [Adapted and summarized from California State Legislature, Joint Committee on Seismic Safety, 1972] Subject Area Recommended Actions Land-use planning -_Provide for effective State review of local seismic safety elements. Require geologic and soils reports for subdivision and construction activity of substantial scope. Permit seismic and geologic hazards to be considered "blighting” conditions making an area eligible for redevelopment funds. Provide for preplanning of postearthquake redevelopment. Require evaluation of geolo 'c and seismic hazards in environmenta impact statements. Employ land-use controls to reduce seismic hazards. Discourage public investment in hazardous areas. Provide purchasers of real estate with property reports disclosing seismic and geologic hazards. Building Construction ______ Upgrade engineering standards and building code provisions. Assist local agencies in enforcing building code standards. Develop programs to train building officials and other ocal personnel in seismic design. Provide geologists, engineers, public safety officials, and others with reasonable protection from liability. Abatement of hazardous buildings __________ Develop hazard abatement program concentrating on re-1933‘ buildings. Inventory potential y hazardous buildings. Critical and high exposure facilities __________ Enforce seismic safety measures in construction of schools, hospitals, and emergency facilities. Review safety of high-rise structures and dams. Emergency preparedness measures __________ Ensure that local emergency plans are prepared and maintained as required. Establish procedures for review and approval of such lans. Conduct isaster exercises to test response. Increase allocation to State Emergency Fund. Require communities to prepare postearthquake reconstruction plans. Research ____________ Increase support of basic and applied research. Insurance __________ Require purchasers of residential buildings to carry earth uake insurance. Explore with Fe eral Government the possibility of comprehensive disaster insurance. 1Explained on p. B 10 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION (e) Gathering, analyzing, and disseminating information; (0 Encouraging research; (g) Sponsoring training to help improve the competence of spe- cialized enforcement and other technical personnel; (h) Helping to coordinate the seismic safety activities of government at all levels; and (i) Establishing and maintaining necessary working relationships with any boards, commissions, departments, and agencies, or other public or private organizations necessary to further an effective seismic safety program for the state. To carry out these responsibilities, the Commission (California Government Code, 1974, Sec. 8898) may: (a) Review state budgets and review grant proposals, other than those grant proposals submitted by institutions of postsecon- dary education to the federal government, in earthquake- related activities and to advise the Governor and Legislature thereon; (b) Review earthquake-related legislation proposals, to advise the Governor and Legislature concerning such proposals, and to propose needed legislation; (c) Recommend the addition, deletion, or changing of state agency standards when, in the commission’s view, the existing situa- tion creates an undue seismic hazard‘ or when new devel- opments would promote seismic safety, and conduct public hearings as deemed necessary on the subjects. Beginning in January 1977, the Commission’s duties were expanded to include advising the State Mining and Geology Board regarding Special Studies Zones (see p. B25) and the State Geologist regarding the State Strong-Motion Instrumentation Program. RESEARCH AND INFORMATION The CDMG (California Division of Mines and Geol- ogy) within the Department of Conservation (Califor- nia State Legislature, Joint Committee on Seismic Safety, 1974, p. 190) has the responsibility to provide: (1) information pertaining to earthquake and other geologic hazards, (2) to conduct, with city and county governments or Federal and other State agencies, large-scale geologic investi- gations "'to identify and delineate*"geologic hazards in and adjacentto metropolitan areas, (3) to organize and monitor a strong-motion instrumentation program in the State, and (4) to quickly identify potential post-earthquake geologic hazards, particularly weakened slopes that could be activated by after- shocks. The division is headed by the State Geologist and operates under the policy direction of the State Mining and Geology Board. California Division of Mines and Geology produced the Urban Geology Master Plan for California (Alfors, 1973). This report estimates, for the period 1970—2000, losses due to geologic hazards, the amount of losses that could be averted by applying current information and technology, and the cost of applying loss-reduction measures. Earthquake losses are estimated to be $21 billion from 1970 to 2000. Approximately half of the losses could be averted by applying existing risk- reduction measures. The cost of applying loss- reduction measures is estimated to be about 10 percent SEISMIC SAFETY AND LAND-USE PLANNING of the total projected losses. The study was intended to help establish State priorities for measures to reduce losses from geologic hazards. Calfornia Division of Mines and Geology also ad- ministers the Alquist-Priolo Special Studies Zones Act (Chapter 7.5, Division 2, Public Resources Code, 1972 as amended 1974 and 1975). Under this Act, the State Geologist maps special studies zones along potentially active and recently active fault traces. The zones are ordinarily less than 396 meters (a quarter-mile) wide unless special considerations indicate the need for a wider zone. Once the Special Studies Zones maps have been officially issued by CDMG, local jurisdictions must require geologic reports prior to approval of most new construction within the zones. Individual geologic reports are not required, however, for projects consist- ing of no more than one single-family, wood-frame home not exceeding two stories. The California Division of Mines and Geology, under the direction of the State Mining and Geology Board, establishes criteria and policies for content and review of the geologic reports, for revising the Special Studies Zones maps to reflect new geologic information, and’for city and county compliance with provisions of the Act. Through contracts with cities and counties, CDMG also provides geologic information for seismic safety and safety elements of the general plan and for other plan- ning purposes. STRUCTURAL STANDARDS The State OAC (Office of Architecture and Construc- tion) in the Department of General Services adminis- ters the California Field Act (Education Code Sections 15451—15465), which was passed after the 1933 Long Beach earthquake destroyed or seriously damaged many buildings and almost all public schools in the area. The Field Act specifies structural standards for construction of new public school buildings and re- quires that an architect or structural engineer prepare plans and supervise construction of school buildings. The act does not apply to State colleges and univer- sities or private schools. Similar provisions relating to other major buildings are set forth in the Riley Act, also adopted in 1933. The Field Act has been amended several times as the technology of designing and building earthquake- resistant structures has improved. Sections added to the Education Code in 1967 require inspection of pre- Field Act school buildings. Those found to be unsafe were to be replaced or brought up to code standards by June 1975, but some school districts have been granted additional time to meet the new requirements because of financial problems. Schools built since the Field Act B27 was passed in 1933 have performed well during earth- quakes (U.S. Office of Emergency Preparedness, 1972, p. 76). California legislation, also adopted in 1967, requires geologic and engineering investigations of any site proposed for a school building “to preclude siting of a school over or within a fault, on or below a slide area, or in any other location where the geological charac- teristics are such that the construction effort required to make the site safe for occupancy is economically un- feasiblem” (Education Code, Section 15002.1 1967). The OAC (Office of Architecture and Construction) also develops and enforces standards for hospital con- struction under contract with the Department of Health as required by legislation enacted in 1972 (California Health and Safety Code, 1972, Sec. 15000—15023). This legislation was passed after the San Fernando earthquake damaged four major hospi- tals so severely that they had to be evacuated; fifty of the fifty-eight deaths attributed to the San Fernando earthquake resulted from collapse or damage to hospi- tal buildings. OAC provides architectural and en- gineering services to State departments in the design and construction of State buildings and other facilities and prepares and administers the State’s building reg- ulations contained in Titles 17, 21 and 24 of the California Administrative Code. Since 1971, the Uni- form Building Code has been adopted by reference as part of Title 24 of the California Administrative Code. The Appendix of the 1927 Uniform Building Code, published by the International Conference of Building Officials, included suggested lateral-force design re- quirements to increase structural resistance to earth- quake ground motion. Lateral-force provisions have been modified several times subsequently, largely in accord with recommendations of the SEAOC (Struc- tural Engineers Association of California). Sections 17958 and 17922 of the California Health and Safety Code, enacted in 1975, require cities and counties to adopt the most recent edition of the Uniform Building Code. Section 2312, Chapter 23, Earthquake Regula- ‘tions, of the 1976 Uniform Building Code (Interna- tional Conference of Building Officials, 1976, p. 132— 150) contain the lateral-force requirements which apply in seismically active areas of the country. Building code requirements typically are minimum standards which apply to all structures regardless of differing geologic conditions. Local jurisdictions may enact requirements more stringent than those of the Uniform Building Code and some jurisdictions, notably Los Angeles and Long Beach, have attempted to relate building standards to geologic conditions of the site. Such codes are technically more difficult to prepare and B28 administer, but as stated by Yanev (1974, p. 53) “wit makes no sense to continue to build seemingly sound structures on unsound ground.” CRITICAL FACILITIES DEPARTMENT OF WATER RESOURCES The DWR (State Department of Water Resources) is responsible for constructing and operating the State Water Project and for the safety of non-Federal dams in California (California State Legislature, Joint Com- mittee on Seismic Safety 1974). Under the Alquist Dam Safety Act (Government Code, Section 8589—5, 1973), DWR and the OES (Office of Emergency Serv- ices), identify dams whose failure might lead to injury or loss of life. The owner of a dam so identified must prepare a map showing the extent of potential flooding from dam failure at full reservoir capacity. OES must review and approve all such maps, which then serve as the basis for emergency evacuation plans drawn up by local governments with advice from the State. CALIFORNIA DEPARTMENT OF TRANSPORTATION The CalTrans (California Department of Trans- portation), within the Business and Transportation Agency, is responsible for building and maintaining the State highway system and for planning a balanced transportation system. Earthquake-resistant design of highway facilities, particularly overpasses and bridges, is essential to prevent collapse and possible loss of life during an earthquake and to maintain the flow of traffic following an earthquake. The vulnerabil- ity of freeway overpasses was dramatically illustrated by the San Fernando earthquake. As a result, more stringent design standards for new construction and reconstruction were instituted by the Department of Transportation, and additional engineering research was strongly recommended (California Division of Highways, September 1971). In addition, existing highway structures are being evaluated and strengthened as funds permit. More directly related to land-use planning, the De- partment also recommends that more attention be paid to seismic hazards in locating highways and inter- changes. As stated in its report (California Division of Highways, 1971, p. 5—6) on the San Fernando earth- quake: Early in the route location process, active and inactive faults should be mapped. A general assessment of the seismic risk of various areas within the study zone should then be prepared. Consideration must be given to the location of major interchanges. They should be sited outside of heavily faulted areas wherever feasi- ble. Where seismic activity is highly probable, consideration should be given to avoiding complex multi-level interchanges in favor of simple designs with short span structures and maximum use of em- bankment. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION Early recognition of seismic risk might lead the planner to modify alignment or grade in order to minimize high cuts, fills, and bridge structures in a given area. Where a freeway must pass through a highly seismic area, the best and safest plan will generally be the simplest: close to the original ground, with simple, square bridge structures. These recommendations were incorporated into the Department’s Highway Design Manual of Instructions in March 1975 (Section 7—1104). EMERGENCY PREPAREDNESS AND DISASTER RELIEF The OES (Office of Emergency Services), within the Governor’s office, was established by the Emergency Services Act (Chapter 7, Division 1, Title 2 of the Gov- ernment Code, 1970). The act requires OES to coordi- nate the emergency activities of all State agencies. The OES develops and maintains the State Emergency Plan as required by Section 301b of the Federal Disaster Relief Act. This plan provides a framework for individual State agency and local gov- ernment plans as well as specifying procedures for de- livery of Federal aid. The Emergency Plan also requires that contingency plans be prepared for specific potential emergencies, including earthquakes. The State Earthquake Re- sponse Plan, published by OES, meets this require- ment. OES also coordinates postdisaster damage as- sessment and provides the Governor with information needed to declare an emergency or request Federal dis- aster assistance. The California Earthquake Predic- tion Evaluation Council, an advisory body to OES composed of geologists, seismologists, and geophysi- cists, reviews and evaluates specific information which could lead to an earthquake prediction. If the COuncil finds a significant possibility that an earthquake is imminent, OES provides preparedness and response information to State agencies and local governments and may provide public information to help individuals prepare for an earthquake. LAND-USE PLANNING AND REGULATION Many states have authorized local units of govern- ment to plan and regulate future development, but few states require local planning. In California, all cities and counties are required by State law to prepare and adopt a general plan which includes at least the follow- ing elements: land use, circulation, housing, conserva- tion, open space, seismic safety, noise, scenic highways, and safety. California law further requires that zoning and subdivision of land be consistent with the adopted general plan. The State Attorney General, a resident, or a property owner may bring suit against a city or county to force compliance with the consistency provi— sion of State law. SEISMIC SAFETY AND LAND-USE PLANNING In accord with a recommendation of the Joint Com- mittee on Seismic Safety, the requirement for a seismic safety element was enacted soon after the San Fer- nando earthquake. Section 65302(f) of the Government Code requires: A seismic safety element consisting of an identification and appraisal of seismic hazards such as susceptibility to surface ruptures from faulting, to ground shaking, to ground failures, or to the effects of seismically induced waves such as tsunamis and seiches. The seismic safety element shall also include an appraisal of mudslides, landslides, and slope stability as necessary geologic hazards that must be considered simultaneously with other hazards such as possible surface ruptures from faulting, ground shaking, ground failure and seismically induced waves. This legislation provides the basic framework in California for local seismic safety planning requiring, in effect, that cities and counties consider seismic hazards in formulating and implementing the general plan. The CIR (Council on Intergovernmental Rela- tions) (now disbanded) issued guidelines to assist local governments in preparing State-required general plan elements. The guidelines (California Council on Inter- governmental Relations, 1973, p. IV—24, 25) for pre- paring seismic safety elements suggested: A. A general policy statement that: 1. Recognizes seismic hazards and their possible effect on the community. 2. Identifies general goals for reducing seismic risk. 3. Specifies the level or nature of acceptable risk to life and property (see safety element guidelines for the concept of "acceptable risk”). 4. Specifies seismic safety objectives for land use. 5. Specifies objectives for reducing seismic hazard as related to existing and new structures. B. Identification, delineation, and evaluation of natural seismic hazards. C. Consideration of existing structural hazards. Generally, existing substandard structures of all kinds (including substandard dams and public utility facilities) pose the greatest hazard to a community. D. Evaluation of disaster planning program. For near-term earthquakes, the most immediately useful thing that a community can do is to plan and prepare to respond to and recover from an earthquake as quickly and effectively as possible, given the existing condition of the area. The seismic safety element can provide guidance in disaster planning. E. Determination of specific land-use standards related to level of hazard and risk. The seismic safety element is related to several other required plan elements. As stated in guidelines pre- pared by the Council on Intergovernmental Relations (California Council on Intergovernmental Relations, 1973, p. IV—27): The seismic safety element contributes information on the compara- tive safety of using lands for various purposes, types of structures, and occupancies. It provides primary policy inputs to the land use, housing, open space, circulation and safety elements. Within this legislative context, several State agen- cies and programs influence land-use planning with respect to seismic hazards. Of particular importance B29 are the Office of Planning and Research and the guidelines of the California Environmental Quality Act, as discussed in the following sections. OFFICE OF PLANNING AND RESEARCH The OPR (Office of Planning and Research), respon- sible to the Governor, develops long-range State goals and policies for land use and environmental quality, evaluates State agency plans and programs for en- vironmental impact, issues guidelines for preparing mandatory general plan elements (a function assumed from CIR), and provides assistance to local gov- ernments in preparing general plans. In 1972, the office published Environmental Goals and Policies setting forth recommended State actions to reduce environmental pollution and to protect en- vironmental resources. The report recommends that areas subject to strong earthquake shaking, tsunamis, and fault displacement be designated as areas of "criti- cal concern”. In such areas guidelines should be formu- lated “to encourage orderly development and protec- tion from natural calamities while minimizing adverse impact upon people or resourcesm” (California Office of Planning and Research, 1973, p. 3). CALIFORNIA ENVIRONMENTAL QUALITY ACT The California Environmental Quality Act of 1970, based on the National Environmental Policy Act, re- quires an EIR (Environmental Impact Report) for all public and private projects or actions which may have a significant effect on the environment and which in- volve a discretionary decision by a public agency. State guidelines and procedures for preparing EIR’s are is- sued by the California Resources Agency. The guidelines specify that impacts which “pose long-term risk to health or safety” be evaluated (California Re- sources Agency, December 1974, Section 15143, p. 19). Although not specifically mentioned in the act or guidelines, seismic hazards are usually considered in the environmental impact assessment. AREA-WIDE PLANNING—SAN FRANCISCO BAY REGION The nine-county San Francisco Bay region is highly vulnerable to earthquake damage. A major earthquake on any of the faults traversing the region would have devastating impact on the entire area. Thus, planning to reduce seismic risk and to increase the ability to respond to an emergency is appropriately a regional concern. The responsibilities of regional agencies (those with a jurisdictional area encompassing parts of more than one county) related to seismic safety are briefly described in the following sections. B30 ASSOCIATION OF BAY AREA GOVERNMENTS ABAG (Association of Bay Area Governments) is the only regional agency covering the entire nine-county bay area that is responsible for comprehensive plan- ning. Established in 1961 to develop plans and policies pertinent to region-wide problems, ABAG is a volun- tary association of city and county governments. Im- plementation of ABAG’s regional plans and policies depends on decisions by State and Federal agencies, other regional agencies, and local governments. How- ever, because ABAG is the A—95 review clearinghouse agency for the San Francisco Bay region, it can indi- rectly influence other governmental decisions through reviewing requests for Federal funds. Because many projects are competing for limited funds, a negative finding by ABAG, although advisory, is likely to be heeded by the funding agency. ABAG also reviews Federal projects proposed for the region for consistency with areawide plans. ABAG is giving increasing emphasis in its planning program to seismic concerns. A report presented to the Regional Planning Committee in May 1976, Areas of Critical Environmental Concern (ABAG, 1975a), lists policies and criteria for identifying critical land and water areas. Areas with known earthquake-related problems are among the critical areas and, according to policy, should be protected from premature or ex- tremely dense development (ABAG, 1975a, p. 44). These areas include: 1. Lands within 50 feet of a known active fault trace, as shown on Special Studies Zones Maps; 2. Lands subject to severe ground shaking shown as categories A and B on the map of max- imum earthquake intensities (Borcherdt and others, 1975); 3. Lands likely to liquefy in a major earthquake as mapped by Youd, Nichols, Helley, and Lajoie (1975). The report further recommends that land uses within 50 feet of a fault trace be limited to agriculture, recreation, secondary streets, and parking and that de- velopment in areas of severe ground shaking meet or exceed the requirements of the most recent uniform Building Code. Critical structures should be located, whenever possible, in the less hazardous areas. A land capability analysis (Laird and others, 1979) prepared by ABAG as part of the San Francisco Bay Region Study illustrates a method of analyzing land capability for a demonstration area of about 100 square miles in the Santa Clara Valley. The cost of damage or mitigation measures per acre which can be expected from geologic hazards is estimated for selected land uses. Such seismic hazards as ground shaking, surface REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION rupture, dam failure, dike failure, liquefaction, and landslides are considered. The report relies on informa- tion similar to that compiled for the entire San Fran- cisco Bay region by Borcherdt, Gibbs, and Lajoie (1975). The report is described more fully on pages B72—B73. In a related effort, ABAG undertook a project spon- sored by the Federal Civil Defense Preparedness Agency to assemble information on disasters and dis- aster mitigation, to outline a method of evaluating risk, and to clarify ABAG’s role in civil preparedness. (ABAG, 1975b) A committee formed to prepare a civil preparedness plan for ABAG concluded that ABAG should have no operational role in disaster response, but that ABAG should offer technical assistance to member jurisdic- tions in ways to reduce hazards, prepare for disasters, and plan for postdisaster recovery. ABAG prepared “A Method for Evaluating Hazards” (ABAG, 1975b) to help local governments set priorities for disaster preparedness. It sets forth a procedure for describing and analyzing hazards and suggests ways a local jurisdiction can de- cide which hazards are important, what measures can effectively reduce them, and appropriate priorities for action. The theme of ABAG’s annual General Assembly in February 1976 was earthquake preparedness and re- sponse. Following a two-day conference (including Federal, State and local government staff members, elected officials, representatives from private indus- tries and citizens) the Assembly adopted a resolution “making earthquake preparedness and response a high ABAG program priority” (Laird and others, 1979) and directing the Executive Board to define an ABAG pro- gram emphasizing legislation and advocacy; planning and technical assistance; and public information and education. As a result, ABAG has appropriated $30,000 for 1976—77 for the following work program: Legislative advocacy a. Monitoring proposed earthquake related legislation; preparing comments as appropriate. b. Working with staffs of State legislative committees, advocating legislation that would assist local governments’ preparedness efforts. Technical assistance to local governments a. Offering information and assistance to member governments in using previous ABAG work to upgrade and improve their seismic safety programs. b. Assisting local governments in using the findings and methods contained in the ABAG Land Capability Analysis Report to improve local seismic safety programs. Plan and project review a. Completing plan and project review procedures and policies on seismic safety. b. Conducting plan and project reviews, and preparing review comments in relation to seismic safety policies and programs. SEISMIC SAFETY AND LAND-USE PLANNING Legal research a. Cataloging legal research into responsibilities and liabilities of local jurisdictions for earthquake damage. ABAG is also seeking additional funds to augment the program and to implement some of the recommen- dations. To carry out policies and criteria related to seismic safety, ABAG relies on its project review powers. ABAG’s (1973) procedures for regional clearinghouse review of environmental impact statements contain checklists of environmental impacts associated with eight different types of projects and include an inven- tory of mapped environmental information. The proce- dures also set criteria for determining regional impact. The importance of seismic hazards, such as historically active faults, high ground-shaking potential, and liquefaction potential are recognized, and sources of geologic and seismological information are listed. SAN FRANCISCO BAY CONSERVATION AND DEVELOPMENT COMMISSION The BCDC (San Francisco Bay Conservation and Development Commission) is a State agency created by the State Legislature. BCDC was authorized to prepare a comprehensive plan for San Francisco Bay and its shores and to control development within its area of jurisdiction. The plan was adopted by the State Legis- lature, and BCDC became a permanent agency charged with carrying out the plan. The adopted plan has legal status and serves as a guide in the review of projects. BCDC shares jurisdiction over land-use deci- sions with the cities and counties which retain normal land use and building-permit controls. However, with certain minor exceptions, a permit from BCDC is re- quired for all projects within its area of jurisdiction. Thus it, in effect, holds veto power over any project proposal in conflict with the San Francisco Bay Plan. The BCDC plan and its project-review activities re- flect a strong concern for seismic safety. The agency’s jurisdiction consists primarily of tidelands, marshes, salt ponds, and diked and filled land underlain by bay mud. Such land is subject to particularly severe seismic ground shaking, liquefaction, differential settlement, and flooding. Yet in spite of these hazards, diking and filling of the baylands to accommodate urban uses has occurred and continues to be a problem. As stated in the San Francisco Bay Plan (San Francisco Bay Con- servation and Development Commission, 1969, p. 2): As the Bay Area’s population increases, pressures to fill the Bay for many purposes will increase. New flat land will be sought for many urban uses because most, if not all, of the flat land in communities bordering the Bay is already in use—for residences, businesses, in- dustries, airports, roadways, etc. Past diking and filling of tidelands and marshlands has already reduced the size of the Bay from about 680 square miles in area to little more than 400. Although some of this diked land remains, at least temporarily, as salt ponds or man- B31 aged wetlands, it has nevertheless been removed from the tides of the a . )Despite the risks involved, the State recognizes that some bay filling may be desirable or necessary if the benefits outweigh the disadvantages. The San Fran- cisco Bay Plan recommends approval to fill if one of the following four conditions is met: (1) the filling is in accord with the bay plan policies as to the bay-related purposes for which filling may be needed ( such as, ‘ ports, water-related industry, and water-related recre- ation) and is shown on the bay plan maps as likely to be needed, (2) the filling is in accord with bay plan policies as to purposes for which some fill may be needed if there is no other alternative (such as, airports, roads, and utility routes), (3) the filling is in accord with the bay plan policies as to minor fills for improving shoreline appearance or public access, (4) the filling would provide for new public access to the bay on pri- vately owned property and for improvement of shoreline appearance—in addition to what would be provided by the other bay plan policies—and the filling would be for bay-oriented commercial recreation and bay-oriented public assembly purposes. The question of safety of the fill must also be addressed before BCDC can issue a permit for filling. With respect to the safety of fills, the plan (San Fran- cisco Bay Conservation and Development Commission, 1969, p. 15) makes the following findings: Virtually all fills in San Francisco Bay are placed on top of Bay mud which presents many engineering problems. The construction of a sound fill depends in part on the stability of the base upon which it is placed. Safety of a fill also depends on the manner in which the filling is done, and the materials used for the fill. Similarly, safety of a structure on fill depends on the manner in which it is built and the materials used in its construction. Construction of a fill or building that will be safe enough for the intended use requires (1) recognition and investigation of all potential hazards—including (a) settling of a fill or a building over a long period of time, and (b) ground failure caused by the manner of constructing the fill or by shaking during a major earthquake—and (2) construction of the fill or building in a manner specifically designed to minimize these hazards. While the construction of buildings on fills overlying Bay deposits involves a greater number of potential hazards than construction on rock or on dense hard soil deposits, adequate design measures can be taken to reduce the hazards to acceptable levels. Policies to reduce potential earthquake damage to structures built on filled land include (San Francisco Bay Conservation and Development Commission, 1969, p. 17):. 1. The Bay agency should appoint a Fill Review Board consist- ing of geologists, civil engineers specializing in soils en- gineering, structural engineers, and architects competent to and adequately empowered to (a) establish and revise safety criteria for Bay fills and structures thereon, (b) re- view all except minor projects for the adequacy of their specific safety provisions, and make recommendations con- cerning these provisions, (c) prescribe an inspection system to assure placement of fill according to approved designs, and (d) gather, and make available, performance data de- B32 veloped from specific projects. These activities would com- plement the functions of local building departments and local planning departments, none of which are presently staffed to provide soils inspections. 2. Even if the Bay plan indicates that a fill may be permissible, no fill or building should be constructed if hazards cannot be overcome adequately for the intended use in accordance with the criteria prescribed by the Fill Review Board. 3. To provide vitally needed information on the effects of earth- quakes on all kinds of soils, installation of strong-motion seismographs should be requried in all future major land fills. In addition, the Bay agency should encourage installa- tion of strong-motion seismographs in other developments on problem soils, and in other areas recommended by the US. Coast and Geodetic Survey, for purposes of data com- parison and evaluation. The proposed Fill Review Board was established as the Engineering Criteria Review Board, composed of geologists, structural engineers, civil engineers, soils engineers, and other professionals as recommended in the plan. The board reviews and evaluates soils and geologic reports submitted by applicants for permits to fill. Significant improvement in the seismic engineer- ing of fills and design of structures has resulted from the board’s insistence on a thorough evaluation of geologic hazards at a project site (San Francisco Bay Conservation and Development Commission, 1974a, p. 8). The policies and review procedures incorporated into the bay plan provide the means to assure that devel- opment within BCDC’s jurisdiction is carried out in accordance with an acceptable degree of risk. However, as stated in a report of the Bay Plan Evaluation Project (San Francisco Bay Conservation and Development Commission, 1974b, p. 19) a “more precise definition of what level of risk is acceptable for design of structures on the Bay” is needed. Such a definition requires a detailed risk analysis involving both seismic and non- seismic hazards. CALIFORNIA COASTAL ZONE CONSERVATION COMMISSION The CCZCC (California Coastal Zone Conservation Commission) and subordinate regional commissions were created by State legislation adopted, by initiative, in 1972. The CCZCC, working with the six regional commissions, prepared a plan for the future of the California coastal zone. While the plan was being pre- pared, the commissions controlled all development, through a permit process, to ensure consistency with the objectives of the establishing legislation and the emerging plan policies. Coastal areas of the bay region are represented by two regional commissions: Central (San Mateo County) and North Central (San Francisco, Marin, and Sonoma Counties). The California Coastal Plan was adopted by the CCZCC in September, 1975, and forwarded to the Gov- ernor and State Legislature in December, 1975. In REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 1976, the California Coastal Act was enacted, estab- lishing the policies and governmental mechanism for ensuring wise use of the State’s coastal areas. The act requires local governments within the coastal zone to adopt local coastal programs to implement the policies of the Coastal Act. It is stated in one policy that new development shall “minimize risks to life and property in areas of high geologic, flood, and fire hazard,” and shall "assure stability and structural integrity, and neither create nor contribute significantly to erosion, geologic instability, or destruction of the site or sur- rounding aream” (California Public Resources Code, 1976, Sec. 30253). Local coastal programs are to be submitted to the appropriate regional commission for certification. After the local coastal programs in a region have been certified, or by January 1, 1981, the regional commis- sions are to be disbanded. The State Coastal Commis- sion is to designate sensitive coastal resource areas which require special protection in local coastal pro- grams. To remain in force after two years, the des— ignations must be affirmed by the State Legislature. Permits for specific coastal zone developments will be required, and the Coastal Act establishes proce- dures to be followed before and after a local coastal program is certified. Guidelines for preparing local coastal programs will be issued by the State Coastal Commission in Spring 1977. In keeping with the policy framework in the act, full consideration of seismic and other geologic hazards is likely to be required of local programs. METROPOLITAN TRANSPORTATION COM MISSION The MTC (Metropolitan Transportation Commis- sion) was created to coordinate development of regional transportation facilities. It was charged with preparing and adopting a Regional Transportation Plan dealing with major highways, mass transit, transbay bridges, airports, and harbors. It must also develop a transportation improvement program and a financial program for carrying it out. MTC also cooperates with ABAG in the A—95 review process by providing comments related to transporta- tion. MTC’s approval is required for certain projects including transbay bridges, public multicounty transit systems on exclusive rights-of-way, all applications from local governments or districts for State or Federal funds related to transportation, and construction of the State Highway System. In addition to the project- review function, MTC administers the public transit funds derived from State and local sales taxes on gasoline for the nine bay region counties. ' MTC adopted the Regional Transportation Plan in 1973 and revisions on several subsequent occasions. SEISMIC SAFETY AND LAND-USE PLANNING With respect to seismic safety, the plan recognizes the importance of the transportation system to postearth- quake evacuation, rescue, and relief efforts. Accord- ingly, the following policy was adopted: "Earthquake and seismic technology shall be used in the planning, location and construction of new transportation facilities” (Metropolitan Transportation Commission, 1974, p. 14). This policy ensures that seismic hazards are considered in the review of transportation projects undertaken by MTC. EVALUATING SEISMIC RISK Seismic safety planning is the process of evaluating seismic risk and formulating public policy to reduce that risk. Methods of evaluating seismic risk are de- scribed in this section; formulating public policy is dis- cussed in the following section. To evaluate risk, it is necessary to understand the distinction between hazard and risk. A seismic hazard is an effect of an earthquake such as surface faulting, ground shaking, a tsunami, liquefaction, landsliding, and other forms of ground failure. Seismic risk is the exposure of individ- uals and structures to potential injury or damage from seismic hazards. For example, the presence of an active fault is clearly a hazard; however, the degree of risk depends on the location, type of construction, and occu- pancy of structures with respect to the fault. Given present knowledge of seismic phenomena, little can be done to modify the hazard, that is, control tectonic pro- cesses, but much can be done to control risk or exposure to seismic hazards. This is the purpose of seismic safety planning. Risk evaluation consists of: evaluating seismic hazards, and assessing the degree of exposure of indi- viduals and structures to those hazards. The tech- niques and specifics of the evaluation may differ, but the basic procedure is becoming fairly well established. Figure 14 outlines the usual steps in evaluating risk from seismic hazards. Each step is described briefly in the following sections. IDENTIFYING SEISMIC HAZARDS The first step in evaluating risk in any area is to determine the potential for damaging earthquakes by reviewing the seismic history of the area and identify- ing any active or potentially active faults. Such faults are identified from historic, geologic, or seismic evi- dence of surface displacement (Borcherdt, 1975, p. A5). Evaluating earthquake potential or seismicity of an area requires information concerning: (1) the location of faults capable of generating damaging earth- quakes, (2) the magnitude of earthquakes anticipated on these faults, (3) the amount of fault displacement anticipated, (4) the na- ture and areal distribution of deformation accompanying earth— B33 quakes or fault movement; and (5) the frequency of recurrence of earthquakes on a known fault (Borcherdt, 1975, p. A29). Work done in the bay region provides an example of the necessary first step in evaluating regional seismic-- ity. Here some 30 faults have been identified as being active or potentially active and therefore potentially capable of producing damaging earthquakes (Bor- cherdt, 1975, p. A30). These faults have been mapped at scales ranging from 1:250,000 to 124,000 and their earthquake potential evaluated. These data provide a detailed description of the seismicity of the bay area (Borcherdt, 1975, fig. 3 and table 1). As in this exam- ple, where damaging earthquakes can be expected, the various seismic hazards need to be identified and eval- uated. The following discussion of individual seismic hazards in the San Francisco Bay region illustrates this process. The discussion is based almost entirely on the technical data in Part A of this report, and major topics are keyed to pages in Part A (Borcherdt, 1975). SURFACE RUP'TURE (A6—A12; A25—A30) Faults which have displaced the surface of the earth in the recent geologic past can be expected to do so again and are classed as active or potentially active. Not all earthquakes result in surface rupture and, in any one earthquake, surface rupture is unlikely to occur along the full length of a major fault. Also the likelihood and amount of potential surface displace- ment vary for different faults and even for different segments of the same fault. However, because even small vertical or horizontal displacements can severely damage structures astride a fault, planners should consider rupture a hazard on all the identified active or potentially active faults in the San Francisco Bay re- gion. Special geologic investigations to determine the nature and amount of anticipated displacement are needed to locate and design those utility lines and other lifelines which must cross a fault. Surface rupture along active faults may also result from fault creep—a process consisting of slow, inter- mittent, or fairly continuous fault movement which can amount to as much as an inch per year. It is usually recognized by surface evidence such as offsets and breaks in curbs, sidewalks, streets, fences, and Select a desrgn earthquake Identify seismic hazards Predict QBOIOQVC efiscts (seismic zonation) Assess Judge seismic acceptable risk risk Inventory T cultural features FIGURE 14.—Steps in evaluating seismic risk. B34 other structures. The presence of creep and its rate can be verified by installing and monitoring instruments along the fault. Another aspect of evaluating hazard from surface rupture is defining the width of the zone of surface deformation associated with fault displacement (Bor- cherdt, 1975, p. A25). Precise delineation of this zone may require extensive subsurface investigation (usually including trenching) to locate all active traces and other evidences of surface deformation. Such in- formation is seldom available, so the width Of the zone is commonly estimated from geologic evidence or from historic records. Widths of zones of deformation are discussed (Bor- cherdt, 1975, p. A25) in these terms: Until proved otherwise by geologic site investigations, prudence sug- gests zone widths of 184 m (600 ft) for the largest strike-slip faults and 1,800 m (6,000 ft) for the largest dip-slip faults. In the San Francisco Bay region, most dip-slip faults are relatively short (less than 16 km or 10 mi), and for these, narrower zone widths are appro- priate. GROUND SHAKING (A52—A57) Ground shaking is a major cause of earthquake damage. The severity of shaking depends on the mag- nitude and type of movement, distance from the fault, and local geology. The most violent ground shaking generally occurs in a fairly narrow band adjacent to the fault and the intensity of shaking tends to decrease with distance from the fault, but local geologic condi- tions may modify this pattern. That unconsolidated sedimentary deposits may amplify bedrock motion and produce strong ground shaking far from the fault is evident in the damage patterns of many major earth- quakes. Borcherdt (1975, p. A64) states: ***The effects of ground shaking are expected to be least for sites underlain by bedrock, intermediate for those sites underlain by al- luvium, and greatest for those sites underlain by artificial fill and bay mud. In the bay region, the relative potential ground shaking can be estimated from the damage patterns of the 1906 earthquake, from empirical studies of amplification of bedrock motion in different earth ma- terials, and from the predicted relationship between distance from the fault and intensity. Current esti- mates of the potential for shaking are given by Bor- cherdt, Gibbs, and Lajoie (1975). LIQUEFACTION (A68—A74) Liquefaction is the transformation of a loose, water- saturated, granular material (such as sand) from a solid to a liquid state. It can be caused by ground shak- ing and may in turn cause major ground failure. The relative potential for liquefaction in the southern San Francisco Bay region was mapped (Borcherdt, 1975, p. A70) using the following criteria: REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION The liquefaction-potential criteria can now be summarized as fol- lows: Saturated clay-free granular sediments with relative densities less than 65 percent are considered to have high liquefaction poten- tial, even in a moderate earthquake; clay-free granular sediments with relative densities greater than 90 percent are considered to have low liquefaction potential, even in a major earthquake; and saturated clay-free granular sediments with relative densities be- tween 65 and 90 percent have moderate liquefaction potential that depends on intensity and duration of ground shaking and textural properties of the sediments. “Potential” is the key word here. For liquefaction to occur, liquefiable materials must be ’within about 30 meters (100 feet) of the surface, saturated, and sub- jected to strong ground shaking. In addition, ground failure from liquefaction occurs only if the liquefied materials are not confined. For some geologic units, like bay mud, geologic site investigations are neces- sary to determine that a particular site is not under- lain by liquefiable materials. LANDSLIDING (A75—A87) Earthquakes may trigger many landslides, particu- larly during the wet season. The potential for landslid- ing is a function of basic slope stability and is highest in unconsolidated, soft sediments or surficial deposits; on steep slopes; where seasonal rainfall is high; where vegetation is shallow rooted or sparse; where erosion rates are high; and where ground shaking is intense. Maps showing relative slope stability for the entire San Francisco Bay region are available at a scale of 1:125,000 (Nilsen and Wright, 1979). These maps evaluate relative landslide potential on an areawide basis. Although they do not predict which landslides will move in an earthquake, they do show those areas in which landslides are most likely. Geologic site in- vestigation isneeded to pinpoint the landslide poten- tial within these areas. FLOODING (A93—A94) Earthquakes may cause flooding from tsunamis and seiches. The susceptibility of a coastal area to tsunami damage depends on local topography and elevation with respect to the potential size and direction of in— coming waves. Potential tsunami runup areas can be generally delineated, and are often based on a maximum probable event. Potential runup areas in the San Francisco Bay region are mapped by Ritter and Dupre (1972). The hazard from seiches is more difficult to evaluate. Generally speaking, any area adjacent to a large reser- voir, lake, or other enclosed body of water is susceptible to flooding from seiches. Overtopping of dams Or ‘ shoreline flooding can also be generated by landslides falling into bodies of water. The potential depends on the location of unstable slopes with respect to lakes, reservoirs, and bays. SEISMIC SAFETY AND LAND-USE PLANNING Major flooding may be caused by the failure of dams or dikes during an earthquake. Areas around the southern San Francisco Bay are particularly susceptible to flooding from dike failure. Many dikes are built mostly of fine-grained sediments dredged from the bay and are located on deposits of bay mud. Such dikes are particularly prone to failure during an earthquake. The areas susceptible to flooding from dike failure vary with tidal level at the time of an earthquake, but they have increased in size over the years because of ground subsidence brought about by the withdrawal of ground water in the south bay area. In accordance with the Alquist Dam Safety Act (see section on "Critical facilities”), areas which would be flooded in the event of dam failures have been mapped throughout California. These areas are extensive in the San Francisco Bay region and are of significant concern because of their size and location. Studies are being made to identify more specifically the likelihood of individual dam failure in the event of a major earth- quake. It is also essential to evaluate the probable depth and velocity of flood waters and, where areas below dams are developed, the length of warning time residents may have. SELECTING THE DESIGN EARTHQUAKE The severity of seismic hazards is directly related to several characteristics of earthquakes. Information concerning possible earthquake magnitude and Idea- tion are needed to estimate the possible surface rupture, ground shaking, ground failure, and flooding in an area. The hypothetical earthquake that is used as the basis for assessing seismic effects is called the de- sign earthquake. Criteria for establishing the mag- nitude of the design earthquake are described below. Maximum earthquake magnitude and frequency can be estimated, based on: (1) the geologically determined rate of slip and historic records of ground deformation, (2) the seismic history of the fault and the sur- rounding tectonic regime, (3) geologic evaluation of the tectonic set- ting, and (4) the empirically derived relation between magnitude of earthquakes and fault length or other parameters (Borcherdt, 1975, p. A17). It is realistic to assume that the largest historic earth- quake can recur on the same fault or a geologically similar fault and that potential magnitude increases with fault length. Based on seismic history and fault- length relations, assuming that half the fault length would break in a maximum magnitude earthquake, the largest expected earthquake on the San Andreas fault is 8.5 on the Richter scale and, on the Hayward fault, 7.0—7.5 (Borcherdt, 1975, p. A10). A magnitude at, or close to, the maximum expected is usually chosen for the design earthquake in evaluat- B35 ing risk for planning purposes. Because earthquake ef- fects cannot yet be predicted in detail, a conservative approach based on the largest magnitude foreseen by competent geologists is prudent, especially when plan- ning for areas or structures with intensive use or for facilities which are critical to the safety and continued functioning and recovery of a community during and after an earthquake. Choice of size of the design earthquake is also influ- enced by projected frequency of occurrence. If an earth- quake of maximum magnitude can be expected to occur once every thousand years, for example, one of lesser magnitude may be reasonably chosen for the design earthquake. However, recurrence intervals, par- ticularly for major earthquakes, are difficult to deter- mine; the historic record is too short, and even careful geologic studies do not always clearly define recurrence intervals. Although magnitude and frequency appear to be re- lated linearly—small earthquakes occur more often than large ones (Chinnery and North 1975, p. 1198)— this relationship is a poor guide for evaluating risk, because in a given area the pattern is highly variable. Different segments of the same fault may behave dif- ferently. For example, the segment of the San Andreas fault running through Marin County and the San Francisco peninsula has been relatively quiet since 1906, whereas the same fault near Hollister is the source of frequent relatively small earthquakes. Fig- ure 15 shows the maximum magnitude, maximum strike slip (horizontal displacement), and recurrence interval estimated for different segments of the San Andreas fault (Wallace, 1970, p. 2881). The magnitude chosen for the design earthquake should not necessarily be the maximum magnitude that may be expected on the fault segment closest to the planning area. For example, the Hollister area may experience more damage from a magnitude 8 on the fault segment to the north than from a magnitude 6 on a closer segment of the fault. All faults and fault seg- ments near the planning area need to be carefully evaulated in selecting the design earthquake. Because damaging effects are related to the length of fault dis- placement which in turn is related to the length of the fault, the design earthquake is usually the maximum event expected on the largest active fault affecting an area. A design earthquake may represent the expected ef- fects of a single large earthquake or a series of earth- quakes of different magnitudes. The design earth- quake does not indicate the overall seismicity or susceptibility to damage» from lesser magnitude earth- quakes or from earthquakes on other nearby faults. However, if appropriate measures are taken to reduce B36 _____ 1 Cape Mendocino l L\ "1 '7 Angheles \ \- ( 4’ s 2,, 1) o 100 200 MILES ’ 2 0 100 200 300 KILOMETERS Behavior category W Maximum magnitude 7-8+ 6-7 5-6 <5 Maximum strike slip (m) 1.2—10 0.3-1.2 0.1-0.3 <0.1 Recurrence interval (yrs) 100-1000 10-100 <10 >50 FIGURE 15.—Behavior of different segments of San Andreas fault (Wallace, 1970, p. 2881). risk from the design earthquake, risk from lesser events are correspondingly reduced. PREDICTING GEOLOGIC EFFECTS Predicting the effects of a design earthquake in- volves relating information available on the seismic hazards to the design earthquake. Since the seismic hazards are closely related to geologic conditions, the evaluation depends on the level of detail and accuracy of the basic geologic mapping. Predicting the geologic effects of an earthquake in a given area is termed “seismic zonation”, which is defined as "the delineation of geographical areas with different potentials for sur- face faulting, ground shaking, flooding, liquefaction, and landsliding during future earthquakes of specific size and location” (Borcherdt, 1975, p. A1). Such studies are essential in evaluating risk for planning purposes. The geologic effects of a postulated earthquake of magnitude 6.5 on the San Andreas fault are predicted (Borcherdt, 1975, p. A88—A95) along a demonstration profile extending from Sky Londa west of the fault on the San Francisco peninsula across the bay to Coyote Hills (fig. 16). The moderate magnitude of 6.5 was cho- sen because reliable strong-motion data obtained within 50 km (31 mi) of the causative fault were avail~ REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION able for earthquakes of moderate size (magnitude 5.0—6.9) but not for magnitude 7.0 and larger earth- quakes. With such data it is possible to statistically predict ground-motion values for competent geologic materials (ranging from bedrock to firm alluvium) for sites at distances greater than 10—20 km (6—12 mi) from the causative fault (Borcherdt, 1975, p. A32). The maximum magnitude expected along this seg- ment of the fault is 8.5, and a magnitude closer to this would usually be used for the design earthquake for planning in the area of the demonstration profile. However, the methods used to predict the geologic ef- fects of a 6.5 M earthquake are the same as for an 8.5, and any ranking of geographic areas on the basis of relative hazard would be the same. An 8.5 M earth- quake would cause more severe effects over a larger area than a 6.5. Figure 17 shows the predicted geologic effects of the postulated 6.5 M earthquake along the demonstration profile. The method described can be applied to other large areas in the San Francisco Bay region where comparable geologic information is available. This in- formation was translated into a series of hazard maps for use in assessing seismic risk by extending the sev- eral geologic effects to areas with similar underlying geologic material in an area roughly centered along the profile. ' Figure 18 shows the generalized geology of the area crossed by the demonstration profile and zones of po— tential surface deformation for the postulated earth- quake. A surface-rupture length of 40 km (25 mi) plus or minus about 10 km. (6 mi) is postulated on the San Andreas fault (see fig. 16). Displacement on the San Andreas of about 1 m (3 ft) is estimated. The zone of potential surface deformation can vary in width from a few meters to a few tens of meters. The hatched lines in figure 18 showing deformation zones are not to scale; they simply indicate that surface deformation is not necessarily confined to the line depicting the fault loca- tion. The zone of predicted surface deformation should be considered highly hazardous in a risk evaluation. Structures within it could experience severe damage from displacement of the ground and from intense shaking; however, detailed investigations may reveal sites within the zone which can accommodate struc- tures with acceptable safety. GROUND SHAKING Predicting relative severity of ground shaking is one of the most difficult tasks of seismic zonation. Two steps are involved in estimating relative ground shak- ing at the surface: predicting bedrock shaking, and predicting amplification of bedrock shaking in uncon- solidated deposits. Shaking was predicted for four sites SEISMIC SAFETY AND LAND-USE PLANNING B37 122°45' 15, 122°00’ 37°45’ — PA CIFI C OCEAN 30! _ EXPLANATION Estimated surface rupture for postulated earthquake (magnitude 6.5) 15' — \\ Fault trace Demonstration profile A Site for which ground shaking was calculated for postulated earthquake (magnitude 6.5) 0 5 10 MILES l_'_—_|—_l—l——J 0 5 10 KILOMETERS I 370%] ._ FIGURE 16.—Location of demonstration profile and estimated length of surface rupture associated with a postulated earthquake of magnitude 6.5 on the San Andreas fault, southwestern San Francisco Bay region (Borcherdt, 1975, p. A89). REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION B38 mmoh. 0%380 .EEBME Sufism marmimmis mo ab 95 so 959% bums—8w Eoubw exwzvfibwv 132 .3m< d .mp3 522.838 2%“: .m .M ma vflfifioo ~55on «.95 Q5 mo bigwm 25. .flwom 3:2me « no 555:8 omEmEm .8 fin m5 mo 33m Ego 05 >3 38¢:th €wa 2.5 3 gunman”. can 95:13:: mmwfiofiu a: 3353?”? Emflvfi mm “880 oxmsnfifimw none mo Eigwm 25. A839: confidm mo 5» :3 gmfiflmm was wEoE cosmhumaofimw Mo 5582 .8.“ m: .wc womv 25mm mafia: :mm 9: 5 am 035 MOZ<._.m_D mePmZZOJ _v_ ON _ _ _ _ _ _ _ _ _ _ _ _ . _ q _ _ . _ _ _ _ _ _ _ 1 mm.:_>_ ma Emacs mxmsvnfimo gflamonmmo 88am uRoEmE vBflvngHH $505 3 : £33 “335 II , 7‘ 00m $5M Smwvxgk :am. i333 NI ‘NOIiVAI-I'IEI 2:: 228 00ml L I‘ll, cozuflmzcz 5:5 owgfioo‘mwm mcfimmam .823 :moawv mum—mccmq / msfi...,fl.. . H fieugbafifi ., Ea ififiwfiéum D mfiwommu 252565 3.8w 92.. 382E B 532:? mcooSmE—m 3mg g Eng—E w:8o_o= g was. mam I ZO_._.32 2’7 .903 Q [l 2;?! " ‘\\\ \‘X lg" 2222. “as? AlRPORT 4/ \ NW DARK 2’!" \. o O ‘1 i- ." fl' I ' I / ‘C/frl/Q/Q 47%{2164/ . 222%/2’2/ ”/22" 1’" zéé7l/Z/f/figgfi ,/ ~/ 1 ll, .2 7”?“ 2222/ 222/ a» I, I 1 2 '9 1.3 %I% l' ' ’1 {ll/,2??? 1W / .2 . «xegggflggg $3 2. :e\ .2 0‘ ®N§w§ ffifi’ , ./ ' ‘ ' . ‘1' 1/ 2.2 , 2 2 22.; 2 ~- 22w .. * v. t ,/2/’ - ,2 226/7 ”WW/g," / . . /.. in, 42" f6‘\‘?’ 94", 5 /¢ . , 4 - ;:.Ev /. 2» . ”42.22%” 7 ’ 2 /.2 a A E: 5 2 :2. {/xfia i2 I, rigg/é? \ '-'-'-' 222/ 22/2 flaw/1.41 ./// A If!!! 4441/4”): , ' EXPLANATION High potential for liquefaction and Low potential for liquefaction and 0 1 2 3 MILES lateral spreading where clay-free lateral spreading where clay-free granular layers are present granular layers are present 0 1 2 3 4 K] LOMETE RS Moderate potential for liquefaction % Land areas underlain by bedrock with and lateral spreading where clay- ‘ little potential for liquefaction free granular layers are present FIGURE 20.—Relative potential for liquefaction and lateral spreading from the postulated earthquake. SEISMIC SAFETY AND LAND-USE PLANNING 122°15' 122°07'30" oster City 37° 30' o 37 . 22. ' 30" EXPLANATION Stable Generally stable 0 1 2 3 4 MI LES L l I I | I I | - Unstable D I I I If 6 KILOMETERS Generally stable to marginally stable FIGURE 21,—Re1ative stability of upland slopes. B43 B44 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION 122°15' I \ I I '{Faster my“ 7 EXPLANATION - Partial flooding possible FIGURE 22.—Possible flooding throughout the bay mud unit from dike failure. When an area can be assigned to more than one hazard zone, the highest hazard zone is shown. The resulting map (fig. 23) shows relative hazards from all the seismic effects considered. It does not show which hazard dominates at a given location. Because differ- ent measures are effective in mitigating different seismic hazards, maps showing the relative severity of 122° 07'30" dthaint 1 2 3 4 5 MILES I I I I l I I I I | 1 2 3 4 5 6 7 KILOMETERS Oq—O 0 individual hazards are almost always needed to formu- late plans for reducing seismic risk._ The composite seismic hazard zone map provides a general overview of relative hazards from the postulated earthquake. Preparing a composite map identifies areas with multiple hazards and areas where, because the hazard is greatest, the most severe damage from SEISMIC SAFETY AND LAND-USE PLANNING B45 122°15' "I2 1: Ilia/”git '1’,- 1:3! \ - (Q: h 5 i by salt a: 91.4, . 5"! v u‘ . ,1 ‘1’ .A’zétizm 1 2 3 4 5 MILES l | l l l I l l 1 2 3 4 5 6 7 KILOMETERS Z O O. m -. w .-. m 3' m N n: -. 0. 0—70 . FIGURE 23.—Seismic hazard zones. the earthquake can be expected. These areas should be carefully investigated before making land-devel- opment decisions and given particular consideration when planning for emergencies. Such a map cannot substitute for detailed investigations, but it can indi- cate where such investigations are most needed. The seismic hazards zone map is only as accurate as its elements. Because ground shaking is difficult to pre- dict regionally, it is less accurately reflected in the composite map than other seismic hazards. As pre- B46 sented here, the map shows relative differences in amplification of bedrock motion in unconsolidated de- posits, but it does not indicate differences in bedrock motion. Areas near the fault shown in the low hazard zone may, because of strong bedrock shaking, be more hazardous than indicated. Another approach to seismic zonation, which partly avoids this problem, was suggested by Borcherdt, Gibbs, and Lajoie (1975), whose map of the South San Francisco Bay region (scale 1:125,000), shows maximum earthquake intensity for a large (7.5M— 8.3M) earthquake on the San Andreas or Hayward fault. Intensities of the 1906 San Francisco earthquake are related to geologic units and distance from the fault to predict the maximum intensity throughout the re- gion. Intensity is expressed in terms of the San Fran- cisco Intensity Scale (table 7) developed by H. 0. Wood (1908, p. 224, 225) to describe the 1906 San Francisco earthquake. Figure 24 shows the maximum intensities predicted for the area shown in figure 23. It is not possible to tell from this map which haZards are present in a given location. High intensities may be from strong ground shaking, ground failure, or some other hazard. How- ever, the intensity map fairly accurately depicts relative ground shaking from a large earthquake. Comparing this map with the seismic hazard zone map indicates that the intensity map probably understates landsliding and flood potential. Hazard evaluation for risk assessment depends, first of all, on the detail and accuracy of the geologic infor- mation available for an area, but it also depends on the purpose of the evaluation, the size and diversity of the planning area, and the power of the agency undertak- ing the evaluation. If geologic maps are highly gen- eralized, hazard evaluation must, of necessity, first focus on identifying those areas most likely to be hazardous, then further data can be collected and ap- propriate geologic information submitted with any major development proposals. Better hazard evalua- tions can be made as more detailed geologic data be- come available. If the hazard evaluation is to be used primarily for earthquake preparedness, hazard zones based on in- tensity may be appropriate. For such purposes, it is ' more important to know the expected level of damage than the exact cause of the damage. If, on the other hand, the evaluation is to be used primarily to prepare land-use plans and regulations and to suggest meas- ures to reduce seismic risk, it is important to evaluate each seismic hazard individually and in as much detail as the information permits. Hazard evaluation for a geologically homogeneous region, regardless of size, can be quite generalized. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION TABLE 7.——San Francisco Intensity Scale for 1906 Earthquake [From Borcherdt, 1975, p. A31 Grade Intensity Description A Very Violent The rendin and shearing of rock masses, earth, tu , and all structures along the line of faulting; the fall of rock from mountainsides; many landslips of great magnitude; consistent, deep, and extended fissurin in natural earth; some structures totally estroyed. B Violent Fairly general collapse of brick and frame buildings when not unusually strong; serious cracking of brickwork and masonry in excel- lent structures; the formation of fissures, step faults, sha compression anticlines, and broad, wave ike folds in paved and asphalt- coated streets, accompanied by the ragged fissuring of asphalt; the destruction of foun- dation walls and underpinning structures by the undulation of the ground; the breakin of sewers and water mains; the lateral is- placement of streets; and the compression, distension, and lateral waving or displace- ment of well-ballasted streetcar tracks. C Very Strong Brickwork and masonry badly cracked, with occasional collapse; some brick and masonry gables thrown down; frame buildings lurched or listed on fair or weak underpinning struc- tures, with occasional falling from underpin- ning or collapse; general destruction of chim- neys and of masonry, brick, or cement veneers; considerable cracking or crushing of foundation walls. D Strong General but not universal fall of chimneys; cracks in masonry and brickwork; cracks in foundation walls, retainin walls, and curb- ing; a few isolated cases of urching or listing of frame buildings built upon weak under- pinning structures. E Weak Occasional fall of chimneys and damage to plaster, partitions, plumbing, and the like. More detail is needed to evaluate seismic hazards in geologically diverse areas where hazard potential will vary significantly throughout the area. The power of the agency evaluating seismic hazards also affects the scope and detail of the effort. For exam- ple, a regional council of governments, with powers limited to planning and reviewing applications of local governments for Federal funds, may find generalized hazard evaluation sufficient for framing broad policies and determining what information should be submit-‘ ted with applications for funds. Local governments, on the other hand, need a more detailed evaluation as basis for land-use plans, land-development regula- tions, project-review criteria and procedures, building-code requirements, plans for public facilities and emergency responses. INVENTORYING CULTURAL FEATURES Evaluating seismic hazards is only part of assessing seismic risk. The other part is assessing the vulnerabil- ity of land uses and building occupancies to earthquake SEISMIC SAFETY AND LAND-USE PLANNING B47 122° 15' 122° 073°" l x I EXPLANATION 1 2 3 4 5 MILES l Very violent Very strong - Violent l:' Strong FIGURE 24.—Maximum earthquake intensity predicted on a regional scale. damage. The next step in assessing seismic riVSk in the used to determine the exposure of structures and area crossed by the demonstration profile would be to people to damage or death and injury from an earth- , inventory cultural or manmade features. quake. This information, considered in relation to the indi- Because of limitations of time and budget and the vidual and composite seismic hazards maps, would be fact that the demonstration profile crosses several 0“”‘0’ 1 2 .3 4 5 6 7 KILOMETERS B48 jurisdictions, an inventory of cultural features was not undertaken as part of this study. Instead, the key ele- ments of such an inventory are first described in gen— eral terms, which are then followed by examples of various methods of combining hazard data with cul- tural data to assess seismic risk. Risk depends on the uses of land and buildings and on the ability of a public agency to respond to a disas- ter. Evaluating seismic risk thus requires an inventory of (1) current land use; (2) structures with high occu- pancy; (3) structures that are hazardous because of age or type of construction; and (4) critical facilities includ- ing lifelines, facilities or structures needed for emergencies, and facilities and structures whose fail- ure would be catastrophic, such as dams or nuclear power plants. CURRENT LAND USE Maps or aerial photographs of current land use, viewed in conjunction with hazard maps, provide an overview of risk. Data concerning the number of dwell- ing units, rate of occupancy, location of businesses, and number of employees are used in estimating the day- time and nighttime population of specific areas. Land use, population density, and hazard maps should be prepared at the same scale. Map overlays are particu- larly useful. In populous areas, computer modelling may be justified. The distinction between land uses should be fine enough to separate structural types, heights, and intensity of use. For example, residential uses should be broken down at least into single-family and multiple-family categories and into height and structural categories. STRUCTURES WITH HIGH AND INVOLUNTARY OCCUPANCY High-occupancy structures such as large apartment buildings, office buildings, major employment and shopping centers, theaters, auditoriums, and stadiums should be identified and noted on the hazard map. Buildings with high involuntary occupancy such as hospitals, schools, prisons, and convalescent homes form a separate and particularly vulnerable group. Most discussions of risk distinguish between a risk that is voluntarily assumed, such as the choice of a home site, and a risk that is involuntary, such as being in school or jail. Presumably structures occupied in- voluntarily should be safer than those voluntarily oc- cupied. The distinction is especially important in those cases where public policy or laws require certain classes of people, such as prisoners or students, to oc- cupy structures which have fairly high occupancy. As a practical matter, only limited volition is possi- ble in choosing the structure in which to work, live, or REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION even spend leisure time. A choice of working in a rela- tively unsafe building or not working at all, or living in a structurally unsound house or leaving the area is often not a real choice. Most people fall somewhere between being voluntary or involuntary occupants of a building. In addition, information concerning the rela- tive safety of structures is often not known. The rela- tive safety of buildings is rarely considered by the job seeker, and the home seeker who would like a house designed as earthquake resistant may not be able to afford it. Similarly, determining what occupancy rate should be classed as "high” depends upon the character of the development area. A community with predomi- nantly low-density residential development might 10g- ically class two-story garden apartments as high occu- pancy in its risk evaluation. Conversely, a central city might apply that term only to structures more than 10 stories high. The location of structures with high or involuntary occupancy can be shown on the land-use map or sepa- rately, depending on the graphic methods used. Infor- mation concerning whether occupancy is for 24 hours, daytime, or nighttime should also be noted. HAZARDOUS STRUCTURES Structures built before seismic safety requirements were imposed by local building code or State law need to be identified, noted on the map, and evaluated to determine if they were constructed with unsafe mate- rials or methods. Particular attention should be given to masonry buildings. Also, poorly attached parapets, cornices, and other appendages should be noted. Fail- ure of a building or parts adjacent to a street may be hazardous not only to occupants but to passersby. This type of information may be available from the agency or department responsible for building inspection; if not, it will need to be obtained. LIFELINES Lifelines are the utility services and communication and transportation lines necessary for the continued functioning of the community. Water supply lines, gas lines, electric transmission lines, telephone lines, major highways,‘and railway lines should all be in- cluded on the maps. Related facilities such as tele- phone exchanges, water and natural gas storage areas, airports, harbors, bridges, highway interchange struc- tures, and power stations should also be identified. The location of shut—off valves, auxiliary suppliers, emer- gency power generators, and back—up communication systems should be noted where applicable. In addition, information on age, condition, and other factors will be needed in order to assess the likelihood of failure dur— ing an earthquake. SEISMIC SAFETY AND LAND-USE PLANNING FACILITIES FOR EMERGENCY RESPONSE The degree of risk partially depends on a communi- ty’s ability to respond to a disaster situation. Command and communication centers, hospitals, medical offices and supply centers, fire stations, and police stations should all be noted on the maps. Buildings such as schools, churches, and theaters which could be used to provide temporary shelter or centers for dispensing emergency aid should also be identified. Emergency resources may also be available from nearby communities under mutual aid agreements. However, because a major earthquake is likely to dam— age highways, communication lines, airports and other links to nearby communities, local facilities need to be sufficient to sustain a community until aid from out- side can be obtained. Areas within a community which could become isolated should also be identified and evaluated for an emergency situation. OTHER CRITICAL FACILITIES Structures whose destruction or damage could have catastrophic effects include nuclear power plants, large dams, and storage facilities for toxic materials. These and similar facilities in, or potentially affecting, a planning area should be identified, and the area of po- tential damage mapped. For example, areas subject to flooding from dam failure or dike collapse should be shown on the map. ASSESSING SEISMIC RISK When information describing seismic hazards, land use, and the structural and disaster response charac- teristics of a planning area has been assembled, deci- sions concerning the nature and degree of seismic risk can be made. Risk can be expressed in a variety of ways and with varying degrees of precision. DOLLAR LOSS Estimates of the dollar loss from all earthquakes over a period of time, or from a design earthquake, can be used to express seismic risk. Alfors, Burnett and Gay (1973) have estimated the total dollar loss from earthquake shaking, fault displacement, landsliding, tsunamis, and other natural hazards which can be ex- pected in California from 1970 to 2000 assuming current hazard mitigation practices. The total loss includes the cost of property damage, life-loss, injury, and intangi- ble loss. The analysis explicitly. relates hazard zones to population levels. Standard 71/2 minute topographic quadrangles were selected as unit cells and were as- signed to high, moderate, and low—hazard severity zones. The percent of population in each zone during the 30-year period is estimated, yielding an estimate of B49 person-years exposure from 1970 to 2000. This figure is multiplied by the expected average total loss per capita per year to attain the total loss figures. The calcula- tions for earthquake shaking are shown in table 8. Projected total losses, 1970—2000, are $76,000,000 from fault displacement, $9,852,000,000 from landslid- ing, and $40,800,000 from tsunamis. Similar estimates were made of losses from other natural hazards. Using dollars to express loss allows a comparison of total risk from several different hazards, including risk of life loss, injury, and property damage. This compari- son is important in assigning priorities for public action and expenditure. It also provides a basis for de- termining the benefits of various risk-reduction meas- ures if the costs of applying such measures are known. The Alfors study also estimates the cost of applying risk-reduction measures to arrive at a benefit/cost ratio for each hazard. Alfors, Burnett, and Gay (1973) assess Statewide risk in order to maximize the benefits of State actions to reduce risk from natural hazards. More specific studies of natural hazards would be needed if such an analysis were made at the regional or local level. But in many cases the time and expense necessary to do a thorough assessment in terms of dollar losses is not justified. In addition, lack of data and lack of ability to evaluate hazards may be significant barriers to risk assessment. DEATHS AND INJURIES Risk can also be expressed as the expected loss of life and injury from a hazard over a period of time, or from a single event such as an earthquake of specified mag- nitude. The risk may be stated as a total of the ex- pected deaths and injuries or as a rate per unit of popu- lation. Algermissen’s (1972) estimates of the expected loss of life and injury from an 8.3 magnitude earth- quake on either the San Andreas or Hayward fault were based on death and injury rates from historic earthquakes adjusted for types of structures and for daytime and nighttime conditions in the bay region. He estimated 2,300 deaths would result from damage or collapse of residential structures if the earthquake occurred at 2:30 am. when most people are at home. An additional 550 deaths would occur in hospitals if the earthquake were on the San Andreas fault, and 820 if it were on the Hayward. The largest number of deaths, 10,360, would occur with an 8.3 magnitude earthquake on the San Andreas fault at 4:30 pm. dur— ing the evening commute period. A comparable earth- quake at the same time of day on the Hayward fault would cause 6,650 deaths (table 9); deaths and injuries due to dam failure are not considered .in these esti- mates. B50 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION TABLE 8.—Projected total loss from earthquake shaking, 1970—2000 [Alfors and others 1973, p. 96 and Alfors, John, oral commun., 1977] Number of urban 7— I4-minute quadrangles of each Estimated percent of Estimated person»years exposure Geology Points (expectable total average loss rate) (dollars per capita Projected total loss Earthquake severity zone severity total population 1970—2000 per year) 1970—2000 High ________________________ 181 371/2 289,050,000 31 $ 8,961,000,000 Moderate ____________________ 242 50 385,400,000 27 10,406,000,000 Low ________________________ 57 121/2 96,350,000 14 1,349,000,000 Total ______________________________________________________ 770,800,000 $20,716,000,000 The potential number of deaths from dam failure is horrendous. Table 10 shows the number of people ex- posed to risk from the failure of major bay region dams and the maximum possible and probable deaths should a dam fail during the night or day. TABLE 9.—Deaths and hospitalized injuries [Adapted from Algermissen, 1972, p. 121] Total . . T‘ f T t l H 't lized Estlmates such as these can be used to express risk Magnitude 13:; Del;ls 01:23,,5 as a death rate per unit of population per unit of time. San Andreas fault ABAG (1975b, p.14) uses estimates derived from the 8.3 —————————————— 33388112- 3,228 {132,338 Algermissen study to calculate risk from an 8.3 mag- 42303111: 10360 40360 nitude earthquake on either fault. Assuming 6,600 7 0 2.30 am 500 1 900 deaths, a population of 5 million, and a recurrence ' """""""" 2200p1m: 1,640 6:200 interval of 170 years for an 8.3 magnitude event, the 4‘30 Pm- 1,990 11,680 following calculation is made: 61) ______________ 2:30 am 35 {£8 2:00 pm. 0 . 6,600 deaths 1 uake 4:30 pm. 100 390 Risk = . . . q 5 milhon population 170 years Hayward fault , , 8.3 ______________ 2:30 a.m. 3,120 11,600 = 7.7 deaths per year per 1 mlllion persons 2:00 p.m. 7,200 28,500 4:30 p.m. 6,650 24,900 Such calculations are useful primarily in comparing 7.0 ______________ 2:30 a.m. 1,040 3,860 the risk of death from earthquakes with other natural Egg 32$: 3:328 gfgg and man-made rlsks. For example, the actual rlsk of death from all causes is about one in 100 persons per 6'0 ------------- 3:38 3:3: 338 i338 year (ABAG, 1975b, p. 14). 4:30 pm. 700 2,550 POPULATION AT RISK One of the simplest ways to express risk is in terms of the number of people exposed to a hazard. This can be done simply by estimating the population of each delineated seismic hazard zone. Ayre (1975) uses the generalized seismic risk map for the United States (see fig. 4, p. 10) as the basis for such an estimate. The results, shown in table 11, indicate that approximately 31 million people (15 percent of the population) live in risk zone 3—the zone with the highest seismic risk. Over half of these people live in California (17,000,000). Assessment of this type is useful only to provide an overview of risk exposure. To develop risk- reduction policies and programs at the regional or local level, hazardous areas must be delineated more pre- cisely, and types of structures and patterns of occu- pancy must be studied in relation to the hazards. In order to study risk, the City of Palo Alto (1976) took a census of population density during the day and during the night (fig. 25). Buildings with high occu- pancy were then located on a map together with gen- eralized hazard zones. It became clear that buildings with high occupancy either during the day or night were located in zones of moderate or low hazard. The Palo Alto study (1976, p. 55) states: Measures to lessen risk to human life and property should focus upon identified areas of population concentration and be keyed to areas of greatest natural hazard and areas of known or suspected structural hazard. RELATIVE RISK Levels of seismic risk are commonly expressed in relative terms. The important point to remember is that a map of seismic hazard zones does not show expo- sure to hazards; information must include types of structures and occupancy characteristics. The term seismic risk map or zones is often used in a misleading sense. ‘ The report by the Tri-cities Seismic Safety and En- vironmental Resources Study (Armstrong, 1973), illus- SEISMIC SAFETY AND LAND-USE PLANNING B51 TABLE 10.—Life loss from dam failure [The figures represent the worst conditions as they may lfcun'entli" exist, assuming unsafe dams All of these dams are (or will be) re—evaluated for safety, and appropriate corrections will be mad e 1funsafe rom Algermissen, 1972, p 132. Asterisks indicate figures not available] Maximum Possible Maximum Possible Estimated Individuals Exposed Deaths Probable Deaths Dam Day Night Day Night Day Night Lafa ette ____________________________________ 95,000 91,000 11,000 7,000 7,000 5,000 San ablo, or Briones and San Pablo ______________________________ 49, 000 51,000 29,000 30,000 20,000 25,000 Upper San Leandro and Chabot __________________________________ 86, 000 109,000 50,000 52,000 30,000 35,000 Lower Crystal Springs ________________________ 61 ,000 57,000 30,000 31,000 20,000 25,000 Calaveras, James Turner, and Del Valle ________________________________________ 125,000 136,000 30, 000 34,000 20,000 24,000 Only Calaveras ________________________________ 35, 000 40,000 8 0,00 7,000 5,000 5,500 Only Del Valle ________________________________ 21, 000 24,000 15, ,000 19,000 10,000 13,000 Lexington ____________________________________ * 72,000 20,000 * 15,000 Andrson and Coyote __________________________ * 18,000 * 5,000 * 3,000 trates a means of evaluating risk for a given structure or proposed project. Each of the following four factors is assigned a high, medium, or low-risk rating: geology of the general area, geology of the site, structures, and building uses. The rating for the geology of the general area is determined from the risk zones shown by Algermissen (1979). A rating of local geologic condi- tions is obtained by using the best information avail- able on risk from active faults, slope stability, lique- faction, tsunamis, seiches, and ground shaking. Structural hazards are rated according to type of build- ing construction. Table 12 lists common structural types of buildings in order of increasing susceptibility to damage in an earthquake. Table 13 provides the basis for rating various building uses. The term “ordi- nary” risk in table 13 applies to structures which would: resist minor earthquakes without damage; re- sist moderate earthquakes without structural damage, but with some nonstructural damage; resist major earthquakes of the intensity of severity of the strongest experienced in California, without collapse, but with some structural as well as nonstructural damage. In most structures it is expected that structural damage, even in a major earthquake, could be limited to repair- able damage (Armstrong, 1973, p. 162). Using the four risk levels, the Tri-cities study gives several examples of risk assessment for particular uses of specific sites. Table 14 is an example of one such assessment for a motor inn. Note that this seemingly simple analysis depends on quite specific information concerning seis- mic hazards at the site, and the structural and use characteristics of the structure. The procedure is adaptable to assessing existing risk levels or changes in risk if changes In land use are proposed. Thus risk, although not quantified, is expressed in a way that can be applied directly to many planning decisions. SCENARIOS Another technique for expressing risk is through a scenario, a fictional but realistic description of a disas- ter and the chain of related events, told in the order in which they occur. Scenarios are especially effective in awakening pub- lic concern for seismic safety because they are dra- matic. A good scenario requires accurate, specific information and should clearly identify areas and structures of greatest risk. It can serve as a basis for developing policy, and, more important, it can dramatize the need for reducing risk. However risk is expressed, the assessment should be as clear and understandable as possible. Elaborate statistical studies of risk, although useful for some purposes, are not needed for most planning purposes. As stated in San Diego County’s general plan (San Diego County, 1975, p. v—2): The problem with quantitative approaches lies in the complexity of principles upon which risk judgments are made. Risk is a function of the underlying lithology of the site and its proximity to an earth- quake epicenter (sic) and varies with the use of the structure as well as the type, kind and quality of construction. Given those indepen- dent factors and economic and social impacts, the mechanical deter- mination of an arbitrary level of risk is too simplistic. Statistical expressions of risk based on historic rec- ords are particularly chancy when applied to in- frequent, catastrophic events, such as major earth- quakes, if the historic record is fairly short. One event greatly affects loss, injury, and death rates for years to come. Conversely, if no major earthquake has occurred during the period studied, an unduly optimistic picture of the degree of risk is conveyed. The foregoing examples of risk assessment evaluate existing risk levels, but risk assessment needs to be a flexible tool. It can also be used to assess the risks B52 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION TABLE 11.—U.S. population-at-risk by seismic risk zone and state [From Ayre, 1975] Estimated Population-at-Risk by Seismic Risk Zone State Total Population Zone 0 Zone 1 Zone 2 Zone 3 Alabama ________________________ 3,444,165 1,056,000 1,126,000 1,263,000 0 Alaska ____________________________ 300,382 0 6,000 25,000 270,000 Arizona ________________________ 1,722,482 0 0 1,742,000 30,000 Arkansas ______________________ 1,923,295 0 1,473,000 166,000 284,000 California ______________________ 19,953,134 0 0 2,636,000 17,317,000 Colorado ________________________ 2,207,259 0 2,207,000 0 0 Connecticut ____________________ 3,032,217 0 2,948,000 85,000 0 Delaware ________________________ 548,104 0 548,000 0 0 Florida ________________________ 6,789,443 5,503,000 1,286,000 0 0 Georgia ________________________ 4,589,575 0 1,777,000 2,812,000 0 Hawaii ___________________________ 768,324 30,000 637,000 39000 63,000 Idaho ____________________________ 712,567 0 0 513,000 200,000 Illinois ________________________ 11,113,976 0 9,951,000 895,000 268,000 Indiana ________________________ 5,193,669 0 2,350,000 2,608,000 236,000 Iowa ____________________________ 2,825,041 0 2,825,000 0 0 Kansas ________________________ 2,249,071 0 1,907,000 342,000 0 Kentucky ______________________ 3,219,311 0 1,349,000 1,467,000 403,000 Louisiana ______________________ 3,643,180 0 3,643,000 0 0 Maine ____________________________ 993,663 0 318,000 675,000 0 Maryland ______________________ 3,922,399 0 3,734,000 189,000 0 Massachusetts __________________ 5,689,170 0 0 1,980,000 3,709,000 Michigan ______________________ 8,875,083 0 8,875,000 0 0 Minnesota ______________________ 3,805,069 0 3,805,000 0 0 Mississippi _____________________ 2,216,912 269,000 1,674,000 217,000 57,000 Missouri ________________________ 4,676,501 0 3,079,000 1,389,000 209,000 Montana ________________________ 694,409 0 240,000 313,000 142,000 Nebraska ______________________ 1,483,791 0 1,206,000 278,000 0 Nevada __________________________ 488,738 0 0 300,000 189,000 New Hampshire __________________ 737,681 0 0 738,000 0 New Jersey _______________________ 7,168,164. 0 7,168,000 0 0 New Mexico ____________________ 1,016,000 0 536,000 480,000 0 New York ______________________ 18,236,967 0 13,211,000 2,481,00 2,545,000 North Carolina __________________ 5,082,059 0 2,172,000 2,910,000 0 North Dakota ____________________ 617,761 0 618,000 0 0 Ohio __________________________ 10,652,017 0 7,863,000 2,789,000 0 Oklahoma ______________________ 2,559,253 0 2,399,000 160,000 0 Oregon ________________________ 2,091,385 0 539,000 1,539,000 13,000 Pennsylvania __________________ 11,7 93,909 0 11,347,000 183,000 264,000 Rhode Island ____________________ 946,725 0 84,000 863,000 0 South Carolina __________________ 2,590,516 0 0 1,577000 1,013,000 South Dakota ____________________ 665,507 0 666,000 0 0 Tennessee ______________________ 3,923,687 0 1,165,000 1,810,000 949,000 Texas ___________________________ 11,196,730 9,859,000 1,325,000 13,000 0 Utah __________________________ 1,059,273 0 40,000 48,000 972,000 Vermont ________________________ 444,330 0 0 444,000 0 Virginia ________________________ 4,648,494 0 2,435,000 2,213,000 0 Washington ____________________ 3,409,169 0 0 1,240,000 2,169,000 Washington, DC. ________________ 756,510 0 757,000 0 0 West Virginia __________________ 1,744,237 0 1,509,000 236,000 0 Wisconsin ______________________ 4,417,731 0 4,418,000 0 0 Wyoming ________________________ 332,416 0 308,000 19,000 5,000 Totals __________________ 203,223,000 16,717,000 115,091,000 40,442,000 30,973,000 (100%) (8%) (57%) (20%) (15%) inherent in proposed land use or occupancy change. Procedures for assessing risk should be designed to ac- commodate new information concerning seismic hazards, changes in structural conditions, occupancies, and other risk parameters. DETERMINING ACCEPTABLE RISKS Public actions to reduce risk involve at least an implicit determination of "acceptable risk.” Acceptable risk, from the point of View of the public agency, is that level of risk at which no governmental response is con- sidered necessary. Acceptable risk is rarely expressed in quantitative terms, but is embodied in the risk- reduction policies, regulations, and standards adopted by the public agency. Acceptable risk is a measure of willingness to incur costs to reduce risk. Aiming for a totally risk-free envi- ronment is unrealistic; some balance must be sought between risk and the costs of reducing it. The balance actually struck by a governmental agency represents its choice of an acceptable level of risk. The choice can be made explicitly by public bodies based on evaluation SEISMIC SAFETY AND LAND-USE PLANNING B53 2:00 P.M./1:00 AM. POPULATION ON WEEKDAYS, av CENSUS TRACT Census Day population mm mm popuunm : SO46 5093 5094 5106 5107 5108‘] 5105.2 5108.3 5109 5860 5l10 "00 6300 Sill 0 51l2 550011,100 51].? 5114 SHS Sll6 Sll7 \ \ mum 3“ UN VER we s - ‘ \\\ a \ . r / . Z , / , 4'0. ‘ . z - l . 4/7- ’ ‘ o ., 4 7 r "é o 1 var / ’/ r? ”44 PER 551?}? El. Q4 6* MI A p” JU / Y0 k Areas with buildings or . /l$ high densily occupancy ’90 Q? g Commercial, industrial. or ° ’///// if multrfamily with more than ‘ I0 units - Srhuol 280 - Buildings with 3 or more stories “wt $ Hospitals © Jail Reference dalz gathered in the field and horn building inspection -I- Plunning area boundary records by the Palo Alto Planning Deparlmem B COMPOSITE RISK ZONES EXPLANATION - Highrisk Moderate risk .5“ Low risk III Planning area boundary Reference: compilation by Palo Alto Planning Department of flood risk areas (HUD Flood Insurance Adminis- nation Dala) ground shaking, ground failure. and landslide hazard areas (USC-S data) FIGURE 25.—-Census of population density for Palo Alto, Calif. A, At 2:00 a.m., most people are in woodframe houses, structures which are likely to suffer the least damage from a great earthquake. At 2:00 pm, people are in school, at work, on the roads, and in commercial buildings which are areas of high population density and greater structural hazard. B, Areas of high population density. C, Fault line areas have the greatest risk for manmade structures. High risk is also associated with bay mud, landslide-prone hillsides, and areas susceptible to flooding. Moderate risk areas involve potential liquefaction, ground shaking, and some flooding. Low risk areas are susceptible to some liquefaction and some ground shaking. of the level of risk and the means and costs of reducing that risk. Such an evaluation is particularly relevant when a public agency is considering land-use plans and regulations, siting and design of major public facilities, renewal or rehabilitation of existing built-up areas, emergency-preparedness plans, and building-code re- quirements. LAND-USE PLANNING AND SEISMIC SAFETY Increasing seismic safety through land-use planning is at present primarily a function of local government, and local actions are central to reducing seismic risk. Most of the power to adopt and administer land—use and development regulations and building codes is now B54 TABLE 12. —Earthquake ratings for common building types [This table is not complete. Additional considerations would include parapets, building interiors, utilities, building orientation, and frequency response (Armstrong, 1973, p. 167)] Relative damageability (in order of increasing susceptibility to damage) Simplified description of structural types Small wood-frame structures, i.e., dwellings not over 3,000 sq. ft., and not over 3 stories Single or multistory steel-frame buildings with concrete exterior walls, concrete floors, and concrete roof. Moderate wall openings __________________ 1.5 Single or multistory reinforced-concrete buildings with concrete exterior walls, concrete floors, and concrete roof. __________________ 1 Moderate wall openin s ____________________________________ 2 Large area wood-frame uildings and other wood-frame buildings __________________________________ 3 to 4 Single or multistory steel-frame buildings with unreinforced masonr exterior wall panels; concrete floors an concrete roof ________________________________________________________ 4 Single or multistory reinforced-concrete frame buildings with unreinforced masonry exterior wall panels, concrete floors and concrete roof ____________________________________________ 5 Reinforced concrete bearing walls with supported floors and roof of any materials (usually wood) ______________________________________________ 5 Buildings with unreinforced brick masonry having sandlime mortar; and with supported floors and roof of any materials (usually wood) __________________________________________ 7 up Bearing walls of unreinforced adobe, un- reinforced hollow concrete block, or unreinforced hollow clay tile ______________ Collapse hazards in moderate shocks TABLE 13.—Scale of risks for various building uses [Adapted from Scale of acceptable Risks of the Structural Engineers Association of California, in Armstrong, 1973, p. 162] Level of risk to public Kinds of stmctures Failure of a single structure may affect substantial populations Extremely high ,,,,,, Structures whose continued functioning is critical, or whose failure might be catas- trophic: nuclear reactors, large dams power inter-tie systems, plants manufacturing explosives. High ________________ Structure whose use is critically needed after a disaster: important utility centers, hospi- tals, fire, police and emer ency communca- tion facilities, and critica transportation elements, such as bridges & overpasses; also smaller dams. Failure of a single structure will affect primarily only the occupants Possible high risk to occupants ________ Structures of high occupancy, or whose use after a disaster will be particularly conven- ient: schools, churches, theaters, large hotels and other high-rise buildings hous- ing large numbers of people, other places normally attracting large concentrations of people, civic buildings such as fire stations, secondary utility structures, extremely large commercial enterprises, most roads, alternate or noncritical bridge and over- passes. An “ordinary” level of risk ,,,,,,,,,,,,,, The vast majority of structures: most com- mercial and industrial buildings, small hotel and a artment buildings, and single—family residences. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION TABLE 14.—Risk analysis ofa motor inn [(From Armstrong, 1973, p. 174)] Factors Situation of Risk motor inn Geology ________________ Tri-Cities area, High several fault systems nearby. Site ____________________ On fault zone, High to perhaps directly very over fault trace. high landslide adjacent. Structure ______________ Multistory Medium utilities. Building ________________ Large number Medium use of occupants. to high Total risk for all factors _______________________ High lodged with local government. Also, the primary re- sponsibility for emergency response by police, fire, and public works agencies is local. Local areas may be iso- lated after an earthquake and depend entirely on local emergency services to protect life and property for a significant period of time before outside aid is avail- able. This section describes how local land-use plan- ning and decisionmaking can reduce seismic risk. THE PLANNING PROCESS Planning is the process of devising and carrying out a course of action to reach an objective. As an organized governmental activity, planning seeks to improve the decisions of public bodies and administrators. Com- prehensive planning affects the future development of an area and involves all major determinants of growth and change—economic, political, social, and physical. To be effective for seismic risk reduction, the com- prehensive planning process must result in specific land-use decisions. The process described here provides for a land-use plan as a key component of the com- prehensive plan; it forms a link between more general goals and policies and the pattern of land development. A land-use plan includes objectives, policies, and proposals for the type, pattern, and intensity of land use. It typically specifies the general location of differ- ent types of land uses, transportation lines, and public facilities. A functional plan defines needed facilities and operations for a specific function of government such as transportation, water development, flood con- trol, or emergency response; it is more specific than a comprehensive plan and usually covers a shorter period of time. Any plan, when adopted by the govern- ing body of an agency, becomes official public policy. The development of comprehensive, and functional land use plans generally consists of six steps: (1) iden- SEISMIC SAFETY AND LAND-USE PLANNING tifying problems and general goals and objectives, (2) collecting and interpreting data, (3) formulating plans, (4) evaluating impacts, (5) reviewing and adopting plans, and (6) implementing plans. These steps, shown in figure 26, are all interrelated. Plan formulation often indicates the need for additional information; ad- ditional information may reveal the need for additional information or modification of the plan. The steps in the planning process constitute a ra- tional, systematic approach to informed decisionmak- ing and are applicable to most governmental activities. The product is a logical and internally consistent plan, or set of plans and programs, to guide public and pri- vate decisions. The planning process is ongoing; it produces refinements, revisions, and new plans, and implements programs as additional information is ob- tained, new issues and problems are raised, or changes in public attitudes are recognized. Public participation is essential throughout the planning process. Success in implementing a plan depends on widespread public support which can only be gained if all major segments of the public participate in the planning process, and not always then. Decisions occur throughout the process, ranging from the decision to engage in a planning effort, to the final approval of a plan and adoption of implementing regulations, programs, and procedures. Elected public officials have the final responsibility for most key pol- icy decisions although persons in nonelective positions actually make many important day-to-day decisions. Where earthquakes are a recognized hazard, seismic safety is an important part of comprehensive land-use and functional planning. Comprehensive plans deal with the social, economic, and physical ramifications of seismic risk and methods of reducing it; and functional B55 plans with the procedures, actions, and resources needed for improving seismic safety in a given gov- ernment function, such as transportation, water supply, and fire protection. But because the degree of seismic risk depends on the location of structures and facilities in relation to seismic hazards, the land-use plan is a key document for expressing a community’s response to seismic risk; seismic safety can be ad- dressed in every step of the land-use planning process. IDENTIFY ISSUES AND DEFINE OBJECTIVES Reviewing available information to identify the major land-use issues and problems helps define the scope of the plans and the limits of the planning area. The issues and problems are then analyzed in relation to existing development, current land-use plans and policies, projected economic and population growth, and other anticipated changes. Based on this analysis, a tentative set of goals, objectives, and priorities is formulated. In this step potential seismicity is evalu- ated to identify seismic safety problems and define ob- jectives for reducing risk. Accounts of historic earth- quake damage in the area, or regional seismicity maps such as the Seismic Risk Map of the United States (fig. 4), can indicate the relative importance of considering seis- mic risk in land-use planning in a given area. COLLECT AND INTERPRET DATA Previously compiled data are evaluated for ade- quacy, and a program for acquiring and interpreting new data is prepared. The data needed include de- scriptions of the economic, social, cultural, political, and natural characteristics of the planning area as a basis for estimating the future requirements for spe- Public initiative and response I \JUU 1 2 3 4 5 6 IDENTIFY COLLECT FORMULATE EVALUATE REVIEW IMPLEMENT PROBLEMS AND PLANS IMPACTS AND PLANS AND DEFINE INTERPRET r ADOPT GOALS AND DATA PLANS OBJECTIVES Feedback for review and revision FIGURE 26.—The planning process. B56 cific kinds of land uses. Geologic and other information describing the natural features of the planning area is needed to describe natural hazards and resources rele- vant to land-use decisions. Because collecting and interpreting data is expen- sive and time-consuming, this study should be closely related to the issues and objectives of the planning pro- gram. A careful analysis of available general informa- tion can pinpoint those areas where more detailed data and analyses are needed. In some instances existing development or readily identifiable physical conditions may so restrict the land-use options that little new in- formation is needed. The degree of seismic risk is assessed during this step. Cooperation between planners and earth scien- tists, particularly seismologists and geologists and structural and soils engineers, is needed to identify and evaluate seismic hazards in relation to existing and potential land and building uses. FORMULA’I‘E PLANS Based on data and analyses gathered in the previous step, alternative policies, criteria, standards, and pro— posals are evaluated for responsiveness to the goals and objectives already identified. The environmental, social, political, and economic impacts of the policies and proposals are tested as a part of the process of selecting alternatives for presentation to the policy bodies. The evaluation of the alternatives is the heart of plan formulation. The plan should provide sufficient detail and definition to guide future growth and change in the planning area. The level of detail needed de- pends on the complexity of the planning area, the na- ture of the decisions anticipated, and the authority of the agency. Policies related to seismic hazards should be incorpo- rated into the comprehensive plan. As previously noted, California law provides for this by requiring all city and county general plans to include a seismic safety element. This element usually consists of a de- scription and evaluation of seismic hazards, together with policies and recommendations for improving seismic safety. These policies and recommendations should be reflected in the land use, open space, safety, circulation, and other general plan elements. The land-use/seismic risk relationships of alternative pol- icy options should be thoroughly considered in plan formulation. Policies and critieria for seismic safety should be specific enough to provide a basis for land use and development regulations and building code re- quirements. EVALUATE IMPACTS» The impact evaluation, begun as a part of plan for- mulation, is formalized in this step for the land-use REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION plan (or alternatives) selected for presentation to the policy body responsible for plan approval or adoption. If formal reports on environmental impact are required, by State or Federal law, the scope of the report is well defined. The method of analysis may be either quan- titative or qualitative. California law requires an environmental impact report for any plan or plan ele- ment which might have significant environmental im- pact. The environmental, social, and economic impacts of proposed measures to reduce risk should also be evalu- ated. The loss of life and the amount of damage from an earthquake depends on the land-use pattern. To aid in the decisionmaking, the levels of risk associated with alternative land-use patterns should be stated as ex- plicitly as possible. REVIEW AND ADOPT A PLAN The plan or plan alternatives and the report evaluat- ing impact are reviewed by the legislative body of the planning jurisdiction. Public hearings are scheduled and publicized to encourage the widest possible public response to the plan proposals. In California, cities and counties are required to hold public hearings before official adoption of any element of the general plan. Although comment and criticism by individuals and institutions is sought earlier in the planning process, new questions and issues are often raised at this point and plan policies and proposals may be modified as a result. After plan review, the plan may be adopted as an official statement of policy and a commitment to a fu- ture course of action. All states do not require that local general plans be adopted officially as does California. However, because many Federal grant pro- grams require formally adopted plans as a condition of eligibility, the practice is expected to become more widespread. Measures needed to implement the plan must be specified and clearly understood by the legis- lators and their constituents and, because circum- stances and priorities change, procedures for amending an adopted plan also should be established. IMPLEMENT THE PLAN Implementing a land-use plan depends on coordinat- ing, scheduling, and carrying out a variety of mea- sures. Land-use regulation alone may fail to reduce seismic risk significantly, particularly if the hazard exists primarily in developed areas. Therefore, regula- tions need to be combined with other measures such as public acquisition of land, urban renewal, redevelop- ment, and code enforcement. Implementing a plan is an intensely political process, and it directly affects the legal rights, economic and social status, and living and working environment of individuals in a community. Im- SEISMIC SAFETY AND LAND-USE PLANNING plementation thus depends on the active support of res- idents and organizations within a planning jurisdiction. The powers to implement plans and the methods used to implement them are different for each state and for each jurisdictional level. For example, most regional agencies rely on a project review process, while local governments have broad powers to regulate land use and development, tax, acquire land, and con- struct and operate facilites. Despite this diversity, most implementing measures fall logically into one of three categories: 1. Controlling land use and development through zoning, subdivision, and grading ordinances, and building and housing codes; 2. Reviewing projects, both public and private, for conformity with an approved general plan and for environmental impact pursuant to State or Federal laws or regulations (such as the U.S. National Environmental Policy Act of 1969 or the U.S. Office of Management and Budget Circular A—95); and 3. Developing and executing governmental pro- grams for acquiring land, constructing public facilities, providing public services, or rede- veloping and rehabilitating substandard parts of the community. PLANNING EXAMPLES Examples of plans, regulations, and administrative procedures to reduce seismic risk are described below. These examples are drawn mainly from cities and counties in California. Because the State requires each city and county to adopt a seismic safety element as part of its general plan, many examples of local plans to reduce seismic risk are found in California and illus- trate methods that can be used elsewhere. Most of the examples are from the San Francisco Bay region be- cause it has an array of earthquake hazards, much sci- entific data that can be used for planning, and a diverse assortment of planning agencies. Planning to reduce seismic risk varies for different earthquake hazards. Planning to reduce risk from sur- face rupture is reasonably straightforward where ac- tive faults are recognized and mapped, as they are in California. On the other hand, dealing with problems of ground shaking is more difficult because few maps of potential levels of ground shaking are available. Re- ducing risk requires adjusting both land uses and structural types to the anticipated intensity, fre- quency, and duration of shaking. In addition, areas of severe ground shaking are usually more extensive than the fairly narrow bands which are subject to sur- face rupture. Methods to reduce seismic risk also depend on the degree of development of a planning area. Land-use B57 planning to reduce seismic risk is most effective and least costly in areas just being considered for develop- ment. PLANNING TO REDUCE RISK FROM GROUND SHAKING To reduce risk from ground shaking requires the col- laboration of planners, structural engineers, and earth scientists. Planners provide information about the present and future locations of high-occupancy struc- tures, critical facilities, and hazardous structures, and the planning techniques available to control the future development. Structural engineers provide advice re- garding criteria for safe building design, the safety of existing structures, and techniques to reduce existing structural hazards. Earth scientists evaluate the ground response characteristics of a postulated earthquake. Seismic safety objectives are most likely to be at- tained when they coincide with other planning objec- tives. For example, preserving the margins of San Francisco Bay for ecological and environmental rea- sons is consistent with seismic safety objectives be- cause these areas, underlain primarily by bay mud, are susceptible to severe ground shaking and several forms of ground failure. Land-use regulations are likely to receive wider public support and withstand legal chal- lenge better if they meet more than one objective. The result of such joint efforts can be a land-use plan in which differences in expected ground shaking levels influence the location and intensity of proposed future development. This matching can only be done in areas where there is a significant variation in predicted levels of ground shaking. Even then, it should be rec- ognized that land-use decisions properly reflect eco- nomic, social, and political, as well as other natural conditions; a perfect match of land uses and seismic risk is rarely achieved. Ground shaking problems can often be directly han- dled in undeveloped areas by requiring seismic and geologic site investigations before approving develop- ment proposals and establishing and enforcing build- ing design and construction standards consistent with the seismic risk. In developed areas, the problems can be handled by abating existing structural hazards through removal or strengthening of parapets and other building appendages, basic structural improve- ment, changes in occupancy, or demolition. Examples of such plans and actions follow. SITE INVESTIGATION AND DESIGN REQUIREMENTS The San Jose seismic safety element (Duncan and Jones, 1974) is based on a thorough geotechnical study of the San Jose planning area by Cooper, Clark, and Associates. This study defines the complex relationship between ground-shaking characteristics and structural B58 type and height as follows (Cooper, Clark, and Associ- ates, 1974, p. 63): The effect of ground motion on buildings depends not only on the characteristics of the ground motion, but also on the characteristics of the buildings. The fundamental periods for typical single-story, 10-story, and 40-story buildings, are on the order of 0.2, 1.0, and 4.0 seconds, respectively. If a building is subjected to a series of ground vibrations having the same period as its fundamental period, large amplitude motions and high internal stresses develop. On the other hand, if the same building is subjected to base vibrations having a period very different from its fundamental period, comparatively small internal stresses will be induced. Accordingly, it is desirable to develop as great a difference as possible between the fundamental period of a building and the fundamental period of the estimated ground surface motion. The San Jose geotechnical report divides the plan- ning area into seven ground-response zones based primarily on depth to bedrock. Expected ranges of maximum ground surface accelerations and funda- mental periods were estimated for each zone. This in- formation can be used to highlight areas where ground-shaking characteristics could cause serious damage to particular types of structures. For example, highrise buildings may sustain more damage in areas of deep alluvium than single-story structures. The study indicates these areas where ground response may cause severe problems and where further investi- gation should be undertaken before development deci- sions are made. The San Francisco Community Safety Plan (San Francisco Department of City Planning, 1974) evalu- ates the possibility of ground shaking in a similar way, observing that the effect of ground shaking on build- ings can be compensated for by proper design and en- gineering. The report recommends “special soils- engineering and geologic investigations in areas of po- tentially strong ground shaking” (San Francisco, De- partment of City Planning, 1974, p. 23), and building code standards incorporating safety factors consistent with the building type, use, and site conditions. ABATING STRUCTURAL HAZARDS A particularly difficult and costly problem is the abatement of existing structural hazards. The San Francisco plan estimated the damage that would result from an earthquake similar to the one in 1906 by analyzing data on the age, use, construction type, number of stories, and floor area of existing structures in relation to the geologic conditions that affect ground motion. The damage potential of each block was classified as severe, heavy, moderate, or slight (fig. 27). Individual buildings were not analyzed. Precode, Type C buildings were also noted. Precode buildings are those constructed before 1948 when comprehensive lateral force requirements to resist earthquake shak— ing were included in the San Francisco building codes; Type C buildings have masonry or concrete exterior REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION bearing walls with wood floors and roofs. More than 1,400 residential buildings with nearly 35,000 living units and 2,800 nonresidential buildings were iden- tified as precode, Type C construction. Their density was mapped by census tract (fig. 28). At 1975 construc- tion costs, replacing these buildings would cost more than one billion dollars (San Francisco Department of City Planning 1974, p. 20). Objectives and policies to abate structural hazards pertain to areas where damage levels are expected to be severe such as in precode, Type C structures, (fig. 8), and in Special Geologic Study Areas (fig. 29) which have potential for ground failure or flooding. Priority is assigned to “(1) areas with high concentrations of po- tentially hazardous precode, Type C buildings; (2) areas with high population densities; and (3) those structures for which there is a critical community need” (San Francisco, Dept. of City Planning 1974, p. 42). Eliminating existing structural hazards often con- flicts with other community objectives and is politically difficult to achieve. Although San Francisco adopted an ordinance in 1969 requiring removal or strengthening of unsafe parapets and building appendages, little has been done to enforce the ordinance. Concern over both private and public costs, resis- tance of property owners, the absence of political sup- port, and concern for the effect on the architectural character of the city are reasons for the lax enforce- ment of the parapet ordinance. The Community Safety plan emphasizes preserving architectural character. Voluntary compliance with the ordinance had resulted in ”a severe loss of building character and appearance” (San Francisco Dept. of City Planning, 1974, p. 56). PLANNING TO REDUCE RISK FROM GROUND FAILURE The most damaging forms of earthquake-induced ground failure are landsliding and failures caused by liquefaction. LANDSLIDING Plans and regulations to reduce risk from slope fail- ure are similar under seismic and nonseismic condi- tions. Where unstable slopes are identified, land‘uses can be restricted, geologic investigations can be re- quired before development is allowed, and grading and foundation design can be regulated. The Town of Portola Valley has taken strong actions to reduce future losses from landslides. Spurred by in- cidents in the wet winter of 1969, the town retained a geologist to assemble the information needed to im- prove land-use decisions by avoiding landslide hazards. A geologic map at a scale of 126,000 was complied and was used to prepare a landslide potential map (offi- cially titled the Land Movement Potential of Undis- SEISMIC SAFETY AND LAND-USE PLANNING B59 \ r l GNU? Highway , . - a ,4 . . _ / ‘ BVsuuomhsi DAMAGE EXPLANATION - Severe 0 Heavy 0 Moderate D Slight D Recently demolished or to be demolished. Information not available or not in study Source. John A Blume a Anomalies. Engineers, Juno l974 FIGURE 27.—Estimated building damage levels for a "1906-type” earthquake, San Francisco Calif. (San Francisco Department of City Planning, 1974, p. 18) turbed Land Map) at the same scale. Provisions were added to the local zoning, subdivision, and grading or- dinances, requiring that geologic information be sub- mitted for review and approval by the Town Geologist before development. Even with the establishment of review procedures, it became evident that a consistent policy would be needed to relate the types of permissi— ble land use to the possibilities of landslides. To assist in formulating such a policy, the town council ap- pointed an eight-member geologic committee, chaired by the Town Geologist and composed of three geol- ogists, two engineering geologists, a soils engineer, an attorney, and a planner. The committee recommended criteria to relate land uses to the stability categories shown in table 15. The geology map, landslide poten- tial map, and criteria for permissible land use were adopted by resolution of the town council to guide land-development decisions. The town council felt that land-use regulation through zoning or other specific restrictions was not warranted because the landslide potential of individual parcels within each mapped category may vary, and because site investigation may REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION B60 Units per gross acre DENSITY OF PRECODE (1948) TYPE C RESIDENTIAL UNITS (bv 1970 Census Tracts) Buildings per gross acre RESIDENTIAL BUILDINGS DENSITY OF PRECODE (1948) TYPE C NON (by 1970 Census Tracts) FIGURE 28,—Precode, Type C buildings in San Francisco, Calif. (San Francisco Department of City Planning, 1974, p. 53). SEISMIC SAFETY AND LAN D-USE PLANNING )Presndio B . . Mm F "on ! «nu _ xx \ \ s, - a: “51V B61 . , Potential ground failure hazards ////// Potential inundation hazards FIGURE 29.—Special geologic study areas, San Francisco, Calif. (San Francisco Department of City Planning, 1974, p. 44)_ show that a given parcel is more or less stable than mapped. The resolution provides for incorporating in the official map any new information from site investi- gations. Portola Valley’s response to landslide problems is to avoid hazardous areas—a response consistent with the town’s existing and planned pattern of low-density res- idential development and policies for preserving the natural environment. In jurisdictions fostering urbani- zation or in already intensively developed areas, spe- cial site and building design or engineering to mitigate the risk from slope failure may be emphasized. The San Francisco Community Safety Plan includes landslide-prone areas in Special Geologic Study Areas (fig. 29). The plan (San Francisco Department of City Planning, 1974, p. 43) states: Special site investigations should be required in these potential hazard areas to determine the actual hazard, if any, for all proposed new development. Based upon the finding of the site investigation and determination of type and degree of hazard present, appropriate engineering design should be required to ameliorate the hazard. If proper engineering design is not technically or economically feasible, development of the site should not be permitted. Even if it is technically feasible, mitigating landslide problems is often expensive, not only for the property owner, but also for the public agency which must main- tain roads, utilities, and other essential facilities. Whenever possible, it is wise to encourage open space or low-intensity uses for landslide-prone areas. LIQUEFAC'I‘ION Reducing risk from liquefaction is one aim of the Santa Clara County Baylands Plan (Planning Policy B62 Committee of Santa Clara County, 1972); this plan covers an area subject to liquefaction as well as other types of seismic and nonseismic ground failure. Geologic and structural engineering consultants iden- tified the natural hazards of the planning area and defined their implications for land use. The resulting report divided the planning area into risk zones based on potential for settlement and ground failure under both seismic and nonseismic conditions. Table 16 lists the risk zones and the nature of the hazard in each. Figure 30 is a map of the risk zones. Table 17 relates the land and building uses to the risk zones. The plan adopts these uses with the stipulation that any developer in the baylands provide data from test boring and sample testing in depth, to demonstrate that a proposed development site is not in a higher risk zone than shown. An Advisory Review Board was rec- ommended to advise public agencies on the adequacy of engineering investigations, design, and construction methods in the baylands. Based on the plan, the county adopted an ordinance requiring a soils report for all major subdivisions un- less specifically exempted. Geologic reports and site in- vestigations are required for all subdivisions on or ad- TABLE 15,—Criteria for permissible land use in Portola Valley Land Houses stabilit Roads (parcel acreage) Utili- Water symbo Public Private ‘A-Ac l-Ac 3-Ac ties tanks MOST Sbr Y Y Y Y Y Y Y STABLE Sun Y Y Y Y Y Y Y Sex [Y] Y [Y] Y Y Y [Y] SIS [Y] [Y] [N] [Y] [Y] [Y] [N] PS [Y] [Y] [N] [Y] [Y] [Y] [N] me [N] [N] [N] [N] [N] [N] [N] MS [N] [N] N N N N N Pd N [N] N N N N N Psc N N N N N N N Md N N N N N N N LEAST STABLE Pf [Y] [Y] (Covered by zoning [N] [N] ordinance) LEGEND: Y Yes (construction permitted) [Y] Normal] permitted, given favorable geologic data an /or engineering solutions N No (construction not permitted) [N] Normally not permitted, unless geolo ic data and (or) engineering solutions favora le S Stable P Potential movement M Moving LAND br bedrock within 3 feet of surface STABILITY (1 deep landsliding SYMBOLS: ex expansive shale interbedded with sandstone f permanent ground displacement within 100 feet of active fault zone ls ancient landslide debris mw mass wasting on steep slopes, rockfalls and slumping s shallow landsliding or slumping sc movement along scarps of bedrock landslides un unconsolidated material on gentle slope REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION jacent to potentially hazardous areas as depicted on official county hazard maps. The map “Risk zones for land-use planning” (fig. 30) is one of the official hazard maps. Geologic reports are normally required for de- velopment in risk zones C and D and may be required in risk zones A and B. PLANNING TO REDUCE RISK FROM SURFACE RUPTURE Planning to reduce risk from surface rupture varies with the degree of development in the fault zone. In California, the Alquist-Priolo Special Studies Zones Act (see section on “Research and Information”) defines minimum local actions. The most effective way to avoid risk from surface rupture is to prevent construction of buildings for human occupancy across known active or potentially active fault traces. It is very difficult, if not impossible, to construct a building which can survive significant ground displacement without extensive structural damage. Preventing such construction depends on the accuracy and scale with which the fault has been mapped. As better information is available, actions in addi- tion to those based on preliminary findings can be taken. Potential surface rupture should always be consid- ered in site selection for critical public and private facilities, structures for human occupancy, and struc- tural design for lifelines. Preliminary soils and geologic reports may be required with major develop- ment proposals, and detailed site investigation re- quired if the preliminary report indicates potential hazards. A geologist is often needed to help review these reports and to help determine when more infor- mation is needed. TABLE 16.—Risk zones for settlement and ground failure established by subsurface conditions in the baylands of Santa Clara County [Adapted from Santa Clara County Planning Policy Committee, 1972] Risk Zone Surface Effect Subsurface Cause A ____________ Little risk of settlement or ground failure Bm, __________ Significant settlement Liquefaction of confined granular layer in alluvium (seismic loading) Cs ____________ Moderate to substantial Consolidation of bay mud settlement and/or dif— or soft clay (static ferential settlement loading) D1, ,,,,,,,,,,,, Substantial settlement Consolidation of uncon- and (or) differential trolled dump fill or settlement sanitar land fill (static oading) DSL ,,,,,,,,,, Failure of ground Liquefaction of granular surface surface layer (seismic loading) D15 ,,,,,,,,,, , ,do_ _ Lateral spreading toward free face (seismic load- ing) B63 .Ambma .oBfiEEoO hunch 9:535 3550 «.35 «€me mmiflzmm 55.60 530 Scum @5583 $565: new moses memldm "559m .0323: 3:539 =03a—ESE 33532.». “5:20“. 0.85 an own—Eu 2:8 acoumcwfiou 0:3 vim .cofiaEaefi 33.53.; teeth: co “.32 mm 3:3 v.2.— mo 5.2353“. 2:. uNFOZ SEISMIC SAFETY AND LAND-USE PLANNING wImFmS. OOON 00°F 0 lJLJl. :ounukmxu: 3335. hum". 008 ooov o was >E $.54 53 Auhmflanon :0 326% SK 5:5» uu=ct=o§= Q 0:3 33% 933356 \o 858:» .3 33m 35:5 235222. Kc BEBENEQ :0 336m $363.3? 3533 0 BEN vim xe AEEEHDR no 33% rooN vim . a :otgkmsw: mmmw KO AEBNEGA to «5.3% m “EON vim QQ ! < 25m xix «QED mxwafih w‘HQ i305. 3&5 EE=:°V<\ / B64 TABLE 17.——Land and building uses appropriate for various risk zones, Santa Clara County [Adapted from Santa Clara County Planning Policy Committee, 1972, p. 22] Risk Zones A B C D Land and Building Uses Group A Buildings Hospitals and nursing homes ________________ >< Auditoriums and theatres ____________________ >< Schools ____________________________________ >< Transportation and airport __________________ >< Public and private office ,,,,,,,,,,,,,,,,,,,, >< Major utility ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, >< Group B Buildings Residential-multiple units ____________________ X X Residential- 1 and 2 family __________________ X X Small commercial ,,,,,,,,,,,,,,,,,,,,,,,,,,, X X Small public ________________________________ X X Small schools-one story ______________________ X X Utilities ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, X X Group C Buildings “Industrial park” commercial ,,,,,,,,,,,,,,,, X X X Light and heavy industry _____________________ X x X Small public, if mandatory ,,,,,,,,,,,,,,,,,,, x X X Airport maintenance _________________________ X X X Group D Buildings Water-oriented industry ,,,,,,,,,,,,,,,,,,,, X X X Wharves and docks __________________________ x X X Warehouses ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, X X X Group D Open Space Agriculture, marinas, public and private open spaces, marshlands and saltponds, and small appurtenant buildings SURFACE RUPTURE—UNDEVELOPED AREAS A draft of the San Mateo County seismic safety ele- ment lists ways to reduce risk from surface rupture based primarily on fault mapping at a scale of 1:62,500. A portion of this map is shown in figure 31. The policy options relevant to largely undeveloped areas include: 1. “Restrict development within active or poten- tially active fault zonesm” (San Mateo County, 1975, p. 72) 2. “Encourage the State Public Utilities Com— mission to establish increased design and construction standards for utility systems traversing active or potentially active fault zonesm” (San Mateo County, 1975, p. 75) 3. “Prohibit development of critical use struc- tures in any active or potentially active fault zonesm” (San Mateo County, 1975, p. 77). Geologic, seismic, and soil investigations of individual sites are recommended before making land-develop- ment decisions in the designated fault zones. Figure 32 is an example of a Special Studies Zones map at a scale of 1:24,000. For regulatory purposes, the REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION zone boundaries are very accurately located using a coordinate system, but the faults and the fault zones are much more complex and irregular. The fault loca- tion shown in figure 32 is derived from the 1:62,500 map (fig. 31) which in turn was based on field mapping of the fault and analysis of stereopairs of aerial photos at different scales. Direct regulation through zoning requires maps that are detailed and accurate enough to show the distance of existing and proposed structures from the fault. The fault mapping done for Portola Valley in 1970 by W. R. Dickinson (fig. 33) provides the basis for the town’s fault setback requirements adopted in 1973 (or- dinance 1973— 1 19) as part of the zoning ordinance. The ordinance prohibits structures for human occupancy within 15 m (50 ft) of a “known” fault trace. “Known” locations are based on surface expressions or subsur- face studies which fix the location of the trace. No use more intensive than a single-family, one-story, wood- frame house, or house of similar earthquake-resistant design, is permitted in the band from 15 m (50 ft) to 38 m (125 ft) on either side of a known fault trace. Setback distances for an “inferred” fault trace are larger—no structures for human occupancy are per- mitted within 30 m (100 ft) of the inferred location and only single-family homes are allowed for an additional 23 m (75 ft). “Inferred” locations are based on the pres- ence of a limited number of surface or subsurface in- dications of a fault trace. The actual position of the “inferred” location is subject to greater error than the “known” location, and therefore the width of potential risk band is increased. A property owner may contract for detailed geologic investigation to locate an "in- ferred” trace more precisely. In such cases, the ordi- nance provides that the trace be reclassified as “known” and the setback requirement correspondingly reduced. Outside the setback lines shown in figure 33, all pro- posals for development more intensive than single- family residences are reviewed by an engineering geologist employed by the town to determine if the site might be subject to significant offset or ground warping related to surface rupture. Existing structures in the fault zone are not affected by the setback ordinance. Had the town chosen to make the fault zone into a zoning district rather than requir- ing setbacks, existing structures would have become nonconforming and subject to eventual removal, de- pending on the zoning ordinance provisions covering nonconformity. Generally speaking, as the fault mapping becomes more precise, the area subject to regulation becomes smaller. The Alquist-Priolo Special Studies zones for Portola Valley encompass a significantly larger area SEISMIC SAFETY AND LAND-USE PLANNING 365 m «I ’~ / . WW ,9” , . —Active and probably active faults and fractures and fracture zones for a portion of San Mateo County, Calif. at a scale of 1262,500 (Brown, 1972). The rectangle outlined is shown at larger scale in figure 32. B66 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION r / L \ ,,/%9(3§7“1 01%; 3’ ‘j, 7%“ W/ n F , \ / W/ p, \\l),‘(;;V /// H’ "l/ W ‘ .//// // /// mg FIGURE 32.—Special Studies Zones for a part of San Mateo County, Calif, at a scale of 124,000. The area outlined is shown at larger scale in figure 33. SEISMIC SAFETY AND LAND-USE PLANNING than that subject to the setback regulations of the town. Specific geologic investigation often narrows the area of potential surface rupture, but where site inves- tigations produce uncertain or unforeseen results, a wider zone of potential surface rupture may be in- cluded. Fault maps used for planning should be systemati- cally updated as new geologic and seismic information becomes available. Excavations for road cuts or con- struction projects, geologic and geophysical site studies, and wells drilled near a known or suspected fault may provide new evidence of subsurface faulting. Where the public interest is affected, geologic investi- gations to locate fault traces more exactly may appro- priately become the responsibility of the public agency. All information pertaining to fault location at a given site should become part of the title record for the parcel. SURFACE RUPTURE—DEVELOPED AREAS Where an active or potentially active fault passes through urban or urbanizing areas, the basic planning problem is how to prevent new construction and how to remove existing structures from on or near an active fault trace with equity and a minimum economic im- pact on the community. Some local jurisdictions decide that the effort simply is not worth the economic and social costs, and the risk is accepted. At the least, agencies faced with this risk should prepare a plan to prevent rebuilding after an earth- B67 quake in areas subject to surface rupture. Such a plan can be similar to a redevelopment plan, that is, specific enough to assure relocation of structures away from areas of potential surface rupture, but general enough to allow flexibility in responding to conditions existing after the earthquake. During redevelopment under various Federal, State, or private programs, structures can be removed from an active fault trace. Such removal is appropriate in older areas where there is structural deterioration. The redevelopment area needs to be large enough to retain the fault zone as open space and still provide enough buildable space to make a project economically feasible. Although high-occupancy or critical facilities in ac- tive fault zones should be removed, public investment in such facilities may be so high as to make such action uneconomic. The Hayward fault runs through a highly urbanized and rapidly growing part of the East Bay (fig. 34). Many schools, public facilities, commercial, industrial, and residential buildings are located on or near the fault. Although no major earthquake has occurred on this fault in more than 100 years, tectonic creep makes continuing maintenance and repair necessary. The fault is active and is considered capable of producing an earthquake of magnitude 7.0—7.5 accompanied by surface rupture. The City of Hayward, crossed by the fault, adopted a seismic safety element in 1972. In it fault traces were Sequoias (retirement home) 50’ Setback EXPLANATION Fault trace location known Fault trace location inferred 0 500 FEET 0 150 METERS FIGURE 33.—Known and inferred fault trace locations and setback lines, Portola Valley, Calif, at a scale of 126,000 (Dickinson, 1970). B68 mapped at a scale of 112,000, and a “fault corridor” was defined including the mapped fault traces and 15 m (50 ft) on either side. This corridor was declared an area of high seismic risk, and creation of a zoning com- bining district applying to the corridor was recom- mended. This seismic safety element further recommended that existing structures or portions of structures iden- tified as hazardous be declared public nuisances and, as such, subject to repair, rehabilitation, or removal. Structures within the fault corridor would be subject to the nonconforming use provisions of the zoning ordi- nance upon adoption of the earthquake-fault combin- ing district. In accord with these recommendations, the Hayward City Council considered designating an Earthquake- Fault Combining District within which construction of the following types of structures over an active fault trace would be prohibited: residences, facilities re- quired for emergency response, structures over 23 m (75 ft) in height, or high-occupancy buildings such as schools, churches, and theaters. Seismic, soils, and geologic reports would be required for new structures and additions to existing structures intended for human occupancy. Before the Council adopted the combining district, the Alquist—Priolo Act, with similar provisions, was passed by the State legislature, and so the city dropped the matter. The Hayward fault corridor runs through the middle of the older downtown section of the city. The city hall and police station, formerly located astride the fault, have been relocated to a new city-center complex a few blocks from the fault (fig. 35). Concern over economic decline of the downtown area led to a revitalization plan (Hayward, 1975a). The plan envisions an L- shaped downtown area linking the BART (Bay Area Rapid Transit) station and the new city center. An in- crease in development was recommended in the area bounded by Foothill Boulevard, C Street, Main Street and Hazel Avenue (fig. 35). Because this area is in the Special Studies Zone, no new development or substan- tial redevelopment could be approved without a geologic investigation. The city contracted with a con- sulting firm to conduct geological and geophysical in- vestigations, including trenching, in this part of the proposed redevelopment area. The study concluded that no fault traces were present in the area (Burkland and Associates, 1975). The consultant’s report provided the basis for a re- quest by the city to the CDMG (California Division of Mines and Geology) that the area be removed from the Alquist-Priolo Special Studies Zone. The study, if ap- proved by CDMG, will meet the requirements of the act for geologic investigations, and development may be REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION . 1"; ' I». [c ., [J l. I I 1 FIGURE 34.—Hayward fault traces in urban and urbanizing area (from Nichols and Buchanan-Banks, 1974). permitted in the area. This is one of the few examples of a public agency assuming the responsibility and cost (in this case $18,000) of conducting the geologic inves- tigations required by the State law. In most instances, the burden falls on the individual property owner. Four blocks of the redevelopment area between Main Street and Mission Boulevard are crossed by the Hay- ward fault. Consequently, the plan calls for meeting the needs for additional downtown parking by expand- ing parking lots in this area as parcels become availa- ble. The city proposes to use tax increment financing permitted under the California Community Redevelop- ment Law (California State Legislature, 1963), HUD Community Development Block Grant funds, parking revenues, and other available funds for public im- provements in the redevelopment area. The official redevelopment plan (Hayward, 1975b) does not directly address the risk from surface rupture in the fault zone. However, implementing the plan could significantly reduce risk from surface faulting. If carried out in accordance with State law and the adopted policies of the Hayward Earthquake Study (Hayward, 1972), the plan will meet seismic safety ob- jectives as well as its stated objectives of revitalizing the downtown area and increasing the space available for parking. PLANNING TO REDUCE RISK FROM FLOODING Tsunamis and failures of dams or dikes are the major SEISMIC SAFETY AND LAND-USE PLANNING source of earthquake-induced flooding. Reducing the flood hazard can be accomplished by regulating con- struction in flood-prone area, by the use of warning systems, by planning for evacuation, and by building various structures to confine or control flooding. TSUNAMIS Because damaging tsunamis are infrequent in United States coastal areas, few coastal communities have tsunami-preparedness plans. A tsunami warning system in the Pacific basin, directed by the National Oceanic and Atmospheric Administration, is the major attempt to reduce risk. Sea walls, breakwaters, or other structures designed to protect coastal areas from storm surges may also prevent damage from small tsunamis; but the cost of protecting communities from the largest foreseeable tsunami by engineering works is unacceptably high (Ayre, 1975, p. 106). Planning to reduce tsunami risk may be stimulated by the Federal Coastal Zone Management Act of 1972 (US Congress Public Law 92—583); this act authorizes Federal grants to coastal states to prepare manage- ment programs for coastal zones. These programs must identify and list “areas of significant hazard if devel- oped due to storms, slides, floods, erosion, settlement, etc.” (US. National Oceanic and Atmospheric Admin- istration, 1975, p. 1687). Potential tsunami runup areas are also included in “coastal high hazard areas” and subject to the re- quirements of the National Flood Insurance Program outlines in Section II. Because of Federal and State requirements, the land-use plans of coastal cities and counties are now more likely to include risk from tsunamis than in the past. The Seismic Safety Element for Monterey County is one of the few that contains policy directly relating to tsunami risk. The following policies (Monterey County, 1975, p. 34), based on very general mapping of areas of historic tsunami runup, were adopted by the county: 1. In general, known tsunami runup areas should be avoided by new development except marine installations requir- ing location in proximity to water. 2. Where development presently exists an adequate warning and evacuation system is essential. 3. All reasonable measures will be taken by this jurisdiction to reduce potential damage. Such measures will include es- tablishing and enforcing standards of construction for structures within harbors and known runup areas, and formulating post-disaster plans for debris clearance and emergency repairs to essential facilities. More precise delineation of potential runup areas and estimates of probable frequency of occurrence would make it possible to implement plan policies in tsunami runup areas by zoning for low-intensity or marine-oriented uses, establishing setback or eleva- B69 tion requirements for proposed structures or imposing design and construction standards. DAM AND DIKE FAILURE Reducing risk from dam or dike failure is particu- larly critical in seismically active areas because of the enormous potential for loss of life and destruction of property. Preliminary estimates of property damage from the recent failure of the Teton Dam in Idaho in a sparsely populated area are around $1 billion. Dam or dike safety is chiefly the responsibility of the engineer, builder, and Operator, and, in the past, little effort has been made to control downstream land use as a means of reducing the change of catastrophic losses. In the future, risk of dam failure may be considered in land- use planning by placing restrictions on development of lands below the dams. Dike failure is of particular concern to communities along the edge of San Francisco Bay. Many dikes rest on unstable bay mud and are composed of materials unlikely to withstand severe ground shaking. The area subject to flooding from dike failure depends on the tide level and elevation of the land. The Alviso area of San Jose, situated about 21/2 m (8 ft) below sea level, is particularly vulnerable to severe flooding as well as other environmental hazards. In recognition of this, the comprehensive plan of the City of San Jose recommends that land uses in close proximity to water retention levees or dams with moderate or high potential for seismic failure shall be carefully regulated (San Jose, 1976, p. 19). No residen- tial or employment growth or land-use change through 1990 is projected for the Alviso area. This decision was reached after two basic alternatives for the future of the area were explored: (1) Build flood control levees and maintain and upgrade the existing community, or (2) relocate Alviso residents to other parts of San Jose (San Jose, 1976, p. 35). Based on an analysis of com- parative costs and existing public investment in Al- viso, the planers concluded that Alviso should remain where it is, and flood-protection levees should be pro- vided. Long-term development options for the area are contingent upon a structural solution to the flood hazard. Planning for the Alviso area is a good example of the complex considerations involved in land-use decisions. The result is not optimum from the point of View of reducing seismic risk, but it represents a balancing or risk with other important economic and social factors and objectives. PUTTING IT ALL TOGETHER The effects of an earthquake are varied, and the abil-: ity to predict and evaluate the potential severity and B70 REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION CITY CENTER — CAPWELLS ACTIVITY CENTER (‘00) ly,(< Botany Grounds Exp-mien to Foothill, 0 Rune" and Second 0 o Foothill Boulevard Improvements between HazeI and Jacknon Future Parking Structure 4&3 ‘21? Area of CItyrfLInded geoIogIc Investigation Parking Lot Expansion in Faulr Corridor Ono-wly Couplfl on B and C Street: Transit Service (Within 10 Second) Improvement between BART and City Parking Strucmres Parking Lot ‘ _ in "Key” Block: atBand Wa‘kiM SpeCIal StudIes (Immediam Acquisition) Zone boundary B Street Improvements between Foothill end BART EXPLANATION RETAIL AND OFFICE OFFICE AND RESIDENTIAL RESIDENTIAL PUBLIC OPEN SPACE Fault rupture, 1868 MAJOR STREET Parking Lot Office BART STATION SECONDARY STREET ACTIVITY CENTER PARKING LOT PARKING STRUCTURE 6; 7% U} PEDESTRIAN WAY I 100' I Special Studies Zone boundary BIKEWAY EIEIIIIHI SPECIAL TREATMENT PRELIMINARY ACTION PROPOSAL FIGURE 35.—Hayward redevelopment plan. SEISMIC SAFETY AND LAND-USE PLANNING location of each effect differs, therefore, it is often nec- essary and desirable to respond to risk from each seis- mic hazard separately. A land-use plan, however, re- flects the desires of a community for the future and is based on a multitude of often conflicting objectives and priorities. Reducing seismic risk requires integrating all the related seismic concerns into an overall plan for community development. Three methods are used to achieve this end: land-capability analysis, systematic consideration of seismic risk in land-use policy and re- gulation, and project review requirements within a general policy framework. These methods are com- plementary, not mutually exclusive. Each method re- lates to a particular phase of the planning process. A land-capability study is a way of interpreting data for direct application in plan formulation. Designating land uses involves formulating a plan and regulating land use. Project review is largely reactive, dealing with development proposals as they come rather than prescribing specific uses ahead of time; it emphasizes plan implementation. Each method is discussed below. LAN D-CAPABILITY ANALYSIS In any area, the existing natural features and pro- cesses present a range of constraints and opportunities for different uses of land. Land-capability studies sys- tematically record judgments concerning the effects of these factors on the value of the land for selected uses. The factors considered usually include topography, hy- drology, geology, soils, vegetation, and climate. Methods of evaluating land capability differ. A study may be largely descriptive, pulling together in narra- tive form information concerning the natural features and processes relevant to a particular land use; or a study may involve a fairly sophisticated effort to quarr- tify, weight, and aggregate the factors relevant to specific uses for all lands within a planning area. In any case, judgment is needed, and the studies should be carried out by planners with the assistance of experi~ enced earth scientists. Land-capability analysis in- volves four basic steps: . 1. Defining the scope of the study and the land use or uses to be considered. ‘ 2. Determining the factors affecting capability for the use or uses selected. ‘ 3. Gathering, analyzing, and presenting the per‘ tinent information. 4. Evaluating the relative capability of the land units to support the selected use or uses. The relative importance of each factor, as well as the range of conditions within each, can be expressed nu- merically. The main advantage of numerical analysis is the greater ease in combining many judgments into an overall rating of land capability. Such analyses varyi B71 greatly in precision, however, depending on the quality of data, the qualifications of the analysts, and the method used. Land capability studies vary in focus as well as in method. A common variation is an analysis which rates land within a study area in terms of relative risk from selected natural hazards. A study may be very detailed, dividing an area into small units which are evaluated for a specific use such as a sanitary landfill; or it may be general, dividing an area into large units which are evaluated for a broad category such as urban development. The Seismic Safety Element of Santa Barbara County uses techniques of land-capability analysis to rank areas, on a grid system, in terms of relative seis- mic and geologic hazards. The following hazards were evaluated: ground shaking, tsunamis and seiches, liquefaction, slope stability, expansive soils, soil creep, compressible/collapsible soils, and high ground water. Surface rupture was considered separately because, as an essentially linear phenomenon, it is difficult to in- corporate into a grid analysis. Each grid cell was rated 1—3 for each hazard based on the following system: 1 equals none or low hazard; 2 equals moderate hazard; and 3 equals high hazard. Each hazard was given a weight representing its im- portance relative to the other hazards. The weight was based on three considerations: consequences severe or moderate consequences (such as loss of life or property damage), frequency of occurrence, and difficulty of pre- vention or mitigation. The hazards were then assigned the following weights: Seismic severity (ground shaking) ______ 18 Tsumani-seiches ______________________ 19 Liquefaction __________________________ 15 Slope stability ________________________ 23 Expansive soils ______________________ 7 Soil creep ____________________________ 4 Compressible/collapsible soils __________ 11 High groundwater ____________________ 3 Total ______________________________ W (100 is the lowest possible score assuming a rating of 1 for all hazards) For each grid cell, a weighted rating for each hazard was obtained by multiplying the weight by the rating. The weighted ratings for all hazards are then totalled for each grid cell. This total is called the GPI (geologic problem index). The GPI was calculated for each 90- acre grid cell county-wide and for each 5-acre grid cell in four urban areas. The range of GPI’s was 100—236 (300 max.) No cell received a maximum GPI because some problems are confined to flatlands or hillsides, and no one cell had a high rating for all hazards. For convenience, GPI’s were grouped into five categories; B72 GPI range Category Severity 100— 125 __________ I low 126— 145 __________ II low-moderate 146— 180 ____________ III moderate 181—210 ____________ IV moderate-severe 210-up ____________ V severe Computer mapping of the five categories was used to show the relative severity of geologic hazards through- out the county and in the four urban areas in greater detail. The system employed in Santa Barbara County includes a “variability number” to indicate differences in reliability of the hazard ratings for particular grid cells resulting from potential local variations, quality of data, and other factors. Areas with the same GPI rating or in the same severity category may have dif- ferent variability numbers which can affect planning recommendations. The Santa Barbara study is a good example of the incorporation of seismic considerations into a land capability analysis. Although the hazard ratings were not related to particular land uses, the GPI does pro- vide an overview of relative seismic and geologic hazards throughout the county. The ratings can be re- lated to levels of acceptable risk for different categories of use, thus providing a guide for land-use planning. Based on the GPI, the seismic safety element rec- ommends that the county: 1. Consider areas in category V for open space, recreational or agricultural use, or possible low-density use, because cost of safe devel- opment may be high. 2. Consider areas in category IV for low-density use or nondevelopment. The relative costs of measures needed to mitigate adverse natural conditions affect the values assigned in a land-capability study. In a pilot study of a part of the Santa Clara Valley, the Association of Bay Area Governments (Laird and others, 1979), expressed land capability in terms of the dollar costs required to miti- gate hazards or to compensate for property damage and loss of natural resources. The ABAG study included such geologic and hy- drologic hazards as ground shaking, surface rupture, flooding, bearing-materials problems (potential for shrink/swell, settlement, liquefaction, and subsidence), slope stability, erosion/sedimentation, and septic-tank limitations. The study also included an evaluation of natural resources. Lands in the study area were evalu- ated for a range of uses: agricultural or rural, semi- rural residential, single-family residential, multi- family residential, regional commercial, downtown commercial, industrial manufacturing, and freeway. The total cost associated with each natural con— straint and resource for each land use was estimated. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION Table 18 lists for each land use the estimated costs associated with different intensities of ground shaking. Estimated costs per acre are obtained by multiplying the value of buildings, personal property, and utilities by the percent damage expected by the annual fre- quency of occurrence, and dividing by a discount rate to reduce future values to present levels. In this case costs were based on an anticipated damage level associated with each land use. Cost information for the identified natural resources and hazards for each 24.9-acre grid cell was aggregated for each land use. The resulting number indicates for each cell the estimated dollar cost per acre of develop- ing that cell with that land use. The range of total costs was divided into six capability levels and a land- capability map for each use was printed by computer. Figure 36, a land-capability map for a part of the Santa Clara Valley study area, is derived from table 19, which shows the costs associated with all the hazards, constraints and resources for multi-family residential use. Analysis of land capability provides only part of the information needed for land—use decisions. Economic, social, political, and esthetic considerations are also important. The physical capability of a parcel of land to support an intensive use may be poor, but other factors, such as location and accessibility, land cost, absence of alternative lands, or overriding public need, may well indicate that the parcel should be intensively devel- oped. A study which systematically evaluates economic, social, and political factors, in addition to physical capability, is often called a "land-suitability study”. A land-capability study can be undertaken as part of a broader land-suitability study. However, on occasion, capability is, or should be, the determining factor. Areas with very low capability for sustaining a par- TABLE 18.—Costs associated with ground shaking resulting from events on the San Andreas, Hayward, or Calaveras faults [Laird and others. 1979] Cost per acre (in dollars) San Francisco Intensity Scale Land use A B C D E Rural or agricultural - V, 40 30 10 5 1 Semi-rural residential ,,,,,, 300 300 100 40 4 Single-family residential ,,,,,, 4,000 3,000 1,000 500 50 Multi-family residential ,,,,,, 20,000 20,000 5,000 2,000 200 Regional shopping centers 50,000 40,000 10,000 4,000 800 Downtown commercial ,,,,,, 70,000 50,000 20,000 5,000 1,000 Industrial ,,,,,,,, 40,000 30,000 10,000 3,000 700 Freeways ,,,,,,,,, 10,000 10,000 10,000 0 0 SEISMIC SAFETY AND LAND-USE PLANNING TABLE 19.—Summary of costs for multifamily residential use [Laird and others, 1979] B73 Hazard, constraint, or resource Costs for map categories (in do lars per acre) (scale from severe to slight) Severe Slight Surface rupture __________________________________________________ 800 800 0 _ __ ____ Ground shaking, San Andreas, Hayward ,,,,,,,,,,,,,,,,,,,,,,,, 20,000 20,000 5,000 2 ,000 200 Ground shaking, Southern Hayward ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 2,000 2,000 500 200 20 Ground shaking, Calaveras ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 20, 000 5,000 2,000 200 ____ Stream flooding __________________________________________________ 40, 000 "d A”- ____ Dam failure ________________________________________________________ ,,,_ A“, ____ ____ Dike failure __________________________________________________ 80, 000 O ____ ____ ,1" Shrink/swell soils ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 20,000 7,000 0 O ____ Settlement ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 30,000 30,000 20,000 2,000 -U- Li uefaction ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 4,000 3,000 300 20 0 Su sidence __________________________________________________ -_-_ ,_-_ ,v,_ A-“ Landslides ____________________________________________________ 200,000 100,000 50,000 9,000 0 Soil creep ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 40,000 40,000 0 0 1 u- Erosion and sedimentation ________________________________________ 200 30 10 0 “a Septic tanks ________________________________________________________ 0 ____ __‘_ ____ ____ Sand and gravel ______________________________________________ 20,000 0 ,_,_ d" ,_,- Mercury ____________________________________________________________ 0 ____ ____ ____ ____ Agricultural land ______________________________________________ 5,000 0 ,_s_ _-__ ____ ticular use can sometimes be eliminated from further consideration, allowing the planner to focus attention on more realistic options. Land-capability studies are becoming increasingly important to land-use planners at all governmental levels. They assure that physical characteristics of the land will be systematically considered in land-use planning. The earth-science information requirements for such studies vary with the total land area and the specific use to be studied. For example, fairly general data may be appropriate for an initial analysis of land capability for regional open space. On the other hand, a study undertaken, at any governmental level, to locate specific sites with good capability for sanitary landfill will require detailed information. Land-capability analysis allows seismic-risk con- straints to be considered along with other natural characteristics in making land-use decisions. The hazard information developed in the process of seismic zonation is combined with other natural characteris- tics and given a numerical weight. Capability analysis relates seismic and other hazards directly to potential land uses. Land-capability studies may usefully serve as an intermediate step in identifying areas where hazards are present and where detailed consideration of risk is needed in deciding land uses, structural de- sign, and occupancy. LAND-USE POLICY AND REGULATION Many cities and counties in California have com- pleted the seismic safety elements of their general plans. For many of these jurisdictions, the preparation of these elements was their first experience in using geologic information systematically in a planning task. The typical seismic safety element is a preliminary step toward developing a comprehensive program to reduce seismic risk. Because of limited experience in using existing geologic data, most seismic safety ele- ments emphasize the need for more data. If detailed data were available and the planning staff was experi- enced, more specific recommendations for immediate action resulted. While most seismic safety elements contain recommendations for land-use policy, few jurisdictions have yet integrated seismic safety policies and programs into their comprehensive plans. This kind of planning will come about when more cities and counties combine the various required general plan elements into a comprehensive document. The recently adopted General Plan 1975 for the City of San Jose is one of the first to consider seismic risk as an integral part of a comprehensive plan. The land-use pattern of San Jose is a classic example of urban sprawl resulting from very rapid development following World War II. An aggressive annexation policy and a growth-oriented political climate led to more than a fivefold increase in city population from 1950 to 1975—from just under 100,000 to 547,500. During the five-year period 1969— 1974, an average of 769 hectares (1,900 acres) was converted to urban uses each year (San Jose, 1976, p. 7). Increasingly, land with devel- opment constraints was being pressed into urban use. The San Jose General Plan 1975 is a guide for managing growth to match the city’s ability to extend urban services, avoid development of unsuitable lands, and achieve a more efficient urban form and a better balance of land uses. The first step was the adoption in 1970 of a set of urban development policies. These policies are now incorporated into the General Plan REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION B74 . 1.1Lluv 3.1.1. 1.1.1,.Iuh4ndn \. , itlfilflfiififliitiluhqwfi‘lvi. ,111}%.L,/1%4n$awilyfbw . ‘Ifltylllvqniflinwnv... 1.. .th. Si Ayyfltfifilttfit.ianbdu1w iiv lint-Illnt‘lllhiw!h,lr./VVII1J V, lull. , ,.Cr.,.:..r4n1v.l.l iWIWLWIJIIhWi.“ FIGURE 36.—Land—capability map for multi-family residential use (Laird and others, 1979, p. 75). SEISMIC SAFETY AND LAND-USE PLANNING 1975. Future development is to be limited to a desig- nated urban service area for the 15-year time span of the plan. Within the urban service area approximately 8,134 hectares (20,100 acres) were undeveloped in 1975. Of these, 1,052 hectares (2,600 acres) were al- ready approved for private development; 425 hectares (1,050 acres) committed to public uses; 399 hectares (985 acres) under Williamson Act Contract‘, 301 hec- tares (745 acres) programmed for public acquisition; and 770 hectares (1,920 acres) considered ill-suited for development because of size, shape, slope, soil subsi- dence flooding, or location in an airport safety zone (San Jose, 1976, p. 7). The plan contains specific policies related to lands considered unsuitable for urban development within the urban service area as well as throughout the area. Based on a goal of striving to minimize risk from natu- ral hazards, the plan (San Jose, 1976, p. 19) contains the following general policies: 1. The City shall not permit urban development in those areas where such development would constitute a significant potential danger to the health, safety, and welfare of the resi- dents. ‘The California Land Conservation Act of 1965 (California Government Code, 1965), also called the Williamson Act, permits landowners to enter into contracts with cities and coun- ties in which the landowners agree to maintain land in agricultural use in exchange for tax assessments based on the economic return from agricultural use of land. EXPLANATION Total Cost Range No. Of Cells Level Symbol (in dollars per acre) In Level 1 " " 0.01 10.00 0 2 33335 10 01 100.00 0 3 100.01 1000.00 0 §§§§§ 4 fix; 1000.01 10000.00 16 §§§§§ 5 10000.01 100000.00 [+644 6 lOOOO0.0llOOOOO0.00 2877 FIGURE 36.—Continued. B75 2. Low levels of ‘acceptable exposure to risk’ shall be established for land uses and structures in which failure would be catastrophic, which are required during emergencies, or which involve involuntary or high human occu- pancy. 3. Risks from natural hazards shall be reduced as much as possible in areas where human ac- tivity is necessary or already exists, and where the natural and man-made environ- ment can be safely integrated. 4. Preventative measures for known natural hazards shall be taken simultaneously with new development. 5. Site-specific information on natural hazards shall be required for proposed new develop- ment and Where identified hazards preclude safe human interaction, development shall yield to natural processes. 6. Provision shall be made for the continuation of essential public services during natural catastrophes. 7. The City shall promote an awareness and cau- tion among San Jose residents regarding pos- sible natural hazards including soils condi- tions, earthquakes, flooding, and fire hazards. Specific policies regarding seismic safety call for re- habilitating or removing structural hazards “without creating undue hardship or relocation policy problems” (San Jose, 1976, p. 19); restricting construction near creek channels where liquefaction is a hazard; requir- ing geotechnical studies to determine the extent of seismic hazards prior to approval of development pro- posals; regulating land uses in areas prone to flooding from dike or dam failure; and requiring detailed dynamic ground motion analysis and suitable struc- tural design for critical facilities (San Jose, 1976, p. 19). These and other policies apply to areas designated as hazardous on maps which are part of the Geotechnical Report prepared by Cooper, Clark, and Associates (1974) as background for the Seismic Safety Element. A generalized natural hazards map (fig. 37) is incorpo- rated in the General Plan 1975. The importance of avoiding development in hazard- ous areas is reflected in the land-use policies and land-use diagram. For example, the following policies of the General Plan 1975 (San Jose, 1976, p. 21, 25) are related to specific land uses: Solid waste disposal land fill sites shall be discouraged on lands which are susceptible to landslides, seismically induced ground fail- ure, *"dam inundationm. Residential development shall not be al- lowed to occur in areas where such development might be hazardous to human habitation'“Densities permitted by the General Plan on B76 slopes greater that 15 percent may be allowed to be transferred***. The land-use diagram shows the area underlain by bay mud as open space, agriculture, and light industry. Areas adjacent to major creeks which may be subject to ground failure are shown as open space. Hillside areas to the northeast and southwest of the valley floor are designated for nonurban uses. In these areas, slope failure and surface rupture during an earthquake are potential hazards. To implement the plan, new zoning districts limiting the density of residential development in hillside and other outlying areas are to be prepared. Geologic hazards are to be systematically considered in the re- view of development proposals. Where warranted by geologic investigation, designated land uses can be al— tered because the geologic hazards may be overriding. The 1975 plan evaluates the full range of factors affect- ing the future development of the San Jose area. Conflicts among economic, social, and environmental objectives are recognized and resolved into a plan and an implementing program that explicitly incorporates seismic safety con— cerns into the decisionmaking process. PROJECT REVIEW Another means of integrating seismic risk and land-use planning is to develop, within a general policy framework, project review requirements and proce- dures. Such requirements and procedures are appro- priate when detailed data on seismic hazards are not available. Generalized data can be used to alert plan- ners and decisionmakers to potential problems. Such a system generally identifies areas where seismic, geologic, or soils investigations are required before development proposals are approved. Specific report requirements, procedures for evaluating reports and requiring hazard mitigation, and criteria for determin— ing the acceptability of proposed projects can be devel- oped to incorporate seismic safety concerns. Project review can be very effective if it assures that seismic risk is considered in site selection, structural design,a nd occupancy of major development proposals. Although the developer has the responsibility for col- lecting data, the public agency must have sufficient information and geologic expertise to evaluate the geologic and seismic reports submitted with the proposals. Santa Clara County emphasizes project review. The Seismic Safety Plan (1975) describes the seismic and geologic hazards in the county and general policies to mitigate or avert undue seismic risk in existing or fu- ture development. The essence of the plan, however, is contained in the recommendations for geotechnical site investigations (Santa Clara County, 1975, p. 19—20): In order to maximize public safety and minimize seismic hazards, additional local geotechnical studies should be performed prior to further development in many areas of the County. These studies REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION should consider the data in this report as general background and regional material and should determine the extent of particular seismic hazards on each site in relation to the specific intended use. These geotechnical investigations should be multidisciplinary, in- cluding component studies of seismology, engineering geology, plan- ning, hydrology, architecture, design engineering, structural en- gineering, and soil engineering. These interrelated components should be coordinated so that all pertinent factors are considered. To review and approve these geotechnical investigations, it is rec- ommended that the County should develop an adequately trained and funded staff team including the various disciplines mentioned above. To help decide if a geologic or geotechnical investiga- tion should be required, the county uses a Relative Seismic Stability Map prepared by the California Divi- sion of Mines and Geology at a scale of 1:62,500. Figure 38 is a part of the reduced version of this map which is included in the Seismic Safety Plan. The original map is incorporated, by reference, in a county ordinance set- ting forth soils and geologic report requirements (Santa Clara County Board of Supervisors Ordinance No. NS—1203.31, December 1974). Soils and geologic reports may be required when applications are submit- ted for subdivisions, building site review, grading per- mits, and building permits. Soils reports are to be prepared by a civil engineer registered by the State and geologic reports by an en- gineering geologist certified by the State. The county staff includes an engineering geologist and other ex- perts competent to evaluate the reports and the mitigating measures proposed. The report require- ments and evaluated procedures are part of the total review process, and they ensure that geologic and seismic hazards are considered adequately in land-use decisions and land-development practices. POSTEARTHQUAKE RECONSTRUCTION After a damaging earthquake, economic, social, psy- chological, and political pressures coalesce to hasten rebuilding. Often this leads to restoring area, build— ings, and services to their previous condition without regard for site or structural hazards revealed by the earthquake, or hazards previously identified, but not heeded. If properly planned and carried out, recon- struction following an earthquake can greatly reduce risk from future events. The San Francisco Community Safety Plan (San Francisco Department of City Planning, 1975) stresses the opportunities presented during reconstruction to carry out the objectives of the comprehensive plan. The plan (p. 38) recommends that the city: Adopt contingency legislation to provide for anticipated needs follow- ing a disaster and to reduce pressures for unnecessarily rapid recon- struction. Create a reconstruction planning committee to insure that devel- opment following a major disaster takes place in a timely fashion according to established objectives and policies. EXPLANATION High landslide susceptibil- ity (high landslide suscep- tibility bordering creeks omitted due to scale) Fault traces Potential salt water inun- dation High ground failure sus- ceptibility Soil creep (highly expan- sive soils above 15 per- cent slope) Very weak soils/recent young bay mud Potential major destruction due to dam inundation (any structure left stand- ing must be considered a total loss) SEISMIC SAFETY AND LAND-USE PLANNING FIGURE 37.—Natural hazards map, City of San Jose .-.-.-.-. Sphere of influcncc 0 5000 METERS |———l—# 10,000 FEET B77 B78 The proposed Reconstruction Planning Committee would have the following duties (San Francisco De- partment of City Planning, 1974, p. 63—64): 1. Insure that postearthquake building code and design standards are as advanced in terms of seismic safety as possible. 2. Implement objectives, policies, and criteria of the Comprehensive Plan. 3. Recommend contingency legislation to be enacted now, but taking effect after an earth- quake to authorize such actions as provision of temporary housing. 4. Determine priorities for allocating resources, particularly building materials. 5. Seek joint agreements with lending institu- tions, insurance companies, and Federal dis- aster assistance agencies to require a valid building permit before money for new con- struction is released. 6. Develop an information booklet setting forth all requirements pertinent to reconstruction and sources of financial assistance. EARTHQUAKE PREDICTION Current research in earthquake prediction appears quite promising and has attracted considerable public attention. Although the science of prediction is still in its infancy, many seismologists in the United States believe that within the next decade or two, they will be able to predict at least some earthquakes. However, predictions that specify the location, magnitude, and time of damaging earthquakes with accuracy and enough lead time to take measures to reduce risk are probably a long way off. Credible earthquake prediction may have short-term adverse economic and social impacts, but the pos- sibilities for reducing loss of life, injury, and substan- tial property damage make prediction a worthwhile re— search objective. The value of such predictions is well stated in a recent article by Frank Press (1975, p. 14— 15): Preliminary results of current investigations indicate that predic- tions of strong earthquakes could be made many years in advance. It also appears likely that a method for making short-term predictions, as short as weeks or even days, will be developed. With this dual capability it should become possible to devise a remedial strategy that could greatly reduce casualties and lower property damage. For example, the long-range prediction of a specific event could greatly reduce casualties and lower property damage. For example, the long-range prediction of a specific event could spur the strengthening of existing structures in the threatened area and motivate au- thorities there to enforce current building and land—use regulations and to revise such codes for new construction. A public—education campaign on safety procedures could also be instituted. Short-term prediction could mobilize disaster-relief operations and set in motion procedures for the evacuation of weak structures or particularly flammable or otherwise hazardous areas. The shutdown REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION of special facilities, such as nuclear power plants and gas pipelines, and the evacuation of low-lying coastal areas subject to tsunamis, or "tidal waves,” could also follow a short-term forecast. In brief, long-term earthquake prediction could spur public agencies to take those actions which are recom- mended in seismic safety and emergency preparedness plans. High priorities could be assigned to such meas- ures if an earthquake were predicted, thus substan- tially reducing the risks. Short-term predictions could avert little property damage but could reduce dramat— ically the risk of death or injury. Planning responses to seismic risk are just as, or even more, relevant if accurate prediction becomes possible. The jurisdiction with a development pattern which avoids intensive use of hazardous areas and which provides for sound structural design and con- struction and carefully located and designed emergency response facilities will be well prepared for that inevitable earthquake whether or not it is predicted. CONCLUSIONS Over the past decade, remarkable progress has been made in understanding the nature, cause, and effects of earthquakes. Planners are still developing methods and procedures to make effective use of the data. Methods which more fully integrate safety concerns in land-use planning and decisionmaking can be expected to emerge. Success in seismic safety planning requires public awareness of the nature of seismic risk and the poten- tial for reducing it. Historically, safety has not been an important factor in locating urban settlements, and overcoming apathy toward seismic risk is a formidable task. Typically, the spurt of public interest following each damaging earthquake gradually dwindles away. Recent California seismic safety planning has occurred because of the State law adopted after the San Fer- nando earthquake of 1971. It is to be hoped that these planning efforts will continue and ultimately will re- sult in significant reduction of losses in future earth- quakes. Success in reducing seismic risk requires a com- prehensive program including preparing for disaster response, establishing and enforcing structural design standards, and planning for safe land and building uses. The land—use planning component of the program is particularly important because if seismic hazards are properly recognized in the land-use patterns, disas- ter response and structural design requirements can be correspondingly reduced. Reducing risk through land-use planning requires an interdisciplinary effort involving earth scientists, engineers, and planners. A land-use plan provides the framework for many public actions including land—use and development re- gulations, building code provisions, and project review B79 SEISMIC SAFETY AND LAND-USE PLANNING 2:: 9.0 236m cuE 9.5 *0 Snow a pm am_>_ >tznmum 252mm m>_wm_wm +0 $2: 93 50:3 >ucmm< Coimtoa .mcmC. 5:on .220 3th 9.: am @923 :mEm w Lo» w_nm__m>m PB 2320 xmwmmhm 5 mar: 35 +0 3300 mEBJUoEom Dots—o9. >:m:to: yo: 2 :o_wmm_umm>c_ 050.com 3ng “5:3va mp >9: co_umm_umm>:_ 052me 32:5 00:3wa >:m:.:o: m_ :oCmmtmet 050.9% 3ng ZO_._._ m B80 procedures and criteria. The basic objective of land-use planning for seismic safety is to reduce exposure to seismic hazards by relating land uses to degrees of seismic hazards. In formulating a land-use plan, the social and economic benefits of particular locations for particular uses must be weighed against the costs for structural measures needed to reduce risk to accept- able levels. In developed areas, the value of existing buildings and infrastructure must be weighed against costs of damage and injuries from earthquakes. In the future, more effective land-use planning to reduce seismic risk will be possible as more accurate and detailed information becomes available and plan- ners become more experienced in its application. Maps showing where ground shaking, slope instability, and liquefaction may occur will be particularly useful to land—use planners. Further research in structural re- sponse to seismic forces is needed to develop more realistic building code requirements. The relative costs of applying various risk reduction measures are usually unknown but are greatly needed. In addition, the public and private costs associated with various land uses and structural types in hazard- ous areas need to be studied. Legal mechanisms and funds are required to help reduce existing structural hazards. If these additional tools are provided, local public agencies in cooperation with knowledgeable citizens and decisionmakers will be well equipped to plan for seismic safety. REFERENCES Alfors, J. T., Burnett, J. L., and Gay, T. E., Jr., 1973, Urban geology master plan for California: California Div. Mines and Geology, Bull. 198, 112 p. Algermissen, S. T., 1969, Seismic risk studies in the United States in Proceedings of the fourth world conference on earthquake en- gineering: Santiago, Chile, v. 1, p. 19—27. ——(principal investigator), 1972, A study of earthquake losses in the San Francisco Bay area: Natl. Oceanic Atmospheric Admin, US. Dept. Commerce, 220 p. Armstrong, Dean, 1973, The seismic safety study for the general plan: Sacramento, Calif, California Council on Intergov- ernmental Relations, 199 p. Association of Bay Area Governments, 1973 Procedures for regional clearinghouse review of environmental impact statements— Phase two: Assoc. Bay Area G0vts., Berkeley, Calif, 225 p. 1975a, Areas of critical environmental concern. Review draft: Assoc. Bay Area Govts., Berkeley, Calif, 83 p. 1975b, Hazards evaluation for disaster preparedness plan- ning: Assoc. Bay Area Govts., Berkeley, Calif, 37 p. 1976, Earthquake preparedness—ideas for action: Assoc. Bay Area Govts., Berkeley, Calif. unpaged. Association of Engineering Geologists, 1973, Geology, seismicity, and environmental impact: Special Publication, Los Angeles, University Publishers, 446 p. Ayre, R. S., 1975, Earthquake and tsunami hazards in the United States—A research assessment: Natl. Sci. Found.—RA—E—75— 005, Univ. Colorado, Boulder, 149 p. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION Baker, L. 0., Jr., 1971, Availability and desirability of earthquake insurance in earthquake risk: Conference proceedings Sept. 22—24, 1971, sponsored by the Joint Committee on Seismic Safety to the California Legislature, report of the Special Sub- committee to the Joint Committee on the San Fernando Earth- quake Study, p. 31—33. Borcherdt, R. D., ed., 1975, Studies for seismic zonation of the San Francisco Bay region—Basis for reduction of earthquake hazards, San Francisco Bay region, California: US. Geol. Sur- vey Prof. Paper 941—A, 102 p. Borcherdt, R. D., Gibbs, J. F., and Lajoie, K. R., 1975, Maps showing maximum earthquake intensity predicted in the southern San Francisco Bay region, California, for large earthquakes on the San Andreas and Hayward faults: U.S. Geol. Survey Misc. Field Studies Map MF~709, 3 sheets, scale 1:125,000. Brown, R. D., Jr., 1972, Active faults, probable active faults, and associated fracture zones, San Mateo County, California: US. Geol. Survey Misc. Field Studies Map MF—355, scale 1:62,500. Burkland and Associates, 197 5, Geological and geophysical investi- gation in downtown Hayward: Burkland and Assoc, Mountain View, Calif, 22 p. and appendix. California Council on Intergovernmental Relations, 1973, Guidelines for local general plans, Sacramento: Calif. Council on Intergovernmental Relations, Sacramento, 106 p. California Division of Highways, 1971, The effectton state highways of the San Fernando earthquake February 9, 1971: Calif. Div. Highways, Sacramento, Calif, 62 p. and appendixes. 1975, Highway design manual of instructions: Calif. Div. Highways, Sacramento, Calif, Section 7—110. California Division of Mines and Geology, State Mining and Geology Board, 1975, Policies and criteria of the State Mining and Geol- ogy Board with reference to the Alquist-Priolo Special Studies Zones Act, 3 p. California Education Code, 1933, secs. 15451—15465. 1967, sec. 15002, Field Act. California Government Code, 1965, California Land Conservation Act of 1965 (Williamson Act): California Govt. Code, secs. 51200—51295. 1970, sec. 65302, Emergency Services Act. 1973, sec. 8589—5, Alquist Dam Safety Act. 1974, secs. 8890—8899, Seismic Safety Commission Act. California Governor’s Earthquake Council, 1972, First report of the Governor’s earthquake council: Sacramento, Calif, 65 p. California Office of Planning and Research, 1973, Summary re- port—Environmental goals and policy: Sacramento, Calif, 29 p. California Public Resources Code, 1972, secs. 2621—2625, Alquist/ Priolo Special Studies Zones Act. 1976, California Coastal Act: secs. 30000 to 30900. California Resources Agency, 1974, Amendments to guidelines for implementation of the California Environmental Quality Act, December 1974, secs. 15001—15163, 23 p. California State Legislature, 1963, California Health and Safety Code, California Community Redevelopment Act, secs. 33000— 33738. Joint Committee on Seismic Safety, Gates, George 0., ed. 1972, The San Fernando earthquake of February 9, 1971 and public policy: Calif. Joint Comm. on Seismic Safety, Sacramento, Calif, 127 p. California State Legislature, Joint Committee on Seismic Safety, 1974, Meeting the earthquake challenge, final report to the Legislature, January 1974: Calif. Div. Mines and Geology, Spe- cial Publication 45, 223 p. California State Senate, 1969, Concurrent resolution no. 128 of 1969. Chinnery, M. A., and North, R. G., 1975, The frequency of very large earthquakes: Science, v. 190, no. 4220, p. 1197—1198. SEISMIC SAFETY AND LAND-USE PLANNING Cooper, Clark, and Associates, 1974, Technical report, geotechnical investigation, city of San Jose’s sphere of influence for the city of San Jose: Cooper, Clark, and Associates, San Jose, Calif., 233 p. and appendix. Council on Environmental Quality, 1973, Preparation of environ- mental impact statements—guidelines: 38 Fed. Reg, August 1, 1973, p. 20562. Dewey, J. F., 1972, Plate tectonics: Scientific American, v. 226, no. 5, p. 56—66. Dickinson, W. R., 1970, Commentary and reconnaissance photo- geologic map, San Andreas rift belt, Portola Valley, California: Town of Portola Valley, Calif., public document, 51 p. Duncan & Jones, 1974, Seismic safety element, San Jose, California: Duncan and Jones, Berkeley, Calif., 48 p. Elders, W. A., 1974, ”Is there an earthquake in your future?”, Cry California, Jour. of California Tomorrow, v. 9, no. 2, p. 20—25. Hayward, 1972, Hayward earthquake study: Planning Commission Subcommittee on Land Use and Development Regulations, Hayward, Calif., 50 p. 1975a, Redevelopment plan, Downtown Hayward redevelop- ment project: Redevelopment Agency of the City, Hayward, Calif., 24 p. 1975b, Preliminary action proposal for downtown revitaliza- tion: Executive Comm. of the Downtown Planning Team, Hay- ward, Calif., 30 p. International Conference of Building Officials, 1976, Uniform build- ing code: Whittier, Calif., lnternat. Conf. Bldg. Officials, 728 p. Interprofessional Council of Environmental Design, 1974, Guide to interprofessional collaboration in environmental design. Jennings, C. W., 1975, Fault map of California with locations of volcanoes, thermal springs and thermal wells: California Div. Mines and Geology, Geologic Data Map No. 1, scale 1:750,000. Laird, R. T., Perkins, J. B., Bainbridge, D. A., Baker, J. B., Boyd, R. T., Huntsman, Daniel, Zucker, M. B., and Staub, Paul, 1979 Quantitative Land Capability Analysis, US. Geol. Survey Prof. Paper 945, 115 p. Lakshmanan, T. R., 1972, Integration of physical and economic planning: Prepared for United Nations Centre for Housing Building and Planning, New York, N. Y., 48 p. Livingston and Blayney, 1974, Santa Barbara County Comprehen- sive Plan, seismic safety element: Livingston and Blayney, City and Regional Planners, San Francisco, Calif., 93 p. Mader, G. G., 1974, “Earthquakes, landslides and public planning”, Cry California, Jour. of California Tomorrow, v. 9, no. 3, p. 16—22. Metropolitan Transportation Commission, 1974, Regional transportation plan: Metropolitan Transp. Comm., Berkeley, Calif., 126 p. and appendix. Monterey County, 1975, County-wide seismic safety element for the General Plans of the county of Monterey and participating municipalities: Prepared by William Spangle and Associates, Monterey County, Monterey, Calif., 50 p. Nichols, D. R., and Buchanan-Banks, J. M., 1974, Seismic hazards and land-use planning: U.S. Geol. Survey Circ. 690, 33 p. Nilsen, Tor, and Wright, R. H., 1979, Relative slope stability and land-use planning: selected examples from San Francisco Bay region, California: US. Geol. Survey Prof. Paper 944, 96 p. Palo Alto, City' of, 1976, Palo Alto comprehensive plan: Palo Alto, Calif., 62 p. (in press). Planning Policy Committee of Santa Clara County, 1972, A policy plan for the baylands of Santa Clara County: County of Santa Clara Planning Dept., San Jose, Calif., 76 p. Press, Frank, 1975, Earthquake prediction: Scientific American, May 1975, v. 232, no. 5, p. 14—23. Ritter, J. R., and Dupre, W. R., 1972, Maps showing areas of poten- B81 tial inundation from tsunamis in the San Francisco Bay Region: U.S. Geol. Survey Misc. Field Studies Map MF—480. San Diego County, 1975, Seismic safety element, San Diego County general plan: San Diego, Calif., 13 p. and appendixes. San Francisco Bay Conservation and Development Commission, 1969, San Francisco Bay plan: San Francisco Bay Conserv. and Devel. Comm., San Francisco, Calif., 43 p. and appendix. 1974a, Bay plan evaluation project, safety of fills: San Fran- cisco Bay Conserv. and Devel. Comm., San Francisco, Calif., 23 p. 1974b, Bay plan evaluation project, final report—Guidelines for Bay plan revision: San Francisco Bay Conserv. and Devel. Comm., San Francisco, Calif., 24 p. San Francisco Department of City Flaming, July 1974, A proposal for citizen review, community safety, the comprehensive plan of San Francisco, Dept. of City Flaming, San Francisco, Calif. 69 p. San Jose, City of, 1976, The general plan 1975: San Jose, Calif., 57 p. and land use/transportation diagram. San Mateo County, 1975, City-county planning task force, seismic and safety elements of the General Plan, v. 1, Redwood City, Calif., 95 p. Santa Clara County, 1975, Seismic safety plan, An element of the general plan: San Jose, Calif., 75 p. and appendixes. Santa Clara County Board of Supervisors, 1974, San Jose, Calif., Ordinance No. NS— 1203.31, 11 p. Santa Clara County Flaming Policy Committee, 1972, A policy plan for the baylands of Santa Clara County: Santa Clara Co. Plan- ning Dept., San Jose, Calif., 76 p. Scott, Nina H., 1973, Felt area and intensity of San Fernando earth- quake in San Fernando, California, Earthquake of February 9, 1971, v. III, Geological and Geophysical Studies: Natl. Oceanic and Atmospheric Ad., US. Dept. Commerce, p. 23-48. Spangle, William and Associates and F. Beach Leighton and Associ- ates, and Baxter, McDonald and Smart, Inc., Feb. 22, 1974, Ap- plication of earth science information in urban land-use plan- ning, state-of-the-art review and analysis: Natl. Tech. Inf. Ser- vice, U.S. Dept. Commerce, Springfield, Va. (NTIS PB 238— 081/AS), 330 p. Steinbrugge, K. V., Schader, E. E., Bigglestone, H. C., and Weers, C. A., 1971, San Fernando earthquake February 9, 1971: Pacific Fire Rating Bureau, San Francisco, Calif., 93 p. Strahler, A. N ., 1971, The earth sciences: New York, Harper and Row, 824 p. US. Civil Defense Preparedness Agency, 1974, Assumptions, pre- cepts, and objectives, program paper: Civil Defense Prepared- ness Agency, v. 1, 32 p. US. Congress, 1968, Intergovernmental cooperation act of 1968: Public Law 90—577. 1970, National environmental policy act of 1969: Public Law 91—190, 5 p. 1972, Federal coastal zone management act of 1972: Public law 92—583. 1973, Flood disaster protection act of 1973: Public Law 93— 234, 9 p. 1974, Disaster relief act of 1974, Public Law 93—288, 22 p. US. Department of Housing and Urban Development, 1973, Minimum property standards for multifamily housing: Wash- ington, D.C., v. 2, unpaged. 1975a, Comprehensive planning assistance; General applica- bility: Fed. Reg, v. 40, no. 164, p. 36856—36865. 1975b, Environmental review procedures for community de- velopment block grant program: Fed. Reg, v. 40, no. 4, Part II, January 7, 1975, Washington, DC, p. 1392—1399. US. Federal Insurance Administration, 1975, National flood insur- B82 ance program, proposed criteria, Part 11: Washington, D.C., Fed- eral Register, March 26, 1975, v. 40, no. 59, p. 13420—13433. U.S. Geological Survey, 1971, Earthquakes: Washington, D.C., U.S. Govt. Printing Office, 19 p. U.S. Geological Survey and U.S. National Oceanic and Atmospheric Admi istration, 1971, The San Fernando, California, earth- quak of February 9, 1971: U.S. Geol. Survey Prof. Paper 733, 254 p. U.S. National Oceanic and Atmospheric Administration, 1975, Coastal zone management program administrative grants, Notice of final rulemaking: Fed. Reg, v. 40, no. 6, part I, p. 1687. U.S. Office of Emergency Preparedness, 1971, X-day + 100, the Fed- eral response to the California earthquake of February 9, 1971, an interim report as of May 20, 1971, in Governmental response to the California earthquake disaster of February 1971, Hear- ings before the Senate Committee on Public Works: U.S. 92d Cong, Calif, Serial No. 92—H22, 21 p. 1972, Disaster preparedness, report to Congress: U.S. Govt. Printing Office 4102—0006, 3 vols. 353 p. U.S. Office of the Federal Register, 1973, United States Government Manual ,1973/74: U.S. Govt. Printing office, 794 p. U.S. Office of Management and Budget, 1973, Evaluation, review. REDUCTION OF EARTHQUAKE HAZARDS, SAN FRANCISCO BAY REGION and. coordination of Federal and Federally assisted programs and projects: Washington, D.C., Govt. Printing Office, Circ. A—95 (rev.), Federal Register, V. 38, no. 228, p. 32874—32881. U.S. Office of Science and Technology, 1970, Earthquake hazard re- duction, Report of the task force on earthquake hazard reduc- tion: U.S. Govt. Printing Office, 54 p. U.S. Senate, 1971, Hearings before the Committee on Public Works, Governmental response to the California earthquake disaster of February 1971: U.S. Govt. Printing Office, 985 p. Wallace, R. E., 1970, Earthquake recurrence intervals on the San Andreas Fault, Geol. Soc. Am. Bull, v. 81, no. 10, p. 2875—2890. Wood, H. 0., 1908, Distribution of apparent intensity in San Fran- cisco, in The California earthquake of April 18, 1906, Report of the State Earthquake Investigation Commission: Carnegie Inst. Washington Pub. 87, p. 220~245. Yanev, Peter, 1974, Peace of mind in earthquake country: San Fran- cisco, Chronicle Books, 304 p. Youd, T. L., Nichols, D. R., Helley, E. J., and Lajoie, K. R., 1975, Liquefaction potential, in Studies for seismic zonation of the San Francisco Bay region, California: U.S. Geol. Survey Prof. Paper 941-A, p. A68—A74. [U \ II II v?" f o > 3430 We