, Equrvalent Uranlum and Selected ManI‘ Elements 1n 1 1, 1;; ‘ , Magneuc Concentrates from the i Candle Quadrangle ' W ‘ ” Solomon Quadrangle and Elsewhere 1n Alaska ” GEO Lee I C A L7 suéEy E y 71> R Q FT: 3 s I 0 N‘A’VL :5 APE R 11175 ‘ 1‘ V ‘ ‘ ”-3? ::QEPC§SITO§Y :1 1 HAJUG 225519811 ~ SEP 03198Q 1 ‘ “Emsctsncufii Equivalent Uranium and Selected Minor Elements in Magnetic Concentrates from the Candle Quadrangle, Solomon Quadrangle, and Elsewhere in Alaska By KUO-LIANG PAN, WILLIAM C. OVERSTREET, KEITH ROBINSON, ARTHUR E. HUBERT, and GEORGE L. CRENSHAW GEOLOGICAL SURVEY PROFESSIONAL PAPER 17135 An evaluation of magnetic concentrates as a medium for geochemical exploration in artic ana' subartic regions UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTONzl980 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Cataloging in Publication Data Main entry under title: Equivalent uranium and selected minor elements in magnetic concentrates from the Candle quadrangle, Solomon quadrangle, and elsewhere in Alaska. (Geological Survey Professional Paper 1135) Bibliography: p. 103 Supt. of Docs. no.:119.16:1135 1. Geochemical prospecting—Alaska. 2. Ore-deposits—Alaska. I. Pan, Kuo-liang. II. Title: Magnetic concentrates from the Candle quadrangle, Solomon quadrangle, and elsewhere in Alaska. 111. Series: United States Geological Survey Professional Paper 1135. TN270.E78 622'.13’o9793 79—607131 For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 CONTENTS Abstract Introduction Purpose of investigation ----------------------------------------- Acknowledgments Magnetic fractions of concentrates ------------------------------ —-- Source and preparation Location and distribution ----------------------------- Randomization Analytical procedures and reliability of the chemical data ----- Control samples Equivalent uranium Eight elements by atomic absorption --------------------- —- Tests of procedure Internal replicate analyses in the full data set ------------ Gold, indium, and thallium by atomic absorption ------- Distribution of the elements -------------------------------------------- Background review Regional results Areas discussed Statistical treatment ---------------------------------------------- Equivalent uranium ----—- ------------------------------- Abundance Mineralogical sources ----------------------------------------- — Possible use Copper. lead, zinc. and cadmium —— -------- - Southeastern Alaska -------------- Southern Alaska ----------- - ----- Southwestern Alaska ------------------------------------ West-central Alaska ----------------------------------- East-central Alaska ------------------------------------------- - Silver and gold Southeastern Alaska --------------------------------------- Southern Alaska ----------------- Southwestern Alaska ----- West-central Alaska -- ---------- - East-central Alaska --------------------------------- Bismuth West-central Alaska - ------------------------------------------- East-central Alaska -------------------------------- - Cobalt and nickel Southern Alaska --------------------------------------------------- Southwestern Alaska --------------------------------- West-central Alaska -- East-central Alaska -- - Indium and thallium ------- Southern Alaska ------ «- Southwestern Alaska -- West-central Alaska ----------------------------------------------- East-central Alaska -- ----------------------------------------- Candle quadrangle results ------ Equivalent uranium ----------------------------------------------------- Page WWCDWCDCDWNNI—l Page Distribution of the elements—Continued Candle quadrangle results—Continued Copper, lead. zinc, and cadmium ------------------------ —- 51 Silver and gold 58 Bismuth 59 Cobalt and nickel 59 Indium and thallium ---——--—--—--—-—-—- ------------------ 64 Solomon quadrangle results -—---—————-—------- ----- ....... 64 Equiyalent uranium ------------------------------- -— 65 Copper, lead. zinc, and cadmium — -------------------------- - 69 Silver and gold 77 Bismuth 77 Cobalt and nickel 80 Indium and thallium -—-—--- --------------------------------- 83 Sources of the anomalous elements -..-_.__........._....... ------- 83 Mineralogical composition of magnetic concentrates —-—- 83 Magnetite, ilmenite, and rutile -------------------------------- 86 Hematite 86 Sulfide minerals and gold -----—------—---—-—--—-—--—-- 86 Quartz and common silicate minerals -------------------- 87 Other minerals 87 Metallic spherules and tramp iron mmmmmmm 87 Mineralogical composition of insoluble residues ~--—----------- 89 Frequency of association of anomalous amounts of elements with specific minerals - --------------- - 89 Copper, lead. and zinc -------------------------- ~- 92 Silver 92 Bismuth and cadmium -—--- --------------------------------- 92 Cobalt and nickel 92 Equivalent uranium -------------------------------------- -« 94 Gold, indium. and thallium ------------------------------------------- 94 Role of anomalous environments ----------------------------- 94 Relations among the elements ----------------------------- —-- 95 Correlations 95 Regional 95 Candle quadrangle 96 Solomon quadrangle ---------------------------------------- 97 98 Copper and silver in southeastern Alaska --------------- - 99 Multielement highs in southern Alaska -— ------------- 99 Base metals in southwestern Alaska ------------------------- 99 West-central Alaska — ------------------------------------------ 99 Equivalent uranium in the Bendeleben, Candle. and Solomon quadrangles ----------------------------- 99 Lead, cobalt, bismuth. and other elements ---------- 100 Silver and gold in east-central Alaska ------------------- 100 Geologic and‘geochemical interpretation -------------------- «- 100 Regional 100 Candle quadrangle 102 Solomon quadrangle 102 Selected references 103 III IV CONTENTS ILLUSTRATIONS Page PLATE 1. Map of Alaska showing generalized outlines of areas represented by magnetic fractions of concentrates and topographic quadrangle maps used in this study In pocket FIGURES 1-10. Concentrations and cumulative frequencies of equivalent uranium and minor elements in Alaskan magnetic concentrates: 1. Equivalent uranium 24 2. Copper 25 3. Lead 26 4. Zinc 26 5. Cadmium 26 6. Silver 27 7. Gold 28 8. Bismuth 28 9. Cobalt 28 10. Nickel 29 1 1—20. Histograms for equivalent uranium and minor elements in Alaskan magnetic concentrates: 11. Equivalent uranium 30 12. Copper 32 13. Lead 32 14. Zinc 33 15. Cadmium 33 16. Silver 37 17. Gold 37 18. Bismuth 39 19. Cobalt 41 20. Nickel 41 21—22. Geologic maps showing the distribution of analyzed magnetic concentrates from the Candle quadrangle. Alaska: 21. Northwestern part 48 22. Southwestern part 50 23—30. Maps showing equivalent uranium and minor elements in magnetic concentrates from the Candle quadrangle, Alaska: 23. Equivalent uranium and bismuth. northwestern part 54 24. Equivalent uranium and bismuth. southwestern part . 55 25. Copper, lead, zinc, and cadmium. northwestern part 56 26. Copper, lead. zinc. and cadmium, southwestern part 57 27. Silver, gold. indium, and thallium. northwestern part 60 28. Silver, gold, indium, and thallium, southwestern part 61 29. Cobalt and nickel, northwestern part 62 30. Cobalt and nickel. southwestern part 63 31—32. Geologic maps showing the distribution of analyzed magnetic concentrates from the Solomon quadrangle, Alaska: 31. Northeastern part ' 66 32. East-central part 68 33—34. Maps showing equivalent uranium and bismuth in magnetic concentrates from the Solomon quadrangle, Alaska: 33. Northeastern part 72 34. East-central part 73 35. Graph showing relation between equivalent uranium in original heavy-mineral concentrates and equivalent uranium in their magnetic fractions in Alaska 74 36-41. Maps showing minor elements in magnetic concentrates from the Solomon quadrangle, Alaska: 36. Copper, lead. zinc. and cadmium. northeastern part 75 37. Copper, lead. zinc, and cadmium, east-central part 76 38. Silver, gold. indium, and thallium, northeastern part 78 39. Silver, gold. indium, and thallium, east-central part 79 40. Cobalt and nickel, northeastern part 81 41. Cobalt and nickel, east-central part 82 TABLE 1. 2. 3. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. . Minor elements in a particle of placer gold from Solomon Gulch in the Ruby quadrangle, Alaska ------------------------------------------- 24. 25. 26. 27: Regional threshold values for eight elements in Alaskan magnetic concentrates compared to possible source minerals -------- 29. 30. 31. 32. 33. 34. 35. 36. . Selected references on the composition of magnetite . Physical properties of 1 1 trace elements in magnetic concentrates from Alaska, compared to the same properties of ferrous CONTENTS TABLES Locations and results of analyses of 347 magnetic concentrates from Alaska, related to known mineral deposits Replicate analyses of upsplit and unsized control sample of magnetic concentrates from Hampton Creek placer, Nevada, for Ag, Bi, Cd, Co, Cu, Ni, Pb, and Zn Replicate analyses of upsplit but sized control sample 300DL of magnetic concentrates from Hampton Creek placer, Nevada, for Ag, Bi, Cd, Co, Cu, Ni, Pb, and Zn . Replicate analyses of split but unsized control sample of magnetic concentrate from Lyle Creek, North Carolina, for Ag, Bi, Cd, Co, Cu, Ni, Pb, and Zn . Replicate analyses of split but unsized subsamples of file number 3799 from Portage Creek placer, Alaska, for Ag, Bi, Cd, Co, Cu, Ni, Pb, and Zn and ferric iron in magnetite . Statistical summaries of the regional geochemical data for 347 magnetic concentrates and of the data for the Candle and Solomon quadrangles, Alaska . Background and threshold values for equivalent uranium and nine elements in magnetic concentrates from Alaska . Comparison of equivalent uranium in magnetic concentrates and original source concentrates in the Circle, Ketchikan Medfra, and Mount Hayes quadrangles, Alaska Equivalent uranium in magnetite and in hematitic coatings on the magnetite, Jump Creek placer, Bendeleben quadrangle, Alaska Values used to plot equivalent uranium and 11 elements in magnetic concentrates from 34 localities in the Candle quadrangle, Alaska Variation in chemical composition of multiple magnetic concentrates from five areas represented by single plotted localities in the Candle quadrangle, Alaska Relative standard deviations, in percent, for eight elements in five sets of samples from single plotted localities in the Candle quadrangle, Alaska, compared to relative standard deviations for subsamples of file number 3799 ---------------------- Average abundances of copper, lead, zinc, and cadmium in the crustal materials of the earth compared with their abundances in magnetic concentrates from the Candle quadrangle, Alaska Cobalt, nickel, and Co/Ni ratios in magnetic concentrates from various geologic provenances in the Candle quadrangle, Alaska Values used to plot equivalent uranium and 11 elements in magnetic concentrates from 41 localities in the Solomon quadrangle, Alaska Variations in chemical composition of multiple magnetic concentrates from six areas represented by single plotted localities in the Solomon quadrangle, Alaska Relative standard deviations, in percent, for eight elements in six sets of samples from single plotted localities in the Solomon quadrangle, Alaska, compared to relative standard deviations for subsamples of file number 37 99 ------------------- Comparison of equivalent uranium in magnetic concentrates and original source concentrates from the Solomon quadrangle, Alaska Appearance, metal content, and mineralogical composition of a subset of 67 magnetic concentrates from Alaska Minor elements in hand-picked magnetite from Alaskan placers Sources of magnetic concentrates containing metallic spherules and tramp iron, Alaska Results of laser-probe analyses of hand-picked metallic spherules from placers in Alaska Results of semiquantitative spectrographic analyses of tramp iron from a placer concentrate from lower Rhode Island Creek, Tanana quadrangle, Alaska Frequency of association of specific minerals with anomalous element content in magnetic concentrates from Alaska --------- Residuals of association of elements present in anomalous amounts with specific minerals in magnetic concentrates from Alaska Relations between values for metals in normal and abnormal magnetic concentrates and between normal magnetic concentrates and hand-picked magnetite from the abnormal concentrates, Alaska Correlation coefficients of the logarithms of the concentrations of equivalent uranium and 11 elements, and number of pairs, in 347 magnetic concentrates from Alaska Correlation coefficients of the logarithms of the concentrations of equivalent uranium and 11 elements, and number of pairs, in 85 magnetic concentrates from the Candle quadrangle,\Alaska Correlation coefficients of the logarithms of the concentrations of equivalent uranium and 11 elements, and number of pairs, in 101 magnetic concentrates from the Solomon quadrangle, Alaska Summary of probable modes of occurrences elements present in anomalous amounts in magnetic concentrates from Alaska Number and type of known mineral deposits or occurrences located near anomalous magnetic concentrates in Alaska -------- Number and type of polymetallic deposits or occurrences located near anomalous magnetic concentrates in Alaska ------------ Page 4 17 18 18 19 20 22 24 29 30 3 1 52 53 53 58 64 70 7 1 7 1 74 84 86 87 88 89 89 90 90 91 93 96 97 98 100 101 102 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS IN MAGNETIC CONCENTRATES FROM THE CANDLE QUADRANGLE, SOLOMON QUADRANGLE, AND ELSEWHERE IN ALASKA By KUO-LIANG PAN, WILLIAM C. OVERSTREET, KEITH ROBINSON, ARTHUR E. HUBERT, and GEORGE L. CRENSHAW ABSTRACT Equivalent uranium and 11 minor elements in the magnetic frac- tions of 347 panned concentrates from the Candle and Solomon quadrangles, Alaska, and elsewhere in the State, were determined in the first investigation bythe US. Geological Survey of the utility of magnetic concentrates as a geochemical sample medium for sub- arctic and arctic regions. The magnetic concentrates were obtained from the Survey’s Alaskan placer concentrate file. Magnetic sep- aration from the nonmagnetic parts of the panned concentrates was done in randomized order, as were radiometric and atomic absorp- tion analyses. The elements determined by atomic absorption are silver, bismuth, cadmium, cobalt, copper, nickel, lead, zinc, gold, indium, and thallium. Replicate analyses of standard samples were made as a check on the analytical results. Sixty-seven of the mag- netic concentrates were examined by microscope and by X-ray dif- fraction to determine their mineralogical composition and to deter- mine what parts of the concentrates were taken into solution during analysis. Most of the magnetic concentrates contain more than 50 percent magnetite, and in 60 percent of the samples the magnetic con- centrates consist of 90—99 percent magnetite. Many of the samples, especially the few that contain less than 50 percent magnetite, are diluted by ilmenite, rutile, sulfide minerals, gold, quartz, hematitic coatings, silicate minerals, metallic spherules, and tramp iron. Because of intergrowths between minerals. coatings, and the mech- anical trapping of nonmagnetic grains in clusters of magnetic grains, no effective method is available for obtaining monomineralic detrital magnetite as a geochemical sample medium. Metallic spherules are present in seven samples, five of which also contain slivers of tramp iron. Both the spherules and the tramp iron adhere to grains of detrital magnetite by ferromagnetic attraction, or are cemented to the magnetite by hematite or limonite, or are present as loose intergranular particles. The tramp iron is derived from machinery used in placer mining, and the metallic spherules may originate as welding spatter or fly ash from mining activities. However, the metallic spherules may have formed in several natural ways: they may be fusion products from volcanic activity, lightning, or forest fires; or they may in part be extraterrestrial material such as meteoric dust or ablation products from iron meteorites. This last possibility adds immensely to the scientific interest generated by the metallic spherules. Chemical digestion of the magnetic concentrates in preparation for analysis by atomic absorption was not complete. Magnetite was largely but not entirely digested. Ilmenite and rutile were slightly leached on the surface. Hematitic coatings, sulfide minerals, and gold were completely dissolved. No effect was noticed on quartz and the common silicate minerals except that they attained a high gloss indicative of the removal of surface coatings. The metallic spherules and tramp iron were taken into solution. Thus, the part of the magnetic concentrate that went into solution was magnetite. hema- tite and other coatings, sulfide minerals, native gold, metallic spherules, and tramp iron. The minor elements contained in the magnetic concentrates are present in substitution for major elements in magnetite, sorbed on the surface of magnetite, in trace minerals included in the magnetite, and in accessory minerals trapped among grains of magnetite. The most probable sources of anomalous amounts of minor elements in these concentrates are: (1) silver, copper, lead, zinc, cobalt, and nickel substituted for iron in the magnetite struc- ture; (2) equivalent uranium, copper, lead, and zinc held in surface sorption on magnetite; (3) copper, cadmium, indium, and thallium in trace minerals; and (4) equivalent uranium, silver, bismuth, cad- mium, copper, gold, indium, and thallium in accessory minerals. Vir- tually all measured equivalent uranium comes from hematitic coatings that constitute an accessory mineral in the concentrates. High values for the elements that substitute for iron in the magnetite structure are increased by accessory particles of sulfide minerals, native gold, metallic spherules, and tramp iron. Native gold is the source of much of the anomalous gold and silver. The least understood aspect of minor elements in magnetite is the role of surface sorption of the elements. The enrichment of trace elements in detrital magnetite or in mag- netic concentrates does not necessarily mean that the source rocks have a superior economic potential. However, anomalously high contents of minor elements are guides to areas deserving further exploration. For example, magnetites with radioactive hematitic coatings clearly identified areas of alkalic intrusive rocks containing abnormal amounts of thorium and uranium. Of 36 previously known copper deposits or occurrences, 24 yielded magnetic concentrates containing anomalous amounts of copper, and the other 12 provided magnetic concentrates with various anomalous lead, zinc, silver, cobalt, or nickel contents. Nineteen previously known lead-bearing deposits or occurrences were the sources for eight magnetic concen- trates with anomalous lead content, and all 19 of these deposits were reflected by various combinations of anomalous amounts of base metals, silver, or gold. Ten of the thirteen areas previously known to have zinc mineralization were the sources of magnetic concentrates with anomalous zinc content, and the others had anomalous amounts of other metals. Magnetic concentrates from polymetallic deposits seldom failed to have anomalous trace- 1 2 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA element contents. Therefore, magnetic concentrates may be used satisfactorily as a geochemical sample medium in subarctic and arctic environments. Anomalous amounts of copper and zinc in magnetic concentrates indicate sulfide mineralization. Where anomalous amounts of cop- per and zinc are combined with anomalous amounts of silver. bismuth, and lead in magnetic concentrates. polymetallic sulfide deposits are indicated. Anomalous silver content usually indicates silver and gold deposits, mainly gold. Anomalous cobalt and nickel content indicates the presence of chromite and, locally, sulfide deposits associated with mafic and ultramafic rocks. Anomalous lead and gold content usually indicates lead sulfide deposits and gold. but not all lead and gold deposits have corresponding anomalous values for lead and gold in the magnetic concentrates. The abundances of these two elements are more affected by chance in collecting and in preparing the sample than are the previously cited elements. However, lead and gold deposits can be detected from anomalous amounts of copper, zinc, and silver in magnetic concentrates. Of interest is the number of tungsten deposits or occurrences that yield magnetic concentrates containing anomalous amounts of copper. This geochemical association is one that requires ap- propriate follow-up investigations to determine if skarn-type tungsten-copper deposits or porphyry-type deposits may be related to the presence of copper in magnetic concentrates. Several prominent regional geochemical highs are indicated by the varied distribution of the anomalous amounts of elements in magnetic concentrates from Alaska. Large copper and silver anomalies are present in the Ketchikan quadrangle in southeastern Alaska. In southern Alaska, high values are found for copper, zinc. and gold in the McCarthy quadrangle. Copper and silver concentra- tions are anomalous in the Valdez quadrangle, and anomalous zinc content is found in magnetic concentrates from the Anchorage, Talkeetna, and Talkeetna Mountains quadrangles. Cobalt and nickel attain high values in the Mount Hayes quadrangle. Magnetic concentrates from the Bethel and Iliamna quadrangles in southwestern Alaska contain anomalous amounts of base metals. In west-central Alaska, anomalous equivalent uranium is associated with magnetic concentrates from the Bendeleben. Candle, and Solomon quadrangles. Extremely high values for lead, cobalt. bismuth, and other metals are found in samples from the Ruby quadrangle. Multielement highs, indicating various combinations of silver, bismuth, copper, nickel, and zinc, are obtained from magnetic concentrates from the Iditarod and Teller quadrangles. and anomalous amounts of cobalt and nickel are found in the Bendeleben quadrangle. West-central Alaska appears to be a bismuth province. In east-central Alaska, high values for silver and gold are common in magnetic concentrates from the Circle, Eagle, Livengood. and Tanana quadrangles. Nickel and zinc are enriched in samples from the Livengood quadrangle, and bismuth attains a high value in one concentrate from the Circle quadrangle. INTRODUCTION PURPOSE OF INVESTIGATION The use of magnetic fractions of panned concen- trates—called here the magnetic concentrate—for a geochemical sample medium has been tested by the US. Geological Survey in the humid temperate zone (Theobald and Thompson, 1959; Theobald and others, 1967) and the arid tropics (P. K. Theobald, Jr., oral commun., 1971). The magnetic fraction of coarse- grained alluvium has been similarly used for geochem- ical exploration in humid, tropical, central Ecuador (de Grys, 1970), as well as in the subarctic glaciated region around Churchill Falls, Labrador (J. E. Calla- han, written commun., 1973; 1974). The trace-element distribution in accessory magnetites from quartz mon- zonite stocks in the Basin and Range Province of Utah and N evada has been studied for its relation to sulfide mineralization (Hamil and N ackowski, 1971). An un sual opportunity to test the possible value of the magnetic concentrate as a geochemical sample medium i the subarctic and arctic environments was afforded y an investigation initiated in 1970 by C. L. Sainsbury, US. Geological Survey. His interest was the minor elements in the nonmagnetic fractions of heavy minerals from the Alaskan placer concentrate file. N onmagnetic fractions were prepared in 1971 by W. R. M rsh from 1,072 of the 5,000 concentrates in the file. A byproduct of 682 magnetic fractions resulted, which W. C. Overstreet thought should be analyzed ‘both to continue the Survey’s research on minor elements in magnetite, and to provide a compar- ison with the results of the spectrographic analyses of the nonmagnetic concentrates (Hamilton and others, 1974). Two avenues for the analysis of these magnetic con- centrates were available: (1) a standard semiquan- titative spectrographic procedure for 30 elements, and (2) an analytical method developed in the US. Geolog- ical Survey by H. M. Nakagawa (1975) to determine the abundances of silver, bismuth, cadmium, copper, cobalt, nickel, lead, and zinc by atomic absorption techniques on single solutions of iron-rich materials. The spectrographic procedure requires only 10—20 mg, of sample; thus, all 682 magnetic concentrates could be so analyzed (Rosenblum and others, 197 4). One gram is needed for the method employing atomic absorption. Many of the Alaskan concentrates were lean in magnetite, so only 347 magnetic fractions were large enough for study by atomic absorption. However, several other tests could be carried out on the large samples of magnetic concentrates: (1) equivalent uranium by radiometric counting; (2) gold by atomic absorption; and (3) mineralogical examination. There- fore, the atomic absorption analytical method was selected. An analytical procedure developed in the Survey (Hubert and Lakin, 1973) made possible the analyses for indium and thallium on part of the set of magnetic concentrates. The present report is an account of the results of these various analyses on the 347 magnetic concen- trates from Alaska to which 30 replicate subsamples of one sample were added for internal control. The report ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA 3 shows the relation of variations in minor elements to the source and mineralogy of the magnetic concen- trate, and illustrates the use of this material as a geochemical sample medium in the subarctic and arctic environment. ACKNOWLEDGMENTS The senior author, a geologist/mining engineer with the Branch of Natural Resources, Institute of Nuclear Energy Research, Republic of China, expresses his appreciation to the International Atomic Energy Agency, Vienna, for an IAEA Fellowship in mineral exploration which made possible his work at the US. Geological Survey in 1971—72, of which this report is a product. Acknowledgment is also made of the cooperation of the US Geological Survey, exemplified in the helpfulness of H. M. Nakagawa, who suggested the use of his then unpublished method for the deter- mination by atomic absorption of eight elements in a single solution from iron-rich samples; the assistance of R. L. Turner, who instructed the senior writer in N akagawa’s method; and the courtesy of Wayne Mountjoy, who placed his laboratory at the senior writer’s disposal for radioactivity measurements. Ap- preciation must also be mentioned for the help of W. R. March and A. L. Larson, who prepared the magnetic concentrates for this study. The technical problem in- vestigated was suggested by W. C. Overstreet. MAGNETIC FRACTIONS OF CONCENTRATES SOURCE AND PREPARATION The magnetic fractions of the concentrates described in this report are contained in the US. Geological Survey’s Alaskan placer concentrate file. This file of concentrates was largely acquired by Survey geolo- gists during the period 1895—1953. It numbers about 5,000 concentrates from gold placer districts, mineral- ized regions, and other areas in Alaska, and each con- centrate has its own file number. Most of the material consists of concentrates panned from alluvium, but a small percentage is sluice-box concentrates donated to the Survey by prospectors and miners, and a few con- centrates are from crushed rock, drill core, or beach placers. A split of the raw concentrate finer than 20 mesh was separated into magnetic and nonmagnetic frac- tions with a hand-held permanent magnet. The magnetic fraction from the first separation was further separated magnetically twice to insure that, as far as practical in the context of a feasible exploration tech- nique, the magnetic fraction was reasonably free of nonmagnetic minerals. In general, the magnetic frac- tions were found to consist of more than 90 percent magnetite. A full discussion of the actual mineral com- position of the magnetic fraction is given in the section called “Distribution of the elements.” Inasmuch as the final magnetic product was not entirely magnetite, the material analyzed is referred to here as the magnetic concentrate. LOCATION AND DISTRIBUTION The file number and location of each magnetic con- centrate is given in table 1, where the description follows that given in the record of the Alaskan placer concentrate file. The codes in table 1 show the source material for the magnetic concentrate. In several in- stances more than one concentrate came from the same locality, reflecting the fact that certain gold placers were sampled intermittently through their productive life by various geologists with the US Geological Survey. Some of the older locality descriptions, made before Alaska was systematically covered by topo- graphic quadrangle maps, give the general source only. Quadrangles represented by the magnetic concen- trates are shown on plate 1. RANDOMIZATION The original group of 1,072 concentrates from which these magnetic fractions were taken was randomized using tables of random permutations (Moses and Oakford, 1963), and the concentrates were processed in random sequence to prepare them for analysis. Throughout all following treatment the magnetic frac- tions were handled in this random sequence. ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA CONTROL SAMPLES Three samples of magnetic concentrates were used for control in connection with the analyses made in this investigation. Two of the control samples came from outside of Alaska: one from Nevada and one from North Carolina. These two control samples were analyzed in replicate by atomic absorption to test the reliability of the analytical procedure for silver, bismuth, cadmium, copper, cobalt, nickel, lead, and zinc before it was adopted for this investigation. The third control sample came from the Alaskan placer concentrate file. Thirty replicate subsamples of the third control sample were intercalated into the ran- domly arranged set of 682 magnetic fractions obtained EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA .5 3:86 m>onm 4&6; -I I- I- § me 8 ._ cm 6. S w. o: m XIRSDE 3 975 : 6:N o m: Nm: 0 m mo $8 I .5 3:95 26% 5&6; I. I- -- ow cm 3 ._ am e. E N. cm a x5334; 3 at; t #2 om 2 NS 0 N mo Rom 4&6; -I I- I- mm B N 3 mm .4. ON N. oINw .6 3m 8,86 E mm N; 6225454; On 2 N3 0 N mo 08m 4&6; -I I- -I mo co 3 ._ mm TI m N. E ----.-mem m x2334; .5 3m 0m 3 NS om _ mo mnom 4&6; --- -- I- mm mm N 0— mm P S N. of III-sung gens E Nm 5 3m on 2 NE 0 N we flow I go xuo; m>oam .mzoz --- --- I. m; cm ON m on v. S v. 93 ; $2533 3 5.5 “I a: o N_ NS 3 _ mo NBm I I I :u 6.81 m>onm 4&6; -- I- I. ma ow on 9 ca o6 0N To 03 gr: f. 3; 333 m x3333 cm 3 NE on 4 mm 33 I I .5 goo; £5 4&6; ._ N. 4 0; mm 8 9 mm ._ f ._ 2.: t ufl t. 56 8&8me 68 Om 3 NM: em _ mo moon I .5 345.6 4&6; I. -I --- 00— mm N m ow v. S v. of m>onm E cow m i362; cm 2 N3 3 m mm 33 £0 wm30Lu Ln 4&6; A ._ -_ om m: m: m N v. S N. o; t 80 i : fizoE m>onm E mm om NF NS 4.: m we moom I I I .5 xuox mo 4&6; -I -I -- 03 on: on m me 6. m4 .4. o? 530E m>onm E m3 m x2332. on 2 NS me _ me mmom .l I go to; 4&6; I. I- I- mm mm N S ow -_ on N. of .6 53E 3.388 1 £23.54: om : NS me _ mo vmom I. .5 V30; 6:02 .I I- -I mm £4 3 8 mm .4. mv m. cm a: E KN 8.6 a 1:534; cm 2 N2 .3 _ mo Dan I 66 50; 520E 4&6; I- -- mm cm 8 2 mm ._ S ._ E 36%. E N 63533 68 cm I N3 me _ mm mvom 4&6; -- --- I- mm mm 8 5 mm a. S 02 I-LtzoE mo Sm: E N Lu 6.8; cm 2 NS m; _ mo 38 I .36 ,5 56:0 Cu 4&6; m. N. ._ mm N m: 3 CM 9. S N. of V33; .6 5:9: got g: cm 2 NE me 4 me 348 4&6; I- I- -I on: em 94 2 me .4. mm. w. R: EtEE‘ 323 «2% m 6:25.53 9 a N9 om N mo Neon I I .5 EEE 4&6; I- -- --. SN 3 02.— 9 NF 6. E ._ z 323 PT: t “2 ; x2333 o w Nm: we _ mm :48 .Imcoz --- I. In mm N 0— I 4 LI o. E N. 8 ....... II- ..... - .6 625sz me N 3: on mm mm NNON 4&6; A 4 _ om mm ooN.N S o: -_ op N.o 2 III ......... 232; .6 Eco om N6 «3 an _N we son A: 6:6 .5 w meumbcwucou 0N8 imam-64.54. $3333 3.34.93 I H 2 a % o. z :.-----5 222,33 .43: :55 o 4 m: 0 mm G SNN .302 I. I- -I E mm 2: m c. z .---I----I--.-..-I-Lu :3; o wv m3 3 N S NONN 4335593 I. I. I- omm 3 o: Im N. z III-III--- ..... Lu m.538; o N m: .24 «N S $3 .202 4 -_ 4 mm 2 me oI N. 2 III. ..... I; £5.53 33.3 cm 3 8: o ow $ 83 I 4....g6; I- .I I. o: N mm Im TI 2 III-L; 3.2236: we PC; 5 m: o 91 em mm 5 «m3 4&6; ”€64 -_ 4 -_ mm 2 mm mI _ 2 III--- I--- I5 Emcmzué o I 9: 3 we 8 NSN I.3.3.:I<6; -I I. I- mw op PI mm 44. z -----------m 65.5% 23.5 on I a: 3 3 8 SN 4&6; ”ENE ._ ._ ._ o: m.— ooN E N. z ---.-.--.-..-- ..... S 33:: 3 mm 3; Om me S 83 4,6258m=.=N.u<.3.=<6._ ._ A A 3 S S m ._ z I. ........ --- ..... S 392$; am 2 m3 0 :4 5 B3 .982 -- -I -- 8 3 mm m N. z -...--.---.6 3.22m as: 5&8 o E w: 3 me 8 $3 4664 I- -. I. coo; 84 mN & 3 To m «.0 z -.--------;: mZE as: 55$ 3 MN 2: m; S 5 MEN A2. 6S.' .5 .:< .6; 634293 6 $363388 :8 2626633 wmmgosué 3.8 3.8 34: E E E E E 3.8 $8 3.8 88 E E =< :N a; .E 3 co 8 E 3 2w 25.583 L0 5.55.38 _. _ c __ _ c .4328: 35036 633.3 63.36 LwnEaz ELwEE c305. AcoIZE .6; 3.63 305955280 acmEmE umwz 562 6:; 8:: 458 Ea .52 .38 65 man 524 84 £836 Essa £4.29 .22....» as: .2 .5556 «o 88: 533 2: :33 $2 4 5 a3 a . 2.8053 352.2: :325 8 $353 6%..va Set muggzwouou orgasms”: Audio «.3335» .3 33me ES 23.58041.“ @4549 =3..ng «a :5: .533 2: E 4354?. anon—2» :28 53:3 manage—$.33 E .3552 4 .- mui . J a 4 4 .- 3: 25 .48 tau—c3 :3 “.552”. .2 $333.3 3% 0: $5335 .-.. 6.9.0238 non—5w .3»: £153 GASES. wfian 295.625 2... 8 axon 52m mm 3550? $5... $2: .5“ @395 fiwauncuau Eon =a .5“ tuna—a5 95;» 81.5.3 =< 4—me on 5.45 65h mun moi—:3 =< .SN 55 3:98 43:56 a 93 6:5 mfishanwocou 393: {avg .351va E 251595 c E moi—nun No .355: 2: ”E. can .5 .:< 38% «5.3: 3.5830 .5. 535m won—En Eod 32.56 PE 83.5% 63°: 0&3..on 32.5. as 3.. u... ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA .262 4 4 4 OO. ON OO O. ON O. O. 4 z 8 OT: N. EN .8: LO 25 O. .. .O. OO ON OO NOON .26.. 4 4 4 ON. ON ON 4 OO O. O. N. z ------B 3.. t .8: .6 as O. .. .O. OO ON OO NOON . .4. . .262. 4 4 N. ME ON ON O OO N. O N. z E .O.. 3 Or: 2 OON .3 25 O. .. .O. OO ON OO NOON 5 OS .26.. -- -- -- % ON OO. O. OO O. E 4 OO O6 .5 3.. t 8 6...; O. OON O. .. .O. OO ON OO OOON E”... N. 4 4 OO. OO OO. ON 8 4 ON 4 OO --.O as .6 .5 3.. t. O .. .2 OO ON OO NOON Eu... -- -- ON. B OO. E ON 4 O. 4 ml .3 E E... .8: LO .48 O .. .O. OO ON OO NOON SHE -- --- ON. OO OO. ON ON O. O. O. ON. -----.:4 3O E... .8: .5 .48 O .. .O. OO ON OO ONON .4. E O. .56.. -- -- -- NO- ON OO % ON O. % O. z 8 9.: O. um. 62; LO as O .. .O. OO ON OO ONON .334 -- -- :- Em OO OO .Om OO N. O. O. z ---fi.6=_ 26% .5. O.. NO .45 O OO OO. O. .N OO N.ON .264 -- -- -. OON m OO ON OO N. O. O. z -----fi_6___ 26% 5 ON .5 as O OO OO. O. .N OO 0.0N 5.26.. 4 4 fl OO ON ON O. OO O. O. O. z ----LO 22.... .3: L6 6&4 O O .O. ON ON OO NOON 5 3% 6 5.3.6.. 5.3.64 -- -- 2. ON OO mlm OO O. O. O. z 5.3 5.2635 .3: L5 .624 O O .O. OO NN OO OOON 3.3.6.. 5.3.64 -- ON ON OO OO OO O. O. O. z - ....... LO ...OO .8: a “.624 OO N .O. OO NN OO NOON .5 O... O .5136; 25.3.53 4 4 4 cm. m. ow Mm om e. m N. 2 mo 5:05 26.3 E cm .5 :3 o m S. o mm mm NOON . .5 OO.. . .3562 25.3.03 OO. fl ON fl OO O. O. O. z 36.3 a: t .O. CO :2. O . .O. O NN OO .OON .Nxon 8.3m 45.2.2 26.3.3 4 4 4 mm ON S w mm 4 I. O. 2 3.3.8; 32.»: .5 3% o N G. o mm .3 3E Naif“... -- OO. mm OO OO ON O. O. O. 2 -----..O OON .5: LO :2. ON OO OO. OO .N OO OOON I .53... 33w... 2.2.5.2 -- -- OO OO ON % ON 4 l. . z O. OO .5 OOO .3: O :2. ON NO OO. OO .N OO NOON NO OOO 5.2.56“. ON ON 9.4m W OO O. fl O. 2 so: :3 £5 2 .O. LO :24 ON OO OO. OO .N OO .OON 2.2.56“. N. 4 4 2. ON ON OO OO O. O. O. 2 PC. t EN 6.2. 5 OO .5 :24 ON OO OO. OO .N OO NOON OO OOO .26: E ON ON W ON N. E O. z t 5.6... 5.2. 5 ON .5 62. ON .O OO. O. NN OO NOON .ENE -- -- ON ON OO OO OO O.O O. O.O 2 £3 t .O. 5.2. E ON LO #5.. OO NO OO. OO NN OO NOON | .O.: 4&6". -- ON. ON OO OO OO 4 O. O. z N. .O. 5.2. O. OO LO 6.5 ON NO OO. OO NN OO OOON 3.6.45.3. I 1 5.5353566 .364 -- OO. OO ON. ON ON N. O. O. 2 -------O.:. O... um. é 38.. O N .O. O NN OO NNON 5625.3. .I f. $ .O.... O. .E.=O£w..N6N SO64 OO. OO ON ON OO 4 ON O. 2 .O. 8 at; t .O. .m 386 O N .O. O NN OO NOON . 4.6.45.3. l l | 5.5.5.56“. .564 ON ON. ON ON OO O. . N. z --::-z 88“. 3 OZ. O. OO. O N .O. O NN OO OOON .O<.OO._.§.=O f l | 5.3.3.56“. .564 4 4 4 F OO ON ON OO 4 O. 4 z .1 88.. 3 £5 O. OO. O N .O. O NN OO OOON .O....Oénz J- E E OO. E F N. O. O. z 5...: $63.62; 264 ON N. .O. O O. OO OOO. .262 ONO pm ON OO OO. O. O. O. OO EEO .5638 26% .5 2.... ON O .O. OO O OO OOO .262 4 4 4 ON OM OO O. OO O. E N. z ---------- --:-LO LOOO O O .O. O. NN OO ONN .33... 4 4 N.O ON OO ON O. OO 4 ON 4 z :5 .28.. OO Pr; CO OOOZOOO OO N .O. O. ON OO NON C... .O.; I: .:< ‘8» 85:95 mm .prmgucwucou mm. 1235.33 2.2.3 2.6:.O<.=<.=N.€.=Os4 % fl OO 3 ON 0.0 O. O.O z 2E. 2.22,... O6 2 .3 .O. O O ON. O . OO ONNN A: E... .E .=<,Low 35:23 no: .wufucmucou 5 «3:53.... .35 39.595 -- a OO OO. @ OO O. O. W 2 1O .35. O ON OO. O OO OO .N.N OON O. E B. OO N. O! R z .3: 5:26: O: as 23.5 ON NN OO. O NO OO ON.N .I... E; OO E F ON O. ON OO 2 ---------- L5 .25: O. NN .O. OO NO OO ONO 4 O. O.O OON ON OON OO. OO O. O. O. z ------------ :5 .26... ON NN OO. O. OO 8 ONO l I .mzwcvooc Us... 4 N. O.O EN OO O.. O. ON O. O. O. 2 Us... :3 63:2 NO 8.68 ON N. OO. ON O. OO N.O - -- ONN O. ON ON ON 4 O. N.O z ---8._O 65$ J. 4.6.3.3.. O NN .O. ON OO OO OON A: v.3 .5 .3 Low 35:93 N .qumgucwucou 3 396.624.. 35mm EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA .u: Emyu we .56: O O O O3 3 OO. O. 33 3. O. N. : 3 LO .5; 335: 8 3L3 3. 33. O3 MN .3. O3 33 33 3N3N ll .5 53.5: :3 .25: O O O O3 O3 3 3. ON 3. O. O O3. 3. 8 £5 OL OON 35...: z: 3: O3 3 E O3 .3 3 NN3N .3LO 353.:36: 3 3 3 3. O3 O O. 3. : L325: :3 3. 8 35 M. :3 O MN .3. O .3 33 NEN ' | | .LO 3...: .5535”: O3. ON O3 33 O3 3. 3. 3. : 8 £5 OL 33. 8 35 m. 3L3 O 3N E O .3 33 333 | .LO 3: 3:53.56: 3 3 a 33 3 3. O. 3. : 8 £5 OL 33. 8 3L3 w. EN O MN .3. O .3 33 333 I .5 mm: .52.:33 3 ON O3 O3 3 3.O O. NO : 8 £5 OL 33. 8 £5 3. 33. O MN .3. O .3 3 33 .ANOO:.:EE OO. ON 3 m O3 3. O. 3. : .---.---LO 3...: 3 £5 OL 33. O 3 E O .3 3 33 . 1 .33.: 3. £3 353:3”? O3 ON 33 ON O3 N. O. 3.. : 25 .53 53 LO L325: :3 3O O 3 .3. O .3 3 O3 .29 quczz .25: O.3l. O. 33 O O3 O 3| O : L: 3. 8 £5 3. 3L3 LO 35: O 3 .3. 33 N3 33 33 .25: 33 3 O3 3. 33 3. 3. 3. : :L: OL 25% LO L333: :3 3O O3 3N E O 33 33 33 .5 L325: 35 :3 3L .25: 3 3 33 ON 3N O 3 O : 35 L: 5L t .395 LO 525: 3. 3. E O3 33 33 N33 ’ .LO L525: 35 v: .25: O O O O3 3. 3 3 O3 3. O. N. OO. 3L 30 O3 ,5 3L 8 £5 3L OON 3. 3. E 3 33 33 333 Cu qucsz mo :3 “L .25: 3 O. 3 3 O3 3. O. N. O3 .3 t 2: 3L 3 £5 3L 33. O O. .3. 3 33 3 33 cgu quczx :3 u; .25: N, O O O3 3. O3 O. 3 O O. 3. O3 .5 i .5 OL .323 LO 525: O 3 E 3 33 33 333 Cu quczx .25: 3 3. O3 3 O3 3. 3 3. O3 35 :3 3L LO .5 3: LO O 3. .3. 3 33 3 33 Cu qucax .25: 33 ON 3 O. 3N N. 3 N. : 35 V: 3L 3 £5 3. .53 O3 .N E O3 33 3 33 .L0 swucsx .25: 3 3 O3 3 3 O O. N. : 3O :3 OL 8 £5 OL 5L3 33 .N E O3 33 3 .33 .25: 33 3 33 3 3 N. O 3. : L.-------:-O .525: 3o 3: 3: O3 .N E O3 33 33 333 .25: O3 O3 O3 O. O3 3. 3 N. 3 5-85 LO 525: O. LO 25.: O 33 .3. O 33 33 O33 .mec :0 Lmucsx .25: 3 3 O3 O. 33 O O. O : .LO 3.5.: 3 £5 3. 33. O 33 .3. 3 33 3 33 , .33.; .6 L635: .25: O O O O3 3 33 3 O3 3. O. N. : .LO 25.: 3 £5 3. 33. O3 3 E O3 N3 3 33 .mem Lu quczx .25: O O O 3 3 3 O. O3 3. O. O : .LO 3.5.: 8 £5 3. 33. O3 3 .3. O3 N3 33 33 | . .LO :33 O3 3 33 O3 33 N. O. N. : 85.6 35 i 3.: 33 t a: L: O .. E 33 3N 3 333 Ilr .mmgu Lu NuLmzo .zu3zw :35: O3. 3 3 3 O3 O 3 N. : 3:533 2533 3: 3.523 O .. E 33 3 3 33 l I. .3 3.52.; 3O 5 3L :33 O O O 3. O3 3 O 3 O O 3. O3 LO :3 3: t 3:53 O: 3.255 O .. .3. 33 3 3 :3 ' b: .3 33.2.; :33 O O O OO. 3 ON. O 33 3. O. N. OO. 33 LO 2 3O :3 3: 3L 253: O .. E 33 3 33 33 .3 3:2,; LO 3.. :36: O: 3 OO. 3 3 O O. N. : 3O 3: 5L t 2,83 3: 3.253 O .. E 33 3N 33 333 l i .3 3.33.5. :36: O3. O3 O3. 3 3 3.O O. NO : 2533 3: 3:53 .5 L33 3 .. E O3 3 3 333 > .3 v:33: .25: O O 3. O3. 3 O3. ON O3 3. 3 O : 2533 3: 3.255 .5 L33 3. .. E 33 3N 3 333 cc mEEm .25: pm O3 O3 3 33 3. O. O : 2.83 3: 3:55 .5 53 3 .. E 33 3 33 N33 .25: O O O O3. 3N OO. O. 3 3. 3 N. : :-.----.:-LO OOO .5 L3 3: 3: 3. .. E 33 3 33 .33 .35. .25: 3% O3 OO. 3 3 N. O. O O3 3: t 3 3L3 3. £3 .5 OOO 3. .. .3. 33 3N 33 33 83:33:09-4: can .5 .O< Low 85:35 mm 333355383 33 :Emcmsuwzc 2.9.3 .NOO 3.8 ANA: SO GO A: :O A: 8.8 :3 O 3:8 82 3 O .. O. 5: ON 3 .: 5O oO 8 .5 3: O3 15.583 35 8.33.538 __ . 3 c ._ _ c . £3: 333.30 on m 3.. co 2.3.5.3.. .5252 N 323...... 585. €3.32: L33. 3235 205235.338 3:35; 5 3333 O .518: 3.: gsfiauoolmfimomwfi 3355 53353 8 @839. 33335.3 Sch 33352528... 0.332565 33...? 323.335? 3.3333335 gougcqlé "39¢. ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA .em.eaue4 --- --- huh om mp ow Mb em 4 4 m z «.r-----------oum> 4: mm Lu as on w_ MVP me e we can» .eeoz 4 4 N.N om mm mN mN mm N. 4 _ z --m wwwazugeu 4; m Lam Lwemsp< cm Ne Few m_ m we mwwm II .z e_wexugou 4 4 11m mN m_ ON ON me 4 m_ 4 z :e gazes w>eeu Ex om cm 5— ww— om m— cm FmNN 44.54 E 4 4 _.m ee e4 8 .4| mN 4 4 e. z ----------------- ..... v: 2m»: 2 mm :1 e e we m3 .mméiéfléem 1- --- 1. E 2 mm mm 3 4 e. 4 z .................e 3435 484 .3 Nm :1 e 44 we Nmm kw II .m .o: .mm.2w.3.:m.mm em 4 4 m. cm mw mm mN me 4 m 4 z gwxeeo .mee .z w—wsqu0u 4w m 0 cm Few me m cm NNN .mnuem 4 4 q. 0N m o_ m cc 4 o— 4 2 -----------------------go mum: mw mm _onm E me go cw>mngmm me 0— _o_ me ow me ONmN II go :euoem we .wcez --- --- --- ov 0N mw ow ow 4 0— 4 2 4w 3: w— .emgu go =w>msgwck me o— _o— o _v mo m_wN .mcez --- --- --- om mp ow mN mm 4 0— o. z - --;u=oe Eegw E me Lu :eueem om ow wm— mv ov me o_wN .24 um --. --- --- om mN mm m ON 9. o_ N. z - -Le cw>mggwmu cu ewe» w_ ecN o ow Few om Ne me ewwN cu cw>ngwmm .2; ea --. --- --- ow mN cm 9— mm c. 0— N. z so go wme4 Leesm w>e4m a w_ o N _m— c we we __wN .mwge Lu cm>azgwmu .zmnee --- --- om mN ow m mN «.0 o_ N.o 2 .Le wme4 Lumem we 4w 2; “m cm N— wow 0 we we owwN .Lu wme4 Lemem .24 ea 4 4 4 me am ow 0— mN 4 ow 4 on we 4w 4; .megm go =m>uggwem om N_ wow 0 we mo mowN .Lu wee4 Lemem .2; ea --. -|- --- mm m_ we m am e. m 4 z we 4w 3: ww e4 4w 3: w4 om N— _o_ c we mm NomN .Lu wee4 Lemzm .zu.em --- --- --- mm m4 om m an 4 m N. 2 we 4w 3: ww e4 4w 3: um om NF _ow e we mm oomN .24 ea .-- --- --- mm m4 me m an e. 0— e. z -l---gu weo4 Lamsm we 4w 3; w4 om N_ Few a we we mowN .Lu wme4 .2; we --- --- .-- me ON ow m mN 4 o_ N. z gemem we mxw 3; we Hum 3e_em om Nw Paw o «q me eowN cu wme4 me2m .24 ea 4 4 4 mN m_ om m mN o. o_ N. 2 we nwgu w_ ecN e4 ewgu a; um_ om N_ 40— o «a mo NowN .2u ea - .-- --. mm om me m cm v. m e. ow Lance aegw 2 mm Lu wme4 Lumem on N_ ~m_ 0 we mo mmNN mggwmm .gu .Zwuee --- --- mww m_ eN m mN w. mw N. om szwwu 4w 3; u; e» 4w 3; pm on m _ow c we mo NoNN .zuuee --- -1- eq_ ow cm m_ cm w. m e. em ----Le LmNcwez e4 nwgu ww ;u__ o e_ _o_ m_ Ne me mmNN mLu gmNcwez cu ewe“ .waee --. --- |-- om cm mww 0— om 4 DP 4 z wF zen .mwgm Lo cw>ezgweu o m_ _o_ m_ om mm owNN .2mnem --. --- --. mm ON oww 0— ON 4 m4 4 2 go Ee_u umem .ewgu m nee—xeem o m— —o_ m— mm mm mmNN .Lu Ew—u waz .24 we --- -1- --- om m_ cm o_ mm 4 0— v. z w>oee No xezo e4 nwsu a; cue cm _N _o_ mw «m me mmmN CU v—USD .2“ ea --- .-. --- ow ow oo_ m mm o. 0— N. z e» nwgp 4; saw eu ewe» e; um_ 0 MN _ow om mm mm NmuN go Ew_o 4mm: .2muee 4 4 4 mm mp o__ ow om q. o_ 4 z w>eee so xeza cu ewgu 4; new o MN _ow cm mm mm omNN ago xeeo .Zunea 4 N. 4 co mw om_ ow me 4 o— N. z o» ewe» a; eLm 04 nwgu w_ amp 9 «N _o_ m_ mm mm mNNN '11. 1o ew—u 4mm: .2; ea --- --- -l- om m_ o_N op mN o. o_ N. 2 e4 ewgu w_ ecN .mmgu Le xuzo o «N wow cm mm mm NNNN umscwuceu.-4»h ecu .c4 .:< Low ewNXPmce mm ”mmuugucmuceu mwv mpacmgcmec m—ucmu .3 02.3 :33 .ON.2.ON.ON.=<.=O ON II- --- --- ONN ON ON Nwflm OO O._ ON w 2 NO szN a» ONLN NNNEW OON ON N OON ON mm mm NNNO I I I I I .30 om=U< .EN.z.mm.ON.O<.=O ON --- --- -.- OO ON ON OON mm N. O O N z OONOO NO OWNN ON ONO“ ONO ON O OON ON NO Om . ONNN I II I I I .5 02.2 .cN.:.mm.ON.=<.=O ON --- --- --- OO ON ON ONN Om N. O O z soNOO NO OMNN 0O ONO“ ”NO “ON O. N OON ON Om Om ONON III I II II .33 .ON.=.ON.OO.ON.OOHOO --. .-- --- ON OON Om OWN OO O. ON m N z NNNNO 3mN> a: OONOO LO omcu< ON O OON ON Om Om NNNN 583.35.2583 ._ N ._ mm 2 ON OO OO 4 N N. OO -I----------._O?£ NO 2 3m p2 2 N OON 2 OO mm 32 I I I I 6:553 .ON.=.OO.ON.=<.=OHON --- --- --- ON ON ON OO ON N. N O. OO “O NOONLO ON ONOON LO OONN OO O OON ON Om Om OOON .umo; :NOE mo 2 .OO ON --- --- --- OO ON ON ON OO O A N. 2 EN O.O ON co NOONLO OONOO LO ON NO NON O _N OO ONOO .OONOO II. I-- --- ON ON ON ON ON 4 A N. z -------LO NNNEONO NO 2 LOO OWN ON NO NON O _N OO NOOO III .ONNNOOOmx .OO ON A O O ONN ON ON O_ OO N.O A N.O z N0 2 Ex O NO NONONOI Om NO NON O NN Om Ommm A: use .5 .:< LOO 3629; N 3355588 3 39.5626 22.238. .wcoz ._ ._ .— 2 S 3 mm mm m.o ._ m6 2 IImuLoaLE :mmci. um 35395 0 mm «m— cm 5 mm meow A: EE .5 .:< .8» cgrccm F $3550.59 5 Emcfomac 23:2. Sum .O<.O<.=O ON --- -I- --- mmm ON ON OOO.ON OO O. m N_ z NONOOO Nam: quOmOLO muw___z O OO NON Om NO OO ONNN I . Om .mcoz O N. N. ONN ON ON pm Om N N O z ONNN NO ONO 3: ON gum OmN ON Om OON O OO Om ONNO .ocoz -.- --- --- OO ON Om Om Om A O N. z ..... N ONNN NO ON305 EONN Ex O Om NO NON ON Om Om NNNm IIII .II .OO_w m .wcoz N O O. OWN ON ON mm ON N. A N. 2 co Nam ONNN OONNOOOO Lum “ON ON N. OON O OO Om ONNO .mcoz --I --- --- ONO O. OO Om OO N.O N N.O z ------------- ........ LO MNOOOO O N OON ON ON OO ONNN A_._. use .5 .3 Low 35:25 N $355588 3 39.535 9.53: .m« OOIN=<.OO ON --- -.- --- mmm ON mmm mma ON O. IN me z -..--NO:NE OON>OO OON .NO NONN ON Om NON ON ON NO OOON I 4&va 3.. II I .93 .83: Law: 30 .ON.ON.=O.mm.mN.=<.OO ON A O O ONO ON ON ON OO O. ON 4 z ONENN 3 “LOO NOONO .LO NNNN OO O OO— O ON NO NNON III .OO ON m3 .III III .III III .III II .ON.ON.OO.mm.Mw.=<.Om ON --- --- -.- OON ONO OOO OON ON O.N ON NO 2 ..... --mOOOmNO NONNN .NO NOONO O— N OON ON NN NO NOON I I I2“: M3 I I I NOE... .ON.ON.=O.O:.mm.O<.Om ON -.- --. --- OON Om OON ON ON O. ON O z NOOWON OOLN .OON OOOEONOz Om NO NON O NN NO NOON II II . ”ON .3 II. II .II. II III II I .ON.ON.OO.mm.IK.=<.OO ON --- --- --- OOO OO ONO ON OO— O. ON N 2 L5 ONNOOLO :0 OOO EONOO NONNN O NO NON ON ON NO _OON II.I. .=< ON .3 III II III III. II I .ON.ON.OO.mm.m«.O<.Om ON --- --- --- OON OO OON OON OO O. ON O z ------ ............ -NNO OONUNOO Om OO NON ON NN NO ONON I I II I I I .85 me 5:05 #6: .3HOI.=<.LOHON --- --- --- OON ON ONO ON ON 4 ON O.N z OONE wz< NNNON .NO OONONOO ON Om NON .OO ON NO NOON :N PE .5 .=< N8 RONNIE; N ”Omufucwucou t 39233 nogmfiE .Lu.un_.:8u 5. TN .5 mics—.5 o mm NE 3 am we 2: NE. ucm .5 .3 No», 3529; _ ”395538 5 29.9626 Oxzwnimu. 3.8 3.8 NNOV E E N: N: 3 N8 $8 3.8 88 E E :< :N NE .N2 3 8 3 .5 5 so 20.533 “6 :oEOCummo _. _ __ _ OE; Econ 2:: SO“ Lwaszz «mm: 532 BE 13.23: 3.58% ENOEE $.65. $5225 Nwa 3.63 mcofimbcwuzou ucwswfi vusfiunoolmfimcmmfi 33:8: $385 8 33$? 335$ Sch NOSEONSSS... c.5253: REVS 383353? «:32 ~33 meotuocQIé Quad. ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA .wzoz -- -- -- om ow P S p N. m N. 2 $an ax: 38¢ E om 3 atom o N S: m— S 8 $3 4&an ”3.393 ._ ._ m. mm 8 3m 8 ON— N. m c. 2 --------.-------------5 32m 3 S 3: an 2 mm mmN I I I .525 3—52 .361 ”3.3;: ._ ._ m.N om mN E ON E N. o— m. 2 we 2 2.53 Lmv:2,— iu 32m 3 3 3: on 2 3 SN .=_m=c m ”mmumgucwucou mpv m—oCMLumsc max»: ucsoz .=I<”EIJ=Q.2.=<.£”3 -- -- -- oo— 8 3 m 3 e. E v. 2 I526 $53: c we «3 mv NF me New £13.51: ”5.53564 -- -- -- E N mm 2 cm -_ m-w A z -- ..... ISSo mesm 3 3 e3 3 3 mo com .982 -- -- -- SN 8 ow m ow ed 5 N.o 3 --------u: 3&2 Co fig 5:6 om RN mm: o m mo omN Afih vcm .:H .:< Lo» uo~»—mcm wco: "mmuwgucmucou my o_m:mguc:c mgwuoz 5.2..HHE 4 w. MI E a 8— % a M S q. z ------:-.. -...m5 228 o 3 m2 2 mm 8 :2 3.5.3.: “3 ._ 4 5w 2: ON 8 m mm «.3 2 N6 z .--.-m;_ £32.; .625 3v: S 2. m: E S S m? APP czu .:H .:< gem uw~>_a:m guon ”manusucwocou NV w_m:mgua=c zumgwuz .nm.m<.nn_.:I<.:u€n_ was: ._ N. NA. @ mw cop 8N; om N. S w.l z III-$355 3.528 86 x3. on mm NE on m: S mmwN Ixmauucg -- -- -- bmm mm o—_ bmo.N mo v. m p z ........ --------mLu can o mN Ne. om NN —o m<_N .nm.m<.£_.=<.=u6n_ -- -- -- FE. 0N oN_ Mm cm a. E N. z ..... --- ----------- 5 953 on 3» NE 2» ON 8 meN .313 $56.. ._ N.o -_ 02 mp 03 on me ._ 0— m6 2 ----- ----- L6 :3 we or: pm on wN N: o NN S vaN AP» ucc .cH .:< Low vaA—mcm N ”mwumsucwucou vv w_m:mgum=c AcuLmuuz .:m.m¢.m:.W.ECqu -- -- --I E 0N B DIN E w. 9 z 3.5 9.5.5 59¢ .5 V3338 on «N w: 0 mm me SE .=I5 cm 8 m3 M: Na mm mm: .mm.:.m:.:<£u.nm”§ ._ A .3 ONF 3 00¢ om mv q. 3 z ------- --------m-G Sam on 5 m3 2 5 me 33 II .xgm>0umwc .am.z.:m.=onm m Ewm_u .Lu mxcuns_mk on w. Ne» o v me me— III II .u_E__ .nmiém. -. -- -.I o2 Im cc W am e. o_ z 5 5:3 N_ .02 86 823:2 o m: 5: cm c mo 3: . -_ A S cm % of cm me N6 4 z ---m=£wwz km: S iconic“. cm 2 5: cm v mo 2: . -- -- -- ow ow o: 5 oo q. or z ------- ---- ----- 16 52:5 on mN 2: cm e mm mm I I I I .16 ammo we sing ._ 4 we mm me ooq mo 3 ._ o— ._ 2 5:9: w>onu “man 5 922:2 Om o D: on m mo 3 II II WM .02 movmsa 43415.36; -- -- -- PHI 3 E mu mm 4 o. «.1 z .8 .m .n. 86 5.58 $33 on cm :1 o A mm mm 3.3.2366; -- -- -- 03 mm 8m 3 co 4 9 N z III-G w>Io .3: 25963.. on on m: cm mN me K II II II III II II III wm 3.5.3.5.?2 -- -- -. o: cm: of mm mm N.o mm m6 2 2.98 33.... 8 £5 5 5.1 cm 3 N3 3 _ mm mm A_» ccm .:H .=< Low um~>~mcm v "mwuugu=mucou __V w_o:mgvm:c cooacw>m4 .WHE 4 4 _.: a oIR o: E om v. m o: z I .......... ----.-$8 8.53 cm N .5 cm mm 8 8R .3 E ._ 4 ._ Mn! 2 E m Mel ._ m N. z Iifisoe 59; E mw .6 mama-6m m— _ 3.; cm FN om Not” .282 -- -- -- 0mm m. QNm co 2: ._ ._ N. z --------------- - IS 35:: cm mm mm; 0 MN so 33 I .z €39.95 .mcoz ._ ._ 4 mm m ow om cm 4 A 4 2 No 5:0... gong E o»: o om mg on oN om oak” .QTES 9.3 233 .952 ._ ._ ._ me 9 ow ON mm N. 4 ._ z 235 m octmacw 3m EN cm mm mm_ o mN oo 3% I .235 33 23.3 .932 ._ -_ ._ mloI m_ mN W 0m ._ m N. 2 up; m 9.235 Law “my on 3 m3 o MN ow «mum .982 A ._ ,._ DE. 2 me b-m mo N. m N. z -_ c3225.— .8 E z 3 oN 3: om : om Em .:N.m<.na.:u"3 -- -- -- 02» cm om mm mm N.o m m6 2 ---m in; 5:9: 36%. E cm: 3 3 «3 m_ m— cm mwnm AZ. ucm .E .:< sow 85:53 e ”mwumbcwucoo 3 Emgégc vim—u 9.3 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 10 e6 L335 59:43.56". ._ .. -_ o: 8 o: -_ mm -. S N. z co :35 9.8-5. Low: :5; o —N mm” o N.» we Nmo .5 | . . 4 F a E .I cm I. E N.I 2 Sam 8 952m .82.. $2.52. on 3 mm. o S E 8. I In -- I- ONN Bu; E E coo; m.N 8 E z ----------;u—:w 55 me 5 mm_. m: «N we mm .5532. -- -- -- 3 MK 02 ONE 3 To 2 08 z ------6 .8528 .8. £8: 2 mm 93 o .3 3 mm 2... “Eu .5 .=< .8 85:23 N .mmufucwogu 3 29.2625 33. .2 9:8 .282 -- -- -- mm N mm m mm o. 0— ._ ON we .53 Ev. .3 W. 33.53:.— 0 om SE 3 S E wmmN . E8 .252 -_ -. -_ me m: 3 m cm .3. m .. co .7: an .8 f. ~. 3232.5 cm cm 8" 0m me .5 SN .25... -- -- H 3 ON mm mI Om .3. 0— cc .-- ...... 1.33335 No 532.. me Nm GE me .13 we momN .3.=<.n.m6n. .. . 3— on m: on: om am To m A z ---¢.Imc:= No m. .6 33:3. m.‘ 3 of cm mm vw «om A: new .5 .3. .8 sign N ”mono-5:880 .3 398.33 33 cotoz I.E.nm6n_ -_ -. ._ om mm 8 m on .3. S N. z ---. wEoz “6 PF; 5 >528 om 3 mm: cm 3 «w 9% I 33.3.3.3 ._ A m. N MI ow m cm . E -_ z ---- ............. .6 .532. me 9 m3 om me we Nmm .33.“. “33.66.. -- -- -- m.V of oo 9 mm A m N. z -- ....... -.-- wEoz JG Eonmo o o 3— cm «m 3. EN 5.3.6“. -- -- -- me E 3 m m ._ 2 N. z ------- ...... .6 “SEES. mv wN m2 om mm we RN 3.3.56; -. -- -- om 9 mm 2 mu 4 m 4 z --- ------- ---- .5 mayo—m om km 3— om S we .8 2... cc... .5 .:< go“. Pflbccm N ”mmumgucwucou 3 29.233 mEoz .wcoz -- -- -- an 2 mm 2 o.» .3. E A z -----.- ----------- .6 386 o N N: o S No :2 II to wzwasugoa .mcoz -- -- -- 3 mp mm elm S N. E o. z 26% a 9.2m 3.3.: 3.. 33 o 3 m3 om 3 No BS .56.. -- -- -- mm mN co 3 ca .3. -_ q. z .6 3.x. 5.6:. 3 of.» t. 3— o S m: o E No wow; I I 5 5:33 .252 -- -- -- 9.: mm 2: 3 mm Nd .. 90 2 mo m E. N 5 3255: we at; 0 cm 9: 3 mm No 3.: 2.. BB .5 .3 L8 2339:. 95: 3322359 .33 22.8.2.5: £532. .26.. .. A m. F N cm 2: m ow w. m N. z ------- --- -------- S 38.28 m.» N3 of m: on no mNS .zu—zw :wwuwwu Emu—ma . ..nn_H:<€n. -_ -_ m. ON 2 E m 3 . m -_ 2 “man 3 .0: 520.5 $365 0 mm of me em me No— 3 :< nwuum J A m. om om o—w o. ow A m 4 z - -- - ---------- Lu :onrguu o Nw omp m— cm mo MNO— 3.25mi; -. A -. om mm of o. o.» ._ 9 N. z I... ---------- .6 38.28 om me of on wm no _No_ I I .5 535532.933 . . m. 3 3 8. 3 S .3. S N. z 8:6: £8... :28 5 3828 cm S cm. m.» mm 8 29 III .ON .0: E_mpu .3.m<.=<6n. -- -- -- m; m eh: m 8 «.0 m m._ 2 J. “2.3 Cu 382 3393 on NN of o 3 ma mm. A: use .5 .2 Le “5529; m ammucbcmucou 3 2332.5". 32.5.92 ~56: .an..m<.:u€._ -- -- 3- mm om oImm m mu .3. -. .3. z .6 :32“. 22mg EV. «.0 a 335 O me m: on m: 8 23 I I .fiSoE 38m .fi.a<.=o.3 ._ -_ ._ oow 0N om ON OMN .3. m N. 2 EV. 9m 5 5.7.5. we PC; 3. 3 mm m: on 2 mo :2 I .539: 26% .nn..m<.3uu-_ -- -- -- mm mN em 0: me Q. S q. 2 EV. N S :22... mo £3 E m; an m: om S me SE .wcoz -- -- -- so 3 om m N A -_ c. z 333 3:. gone E MN S “8.26 cm 0 3: mp mm 8 mt: I .me-fi. 3:. .262 -- -- -- 8— mv c: on oe N. m w. 2 36% E ca m 293$ 23.: c we 3: o 3 mo 2.: .982 -- -- -- mm mm m— 9 ON .3. m N. o.» - 33%.. >3; o>onc E mNm-G in. on mm 3: m. S no NE: .o v.75 .ocoz -- -- -- me 2 cm 0— mm A 4 ad 2 >3; 26% EV. v.0 5 $63 om N 3: cm 5 me :3 3:528!“ E 35 .E .=< L8 “.339; m 338.2350 M: 33965:... 3?: age: 3.3 3.3 3.0. E .3 E 3. E 3.3 .m . 3.3 83 ,. _ o ._ _ . NEE-Em: 37582. C E 3 :N an. E 3 8 8 .5 3 3 18.583 No EBB-Gwen. 2.33:3 $3.53 .832. .EoEE 555. 3.3. 5-62 2E 35:2... Lwa 3-33 30.595588 acmewfi. twaiauoOImtmofiuw 3.8.5.: 2325 a... 3832 .333? EEK 3355:3380 uumgwufi «Rio $9333.? 9:3me 3:3 mEctuquIA H.329 11 ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA . .e_ name .esez n 4 mm m_ oN m om e. 0— N. oe em a s=_sm3s 0e nwse es e:N om w NeF om ew we ommN .sevsmeE ememsm_ .esez --- --- om m_ es s mN w. op N. ow Eese Emegemee amen m s=_s_zx om m NeF om ww we mme .i l l .552 .esez --- --- mm ON aw cm bk n op z e>enm Es w.o se xw_=eewswz mw _F Ne— mp ow we mme .esez --- --- om e_ em on ow 0— ow --------se :emsme ee 2 see ecN om Nm Ne_ 0 NM we wme yll. .I. .se somsmu we .esez 1-- --- em e~ :w_ 0— oe n op z 2 sum UCN semen xem seesez om Nm Ne— 0 NM we mme .ecez --- --- ow 0— mN 4 mm s m 4 cm ---se:eE e>enm E mmN L.e sewsme m_ em Ne_ mw eN we “mmN A! go cemsme .esez -.- --- mw o_ eN m cm e N. 2 we m sum “we ee 2 semee m_ em Ne. mw mN we emmN kw Cu .ecez n n mm e_ ON s cm w. o_ n z somsee we m sum em_ we Ln em m» Km Ne_ o eN we mNmN .se :omsme we .esez --- --- ew e_ e_ op om n 0— 4 cm m sew usN semen zem :eesez 0 Km New m_ eN we eNmN ye .esez --- --- :m e_ o_ be mN N. o_ e. z cemsme ee m semen zmm seeLez om em Ne_ o mN we eNmN .esez --- --- mm oN m_ m_ on w. o_ w. z - ....... se semgme we m Lem eLm om em Ne_ o mN we wNmN .esez --- --- om eN e_ 0— cm w. o_ w. ow ........ Le sewsee me e sew sew em mm Ne_ o mN we MNmN II .m_=m:w:ee se_mw_z .esez --- --- oe o. cm op em N. OF w. 2 ea w semen xmm cws>e_ee o kw Nefi ew wN we eFmN .II 2‘ ‘\ .m—emcwcee :o_mmwz we w .ecez n 4 cm e_ om 0— ow w. 0— w. 2 sum esN m semen zem :_s>e_ee om Kw Ne_ ew mN we mmmN . .1 .m_:ms_:ee sowmm_: -esez --- --- ew mP oN m mw e. o_ n z we eeww m semen xmm :_s>efiee om om NeF o RN we __mN .1 \. .m_:mswsee :ewmmwz .esez n n om ON mN m ow w. 0— N. 2 we ee_m z semen xmm cwc>epoe on em Ne~ 0 RN we o_mN 1. ye em30sseesmee—m .esez n n on 0N mN n mm s 0— n 2 we m Es w.o xmm sws>epee on om Ne_ 0 AN we momN .. t . :e emeesseusmee_m .esez n w._ em e_ on m mw n o_ n z we 2 semen >em cws>e_ee o ow Ne_ mw wN we momN 1: I. .momN .esez --- --- cm 0— mm! s mw n o_ w. z e_eEmm ee 3 semen xem s_s>e~ee o mw Ne» mw eN we womN .ecez --- --- en m emN n me n m w. 2 .sn es Le ¥_seese om em Ne_ m_ mm we NmmN II I! .;e swseese .esez n n me 0— oe_ e ex n e N. z ee msn e_ use as een 3e_em on me Ne_ em wm we NeeN . . .n_se *— .esez --- --- mN CF mm s me n n w. 2 see e>enm Le s_=msem¥ Ln em om em Ne_ o mw we mNmN .yw‘. ‘ . .n_su up new .esez --- --- oe oo_._ mm m ce e. 0— w. z Eese sew: Le swemsees Ln em on em Ne_ o mw we wNeN z! . . .1 .Ln us. .ecez --- --- mm m_ om o_ om w. mN n 2 en n_se e; em~.se swamsems 0 am Ne_ e_ Nw we NeeN ,. . .\ an e— we .esez --- --- mm ON ON m_ oe w. 0— z sezes 3e_en ewzw Le swemsees o mm Ne_ o ow we emeN .. .y. -. .se swemsemx .ese: --- --- mm mN ex 0— me n e_ n z ee n_se e_ esN ee se as em 0 0 me_ om em we emmN A_H use Jse .ew Lee eeN»_msm Fw ”meeeseceesee _o_v e_m:msem=e :OEe_em . I. -. .Le .ee.e:.=< ea n e.m oc_ em om_ mm me w. o_ z m~=esem e>onm Es N._ Le Lmem om ew mm_ om m _e mF_N .. .se .mw ea 4 n mm m_ mm c_ ew w. o_ n z Lmem e>enm Es m.o m smw¥=_=p ew em mm_ em H _e FNm New en --- --- oo_ eN oo_ c_ cm w. o_ N. z ................ ------se swseo m_ Fm mm_ cm w_ _e mew :<.ms N.o w.o me e_ co o_ om N.o m w.o z ..... ------ ........... Le L.Eeo 0 me om_ e_ e_ _e Ne A_p uzm .c. .:< so» ee~>Fecm m "meumguceesoe ww e—ecmgemeu cowmm_z cmwmmzm EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 12 .5 :5: 8 PE .ee.3.:m.em.zunua In- --- --- me ow OF e mN 4 my 4 mam my ucN .mmLa m xw_:u=e:p O e— Ne_ my me we mOON ‘1. III .ewgu m_ um— m>oem Ex w.O .ne.z.:m.em.zunun --- --k --- OO ow mfi m Om e. OF N. oee LO Lam—O .uwgm m Ne—zuzezh Om e_ Ne_ mw me we NOON 1|. .LO Lam—O o» .nO,3.:m.wz.zu O; m. 4 4 me Om mm m mN e. e_ N. bob e_Lu m— ecN cu ewe“ u; amp O w— Ne_ ew Ne we OOON .ee.3.:m.em.zenuu -l- --- --- Oe em OF e eN 4 O_ w. Omm -u---.-LO me_o on ewe“ L~ ucN O w— Ne_ ew Ne we meN cu Lemme cu .eu.3.:m.em.zu cm -I- -u- -|- em mN OF m ON w.O O~ N.O Om— e_Lu my OON on ewe“ m_ ucN O e— Ne~ e— me we OOON Ill :O LawFQ cu .ee.3.cm.mz.zeuum --- --- --- me ew m” m mN w. OF e.e ON_ nee“ NF ucN cu e_gu NF um_ Om e_ Ne_ ew me we NOON .LO saw—O ou .eu.3.:m.em.=m ON .I. --- --- mm mm O_ m ON w. O_ N. om e_gu em ucN 0» even my amp Om em New ew me we emmN I ‘ .5 L38 3 .nO.z.:m.em.zm O; -l- --- --- em Om m_ OF Om w. em w. OO_ ewe“ mm OcN on eye» em uw_ Om e_ Ne_ mw me we eemN :O Lewwu on .ee.3.:m.em.2muva -l- --- --- oe mw ep 4 mN N. e e. Oe— a_Lu w_ ucN o» ewe“ m_ em_ Om e_ Ne— ew me we wme .11 w! III .mmgm « xw_=u:ezk .ne.3.=m.em.ze ca --- --- --- oe ON_ O_ m mN e. ON w. OON .LO Lemme ow ewe» $_ ucN O we Ne_ mw Ne we meON .nveu .mx O; N. N 4 mm mm O_ m eN w. or N. owp up “mm m>oem Ex m.O LO Lamme O __ Ne_ Om me we NeON .evgu um_ m>oeu Ex .ez u; 4 4 4 mm eN em m Om 4 O_ 4 OO w.O OwnEamLum u_o m ewfizuapzh O _F Ne_ Om me we _OON 11 go Law—O o» nwgu .mcoz --- -.- --- mm mm ow e em 4 O_ w. oe eh em_ .mwea x 4w_:u=ezm ew m_ Ne_ mp we we OmON l :1 .5 Laws on .mcoz 4 Om ew e_ m Om e.e m_ 4 OO_ ewge $_ umw cu nee“ u; :uw ew m_ Ne_ e_ we we ONON .wcoz --- - mm mm ep 4 Om w. O_ w. Om_ -------LO gmmFO o» n_Lu e_ um_ ew my New e_ we we NNON go Law—O cu .mcoz m. 4 4 ew eN O— m eN w. O— N. O_N ewe“ m_ um_ cu ewe“ u; uem ew m_ Ne_ e_ we we eNmN II :O Law—O cu .wzoz --- --- --- mw Om e_ m mm 4 m_ 4 OF_ ewe“ w_ um_ on a_Lu a; OEN ew m_ Nee e_ we we eNmN 5.75 3 .mm On 4 m. 4 me eN e_ 4 Om 4 e w. Om um_ eo sauce w>oem LG Lam—O O __ New Om me we wNON .: 9:8 t 2:; . .exnum 4 4 4 me eN m_ O_ mN w. OF 4 ow_ ea LO Lemme eo Lee m_nnou e4 O w New ew me we mNmN I1 cu Luw_O .em cm 4 4 4 ex ON mm m Om N. OF w. Om zo_wn Ex w.O m Nw_=u:e=» O O Nep ew me we NNON .mcoz 4 4 4 eN e_ Om e em e. O_ N. z ...... --O_ name an m xw~=uzezm mw m Ne_ Om Oe we ONoN xx .mxe 3; sopmn .mcoz 4 4 4 me my ON 4 Om N. O_ 4 Om m xzw:_zx ow ewe“ a; ucN Om eN Ne_ em ow we meON .mmsm m xawcwzx .3.mm.:e Om N. 4 4 mN mN e_ 4 Om e. O_ w. OON m_ mace m>onm ewe“ m_ Ogm Om NN Ne_ O mw we FemN II .O— name m>oem mccm_mw .3.em.:u O; 4 4 N. mm em m_ e eN e. O_ N. Oe m>oec Lug em m xzecwzg Om NN Ne— O mw we OemN I1 .x xaw:_3x :0 O_ .3.em.:O O; N. 4 me er ON m Om w. OF N. O_F aEmO m>oem e_gu ew um—aw>oe< m_ Om Ne_ O ww we mmON II 5— Ewe .3.em.=enum .:n\: ..n\\. ---1\ ON .;- ON ON O_ em w.O m N.O O__ m>oeu u:m_m_ zoroe m 4mwcw2¥ e_ mm Ne— O ww we OmON .OF Ewe , .3.ez.:Onua 4 4 4 me ep e_ 4 Om 4 OF 4 oe eo ucoge :_ gun e— z xswcwzx mm m_ Ne_ O ww we NeON . .II. II! x! .O_ name zo_wn .3 em.=u On .-- --| --- OOe ow OeN O_ mm e. Ow e. OO_ m xzw:_3¥ 0» are» ew um_ em O_ Nee O ww we eeON . .O_ game gamma .2 ex.:u um - .-- ,-- Om m_ m_ m mm 4 OF 4 ow e_Lu a; saw w>onm m 4:W:_zx mm O_ Ne_ O ww we wemN xv .m xzwc_3x .wcoz --- --- eN ON ON m em w. OF w. Om :0 nee“ u; ugm 30—wn 54 0.0 O w New O ww we FemN wwzcwpcoe--A~H vcm .=H .aw Low em~m_m:m Fw ”mmumgucmucou _O—v ewwmcmgvmac case—om 3.8 3.8 3.: E E E E E 3.8 3 V 3.8 88 E E 2 :N a“. E 3 8 8 .5 3 .5 28.583 3 $.53:me __ _ c __ _ c megcm: mu_moawu wu:u_O:O4 we:u_um4 Lweazz N Acow——WE Lwa muLanv w:o_umgucwucou ucwEme “mm: :uLoz w_we FmLm:_E czacx cosfiunoolmfimcmwfi 335$ 2305‘ 2 3333 .333V 52.x geek—3289 Paw—ES: nvmko mmmaeutukc 33me 359 m=oe§ooQ|A mam—<9 13 ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA .mzuum .mmuua .mmuum .mmuvm .mxuum .mm.um .mmnvn .mznun .wcoz .wcoz .wcoz .wcoz .mcoz .wcoz .wcoz .mcoz .mcoz .m:oz .mcoz .mcoz .wcoz .mcoz .mcoz .wcoz .mcoz .8.=.cm.mm.zn_6n_ .nu.3.:w.mm.z$uum .nu.:.:w.mm.zmnvm .nu.2.cm$mz.2$“um .nu.3.:m.mz.:unva _l_l .J-J .—l_l _J_J ow mo mlw on me am mm 0m mm Cu mm mm mm ON am me am om mx 2 m cm .4. 2 m on 4 m: 2 mm .4 2 m mm o. 2 m 8 4 E 4 cm olN m om 4 2 4 om N. E m on q. 3 m mm m. 2 m m 4 2 4 8 q. S m m «a m: 4 mm .4. 2 m mN N. S S cm 4 S o. 8 c. 2 4 8 q. S m mm 4 DIN m mm w. elm m cm 9. 2 2 cm 4 2 m mm v. m: m ow 4 E m Q q. 8 m WM 44. S m 8 v. 2 m 8 m. S 2 mm 44. 2 S on o. O .— | eel 0' Ln .— r— mom'm ~.—.—-.— mmoomm _|_Jr.__l._ OLD :—v- o— op % S of» fl Q‘Q’WQ N. ow om— omp cow om_ on_ % chm cum ow— om o__ On oo o_N ov— cow GNP oo_ oo— OVF oo— om ok— omp om_ 9.: cl: a a to Lam—u zopma “man m 4$_:u:a=h .nwgu $— sum m>ono E om Lo me_u WLQ gum—u $o n_Lu $_ cum .mmgm x ¥$_:u:n:$ .Lu Law—o $0 D_Lu $— gum $9 guzoz cu Law—u $0 nwgu $— sun .uwxu m x__=u=n:h -------Lu gumFU op nwgu $F can .Lu Low—u $o pwsu $_ can 0» x$ um .Lu Lew—u $0 nmgu $_ sue cu x$ 3; $4 -------Lu Luwpo $o nwgu $F gum .mn_su $F gum ccm saw :44 so Law—u .wFOm m_asmm 30—wn 2 me go meFu .n_Lu $F zu¢ m>onm 5 mm. Lu Lew—u cu Lam—o cu n_gu $F ugm w>onm a_gu as new .50 Lam—u cu a_gu $_ cgm m>onm n_Lu u; um_ cu Lampu ow n$Lu $_ ugm $o 4$ 3: um cu gum—u cu n_$u $_ vLm $o 4$ $4 :u Lam_u ow n_Lu $— uxm $o m4$ 3: $o new 3o_wm go gmmfiu cu nwxu $_ ugm 0» n_$u a; cgm cu Luw_u cu nwgu $_ ugm cu nwgu 4L c:~ $0 m4$ 3am .Lu Law—u cu n_Lu $— ugm op a_Lu a; ucw $o 4$ $4 .Lu me_u cu nwgu $_ ugm cu a_gu u; u:~ .gu Lam—u op a_Lp $_ vLm .mwgm m ¥$_:pznzk ..... --$u Low—u cu nwgu $_ uLm .n_La $_ vLm m>onu Lu mepu .cmgm m x_—=p:n:h :u Law—u cu nwgp $_ vgm sopmn Ex w.o v:m_mH ----- ..... N_ game um Lam ~_mEm .Lu Lumpu ou n_Lu $— ucm $o Ln 3; um wLu Lam_u on n_su $_ ch $o gs 3; $4 wLU gmw_u on nwgu $_ ucm op nwLu $F cuv w>one onFm no gum—u op nwsu $_ new op n_Lu $_ :uv o w No— on om em on cm on on m— m_ m— m— mp mfi mp m— cm om ow $— up a— op mp @— mp wF RF A— m— m_ NF $_ n_ n— w— w_ «P on @— wh NwF ww— Nap ww— mop Nmp Ne_ mm— mm— No_ mm— Nap mofi Ne— ww— No_ mm~ mmp No— No— Nay Nm~ Nop No— No— Nap No— No— No~ 0m om cm on em OOODOC LDC) m¢ mv me mv mp mp m— mm mm Nm Fm ~m mm mmom ¢Nom mmom mmom Fwom omom mpom w_om NFom m—om mpom <_om mpom mpom __om opom ween woom meow meow mQOM voom Noon _oom ooom wmmm mmaw comm mamm vomm umzcwucou-‘flph van .:4 .:< Lo$ vaxpm:m _< mmwumgucmucou Fo—V wmpmcmgvmzc coeoFom EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 14 .36; -- -- -- OOO.N mN OO ON OO O O_ N. z .................... LO umxoo-LO Om «_ N: m— N NO NOE I I I .rLO .Om.=8m E OO O i332: Om O NO_ m_ mm «O OmOm .282 I- I- -- mm mN Op ._ mN «. O_ «. O: ...... LO :82; cu PC; OL :3 Om 2 NO_ Om Om «O mmOm I I .LO :82; O0 PCO H... .952 I- I- I- mO Om ON m Om O. m— «. ON OLm m>onm 8.; m LIES?» Om 2 NO_ Om Om «O «mOm I ATS Mai-LEONE I- I- -- Om mm or -_ Om «. OF «. OON .LL OLm 323 E mNN LO :32; m« 2 NO_ O Om «O mmOm I .qum 1 x1333 @155qu I- I- I- QI— mN m_ m mN «. O_ N. OON LO :32; 3 f: “L OLm m« m_ NOF O Om «O NmOm I .92» .282 -- -- I- ON m« m_ m Om O. S N. ONN OL ch :38 Oman LO c325 Om 2 NO_ Om Om «O FmOm .252 I- -- -- Om ON 3 OF Om O O_ O ONP LO :32; £86 m i332; O N NO_ Om Om «O OmOm .fLO .OmiudOuE I- I- I- OO Om mL m mN O. O_ N. ON_ OL OcN Lo O8; L8: LO :32; m« 2 NO_ O Om «O ONOm I I .chm x . .wcoz -_ -_ O 9 OF ON 5 O_ «. m N. OO i338; .LO :82; a: E ONN Om O NO_ m_ mm «O ONOm I .LO Lam—O mo .wcoz -.- -- -- mm mN mm m mm -_ O_ ._ Om 5:05 58w Lama: 32$ EN Om O NO_ 2 mm «O ONOm OOOOCOOOIA C use .5 .:< L8 3532; .« $322830 85 OmpmcmLumac coca—om ANOO NOV ANOO A: my CO A: A: ANOO Amv 8.8 88 i S 2 ON an E :O 0O OO .5 c< Ow 18.583 mo 5.59ng __ _ O a __ _ o \ELam: 3.582,. mg“: :3 33.53 33:52 V ELOEE ESE. 2057:: L3 3,33 20552350 2255 3%. 532 BE wonfiunoolmhmommw 335E £393 8 3333 «imam—x Set 33:59:?» otugmufi «v58 mumbaeuko 33me NE: anotaoofllé mqmfir 15 ANALYTICAL PROCEDURES AN D RELIABILITY OF THE CHEMICAL DATA .wawwu cw—m —sz m¢=~u> mzopueo:m 0m .Ao_ w_amu wwmv wgmgzmew mmocu EoLm ucwgwmwwu wLm m—mcmgvmzc :oeopom msu cw mw=~m> UFOsmwgsh c .m.m :mgu Lmumem >~?mcwo m .memmu em_m ——wz mw:_a> mac—mac:m om .Ao_ m—nmu wwmv wgmgzmw_w wmozu 50L» ucmgmmwwu mgm m—mcmgumzc vacnu wnu :_ mwspm> uFosmmggh 1 .wuwgucwocou wmumgu go xon muw=_m Eosw um>_Lwo m .m_c_gwans w_nm:o_mm_w n z; memmewE :ugwm wgag H mm uuwmonmu Lwom_a n ca mummoawu muo_ u cg ”cm:w_gwu:= mgm :owuoacoga mo ugouwg m saw: mucwsw_m N .uum mgugzow “saw ”cgwsu .vLm Mucouwm .vcm "umgww .ump ”ume .3 uewwguma: .Luma mxgmpsnwgg .nwgu ”Emwgum .Lum «macmgqm .mmam ”zu30m .m mugmwg .u; acme; .v; mgm>wm .m mu:_oa .ua mgmnE=: ..o: usage: .2 ”cwmuczos .HE mm_vu_e .qu mummy .m_ ”cowuuczn .u0m ”Ugo—m? .m_ mxngmw; .Azc mflmwgmucsnmm; .Hmvzg ”Amvxsow mummm .m “Emmgumczou .Lumu ”Amvxmmgu .vagu ”oamgucwucou .ocoU uxcmanu ..ou “cwmzuwn .zun ”Awmvcocmgn .Amvgn ”:o_umuo— mnwgummu cu now: mcomum_>wgnn< x .mcoz 4 blm m_ on WW om N. m N. z -------- ..... m mcw:u_wz w—up_4 0 cm me_ me am Fe NcFN genuA -l‘ -- bah mm om— mm mm N. 4 w. z ....... -..----------su ogmcgwm om N map on Am _o mmpm .cN.=uuc4 ‘-- -- oo_ m_ om? ow ON 4 m ~.o 2 .-----Lu xuom om m mv_ om «w ”o _mp~ A_h ucm .:H .=< Lo» uw~>rmcm F mmmumgucmucou mv m_m:mgumac Nmu_n> .cmuua 4 oo—._ m_ oo_ 4 ON V. mm q. z -l----------wxp_u :wh .Lo mama m_ Km mo_ om mm mm Now {I .mmcwe Lmum zugoz .cm.nu um --- -. om op oo_ m om ¢.o 0— 4 2 Law: Ewmflu Joe—Lm>o .Lu wamu 0 mm Rwy me vm mo cmq ALF .cH .=< Lo» vm~»_mcm _ mmwumguzmucou NV w_a:m;cm=c Lm__w» .342; -- 8 w a a E q. E a 2 ..........:ms 3 2.9; .533 o S 9: cm a 3 EN A_H vcm .:H .=< Low um~zpmca Ho: ”wumgucwu:0u Fv mpmszcmzc mcmcmh . II .1 xop .mcoz --- .. mph mN om mm owl m. A e. z w_uu_4 Co gazes um uau cmom om mp me, o m mo oomF wcoz --n -- onw mm omy om o<_ N. a N. 2 wmv_ we mz Ex m z xop me n_;» om v_ me" on N no mav— ‘1. .m uwgn .mcoz --- -- mm m_ mm mN ow N.o 4 4 2 as: m>onm 5 cm Lo xuwggm> 0 mm mwp mw mm mm mm<_ A—P uzm .:~ .=< Log um~z_m:w mzo: mmwumguzmucoo my mchchmac mmoLumcmh 16 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA from the original suite of 1,072 concentrates to test the repeatability of analytical results. Hence, this sample appeared 31 times in the set of magnetic concentrates weighing 1 g or more: once as a sample and 30 times as a replicate subsample. The control sample (number 300DL) from Nevada was detrital magnetite collected in 1964 by D. E. Lee, US. Geological Survey. The magnetite was separated from a table concentrate of a placer in the canyon floor at the mouth of Hampton Creek at an altitude of 1,860 m on the east side of the northern Snake Range in Humboldt National Forest, Nevada. The principal source rocks for the placer concentrate were Lower Cambrian clastic sediments metamorphosed to garnet- staurolite-muscovite schist (Hose and Blake, 1970). Carbonate rocks of Cambrian age are locally present, and small areas of the basin are underlain by igneous rocks. Thus, the magnetite in the Hampton Creek con- trol sample was probably derived from several kinds of source rocks. The control sample (number 52—WE—819) from North Carolina was panned in 1952 by A. M. White, US Geological Survey, from gravel in the bed of Lyle Creek, Catawba County. The basin of Lyle Creek is underlain by metamorphosed sedimentary rocks, mafic rocks, and intrusive quartz monzonite (Theobald and others, 1967). Separation of the magnetic fraction from the panned concentrate and division of the magnetic fraction into appropriate subsamples for replicate analyses were done in 1971 by A. L. Larson, US. Geological Survey. The control sample of magnetic concentrate (number 3799) from Alaska was prepared in 1971 by A. L. Lar- son from a sample in the Alaskan placer concentrate file. This sample was collected in 1949 by A. E. Nelson, US. Geological Survey, in the Bowman Out of the gold placer on Portage Creek at an altitude of 300 m on the north side of Lake Clark in the Lake Clark quadrangle (table 1). The placer was exploited from 1910 to 1912 and for a few years after World War II (Cobb, 1973, p. 11—12). Portage Creek drains an undivided complex of mafic lava and tuff with considerable metamor- phosed sediments and some intrusive rocks locally present (Capps, 1935, pl. 2). This complex was re- garded as Early Jurassic to Cretaceous in age. Doubt- less the lavas were the main source of the abundant magnetite in the concentrate. EQUIVALENT URANIUM The radioactivity of 347 magnetic concentrates and 30 replicates was measured by K.-L. Pan to determine the equivalent uranium before the samples were dis- solved for chemical analyses. To measure the equiv- alent uranium, about 1 g of the magnetic concentrate was placed in a lead shield 4 cm thick in which an end- window Geiger-Muller tube was mounted. Radiation was recorded in counts per second by a scaler~type in- strument operated at 1,450 volts. A counting time of 256 seconds marked by automatic timer was used for the samples, and background counts were made over the longer interval of 2,560 seconds. By comparing the counts recorded for the magnetic concentrates with those recorded for a standard sample with 0.1 percent uranium obtained from the US. Atomic Energy Com- mission, and correcting for background, the radiation of each magnetic concentrate was recorded as percent equivalent uranium (eU). Because of the high random fluctuation in both intensity and direction of l3-particle and y-ray emission measured by the instrument (Wayne Mountjoy, oral commun., 1971), the cutoff value was defined as twice the standard deviation of the counts, and was determined to be 0.003 percent eU (30 ppm). The results of these analyses are listed in table 1. Higher-than—background radioactivity was not de- tected in any of the 30 subsamples of the replicate magnetic concentrate (3799) from the Portage Creek placer. Thus, it is difficult to draw a conclusion about the precision of the eU determinations. However, they are thought to be reliable because of the consistent results from many repeated measurements of the 0.1 percent uranium standard and from the background counting. EIGHT ELEMENTS BY ATOMIC ABSORPTION The method used by K.-L. Pan to determine the abundances of silver, bismuth, cadmium, copper, cobalt, nickel, lead, and zinc in the magnetic concen- trates was modified from the single-solution, atomic absorption procedure developed in the US. Geological Survey by ‘H. M. Nakagawa (1975) for use with iron- rich samples. The modification, worked out with R. L. Turner, employed the following steps. One gram of magnetic concentrate was decomposed without grinding (as a way to save time in geochemical exploration) in a test tube in 15 ml of concentrated HCl under moderate heat. After evaporation to dryness, the residue was dissolved in 4 ml of concentrated HNOa and brought to dryness. The residue was then dissolved again under warm heat in 1 ml of concen- trated HNO8 and 9 ml of 6N HCl. After filtration, the abundant iron in the sample solution, which interferes with the determination of the minor elements, was extracted by 0.5 ml of concentrated HBr and 10 ml of methyl isobutyl ketone (MIBK). This extraction was repeated several times, either until the organic layer ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA 17 that was discarded was colorless or until the color ceased to change. The iron-free sample was then read on a Perkin-Elmer1 290 atomic absorption instrument. The concentrations were determined in parts per million (ppm) in the same single solution, with lower limits of detection as follows: silver, 0.2 ppm; bismuth, 5 ppm; cadmium, 0.2 ppm; cobalt, 1 ppm; copper, 1 ppm;'nickel, 1 ppm; lead, 5 ppm; zinc, 1 ppm. About 15 samples (120 determinations) were analyzed in one man-day. 'FliS'FEi()l? I’I{()(IIEI)IJI{IE Three tests of the reproducibility of the results by the eight-element procedure were carried out on two of the control samples of magnetic concentrates. The pur- poses of these tests were to determine the suitability of the procedures for typical applications in geochemical exploration, and to consider aspects of sample prepara- tion prior to analysis. Because the acid digestion used in the procedure takes unground magnetite into solu- tion, the grinding of the magnetic concentrate could be eliminated to reduce time needed for the analysis, pro- vided unacceptable bias is not introduced. Inasmuch as the samples were dissolved twice in different con- centrations of acids over moderate heat, the efficiency of the digestion could have been affected by many physical and chemical factors. The only way to lessen this kind of bias was to maintain identical conditions of heating and digestion, so far as possible, for each lot of samples. An effort was made to maintain identical conditions in the three tests described below. The first test was designed to evaluate reproducibil- ity for each of the eight elements in unsplit and unsized samples. The control sample from the Hampton Creek placer was chosen for this test because it was large and the magnetite consisted of poorly sorted grains. Twenty-four unsplit and unsized specimens were vol- umetrically scooped out of the sample container, and each specimen was analyzed just as it was scooped. The results are given in table 2, which shows that the variance for all elements except silver and zinc is relatively acceptable. Silver and zinc show great varia- tions, and several individual values approach or are greater than twice the arithmetic mean. These varia- tions may be caused by: (1) inhomogeneity of the sub- samples, constituting a sampling bias; (2) variations in the content of minor elements in the magnetite related to grain size and source; and (3) errors in analysis. The chemical variations shown in table 2 probably reflect differences in the relative amount of coarse and fine grains of magnetite in the subsamples. 1The use of trade names in this publication is for descriptive purposes only and does not constitute endorsement by the US. Geological Survey. TABLE 2.—Replicate analyses of unsplit and unsized control sample of magnetic concentrates from Hampton Creek placer; Nevada, for Ag, Bi, Cd, Co, Cu, Ni, Pb, and Zn [Data are in parts per million] Elements Subsamp1es of 300DL Ag Bi Cd Co Cu Ni Pb Zn 1 0.96 12 1 2 48 59 48 420 60 2 1.44 12 1 6 48 39 48 420 40 3 1.20 9 1 6 47 54 48 420 40 4 1.92 9 1 2 47 44 36 420 44 5 1.80 12 1 6 48 52 48 360 112 6 2.64 9 1 6 47 54 48 420 44 7 1.68 12 1 6 44 44 48 420 48 8 1.44 12 1 6 44 38 36 540 56 9 1.44 12 1 6 42 36 48 540 36 10 5.10 15 1 6 48 49 48 354 133 11 1.92 12 1 2 47 53 48 420 60 12 1.44 12 1 6 44 41 48 420 36 13 1.44 12 2 0 47 42 36 480 100 14 .72 12 1 6 41 45 36 420 112 15 2.70 12 1 6 48 42 48 396 119 16 .96 12 1 6 50 38 36 420 64 17 1.20 9 1 2 48 40 48 480 68 18 .72 9 1 2 42 40 36 420 64 19 .96 9 1 6 48 48 36 480 76 20 1.80 12 1 6 46 50 48 336 144 21 .96 9 1.2 42 42 36 420 56 22 1.68 12 1.2 47 60 36 420 80 23 1.44 12 1.6 47 52 48 480 92 24 1.44 12 1.2 45 49 36 480 84 Mean 1.15 11 1.5 46 46 43 433 74 Standard deviation 1 oo \ 1.5 22 2 4 6 9 6 0 50.4 31 9 Re1ative standard -deviation (percent) 87.0 14.5 15.3 5 2 15.0 14.0 11 6 43.1 Detrital magnetite in replicate sample 300DL from Hampton Creek is a mixture of grains derived from several source rocks. The minor elements in a detrital mineral display a tendency to vary by grain size, and the grain size tends to vary by source (Overstreet and others, 1970). Distinctive populations of minor elements have been found to characterize magnetites from individual intrusive rocks and mineralized districts (Hamil and Nackowski, 1971). Thus, the physical and chemical properties of grains of magnetite in a detrital mixture will be irregular and in- homogeneous. Variations in grain size of the magnetite in the 24 subsamples from the Hampton Creek control sample probably caused the great variations in the results of the replicate analyses. The second test was designed to examine possible variation in the composition of the unsplit magnetic concentrate related to particle size. For this test the Hampton Creek control sample was sieved to give four sized fractions (mesh): +42; -42+80; —80+170; and —170. One subsample, called here the original sub- sample, was scooped from each size fraction, which ex- hausted the supply of the coarsest material but left 18 enough of the smaller sizes to permit replicate analyses to be made (table 3). The data from this test show con- vincingly that the content of minor elements in the coarsest grains is substantially different from that in the finer fractions. The data in table 3 suggest that in a well-sized sample the material can be volumetrically scooped instead of split to acquire a suitable sub- sample for analysis. However, further study of the relations of the minor-element content, grain size, and source of detrital magnetite is evidently needed. TABLE 3.—-Replicate analyses of unsplit but sized control sample 300DL of magnetic concentrates from Hampton Creek placer, Nevada, forAg, Bi, Cd, Co, Cu, Ni, Pb, and Zn [Data are in parts per million] Sieve fraction Subsamp1e Ag Bi Cd Co Cu Ni Pb Zn (mesh) +42 0rigina1- 2.2 76 2.9 91 67 652 362 106 -42+80 0rigina1- .5 13 1 50 41 60 233 128 Replicate .5 13 1 50 40 65 233 130 --do ----- .5 13 1 50 43 60 240 130 --do ----- .5 13 1 50 40 60 233 142 —80+170 Origina1- .5 13 1 48 39 50 300 143 Rep1icate .5 13 1 50 41 50 300 145 --do ----- .8 13 1 50 50 60 300 150 --do ----- .8 13 1 50 40 60 300 167 -170 0rigina1 .5 13 1 50 50 50 400 140 Rep1icate .5 13 1 50 50 60 400 157 Sizing by screens is less convenient than sizing by grinding; therefore, grinding is indicated unless the original concentrate can be split into a representative magnetic fraction containing essentially the same dis- tribution of particle sizes as the original concentrate. Also, sizing by screens would eliminate from analysis some extra large or extra small grains whose composi- tions would be needed in a geochemical survey to rep resent a group of rocks or ore deposits that would otherwise be missed. The inhomogeneity of magnetic fractions should be easier to reduce than many of the other biases that lead to variance in analytical results. The standard splitting procedures used in sedimentary petrology should give subsamples of a magnetic concentrate that are of appropriate weight for analysis and contain represent- ative parts of all grain sizes in the original sample. A split of this sort could be taken into solution without grinding. A test was made on the control sample from Lyle Creek, North Carolina, to examine the results of replicate analyses of an unsized sample split into sub- samples by the use of a CARPCO microsplitter having 3.17-mm (1/8-in.) chutes. The bulk sample of magnetic concentrate used for the test was first cleaned five EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA times with the hand magnet, then it was split into 32 subsamples with the CARPCO microsplitter. The first 10 subsamples were further split into 20, so that the original control sample was divided into 42 sub- samples. Each subsample was assigned a number and every fifth was analyzed (table 4). The data in table 4 show that the metal contents of the l/64-split speci- mens (numbers 5, 10, 15, and 20) have no major differ- ences from those of the 1/32-split specimens (numbers 25, 30, 35, 40), though some minor differences can be seen. The results of the replicate analyses of the sub- samples of the carefully split but unsized control sample have acceptable standard deviations for geo- chemical exploration. They show that preparation by splitting, digestion of the unground subsample, and single-solution determination of the eight elements constitute an acceptably accurate method for geo- chemical exploration. TABLE 4.—Replicate analyses of split but unsized control sample of magnetic concentrates from Lyle Creek, North Carolina, for Ag, Bi, Cd, Cu, Ni, Pb, and Zn [Data are in parts per million] Subsamp1es E1ements of 52-wE-819 Ag Bi Cd Co Cu Ni Pb Zn 5 0.9 9 1. 2 46 44 48 280 136 10 0.9 9 1.2 48 47 48 280 132 15 0.9 9 1.2 46 47 48 280 132 20 0.9 9 1.2 45 45 48 280 132 25 0.6 9 1. 2 48 45 48 280 136 30 0.9 12 1. 2 50 47 60 280 136 35 0.9 9 1.2 46 46 48 240 136 40 0.9 9 1.2 45 44 48 280 128 Mean 0 9 9 1.2 47 46 50 275 133 Standard deviation 01 1.06 0 1.75 1.30 4.24 14.14 2.98 Re1ative standard deviation (percent) 11.1 11.8 0 3.5 2.7 8.0 4.8 2.1 INTERNAL REPLICATE ANALYSES IN THE FULL DATA SET The results of the analyses by K.-L. Pan of the full set of samples for silver, bismuth, cadmium, cobalt, copper, nickel, lead, and zinc, are given in table 1. One of the samples (3799) in table 1 was also inserted as an additional 30 subsamples in the 347 samples of the set. The subsamples of 3799 were prepared in the same manner as the subsamples of 52—WE—819 (table 4). Thus, several subsamples are in each lot of 50 that was analyzed. The advantage of this array is that the reliability of the analytical data can be evaluated as a whole, or by the lot. Inasmuch as the replicate sub- samples were hidden in the. numbering system pro- vided the analyst, any factor of operator bias during ANALYTICAL PROCEDURES AND RELIABILITY OF THE CHEMICAL DATA 19 TABLE 5.—Replicate analyses of split but unsized subsamples of file number 3799 from the Portage Creek placer, Alaska, for Ag, Bi, Cd, Cu, Ni, Pb, and Zn [Data are in parts per million. L=detected but less than lower limits of determination] Subsamples Elements of 3799 Ag Bi Cd Co Cu Ni Pb Zn 1 0.8 15 0.2L 100 40 100 270 210 2 0.8 15 0.2L 100 50 100 270 210 3 0.4 15 0.6 100 30 100 270 210 4 0.4 10 0.4 100 30 100 270 200 5 0.4 10 0.6 110 35 100 270 210 6 0.4 15 0.6 100 35 100 280 210 7 0.2 10 0.6 90 30 100 260 190 8 0.4 10 0.4 90 30 110 270 210 9 0.4 10 0.4 85 30 100 270 190 10 0.6 15 0.4 90 30 140 240 230 11 0.6 10 0.4 100 35 100 260 220 12 0.6 10 0.4 100 35 110 270 220 13 0.4 10 0.4 85 35 100 290 220 14 0.6 10 0.4 100 35 110 260 220 15 0.4 10 0.4 90 35 100 260 210 16 0.6 10 0.4 90 35 110 270 220 17 0.6 10 0.4 90 35 110 260 210 18 0.6 10 0.4 90 35 110 270 210 19 5.5 5 0.4 90 40 100 250 250 20 0.6 10 0.4 90 35 100 250 230 21 0.6 10 0.4 90 30 110 270 230 22 0.6 10 0.4 100 40 100 270 240 23 0.6 5L 0.4 90 45 100 250 250 24 0.6 5 0.4 90 30 100 260 190 25 0.6 5 0.4 100 35 110 270 200 26 0.6 5 0.4 100 35 100 270 210 27 0.6 5 0.4 90 35 110 270 220 28 0.6 5 0.4 90 30 100 260 200 29 0.8 5 0.2 100 35 100 270 210 30 0.8 5 0.4 100 65 110 270 220 Mean 0.7 9.3 0.4 94.7 35.8 104.7 265.7 215.0 Standard deviation 0.91 3.5 0.1 6.1 7.2 8.2 9.7 15.3 Relative standard deviation (percent) 130.0 37.6 25.0 6.4 20.1 7.8 3.6 7[ the analyses was reduced to a minimum. The results of the replicate analyses of the 30 subsamples are given in table 5. The relative standard deviations shown in table 5 are larger than those reported in table 4 for all elements determined except lead and nickel. It is thought that these poorer results are probably due to instrumental noise and variability. For example, the amounts of silver and cadmium in the magnetic concentrates are low, with the exception of one aberrant subsample unusually rich in silver, but the smallest reading that could be made on the test instrument for these two elements was 0.2 ppm, regarded as the lower limit of detection for each. Thus, the large variance may be partially a function of size of reporting interval and concentration present. The lower values for silver and cadmium were not easy to read with certainty, because of the fluctuations of the meter (instrumental noise). Thus, the lower limit of detection for silver and cad- mium probably should have been set at 1 ppm each, which would have greatly reduced the relative stan- dard deviation for these elements. Indeed, by 1975 the lower limit of detection for silver and cadmium was raised to this level (Nakagawa, 1975). Another ap- parent reason for the larger standard deviations in table 5 is that the replicate subsamples in that table were digested at different times in different lots, whereas the subsamples in table 4 were digested at the same time. However, at the 99.7-percent confidence level the means are not far from the true concentra- tions of these eight elements. GOLD, INDIUM, AND THALLIUM BY ATOMIC ABSORPTION Gold, indium, and thallium contents were deter- mined by A. E. Hubert and G. L. Crenshaw on the 140 magnetic concentrates weighing 2 g or more remaining after the eight-element analyses. Thus, the data for these three elements are for part of the set only: 131 of these samples are from the main set, including file number 3799, and 9 are replicate subsamples of 3799. The three elements were measured by atomic absorp- tion with a lower limit of detection of 0.2 ppm each in a single solution of 2 g of sample. The procedure was modified from the technique for gold described by Thompson, Nakagawa, and Van Sickle (1968) and the technique for indium and thallium presented by Hubert and Lakin (1973). After repeated digestions of the unground sample in hot HCl, the mixture was evaporated to dryness, and the residue was then dissolved in a Br-HBr solution. This solution was heated to eliminate excess bromine, then diluted with water, and the metals were extracted in methyl iso- butyl ketone (MIBK). Indium and thallium were ex- tracted from the MIBK layer with 1.5 N HBr and were determined by atomic absorption spectrophotometry. After this determination, the remaining MIBK solu- tion was shaken with 0.1 N HBr to remove iron, and the gold in the MIBK solution was measured by atomic absorption. Seventy-five determinations were made per man-day. The results are given in table 1. Owing to the small amounts of gold, indium, and thallium in these magnetic concentrates, too few replicate subsamples of sample 3799 were left in the set to permit a study of variance in the analytical results. One of the nine subsamples was found to con- tain 0.3 ppm gold and 0.3 ppm indium, and another subsample had 0.2 ppm indium. The amounts of gold, indium, and thallium in all other subsamples were below the limit of detection (<0.2 ppm), except that 20 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA sample 3799 itself registered 11.1 ppm gold. The great difference in the amount of gold determined for sample 3799 and for its nine subsamples is attributed to the variable presence of particulate gold, even in carefully split samples. This problem is further discussed in other sections. DISTRIBUTION OF THE ELEMENTS IiAl31((}I{()IJI¢I) 1113\7113VV Almost onethird of the 347 magnetic concentrates are from the Solomon quadrangle, and one-fourth are from the Candle quadrangle (table 1). The remaining samples are scattered in 31 other quadrangles (pl. 1). Owing to this lack of balance in the distribution of the samples, the discussion of the results of the analyses is given under three headings instead of by 33 quad- rangles: (1) Regional results; (2) Candle quadrangle results; and (3) Solomon quadrangle results. Note that the regional results encompass data from the Candle and Solomon quadrangles. The distribution of the elements is described by methods of analysis and by the geochemical associa- tions of the elements in the region and in the Candle and Solomon quadrangles. Thus, equivalent uranium, which was determined by radiometric counting, is discussed separately from the 11 elements determined by atomic absorption spectrophotometry. These 11 elements are discussed in five geochemical associa- tions: (1) copper, lead, zinc, and cadmium; (2) silver and gold; (3) bismuth; (4) cobalt and nickel; and (5) indium and thallium. The magnetic concentrates used to determine the distribution of these minor elements are not monomineralic separates of magnetite; rarely they may contain as much as 50 percent of other minerals, including some recognizable sulfides of the base metals. Thus, some of the minor elements contained in the magnetic concentrate may be in minerals other than the dominant magnetite. Variations in the minor- metal content of the magnetic concentrates from a given district, or from several districts, may reflect in part—possibly in large part—variations in the abun- dance of associated minerals other than magnetite. Keeping this condition in mind, and recognizing that the discussion will return in detail farther along to the roles of the various minerals as sources for the minor elements, this introduction is concerned with the prob- lem of minor elements in magnetite—a subject on which there is an extensive literature relevant to ex- ploration geochemistry (table 6). Minor elements in magnetite have been investigated in studies of the genesis of ore deposits, principally TABLE 6.-—Selected references on the composition of magnetite Subject Authors and date Origin of the deposit --------- Carstens, 1943; Chakraborty and Majumdar, 1971; Chistyakov and Babanskiy, 1971; Deb and Banerji, 1967; Duncan and Tay1or, 1968; E1sdon, 1972; Green, 1960; Green and Carpenter, 1961; Hami] and Nackowski, 1971; James and Dennen, 1962; Kise1eva and Matveyev, 1967; Komov, 1968, 1969; Ksenofontov and Davydov, 1971; Landergren, 1948; Lewis, 1970; Linds1ey and Smith, 1971; Lipman, 1971; McKinstry and Kennedy, 1957; Marmo, 1959; Nagaytsev, 1971; Némec, 1968; Neuerburg and others, 1971; Newhouse, 1936; Oshima, 1972; Pavlov, 1969; Ramdohr, 1940, 1962; Rost, 1940; Sastry and Krishna Rao, 1970; Shangireyev, 1969; Shcherbak, 1969, 1970; Vartanova and Zav'yalova, 1970; Vassi1eff, 1971. Be1kovskiy and Fominykh, 1972; Bocchi and others, 1969; Borisenko and Zo1otarev, 1969; Boyadzhyan and Mkrtchyan, 1969; Deb and Ray, 1971; Duchesne, 1972; F1eischer, 1965; Fominykh and Yarosh, 1970; Frietsch, 1970; Howie, 1955; Kisvarsanyi and Proctor, 1967; Lauren, 1969, Leung; 1970; Lopez M. and others, 1970; Santos and Na1ters, 1971; Sen and others, 1959; Sklyar, 1972; Theoba1d and others, 1967; Theoba1d and Thompson, 1962; Unan, 1971; Vergi1ov, 1969; Vincent and Phi11ips, 1954; Yamaoka, 1962. Geo1ogic setting—— Temperature of crysta11ization- Abdu11ah and Atherton, 1964; Oshima, 1971; Shi1in, 1970. Avai1abi1ity of e1ements -------- Borisenko, 1968; Green and Carpenter, 1961. Thermodynamic and crysta1 chemica1 factors --------- Dasgupta, 1967, 1970. Postdepositiona1 metamorphism---- Abovyan and Borisenko, 1971; Leb1anc, 1969; Ogorodova, 1970; Sergeyev and Tyu1yupo, 1972; Shteynberg and Chasechukhin, 1970. Sorption of minor meta1s ---------- Co1ombo and others, 1964; Fujigaki and others, 1967; Hegemann and A1brecht, 1955. those of iron and copper, but with conflicting results. The conceptual basis for the interpretation of the analytical data is that the major elements in magnetite are generally accompanied by minor amounts of other elements. The presence and quantity of the minor elements are affected by: (1) the origin of the deposit; (2) the geologic setting; (3) the temperature of crystallization; (4) the availability of the elements to the crystallizing magnetite; (5) thermodynamic and crystal chemical factors of the structure of magnetite; (6) postdepositional metamorphism; and (7) surface sorption phenomena, during hypogene and supergene processes, including events after the magnetite has been eroded from its source rocks and is being trans— ported as detrital grains in streams (table 6). The minor elements are contained in minerals as DISTRIBUTION OF THE ELEMENTS 21 trace elements and as trace minerals (Haberlandt, 1947). In magnetic concentrates, minor elements are also contributed by accessory minerals mechanically trapped among grains of magnetite. As trace elements, the minor elements are held in isomorphous substitu- tion for major elements, or are held by sorption or other chemical means permitted by thermodynamic and crystal chemical considerations. As trace minerals, the minor elements are present in minute inclusions, intergrowths, or overgrowths caught up in the host mineral and not cleared by subsequent geo- logic events affecting the host. Several minor elements are common in magnetite. These minor elements usually occur in subordinate amounts as substitutes for other elements in the mineral. The only maj or cations in magnetite that are available for substitution are Fe+3 and Fe”, as shown by the conventional formula for magnetite, Fe+3 (Fe’z, Fe‘3)04. The term diadochy has been used to describe the ability of different elements to occupy the same lattice position in a mineral. Several cations are known to sub- stitute for ferrous and ferric iron in magnetite, but the diadochy between ferric iron and titanium is most prevalent (Dasgupta, 1967). Diadochy always refers to a given structure; thus, two elements may be diadochic in one mineral but not in another. A minor element may substitute for a major element diadochically if the difference in the size of their radii does not exceed about 15 percent (Goldschmidt, 1954). However, this rule is not always valid, but it gives a rough approxi- mation of the magnitude of the difference tolerated. Substitution is also affected by the ionic charge. Ions of similar radii whose charges differ by one unit may substitute readily for one another, but the substitution is only slight if the difference in the charges is greater than one. Of two ions that compete for a lattice site, the one that forms the stronger bonds with its neighbors is the one with the smaller radius or the higher charge or both. Ionization potential also influences the substitution between elements with the same ionic charge and similar ionic radii (Ahrens, 1953; Goldschmidt, 1954). Some elements occupying identical positions in the structures of minerals, and therefore being geochemi- cally closely related, have similar ionization potentials. Ringwood (1955) proposed using values of electro- negativity in addition to Goldschmidt’s rules to predict substitution of one ion for another. The electronega- tivity of an element is clearly related to its ionization potential (for cations) and to its electron affinity (for anions). A bond formed between two atoms is almost purely ionic if the electronegativities are very different. Substitution of one ion for another may be very limited even where they have similar radii and charges, if there is a marked difference in electronegativities between the two ions. Temperature affects the degree of diadochy; thus, high temperatures of formation usually favor diadochic substitution. Ions or atoms may fit into interstices in the lattice instead of replacing another element diadochically, or the ions or atoms may occupy lattice defects where some atoms are missing and lattice positions are ‘vacant. The sorption of trace elements on the surfaces of minerals also accounts for the presence of minor :elements. Sorption has been intensively studied in clays, for example, by soil scientists for plant nutri- tion, by environmental geologists for the disposal of radioactive wastes, and by geochemists for mineral ex- ploration. The sorption of minor elements on minerals, and the possible application of this phenomenon in ore deposition, was discussed by Sullivan (1907), but his experiments did not include magnetite. Green and Carpenter (1961) identified an activity gradient de- creasing from the surface to the center of magnetite grains studied for distribution of radioactivity. They interpreted the gradient as an effect produced either by exsolution during crystallization or by later surface sorption from external sources of uranium and thorium. Later work by Fujigaki and others (1967) showed how magnetite sand could remove the metal ions Mn”, Cu”, Pb”, and Zn+2 from industrial waters. This observation opened the possibility that detrital grains of magnetite downstream from base-metal deposits might adsorb ions of the base metals from stream water. Half of the ferric ions in magnetite are coordinated to four oxygen ions (Bragg and others, 1965), whereas the remaining ferric ions and all the ferrous ions are surrounded by six nearest-neighbor anions. Each oxy- gen ion is bounded by one tetrahedrally coordinated cation and by three octahedrally coordinated cations: The physical characteristics of the 11 minor elements determined in this investigation of magnetite are listed (table 7) to give an idea of the capability for substitu- tion of these elements for ferrous and ferric iron in magnetite. Taking the major cations Fe‘2 and Fe+8 in six-fold coordination, and accepting Goldschmidt's 15-percent tolerance in ionic radius for easy substitution, cations of radius 0.63—0.85 A might replace Fe", and those of ,radius 0.54—0.74 A might replace Fe”. Among the ’minor elements determined here, the divalent ions )Cu”, Zn”, Co”, Ni‘2 and the trivalent ions Co+3 and N i+3 ‘all have radii within the limits set above. Commenting on the substitution of these ions in the magnetite lat- 22 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA TABLE 7.— Physical properties of 11 trace elements in magnetic concentrates from Alaska, compared to the same properties of ferrous and ferric iron in magnetite [—means no data available] Ionic radius Observed Ionic Ionization Ion for 6- coordination Electro- bond with potentia1 coordination number negativity 2 (eV) (percent) Fe:§ 0.74 6 1.8 69 16.16 Fe 0.64 6 1.9 54 30.8 Ag:3 1.26 8, 10 1.9 71 7.57 Bi+5 0.96 6, 8 1.9 66 25.6 Bi 0.74 -- -- -— 55.7 6d:§ 0.97 6, 8 1.7 66 16.84 Co+3 0.72 6 1.8 65 17.3 Co 0.63 —- -- —- 33.6 60:2 0.96 6, 8 1.9 71 7.72 Cu+2 0.72 6 2.0 57 20.34 Ni+3 0.78 6 1.8 60 18.2 Ni+2 0.62 —— —- -- 35.2 Pb+4 1.20 6-10 1.8 -- 14.96 Pb 0.84 -— -- 72 42.4 Zn:2 0.74 4, 6 1.7 63 17.89 Au+3 1.37 8—12 2.4 62 9.22 In+3 0.81 6 1.7 62 28.1 T1 0.95 6 8 1.8 58 29.9 ‘Compiled from: Mason (1966), Krauskopf (1967), Summers (1970), and Whittaker and Muntus (1970). tice, Frietsch (1970) noted that Cu‘2 is similar in ionic radius to Fe‘2 but has a higher electronegativity, thus it does not readily substitute although it tends to in- crease in magnetites from late felsic magmatic rocks and in magnetites from hydrothermal, contact pneumatolytic, and skarn deposits. Zn” has the same ionic radius as Fe” and slightly lower electronega— tivity, thus it tends to substitute for ferrous iron and is camouflaged in minerals like magnetite. Zinc is par- ticularly common in magnetites from late magmatic differentiates, where the ratio Zn”/Fe” increases (Frietsch, 1970). According to Frietsch, zinc is most abundant in magnetites from the contact pneumato- lytic deposits of rare mafic late magmatic rocks. The Co” content was thought by Frietsch to be high in magnetites from early mafic magmatic rocks, as is Ni”, and the Ni/Co ratio in magnetite was thought to fall during magmatic differentiation. For some unusually cobalt-rich magnetites, the source of the cobalt was inferred to be trace minerals, probably pyrite and pyrrhotite. Thus, the presence of copper, cobalt, nickel, and zinc in diadochic substitution in magnetite is identified by theory and confirmed by the literature. However. these four elements may also be present in minor minerals or in trapped accessory minerals, both of which were discussed above. The im- portance of these two mineral forms, compared to diadochic substitution, will be reviewed below. The ionic radii of lead, cadmium, silver, and gold are too large to permit them to substitute for either Fe” or Fe”. Lead tends to occur in silicate structures as the Pb‘2 ion (Rankama and Sahama, 1950); the smaller Pb"4 ion is rarely found in mineral systems (Taylor, 1965). The major ions which are possibly replaced by Pb” in minerals are Ca” (0.99 A) and K‘ (1.33 A). Calcium has commonly been reported in magnetite (Vincent and Phillips, 1954), and as much as 0.94 percent CaO has been found in magnetite (Deer and others, 1962). Thus, Pb” might replace Ca*2 in magnetite. The presence of lead in magnetite might also be explained by lattice defects or minor mineral inclusions, principally sulfides. In spite of the similarity of the ionic radii of Cd‘2 (0.97 A) and Ca” (0.99 A), cadmium appears to follow iron instead of calcium (Vincent and Bilefield, 1960). The second ionization potential of Cd” (16.84 eV) is much higher than that of Ca” (11.90 eV) but is similar to that of Fe” (16.16 eV), which may explain why cadmium follows iron. Cadmium also has a notable tendency to be concealed in zinc minerals despite the difference in ionic radii of cadmium and zinc (table 7). Cadmium may replace zinc diadochically at elevated temperatures; however, cadmium is readily separated from zinc minerals during weathering at ambient temperatures. Doubtless much of the small amount of cadmium found in the Alaskan magnetic concentrates is in minor or accessory minerals. Silver belongs to the same subgroup in the periodic table as copper; therefore, the geochemical distribution of silver tends to resemble that of copper. However, the ionic radius of Ag” is too great for silver to sub- stitute directly for iron in magnetite. Ag” (1.26 A) has a suitable radius to replace K‘ (1.33 A) and, with in- creased tolerance, perhaps under conditions of high temperature, Ag+ might replace Ca”. Frietsch (1970) suggested that traces of silver he found in magnetite possibly resulted from diadochic substitution, but he recognized the role of silver-bearing minor minerals in the magnetite as an alternate explanation. Gold exhibits a different geochemical behavior from silver (Rankama and Sahama, 1950). One marked dif- ference is that the first ionization potential of gold is much greater than that of silver (table 7); therefore, gold ionizes with difficulty. Gold might be incorpo- rated into minerals in the form of uncharged atoms (Vincent and Crocket, 1960). However, it is not likely that the gold atom could be accommodated in unoccu- pied structural sites (interstitial solid solution) because its radius is large (1.44 A). Nor could the Au‘ ion substitute for the ferrous or ferric ions in magnetite. Probably much of the gold in magnetite is in the form of a minor mineral (native gold) or serves as an element DISTRIBUTION OF THE ELEMENTS 23 in other minor minerals such as sulfides. The presence of native gold embedded in grains of magnetite from the Innoko district, Alaska, has long been known (Eakin, 1914, p. 28). The trivalent ions Bi+3 (0.96 A), In+3 (0.81 A), and T1” (0.95 A) might replace Ca*2 (0.99 A), but not readily (N ockolds, 1966). Because it has a smaller radius than the other two trivalent elements, In"3 might substitute for Fe”, Zn”, or Mn*2 (0.80 A), where the latter two are also present in magnetite. Bismuth and thallium replace lead in lead minerals; thus, they may be incor- porated as minor minerals. _ Copper and zinc can substitute diadochically for bivalent iron in magnetite. Lead is usually associated with copper and zinc in sulfide deposits, and cadmium has a geochemical affiliation for zinc. Because of these relations this group of four elements is discussed together in the interpretations of the analyses from the magnetic concentrates. The precious metals silver and gold are discussed together, and bismuth is set aside for individual treatment. The ferrides cobalt and nickel are usually associated in nature with iron; thus, they are described as a group. The dispersed elements in- dium and thallium, which rarely form their own minerals but occur in host minerals such as sphalerite (indium) and galena (thallium), compose the last group for discussion. REGIONAL RESULTS AREAS DISCUSSED The regional results of the determinations of equiv- alent uranium and the atomic absorption analyses for 11 elements in 347 magnetic concentrates from Alaska are given in table 1. These regional results are dis- cussed by five geographic subareas, which are listed below, along with the 1:250,000-scale quadrangles that represent them: 1. Southeastern Alaska; Bradfield Canal, Juneau, Ketchikan. 2. Southern Alaska; Anchorage, McCarthy, Mount Hayes, Mount McKinley, Nabesna, Talkeetna, Talkeetna Mountains, Valdez. 3. Southwestern Alaska: Bethel, Goodnews, Hage- meister Island, Iliamna, Lake Clark, Russian Mission. . West-central Alaska: Bendeleben, Candle, Idita- rod, McGrath, Medfra, Nome, Norton Bay, Ruby, Solomon, Teller. 5. East-central Alaska: Circle, Eagle, Fairbanks, Livengood, Tanacross, Tanana. J; STATISTICAL TREATMENT Statistical summaries of the minor elements in the magnetic concentrates are given in table 8. The table lists the regional data and data from the Candle and Solomon quadrangles. The range of abundance for each element in the magnetic concentrates is shown by its minimum, maximum, and geometric mean. Also listed are the geometric deviations and the percentage of samples having less than the limit of detection of an element. The estimates of mean and geometric devia- tion are based on censored data—that is, data which are qualified on table 1 with L (less than the limit of detection), N (not detected), and (not determined). The samples with qualified values were not used in computations of mean and standard deviation; there- fore, the summaries in table 8 are not suitable for general estimates of abundance. The frequency distributions of the abundances of equivalent uranium, silver, bismuth, cadmium, cobalt, copper, nickel, lead, zinc, and gold were studied in the form of histograms and cumulative frequency curves. Because most of the magnetic concentrates were found to have indium and thallium below the limits of de- tection, cumulative frequency curves for these two elements could not be satisfactorily constructed. The concentration and cumulative frequency were plotted on log probability paper with the frequencies cumulated from the highest to the lowest concentra- tions (figs. 1—10) in order to use Lepeltier’s method to identify threshold and anomalous values for the metals (Lepeltier, 1969). In this procedure, as it is generally applied to stream-sediment samples, the background is given by the intersection of the straight-line cumulative frequency distribution curve on the log- probability plot with the 50-percent ordinate, and the threshold is given by the intersection of the same line with the 2.5-percent ordinate. These considerations are not applicable to the Alas- kan magnetic concentrates, because the concentrates are strongly biased toward anomalous values as they were collected mostly from mineralized areas. There- fore, the percentage of anomalous magnetic concen- trates is much greater than what would be expected from the usual regional program of geochemical explo- ration based on stream sediments. A normal distribu- tion is completely determined by the arithmetic mean and the standard deviation. Because a normal curve is both unimodal and symmetrical, the arithmetic mean, the mode, and the median coincide. The median—the value that divides the area under the curve in half—is usually taken as the background value. For a perfectly 24 normal distribution, 4.54 percent of the distribution will fall outside the limits indicated by a distance equal to two standard deviations measured on the X-axis on both sides of the arithmetic mean. For a moderawa skewed distribution, the percentage is often taken as approximately 5 percent, only half of which will fall on each side of the curve tails. Therefore, the 2.5-percent ordinate is drawn to give the threshold (Hawkes and Webb, 1962). Unfortunately, the cumulative frequency curves for most of these elements show polymodal lognormal distribution (figs. 1—10). Therefore, no at- tempt is made to draw the background and threshold‘ based exactly on the above definitions. For some elements the threshold in the magnetic concentrates is given by the intersection of the cumulative curve with the 16-percent ordinate, a limit that seems to insure detection of strongly mineralized areas but may permit some weakly mineralized areas to be overlooked. Another reason that these statistical procedures are not satisfactory for these data is that only a few samples are represented for quadrangles other than the Candle and Solomon quadrangles. The values selected as background and threshold for the region and for the Candle and Solomon quad- rangles are listed in table 9. The arguments for these values are set out in appropriate sections that follow. EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 2 I TOTAL OF SAMPLES HAVING DETECTABLE EQUIVALENT U a’ o T lJ_LlL 500 1000 I I [II J L 100 eU, IN PARTS PER MILLION 98 I I 10 50 FIGURE 1.—Concentration and cumulative frequency of equivalent uranium in Alaskan magnetic concentrates. TABLE 8.-—Statistical summaries of the regional geochemical data for 347 magnetic concentrates and of the data for the Candle and Solomon Quadrangles, Alaska [n.d. indicates no data available] Statistic eU Ag Bi Cd Co Cu Ni Pb Zn Au In T1 REGIONAL Minimum (ppm) 40 0.2 5 0.2 5 5 10 5 10 0.2 0.2 0.2 Maximum (ppm) 560 600 90 5.5 1,000 25,000 2,200 4,700 2,800 640 0.5 1 Geometric mean (Ppm) 108 0.44 10 0.39 44 16 50 26 85 2.13 0.23 0.27 Geometric deviation 1.9 3.32 1.6 1.59 1.7 3.3 3.0 2.1 2.3 7.6 1.39 1.54 Percent of samples below detection 64 21 9 30 0 9 0 O 0 1('39 185 183 CANDLE QUADRANGLE Minimum (ppm) 40 0.2 5 0.2 20 5 20 5 30 0.2 <0.2 0.2 Maximum (ppm) 160 1.5 20 0.8 150 90 570 120 630 1.5 0.2 0.3 Geometric mean (ppm) 65 0.34 10 0.38 48 15 66 27 79 0.5 <0.2 0.23 Geometric deviation 1.5 1.7 1.4 1.45 1.5 2.4 1.8 1.8 1.9 2.5 n.d 1.6 Percent of samples below detection 80 24 2 32 0 5 0 0 0 284 296 288 SOLOMON QUADRANGLE Minimum (ppm) 40 0.2 ‘5 0.2 10 5 10 5 10 0.2 0.2 0.2 Maximum (ppm) 560 6.5 40 5.5 95 30 280 1,100 500 1.4 0.3 1 Geometric mean (ppm) 124 0.32 11 0.44 31 7 20 26 58 0.3 0.22 0.28 Geometric deviation 1,9 1.8 1.4 1.6 1.4 1.5 2.0 1.9 1.5 3.1 1.2 1.6 Percent of samples below detection 21 23 1 30 0 20 0 0 0 393 388 354 1Percentages computed 2Percentages computed 3Percentages computed for 131 ana1yzed samples. for 25 analyzed samples. for 41 analyzed samples. DISTRIBUTION OF THE ELEMENTS 25 15— (A) T b O | COPPER (Cu), IN PERCENT a) 01 o o I I TOTAL OF SAMPLES HAVING DETECTABLE \l ‘1’ 90— 95— 98 / IDIIIIIII ll[lfl|l[ l L IIIIII l I l 50 1 00 Cu, IN PARTS PER MILLION FIGURE 2.—Concentration and cumulative frequency of copper in Alaskan magnetic concentrates. EQUIVALENT URANIUM ABUNDANCE Partly as a result of the number of samples analyzed, most of the radioactive magnetic concentrates are from four quadrangles: Bendeleben, Candle, Norton Bay, and Solomon (table 1). Other areas where radio- active magnetic concentrates were found are the Circle, Ketchikan, Medfra, and Mount Hayes quad- rangles, where the areas yielding radioactive magnetic concentrates are few and the levels of radioactivity are low, generally near the lower limit of detection of 30 ppm eU. The cumulative frequency curves marked “regional” on figures 1—10 encompass data from all 347 magnetic concentrates. For eU (fig. 1), this curve appears to be bimodally lognormally distributed. The low-value frac- tion of this distribution is contributed mainly from the Candle and Solomon quadrangles. This straight line in- tersects the X-axis outside the figure. Most of the magnetic concentrates have values for eU below the limit of detection. The background value for these samples (not the radiation background) is established at the point of intersection between the extended line and the 50-percent ordinate, and is less than the limit of detection (table 9). Above 120 ppm eU, the line abruptly turns toward low values (the slope of the line increases). Thus, for the regional distribution, the histogram of eU is negatively skewed (fig. 11). The high-value fraction is contributed mostly by samples from the Bendeleben and Solomon quadrangles. Inter- estingly, both the high values and the low values of the 26 N IIIIl I ll SOLOMON OUADRANGLE o // I 7 , \CANDLE H // / QUADRANGLE / _I 01 I \Immhw to..- 00000 0010 I I I I I III 8 I 85— 90— TOTAL OF SAMPLES HAVING DETECTABLE LEAD (Pb), IN PERCENT 95*“ | I | | I II I l L 50 100 Pb, IN PARTS PER MILLION 98III FIGURE 3.—Concentration and cumulative frequency of lead in Alaskan magnetic concentrates. I I I I I I I I/ I III / / 5 / ,0 ~ SOLOMON D , w OUADRAN Ed 10_ GLE\// _ E U 15— / _ .“J 20— U REGIONAL — a » / , c, E 30 — // /O — 2 <5: Ew— /O\CANDLE — I z 50 _ QUADRANGLE _ 8 :60 a g — 2 E _ < E 70 / 0') N 0 g 80— o/ — El 85— / — / I5 90— // __ I— l/ 95— II/q/ E // 98 I/ 42/ I I I I I I I I I I I I I I I 10 50 100 500 1000 Zn, IN PARTS PER MILLION FIGURE 4.—Concentration and cumulative frequency of zinc in Alaskan magnetic concentrates. regional bimodal distribution of eU are represented in the Solomon quadrangle, whereas the samples from the Bendeleben and Norton Bay quadrangles are typi- cally highly radioactive and those from the Candle EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 2 I I I | I /I//I / I I I I I I 5_ CANDLE / / _ QUADRANGLE / / 10— \/ / — 15— — b h) N o 0 cl) I I \I o 8 I / A SOLOMON OUADRANGLE _ CADMIUM (Cd), IN PERCENT co co a: 01 0 cl: TOTAL OF SAMPLES HAVING DETECTABLE to o I I 95— ~ 98 I LIIIIIII IIIII 0.1 0.5 1 5 10 Cd, IN PARTS PER MILLION FIGURE 5.—Concentration and cumulative frequency of cadmium in Alaskan magnetic concentrates. quadrangle tend to have the lower values. The value for the radioactivity of the breaking point on the regional cumulative curve, 120 ppm eU, is taken as the threshold between background and anomalous values. About 16 percent of the magnetic concentrates with measurable radioactivity are above this threshold. The radioactive magnetic concentrates from the Bendeleben, Candle, Norton Bay, and Solomon quad- rangles come from a single geologic area divided by the arbitrary boundaries of the quadrangles. All the radioactive magnetic concentrates from this area were obtained from streams that drain the Darby Moun- tains, which are underlain in part by granite, quartz monzonite, and potassium-rich alkaline intrusive rocks (Miller, 1972; Miller and others, 1971; Miller and others, 1972; Elliott and Miller, 1969; Cass, 1959). The most radioactive magnetic concentrates are derived from the alkaline rocks with ultrapotassic character in the intrusive complex of the Darby Mountains, and come mainly from the Bendeleben, Norton Bay, and Solomon quadrangles. Samples from the Candle quad- rangle tend to be somewhat less radioactive than those from the other three (tables 8—9). The relations between the radioactivity of magnetic concentrates and the source rocks for the concentrates can be most clearly detailed by a study of the 85 samples from the Candle quadrangle and the 101 samples from the Solomon quadrangle. Details of these two areas are discussed separately below. The rela- DISTRIBUTION OF THE ELEMENTS 10 15 20 (A) O 8 O) 0 TOTAL OF SAMPLES HAVING DETECTABLE SILVER (Ag) IN PERCENT \l m o 0 so 85 // 907/ 98 llIIIIJl NM 536‘? ll IIIJII 1 [I 0.1 A9, IN PARTS PER MILLION FIGURE 6,—Concentration and cumulative frequency of silver in Alaskan magnetic concentrates. tions found in the Candle and Solomon quadrangles apply also in the Bendeleben and Norton Bay quad- rangles, where only 20 and 4 samples, respectively, were measured for radioactivity. The equivalent uranium in the magnetic concen- trates from the Circle, Ketchikan, Medfra, and Mount Hayes quadrangles is compared in table 10 with the equivalent uranium cited for the original concentrate in the records of the Alaskan placer concentrate file, as determined in the late 1940’s and early 1950’s by John J. Matzko, US. Geological Survey. Generally similar equivalent uranium was found for the two materials, but a tendency exists for the magnetic concentrate to have slightly greater equivalent uranium than the whole concentrate. The reverse was found to be charac- teristic of concentrates from the Solomon quadrangle, described later in the report. The magnetic concentrate from the Circle quad- rangle was separated from a sluice-box concentrate from the H. C. Carstens mine on Portage Creek. Allanite, garnet, scheelite, sphene, topaz, uranothoria— nite, and zircon occur in this placer, and apatite, fluorite, garnet, limonite, scheelite, and zircon are in the granitic source rock for the placer (Nelson and others, 1954, table 9). Uranium is present in allanite, sphene, uranothorianite, zircon, apatite, and limonite at this locality. The two concentrates from the Ketchikan quad- rangle, listed in table 10, are from streams draining mainly the Eocene Hyder Quartz Monzonite and Up- per Triassic or Lower Jurassic Texas Creek Granodio- rite (West and Benson, 1955, pl. 7). No independent uranium-bearing mineral was recognized by West and ’ Benson (1955, table 1) in concentrate 3339, but in 3343 N on EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 2IIIIIIII l _| 0 GOLD (Au), IN PERCENT N 7" o 01 TOTAL OF SAMPLES HAVING DETECTABLE w o I Tlil‘ljl I | I III 40—~ _ 50 I I I I I I I I I I I I l I I I I I I I l L 0.5 1 5 10 50 100 Au, IN PARTS PER MILLION FIGURE 7.——Concentration and cumulative frequency of gold in Alaskan magnetic concentrates. 2 I I I I I I I III I I I 2 I I I I I I I I /I I I I I I I /D CANDLE I 5 7/ / 5‘ ADRA — _ — m cm NGLE SOLOMON / REGIONAL -’ 10— _ OUADRANGLE\7 gt: LIJ 10 — — '5 ‘5‘ — ‘5' 15 / ”" I— — < — — I- 20 — I— u.I 2 U 20— — D '-'-' u.I u 30, _ I— I— o I: Lu 2 I z I” D “J 30— — _ n. U l > 40— — 0 n: < Z z “-I I :I: A. 50_ _ S n- 40— —- a, 25 SOLOMON < z / ‘iI " 60— QUADRANGLE — I ‘50— I — a E a 8 <§z 3 70* — 1' v60» I I _ (n 2 2 I— / :5 u_ (L) < a] o /\ 0 m 80— REGIONAL — w 3 7°- / ’ CANDLE T _I 85 — u. U I / QUADRANGLE 4: O 80 _ / _ I5 90— — _, I— E 85 — — 95 w _ I- 90 __ — 93 I I I I I I I I 1 5 10 100 95— — Bi, IN PARTS PER MILLION FIGURE 8.—Concentration and cumulative frequency of bismuth in 98 ' I ' ' ' ' I ‘ ' 1 Alaskan magnetic concentrates. 10 50 100 500 C0, IN PARTS PER MILLION uranium was identified in sphene. In the same general F 9 C . . . area as the sources of these two concentrates, hematite IGURE .—- oncentration and cumulative frequency of cobalt in . . . . . . Alaskan magnetic concentrates. and hmomte, elther as 1ndependent minerals 1n concen- trates or as coatings on other minerals, were shown to Appel Mountain, but it is not described in the litera- contain uranium. ture. Concentrates from the Nixon Fork mining The concentrate from the Medfra quadrangle listed district about 30—40 km to the northeast of Appel in table 10 comes from a gulch on the north side of Mountain were discussed by White and Stevens (1953). DISTRIBUTION OF THE ELEMENTS 30— 60— 70-— TOTAL OF SAMPLES HAVING DETECTABLE NICKEL (Ni), IN PERCENT S l 98 I 1 I I FIGURE 10.—Concentration and cumulative frequency of nickel in Alaskan magnetic concentrates. Ni, IN PARTS PER MILLION 50 100 500 1000 TABLE 9,—Background and threshold values for equivalent uranium and nine elements in magnetic concentrates from Alaska [Data are in parts per million] Area LeveI eU Ag Bi Cd Co Cu Ni Pb Zn Au RegionaI Background <30 0.2 8 0.24 45 IO 45 25 75 0.2 ThreshoId I20 I I4 I 95 25 240 60 120 I Candle quadrangIe Background <30 0.27 9 0.25 55 IO 65 25 80 n.d. ThreshoId I00 I 15 I 90 22 I70 50 I40 I SoIomon quadrangIe Background 90 0.27 IO 0.30 30 6 IS 25 60 n.d. ThreshoId 220 I IS I 40 15 20 50 80 I They stated that most of the radioactivity in concen- trates from granite and from contacts of granite was from thorium-bearing minerals or from uranium in thorium minerals, but locally hematite was found to be uraniferous. The radioactive magnetic concentrate from the 30 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA T l l | 10 _ Candle quadrangle — Not detected=68 0 20 ~ * Solomon quadrangle Not detected=21 a) z 9 '2 Z 10 — — E n: Ll.l l— Lu 0 LI. 0 0 n: UJ no 5 Regional area 2 _ Not detected=225 20 ~ — 10 — __ 0 l 10 50 100 500 1000 ppm eU FIGURE 11.—Histograms for equivalent uranium in Alaskan magnetic concentrates; shaded areas anomalous. Mount Hayes quadrangle was separated from a con- centrate panned from 100 pounds of gravel from Dry Creek about 80 m upstream from the bridge on the Alaska Highway (Wedow and others, 1954, table 1, fig. 8). Granitic intrusive rocks were the source for gravel in Dry Creek. Tests by Wedow and others (1954, p. 16) showed that most of the radioactive minerals in con- centrates similar to 1472 were in the nonmagnetic frac- tion; thus, removal of the magnetic fraction causes a relative enrichment in the equivalent uranium of the nonmagnetic fraction owing to the presence of zircon, the main radioactive mineral from the granite. The role of radioactive limonite or hematite in these concen- trates was not assessed. The original literature did not report whether the radioactive magnetic concentrates listed in table 10 were derived from potassium-rich intrusives like those in the Darby Mountains. TABLE 10.—Comparison of equivalent uranium in magnetic concen- trates and original source concentrates in the Circle, Ketchikan, Medfra, and Mount Hayes quadrangles, Alaska [Equivalent uranium of original concentrate determined 1949—53 by John J. Matzko, US. Geological Survey; equivalent uranium of magnetic concentrate determined by K.-L. Pan, 1971] Equivalent uranium (ppm) File Original Magnetic Quadrangle number concentrate concentrate Circle 3646 l20 150 Ketchikan 3339 20 50 Do ------- 3343 lOO 40 Medfra 296 10 40 Mount Hayes 1472 80 40 MINERALOGICAL SOURCES The reports of the investigations of radioactive deposits in Alaska cited above called attention to uranium-bearing hematite and limonite as one source for radioactivity in concentrates. Radioactive uranifer- ous iron oxides are widely reported in the literature (Lovering, 1955; McKelvey and others, 1956; Karkhanavala, 1958; Levering and Beroni, 1959; Green, 1960), and the intimate association of hematite with certain uranium deposits has been used as a guide in geologic prospecting (Nininger, 1956, p. 116). The surface contamination of magnetite by uranium is known (Damon and others, 1960; Green and Carpenter, 1961), and the coating of magnetite with hematite, or the alteration of magnetite to hematite, is a common phenomenon. Many, if not most, of the radioactive magnetic con- centrates are less splendent than the nonradioactive ones, and tend to be dull brown or brownish black in- stead of bright black. A test was made to determine if the dull brown color was attributable to a coating, and if the coating was more radioactive than the grains on which it was deposited. For this test, sample 299 from Jump Creek (65°51’15" N.; 162°01'15" W.) in the Bendeleben quadrangle was chosen, although it is not one of the magnetic concentrates otherwise analyzed here. Jump Creek is north of the Darby Mountains in the northwestern headwaters of Candle Creek where silicified intrusive breccia of Cretaceous age is reported (C. L. Sainsbury, oral commun., 1972). The magnetic concentrate was measured for radioactivity before removal of the coating. Then the coating was removed by ultrasonic cleaning and the cleaned magnetite and its coating were separately measured for radioactivity. Because the coating was thin, only enough could be obtained to make one sample DISTRIBUTION OF THE ELEMENTS 31 for counting, but the cleaned magnetic residue was large enough to be divided into eight subsamples for counting. The counting followed the procedure used for the other magnetic concentrates, with the same stan- dard for comparison. The results of these radiometric analyses are given in table 1 1, where it can be seen that about two-thirds of the radioactivity of the magnetic concentrate is attributable to the brown coating. TABLE 11.——Equivalent uranium in magnetite and in hematitic coatings on the magnetite, Jump Creek placer, Bendeleben quadrangle, Alaska [Measured by Wayne Mountjoy, U.S. Geological Survey. January 26, 1972] Material analyzed Equivalent uranium (ppm) Magnetic concentrate before cleaning --------- 98 Magnetic concentrate after cleaning Subsample l .............................. 55 Subsample 2 ....................... 57 Subsample 3---- ________ 35 Subsample 4- 39 Subsample 5- ..... <30 Subsample 6-- ------ 41 Subsample 7---- .................... <30 Subsample 8 .............................. 30 Nonmagnetic coating removed by ultrasonic cleaning ................................... 315 X-ray diffraction studies, by Keith Robinson, of the brown coating from sample 299 showed that it is main- ly hematite. A small percentage of analcime is mixed with the hematite. The grains on which the coating was deposited are magnetite. It seems probable that much of the radioactivity of the other magnetic concentrates is in hematitic coatings on grains of magnetite. Also, the source of the radioactivity is probably mainly uranium instead of thorium. POSSIBLE USE Magnetic concentrates from streams draining alkalic rocks in the Seward Peninsula, Alaska, are radioactive, whereas most magnetic concentrates from other geologic provenances in Alaska are not radioactive. This distinctive association may afford a method for recognizing the presence of otherwise hidden alkch rocks or uranium deposits in areas with a heavy cover of vegetation or with deeply weathered rocks. The relative ages of different intrusive rocks in a sequence might be distinguished by different degrees of radio- activity of the magnetic concentrates. Much work to test these concepts is needed. One test made in connection with the present investigation failed to show radioactivity in magnetites separated from rocks associated with two alkalic complexes in Brazil. These samples, contributed by D. B. Hoover, US. Geological Survey, are a specimen of olivine gab- bro from J ose Fernando about 20 km south of Ribeira, $50 Paulo, and a specimen of carbonatite from Jacupiranga, Parana State. These specimens were crushed and sieved, and the magnetite was removed with a hand magnet following procedures used for making magnetic concentrates. A clean sample of magnetite (AP—5) was obtained from the olivine gabbro, and two samples of magnetite were recovered from the carbonatite, of which one was clean magnetite .(J P—1) and one is magnetite with intergrown carbonate minerals (JP—1a). None of these magnetites is coated with hematite. They were analyzed for radioactivity using twice the counting time previously employed. The three magnetites have no measurable radioactiv- ity at a lower limit of detection of < 10 ppm eU (Wayne Mountjoy, written commun., March 27, 1972). COPPER, LEAD, ZINC, AND CADMIUM The geometric means of copper, lead, zinc, and cad- mium in the magnetic concentrates from Alaska (table 8) show that zinc (85 ppm) is the most abundant of these elements, followed by lead (26 ppm), copper (16 ppm), and cadmium (0.4 ppm). Copper appears to have three populations in the regional samples (fig. 2). The abscissa of the inflection points on the cumulative frequency curve indicates the limit above which there is a departure from lognormal distribution. The low-value branch corresponds to the background, the central branch to the weakly mineralized population, and the high-value branch to the highly mineralized population. However, the cen- tral branch seems to be a mixture of the other two populations instead of an independent population; therefore, the threshold for anomalous copper was taken as the abscissa of the middle of the central branch at 25 ppm copper. Values for copper greater than 100 ppm, marked by the second inflection point on the curve (fig. 2), are probably indicators of strong mineralization. The histograms in figure 12 show a positive skewness for the regional distribution of copper, but many samples have low values. The gap be- tween the two lowest value bars is caused by the 5-ppm reporting interval used for copper. The inflection points on the regional cumulative fre- quency curves for lead (fig. 3) and zinc (fig. 4) are taken as threshold levels of concentration for these two elements in the Alaskan magnetic concentrates. The second inflection at 60 ppm is used for ‘the lead thresh- old, and the inflection at 120 ppm is used for zinc. Both lead and zinc have positively skewed distributions in the histograms (figs. 13—14), but lead has three un- 32 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA | I I I I I I I I 20" Candle quadrangle ’ Not detected=4 0 Solomon quadrangle 40 _ Not detected=20 # w _ _ z 20 Q [— < Z 5 E ° " g _ u. Regional area 2 Not detected=32 m 2 D _ z 60 — fl 40 — _ 20 — — 01 5 5000 10,000 50,000 ppm Cu FIGURE 12.—Histograms for copper in Alaskan magnetic concen- trates; shaded areas anomalous. distinguished populations whereas zinc displays two strongly marked populations. The positive skewness appears mainly to be caused by the fact that some samples were taken from placer mines where the magnetic concentrates contain abundant minor elements. This skewness may be an indication of base- metal mineralization in the source areas of gold placers, and where an excess of high values is found (strong positive skewness), the indication for future prospecting should be greatest. For the region, the abrupt break in the cumulative frequency curve for zinc (fig. 4) at 120 ppm is interpreted to mean that anomalous values above 120 ppm zinc may be an in- dication of mineralization. | I l I I I I I l 20 _ Candle quadrangle _ Not detected=0 0 l_‘l Solomon quadrangle Not detected=0 g 20 — _ Q _ '2 Z 2 l_l If — E 0 E Q 80 — _ — u. T Regional area 0 Not detected=0 n: LLI _ g 60 — 3 F 2 40 — — 20 — — 0 H l 9522' . . 1 5 10 50 100 500 1000 5000 10,000 50,000 ppm Pb FIGURE 13.—Histograms for lead in Alaskan magnetic concen- trates; shaded areas anomalous. In about 30 percent of the magnetic concentrates the cadmium content is less than the limit of detection for the analytical method used. About 40 percent of the values for cadmium (table 1) cluster in the range be- tween 0.4 and 0.6 ppm (figs. 5 and 15). Only five samples have cadmium contents greater than the threshold level of 1 ppm given by the intersection of the regional cumulative frequency curve with the 2.5-percent ordinate. The maximum cadmium content is 5.5 ppm. SOUTHEASTERN ALASKA The Bradfield Canal and Ketchikan quadrangles in southeastern Alaska have strong positive anomalies for base metals in magnetic concentrates, but only a weak anomaly for copper was noted in the Juneau quadrangle (table 1). In the Bradfield Canal quad- rangle, sample 3338, from the Salmon River at a point 4.8 km south of Mineral Hill, has anomalous copper DISTRIBUTION OF THE ELEMENTS 33 l l l l l I | | | 20 — Candle quadrangle _ Not detected=0 0 m 40 ’— . 7 Solomon quadrangle Not detected=0 ‘£ 20 — _ Q E i E 2 .— m —_ 0 LL Regional area 0 80 — Not detected=0 a E m E D z .—. 60 — _ 40 — _ 20 —— _ 0 I ., 1 5 10 50 100 500 1000 5000 I0,000 50,000 ppm Zn FIGURE 14.—Histograms for zinc in Alaskan magnetic concen- trates; shaded areas anomalous. and lead content and highly anomalous zinc content, but the cadmium content is low. A few kilometers farther downstream, six magnetic concentrates from Fish Creek, a tributary to the Salmon River in the Ketchikan quadrangle, contain highly anomalous amounts of copper and generally large amounts of lead and zinc (3339—3377 in table 1). Only one sample (3377) has anomalous cadmium content; it is also anomalously rich in zinc. Base-metal and precious- metal deposits have long been known in these areas (Buddington, 1929; Sainsbury, 1957). SOUTHERN ALASKA All the quadrangles sampled in southern Alaska have weak anomalies for one or more of the base metals, but only in the Talkeetna quadrangle is the cadmium content anomalous (table 1). The strongest lllllll llll 140— _ — Regional area Not detected=106 120— — (n S — 100 — — '2 E E 5 80— — .— Lu 0 8 60 — _ I: g Candle quadrangle 2 Not detected= 27 :> 40— — 2 Solomon quadrangle Not detected=30 20— — o H I\ I H I L I I 0.1 0.5 1 0.1 0.5 1 5 10 0.1 0.5 1 5 10 ppm Cd FIGURE 15.—Histograms for cadmium in Alaskan magnetic concen- trates; shaded areas anomalous. anomalies are for zinc. In the Anchorage quadrangle the area around the upper part of Knik Arm yielded magnetic concentrates markedly richer than back- ground in copper and zinc with associated anomalous amounts of cobalt and nickel (samples 2199, 2202, and 2203). Zinc has the strongest anomalies in the An- chorage quadrangle, and copper content is weakly anomalous, but lead and cadmium content is low. The presence of lode deposits for copper, lead, and zinc in this region has long been known (Landes, 1927), and the chromite-bearing rocks near Eklutna, doubtless the source of the anomalous amounts of cobalt and nickel, were studied by Rose (1966). Three samples with anomalous copper content (2187, 2191, and 2192) identify the area of base-metal deposits on the south flank of the Talkeetna Mountains in the basin of the Little Susitna River, and sample 2163, which has weakly anomalous copper content and strongly anomalous zinc content, is from a locality about 3.2 km northeast of the Sheep Mountain copper deposit (Cobb, 1972a) All four of the magnetic concentrates (2134, 2136, 2148, and 2438) from the McCarthy quadrangle have anomalous contents of copper and zinc but only background amounts of lead and cadmium. The Nikolai Butte copper deposit (MacKevett and Smith, 1968; 1972) is identified by the very high copper con- tent (2000 ppm) in sample 2148 plus the weak copper anomaly in sample 2134, and the Kennicott copper deposits (MacKevett, 1971) are shown by a high 34 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA positive anomaly (1700 ppm Cu) in sample 2438. No mineral deposit is reported in the immediate vicinity of sample 2136, but a number of base-metal and gold deposits are located upstream to the east along Young Creek and in the Mount Holmes area (MacKevett and Cobb, 1972). A few magnetic concentrates from the Mount Hayes quadrangle have weakly anomalous copper (1473, 1510, and 1513) and zinc contents (1511), but the con- tent of lead and cadmium is low (table 1). No mineralization is reported for the area represented by 1473, but the three other anomalous samples are from the Rainbow Mountain copper deposits (Cobb, 1972b). The only magnetic concentrate from the Mount McKinley quadrangle with anomalous base metals is 1019, which has threshold amounts of copper and lead associated with background zinc and cadmium (table 1). The sample is from Last Chance Creek in the gold, antimony, and base metal district at Glacier Peak and Spruce Peak (Cobb, 1972c). Three of the four magnetic concentrates (1493, 1504, and 1507) from the Nabesna quadrangle are weakly anomalous for copper and sample 1493 is weakly anomalous for zinc (table 1), but none is anomalous for lead or cadmium. None of the three comes from a recognized mineralized area (Richter and Matson, 1972). Copper content is not anomalous in any of the five magnetic concentrates from the Talkeetna quadrangle (table 1), but four samples (254, 482, 1304, and 1336) contain zinc in greatly anomalous amounts, and the lead content in sample 1336 is weakly anomalous, as is the cadmium content in 482. The cadmium-rich sample also contains the most zinc. Samples 482, 1304, and 1336 are from the gold placer district between Peters Hills and Dutch Hills and downstream from the lode gold deposits on Dutch Hills (Clark and Cobb, 197 2). Zinc deposits are unreported. Sample 254, which also contains anomalous amounts of gold, is from a locality downstream from the Big Boulder Creek, Chicago Gulch, and Twin Creek gold placers at Fairview Mountain. The three magnetic concentrates (2179, 2181, and 2182) from the Talkeetna Mountains quadrangle also have highly anomalous zinc content and two (2179 and 2181) have weakly anomalous copper content, but none has anomalous concentrations of cadmium (table 1). One sample (2179) is weakly anomalous for lead. They come from an area that includes gold placers along Crooked Creek on the southeast side of the Horn Mountains, but base-metal deposits are unreported (Cobb, 1972d). The extremely high zinc content (2800 ppm) of sample 2182 is notable. This concentrate was collected farther downstream along Crooked Creek than the other samples, thus the greatest enrichment in zinc might be even farther downstream where magnetic concentrates have not been collected. The three magnetic concentrates (2131, 2156, and 2162) from the Valdez quadrangle contain anomalous amounts of copper and zinc but have only background amounts of lead and cadmium (table 1). The source localities of these samples are widely separated. Only 2131 is near a known mineral deposit: it comes from a point about 2.7 km south-southeast of the Willow Mountain copper-zinc lode deposit (Berg and Cobb, 1967, p 52). SOUTHWESTERN ALASKA The magnetic concentrates from the Goodnews quadrangle lack anomalous copper, lead, zinc, and cad- mium content (table 1). One of the two from the Hagemeister Island quadrangle is weakly anomalous for copper, and those from the Russian Mission quadrangle are weakly anomalous for copper and lead. Magnetic concentrates strongly anomalous for copper and zinc were obtained from the Bethel, Iliamna, and Lake Clark quadrangles. In the Bethel quadrangle, magnetic concentrates (928, 929, 2120, and 2121) with strongly anomalous copper and zinc content were obtained from Marvel Creek and Cripple Creek, both tributary to the Salmon River and both exploited for placer gold (Cobb, 1972e). Samples 240 and 918, anomalous for zinc, come from localities on the Kuskokwim River and Canyon Creek where zinc deposits have not been reported. The single magnetic concentrate (553) from the Hagemeister Island quadrangle with weakly anomalous copper content comes from the Platinum Creek placer, where platinum, gold, and chromite are reported (Cobb, 1972f). The copper may be accounted for by the great amounts of basic and ultrabasic ig- neous rocks near the sources of the detrital magnetite. All five of the magnetic concentrates from the II- iamna quadrangle have anomalous copper content, and four have anomalous zinc content (table 1). The five samples were collected from the southern, eastern, and northern sides of Iliamna Lake, an area that has many lode deposits of copper (Detterman and Cobb, 1972). However, only one sample is close to a known deposit: sample 3779 is from Millets copper prospect near Chekok Bay on the north shore of the lake. This sample has 25,000 ppm copper, the largest value reported for this set of magnetic concentrates. Copper content is anomalous in all the magnetic con- centrates from the Lake Clark quadrangle, and four (3783, 3791, 3797, and 3799) contain highly anomalous amounts of zinc (table 1). These four are downstream DISTRIBUTION OF THE ELEMENTS 35 from gold and base-metal deposits on the north and south shores of Lake Clark (Cobb, 1972g). The other samples, containing anomalous amounts of copper but not of zinc, are from the northeastern end of Little Lake Clark, where ore deposits are unreported. Only one (2119) of the four magnetic concentrates from the Russian Mission quadrangle contains anomalous amounts of base metals (table 1). It is from the gold placer on Bear Creek (Hoare and Cobb, 1972). WEST-CENTRAL ALASKA The strongest anomalies found for copper, lead, and zinc in the magnetic concentrates from west-central Alaska are in the Iditarod and Ruby quadrangles. Magnetic concentrates from west-central Alaska are more commonly anomalous for cadmium than are those from any of the other areas (table 1). None of the magnetic concentrates from the Bendeleben quadrangle has anomalous copper content (table 1), and relatively few have anomalous amounts of lead (3056, 3070, and 3075), zinc (3041 and 3078), or cadmium (3074). All these samples are in the Darby Mountains and come from streams that drain granitic plutons (3070, 3074, 3075, and 3078) or sedimentary rocks (3041 and 3056) in the vicinity of known mineral deposits (Cobb, 1972b). Sample 3056, which contains anomalous amounts of lead, comes from the Grouse Creek gold placer. Samples 3070 and 3075 also have anomalous lead content; 307 0 comes from the drainage southwest of the Grouse Creek placer, and 3075 was collected between the Grouse Creek gold placer and the Otter Creek tin placer. This latter location is also the source of the zinc-bearing sample 3078 and the sample with anomalous cadmium content (307 4). Sample 3041 with anomalous zinc content was collected near the Camp Creek gold placer. Anomalous amounts of copper and lead are less common in magnetic concentrates from the Candle quadrangle than is anomalous zinc content, and cad- mium occurs only in background amounts (table 1). Most of the samples that are anomalous for copper are not anomalous for zinc. The three samples (2468, 2473, and 2501) with anomalous lead content also have anomalous amounts of copper but not of zinc. These relations are discussed in the section on the Candle quadrangle. Magnetic concentrates from the Iditarod quadrangle that are anomalous for zinc also tend to be weakly anomalous for copper and lead. Thus, samples 1804 and 1815 from the gold placers in the Willow Creek area have anomalous zinc and copper content, and 1815 also has anomalous lead content. Gold placers in the Flat area are the source of zinc-rich samples 1831, 1838, 1867, 1877, and 1883. These samples also con- tain anomalous amounts of other metals: copper and lead in 1831; copper, lead, and cadmium in 1867; and copper in 1883. Most of the samples from the Flat area were taken from Otter Creek and Flat Creek, where lode deposits containing gold, silver, copper, lead, zinc, tungsten, antimony, and mercury are reported (Cobb, 1972i). Only two magnetic concentrates come from the McGrath quadrangle (table 1). Number 483, which con- tains anomalous amounts of zinc, was collected at the lode deposit of antimony, bismuth, gold, and tungsten on the south side of Vinasale Mountain; and number 1917, from the Candle Creek gold placer on the north- western flank of Roundabout Mountain (Cobb, 197 2j), has anomalous copper, lead, zinc, and cadmium content. In the Medfra quadrangle, magnetic concentrate 902, which contains anomalous amounts of copper (table 1), is from the Hidden Creek mineralized area be- tween Greens Head and Jumbo Peak, where placers and lode deposits of bismuth, gold, tungsten, and copper are known (Cobb, 1972k). No ore deposits are reported in association with sample 296, which con- tains anomalous amounts of zinc. Only one magnetic concentrate of the five analyzed from the Nome quadrangle (table 1) has any anomalous base metal content. Lead is weakly anomalous in sam- ple 279. Gold placers together with lode deposits of copper and gold are reported in the vicinity of the sample (Cobb, 19721). Copper content is weakly anomalous in only one of the four magnetic concentrates analyzed from the N or- ton Bay quadrangle (table 1). The sample (304) is from the Bonanza Creek gold placer 6 km east of Ungalik (Cobb, 1972m). The placer contains anomalous amounts of antimony and tungsten. Copper and lead contents are strongly anomalous in two magnetic concentrates (56 and 59) from the Ruby quadrangle, and sample 59 also has weakly anomalous amounts of zinc and cadmium (table 1). The 4,700 ppm lead in sample 59 is the largest value for lead found in any of the 347 magnetic concentrates. The sample is from the area of the Flint Creek placer gold and tin deposits east of Long (Cobb, 1972n). Sample 56 is from the mouth of Solomon Creek in the area of the Poorman Creek gold and tin placers. The weakly anomalous values noted for copper, lead, zinc, and cadmium in magnetic concentrates from the Solomon quadrangle are widely scattered (table 1). Of the 101 concentrates analyzed, only 10 samples have any anomalous base-metal content at all: six for lead, two for zinc, and one each for copper and cadmium. Of the two analyzed magnetic concentrates from the Teller quadrangle (table 1), only one (497) contains an 36 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA anomalous amount of a base metal: 1,100 ppm zinc. This value is one of the highest for zinc in the set of 347 concentrates. The sample is from Cape Creek at Tin City (Cobb and Sainsbury, 1972). The source area for the detrital magnetite seems to be upstream along Cape Creek, southeast of the Bartels tin mine (Mulligan, 1966, fig. 6). However, detailed petro- graphic descriptions of minerals in tin placer concen- trates from this area do not mention the presence of in- dependent zinc minerals, nor did spectrographic analyses of core from the area show a trace of zinc (Mulligan, 1966, tables 10, 14—18). EAST-CENTRAL ALASKA The concentrations of copper, lead, zinc, and cad- mium are below threshold anomalous values in magnetic concentrates from the Fairbanks quadrangle in east-central Alaska, and cadmium does not reach anomalous abundances in any of the magnetic concen- trates from east-central Alaska (table 1). The largest base-metal anomalies in these concentrates are found in the Livengood quadrangle. The single sample (3646) of magnetic concentrate from the Circle quadrangle contains greatly anomalous amounts of copper and zinc but has little lead or cadmium (table 1). This sample comes from Portage Creek, about 2.4 km upstream from a known occur- rence of zinc (Cobb, 19720), but its copper content is more anomalous than its zinc content. Only four of the ten magnetic concentrates (277, 532, 3689, and 3704) from the Eagle quadrangle contain anomalous amounts of copper, and one (3689) has anomalous lead content. None of the anomalous values is very large (table 1). Cadmium is not present in anomalous amounts. The first three copper-rich samples were collected along the South Fork Fortymile River in or downstream from the Chicken district, an area known to have placer deposits of gold, minor tin, and minor tungsten, and several lode deposits of lead and silver (Cobb, 1972p). Sample 3704 is from My Creek just downstream from some small prospect occurrences of antimony and lead (Cobb, 197 2p). The magnetic concentrates containing anomalous amounts of base metals in the Livengood quadrangle (table 1) come from two areas: the gold placers east and north of Cleary Summit, where tin and tungsten are also reported (Cobb, 1972q); and the placer and lode gold area at Livengood. Copper and zinc are about equally distributed in the Livengood samples, where lead is sparse, and copper is more common than lead or zinc in the material from the Cleary Summit area. Three magnetic concentrates from the Tanacross quadrangle have weakly anomalous copper content (table 1), yet none of these represents an area of recog- nized mineralization (Cobb, 1972r; Foster, 1970). Lead content is below the anomalous threshold in all three samples, but one sample (1499) contains strongly anomalous amounts of zinc. The single magnetic concentrate (2418) from the Tanana quadrangle has anomalous amounts of copper and lead (table 1). It is from Rhode Island Creek down- stream from gold placers in which lead and mercury are reported (Cobb, 197 23). SILVER AND GOLD The distribution of silver in the magnetic concen- trates-from Alaska (fig. 6) is more or less similar to that of copper (fig. 2). However, silver has two distinct populations and a positive skewness. Gold, however, has a linear distribution (fig. 7). The anomalous thresh- old value for silver, 1 ppm (table 9), is taken from the regional cumulative curve in much the same way as was used for copper. Establishing the anomalous threshold value for gold is a problem. The cumulative frequency curve for gold (fig. 7) displays a nearly linear distribution. The slope of the line is such that if the threshold value is taken from the 2.5-percent ordinate, the minimum anomalous value would be as great as 50 ppm gold, which would lead to misinterpretation of the data. Therefore, the threshold value for gold was taken to be the same as silver, 1 ppm. This value might be rather high for gold, because silver is 200 times as abundant as gold in the average igneous rock (Hawkes and Webb, 1962, table 2—7). However, in the present work the lower limit of detection was the same (0.2 ppm) for both gold and silver; thus, similar thresholds are used for anomalous values, but it might reasonably be assumed that any value above the lower limit of detection would be anomalous for gold. The upper values for silver (600 ppm) and gold (640 ppm) in the magnetic concentrates are close (figs. 16 and 17), but the ranges in abundance vary so much that the gold must occur either as native gold particulates (trace minerals) in the magnetite or as accessory minerals in the magnetic concentrates. Many of the magnetic concentrates were collected from sites near gold placer mines; thus, some anomalous silver and gold was found in about half of the quadrangles (table 1). The largest number of magnetic concentrates containing anomalous amounts of silver and gold are from the Bethel, Circle, Eagle, and Livengood quadrangles, where the contents of silver and gold are several times to more than 3,000 times background. SOUTHEASTERN ALASKA Four magnetic concentrates (3372, 3373, 3375, and 3377) from tributaries to Fish Creek in the Ketchikan DISTRIBUTION OF THE ELEMENTS 37 _ l | T | 20 _ Candle quadrangle _ Not detected=20 0 “l __ Solomon quadrangle Not detected=23 20 — _ 0 L n m 2 120 — _ — ,9 Regional area <2: Not detected=75 S n: ".1 g 100 - —— LL 0 a: u: m g _ 60 — . __ 40 — _ 20 —— — 0.1 0.5 1 5 10 50 100 500 1000 ppm Ag FIGURE 16.-Histograms for silver in Alaskan magnetic concen- trates; shaded areas anomalous. quadrangle, southeastern Alaska, contain anomalous amounts of silver. Insufficient material was available for analysis of gold. Therefore, it is not known if these samples also have anomalous gold content (table 1). | | | l l l l | | Candle quadrangle Not detected=21 10 — — 5 _, _ m CE) 0 m l_l L '2 E E Solomon quadrangle E 10 — Not detected=38 _ Lu 0 ‘c‘; 5 — — E 0 l_l l_l m m E :> 2 Regional area Not detected=91 10 — — 5 —— _ 10 50 100 500 1000 ppm Au 0 0.01 0.05 0.1 0.5 1 FIGURE 17.—Histograms for gold in Alaskan magnetic concen- trates; shaded areas anomalous. The area has lode deposits of copper, lead, zinc, gold, and silver (Cobb, 197 2t). SOUTHERN ALASKA Silver and gold are absent in most of the magnetic concentrates from southern Alaska; the few samples that contain silver have weakly anomalous amounts of it, whereas gold is present in strongly anomalous amounts or not at all. Sample 2192 from the Anchorage quadrangle, which contains threshold amounts of silver but lacks gold at the lower limit of detection (table 1), comes from Arch- angel Creek in the vicinity of many lode deposits of gold, some of which contain silver and lead (Cobb, 1972a). The creek is a tributary to the Little Susitna River on the south flank of the Talkeetna Mountains and is located along the southern border of a granitic batholith (Dutro and Payne, 1954; Cobb, 1972a). In the McCarthy quadrangle, threshold silver con- tent is determined in magnetic concentrate 2148, but insufficient material is available to permit analysis for gold (table 1). The concentrate is from Dan Creek im- mediately downstream from the Nikolai Butte copper deposit (MacKevett and Cobb, 1972) and from lode deposits that contain antimony, copper, gold, silver, and tungsten. The gold-rich magnetic concentrate 38 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 2438 (table 1) is from the Kennecott copper mines area. Silver is present in the magnetic concentrate, but its concentration (0.8 ppm) is just under the threshold value. The two gold-rich magnetic concentrates (232 and 241) from the Mount Hayes quadrangle (table 1) come from gold placers in the Slate Creek area where both placer gold and lode deposits of gold, silver, and copper are known (Cobb, 1972b). Silver is present in the con- centrates, but only in background amounts (table 1). Weakly anomalous silver content is detected in magnetic concentrate 196 from a gold placer on Little Moose Creek in the Mount McKinley quadrangle (table 1), but insufficient sample is available for a determina- tion of‘ gold. In sample 1028 from the Caribou Creek gold placer (Cobb, 1972c) the highly anomalous value ’of 10.5 ppm gold is determined, but silver content is at background level (table 1). The Talkeetna quadrangle yielded one magnetic con- centrate (254) that has strongly anomalous gold con- tent and background silver content, and one sample (482) that has threshold amounts of silver but is too small to provide material for the determination of gold (table 1). The gold-rich magnetic concentrate is from Mills Creek at a site downstream from the Big Boulder Creek and Chicago Gulch gold placers near Fairview Mountain (Clark and Cobb, 1972). The sample with threshold silver content is from Canyon Creek, a tributary to the Long Creek placer deposits of gold, platinum, and tin (Clark and Cobb, 1972). SOUTHWESTERN ALASKA Anomalous and nearly equal amounts of gold are detected in two magnetic concentrates (918 and 928) from the Bethel quadrangle in southwestern Alaska (table 1), and both have equal background quantities of silver. Placer gold deposits are reported about 19 km upstream from the area of sample 918, and 928 is from the Marvel Creek gold placer (Cobb, 1972e). Of the two magnetic concentrates from the Good- news quadrangle (table 1), sample 149 contains anomalous amounts of silver and background amounts of gold; and sample 268 has strongly anomalous gold content, but its silver content is just below threshold. Sample 149 is from the discovery claim on the Wat- tamuse Creek gold placer in the Slate Creek area (Cobb and Condon, 1972), and 268 is from the Snow Gulch gold-platinum placer on the Arolik River. Magnetic concentrate 3779 is from Millets prospect near Chekok Bay in the Iliamna quadrangle where cop- per, gold, and silver are reported (Detterman and Cobb, 1972). The sample has 12 ppm silver (table 1), but it was of insufficient size for an analysis of gold. The Bowmen Cut at the Portage Creek gold placer (Cobb, 1972g), in the Lake Clark quadrangle, is the source of magnetic concentrate 3799, which has back- ground silver and anomalous gold content (table 1). In the Russian Mission quadrangle, magnetic con- centrate 2119 from the Bear Creek gold placer in the Bonanza Creek area (Hoare and Cobb, 1972) contains anomalous amounts of gold and low background amounts of silver (table 1). WEST—CENTRAL ALASKA Analytical results from two of the sampled quadrangles in west-central Alaska, the Candle and the Solomon quadrangles, are discussed in other parts of the text. Neither quadrangle contains a notable number of magnetic concentrates that have anomalous silver and gold content. Only two concentrates out of the 85 analyzed for silver and one of the 25 analyzed for gold in the Candle quadrangle have anomalous amounts of these metals (table 1). Of the 101 magnetic concentrates from the Solomon quadrangle analyzed for silver, three are anomalous, and of the 41 analyzed for gold, one is anomalous. High values were found for silver in two samples from the Ruby quadrangle, and strongly anomalous amounts of gold were detected in two samples from the McGrath quadrangle. The Dahl Creek gold placer in the Bendeleben quad- rangle (Cobb, 1972b) is the source locality of magnetic concentrate 400, which has threshold gold content and low background silver content (table 1). The five magnetic concentrates from the Iditarod quadrangle that have anomalous silver content (table 1) were too small to permit analysis for gold; thus, it is not known if their gold content might also be anomalous, but the five come from gold placers (Cobb, 1972i). Magnetic concentrates 483 and 1917 from the McGrath quadrangle have strongly anomalous gold content and low background silver content (table 1). Both are from gold placers (Cobb, 1972j). Magnetic concentrate 304 from the Norton Bay quadrangle contains threshold amounts of gold, but its silver content is below the limit of detection (table 1). The sample comes from a gold placer 6 km east of Ungalik. This placer contains minor amounts of anti- mony and tungsten (Cobb, 1972m). Two magnetic concentrates (56 and 59) from the Ruby quadrangle, which have highly anomalous silver contents (table 1), are associated with lode and placer deposits of gold and tin (Cobb, 1972n), but both samples were too small to permit the determination of gold. DISTRIBUTION OF THE ELEMENTS 39 EAST-CENTRAL ALASKA East-central Alaska yields magnetic concentrates with strong anomalies in silver and gold, but relatively few samples are represented. All the concentrates from this area were collected from streams that drain granitic rocks except for samples 74, 1446, and 1455 from the Livengood quadrangle, which were collected from areas underlain by mafic, ultramafic, sedimen- tary, and metasedimentary rocks. Concentrates from both terranes have anomalous silver and gold con- tents, but the samples from the granitic areas in the Livengood quadrangle have higher values for silver and lower values for gold than samples from areas of mafic rocks. The magnetic concentrate (3646) from the Portage Creek gold placer in the Circle quadrangle (Cobb, 19720) has anomalous silver content and strongly anomalous gold content (table 1). The amount of gold, 640 ppm, is the largest value determined for that element in the magnetic concentrates analyzed for this study. Four magnetic concentrates (1, 27, 535, and 3689) from the Eagle quadrangle are variously anomalous for silver and gold (table 1). All are associated with gold placers (Cobb, 1972p). Sample 1 with 9.6 ppm gold is from Wade Creek. Sample 27, from the Chicken Creek area, has 28 ppm silver but was too small for a gold analysis. The Myers Fork material (sample 535) has anomalous gold content (5.1 ppm) but only background silver content, whereas sample 3689 from the Atwater Bar of the South Fork Fortymile River has anomalous contents of both silver (1 ppm) and gold (2.7 ppm). Three magnetic concentrates from the Livengood quadrangle contain anomalous amounts of silver (36, 74, and 100), and four (97, 100, 1446, and 1455) have anomalous to highly anomalous gold content (table 1). Samples 36 and 74 were not analyzed for gold owing to their small size, but are probably auriferous because they are from areas of gold lode and placer deposits (Cobb, 1972q). Magnetic concentrate 2418 from Rhode Island Creek in the Tanana quadrangle has strongly anomalous silver content (table 1), but was not analyzed for gold. The source locality of the sample is downstream from speculative and unproven lode deposits of gold, lead, and tin (Cobb, 19723). BISMUTH Bismuth shows three populations in its regional distribution (fig. 8). More than half of the magnetic concentrates analyzed contain 10 ppm bismuth, and the maximum concentration reaches 90 ppm (fig. 18). The histogram shows a positive skewness. From the cumulative frequency diagram (fig. 8) it appears that the regional threshold for bismuth is about 14 ppm (table 9). A clear geographic control is seen for the regional distribution of bismuth in the magnetic concentrates (table 1). Anomalous amounts of the element are lacking in samples from southeastern Alaska. Only one quadrangle in southern Alaska—the Talkeetna Moun- tains quadrangle—yielded a magnetic concentrate with anomalous bismuth content, and that sample (2181) is only weakly anomalous (20 ppm). In southwestern Alaska only two samples were found to have anomalous amounts of bismuth. Both have the weakly anomalous content of 20 ppm; one (929) is from the Bethel quadrangle, and the other (553) is from the Hagemeister Island quadrangle. The quadrangles in west-central Alaska and east-central Alaska have the greatest amounts of bismuth in magnetic concen- trates. Two of these, the Candle and Solomon quadrangles, are discussed in separate sections of the text and are not considered here. Elsewhere in west- 200 ll lllll lll ll II 180 _ Regional area _ Not detected=30 160— _ (2 140 — _ Q '2 _Z_ 120 — ._ E n: LIJ E 1 _ o 100 LI— 0 Solomon quadrangle E 80* Not detected=1 — m g Candle quadrangle z 60— Not detected=2 I H _ 0 1 5 10 50100 1 5 10 501001 5 10 50100 ppm Bi FIGURE 18.—Histograms for bismuth in Alaskan magnetic concen- trates; shaded areas anomalous. 40 central Alaska, bismuth-rich magnetic concentrates are characteristic of the Bendeleben, Idatarod, Medfra, Ruby, and Teller quadrangles. In east-central Alaska, samples with anomalous bismuth content are found in the Circle, Eagle, and Livengood quadrangles. WEST-CENTRAL ALASKA Magnetic concentrates 2027, 3042, 3047, 3054, 3056, 3069, 3070, and 3076 contain bismuth in anomalous amounts ranging from 15 to 45 ppm (table 1). All ex- cept 2027 are from a tight cluster of sample sites on the east side of the Darby Mountains, where small lode gold deposits have been reported (Cobb, 1972h), and where large anomalous values for bismuth—more than 10,000 ppm—were discovered in stream sediments by Miller and Grybeck (1973, p. 4). The area was regarded by Miller and Grybeck to be highly mineralized, a conclusion also supported by the anomalous values found for many metals in these magnetic concentrates. Sample 2027, which contains anomalous amounts of bismuth, is from a site 5.5 km north of a gold and tin placer reported on Humboldt Creek (Cobb, 1972h). Only one of the seven magnetic concentrates from the Iditarod quadrangle lacks anomalous amounts of bismuth (table 1); the rest have from 15 to 25 ppm bismuth. The samples are from the mineralized area around Flat and southward to Chicken Creek, where placer deposits of gold, silver, antimony, chromium, and tungsten and lode deposits of antimony and gold with associated cobalt are known (Cobb, 1972i). No mineral deposits have been reported (Cobb, 1972k) for the Medfra quadrangle in the vicinity of magnetic concentrate 296, which has threshold bismuth content (table 1). The other concentrate (900) with threshold bismuth is from Greer Gulch near Jumbo Peak. Greer Gulch drains a small mass of granite that hosts several lode deposits of bismuth, copper, gold, and silver about 5 km to the southwest (Cobb, 1972k). Two magnetic concentrates from the Ruby quad- rangle contain anomalous amounts of bismuth (table 1): sample 56 has 15 ppm and sample 59 has 90 ppm, which is the largest value found for bismuth in any of the 347 magnetic concentrates. These are unusual samples in that they contain anomalous amounts of five and eight different metals, respectively. Only gold placers are reported in the vicinity of sample 56 (Cobb, 1972n), but both gold and tin placers have been men- tioned in the vicinity of sample 59 (Chapman and others, 1963, p. 48). Magnetic concentrate 497, from Cape Creek at Tin City in the Teller quadrangle, contains 45 ppm bismuth (table 1). However, bismuth is not detected in EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA drill core from the Bartels tin mine area at the head of Cape Creek (Mulligan, 1966, table 10). EAST-CENTRAL ALASKA One magnetic concentrate each from the Circle (3646), Eagle (2251), and Livengood (36) quadrangles in east-central Alaska contains anomalous amounts of bismuth (table 1). The sample from the Circle quadrangle not only yields the highly anomalous value of 70 ppm bismuth, but also is anomalous for six other elements. Its source is the H. C. Carstens placer gold mine on Portage Creek, where bismuth, copper, gold, tin, and tungsten have also been reported (Cobb, 19720). The sample from the Eagle quadrangle con- tains only threshold amounts of bismuth. It is from the Fortymile River about 2.4 km downstream from the nearest gold placer of a group of placers scattered upstream to the head of the river (Cobb, 1972p). Sam- ple 36, from the Livengood quadrangle, contains 55 ppm bismuth and comes from a site on Fish Creek about 3 km downstream from placers where antimony, bismuth, gold, tin, and tungsten have been reported (Cobb, 197 2g). A small gold-bismuth-quartz lode is also located in Melba Creek, a headwater tributary of Fish Creek. COBALT AND NICKEL The distributions of the ferride elements cobalt and nickel are the least skewed from lognormal of all the metal distributions studied in these magnetic concen- trates. Although cobalt shows two populations and nickel shows four (figs. 9 and 10), the cumulative fre- quency curves for cobalt and nickel do not depart much from a linear distribution. An estimate of the backgrounds for these two metals is given by the inter- section of the 50-percent ordinate and the cumulative frequency curves, which is at 45 ppm for both metals (figs. 9 and 10). This is close to the geometric means of 44 ppm for cobalt and 50 ppm for nickel listed in table 8. From the geometric deviations given in table 8 and the histograms in figures 19 and 20, it is seen that values for nickel have a wider dispersion than those for cobalt. However, both have a slight positive skewness. The threshold value for each element is derived by Lepeltier’s (1969) method from inflections of the regional cumulative frequency curves. On the curve for cobalt (fig. 9), concentrations above the 95-ppm threshold are considered anomalous. The high-value branch thus defined deviates to the right slightly from the lognormal distribution because of an excess of high values. From the curve for nickel (fig. 10), a threshold value of 240 ppm is taken from the midpoint of the third branch. DISTRIBUTION OF THE ELEMENTS 41 I I I | Candle quadrangle Not detected=0 20— — Solomon quadrangle Not detected=0 m z 9 '5: E E n: E ._._ g 100 _ Regional area _ “- Not detected=0 O _ n: u: m 5 80 _ _ z 60 — — 4o — _ 20 — — 0 H rh— l .l. | 1 5 10 50 100 500 1000 5000 . ppm Co FIGURE 19.—Histograms for cobalt in Alaskan magnetic concen- trates; shaded areas anomalous. The geometric mean contents of 44 ppm and 50 ppm for cobalt and nickel, respectively, in the magnetic concentrates from Alaska are not greatly differ- I I I I | I | Candle quadrangle 40 — Not detected=0 _ 20 — — 0 (I) (S 40 Solomon quadrangle _ LE Not detected=0 Z 2 cc LIJ In 20 — o u. 0 n: LU an 2 D 0 z I— _ I— Regional area Not detected=0 4o — _ __ 20 ~ _ O I | I . 1 5 10 50 100 500 1000 5000 ppm Ni FIGURE 20,—Histograms for nickel in Alaskan magnetic concen- trates; shaded areas anomalous. ent. However, in magmatic differentiation, cobalt shows only a slight early enrichment whereas nickel is strongly enriched in the early differentiates, par- ticularly the ultramafic rocks, yielding Co/Ni ratios that increase from 0.08 for peridotites to 3.3 for granites (Rankama and Sahama, 1950, p. 681—683); Nickel is thus depleted at a faster rate than cobalt during magmatic differentiation. Owing to the substi- tution of these elements for iron in magnetite, the Co/Ni ratio in magnetites precipitated at different stages of differentiation has been found to increase as the stage of fractionation of the magma becomes more advanced (Wager and Mitchell, 1951; Howie, 1955). 42 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA The value of Co/Ni ratio in magmatic iron ores has been suggested as an index of the degree of fraction- ation (Landergren, 1948, p. 122—129; Davidson, 1962, p. 79); however, Sen, Nockolds, and Allen (1959) and Frietsch (1970, p. 93—104) reported that this ratio doesn’t always increase in magnetite with increasing magmatic fractionation. Frietsch further noted that the ranges of the Co/Ni ratios for magnetites of magmatic, volcanic-sedimentary, and metasomatic origin were similar, suggesting that the ratio is unrelated to origin. The Co/Ni ratio for this group of Alaskan magnetic concentrates, as derived from the geometric means (table 8), is slightly below unity. This suggests an in- fluence from sources in ultramafic magmatic rocks, because many of the concentrates from this source have remarkably high nickel content, locally as great as 20—40 times the mean content. Where the cobalt contents of the magnetic concentrates are consistently higher than the nickel contents, the abundance data are a strong confirmation that the magnetites were derived from silicic igneous rocks. The inflections on the regional cumulative curve for nickel (fig. 10), which suggest four populations of nickel-bearing magnetic concentrates, may be in- terpreted to show that the data for nickel are more sensitive to the geologic source of the concentrates than are the other chemical data. The samples are from geologically diverse source materials. The four branches of the curve, from low values to high values for nickel, may reflect derivation of the concentrates respectively from silicic rocks, intermediate rocks, ultramafic rocks, and mineralized rocks. The presence of threshold amounts of nickel, where the magnetic concentrates are derived from ultramafic or mafic rocks, probably should be regarded as normal instead of being thought of as possibly indicating mineraliza- tion, because of the natural enrichment of nickel in these rocks. Although most of the cobalt and nickel deposits re- ported by Berg and Cobb (1967) are in southeastern Alaska, none of the concentrates came from the vicini- ty of these deposits, and none of the concentrates from the Bradfield Canal, Juneau, and Ketchikan quadrangles is anomalous for either cobalt or nickel (table 1). Other areas in which the analyzed magnetic concentrates lack anomalous cobalt or nickel content are the McCarthy, Nabesna, and Valdez quadrangles in southern Alaska; the Iliamna and Russian Mission quadrangles in southwestern Alaska; the Medfra, Nome, Norton Bay, and Teller quadrangles in west- central Alaska; and the Eagle and Fairbanks quadrangles in east-central Alaska. SOUTHERN ALASKA One magnetic concentrate (2199) from the Anchor- age quadrangle has a weakly anomalous cobalt con- tent, and two (2190 and 2203) have anomalous nickel content (table 1). Strangely, the cobalt-enriched sam- ple, the source of which is downstream from exposures of ultramafic rocks in the western Chugach Mountains (Clark and Bartsch, 1971; Clark, 1972), has no anomalous nickel content, even though nickel was shown by Clark and Bartsch (1971, p. 14) to be six times as abundant as cobalt in these rocks. The nickel- rich magnetic concentrate from Wolverine Creek (2203), which lacks anomalous cobalt content, was collected downstream from the Wolverine ultramafic complex, where dunite and peridotite were found by Clark (1972, p. 10) to contain 1500—3000 ppm nickel and 150—300 ppm cobalt. Quartz diorite and mica schist (Capps, 1915, pl. 3) would appear to be the source rocks for the magnetic concentrate (2190) from Willow Creek with weakly anomalous nickel content, but glacial erosion in the area may have contributed detritus, including nickel-bearing magnetite, from otherwise unrecognized ultramafic source rocks (Paige and Knopf, 1907, p. 65—67; Capps, 1915, p. 38—39). Of the magnetic concentrates from the Mount Hayes quadrangle, five from the Slate Creek area (230, 232, 238, 241, and 293) have anomalous contents of both cobalt and nickel (table 1), one from the Rainbow Mountain area (1511) has anomalous cobalt content, and another from the same area has anomalous nickel content. Gabbro is in part the source for samples 230 and 232 (Rose, 1967, fig. 1). Possibly similar rocks, or even pyroxenite, peridotite, and hornblendite locally rich in magnetite, are partial sources for the other anomalous magnetite concentrates from the Slate Creek area. Details of the bedrock in the eastern part of this area are lacking, but an extension of ultramafic rocks toward the east is probable (Rose, 1967, p. 8 and table 3). Samples 1511 and 1513 are from the vicinity of copper, lead, gold, silver, and nickel prospects in the Rainbow Ridge area (Hanson, 1963, p. 67—70), where the nickel and part of the copper are associated with ultramafic igneous rocks. One magnetic concentrate (1023) from the Mount McKinley quadrangle contains anomalous amounts of nickel (table 1). The sample is from Caribou Creek at the mouth of Last Chance Creek, a tributary from the southeast. Quartz lodes with stibnite are found in con- torted hornblende gneiss at this junction (Capps, 1919, p. 108; Wells, 1933, p. 353), and magnetite is a common mineral in sluice-box concentrates from Caribou Creek (Capps, 1919, p. 92). The source of the magnetite that DISTRIBUTION OF THE ELEMENTS 43 has anomalous nickel content may be the hornblende gneiss. Magnetic concentrate 1290 has anomalous cobalt and nickel contents, and sample 1336 has anomalous nickel content (table 1). These samples come from two gold placers about 1.5 km apart on Cache Creek in the Talkeetna quadrangle. Cache Creek lies in a glacial trough (Mertie, 1919, p. 242—248) between hills of Mesozoic slate and graywacke and Eocene gravel, sand, clay, and lignite (Capps, 1913, fig. 6). Glacial deposits mask the sources of the gold, platinum, cassiterite, scheelite, and magnetite found in the placers, but these minerals have been interpreted to be of local provenance (Mertie, 1919, p. 245—246). However, other studies have shown that the gravels above the slate and graywacke are derived from dis- tant sources in the Alaska Range (Robinson and others, 1955, p. 14). The presence of the detrital platinum indicates that the source area contained some ultramafic rocks, to which these nickel-enriched magnetic concentrates might be attributed. Anomalous cobalt content is present in a magnetic concentrate (2181) from the gold placer on Albert Creek in the Talkeetna Mountains quadrangle (table 1). Albert Creek drains an area underlain by Mesozoic volcanic and sedimentary rocks and generally covered by glacial and fluvioglacial deposits (Cobb, 1973, p. 29). A little detrital platinum has been reported with the gold, but local sources have not been identified for these placer minerals. The lack of anomalous nickel content in the magnetic concentrate suggests that silicic rocks more than ultramafic rocks have been a source for the detrital magnetite. SOUTHWESTERN ALASKA Two magnetic concentrates (928 and 929) from gold placers on Marvel Creek in the Bethel quadrangle of southwestern Alaska have anomalous amounts of nickel but lack anomalous cobalt content (table 1). The stream drains an area underlain by Cretaceous sedi- mentary rocks, chiefly interbedded graywacke and silt- stone with lesser amounts of conglomerate, into which are intruded small stocks of granite and dikes and sills of gabbro and basalt (Hoare and Coonrad, 1959). Other metals such as copper, zinc, silver, and gold are also anomalous in these concentrates, but the sources of the nickel-rich magnetite are uncertain; seemingly the mafic rocks are insufficiently abundant to account for it. The magnetic concentrate (268) from the Arolik River in the Goodnews quadrangle, which has anom- alous nickel content and background cobalt content (table 1), was collected at a site about 11 km down- stream from a gabbroic stock and exposures of mafic volcanic rocks (Hoare and Coonrad, 1961a). Doubtless these rocks are the source of this threshold anomaly in the magnetic concentrate. Cobalt and nickel contents are each anomalous in magnetic concentrates 542 and 553 collected south of Red Mountain in the Hagemeister Island quadrangle (table 1). Neither element is strongly anomalous, but the source is probably the nearby pluton of ultramafic rocks (Hoare and Coonrad, 1961b). The area is the site of major platinum placers (Mertie, 1940a, pl. 8). Weakly anomalous values for cobalt and nickel are found for sample 3797 from Hatchet Creek in the Lake Clark quadrangle (table 1). Mineral deposits have not been identified in the area (Cobb, 1972g). Hatchet Creek rises in a terrane of Mesozoic mafic lavas and traverses a sequence of Paleozoic metamorphosed sedimentary rocks (Capps, 1935, pl. 2) to enter the northern end of Lake Clark. Had the magnetic concen- trate been anomalous only for cobalt and nickel, then its source might have been attributed to the lavas, but sample 3797 also contains anomalous amounts of cop- per and zinc. Thus, a source including some base-metal sulfide minerals is not wholly improbable. WEST-CENTRAL ALASKA Magnetic concentrates 400 and 3041, from the Bendeleben quadrangle in west-central Alaska, have weakly anomalous cobalt content and strongly anomalous nickel content (table 1). Both samples come from streams that drain areas underlain by Precam- brian carbonaceous siltite, although the stream that provided sample 3041 also drains an area of Paleozoic limestone (C. L. Sainsbury, oral commun., 1972). Gold content is anomalous in sample 400 and zinc content is anomalous in 3041. Gold placer prospects have been noted in the areas (Cobb, 1972b). Relatively large amounts of cobalt (geometric means of 13—17.5 ppm) and nickel (42—70 ppm) are reported (Miller and Grybeck, 1973, table 4) in stream sediments from areas underlain by the Paleozoic limestone and Precambrian metamorphosed sedimentary rocks east of the Darby Mountains and a few kilometers south of the source for sample 3041. However, Miller and Grybeck (1973, p. 6) noted that these sedimentary rocks are characterized by low geometric mean values for cobalt and nickel. Therefore, they postulated that the high tenors for these elements in the alluvium must be caused by the introduction of debris from the diabase stocks and plugs, which are common to the area and in which high values were found for cobalt and nickel. The strongly 44 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA anomalous nickel content in the magnetic concentrate from a tributary to the Tubutulik River (3041) east of the Darby Mountains appears to confirm this inter- pretation of the source. Presumably the presence of diabase is to be expected near the source of sample 400, or possibly the lava field east of the sample site supplied nickel-rich magnetite. The amount of cobalt is anomalous in one magnetic concentrate (444) in the Candle quadrangle (table 1), and the nickel content is anomalous in two (2491 and 2696), but the three localities are scattered in the western part of the quadrangle. The cobalt-rich concen- trate is from Dime Creek at the discovery claim of placer deposits in which chromium, gold, and platinum are reported (Cobb, 1972u). The sample locality is in a small plug of basalt (Patton, 1967). Normal stream sediment from this part of Dime Creek was found by Elliott and Miller (1969, p. 28) to have threshold cobalt content, about 50 ppm, and low amounts of nickel. Magnetic concentrates 2491 and 2696 come from the Bear Creek gold and platinum placer and from east of the Spruce Creek gold placer, respectively (Cobb, 197 2u), which are both close to and downstream from intrusive masses of basalt (Patton, 1967 ). Stream sedi- ments from these localities were shown by Elliott and Miller (1969, p. 13) to have low background values for cobalt and high background values for nickel. For sample 2696, the anomalous value for nickel appears to be influenced by the presence of copious tramp iron in the magnetic concentrate, because five other samples from the vicinity (2690, 2693, 2695, 2698, and 2712) lack anomalous nickel (table 1). One magnetic concentrate with anomalous amounts of cobalt and nickel (1831) and five with background cobalt content and anomalous nickel content (1804, 1815, 1838, 1867, and 1883) were obtained from the Flat and Willow Creek areas in the Iditarod quadrangle (table 1). Sample 1831, which had anomalous cobalt content and highly anomalous nickel content, is from the Granite Creek gold placer; and 1838, which has anomalous amounts of nickel, is from a tributary placer. Mertie (1936, p. 221) remarked that the creek was not well named, because its valley is underlain mainly by sandstone and argillite with only a few dikes of granite. The fact that chromite is present with the placer gold (Cobb, 1972i) indicates an in- fluence on the detrital minerals by some mafic or ultramafic rock, as do the anomalous values for nickel in the magnetic concentrates. Possibly this source is the two small stocks of pyroxene diorite and gabbro shown by Mertie and Harrington (1916, pl. 11) as in- trusive into the metasedimentary rocks between Flat and Granite Creek on the north and south sides of Otter Creek. The concentrate from Otter Creek (1867) is evidently influenced by the northern mafic stock, and those from Flat Creek (1883) and Chicken Creek (1804 and 1815) apparently are influenced by the southern stock. The presence of mafic intrusives was mentioned by Maloney (1962, p. 8, figs. 2 and 3), but their distribution was not shown. Faintly anomalous cobalt content is detected in one magnetic concentrate (1917) from Candle Creek in the McGrath quadrangle (table 1), but the sample has only background amounts of nickel. The walls of the Candle Creek valley are reported to be largely composed of sandstone and shale intruded in the upper reaches of the valley by quartz monzonite, and the divide at the head of the valley is capped by basalt (Mertie, 1936, p. 197). Locally derived detrital gold, Cinnabar, and scheelite are present in the Candle Creek placer; their source is thought to be the intrusive quartz monzonite (Mertie, 1936, p. 197 ). The source of the magnetite with anomalous cobalt content is less certain. Because cobalt is enriched and nickel is not, the material might well be derived from the quartz monzonite, but the in- fluence of the basalt is uncertain. A magnetic concentrate (56) from the mouth of Solomon Creek in the Ruby quadrangle contains anomalous amounts of nickel, and one from Glen Gulch (59) has highly anomalous cobalt and nickel con- tent (table 1). Both sites have gold placers developed on phyllitic bedrock (Mertie, 1936, p. 157, 164; White and Stevens, 1953, p. 1). Granite is intrusive into the phyllites, but it has not been identified at the im- mediate sites of these placers. Gabbroic greenstone is present in the general area (Chapman and others, 1963, p. 37 —38) but seemingly too far south to be a possible source for the extremely anomalous sample from Glen Gulch. The origin of these anomalous samples is thus unresolved. The one from Glen Gulch probably is worth further investigation to account for its high cobalt content. A magnetic concentrate (2956) from the Kwiniuk River in the Solomon quadrangle (table 1) has weakly anomalous cobalt and nickel content and one (2887) from Cheenik Creek has threshold amounts of nickel. The drainage basin of the Kwiniuk River is underlain by limestone, dolomite, and black shale (West, 1953, pl. 1). The upstream end of the basin reaches the contact betweeen these sedimentary rocks and an in- trusive body of granite. The Cheenik Creek sample was collected at the contact between an undivided igneous complex, consisting mainly of granite with some diorite and greenstone, and a unit of metamorphic rocks composed of schists and limestone (West, 1953, pl. 1; Miller and others, 1972, map). The specific source of the anomalous magnetite is not apparent. DISTRIBUTION OF THE ELEMENTS 45 EAST-CENTRAL ALASKA Weakly anomalous nickel content was recorded in a magnetic concentrate (3646) from upper Portage Creek in the Circle quadrangle in east-central Alaska (table 1), where mica schist, quartz-mica schist, and chlorite schist are intruded by biotite granite (Nelson and others, 1954, p. 11, fig. 4). Stream and bench gravels in Portage Creek were mined for placer gold, and detrital minerals containing, variously, bismuth, copper, rare earths, tin, and tungsten have been noted in the con- centrates (Cobb, 19720), but nothing has been reported that suggests the slight nickel enrichment of this magnetic concentrate is other than a normal increase related to some as yet unknown mafic source rock. Three (74, 1446, and 1455) of the four magnetic con- centrates with anomalous nickel content from the Livengood quadrangle (table 1) are from an area in- fluenced by the abnormally nickeliferous serpentinite near Livengood (Foster, 1969, p. 2, fig. 3). Doubtless the anomalous nickel content in these concentrates reflects this source. The most nickel rich of the three (1455) also contains weakly anomalous amounts of cobalt. Sample 97 from Fairbanks Creek in the south- eastern part of the quadrangle contains anomalous amounts of nickel but only background amounts of cobalt (table 1). Gold, tin, tungsten, and bismuth minerals have been recognized in the Fairbanks Creek gold placer, but independent nickel minerals have not been noted (Cobb, 1972q); the concentrates are dominated by garnet, ilmenite, magnetite, and rutile, and also contain native bismuth, galena, arsenopyrite, wolframite, and cassiterite (Prindle and Katz, 1909, p. 187-190). Possibly the source of the nickel-enriched magnetic concentrates is the basalt on Fourth of July Hill adjacent to the placer (Prindle and Katz, 1909, p. 186; Prindle, 1913, pl. 8). The magnetic concentrate (1499) that has weakly anomalous cobalt content from the Tanacross quadrangle (table 1) is from a tributary to the Tok River that drains an area underlain by phyllite, schist, and metadiorite (Foster, 1970). The same sample is also anomalous in copper and zinc, but no mines are re- corded for this area (Cobb, 197 2r). However, samples of intrusive rocks and altered zones nearby were shown to be highly anomalous in copper, gold, silver, and zinc but were rather low in cobalt (Clark and Foster, 1969, p. 11 and fig. 1). Anomalous nickel content in magnetic concentrate 2418 from the gold placers in Rhode Island Creek in the Tanana quadrangle (table 1) is accompanied by anomalous amounts of gold, copper, and lead. Gold, lead, tin, and mercury have been found in the placers (Cobb, 19723) along with other detrital minerals among which pyrite is particularly common (Waters, 1934, p. 237-238). The nickel in this sample may be related to sources for the magnetite in small masses of serpen- tinized intrusive rocks similar to those exposed farther- east in the Tofty area (Wayland, 1961, pl. 40), but this is uncertain. A more likely possibility is that tramp iron is present in the magnetic concentrate, or that gabbroic and similar rocks occur in the headwaters of Rhode Island Creek. INDIUM AND THALLIUM Eighty-five percent of the 131 magnetic concen- trates analyzed for indium and thallium contain less than 0.2 ppm of each (table 1). The maximum value found for indium is 0.5 ppm and the maximum for thallium is 1 ppm. For the samples with values above the lower limit of detection (0.2 ppm), the geometric mean for indium is 0.23 ppm and for thallium is 0.27 ppm (table 8); these means are, respectively, twice as great and half as great as the average abundance of these elements in the Earth’s crust (Krauskopf, 1967). Considering that the largest part of the analyses (shows less than 0.2 ppm indium, it is probable that the average magnetic concentrate from Alaska is not enriched in indium over the crustal abundance of 0.1 ppm. For similar reasons, the average magnetic concentrate from Alaska probably contains less than one-fourth of the normal crustal abundance of 0.45 ppm thallium. Thus, magnetic concentrates are not ac- cumulators of these elements. Indium and thallium rarely form independent minerals, but tend to be dispersed in silicates or to be in sulfides (Rankama and Sahama, 1950, p. 723—7 27). The most important sulfide hosts are sphalerite for indium and galena for thallium. The correlation co- efficients, discussed later, show that indium correlates positively with zinc and thallium with lead. From these observations it may be concluded that indium and thallium, in the few magnetic concentrates Where they rise above the limits of detection, most likely are pres- ent in sulfides incorporated as minor minerals in the magnetite or as accessory minerals trapped between grains in the magnetic concentrates. Too few samples contained indium and thallium in amounts greater than the limit of detection to permit appropriate statistical treatment of these data. No anomalous values are given for these elements in table 9, both to conform to the lack of statistical treatment and in recognition that the abundances of indium and thallium in the magnetic concentrates seldom exceed crustal abundance for these elements and the sample medium is not a collector for them. However, some in- teresting chemical relations are lost thereby that can 46 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA be brought out by regarding all values of 0.2 ppm or more as weakly anomalous for indium and thallium in these magnetic concentrates. Southwestern and west- central Alaska are the principal centers for these elements. Indium is present above the limit of de- tection in magnetic concentrates from eleven quadrangles, but thallium is so determined in Only three quadrangles—Bendeleben, Candle, .and Solomon—where the thallium-bearing concentrates ap- pear to be derived from sources in intrusive granitic and alkalic rocks. Thallium is particularly uncommon outside these three quadrangles, and it is most com- mon in the Solomon quadrangle. Only three concen- trates, 3044 from the Bendeleben quadrangle, and 2959 and 2982 from the Solomon quadrangle, contain both indium and thallium above the limit of detection. The association of indium-bearing magnetic concen- trates with zinc-rich samples is shown in table 1; all 18 indium-bearing concentrates contain zinc and eight have anomalous zinc content. The 23 thallium-bearing samples all contain lead, but only two have anomalous lead content. A, strong correlation also exists between thallium and equivalent uranium. Of the 23 thallium- bearing concentrates, 22 also contain equivalent uranium, of which 17 are anomalous. Inasmuch as the equivalent uranium is dominantly associated with hematitic crusts on magnetite and the thallium is prob- ably in inclusions of galena in the magnetite or in detrital particles of galena associated with the magnetic concentrates, the relation between thallium and equivalent uranium is not chemical. It is geologic, and is caused by derivation from the same sources in alkalic rocks. The distributions of indium and thallium in magnetic concentrates from the Candle and Solomon quadrangles are discussed in other sections of the text. What follows is a review of the sparse data on other parts of Alaska. SOUTHERN ALASKA Two magnetic concentrates (2134 and 2438) from the McCarthy quadrangle in southern Alaska and one (2181) from the Talkeetna Mountains quadrangle con- tain indium above the limit of detection but lack thallium (table 1). Zinc content is anomalous in all three samples. Samples 2134 and 2438 are associated respectively with the Nikolai Butte copper deposit (MacKevett and Smith, 1968; 1972) and the Kennicott copper deposits (MacKevett, 1971). Sample 2181 is from a group of concentrates from gold placers along Crooked Creek (Cobb, 1972d). All these concentrates from Crooked Creek are enriched in zinc, but base- metal deposits are unreported. SOUTHWESTERN ALASKA Two magnetic concentrates (918 and 928) from the Bethel quadrangle in southwestern Alaska contain indium above the limit of detection, as does one con- centrate each from the Goodnews (149), Iliamna (3778), and Russian Mission (67) quadrangles. None has de- tectable thallium (table 1). Anomalous amounts of zinc are present in the two concentrates from the Bethel quadrangle, one of which (928) is from a gold placer (Cobb, 1972e), but neither comes from an area known for base-metal mineralization. Sample 149 was col- lected from a gold placer in the Goodnews quadrangle (Cobb and Condon, 1972), and the concentrate has anomalous silver content but lacks anomalous zinc content. Sample 3778 has anomalous amounts of zinc and is from a region in the Iliamna quadrangle where many copper lodes have been identified, but none has been described at the site of this sample (Detterman and Cobb, 1972). The magnetic concentrate (67) with detectable indium from the Russian Mission quad- rangle otherwise has no anomalous metal content. The sample is from the Ophir Creek gold placer, in which base metals are unreported (Hoare and Cobb, 197 2). WEST-CENTRAL ALASKA Indium- and thallium-bearing magnetic concentrates from west-central Alaska (table 1), other than those from the Candle and Solomon quadrangles (see below), include two concentrates (3044 and 3069) from Rock Creek in the Bendeleben quadrangle and one (1917) from the McGrath quadrangle. Both samples from the Bendeleben quadrangle show the presenCe of indium, but only 3044 has thallium above the limit of de- tection. Mineral deposits have not been reported for the two localities on Rock Creek, nor have geochemical anomalies been observed (Miller and Grybeck, 1973, p. 30—31), but several gold placers have been noted in the general area (Cobb, 1972b). The sample from the McGrath quadrangle has anomalous amounts of base metals and cadmium and is from a gold placer (Cobb, 1972j). EAST-CENTRAL ALASKA The only magnetic concentrate from east-central Alaska with detectable indium is sample 3646 from the Circle quadrangle (table 1). A variety of elements, in- cluding zinc, are present in anomalous amounts in the concentrate, and the sample is from a locality about 2.4 km upstream from the Portage Creek zinc lode oc- currence (Cobb, 19720). Thallium is below the limit of detection, and lead content is not anomalous. DISTRIBUTION OF THE ELEMENTS 47 CANDLE QUADRANGLE RESULTS Eighty-five magnetic concentrates analyzed for this study come from the western half of the Candle quad- rangle. The locations of the samples and the results of the analyses are listed in table 1. The geology of the area sampled in the Candle quad- rangle is presented in figures 21 and 22, adapted from Patton (1967). The area is predominantly underlain by younger volcaniclastic rocks, flows, and granitic in- trusives, though parts of it are underlain by Paleozoic limestone and schist. Three units of volcaniclastic and volcanic rocks were recognized by Patton (1967). The oldest unit consists of Jurassic(?) and Cretaceous andesitic and trachyandesitic crystal tuff, lithic tuff, tuffaceous volcanic graywacke, massive andesitic brec- cia, agglomerate, conglomerate, and intercalated flows of porphyritic pyroxene andesite and basalt. A unit of slightly younger age comprises Cretaceous volcanic graywacke and conglomerate, which is composed mainly of andesitic rock detritus and which locally con- tains large quantities of granitic debris and fine- grained tuff. The youngest of these units consists of Tertiary(?) and Quaternary flows of gray to dark-red vesicular olivine basalt with some black, dense, glassy basalt. Cretaceous granitic rocks, principally horn- blende and pyroxene monzonite, syenite, and biotite- quartz monzonite, intrude the older rocks. Much of the granitic detritus in the Cretaceous volcanic graywacke is lithologically identical to the granitic plutons. Where these plutons intrude the Jurassic(?) and Creta- ceous andesites, the volcanic rocks are said (Patton, 1967) to be hornfelsic and propylitically altered to a hard, pale-green aggregate of chlorite, epidote, calcite, and sodic plagioclase. Many small intrusive bodies of hybrid diorite, syenite, and monzonite cut the volcanic rocks in the vicinity of the granitic plutons. In ques- tionable Tertiary and early(?) to middle Pleistocene time, the flows of olivine basalt were extruded and spread out over a terrain of moderate relief (Patton, 1967). Only 34 localities are shown on figures 21 and 22 for the 85 magnetic concentrates from the Candle quad- rangle, because the scale of the maps causes many in- dividual sample localities to overlap. This distribution results from the fact that some localities were sampled in detail, or were visited several times over the years by geologists of the US. Geological Survey, each of whom collected one or more concentrates at the site. Discrimination of the closely adjacent samples listed for one locality on figures 21 and 22 can be made by reference to table 1. Five of the sample localities shown on figures 21 and 22 each provided 5 to 10 magnetic concentrates. The individual concentrates were collected mainly from closely adjacent localities, but several concen- trates are duplicates from the same site. Gross varia- tions in the trace-element composition of concentrates from the same locality could be expected if the samples represented highly diverse sources, or if the magnetite contained unusually diverse inclusions or un- systematic intermixtures of accessory minerals. In order to prepare maps showing the distribution of the minor metals in the magnetic concentrates from the Candle quadrangle, it was necessary to reach an average value for each metal at every locality where two or more samples are represented. These averages are given in table 12, which also identifies the sample numbers at each locality. The statistical procedures used to obtain single values for each metal at the localities that had 5 to 10 samples afford an opportunity to test the variance for each element at each site (table 13), and to compare this variance with the relative standard deviation of the analytical method, as shown in table 5. The relative standard deviations in percent are shown in table 14. In the data from localities A—E (table 13), all the relative standard deviations shown for silver are much smaller, and those for bismuth average slightly smaller, than the relative standard deviations found for these elements in the 30 replicate subsamples of sample 3799 (table 5). For the other elements, the relative standard deviations obtained from the 5 to 10 samples that represent each of these localities are all much greater than the comparable figures obtained for the subsamples of 3799. The largest variations within table 13 are for equivalent uranium, silver, cadmium, and copper, and the smallest variations are for cobalt. The relative standard deviation for equivalent uranium at locality D of table 13 is zero because all values were below the limit of determination. Other- wise, the high relative standard deviations for equivalent uranium further substantiate that the hematitic coatings on magnetite, which vary greatly in thickness, are the predominant source of the radio- activity (see table 11). The large relative standard deviations for silver and cadmium in table 13 are caused by setting the values too low for the lower limit of determination for these two elements. On both tables 5 and 13, the contents of silver and cadmium in the concentrates were in the lowest ranges of reported values. This is the range most affected by instrumental noise. The average relative standard deviation for bismuth at localities A—E, table 13, is a little less (33 percent) than the relative standard deviation for the analytical procedure (36.6), but at one locality it is considerably greater, and overall there is a range from 19 to 51 per- EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 48 we“; V. r t 2. 1.»... h 15 KILOMETEHS 10 MILES howing the distribution of analyzed magnetic concentrates from the northwestern part of the Candle ogre map 5 .—Geol FIGURE 21 quadrangle, Alaska. DISTRIBUTION OF THE ELEMENTS 49 CORRELATION OF MAP UNITS Geology generalized from Patton (1967) SEDIMENTARY ROCKS IGNEOUS ROCKS QUATERNARY AND TERTIARY(?) CRETACEOUS CRETACEOUS AND JURASSICC’) 0a } QUATERNARY SEDIMENTARY ROCKS Qa ALLUVIUM (QUATERNARY)—Flood-plain deposits of gravel, sand, and silt, and terraced alluvial silt IGNEOUS ROCKS BASALT (QUATERNARY AND TERTIARY?)—Flows of gray to dark-red vesicular olivine basalt, some black dense glassy basalt Kg GRANlTIC ROCKS (CRETACEOUS)—Chiefly hornblende and pyroxene monzonite and syenite; some biotite-quartz monzonite ANDESITIC VOLCANIC ROCKS (CRETACEOUS AND JURASSIC?)—lncludes a wide variety of andesitic flows and volcaniclastic rock 0Tb KJv Contact—Dashed where approximately located, inferred, or indefinite Fault—Dashed where approximately located or inferred . 2502 Locality of magnetic concentrate—Shows Alaskan placer concentrate file number. Where two or more concentrates are from the same locality, the range in file numbers is given cent. These differences are not great, but they are interpreted to indicate an erratic presence of bismuth in magnetic concentrates from the same general area, possibly attributable to minor minerals included in the magnetite or accessory minerals in the magnetic concentrate. The departure of the relative standard deviations for cobalt, copper, nickel, lead, and zinc from the relative standard deviations for these metals in the replicate analyses (table 14), shows the influence of the locali- ties, methods of collection, and methods of preparation on the composition of these magnetic concentrates. The large relative standard deviation for copper at localities A—E, compared to the relative standard deviation for copper in the replicate analyses (table 14), probably indicates the variable presence of chal- copyrite or other copper sulfide minerals as minor minerals included in magnetite, or the variable presence of such sulfides as accessory minerals in the magnetic concentrates. Remarkable similarities exist in the relative stan- dard deviations for nickel, lead, and zinc at localities A—E, but most are much greater than the relative stan- dard deviations for these elements in the subsamples of 3799 (table 14). Nickel fits in this group only because the single aberrant value for sample 2696 at locality D has been excluded as caused by abundant tramp iron (see discussion of mineralogy). If this value were re- tained in the analysis of variation, then the coefficient of variation for nickel would be 150 percent, making nickel the most erratic of the elements in these concen- trates. The consistency with which nickel otherwise ap- pears suggests that the value reported for 2696 should be dropped, and that the amount probably should have been given as about 55 ppm instead of 570 ppm. The similar relative standard deviations for nickel, lead, and zinc at localities A—E, and the fact that these measures are much greater than those for the replicate subset (table 14), suggest that the mode of occurrence of these elements in the magnetic concentrates (sorp- tion, substitution, minor included minerals, or accessory minerals) is sufficiently consistent that the collecting and processing of the samples do not exag- gerate differences in distribution, as they apparently did for equivalent uranium. The low relative standard deviations for cobalt at localities A, B, C, and E, table 13, are as much as twice as great as the relative standard deviation for that element in the method used for analysis (table 5). The amount of variation probably is as low as can be ex- pected for material from several sources at each local- ity. Indeed, these low relative standard deviations may indicate that the main mode of occurrence for cobalt in magnetic concentrates from the Candle quadrangle is diadochic substitution in magnetite. No other mode would reduce the values resulting from diverse prov- enance, collection, and preparation to so narrow a range. The distribution of the anomalously abundant ele- ments in the magnetic concentrates from the Candle quadrangle is based on threshold values given in table 9, which do not differ greatly from the regional values. Copper, nickel, and lead in the Candle quadrangle have slightly lower thresholds, and zinc has a slightly higher threshold, than the regional values. The order of discussion follows that used for the region. EQUIVALENT URANIUM Most of the low values for detected equivalent uranium in the magnetic concentrates from Alaska are in samples from the Candle quadrangle, where only 20 percent of the samples have radioactivities above 50 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA R14 Wt 151°30’ 13 12 ‘ 161°00' 1’1” R10 We .zwr- 65°3D’ . _‘ _ ‘7“ 55°15 ‘ 15 KILOMEI'EHS l | 10 MILES FIGURE 22.—Geologic map showing the distribution of analyzed magnetic concentrates from the southwestern part of the Candle quadrangle, Alaska. DISTRIBUTION OF THE ELEMENTS 51 CORRELATION OF MAP UNITS Geology generalized from Patton (1967) SEDIMENTARY ROCKS IGNEOUS ROCKS QUATERNARY AND QUATERNARY TERTIARY(?) CRETACEOUS CRETACEOUS CRETACEOUS AND JURASSIC(?) PALEOZOIC SEDIMENTARY ROCKS Qa ALLUVIUM (QUATERNARY)—Flood-plain deposits of gravel, sand, and silt, and terraced alluvial silt Kc MD(ED SEQUENCE (CRETACEOUS)—-Consists of calcare- ous graywacke, calcareous mudstone, volcanic graywacke, and volcanic conglomerate Kv VOLCANIC GRAYWACKE AND CONGLOMERATE (CRETACEOUS)—Composed chiefly of andesitic volcanic rock detritus; locally includes notable amounts of granitic rock detritus and fine-grained tuffaceous material le LIMESTONE (PALEOZOIC)—Chiefly recrystallized limes- tone and dolomite. Includes calcareous mica schist and dark phyllite IGNEOUS ROCKS BASALT (QUATERNARY AND TERTIARY? )——Flows of gray to dark-red vesicular olivine basalt; some black dense glassy basalt Kg GRANlTlC ROCKS (CRETACEOUS)—Chiefly hornblende and pyroxene monzonite and syenite; some biotite-quartz monzonite ANDESITIC VOLCANIC ROCKS (CRETACEOUS AND JURASSIC?)—lncludes a wide variety of andesitic flows and volcaniclastic rocks 0Tb KJv Contact—Dashed where approximately located, inferred, or indefinite Fault—Dashed where approximately located or inferred; dot- ted where concealed; queried where doubtful .2501 Locality of magnetic concentrate—Shows Alaskan placer concentrate file number. Where two or more concentrates are from the same locality, the range in file numbers is given background (table 1). The radioactive samples contain from 40 to 160 ppm equivalent uranium with a geo- metric mean of 65 ppm (table 8). The distribution of equivalent uranium in magnetic concentrates from the Candle quadrangle appears as a straight-line cumulative frequency curve (fig. 1). This line is located left of the regional curve and the curve for equivalent uranium in magnetic concentrates from the Solomon quadrangle. The 17 magnetic concentrates with above-back- ground amounts of equivalent uranium come from only nine localities in the Candle quadrangle (figs. 23 and 24). The highest value for equivalent uranium (160 ppm) was measured in sample 2722 from outside the southeastern contact of the Hunter Creek pluton (fig. 21), the next highest value (120 ppm) was determined on sample 2538 from the Granite Mountain pluton (fig. 22). Other radioactive magnetic concentrates derived from streams draining the granitic rocks of the Hunter Creek pluton are: 2640, 70 ppm; 2665, 80 ppm; 2667, 60 ppm; 2668, 50 ppm; 2669, 100 ppm; 2785, 50 ppm; 2799, 40 ppm; and 2808, 80 ppm. A magnetic concen- trate (2797) from the area of basalt immediately east of the eastern margin of the Hunter Creek pluton showed '50 ppm equivalent uranium and reflects the presence of the pluton. Other radioactive magnetic concentrates from the Granite Mountain pluton are: 2549, 50 ppm; 2556, 40 ppm; 2560, 40 ppm; 2570, 100 ppm; and 2571, 70 ppm. The most southerly magnetic concentrate (444) from the Candle quadrangle is also slightly radio- active (40 ppm equivalent uranium). This sample seems to be derived solely from andesitic volcanic rocks, but the mere appearance of this radioactivity may indicate the presence there of previously unrecognized granitic rocks. Most of the radioactive magnetic concentrates from the Candle quadrangle are derived from the marginal parts of the Hunter Creek and Granite Mountain plutons. Owing to the zoned character of these plutons, the rocks along the margins have syenitic composition (Miller, 1970; 1972). The radioactive magnetic concen- trates are probably from this syenite. COPPER, LEAD, ZINC, AND CADMIUM Copper shows three populations, lead shows two populations, and zinc and cadmium show one popula- tion each in the magnetic concentrates from the Candle quadrangle (figs. 2—5). The cumulative frequency distribution curves of copper, lead, and zinc in this quadrangle are similar to the regional distribution curves except in the high-value portions, where the regional curves show a positive skewness, indicating that the magnetic concentrates with the higher metal contents are not from the Candle quadrangle. The geometric means of copper, lead, and cadmium are close to the regional means, but that of zinc is less than the regional mean (table 8). Of the magnetic concentrates from the Candle quad- rangle, 50 are clearly ascribable to sources in the granitic plutons, and 22 are derived from areas of 52 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA TABLE 12,—Values used to plot equivalent uranium and 11 elements in magnetic concentrates from 34 localities in the Candle quadrangle, Alaska [Data are in parts per million. Numbers in parentheses below the element symbols are the lower limits of determination. N = not “ ‘ Jr L = -‘ ‘ but the ‘ union is below the limit of determination; n.d. = not determined] Co Cu Ni Pb Zn Au F1'1e numbers at eU Ag Bi Cd In T1 each locality (30) (0.2) (5) (0.2) (1) (1) (1) (5) (1) (0.2) (0.2) (0.2) 262 ---------------------- N L 20 L 55 15 30 55 35 0.7 L L 276 ---------------------- N .2 10 .4 50 15 45 55 30 L L L 444 ---------------------- 40 .4 10 .4 150 50 85 50 630 n.d. n.d. n.d. 1060 --------------------- N .4 10 .2 70 15 100 20 270 n.d. n.d. n.d. 2464, 2468, 2469, 2473--— N .4 15 .2 60 25 90 70 90 L L L 2485, 2487 --------------- N .5 15 .4 60 40 65 35 120 n.d. n.d. n.d. 2488 --------------------- N .6 10 .2 75 35 85 35 150 n.d. n.d. n.d. 2489, 2491, 2492, 2494--- N 1 15 .4 65 50 175 40 90 L L .3 2496, 2502 N .3 15 .2 65 50 55 20 85 L L L 2501 --------------------- N .4 10 .4 65 70 85 110 100 n.d. n.d n.d 2503, 2504 --------------- N 4 10 4 55 50 50 30 100 n.d. n.d. n.d. 2509 --------------------- N 6 1O 4 40 15 25 3O 65 1.5 L L 2515, 2517 --------------- N 6 10 2 55 25 60 50 270 n.d. n.d n.d 2528, 2538, 2543, 2549, 2567, 2568, 2570, 2571, 2575, 2587 --------------- 40 .3 10 .3 70 25 100 40 120 L L .3 2556, 2557, 2558, 2559. 2560, 2561, 2562, 2564--— N L 10 .4 65 10 100 35 160 .2 L L 2636, 2637 --------------- N .2 10 .4 30 10 50 30 35 L L L 2639, 2640- ------------- 70 .2 10 .2 50 10 50 30 90 n.d. n.d. n.d. 2657, 2661, 2663 --------- N .2 5 .2 30 10 45 15 45 n.d. n.d. n.d. 2665, 2667, 2668, 2669, 2672 --------------------- 60 .3 10 .3 30 10 45 15 45 L L .2 2673 --------------------- N .4 15 .4 65 15 60 25 85 n.d. n.d. n.d. 2678 --------------------- N L 5 L 50 L 55 10 140 n.d. n.d. n.d. 2690, 2693, 2695, 2696, 2698, 2712 --------------- N .6 10 .3 45 4O 55 25 9O n.d. n.d. n.d. 2722 --------------------- 160 L 10 .4 20 15 35 40 30 L L L 2726, 2727 --------------- N .2 10 .4 65 10 150 15 45 n.d n.d. n.d. 2729, 2730, 2732 --------- N .2 10 .4 60 10 120 15 70 L .2 L 2733 --------------------- N .4 10 L 55 10 50 15 30 n.d. n.d. n.d. 2753 --------------------- N L 15 L 70 10 180 20 85 n.d. n.d. n.d. 2780 --------------------- N L 10 L 30 10 45 3O 50 n.d. n.d. n.d. 2785 --------------------- 50 .4 5 .4 30 15 50 40 140 n.d. n.d. n.d. 2797 --------------------- 50 .2 15 .4 25 5 20 15 75 n.d. n.d. n.d. 2799, 2802, 2804, 2805, 2806, 2807, 2808, 2810--- N .2 10 .3 25 5 40 20 45 L L L 2811 --------------------- N .2 10 .4 55 10 5O 25 80 n.d. n.d. n.d. 2814 ..................... N ,2 10 .4 20 5 35 25 30 n.d. n.d. n.d. 2816, 2818, 2820 --------- N .4 10 L 40 25 40 25 50 n.d. n.d. n.d. andesitic and basaltic rocks (figs. 21, 22, 25, and 26). If the arithmetic means of copper, lead, zinc, and cad- mium in all 85 magnetic concentrates are estimated (Miesch, 1963) from the geometric mean and geometric deviation (table 8), it is seen that the estimated mean contents of the four elements are about twice the average abundances reported for these elements in granite (table 15). These high values are in part attributable to the 22 concentrates from andesitic and basaltic provenances that are included in the total. These 22 concentrates have averages of 38 ppm cop- per, 39 ppm lead, 133 ppm zinc, and 0.3 ppm cadmium. DISTRIBUTION OF THE ELEMENTS 53 TABLE 13.— Variation in chemical composition of multiple magnetic concentrates from five areas represented by single plotted localities in the Candle quadrangle, Alaska [Data are in parts per million. Entries shown in table 1 as not detected are here assigned a numerical value equal to one-third the lower limit of determination: eU = N = 10 ppm here. Entries shown in table 1 as detected below the limit of determination are here TABLE 14.—Relative standard deviations, in percent, for eight elements in five sets of samples from single plotted localities in the Candle quadrangle, Alaska, compared to relative standard deviations for subsamples of file number3799 Sam 1e assigned numerical values equal to one-half the lower limit of determination: Ag = L segl Ag 81 Cd C0 Cu Ni Pb Zn =0.1;Bi=L=2.5_:Cd=L=0.1;Cu=L=0.5ppl:nhere] File number eU Ag Bi Cd Co Cu Ni Pb Zn A 75 51 92 14 120 32 34 34 A. Locality at 65°28‘45" 11.; 151°11'00" v. a 36 32 46 7.6 71 34 32 13 10 0.8 20 o 8 75 so 45 35 55 C 57 34 65 ll 72 3 l5 4‘ 120 .4 10 8 70 30 100 55 120 10 .1 1o 1 75 15 100 40 120 D 30 19 42 23 5] 35 31 4o 50 .1 20 1 60 20 100 45 100 10 .2 10 4 65 15 160 60 160 E 52 32 60 9 7 31 14 33 39 10 2 10 l 75 5 100 15 110 100 2 1o 4 65 o 5 120 33 138 Subsamples 70 4 2.5 1 75 0 5 85 6 2 . 10 2 5 1 70 5 85 45 160 of 3799 130 37.6 25.0 6.4 20.1 7.8 3.6 7.1 10 2 10 2 45 90 65 30 80 Arithmetic mean 40.0 .28 10.7 .31 67.5 24.1 96.0 42.0 126.5 :Letters ”fer to sample SEtS listed in Table 13- Standard deviation--- 42.7 .21 5.5 .28 9.5 29.2 30.8 14 2 42 8 Data from Table 5' Relative standard deviation (percent) 107 75 51 92 14 120 32 34 34 40 0.1 10 0.4 65 10 100 25 150 10 .2 .5 .2 65 5 85 30 180 10 .2 10 .4 60 0.5 70 35 130 10 .1 10 .8 75 10 85 35 160 40 .l 10 .2 70 5 100 30 180 10 .2 5 .4 65 10 100 25 160 10 .1 10 .4 65 5 80 40 180 10 .1 5 .4 60 20 180 60 130 Arithmetic mean 17.5 .14 8.1 .4 65.6 8.2 100.0 35.0 158.8 Standard deviation--- 13.9 .05 2.6 .18 4.9 5.8 34.1 11 3 21 0 Relative standard deviation (percent) 79 as 32 46 7.5 71 34 32 13 C. Locality at 65°44‘15" N.; 161°18'00" w. 0.6 5 0.4 30 5 40 15 35 .4 10 .1 35 10 50 15 30 .2 10 6 30 5 45 10 75 .2 10 4 30 5 45 15 50 .1 5 l 25 20 45 15 35 Arithmetic mean 60 .3 8.0 .3 30.0 9 0 45.0 14 0 45.0 Standard deviation--- 33.9 .2 2.7 .2 3.5 5 5 3.5 2 2 18.4 Relative standard deviation (percent) 57 67 34 66 11 72 8 16 41 D. Locality at 65°41'00" N.; 161°23'00" w. 10 1.5 10 0.2 60 20 65 20 90 10 .6 10 .4 40 35 25 20 100 10 .2 10 .4 35 40 50 20 75 10 .4 10 .4 35 55 570‘ 35 75 10 .4 15 4 40 65 60 20 140 10 .4 10 1 50 15 25 35 35 Arithmetic mean 10 .58 10.8 32 43.3 38.3 55 25 85.8 Standard deviation--- 0 .47 2.0 .13 9.8 19.4 19.0 7.7 34.6 Relative standard deviation (percent) 0 80 19 42 23 51 35 31 30 E. Loca1ity at 65°44'00" 11.; 161°12'30” N. 40 0.4 5 0.4 30 5 45 30 75 10 .2 10 .6 25 5 30 15 75 10 2 10 1 25 5 4O 20 45 10 4 10 4 30 5 45 15 35 10 .2 5 .1 30 5 50 15 35 10 .1 5 .4 30 5 45 15 35 80 .1 10 .1 25 10 40 30 45 10 .2 10 .4 25 5 40 25 30 Arithmetic mean 22.5 .23 8.1 .31 27.5 5.6 41.9 20.6 46.9 Standard deviation--- 25.5 .11 2.6 .19 2.7 1.8 .9 6.8 18.1 Relative standard deviation (percent) 113 52 32 60 9,7 31 14 33 39 ‘Thought to be a reporting error from tramp iron; hence, arithmetic mean, standard deviation, and relative standard deviation calculated without this va1ue. These are greater than the world averages for lead, zinc, and cadmium in basalt (table 15), but are less than the world average for copper in basalt. In addi- tion, the 50 magnetic concentrates from the granitic rocks in the Candle quadrangle have averages for these elements that are about 50 percent greater than the world average for granitic rocks; they contain 16 ppm copper, 29 ppm lead, 78 ppm zinc, and 0.3 ppm cad- mium. Thus, the magnetic concentrates appear to be slight accumulators of these metals on the basis of this general data. However, in the Darby and Kachauik plutons in the Bendeleben and Solomon quadrangles, which are similar to those in the Candle quadrangle, the geometric mean values for copper (5—13 ppm) and lead (53—72 ppm) in sediments from stream—s draining Hie granitic rocks suggest that these plutons are depleted in copper and enriched in lead compared to the average granite (Miller and Grybeck, 1973, table 4). Thus, the magnetic concentrates may accumulate copper, but their high lead content may simply reflect the composi- tion of the source plutons. Equivalent data are lacking for zinc and cadmium. In the area of the Granite Moun- tain pluton in the Candle quadrangle, copper and lead in stream sediments derived from the granitic rocks are about 70 and 100 ppm, respectively (Elliott and Miller, 1969). If these figures are at all representa- tive of the copper and lead content in the rocks of the Granite Mountain pluton, then the magnetic con- centrates are not accumulators of these metals. How- ever, Elliott and Miller (1969) carefully stated that their collections of stream sediments were made in areas of known mineralization with the result that average values from the reported results may be biased upward. A comparison shows that the 50 magnetic concen- trates from streams that drain granitic plutons have a 54 66°00’ R. EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 14 W. l3 _ x T “7 —i \“5' 7 . Ni ‘. 161°30’ 161°00’ 11 65645' 4 \ fie , .ZNH /( (3/ U l g l A3 V‘s/F// (K l l 5 \ \ \ mm WA ‘ Keith" :, @2756 K 15 KlLOMETEBS?° l l 1 MILES ' 10 EXPLANATION N1220 Locality of magnetic concentrate showing eU and Bi in parts per million. Top L—20 figure is equivalent uranium and bottom figure is bismuth; more than one figure indicates range in values where multiple samples taken from the same locality; right figure is average for equivalent uranium and left figure is average for bismuth, where two or more samples are represented. N=not detected; L=present but below the limit of determination FIGURE 23,—Map showing equivalent uranium and bismuth in magnetic concentrates from the northwestern part of the Candle quadrangle, Alaska. DISTRIBUTION OF THE ELEMENTS 55 151mm n_ mo w. 13 12 ,t W K , m7“ . w v WU I _ .. \\\ O ,5 , é b.w 06 \ x a 9 , \ \\ ’ 0 fl 0 1 . Ii; ‘~ ' > "1‘ \° V [D 1 ($3? a 1/3' 6., \f" 2. C56? 1;)“ I / ,V 'Cabin 0%;j‘tofi [15, 903- i /l ._ .° \hCa n5 on Q ( O /)’k\' g; .7 ,‘l ‘ fl” 4*,“ o ‘3 \p Z La" I r of) l \l “5“” _' x ( “cab » g J: 5235 «/ .i_'~ ‘WJgfgi’S/é \\ “I"; ,,- .3 / 65°15’ “ 0— .u H“ 15 KILOMETERS i 7 -~ I '3 v0 MILES ° N-1 20 Locality of magnetic concentrate showing eU and Bi in parts per million. Top 10 . 4O figure is equivalent uranium and bottom figure is bismuth; more than one figure indicates range in values where multiple samples taken from the same locality; right figure is average for equivalent uranium and left figure is average for bismuth, where two or more samples are represented. N=not detected; L=present but below the limit of determination FIGURE 24.—Map showing equivalent uranium and bismuth in magnetic concentrates from the southwestern part of the Candle quadrangle, Alaska. 56 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 66°00' R14W, x‘l51°30' 13 29151511130" ll 5 RJOWY. / T. . < \l. ;;3 dmflf/ L‘\J i Sm % 6 5 65°45’ 4 é - g ’5‘? eq/ ‘ \- I‘ we I ' l‘ m "V WSW/J WM T. 0 5 10 15 KlLOMETERS?‘ 2 -: I I I I 1 I I -)?~c‘ N.*0 V 5 V H _ 10 MILES ‘) 06/7. \W W (J Wfiéflk (5&3; EXPLANATION 15 Locality of magnetic concentrate showing Cu, Pb, Zn, and Cd in parts per 80 2.50'4 million. Top figure is copper; bottom is lead; left is zinc; right is cadmium; average values used at sites having multiple samples; L=present but below the limit of determination FIGURE 25,—Map showing copper, lead, zinc, and cadmium in magnetic concentrates from the northwestern part of the Candle quadrangle, Alaska. m4 w. 161°30’ ‘7 DISTRIBUTION OF THE ELEMENTS 2N7-i N? , ‘1 V mug};- ., ,7 ‘KW 12 “31°00: 11 / l W ° V Vsu M24 37 l , l 7" 4, 34:1»i‘gn.‘ \ I, 19 v V ‘Eflfili‘a‘ vb fut; CK / (14‘, ‘W\ “Lt . I ' I. i J M _ aw D "‘1‘ . \ l '«4 I / \0 a .9 ‘ d(/‘J / 65°30’ I * V , ‘” ' o “Carr ‘1 N {J4 "\L’\ LQES‘QO ()Gi» @SVJ f 5-: 2%?” ~ ” 0 0%?” ' R Q 55/ g‘ T. g \ 3 'M“ K /J/( 65°15’ 15 80 O 0.3 25 Locality of magnetic concentrate showing Cu, Pb, Zn, and Cd in parts per million. Top figure is copper, bottom is lead; left is zinc; right is cadmium; average values used at sites having multiple samples; L=present but below the limit of determination FIGURE 26.—Map showing copper. lead, zinc, and cadmium in magnetic concentrates from the southwestern part of the Candle quadrangle, Alaska. 58 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA TABLE 15.—Average abundances of copper, lead, zinc, and cadmium in the crustal materials of the Earth compared with their abundances in magnetic concentrates from the Candle quadrangle, Alaska [Data are in parts per million] Average abundance Abundance in Candle quadrangle (as shown in Krauskopf, 1967) Element magnetic concentrates Crust Granite Base t Geometric Estimated mean arithmetic mean Copper 55 10 100 l5 23 Lead 12.5 20 5 27 32 Zinc 70 40 100 79 95 Cadmium 0.2 0.2 0.2 0.38 0.46 lower percentage of anomalous values than the 22 concentrates from provenances of andesite and basalt: Percent of anomalous values in concentrates Element Granitic Andesitic and provenance basaltic provenance Copper 20 68 Lead 8 4 Zinc 18 27 Cadmium 0 0 The presence of two principal sources for the magnetic concentrates provides a possible explanation for the multiple population values (figs. 2—3) for copper (three populations) and lead (two populations). Con- centrates from the granitic provenance yield lower mean values for copper, lead, and zinc than the con- centrates from andesitic and basaltic provenances. However, the kind of bedrock does not appear to be a factor in the single populations found for zinc and cadmium (figs. 4—5). The magnetic concentrates containing the largest amounts of copper, lead, and zinc in the Candle quadrangle are from the Bear Creek area and the Granite Mountain pluton. High values for copper are particularly common along Bear Creek, where normal samples of stream sediments are reported to be anomalously rich in lead, zinc, and copper, and where galena, sphalerite, and pyrite have been found in andesite (Miller and Elliott, 1969, p. 14). Only one anomalous value for lead and one for zinc are noted among the nine magnetic concentrates that had anom- alous copper content from Bear Creek (table 1). Greatly anomalous amounts of lead and zinc are seen in the magnetic concentrates from the north-central part of the Granite Mountain pluton (table 1 and fig. 26). The area along Bear Creek that yielded the copper- rich samples is known to have sulfidebearing quartz- calcite veins and disseminated galena, sphalerite, and pyrite in andesite (Herreid, 1965; Berg and Cobb, 1967, p. 114); and it has been a source for placer gold (Miller and Elliott, 1969, p. 14). Lesser anomalies for copper and zinc were found among the magnetic concentrates from the southern part of the Hunter Creek pluton (fig. 25), but concen- trates from the northern part of the pluton are lean in base metals. In the andesites south of the Granite Mountain pluton two magnetic concentrates with anomalous zinc content are shown on figure 26. One concentrate (1060), which lacks anomalous copper and lead content, is from Sweepstakes Creek. The other (444), from Dime Creek, has highly anomalous amounts of zinc and cop- per, and high background amounts of lead (table 1). Gold placers are reported upstream from both the Sweepstakes Creek and Dime Creek localities (Cobb, 1972u). Thus, further exploration for base metals could be productive in the Bear Creek area, the northern part of the Granite Mountain pluton, the southern part of the Hunter Creek pluton, and the area around Sweep- stakes Creek and Dime Creek. The areal distribution of cadmium-bearing magnetic concentrates in the Candle quadrangle is irregular (figs. 25 and 26), and none of the concentrates is anomalous in cadmium (table 1). Most, but not all, of the samples with high values for cadmium are anoma- lously rich in zinc. Possibly the cadmium is present in inclusions of sphalerite in the magnetite, or in ac- cessory sphalerite in the concentrate. SII.\/lElR 18111) (}()I.I) All 85 magnetic concentrates from the Candle quad- rangle were analyzed for silver, but only 25 were analyzed for gold (table 1). The distribution of silver shows one population in the Candle quadrangle (fig. 6). The geometric means of silver and gold are lower for the quadrangle than for the region, and the geometric deviations for these elements in the quadrangle are much lower than those for the region (table 8). Using 1 ppm as the anomalous value for silver and for gold (table 9), only two concentrates are anomalous for silver and and one for gold (table 1; figs. 27 and 28). The concentrates with anomalous silver content are from Bear Creek (2492) and the southwestern part of the Hunter Creek pluton (2690). These are the same areas that yielded concentrates containing anomalous amounts of base metals (figs. 25 and 26). Gold is detected in only four of the 25 magnetic con- centrates analyzed and gold content is anomalous in only one, sample 2509 from the confluence of Eagle Creek with Bear Creek (fig. 27). Several gold placers DISTRIBUTION OF THE ELEMENTS 59 are situated downstream from the site of 2509 (Cobb, 1972u). BISMUTH Bismuth has a nearly normal distribution in magnetic concentrates from the Candle quadrangle; its mean value is 10 ppm (table 8). Thecumulative fre— quency curve for bismuth in magnetic concentrates from the quadrangle coincides with the regional curve (fig. 8) except in the high-value and low-value branches, showing that the magnetic concentrates from the Candle area are neither extremely rich nor extremely lean in bismuth compared to the region as a whole. The threshold anomalous value for bismuth was taken as 15 ppm (table 9). Low anomalous values of 15—20 ppm bismuth are found in magnetic concen- trates from drainage basins underlain by both granitic rocks and volcanic rocks (figs. 23 and 24). In general, samples from the northern part of the Granite Moun- tain pluton (2464, 2468, 2469, 2528, and 2549) tend to be slightly richer in bismuth than samples from the Hunter Creek pluton (2673, 2698, 2753, and 2797) or from the andesite and basalt (262, 2487, 2491, 2492, 2494, and 2502). The Granite Mountain pluton and the volcanic rocks around Bear Creek are thus weakly mineralized with bismuth as well as the base metals. COBALT AND NICKEL Cobalt in magnetic concentrates from the Candle quadrangle shows two populations and negative skew- ness (figs. 9, 19). Nickel has only one population and an indistinct positive skewness (figs. 10, 20). The geometric means of both elements (table 8) in samples from the Candle quadrangle are higher than those for the region as a whole or for the Solomon quadrangle; however, neither element in the samples from the Can- dle quadrangle reaches the extreme anomalous values found for some magnetic concentrates from elsewhere in Alaska (table 1). Threshold anomalous values of 90 ppm cobalt and 170 ppm nickel were set for magnetic concentrates from the Candle quadrangle (table 9). The Co/N i ratio, derived from the geometric means (table 8), is 0.73, somewhat lower than the regional ratio of 0.88 and notably less than the ratio of 1.55 for magnetic concentrates from the Solomon quadrangle. The decreased ratio in the Candle quadrangle results from the high mean value of nickel. As noted above in the regional results section, this high nickel content suggests sources in mafic rather than silicic rocks. However, the distribution shown in figures 29 and 30 seems to conflict with this interpretation. Three categories of rock units are shown in table 16, representing the major sources for the magnetic con- centrates, and the Co/ Ni ratios for the concentrates are listed by provenance. The arithmetic mean values for the Co/Ni ratios in magnetic concentrates from basaltic and andesitic source areas, 0.95 and 1.05, respectively, both resemble the ratio of 1.1 reported for magnetite from mafic magmatic rocks in northern Sweden (Frietsch, 1970, p. 95). The Co/Ni ratio for magnetic concentrates from granitic source areas in the Candle quadrangle is only 0.66. From the general considerations previously given, that the concentra- tion of cobalt varies less with magmatic differentiation than the concentration of nickel, and that nickel tends to be depleted in silicic rocks, it would seem reasonable to expect a larger Co/Ni ratio in magnetic concentrates derived from granite than in those derived from basalt, even if they are not part of the same differentiation sequence. However, the data belie the assumption. The two populations recognized for cobalt (fig. 9) ap- parently reflect the high mean values in the magnetic concentrates from basaltic and andesitic provenances and the low mean value for samples from granitic sources (table 16). The narrow differences found for the mean values of nickel from these three sources ac- counts for the single population recognized in figure 10. The negative skewness in the cumulative frequency of cobalt and the slight positive skewness in the cumulative frequency of nickel in figures 9 and 10 reflect the low values for cobalt and high values for nickel in table 16. The areas where high values are found for cobalt and nickel in these magnetic concentrates are roughly coin- cident, reflecting the common association of the two elements in nature. However, the highest values for the individual elements are not in the same sample. In fact, the sample (444) that has the richest cobalt con- tent (150 ppm) contains only 85 ppm nickel, which is just half the threshold value for nickel (table 9). The sample comes from Dime Creek, which drains an area underlain by andesite. Furthermore, the highest acceptable value for nickel (440 ppm) is in a sample (2491) that contains only 50 ppm cobalt. This sample comes from a provenance of basalt and andesite. Sam- ple 2696 showed greater nickel content (570 ppm), but this value probably reflects the presence of tramp iron. The sample '(444) from Dime Creek is the only one that has anomalous cobalt content (table 1 and figs. 29—30), but values between background (55 ppm) and threshold (90 ppm) are numerous, as would be expected from the means. Anomalous nickel content is present in three other magnetic concentrates: 2564, 2727, and 2753 (table 1). All are from areas underlain by granitic rocks: 2564 is from the central part of the Granite Mountain pluton, and the others are from the northern extension of the Hunter Creek pluton. 60 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA 161°00’ 11 65°00, _ “‘161 30' .Z\‘.—l : 233. I 65°45} 4 > )2 (W 10 Wm? E -} . I if K1IL0METERS\ '= fl ,}.‘ Ni] WV WV 5 V W H 10 MILES _ We" \V‘K‘Ssu/ m (J Wffifllfl {W7 0.6(15) L 0.2 O 0.3 EXPLANATION ocality of magnetic concentrate showing Ag, Au, In, and TI in parts per million.Top figure is silver,averaged where multiple samples were analyzed and showing in parentheses a single anomalous value among a multiple of analyses; bottom figure is gold; left figure is indium; right figure is thallium; lack of figure or letter indicates not determined; L=present but below the limit of determination FIGURE 27.——Map showing silver, gold. indium, and thallium in magnetic concentrates from the northwestern part of the Candle quadrangle, Alaska. DISTRIBUTION OF THE ELEMENTS \ 61 151°00' 66:» ' /’ < I 03 2C” ' .zwri 65°30’ ‘ “w L" w 1 m“ ; ///( f 65°15’ _V ., 7‘1 ’1 «“h- ...1 15 KILOMETEHS l y 7) T - £1:in :1‘ 4‘ | A H .\ '1, 5 V 10 MILES “ f ‘ _ -\ \\ w EXPLANATION 0-6(.1 -5) Locality of magnetic concentrate showing Ag, Au, In, and T1 in partsper 0‘2 L 0'3 million.Top figure is silver,averaged where multiple samples were analyzed and showing in parentheses a single anomalous value among a multiple of analyses; bottom figure is gold; left figure is indium; right figure is thallium; lack of figure or letter indicates not determined; L=present but below the limit of determination FIGURE 28.—Map showing silver, gold, indium, and thallium in magnetic concentrates from the southwestern part of the Candle quadrangle, Alaska. 62 ' EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA 66°m’ ‘7 R14 W. ”161 30’ 161°00’ 11 Rio W. .ZVT“ -‘ ’r‘: 255. gig, i? l ’ uc , _ _:~ . 65°45’ \/ A 1 f2; 4\/1/OL~A 5%A.Efs%7\rf\\¥i\ I l 10 MILES : ' V m \‘I _, > \ mm (J mews» K 15 KILOMEI'ERS 37f”: Wet 1 o“ V V ’1 $7 > ) EXPLANATION 55-75 Locality of magnetic concentrate showing Co and Ni in parts per million. Top figure is cobalt and bottom figure is nickel; more than one figure indicates range in values where multiple samples taken from the same locality; right figure isaverage for cobalt and left figure is average for nickel, 150 O 65 100-210 where two or more samples are represented FIGURE 29.—Map showing cobalt. and nickel in magnetic concentrates from the northwestern part of the Candle quadrangle, Alaska. DISTRIBUTION OF THE ELEMENTS 63 12 O I \J ¥ f“ 15A: . «A f? : ty‘ ‘1- . ”a J ”giA - R.14 W. 151°30’ Lee M X 87,» . 9R7? , v" I’ 2N3-l 0 : wanes; " 65°30’ ‘ ‘ \. c a . SQ? ' LG? {/‘i fiifiw {ZK/ “ {/35 a < 92> A a .2-.—« 55°15’ .‘ :7 «a» .., 2': l\ 15 KILUMETEHS \ I | L A‘ ' La 10 MILES o g o ‘i 5 “Fig 4.- wise! “:~1®>§1%§S§£K k (\\\/‘\/.KCEV\7;/L§//>r> {W ,gg" ’1‘ “/‘L‘ 45 'i c i“ n v ,/ 3 K . v , \r < / ,L/‘~ .43 1l5 KILDMETERS l 10 MILES EXPLANATION 1o Locality of magnetic concentrate showing Cu, Pb, Zn, and Cd in parts per 65 O 0.6 million. Top figure is copper; bottom is lead; left is zinc; right is cadmium; 50 average values used at sites having multiple samples; L=present but below the limit of detection FIGURE 36.—Map showing copper, lead. zinc, and cadmium in magnetic concentrates from the northeastern part of the Solomon quadrangle, Alaska. and is from a fault zone in the Devonian limestone about 3 km northeast of the gossan noted by Miller and Grybeck (1973, p. 39) to have anomalous zinc, lead, copper, and barium content. Virtually all the samples considered anomalous for zinc in the Solomon quadrangle contain less zinc than the threshold amounts used for Alaska as a whole (120 ppm) or for the Candle quadrangle (140 ppm). Anom- alous amounts of other metals are rarely associated with the concentrates having anomalous zinc content (table 1). Samples 2915, 2956, 3036, 3037, and 3048 have anomalous cobalt and (or) nickel content. Anomalous copper content is associated with the zinc in sample 2924, and anomalous lead content is present in 3048. Samples 2992, 3017, and 3032 contain anomalous amounts of equivalent uranium. Six of the magnetic concentrates (2915, 2923, 2924, 3036, 3037, and 3048) that have anomalous amounts of zinc are 76 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA 163°00’ 30’ 162°00’ 64°30' , E4? f \.,\_ m [Ex ,,—.,\\ / f/{J‘ \ \fl’i / / \ , \ , ,», / R (j), //— // \\ . , ' 4 I) - fl, ; J) in Q , \fl. [FAME V‘MK A75 & /I 4 ‘ ,4; . s‘,,/\/ 17;” ,x’ jo/Kéii & {7" ‘ “s i ‘ 16:33"; hm L6 \. 0 5 10 15 KILOMETERS l ' I . a 0 5 10 MILES EXPLANATION 10 Locality of magnetic concentrate showing Cu, Pb, Zn, and Cd in parts per 65 O 0.6 million. Top figure is copper; bottom is lead; left is zinc; right is cadmium; 50 average values used at sites having multiple samples; L=present but below the limit of detection FIGURE 37 .—Map showing copper, lead, zinc, and cadmium in magnetic concentrates from the east-central part of the Solomon quadrangle, Alaska. unassociated with described mineral occurrences (Cobb, 1972v), but the others are from streams which contain placer deposits of various minerals (West, 1953, p. 6—7): copper, rare earth, tin, tungsten, and radioactive minerals are reported in tributaries to Clear Creek (source of 2985, 2992, 3011, 3012, 3014, 3017, and 3025); and columbium, rare earths, and radioactive minerals have been found in tributaries to Vulcan Creek (source of 3032 and 3035). Only one magnetic concentrate (2978) from the Solomon quadrangle has an anomalous amount of cad- mium (5.5 ppm; fig. 36), and its cadmium content is the highest shown in table 1. The source of this sample was not known as a mineral prospect (Cobb, 197 2v), nor did the area yield <80-mesh fractions of stream sediments with detectable cadmium (Miller and Grybeck, 1973, p. 38). However, the lower limit of detection for cad- mium (20 ppm) in the <80-mesh fractions is too high to show low values such as those reported for the magnetic concentrates. Possibly the source of the high DISTRIBUTION OF THE ELEMENTS 77 cadmium in this single sample is a minor sulfide mineral included in the magnetite, or an accessory sulfide mineral trapped in the magnetic concentrates. The area seems unlikely to be a source for cadmium. Anomalous amounts of copper, lead, and zinc tend to be in the magnetic concentrates from the northern Darby Mountains, particularly around the quartz mon- zonite in the northern part of the Darby pluton. These anomalies may reflect a southern expression of mineralization noted in the Bendeleben quadrangle at the Omilak mine and the Foster prospect (Miller and Grybeck, 1973, fig. 1, p. 4). The highest anomalous values, however, are reported from the southern part of the area, and the strongest indications for anomalous contents of lead and zinc are from the fractured Devo- nian limestone and dolomite where they are intruded by the quartz monzonite and various dike swarms. Possibly the area of the Norton Bay Native Reserva- tion deserves further investigation because the strongest copper (2945) and zinc (2956) anomalies are found there. Anomalous amounts of copper and zinc are also present in the southern part of the reservation on the southeastern flank of the Kwiktalik Mountains (2838, 2923, and 2924), and to the north of the reserva- tion as far as Vulcan Creek, a tributary of the 'I‘ubutulik River (2985, 2992, 3011, 3012, 3014, 3017, 3025, 3028, 3032, 3035, 3036, and 3037). The area along the Kachauik River underlain by metamorphic rocks and granitic intrusives appears to be favorable for lead (2874). Anomalous amounts of lead also occur in the central part of the Darby pluton, where equivalent uranium content is also highly anomalous (2983 and 2995), but the area is probably less favorable than the limestone wallrocks to the east. SILVER AND GOLD Silver in magnetic concentrates from the Solomon quadrangle shows a single population of values with a cumulative frequency curve similar to that for the Candle quadrangle but notably different from the regional curve, especially in the high-value tail (fig. 6). Of the 101 samples from the Solomon quadrangle that were analyzed for silver, about 25 percent (table 1) con- tain less than the limit of determination for silver (0.2 ppm). The regional background for silver is 0.2 ppm, and the backgrounds for the Solomon and Candle quadrangles are a bit higher at 0.27 ppm, but the threshold value of 1 ppm for silver is the same for all three areas (table 9). Of the 41 magnetic concentrates from this quadrangle analyzed for gold (table 1), only three have gold content above the limit of determina- tion of 0.2 ppm. Three magnetic concentrates (2987, 3008, and 3024) from the Solomon quadrangle are anomalous for silver (table 1), and these are all from a small area on the east side of the Darby pluton (figs. 31, 32, 38 and 39). This area, underlain by quartz monzonite, is drained by the headwaters and western tributaries of Clear Creek. None of the silver-rich concentrates contains anomalous amounts of any other element, but their content of equivalent uranium is among the high background values for the Solomon quadrangle. The silver-rich magnetic concentrates at the head of Clear Creek (3024) and on a western tributary to Clear Creek (3008) are not associated with any reported pros- pect or mineral occurrence (Cobb, 1972v), but the con- centrate from another western tributary (2987), which also has anomalous silver content, is from a site reported to have detrital radioactive minerals and minerals containing rare earths, columbium, tin, and tungsten (West, 1953, p. 6—7). All three sites are at or near localities sampled by Miller and Grybeck (1973, fig. 1, p. 29) and found to have less than 0.5 ppm silver in the <80-mesh fraction of stream sediments. The anomalous values for silver in the magnetic concen- trates are not much above threshold; hence, it is in- ferred from these low anomalous values and from their sparsity that the area affords scant potential for silver. Gold is detected in only three of the 41 analyzed magnetic concentrates from the Solomon quadrangle (table 1), and one of these (2905) contains the low anomalous amount of 1.4 ppm gold (fig. 39). The source area for this sample (fig. 32) is the contact of the hybrid diorite with the monzonite and syenite units of the Kachauik pluton (Miller and others, 1972, fig. 1) south of Portage Creek. Samples of the <80-mesh fraction of stream sediment from this area were reported by Miller and Grybeck (197 3, p. 41) to have less than the lower limit of determination of gold (10 ppm), and no mineral occurrence or prospect is described in this area (Cobb, 1972v; Miller and Grybeck, 1973). The gold- bearing concentrate 2905 does not have anomalous silver content, but weakly anomalous amounts of cobalt and nickel were found in it (table 1). These factors, combined with the low values for gold reported by Miller and Grybeck (197 3, p. 41), are interpreted to mean that the gold in the magnetic concentrates is probably present as a fortuitous grain and that the area is not especially enriched in the element. BISMUTH Bismuth is present above the lower limit of deter- mination of 5 ppm in 100 of the 101 magnetic concen- trates from the Solomon quadrangle (table 1). By con- trast, bismuth was found to be below the limit of deter- mination in 9 percent of the samples in the regional 78 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA .//;7~\. < L /b‘ 64°45" Mi 162°00’ fi-°$S>../’v' 528$; Eels / «as? o ( U W l m» .r—V/rj‘; 0 ,- ‘, Masks Pam: Landmg Area 15 KILOMEI'ERS l | 5 10 MlLES EXPLANATION 1.6(6.5) Locality of magnetic concentrate showing Ag, Au, 1n, and T1 in parts per L O 0.3 million.Top figure is silver, averaged where multiple samples were analyzed 0-2 and showing in parentheses a single anomalous value among the multiple analyses; bottom figure is gold; left figure is indium; right figure is thallium; lack of figure or letter indicates not determined; L=present but below the limit of determination FIGURE 38.—Map showing silver. gold, indium, and thallium in magnetic concentrates from the northeastern part of the Solomon quadrangle, Alaska. results (table 8). In the Solomon quadrangle, the analytical values for bismuth are divided into two populations showing a positive skewness toward an ex- cess of high values (fig. 8). However, the maximum value observed for bismuth in the quadrangle, 40 ppm in file number 2995 (figs. 33 and 34), is considerably less than the highest value found for the region as a whole, 90 ppm in sample number 59 from the Ruby quadrangle (table 1). Nevertheless, the results of the analyses show that the Solomon quadrangle covers a part of the bismuth-enriched area of the Seward Penin- sula, and the geometric mean value for bismuth in magnetic concentrates from the Solomon quadrangle (1 1 ppm) is slightly greater than the regional geometric DISTRIBUTION OF THE ELEMENTS 79 163°00’ 30’ 162°00’ $7 64°30’ [I] 5 1? 1'5 KILOMETERS l I j 0 5 10 MILES EXPLANATION 0.4lll Locality of magnetic concentrate showing Ag, Au, In, and T1 in parts per L O 0.3 million.Top figure is silver, averaged where multiple samples were analyzed 1.4 and showing in parentheses a single anomalous value among the multiple analyses; bottom figure is gold; left figure is'indium; right figure is thallium; lack of figure or letter indicates not determined; L=present but below the limit of determination FIGURE 39.—Map showing silver, gold, indium. and thallium in magnetic concentrates from the east-central part of the Solomon quadrangle, Alaska. mean of 10 ppm (table 8). The threshold of 15 ppm for bismuth is also slightly greater in the Solomon quadrangle than for the region as a whole (table 9). Twenty-five concentrates contain anomalous amounts of bismuth (table 1), mainly confined at one reporting interval above the threshold of 15 ppm; one sample (2867) contains 25 ppm bismuth and one (2995) has 40 ppm. Equivalent uranium and nickel content are most commonly associated with bismuth, but in nine of the samples the anomalous bismuth content is unaccompanied by anomalous amounts of other elements. Most of the magnetic concentrates with anomalous ' bismuth content are from Clear Creek and its 80 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA tributaries where these streams drain the quartz mon- zonite of the Darby pluton (2975, 2978, 2990, 2993, 2994, 2995, 2997, 3004, 3005, 3006, 3009, 3010, 3019, 3021, 3023, and 3048; fig. 31) or the contact zone be- tween the quartz monzonite and the unit of Devonian limestone and dolomite (2983, 3001, 3014, 3016, and 3017; fig. 31). Several samples with anomalous bismuth content from the same general area are de- rived from areas underlain by fractured limestone (3034 and 3036). The only magnetic concentrates that have anomalous amounts of bismuth and lack de- tectable radioactivity are from the area in the southwestern part of figure 31, which is underlain by Precambrian quartz-mica schist (Miller and others, 1972, fig. 1). Many bismuth-rich samples from Clear Creek and its tributaries come from localities reported to have scheelite, cassiterite, radioactive columbium-bearing minerals, and detrital rare-earth minerals (2983, 2990, 2993, 2994, 2995, and 2997); or just the latter two types of minerals (2975 and 2978); or detrital rare- earth minerals (3014, 3016, 3017, 3019, and 3021) (West, 1953, p. 6—7). Bismuth-bearing minerals were not noted in the earlier literature (Cobb, 1972v). Bismuth anomalies exceeding 10,000 ppm were found by Miller and Grybeck (197 3, p. 4) in <80-mesh frac- tions of stream sediments from a strongly mineralized area at the north end of the Darby Mountains in the Bendeleben quadrangle, and the low-level anomalies in the magnetic concentrates from the Solomon quad- rangle may reflect a southern extension of the metallization. No mineralization has been reported in association with the bismuth anomalies in magnetic concentrates from the Precambrian quartz-mica schist in the southwestern part of the area covered by figure 31 (Cobb, 1972v), nor was bismuth detected in the <80-mesh fractions of stream sediments collected there by Miller and Grybeck (1973, p. 38). COBALT AND NICKEL Both cobalt and nickel show two populations and a positive skewness of values in the 101 magnetic con- centrates from the Solomon quadrangle (table 1, figs. 9 and 10), but their respective geometric means (31 ppm and 20 ppm) are less than those for the region as a whole (44 ppm and 50 ppm; table 8). Threshold values for cobalt and nickel in the samples from the Solomon quadrangle are much lower than threshold values for these elements in magnetic concentrates from the Candle quadrangle or from Alaska as a whole (table 9). Using these low threshold values of 40 ppm cobalt and 20 ppm nickel leads to the classification of 18 samples as anomalous for cobalt and 40 as anomalous for nickel (table 1). The presence of two populations of values for both cobalt and nickel, and the geographic distribution of the anomalous samples, fit closely the geologic features of the Solomon quadrangle (figs. 31, 32, 40, and 41). The low-concentration populations for each element come dominantly from source areas in the granitic rocks, particularly in the Darby pluton; and the high-concentration populations are mainly from source areas in the Precambrian metasedimentary and metamorphic rocks or the Devonian limestone and dolomite. Of the 18 magnetic concentrates that have anomalous cobalt content, nine are from areas underlain by Precambrian quartz-mica schist, metavolcanic rocks, schistose marble, metamorphic complex, and migmatitic rocks (2836, 2838, 2867, 2874, 2879, 2882, 2915, 2943, and 2944), and three are from sources in the Devonian limestone and dolomite (2945, 2956, and 3037). Contacts between the sedimen- tary rocks and the plutonic intrusive rocks are the source of two cobalt-rich samples (2887 and 3015). The gneissic monzonite, monzonite and syenite, and hybrid diorite units of the Kachauik pluton are the sources for four concentrates that have anomalous cobalt content (2904, 2905, 2910, and 2911). Twelve of the 40 concentrates containing anomalous amounts of nickel are from sources in Precambrian quartz-mica schist and metavolcanic rocks, schistose marble, metamorphic complex, and migmatitic rocks (2836, 2838, 2867, 2874, 2879, 2882, 2915, 2916, 2943, 2944, 2963, and 2970), and 12 are also from areas underlain by Devonian limestone and dolomite (2945, 2950, 2951, 2956, 2958, 2959, 2972, 3026, 3028, 3034, 3036, and 3037). Contacts between various sedimen- tary rocks and intrusive granitic rocks are sources for four magnetic concentrates with anomalous nickel con- tent (2887, 2980, 2998, and 3049). Hybrid diorite, mon- zonite, and syenite of the Kachauik pluton provide five anomalous samples (2904, 2905, 2909, 2910, and 2911), and the quartz monzonite of the Darby pluton is the source of seven (2928, 2936, 2937, 3005, 3006, 3019, and 3048). The common association of anomalous amounts of cobalt and nickel with such sedimentary and metasedi- mentary rocks as the Devonian limestone and dolomite, Precambrian schistose marble, and Precam- brian quartz-mica schist seems most likely to be attributable to sources for the magnetite in the numerous dikes and plugs of mafic volcanic rocks that intrude the sedimentary rocks (Miller and others, 1972, p. 4). DISTRIBUTION 64°45" \ ' ‘ i N . .1, ’ 7 ’, :. /,,,\/ RH?“ >\ H15 \J ”G” ‘ i ' , "a o >>>;/r/§ MM Q X} 79>. )w h RAT/4.2: .17 4' I “M. sfiQyth em,“ Fa c, K . , W (4 'V 925/ g Ly‘bk ‘ ' MR \00 ..... xijgng: ‘\ ~ x. 'x. r gr'w‘“\\,\:j 81 162°00’ OF THE ELEMENTS \f/ . {I /:/ V/I} >127? ‘1 23 5, ,‘ , @329) “i” J 0 g Q was; i 4 fiafl$§§ffix .31 L \tfl , k\\ ~4 995437 2: 5 1'0 15 KILOMETERS l l | 5 10 MILES EXPLANATION 25—30 Locality of magnetic conce 85 0 50 Top figure is cobalt and 55—120 indicates range in values locality; light figure is aver ntrate showing Co and Ni in parts per million. bottom figure is nickel; more than one figure where multiple samples taken from the same age for cobalt and left figure isaverage for nickel, where two or more samples are represented FIGURE 40.—Map showing cobalt and nickel in magnetic concentrates from the northeastern part of the Solomon quadrangle, Alaska. Samples that have anomalous cobalt content tend also to contain anomalous amounts of nickel (table 1). The three samples with the most cobalt, 2882 and 2887 from the area of Cheenik Creek (fig. 41) and 2956 from the Kwiniuk River area in the Norton Bay Native Reservation (fig. 40), are also the richest in nickel. All three samples are located outside the areas of the granitic plutons. The plutons were not recognized as source areas for <80-mesh stream sediments anomalously rich in cobalt, but nickel was more than ordinarily abundant in that sample medium at those sites (Miller and Grybeck, 1973, p. 38—39). The magnetic concentrate from the Kwiniuk River area was taken from a sample that had previously been reported to contain copper- and tungsten-bearing minerals (West, 1953, p. 6), but the results of the 82 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA 163°00’ WEE/“JO \lU <1, . , /( 64°30’ // // a} D l ,. \j \o(£ 6/ \/ \wfi W ,«\ V \\°j ‘. J/w' Q76; ," . u mate. #0 \C f/ [J ' ewé 52': /,.»—.,\\ v {/flt VSW/fi do all)” Ow Li’ff‘» \ L / .'. z .\ ‘ r/' /t W 30’ 162°00’ 0 5 1'0 15 KILOMHEHS | 'r l l | D 5 10 MlLES EXPLANATION 10—20 Locality of magnetic concentrate showing Co and Ni in parts per million. 20 O 15 Top figure is cobalt and bottom figure is nickel; more than one figure 20—280 indicates range in valueswhere multiple samples taken from the same locality; right figure is average for cobalt and left figure is average for nickel, where two or more samples are represented FIGURE 41.—Map showing cobalt and nickel in magnetic concentrates from the east-central part of the Solomon quadrangle, Alaska. chemical analyses show it to have only an anomalous zinc content associated with the cobalt and nickel. Most of the magnetic concentrates from localities in the Norton Bay Native Reservation contain anom- alous amounts of nickel, but only four have anomalous cobalt content, and only one each is associated with anomalous amounts of copper (2945; 30 ppm) and zinc (2956; 500 ppm). Most of the anomalous values are low. in contrast to those in samples from other areas in Alaska. Variations in the lithology of the source rocks from which magnetic concentrates are derived account for the low anomalous values of cobalt and nickel. A high value for lead, 110 ppm, is associated with anomalous cobalt and nickel values in sample 2874 from the Kachauik River. Anomalous amounts of cobalt and nickel are also found in the magnetic con- SOURCES OF THE AN OMALOUS ELEMENTS 83 centrates from the Tubutulik River area (fig. 40), where copper and zinc values are also anomalous. INDIUM AND THALLIUM Analyses for indium and thallium were made on 41 of the magnetic concentrates from the Solomon quad- rangle with a lower detection limit of 0.2 ppm for each element. Only five samples were found to contain 0.2 ppm or more indium, and 19 samples had 0.2 ppm or more thallium (table 1, figs. 38, 39). The geometric means for indium and thallium, 0.22 ppm and 0.28 ppm, respectively, are extremely close to the regional geometric means (table 8). Several geologically different source areas yielded ‘ magnetic concentrates with measurable thallium and indium, but the principal sources are in the Darby pluton and the Devonian limestone and dolomite, or along contacts between these units (figs, 31, 32, 38, and 39). The values for indium reflect no essential difference for source, but the high values for thallium, including the greatest found (1 ppm in sample 3005), are identified with sources in the quartz monzonite of the Darby pluton. Except for equivalent uranium and bismuth, other metals are seldom associated in anom- alous amounts with indium and thallium. The presence of indium and thallium in the magnetic concentrates has no apparent relation to known metallic mineral deposits in the Solomon quadrangle, although the distribution of these elements conforms broadly to the more favorable sites recognized for copper and zinc. SOURCES OF THE ANOMALOUS ELEMENTS The results of the radiometric and chemical analyses of the 347 magnetic concentrates from Alaska reveal that about 70 percent of these samples contain anomalous amounts of equivalent uranium, silver, bismuth, cadmium, cobalt, copper, nickel, lead, or zinc (table 1), either singly or in varied combinations. The principal constituent mineral in the analyzed concen- trates is magnetite, but other minerals are present. If the probable source or sources of the anomalous elements in the magnetic concentrates can be estab- lished with some degree of confidence, then the results of selective analyses of this kind of concentrate could be used more effectively to evaluate geochemically the mineral potential of a given area. Accordingly, the mineralogical composition of the magnetic concen- trates was investigated. The mineralogical composition of 67 of the 347 analyzed magnetic concentrates was semiquantita- tively determined by Keith Robinson using optical and X-ray diffraction techniques. The 67 samples were selected on the basis of their chemical characteristics to be a representative subsample of the 347 magnetic concentrates. With a few deliberately chosen excep- tions, such as sample 3779 which contains an unusu- ally high tenor in copper, the distribution of the elements is similar to that in the whole population. Of the 67 samples, 80 percent contain anomalous amounts of one or more of the elements for which the whole 347 were analyzed, and 20 percent contain only back- ground amounts of these elements. Another factor evaluated in the mineralogical ex- amination was the solubility of the various minerals composing the magnetic concentrate in the dissolution procedure used to prepare the sample for analysis by atomic absorption. If one or more minor minerals in- cluded in the grains of magnetite, or one or more accessory minerals trapped among the grains of magnetite in the concentrate, were found to be insolu- ble in the acid digestion, then they could not have been sources for minor metals reported in the results of the analyses. For this evaluation, a mineralogical study was made of the insoluble residues left from the acid digestion and recovered by filtering the leachate. MINERALOGICAL COMPOSITION OF MAGNETIC CONCENTRATES The presence in magnetite of trace amounts of elements such as those discussed here has been substantiated by previous research and is well documented (table 6). It is also generally recognized, as described above, that the trace elements may be chemically hosted in the magnetite itself, or they may be mechanically hosted through their presence in trace minerals or accessory minerals. The procedures used in the present mineralogical evaluation deal mainly with the accessory minerals that are present in the magnetic fraction of the panned concentrates. The practical consequence of preparation is that small and variable quantities of other minerals— even some nonmagnetic minerals—will be associated with the grains of the magnetic aggregate as trapped intergranular particles. The role of these particles in contributing to anomalous values for the elements needs evaluation. The mineralogical composition and anomalous ele- ment content in the 67 magnetic concentrates chosen for this examination are identified in table 21. All but six of the concentrates contain more than 50 percent magnetite, and 39 of the concentrates contain 90—99 percent magnetite. The six samples with less than 50 percent magnetite (56, 2121, 2418, 2696, 2785, and 3646) are diluted by ilmenite, rutile, sulfide minerals, EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS. ALASKA 84 .3 .5 959: 359$: N-vI $5.8 I- n. - II .- m-m I. II II I- ma -II-I:N J: .3 3595 3C 33 momw .3832. 385.5 .. uwiicmuwcz Mo 3?; I- I. I- -I II -I II .fi - .- wmA ......... 8 .m< ..... III ......... III on mmpm 3:35: N w EEG .- a I. II II n. 3 CT; I- II mm III-II-III-I-cN I. ...... I ............ on $5 3:33: my $58 a a - -I n. m I- II II om II-IIII-ICN .8 -IIII-IIIIIIII ------ on $5 .3533: I: ....... I----III--- -- -I m -I I- -I m I- I- I- 8 ..... :N .3 .9: :83 32.52% 9.5 .Aimmmumcm 2:: $32.23 33:38 E 82; I; I- -I I- I- I- II -I m: -I 3 ----- :N .3 .3. 3:83 23:59: lma 3:33: .5 .nm :9: 3.85 ”3.5959: ----- -IIII-I----II-II L; n. -- n. I- n. n. n. I- n. cm. .2 .3 .8 .9: :o 25:8 9.5.5wa :3: .:N A: #2 I- ....... -III---II--I .- -I I- E: : : : I- I- 3. .3 .8 .5 .9: ....... $3.23 3:32: :2: ------- IIIIIII---II II q I. -I a m m II II mm -IIIII- III--8 -IIIIIIIIIIIIIIIIII..-on_ BB .58.; we 8?; I- min t I. I- : I- -I II mm ............. 5 -IIIII- ------ IIIIIIIIIo: m2: III-II --------------- -I m I; n_ I. n_ -I I- I. I. ma -------------- - I IIIIII III--3523: :38 $3 .3580 ESE mo -- m -I II II -I II t. I- II mm IIIIIIIII :N .8 39.... 5.; 35¢:qu :85 $3 .332? 38:5 3:33:32: mo 33.: II I- -I II II - -I I- II m mm 5 ..:z .8 .8 IIIII-IIIucmmmS :9: 9:3: Em: .3533: --I-I IIIIIIIIIIIIII II I- II II II -I S -I I- II om IIIII :N :5 .32 ’5me SC .23 £38 mm: .3533: £3.68: we 8?; II L; I- .I I- I- II I; I- .- mm IIIIIIIIIIIIIII :83 33.523 83 59:58 33.6%: 0.5952: ucm :9..— 9:2» .5 ._z 833:3: 39:6: ------ II-II-I-II-IIII L: L; -I II II II II II : 2 co .8 .8 .5 .3 2m: .8585 2.: 588 38 .3533: uwuczo: .8 -I -- -- I- I- -- -- I- E: mm m: ..... :N J: .3 :m: .8555 2C :8: EN: .8 I- I- I- I- -I -I -- L; -I I- mm -I-I-- IIIIIIIII IIII- IIIIIII 3529:: :38 N8 . .3595? .8 II .fi II II - II II -I II I; mm .32 .8 .8 .5: 355m SC :23 £85 :mmm .mZLwEE 382.5 ECG—:32: .3 3?; -I II II II II -I 07 II -I II cm .I IIIIII I IIIIIIIII I-I-II 3.55sz :38 mmm .2955: 382.3 vwwtucwzwca ucm goo: 3:82; .333 .mmmE 9.28:; we $3.5ng 9:3 5.; .59.? we 30?; I- I. II II II II t. I. I. I- ma IIIII :N .8 .3 35959: cum—o 33.523. va .3»? new .30va C2 IEEC $3328 .353 58...; 59:8 .333ng .32m use £9: 9:95 .323 me $02.: I; L; II t. .- m om .- II I; om I-III-IIIE .8 M3.23:3: :33 33.523 :8N 3:3ng 2 .359: II II II II II t. 2 .fi II II mm IIIIIIIIIIIIIE 3333:: Egumgzm 6380 New IIIIIIIII III-III-I- .- II II II II I; mm II II -I m: II-I-I-II._z .8 {1:352:35 uczo: £38 N8 .353 mo. 39; II II -I - .F .- II -I II I- 3 -I-- IIIIIIII m< II-mtumcmme Egumzsm 535 map .338ng 2:... 31.8 :o 39:. II II -I - 3 II II II II - 3 III-II IIIIIII 352mg: :58 w: IIIIIIII I----I------- I- -- I- E: E: -- -- -- I- -I R. .I..... ------- 8:23: ESE: 58; 3w .5 A: .E .8 .Bom ES .5353: we 39:. .fi -- - II II I; t. a : mm .8 .8 .5 .3 33:33 3239: 2.23:8 :mm .332; :9: Lo 3?.» 3.69.8 mNIoN RE :9: :59; mo 3?; 63.38ng EB 3.7;; Om t. -I -I II II mN II II .C. ON ..:z .5 .5 .3. 39:58 33.5.5; 38 Em .mfiumcmfl: 2:625 IIIIIII -I-II-IIIIII-I II II LP t. II LP II II II II mm III-II-IIIIIIII .uwtOmIZwE .565 F 3.50:; 35:53» v.8: Eom 38 9.7525 $3.633 :9: 3.5 3225.; 8883.2: .8385: $25 #5532 BE 353.5: .550 35oz 3.55 :38me .235 3:: 3.7826 3.255: 355w: 3:32: 9:9: .33: E 3:33: 383538 we 35.832 3.; 333:3ng 39:39:33 553858 $0.328.sz 3:265 3:23.: u I 5:32: N 8 n=.m$=qv:=na E gamma-HI m 3:8wa H nag-u mm.»— 6wuh -H .E. .miam 1.6.3286 .m.D £855.: seal .3 mwmbaau ~3Eoi852. uflmfivy EEK m3§§u§8 QEwEMGE Nb :3 333m 6 k0 :Qumwcmsg NGQNMSEmEE 3:: aefiuoo ~82: .QQEESQRVIAN mqmfib £35 SOURCESOFTHEANOMALOUSELEMENTS .epem ecu .mszgosnm uwumcmme .cexw manna we oucwmwgn mg» on mcwze —mELe:am mm cm m—nmu cw ewmwm—e meemgpcmecee e_uw=ma2H .wumuqu vcm .wpvsumrme .wuwszum we mwemsu mucuugwe m .muwgzeeu—mgu .=eung we weagh .ucwmwgn ceuLwN ecu awcgmm Jacwugwe o— .wuwgwp_mmmo .:eeLwN we wemgu wucwmwgn :mmm—me mucaugme m .m—Tuam .mewcwuzmegom we mung» II IIIIIII uhfIIIIIII .ucwesme m— .mr_u:x .ucwmwge wwumumpwm uw_m_u:mew:= mucmugwe op .e—wuzm .ucwegwa om .mpwuzx .uceugwa m seeusz IIIIIIII .uceugwn m .m—wuzm .Luau::_u we wemgp mID m— o— mN mA op op LP LP LP «um I r)r)n Nam BIN LP Lb LP om LP mm mm om om ow mm« mm oq mmA ma ovv om omA omA wmA mm lullulluulzu -u-----------=u -----=N .=e .m< ----.----=N .39 -. ..... --.---=u -----nm .ae .m< .CN .wz .ge .wm .m< .zm -u--- ...... - :u --- -----=u ---.--- ...... 2w --.---..-----=w ---------wm .am ---- ..... :N .Dw ---------wm .aw ....... --wz .vm ------ ...... -zm ....... --=N .zm -----ue .wm .=w ---- ..... F2 .00 .........,z .09 ---------wz .OU ---------_z ,oe ---------wz .oe ..... a; .wz .oe ------.--_z .se -----=e .wm .m< ---.-----=N .ae --- ----=N .39 -nm ._z .=e .m< .m:_umee :u_= mu_um:mas um:_mgm mcww >ge> -.----u-kuueu mw muvumcmmz .ucmmmge «gm m_mgm:_E LeggOe ”emumee mw muwumcmmz -u----..'---ou_uo:mme cum—u lunulvoo .mcmumee we mange a saw: wumuw:ume cum—u .ucmmmga mmF=stem ew__mums ucm :eLw gauge ”muwgwcmma umucsog ppm: .eecwugm wcwm .-.--------------.----en -.-.-- ........... -----eo .--- ........ - ......... eo wp_um:mms =e mcwumeu ocegum n-----mcwuuee mm; wuwgwcmmz . ........... mpwuwcmce cumpu .mcwquU we wows» saw: muwuwzmme cum—e >Fw>wumpwm -n-----..--.muwuwcmms cum—u .mcwumeu we wumgu ;u_3 eumuw=mme :umpu z_w>_pm_em - ...... -----wu_um:mme :mmpu ----.---.--------.-.ee .mmumuwrvm ecm .m_wH=L .wuwcme_w u:mu:=n< ------m:wpmeu mu: wuwumcwmz .m:_umou ugmw_m mm; .muwuwcmme mgse >Fw>wum_mm ------m=_umoe ma; wuwuwcmmz -----.-.-mmumuv—ww ucmvcza< mumuwcmcs cam—u x—w>wym_wz --..----.----.-.------ee .uceumou ugmw_m guys muvuwzmme cum—u x~m>_um_mm ---.-- ..... -n---.-----oa .acwumou “saw—m ;u_z wuwumcmme maze a—m>wum_wz .:es_ gauge we wemgu amuwumcmms :uer 2—w>_um_wm .wm_=gw;em uw__muws muem “cegw eEmLu apzwm: meme mmnm mumm wumm wonm mwwm .ovom memm mmmm Nmom —mom opom Hpom OFOM woom ommw Nmmw meow vemw meow nmww waN mNmN «NwN —_wm momm mmnm mmNN mNNN omow “new mwmm momm Navm wwvm wmqw HmFVN 86 gold, quartz, hematitic coatings, a diverse array of silicate minerals, metallic spherules, and tramp iron. Many of those samples that contain more than 50 per- cent of magnetite also have these diluents to a lesser degree. MAGNETITE, ILMENITE, AND RUTILE N o magnetic concentrate was found to be 100 per- cent magnetite, although some are very nearly pure (553, 682, 1026, 2192, 2488, 2874, 2879, 2882, 2887, and 3689). In the samples that have the highest per- centages of magnetite, the magnetite itself tends to be very clean and free from hematitic stains and coating. Ilmenite and rutile generally are present as complex crystallographic intergrowths within the grains of detrital magnetite. Magnetite was hand-picked from nine concentrates and analyzed by E. L. Mosier to compare the composi- tion of magnetite to that of the magnetic concentrates. The results are given in table 22 and are discussed in later sections that describe the frequency of associa- tion of elements with specific minerals. EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA HEMATITE Hematite forms surface coatings on detrital grains of magnetite and quartz. It also serves as a cementing agent to bind other grains, including nonmagnetic grains, to magnetite, as noted below. Hematitic coatings are readily removed from the magnetite by ultrasonic cleaning or acid digestion. SULFIDE MINERALS AND GOLD Pyrite, marcasite, and Cinnabar are the principal sulfide minerals identified in the concentrates (table 21). Chalcopyrite is sufficiently abundant in one sample (3779) to give a large copper anomaly and prob- ably to be the source for anomalous silver and zinc con- tent. The sulfide minerals other than marcasite and Cinnabar are associated with the magnetite as crystalline intergrowths. Pyrite, marcasite, Cinnabar, and Chalcopyrite are attached to the magnetite by secondary cementing agents such as hematite. Native gold is found as discrete particles, or as TABLE 22.—Minor elements in hand-picked magnetite from Alaskan placers [Semiquantitative spectrographic analyses of hand-picked magnetite by E.L. Mosier, U.S. Geological Survey, November 13. 1972; data on magnetic concentrates from table 1. All data are in parts per million. n.d. = not determined. N = not detected at lower limits of determination, which for the analyses of the hand-picked magnetite, are Bi, 2ppm; Cd, 2 ppm, Co, 10 ppm; Ni. 10 ppm; Zn, 100 ppm; Au. 5 ppm; In, 2 ppm; T1, 5 ppm; and Sn. 10 ppm] File number Ag Bi Cd Co Cu Ni Pb Zn Au In T1 Sn Hand—picked magnetite 59 1,000 30 N 100 70 100 5,000 150 w1,000 N N 500 928 5 N N 30 20 N 50 1,000 N N N 10 929 10 5 N 20 20 15 50 1,000 10 N N 10 1455 1.5 N N 200 10 1,000 10 500 N N N 50 1831 1 N N 50 7 700 50 2,000 N N N 200 1867 1.5 N N 50 10 300 100 10,000 N N N 300 2121 1.5 N N N 300 30 7 N N N N 30 2148 1 N N N <1 10 10 150 N N N N 3646 0.2 10 N 70 20 200 300 300 N N N 5,000 Magnetic concentrate 59 340 90 2.5 1,000 470 830 4,700 220 N n.d. n.d. n.d. 928 .4 10 .4 65 190 280 35 700 4.9 0.5 <0.2 n.d. 929 68 20 .6 75 70 350 50 1,300 n.d. n.d. n.d. n.d. 1455 .8 5 .4 190 60 970 10 170 135 <.2 <.2 n.d. 1831 1 20 .6 130 25 820 85 930 n.d. n.d. n.d. n.d. 1867 43 10 1.5 70 230 600 530 140 n.d. n.d. n.d. n.d. 2121 33 10 .5 45 220 100 45 180 n.d. n.d. n.d. n.d. 2148 1 5 .4 65 2,000 110 55 230 n.d. n.d. n.d. n.d. 3646 18 70 .2 80 220 300 45 150 640 .2 <.2 n.d. SOURCES OF THE ANOMALOUS ELEMENTS 87 grains cemented to the magnetite by secondary minerals. In one sample the fragments of detrital gold are coated by very fine grained particles of magnetite. Native gold embedded in magnetite, as reported by Eakin (1914, p. 28), is not observed in this group of 67 samples, unless the gold particles coated with magnetite may be so considered. The minor metals in Alaskan placer gold have been reviewed by Mertie (1940b), and one sample (56) of detrital gold in the magnetic concentrates from Solomon Creek in the Ruby quadrangle is analyzed in the present work (table 23). TABLE 23.—Minor elements in a particle of placer gold from Solomon Gulch in. the Ruby quadrangle, Alaska [Laser probe analysis of sample 56 by J.M. Nishi, US. Geological Survey, November 1, 1972] Significant trace elements Minor trace elements Major elements Au Si Ba, Mg, Mn, Pb, Fe, A9 V, Cu, Ti, Ca. QUARTZ AND COMMON SILICATE MINERALS Quartz and common silicate minerals such as chlorite, mica, amphibole, and feldspar are present, and occasionally unexpectedly abundant, in the magnetic concentrates. Neither quartz nor feldspar would be expected, but hematitic coatings and in- tergrowths with other minerals have caused a feeble magnetism, and some particles have been trapped among clots of magnetite grains. Both of these cir- cumstances account for the persistence of these non- magnetic minerals into the magnetic concentrate. Some grains of chlorite, micas (mainly phlogopite and biotite), and amphiboles (commonly tremolite) find their way into the magnetic concentrate largely as the result of entrapment with the magnetite or because of magnetic inclusions. Where hematitic coatings on quartz and magnetite are common, many of the mag- netite grains are cemented to grains of quartz or to grains of the common silicate minerals. Crystalline in- tergrowths of magnetite with quartz and the common silicate minerals are abundant in some concentrates. OTHER MINERALS A number of other accessory minerals are in the magnetic concentrates, including: iron sulfate, garnet, zircon, spinel, scheelite, epidote, hopeite, cuprite, anatase, serpentinite, celsian, cassiterite, azurite, malachite, and calcite. Several unidentified silicate minerals were noted, and volcanic glass, slate, and serpentinite were found in one sample each. As with the common silicate minerals, various intergrowths of these other minerals with magnetite partly account for their presence in the magnetic concentrate, but cemen- tation to magnetite by hematite or other agents and mechanical entrapment among grains of magnetite also are important factors in their presence. There is no practicable or feasible method for the separation of the quartz, common silicate minerals, and other minerals from the detrital magnetite, because of the common occurrence of intergrown grains or cemented particles. The iron sulfate in sample 56 appears to be a second- ary mineral phase produced by the oxidation of pyrite and marcasite. It may have formed in air after the con- centrate was panned. Garnet and zircon are the most common of the other silicate minerals observed in the magnetic concen- trates. Garnet is present in five and zircon in eight samples (table 21). Unidentified silicate minerals are reported in three concentrates, but the remainder of the other minerals, and the volcanic glass, slate, and serpentinite are each recorded only once. Thus, spinel, scheelite, and epidote are in sample 293; cuprite and anatase are in sample 2121; and azurite, malachite, and calcite are in sample 3779. Hopeite, a hydrated zinc orthophosphate mineral, is tentatively identified as a trace amount in sample 1026 from the Mount McKinley quadrangle, but the mag- netic concentrate is not enriched in zinc (table 1). In- deed, this sample lacks anomalous amounts of any metal, and the equivalent nonmagnetic concentrate, where hopeite might be expected to be concentrated, has less than 200 ppm zinc (Hamilton and others, 1974). This casts doubt upon the quantity present and upon the identification of hopeite itself. The mineralogical identifications of scheelite in sample 293, cuprite in sample 2121, celsian (barium feldspar) in sample 3072, cassiterite in sample 3646, and azurite and malachite in sample 3779 are chemi- cally confirmed through the presence of tungsten, barium, and tin, respectively reported for samples 293, 3072, and 3646 by Hamilton and others (1974), and through the anomalous copper content in samples 2121 and 3779 (table 21). Nickel, a common minor element in ultramafic rocks, is anomalous in sample 2887 (table 21), in which accessory serpentinite was identified. METALLIC SPHERULES AND TRAMP IRON Metallic spherules are present in samples 59, 928, 929, 1831, 2121, 2418, and 3646, and tramp iron is in samples 56, 59, 293, 553, 928, 929, 1455, 1867, 2418, 2438, and 3646. Both the spherules and the tramp iron adhere to grains of magnetite by ferromagnetic attrac- 88 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA tion, are cemented to the magnetite as particulate en- crustations by secondary cementing agents such as hematite or limonite, or occur as loose intergranular particles. The metallic spherules are smooth to scoriaceous, light- to dark-colored metallic .particles. The tramp iron is in the form of small metallic slivers evidently derived from the blades and tracks of earth- moving equipment, steel sluice boxes, drill bits, and other tools. Metallic spherules of various sorts have long been noted in heavy-mineral concentrates from many sources. These spherules have been variously iden- tified as industrial byproducts such as fly ash and welding beads or spatter (Handy and Davidson, 1953; US. Army Corps of Engineers, 1961; Fredriksson and Martin, 1963; Charles Milton, written commun., 1972); as fusion products formed by natural processes such as volcanic activity (Fredriksson and Martin, 1963), lightning, and forest fires (Overstreet and others, 1963); or as extraterrestrial material such as meteoric dust, ablation products from iron meteorites, or tektites (Crozier, 1960, 1961, 1962; Finkelman, 1972; Fredriksson and Gowdy, 1963; Kaye and Mrose, 1965; Langway and Marvin, 1964; Schidlowski and Bitz- kowski, 1972; Skolnick, 1961; and Thiel and Schmidt, 1961). Much interest attaches to the spherules and tramp iron for their contributions to the minor-element geochemistry of the magnetic concentrates. The origin of the tramp iron is reasonably apparent, but that of the magnetic spherules affords some room for further research and may not be attributable to a single proc- ess for all spherules. More attention to the general distribution of such spherules in Alaskan .surficial materials is needed. The scope of the present investigation did not permit resolving with certainty the most probable origin of these Alaskan metallic spherules. The presence of all the observed spherules in concentrates from placer mines (tables 1 and 24), and the association of some of the spherules with tramp iron (tables 21 and 24), are tentatively interpreted to indicate that the spherules probably originated through activities related to placer mining, and that they are welding beads. However, the chemical composition of the metallic spherules casts some doubt on a single-source hypothesis and leaves open the possibility of extrater- restrial origin. Samples 2121 and 3646 both contain many metallic spherules, which were hand picked under a binocular microscope for analysis. Spherules from sample 2121 are quite clean and bright, whereas those from magnetic concentrate 3646 are dull and rusty. Six subsamples of spherules were prepared from sample 2121 and given the numbers 2121a—2121f. Five were picked from sample 3646 and given the numbers 3646a—3646e. Laser-probe analyses of these metallic spherules were made by J. M. Nishi, US. Geological Survey, and the results are given in table 25. Many of the elements commonly associated with particulate matter from welding (Brown and others, 1972, table 5) are lacking in these metallic spherules. However, detailed microscopic and chemical analyses are needed to determine if the spherules are manmade or are of ex- traterrestrial origin. TABLE 24.—Sources of magnetic concentrates containing metallic spherules and tramp iron, Alaska File Fig. No. Metallic particles No. Quadrangle Source in Cobb, present (x) or absent (---) 1973 Spherules Tramp iron 56 Ruby ——————— Placer at mouth of Solomon Creek ——————————— 54 --- x 59 ———-do.---- Placer on Glen Gulch ——————————————————————— 54 x X 293 Mount Hayes Placer on Slate Creek —————————————————————— 8 -—— x 553 Hagemeister Squirrel Creek placer ---------------------- 15 —-- x Island.' 928 Bethel ————— Marvel Creek placer ———————————————————————— 12 x x 929 -———do.---- ----do. ------------------------------------ 12 x x 1455 Livengood—— Amy Creek, cleanup from Mr. Nells' placer-— 55 -—- x 1831 Iditarod--- Concentrate from Frank Salem Cut, Granite 49 x —-— Creek placer. 1867 ———-do.———- Concentrate from Riley Dredge on Otter Creek 49 --- x 2121 Bethel ----- Sluice concentrate from Marvel Creek placer- 12 x -—— 2418 Tanana ----- Sluice concentrate from Johnson and Johnson 47 x x placer on lower Rhode Island Creek. 2438 McCarthy--- Sluice box concentrate from Chititu Mines 9 —-- x placer on Rex Creek. 3646 Circle ----- Sluice box concentrate from H. C. Carstens 43 x x placer mine on Portage Creek. SOURCES OF THE ANOMALOUS ELEMENTS 89 TABLE 25.—Results of laser-probe analyses of hand-picked metallic spherules from placers in Alaska [Analyses by J .M. Nishi, US. Geological Survey. November 1, 1972; N = not detected at lower limit of determination. Numbers in parentheses below the element symbols show the lower limits of determination] Subsamples Data in percent Data in parts per million of 2121 Fe Ca Ti Cu Pb Mn Zr (0.05) (0.05) (0.001) (5) (20) (10) (20) a 10 0.1 0.01 70 N 1,000 N b 10 .05 .05 100 N 2,000 N c 2 N .07 15 N 1,000 N d 7 N .002 N N 100 N e 5 N .07 5 N 1,000 N f 3 N .01 N N 1,000 N Subsamples of 3646 a 10 .05 .015 7 30 500 N b 7 N .03 N N 2,000 70 c 20 N .01 20 N 500 N d 10 N .05 5 N 3,000 N e 15 N 007 15 N 300 N 1Elements that were looked for but not detected are listed here with their lower limits of determination: in percent, Mg, 0.02; in parts per million, A9, 0.5; As, 100; Au, 5; B, 5; Ba, 5; Be, 0.5; Bi, 20; Cd, 100; C0, 20; Cr, 5; La, 100; Mo, 2; Nb, 10; Ni, 10; Sb, 50; Sc, 20; Sn, 10; Sr, 10; V, 10; w, 50; Y, 50; and Zn, 50. Tramp iron was hand picked from magnetic concen- trate 2418, divided into three parts, and the parts were analyzed spectrographically by E. L. Mosier, US. Geological Survey (table 26). The iron, chromium, and manganese contents are within ranges of values that would be expected from steels used in machinery, structural elements, and hand tools around placer mines. MINERALOGICAL COMPOSITION OF INSOLUBLE RESIDUES Chemical digestion of the magnetic concentrates in preparation for analysis by atomic absorption was not entirely complete. The mineralogical composition of TABLE 26.—Results of semiquantitative spectrographic analyses of tramp iron from a placer concentrate from lower Rhode Island Creek, Tanana quadrangle, Alaska‘ [Analyses by E.L. Mosier, US Geological Survey. November 13. 1972; G = greater than value shown; L = present, but below limits of determination. Numbers in parentheses below the element symbols show the lower limits of determination] Sub- Data in percent Data in parts per million samples Fe Ti Co Cu Ni Pb Cr Mn Sn Y of 2418 (0.05) (0.002) (5) (5) (5) (10) (10) (10) (10) (10) a (320 L 100 200 300 10 300 1,500 20 20 b 620 L 70 300 1,000 L 500 1,500 20 15 c (320 0.002 30 200 300 10 300 1,500 15 15 1Elements that were looked for but not detected are listed here with their lower limits of determination: in percent, Ca, 0.05; Mg, 0.02; in parts per million, A9, 0.5; As, 200; Au, 10', B, 10; Ba, 20; Be, 1; Bi, 10; Cd, 20‘, La, 20; Mo, 5; Nb, 10', Sb, 100', Sc, 5', Sr, 100', V, 10; H, 50; Zn, 200; and Zr, 10. the residues of the 67 samples was examined in order to determine what minerals and other components of the magnetic concentrates were taken into solution and thereby contributed to the measured content of minor elements, and to determine what components were resistant to digestion and are unlikely to have contributed minor elements. This study showed that the magnetite is largely, but not completely, digested. Ilmenite and rutile, although commonly slightly leached on the surface, are essentially unaf- fected by the chemical treatment. Such leaching as is present may, in part, be attributed to the solution of magnetite intergrown with the ilmenite or rutile. Hematite coatings, sulfide minerals, and gold were taken completely into solution. No effects could be detected on the quartz, common silicate minerals, and other minerals, except that the quartz and silicates at- tained a high gloss indicative of the digestion of various surface coatings. Such coatings themselves have been found by other investigators to be sources for trace elements (Chao, 1972; Goni, 1966). Carbonate minerals were dissolved. Generally, the metallic spherules and tramp iron were taken completely into solution. Thus the significant result of the chemical digestion of the magnetic concentrates is that magnetite, hematite, sulfide minerals, native gold, car- bonate minerals, metallic spherules, surface coatings, and tramp iron are mainly taken into solution and the other minerals are not. Two samples (56 and 293) that had particles of native gold left residues containing more than 1 per- cent silver chloride crystals, formed artificially as a precipitate from solution. However, the analyses of the solutions themselves disclosed far less than 1 percent silver. The solution from sample 56 had 600 ppm silver and that from sample 293 had only 0.4 ppm silver (table 1). Silver chloride crystals from sample 293 were analyzed spectrographically by E. L. Mosier, US. Geological Survey, who reported major silver, minor gold, and traces of silicon, iron, magnesium, manga- nese, titanium, copper, and mercury. FREQUENCY OF ASSOCIATION OF ANOMALOUS AMOUNTS OF ELEMENTS WITH SPECIFIC MINERALS To facilitate the isolation of the probable host minerals or materials for the anomalous concentra- tions of metals found in the 67 magnetic concentrates, the frequency of association of anomalous elements with specific minerals was determined (table 27) from the data on the 67 magnetic concentrates in table 21. Table 27 shows, for example, that all nine magnetic concentrates containing anomalous amounts of equiv- 90 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA TABLE 27.—Frequency of association of specific minerals with anomalous element content in magnetic concentrates from Alaska, in percent [Data on minerals and anomalous elements from table 21] Elements Number Minerals or material present in of Magnetite Tramp Metallic Hematitic Ilmenite Rutile Pyrite Gold anomalous samples iron spherules coating and amounts marcasite eU 9 100 ll 11 100 55 0 0 11 Ag 15 100 40 40 47 60 7 7 4O Bi 12 100 42 33 67 50 17 8 33 Cd 4 100 50 25 75 100 25 0 50 Co 13 100 38 23 23 38 8 8 23 Cu 30 100 33 23 50 47 3 3 23 Ni 22 100 45 27 27 45 4 9 31 Pb 8 100 50 38 38 63 12 12 50 Zn 23 100 30 26 52 61 9 0 22 None 13 100 0 0 46 46 23 8 0 Chlorite Mica Amphiboles Garnet Feldspar Quartz Zircon Other minerals eU 9 55 66 11 ll 78 100 ll 22 Ag l5 l3 13 20 13 13 53 7 80 Bi 12 50 25 25 8 42 75 8 50 Cd 4 50 25 50 O 25 50 25 75 Co 13 38 31 15 8 8 46 8 46 Cu 30 4O 10 17 10 2O 70 3 57 Ni 22 40 18 18 9 14 59 4 54 Pb 8 25 25 25 0 0 50 0 63 Zn 23 35 13 13 13 26 61 9 61 None 13 38 23 38 0 38 38 0 62 alent uranium also contain the mineralogical associa- tion of magnetite and quartz coated with hematite. This observation supports the data previously presented that hematitic coatings on grains in the magnetic concentrates are the main sources of radio- activity. Similarly, the 13 concentrates lacking anomalous metal content all lack tramp iron, metallic spherules, and gold. Table 27 can only be used, however, in connection with the mineralogical study that showed the virtual complete insolubility of ilmenite, rutile, chlorite, mica, amphiboles, garnet, feldspar, quartz, zircon, and other silicate minerals, and with the data in tables 22, 23, 25, and 26 showing the trace-element compositions of hand-picked detrital magnetite, native gold, metallic spherules, and tramp iron (table 28). Magnetite alone could be a sufficient source for anomalous amounts of silver, copper, nickel, lead, and zinc (table 28), but the presence of native gold would add to the values reported for silver, copper, and lead. Further additions to copper and lead would be con- tributed by the metallic spherules, and the presence of tramp iron would notably raise the values for cobalt, copper, and nickel in the magnetic concentrates. Undoubtedly these accessory minerals have added to the values reported for these elements, but the ac- cessories are present in only some, not all, of the anomalous concentrates. Thus, detrital native gold is found in 40 percent of the silver-rich concentrates, 23 percent of the cupriferous concentrates, and 50 percent of those with anomalous lead content. Metallic spherules are present in 23 percent of the cupriferous concentrates and 38 percent of those with anomalous lead content. Tramp iron is in 38 percent of the cobalt- rich magnetic concentrates, 33 percent of those with anomalous copper content, and 45 percent of those with anomalous nickel content. As previously stated, tramp iron, metallic spherules, and native gold are lacking from the 13 nonanomalous magnetic concen- trates in table 21. TABLE 28.—Regional threshold values for eight elements in Alaskan magnetic concentrates compared to possible source minerals [All data are in parts per million; -- indicates no data available] Regional Estimated mean values Element threshold Hand-picked Metallic Tramp Native value magnetite spherules iron gold (table 9) (table 21) (table 24) (table 25) (table 22) An 1 1.5 <0.5 <0.5 Major Bi 14 m2 <20 <10 -- Cd 1 <2 > > > < < > > Difference None None None None Significant Significant None None Hand-picked magnetite (table 22) Total (ppm) 1,038.5 50 8 525 457 2,350 5,567 15,000 No. samples 8 8 8 8 8 8 8 8 Mean (H ppm) 130 6 1 66 57 294 696 1,875 0% 129,904 100 0 3,844 10,000 134,689 3,034,564 11,175,649 02 352 10 0 62 100 367 1,742 3,343 Relation of hand-picked magnetite to normal magnetite concentrate 2 2 01 02 OD 'dN + N 127.43 3.58 0.1 22.16 35.46 132.83 616.24 1,183.22 — 1 2 2WD 254.86 7.16 0.2 44.32 70.92 265.66 1,232.48 2,366.44 N - 8 _ _ -129.6 3 —0.6 -15 —37 -188 -648 -l,724 ZOD> or > < > > > > > Difference None None Significant None None None None None Relation of hand-picked magnetite to abnormal magnetic concentrate 01 02 CD = N; + N—z 138.05 8.55 0.20 80.12 58.39 156.32 726.98 1,187.86 ZOD 276.1 17.1 0.4 160.24 116.78 312.64 1,453.96 2,375.72 A - 8 __ _ -28 18 -0.3 96 121 140 -220 —1,539 20D> or > > > > > > > Difference None Significant None None Significant None None None 94 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA tribute to the anomalous values for nickel in the magnetic concentrates, although nickel is known to be a common trace element in magnetite. This conclusion is supported by the results of the spectrographic analyses given in table 22, where all the values for nickel except one (sample 1455) are clearly less than those reported for the whole magnetic concentrate. EQUIVALENT URANIUM The data on the mineralogical sources for the radioactivity reported as equivalent uranium in the magnetic concentrates from Alaska are given in tables 10, 11, 19, 21, and 27, where it can be seen that the principal source is hematite, which forms coatings and cement on other minerals in the concentrate. The main minerals with hematitic coatings are magnetite and quartz. GOLD, INDIUM, AND THALLIUM Particulate gold, either as a minor mineral included in magnetite at the time of crystallization of the magnetite, or as an accessory mineral trapped in the magnetic concentrate, is thought to be the most prob- able source for the gold reported in the magnetic con- centrates (table 1) despite the very poor correlation between the presence of mineralogically identified native gold and the presence of chemically determined gold (tables 1 and 21). Of the 67 concentrates examined mineralogically (table 21), eight concentrates were found to contain native gold. Only two of these (293 and 3646) had been analyzed chemically, and both con- tained gold. On the other hand, 11 others among the 67 concentrates had chemically detectable gold, but this gold was not observed in the mineralogical study. How much of the chemically determined gold is actually particulate native gold is, therefore, uncertain. Indium is known to be geochemically associated with tin in minerals from cassiterite deposits on the Seward Peninsula, Alaska (Sainsbury, 1963; 1964; 1969). Of the 67 magnetic concentrates examined mineralogically, three samples (928, 2729, and 3646) have detectable indium; none contains indium-bearing magnetite (table 22), although one (3646) contains 10 percent cassiterite (table 21) and the other two are from tin-bearing areas (Hamilton and others, 1974). Evidently most of the indium in the cassiterite-rich sample can be attributed to the cassiterite, although the magnetite itself contains 5,000 ppm tin (table 22). In the magnetic concentrates 928 and 2729, which have 20 ppm and 50 ppm tin, respectively (Rosenblum and others, 1974), the indium is probably not in the magnetite, because magnetite from sample 928 lacks indium (table 22). Each of these magnetic concentrates is derived from sources known to contain some cassiterite. Thallium tends to be associated geochemically with zinc and lead, but among the 67 magnetic concentrates examined mineralogically, neither of the thallium- bearing samples (2978 and 3010) is enriched in zinc or lead (table 1, and Rosenblum and others, 1974). These two samples are from the Seward Peninsula, an area where thallium is generally present in small amounts in rocks and minerals associated with tin deposits (Sainsbury and others, 1968, p. F29), but tin is quite sparse in both the magnetic concentrates (Rosenblum and others, 1974) and in the nonmagnetic concentrates (Hamilton and others, 197 4). Accessory sulfide minerals with which the thallium might be associated were not seen in these magnetic concentrates (table 21). Low concentrations of thallium have been noted in endogenetic iron hydroxides (Vlasov, 1966, p. 521); thus, the thallium may be in the hematitic coatings on these samples. Slight support for this interpretation arises from the fact that both samples 2978 and 3010 have hematitic coatings. The coating on 3010 is described (table 20) as heavier than that on sample 2978, and 3010 has slightly more thallium (0.3 ppm) than 2978 (0.2 ppm). ROLE OF ANOMALOUS ENVIRONMENTS The preceding data insufficiently reflect the role of anomalous environments as contributors to the con- taminants that appear to increase the metal content of magnetic concentrates. Native gold is a contaminant of the magnetic concentrates from placer deposits, which are intrinsically anomalous environments, and it raises the local values for gold and silver in magnetic concentrates. The mining of placer gold results in the further contamination of magnetic concentrates by tramp iron, certainly, and by metallic spherules, possibly. The presence of these contaminants is, then, the result of an anomalous environment, and they tend to raise the values of bismuth, copper, and nickel in the magnetic concentrates above values that might have been obtained if the locality had not been mined. As a consequence, the opportunity for a spurious anomaly in bismuth, copper, or nickel exists, but spurious values can be identified from the mineralogy of the concentrate. Hematitic coatings and cement on grains of magne- tite and quartz are a common contaminant in the magnetic concentrates. These coatings are the princi- pal source for equivalent uranium and may be the source for thallium. However, environments having naturally anomalous radioactivity are necessary to RELATIONS AMONG THE ELEMENTS 95 produce hematitic coatings enriched in equivalent uranium. Thus, the hematitic coatings in the magnetic concentrate indicate whether or not an anomalous en- vironment existed. Magnetite itself is the probable source for anom- alous values in cobalt, nickel, copper, and zinc in the magnetic concentrates. Consequently, environments anomalous in these elements could be directly iden- tified from the magnetic concentrates without need of the additional values for contaminants. RELATIONS AMONG THE ELEMENTS CORRELATION S Geochemically coherent elements tend strongly to be found together, and in polymetallic mineral deposits there is generally a positive correlation of geochemi- cally coherent elements. That is, a sample enriched in one element tends to be enriched also in associated coherent elements. The degree of dependency is usu- ally measured by correlation coefficients which show a value of +1 for a perfect direct correlation, a value of —1 for a perfect inverse correlation, and a value of 0 for no correlation at all. Values between +1 and -1 in- dicate degree of direct (positive) or inverse (negative) relations. Because the correlation coefficient is a measure of the degree of association between two ele- ments, a positive correlation coefficient may be used to indicate the value of one element as a pathfinder for less readily detected elements in geochemical exploration. REGIONAL The correlation coefficients between the logarithms of the concentrations of the elements in the magnetic concentrates from Alaska are given in table 31. Those values shown as “L” (below the limit of determination), “N” (not detected), or “---” (not deter- mined) in table 1 are not included in the computations for table 31; thus, the correlation coefficients are only approximate. Gold, indium, and thallium pairs repre- sent censored data, and only a few pairs are available for treatment, thus severely weakening the signif- icance of the correlations. The dependency of the pairs of elements can be classed as very significant positive correlation at the 99 percent confidence level and significant positive cor- relation of the 95 percent confidence level. These levels are a function of the number of element pairs. An ex- amination of the correlation coefficients in table 31 shows that certain elements tend to be associated in the magnetic concentrates from Alaska: Very significant positive correlation: Cu—Pb, Cu—Zn, Cu—Ag, Cu—Co, Cu—Ni, Cu—Au; Pb—Zn, Pb—Ag, Pb—Cd, Pb-Co, Pb-Bi, Pb—Ni, Pb—eU; Zn—Ag, Zn—Co, Zn—Ni, Zn—In; Ag—Co, Ag—Bi, Ag—Ni; Cd—Bi; Co—Ni; Bi—Au, Bi—Eu. Significant positive correlation: Pb—Tl, Ag—Au, Ni—Au. Cobalt and nickel have a very significant positive correlation coefficient of 0.75, showing the strong positive relation between concentrations of cobalt and nickel in magnetic concentrates. These two elements are typically geochemically coherent. Their strong association in the concentrates probably indicates their substitution for Fe“2 in the magnetite lattice. Copper has very significant positive correlation co- efficients with silver, lead, zinc, cobalt, nickel and gold. Very significant positive correlation coefficients also exist between lead and silver and between zinc and cobalt. These associations suggest a relation between these elements based on their presence in minor sulfide minerals. Equivalent uranium has a very significant positive correlation with lead and bismuth. The source of most of the equivalent uranium in the magnetic concen- trates is hematitic coatings on other minerals. Doubtless some lead is also present in the exogenetic hematite derived from hydrous iron oxides, which are notorious scavengers of heavy metals (Jenne, 1968). Also of interest is the possible association of the original sources for the lead and the equivalent uranium. Lead tends to be enriched in acidic igneous rocks, as do the radioactive elements, and the bulk of the concentrates showing equivalent uranium are from streams draining granitic rocks which have anomalous lead content (Miller and Grybeck, 1973, p. 3). The high positive correlation shown in table 31 for equivalent uranium with gold (0.67) is not significant, as it is based on too few samples. The association of equivalent uranium and gold depicts a placer source for the samples with the magnetic concentrate con- taminated with gold. Zinc and indium have a very significant positive correlation coefficient (0.56), which may reflect the geochemical association of indium and zinc in sphalerite (Rankama and Sahama, 1950, p. 725). Sphalerite is probably a minor mineral included in the 96 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA TABLE 31.-Correlation coefficients of the logarithms of the concentrations of equivalent uranium and 11 elements, and number of pairs, in 34 7 magnetic concentrates from Alaska [Values on the left side show number of element pairs; those on the right are correlation coefficients] eU Ag 81' Cd Co Cu Ni Pb Zn Au In Tl eU eU 0.11 0.27 0.18 -0.24 -0.16 -0.40 0.46 -0.14 0.67 -0.74 0.29 Ag 98\Ag\.22 .10 .31 .54 .30 .36 .21 .43 .11 .27 Bi 120 247 B1'\.29 .03 -.02 -.01 .32 .08 .42 -.23 .36 Cd 94 213 225 Cd 0 —.09 -.06 .33 .01 .09 .39 -.23 Co 123 273 318 242\C0 .45 .75 .20 .41 .23 .38 -.08 Cu 103 247 289 223 316\Cu .39 .21 .43 .49 .28 .01 N1' 123 273 318 242 347 316\N1' .15 .28 .31 .37 -.10 Pb 123 273 318 242 347 316 347\Pb .20 .31 .23 .47 Zn 123 273 318 242 347 316 347 347\ Zn .31 .56 .31 Au ’3 29 35 26 40 38 40 40 40 \Au .04 —- In 8 17 19 12 20 18 20 20 20 10\In -— 11 22 18 23 15 23 20 23 23 23 0 3\ T1 magnetite, because it was not detected as an accessory mineral in the concentrates (table 21). Very significant to significant positive correlations are shown in table 31 for copper and gold (0.49), gold and silver (0.43), gold and bismuth (0.42), and lead and thallium (0.47). Many of these could be expected. Cop- per sulfide minerals may be present in source areas for placer gold, or copper may be alloyed with the native gold. Silver is a common alloy with the Alaskan native gold. The association of bismuth with gold seemingly reflects areas of complex sulfide ores and gold, and the bismuth may be in minor sulfide minerals in the mag- netite; it is probably in galena, although the correla- tion coefficient of bismuth with lead (0.32) is a little lower than that with gold. Bismuth is found in galena but is rarely present in sphalerite (Rankama and Sahama, 1950, p. 740), a condition reflected by the very low positive correlation coefficient found for bismuth with zinc (0.08; table 31). The significant cor- relation coefficient between thallium and lead, com- pared to the nonsignificant correlation between thallium and bismuth, lends a little support to the possibility that thallium is in minor inclusions of galena in magnetite. Cadmium displays no correlation (0.01) with zinc in table 31. This contradicts the well-known natural geochemical association of cadmium with zinc. The reason for this apparent contradiction seems to be analytical bias. Most of the magnetic concentrates con- tain 0.2—0.6 ppm cadmium. Owing to drift and fluctua- tion of the meter on the atomic absorption instrument, the values for cadmium in that range are imprecise. In the data in table 1, some high concentrations of cad- mium are found in zinc-rich samples: for example, file number 59 from the Ruby quadrangle, number 1867 from the Iditarod quadrangle, and number 1917 from the McGrath quadrangle. Strong significant negative correlations are shown in table 31 for equivalent uranium and indium (—0.74) and equivalent uranium and nickel (—0.40). Only eight pairs are represented for equivalent uranium and in- dium; thus, the value of the correlation coefficient may not be reliable. The negative correlation between equivalent uranium and nickel seems readily explained by the types of source rocks with which these two elements are associated. The radioactive magnetic con- centrates come from sources in granitic rocks which are lean in nickel. Magnetic concentrates enriched in nickel are derived from ultramafic rocks poor in radioactive elements. CANDLE QUADRANGLE Correlation coefficients were computed for equivalent uranium and 11 elements in the 85 magnetic concentrates from the Candle quadrangle. Table 32 shows the results of the correlation analysis. Gold, indium, and thallium are not discussed in this section because they each form no more than three cor- related pairs with the other elements. For the other ele- ment pairs, the degrees of correlation are again classed as very significant or significant positive correlation related to the number of element pairs. The observed positive correlations are: RELATIONS AMONG THE ELEMENTS 97 TABLE 32.-—Correlation coefficients of the logarithms of the concentrations of equivalent uranium and 11 elements, and numbers of pairs, in 85 magnetic concentrates from the Candle quadrangle, Alaska [Values on the left side show number of element pairs; those on the right are correlation coefficients] eU Ag Bi Cd Co Cu Ni Pb Zn Au In T] eU eU\-0.12 0.01 0.38 -0.30 0.13 —0.11 0.14 -0.45 —— —— —1.00 Ag 12 Ag\.23 —.27 .29 .50 .07 .29 .17 1.00 —— —— B1‘ 16 61 B1'\.25 .24 .40 .09 .16 0 .68 —- .50 Cd 13 47 55 Cd\-.05 -.03 -.03 -.03 -.25 .66 -— —- Co 17 63 80 56 Co .40 .53 .42 .65 -.95 -- 1.00 \ Cu 15 60 77 54 78 Cu\.27 .45 .32 .52 -- .84 Ni 17 63 80 56 82 78 N1'\.24 .41 -.83 —— .91 Pb 17 63 80 56 82 78 82 Pb\ . 38 -. 15 —— .93 Zn 17 63 80 56 82 78 82 82 Zn\-. 76 —— .99 Au 0 1 3 3 3 3 3 3 Au\-— -- In 0 1 1 1 1 1 1 1 0 In —— T1 2 1 2 2 2 2 2 2 0\T1 Very significant positive SOLOMON QUADRANGLE correlation: Cu—Pb, Cu—Zn, Cu—Ag, Cu—Co, Cu—Bi; Correlation coefficients of minor elements in 101 Pb-Zn, Pb—Co; magnetic concentrates from the Solomon quadrangle, Zn—Co, Zn—Ni; Alaska, are given in table 33. Again, the three C 0— Ni. elements gold, indium, and thallium are excluded from . . _ . . _ . the following discussion because of the few pairs Sigmflcant pos1t1ve Cu—Ni, Pb—Ag, .Pb'Nl’ represented in the data set for the Solomon quad- correlation: Ag—Co, CO‘BL rangle. Using the same two degrees of correlation as The pairs of elements with very significant positive correlations in magnetic concentrates from the Candle quadrangle are the same as for the regional data, ex- previously, it is seen that many pairs have significant positive correlation coefficients: Very significant positive cept that the copper-nickel pair has dropped from very correlation: Co—Ni, Cu—Ni, Pb—Zn, significant to significant positive correlation. A Pb—Bi, Pb—eU, Bi-eU. stronger copper-bismuth association is found in the Significant ositive Candle area than in the whole group, reflecting, correlatiofi: Pb—C d Zn—Co C d—Bi possibly, the presence of polymetallic sulfide deposits (Miller and Elliott, 1969) and, certainly, bias in the samples toward mineralized areas. Equivalent ura- nium shows negative correlations with cobalt, nickel, and zinc in the Candle quadrangle as well as in the region as a whole. The classical trace-element geochemistry of igneous rocks would explain the in- verse relation of the radioactive material with cobalt and nickel, but it is quite unexpected to find that zinc seemingly is not enriched in the hematitic coatings which provide the radioactivity. Inasmuch as the abundances of cadmium in the magnetic concentrates from the Candle quadrangle are in the range of values associated with instrumental noise, its negative cor- relations with most elements in table 32 are not reliable. The very high positive correlation of the pair cobalt- nickel, characteristic of the whole data set, is excel- lently shown in the magnetic concentrates from the Solomon quadrangle, where the correlation coefficient is 0.83 (table 33). These elements are reported (Miller and Grybeck, 1973, p. 6) to be enriched in stream sediments derived from plugs and dikes of diabase. Possibly the strong correlation coefficient for them in the concentrates also is a reflection of sources in mafic rocks. Equivalent uranium shows a consistent high positive correlation with lead (tables 31—33). The cor- relation coefficient for this pair reaches its greatest value, 0.54, in the Solomon quadrangle, from which the largest number of Alaskan radioactive magnetic con- 98 EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA TABLE 33.-Comlation coefficients of the logarithms of the concentrations of equivalent uranium and 11 elements, and numbers of pairs, in 101 magnetic concentrates from the Solomon quadrangle, Alaska [Values on the left side show number of element p'airs; those on the right are correlation coefficients] eU Ag B'l Cd CO Cu Ni Pb Zn All In Tl eU eU\0.lO 0.34 0.17 -0.22 -0.l5 -0.37 0.54 0.08 1.00 -0.79 0.22 Ag 64 A .05 -.22 .07 .Ol -.02 .08 .14 l.00 .Ol - 23 Bi 80 7:\Bl\.25 -.09 .06 -.22 .35 .07 -- -.90 .37 Cd 59 63 7l Cd .07 -.l6 0 .29 0 -- -- .23 \ C0 80 78 l0] 7] CO\.l8 .83 -.l6 .22 1.00 .25 .16 Cu 65 63 83 58 83 CU .27 -.ll -.04 -- -- .05 N1 80 78 lOl 7l 1m 83 N1\-.25 -08 .97 .05 .03 Pb 80 78 lm 7] l0] 83 lOl Pb\.22 -.92 .06 .39 Zn 80 78 1m 71 lOl 83 lOl l0] Zn\—.26 .18 .l7 Au 2 2 4 l 4 4 4 4 4 AU\-- -- In 5 5 5 3 5 3 5 5 5 1 In\-- Tl lg 15 l9 l3 l9 l6 l9 19 l9 0 2 T1 centrates were collected. Anomalous amounts of lead are present in stream sediments derived from granitic plutons and the contact zones of these plutons in areas of above-normal radioactivity (Miller and Grybeck, 1973, p. 5—6). The association is genetic, as the source for both the lead and the radioactive elements is granite, but the correlation is enhanced by exogenetic processes that have added radioactive hematitic coatings to grains of magnetite. Equivalent uranium also has a very significant positive correlation co- efficient with bismuth (table 33). This may be related to the association of bismuth with lead, indicated in table 33 by the very significant correlation coefficient of 0.35 between lead and bismuth, and to the associa- tion of bismuth with the metamorphic rocks that are adjacent to the granitic plutons, as shown by samples of stream sediments (Miller and Grybeck, 1973). The very significant negative correlation between equivalent uranium and nickel in magnetic concen- trates from the Solomon quadrangle (table 33) con- firms the relation brought out in the regional data (table 31), and reflects the geochemical differences be- tween the granitic source rocks of the radioactive elements and the mafic and ultramafic sources of the nickel. Copper in magnetic concentrates from the Solomon quadrangle is negatively correlated with lead and zinc (table 33). Magnetic concentrates from the Solomon quadrangle are notably deficient in copper compared to the regional average (table 8) and have similar to slightly lower means for lead and zinc. These charac- teristics of distribution are borne out by the negative correlation coefficient. However, copper has a very significant positive correlation (table 33) with nickel (0.27). Lead has very significant positive correlations with bismuth and zinc, but lead is negatively cor- related with cobalt and nickel (table 33). Similar relations for copper, lead, zinc, cobalt, and nickel are described for stream sediments from the Solomon quadrangle (Miller and Grybeck, 197 3, p. 5—6), and are related to source rocks. The significant association of copper with nickel is related to sources in diabase dikes and plugs, whereas the lead and zinc are related to sources in granitic plutons. However, in the data from the magnetic concentrates (table 33), zinc shows no significant correlation with nickel, but it has a significant positive correlation coefficient (0.22) with cobalt. Although the magnetic concentrates from both the Candle and Solomon quadrangles are partly derived from granitic rocks, the mean contents of copper and zinc (table 8), as well as the associations of copper, lead, and zinc (tables 32 and 33) are quite different in the two areas. This may indicate fundamental dif- ferences in the compositions of the granitic rocks in the two areas. The magnetic concentrates that have high values for equivalent uranium are derived from alkalic plutons. Alkalic plutons seem to be deficient in copper, and to yield magnetic concentrates that have no cor- relation or negative correlation between copper and lead and between copper and zinc. PROMINENT GEOCHEMICAL HIGHS The varied distribution of anomalous amounts of metals in magnetic concentrates from Alaska is shown RELATIONS AMONG THE ELEMENTS 99 in table 1. The table is not intended for use in defining areas of anomalous metal content, because for some quadrangles only a few concentrates have been ana- lyzed, and most of those samples were taken at known mineralized areas. For several quadrangles, only single concentrates have been analyzed, and they also are apt to be from known mineralized areas. The data in table 1 show that copper anomalies occur in 82 percent of the 33 quadrangles, and zinc anomalies occur in 70 percent; the other 10 elements reach anomalous concentrations in no more than half (17) of the quadrangles. Possibly the ease with which copper and zinc can substitute for Fe"2 in magnetite accounts for their more common occurrence in anomalous amounts in the magnetic concentrates. Their presence may also be related to the geochemistry of the source region. Gold, indium, and thallium are poorly repre- sented in table 1, partly because only 131 of the 347 magnetic concentrates were analyzed for these elements. Of the other nine elements, equivalent uranium, bismuth, and cadmium are the least com- monly anomalous and the least widespread. About half (52 percent) of the quadrangles yielded lead-rich magnetic concentrates. Lead does not readily replace iron in magnetite, but it may be associated with silver in accessory sulfide minerals and gold, as it has ap- proximately the same percent frequency of anomalous occurrences as silver. Cobalt and nickel can readily substitute for iron in magnetite, but despite this geochemical advantage, cobalt anomalies are present in only 42 percent of the quadrangles and nickel anomalies occur in 48 percent. However, these anomalies are more common in samples from mafic provenances, whereas the majority of the analyzed samples are from granitic provenances. Silver and gold contents also tend to be more frequently anomalous in magnetic concentrates from specific areas. Thus, cer- tain general areas in Alaska yield magnetic concen- trates that are characterized by particularly prominent geochemical highs. These localities are summarized below under the major regional divisions used in discussions of the distribution of the elements. COPPER AND SILVER IN SOUTHEASTERN ALASKA Notable anomalies for copper and silver are found in six of the nine samples collected in the Ketchikan quadrangle in southeastern Alaska. MULTIELEMENT HIGHS IN SOUTHERN ALASKA A number of prominent highs for various multi- element combinations of copper, zinc, silver, cobalt, and nickel are present in magnetic concentrates from quadrangles in southern Alaska. High values for copper, zinc, and gold are found in samples from the McCarthy quadrangle (table 1); copper and silver have anomalously high associated values in the Valdez quadrangle; zinc anomalies are present in concentrates from the Anchorage, Talkeetna, and Talkeetna Moun- tains quadrangles; and cobalt and nickel attain high values in samples from the Mount Hayes quadrangle. BASE METALS IN SOUTHWESTERN ALASKA Magnetic concentrates from the Bethel and Iliamna quadrangles in southwestern Alaska yield anoma- lously high values for the base metals (table 1). Promi- nent highs are found for copper and zinc, with associated silver and gold along Marvel Creek and Cripple Creek, which are tributaries to the Salmon River in the Bethel quadrangle. The northern shore of Iliamna Lake in the Iliamna quadrangle is the source of the most copper-rich sample listed in table 1. WEST-CENTRAL ALASKA EQUIVALENT URANIUM IN THE BENDELEBEN, CANDLE, AND SOLOMON QUADRANGLES The principal sources of radioactive magnetic con- centrates are in the Bendeleben, Candle, and Solomon quadrangles, although measurable equivalent uranium was detected in samples from several other areas (table 1). In the Bendeleben and Solomon quadrangles the concentrates have 40 to 560 ppm equivalent uranium, but most of the values are in the range from 120 to 140 ppm. In the Candle quadrangle the equivalent uranium ranges in value from 40 to 160 ppm with a mean of 65 ppm. The magnetic concentrates with the highest radioactivity are from tributaries to Clear Creek in the northeastern part of the Solomon quadrangle. The sources are alkalic granitic rocks of Middle Cretaceous to Late Cretaceous age (Miller and others, 1972, p. 5—7) having affinities with the alkalic rocks of the Candle quadrangle, which are the sources of somewhat less radioactive magnetic concentrates. The most radio- active concentrates from the Candle quadrangle are derived from the northwestern margin of the Granite Mountain pluton. Miller (1970) described this pluton as consisting of a core of equigranular quartz mon- zonite surrounded successively outward by massive to porphyritic monzonite, nepheline syenite, and garnet syenite. Samples from the core lack radioactivity or are only weakly radioactive. The most radioactive samples are from the outer wall of the pluton. Ac- cording to Miller (1970), CaO, MgO, FeO, and Fe203 in- crease outward in the pluton as SiO2 decreases. Possibly the increasing radioactivity of the magne- tites, which is parallel to this outward variation in the 100 composition of the pluton, is related to the changing calcium content of the pluton. LEAD, COBALT, BISMUTH, AND OTHER ELEMENTS Magnetic concentrate 59 from Glen Gulch in the Ruby quadrangle, west-central Alaska, yields the highest values for lead (4,700 ppm), cobalt (1,000 ppm), and bismuth (90 ppm) in this set of samples. Other metals in the same sample, including copper, zinc, cad- mium, nickel, and silver, are also abundant. Poorman Creek in the same quadrangle has yielded anomalously high values in silver, bismuth, copper, nickel, and lead, but no samples from this area are included in this report. There are prominent high values for silver, bismuth, copper, nickel, and zinc in samples from the Iditarod quadrangle. One of these samples is also enriched in lead and cadmium. Unusually high values for cobalt and nickel are pres- ent in two concentrates (400 and 3041, table 1) from the Bendeleben quadrangle. Cape Creek in the Teller quadrangle is the source of a magnetic concentrate (497) which contains high values for zinc and bismuth. Indeed, west-central Alaska is a bismuth province. EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA SILVER AND GOLD IN EAST-CENTRAL ALASKA High values for silver and gold are common in magnetic concentrates from the Circle, Eagle, Liven- good, and Tanana quadrangles in east-central Alaska (table 1). Nickel and zinc are also enriched in samples from the Livengood quadrangle, and bismuth attains a high value in the only concentrate from the Circle quadrangle. GEOLOGIC AND GEOCHEMICAL INTERPRETATION REGIONAL The most probable modes of occurrence for anomalous element content in the magnetic concen- trates are as follows (table 34): (1) silver, copper, lead, zinc, cobalt, and nickel sub- stituted for iron in the magnetite structure; , (2) equivalent uranium, copper, lead, and zinc held by surface sorption on magnetite; (3) copper, cadmium, gold, indium and thallium in trace minerals; and (4) equivalent uranium, silver, bismuth, cadmium, copper, gold, indium, and thallium in accessory minerals. TABLE 34.—Summary of probable modes of occurrence of elements present in anomalous amounts in magnetic concentrates from Alaska [X = occurrence likely from data; -- = occurrence unlikely from data; E = equal probability between substitution and accessory minerals; n.d. = no data or insufficient data] Elements present in anomalous amounts Possible Source mode of of . . occurrence evidence eU Ag 81 Cd Co Cu Ni Pb Zn Au In Tl Substitution in magnetite Literature —- X —- -— X X X -— X -- -— -- Geol. aSSOC.1 -— —- -- -- X X X -— X -- —- -— Mineralogy -— -- -- n d. E -- —- E E n d. n.d. n.d. Correlation2 —— X -- -- X X X —— X -— n.d. n.d. Surface sorption-——- Literature X n.d. n.d. n.d. n.d. X n.d. X X -- n.d. n.d. Geol. assoc.l n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Mineralogy n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Correlation2 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Trace minerals---- Literature X X X X X X X X X X X X Geol. assoc.l X X n.d. X X X X X X X X X Mineralogy n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Correlation2 —- X X n.d. X X X X X X X X Accessory minerals—-—— Literature X X X X n.d. X n.d. X X X X X Geol. assoc.1 X X X X X X X X X X X X Mineralogy X X X n.d. E X X E E n.d. n.d. n.d. Correlation2 X X X n.d. X X X X X X n.d. n.d. 1Geologic association in mineral deposits or occurrences. 2Abundances and correlation coefficients. GEOLOGIC AND GEOCHEMICAL INTERPRETATION Although the mineralogical studies (table 30) suggest that anomalous amounts of cobalt and nickel could oc- cur either in substitution for iron or in accessory minerals, the high correlation coefficient for this pair of elements (table 31) favors the substitution mode. Mineralogical data also show an equal probability that lead and zinc are present in substitution for iron in the magnetite structure or in accessory minerals in the magnetic concentrate. The least probable mode of oc- currence shown in table 34 is substitution of bismuth, gold, indium, and thallium in the structure of magnetite. The data in table 34 indicate that the least understood aspect of minor elements in magnetite is the role of surface sorption. Research is needed to determine the effect of sorption on magnetic concen- trates used as a geochemical sample medium. The enrichment of trace elements in detrital magne tite or in magnetic concentrates does not necessarily mean that the source rocks have a superior potential for economic mineralization. The metal-bearing ore solutions may have been dispersed instead of concen- trated if conditions were not favorable for the deposi- tion of ore. However, the anomalously high contents of minor elements are guides to areas deserving further exploration, because many of the anomalous concen- trates are derived from areas of known lode or placer deposits or mineral occurrences. The mineralized areas are well reflected by the regional data in table 1. The significance of these data is made more manifest in table 35 which shows how many known deposits of each type of metal are associated with anomalous magnetic concentrates, and in table 36, where the data are grouped by associations of metals in known lode deposits or mineral occurrences. Twenty-four of the 36 copper deposits are associated 101 with copper-rich magnetic concentrates (table 35); the other 12 are reflected by anomalous lead, zinc, silver, cobalt, or nickel content. The 19 lead-bearing de- posits are reflected in only eight magnetic concen- trates with anomalous lead content, but they are completely reflected by various combinations of anomalous amounts of base metals, silver, or gold (table 1). Magnetic concentrates collected near 10 of the 13 zinc deposits have anomalous zinc content, and samples from near the other three deposits are anomalous in other metals. Of interest is the number of tungsten deposits or occurrences associated with magnetic concentrates that contain anomalous amounts of copper (table 35). This geochemical associa- tion is one that requires follow-up investigations to determine if skarn-type tungsten-copper deposits or porphyry-type deposits may contribute some of the copper. Very strong correlations are found between poly- metallic lode deposits or occurrences and anomalous metal content in the magnetic concentrates (table 36). Indeed, the polymetallic deposits seldom lack accom- panying anomalous magnetic concentrates. Magnetic concentrates may be used satisfactorily as a sample medium for geochemical exploration in subarctic and arctic environments. The data presented here show that anomalous amounts of copper and zinc indicate sulfide mineralization; where combined with other anomalous amounts of elements such as silver, bismuth, and lead, they indicate po1ymetallic sulfide deposits. Anomalous silver content usually indicates silver and gold deposits, mainly gold. Anomalous cobalt and nickel content usually indicates the presence of chromite and, locally, sulfide deposits associated with mafic and ultramafic rocks. Anoma- lous amounts of lead and gold usually indicate lead TABLE 35.—Number and type of known mineral deposits or occurrences located near anomalous magnetic concentrates in Alaska [Data from table 1; numbers in parentheses represent high-background but nonanomalous samples; RE = rare-earth minerals; FM = minerals with fissionable materials] minlzgl gzposit nlghgl Anomalous e1ement content in nearby samples or occurrence of . . deposits eU Cu Pb Zn Ag 81 Co Ni Cd Au1 RE and (or) FM 24 11(1) 1(1) 3(7) 6(4) 1(4) 11 0(1) 3(1) 0 0(1) Copper -------- 36 2(1) 24 9(1) 20(2) 7(6) 7(1) 6(2) 11 2 3 Lead ---------- 19 0 13 8 9(1) 3(4) 4 3 4 2 0(1) Zinc ---------- 13 1 11 5 10(1) 4(3) 4 2(1) 0 2 L Si1ver -------- 22 0 14 7 11(1) 6(4) 6 4(1) 8 2 1 Bismuth ------- 6 1 4 2 2(1) 2(1) 3 0 1 O 2 Coba1t -------- 1 O 1 0 1 1 1 O 1 0 O Go1d ---------- 77 3 43(3) 23 34(7) 22(12) 21(1) 11(2) 27(1) 5 20(5) Mercury ------- 14 0 10 6 10(2) 5(1) 5 2 7 2 4 Tungsten ------ 33 4(1) 16(1) 10 12(5) 9(2) 8 3(2) 10(1) 3 7(1) lIncomp1ete data. 102 TABLE 36.—Number and type of polymetallic deposits or occur rences located near anomalous magnetic concentrates in Alaska [Data from table 1; numbers in parentheses represent high-background but nonanomalous samples] Type of Total lode number deposits of - - deposits Cu Pb Zn Aq Bi Co Ni Anomalous element content in nearby samples Cu-Zn ------- Cu-Au ——————— Cu-Bi ------- Cu-Pb-Ag---- Cu-Zn-Ag——-— dwawd LNG—ad COONO OO—‘OO ——-—'O——'O Cu—Pb-Zn---- Cu-Ag-Au-——- Cu—Pb—Zn-Ag- Cu-Bi-Au-w-- _._.N L _._._._ OOOO OOOO Oxo-a-d oo—Io Bi—Au-w-Sb-— l Co—Au ------- l Pb-Sb ------- l Polymetallic 9 \1—4.—A.—a ONOOO bo—‘O bO—O Ln \0 Totals ------ 26 19 8 6(3) 6(1) sulfide deposits and gold deposits, but not all lead and gold deposits have corresponding anomalous values for lead and gold in magnetic concentrates. Lead and gold content in the magnetic concentrates is more due to chance than are the concentrations of the elements above. However, lead and gold deposits are often in- dicated by anomalous amounts of copper, zinc, and silver. As this investigation shows, copper and zinc in the magnetic concentrates are useful pathfinders for base-metal and precious-metal deposits; cobalt and nickel are useful pathfinders for ultramafic rocks; and equivalent uranium may indicate alkalic granitic rocks. CANDLE QUADRANGLE The copper, lead, zinc, gold, silver, and bismuth con- tent in magnetic concentrates from the Candle quad- rangle is generally greatest near the contacts of the granitic bodies with their wall rocks or in fractured areas of the wall rocks. A possible explanation is that these are the elements that largely remain in the residual magma throughout the main stage of crystal- lization. In the following stages of crystallization the late solutions enriched in these metals tend to move toward low-pressure areas such as faults and fractures, carrying with them, or precipitating, magnetites enriched in these elements. Thus, the concentrations of the minor elements in the magnetic concentrates can probably be used not only as guides to the loci of possi- ble mineralization, but also for an interpretation of the directions of flow of the metal-bearing solutions. For example, during the intrusion of the Granite Mountain pluton into its wall rocks, the residual solutions appear EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA to have moved toward low-pressure areas to the north and northeast of the intrusive, identified by faults and fractures in the wall rock, and deposited various base and precious metals. In this same pluton, high values for equivalent uranium in the magnetic concentrates are confined to the area of the pluton. This may be because the radioactive elements are incorporated in accessory minerals in the granitic rock rather than in vein minerals. Similar distributions of the minor metals are found in magnetic concentrates from the area of the Hunter Creek pluton. The high values for copper, lead, and zinc in the southern and southwestern parts of the Hunter Creek pluton may indicate the migration of minor elements along pressure and temperature gradients toward these areas, leaving only uneconomic dissem- inated deposits or uneconomic veinlets in the pluton. SOLOMON QUADRANGLE The minor-element contents of the magnetic concen- trates from the Solomon quadrangle are quite different from those of the Candle quadrangle, although granitic rocks are dominant sources in both quadrangles. Ex- cept for bismuth and equivalent uranium, the abun- dances of the minor elements are remarkably low in the concentrates from the Solomon quadrangle. Thus, fun- damental differences exist between the trace-element geochemistry of the granitic plutons in the Solomon quadrangle and those in the Candle quadrangle. The minor elements are so sparse in the magnetic concen- trates from the plutons in the Solomon quadrangle, it seems likely that these rocks are barren of base metal deposits. Even where the base metals are enriched in the wall rocks of the plutons in the Solomon quad- rangle, the values are much lower—particularly for copper—than in similar settings in the Candle quadrangle. Some of the magnetic concentrates with anomalous amounts of minor elements are from metamorphic rock terrane in the Solomon quadrangle. During metamor- phism, most minor elements would be locally released during mineral phase transformations. Although re- maining largely in place, these minor elements can be expected to follow Goldschmidt’s rules for camouflage, admission, and capture by the crystal lattices of newly formed minerals. Owing to the high degree of disorder of crystal structures during metamorphism, a greater tolerance for minor elements may be possible in meta- morphic magnetites than in magnetites of igneous origin. SELECTED REFERENCES SELECTED REFERENCES Abdullah, M. I., and Atherton, M. P., 1964, The thermometric sig- nificance of magnetite in low-grade metamorphic rocks: American Journal of Science, v. 262, no. 7, p. 904-917. Abovyan, S. E., and Borisenko, L. F., 1971, Novyye dannyye a vtorichnom magnetite iz ul’trabazitov Armyanskoy SSR, [New data on secondary magnetite from ultrabasic rocks of Armenia]: Akademiya Nauk Armyanskoy SSR Doklady, v. 52, no. 4, p. 240—245. Ahrens, L. H., 1953, Anion affinity and geochemistry, pt. 2 of The use of ionization potentials: Geochemica et Cosochimica Acta, v.3, no.1, p. 1-29. Berg, H. C., and Cobb, E. H., 1967, Metalliferous lode deposits of Alaska: U.S. Geological Survey Bulletin 1246, 254 p. Blekovskiy, A. I., Fominykh, B. G., and Loktina, I. N., 1972, Tip- omorfnyy magnetit matamorfogennykh migmatitov, [Typo- morphic magnetite from migmatites of metamorphic origin]: Akademiya Nauk SSR Doklady, v. 204, no. 4, p. 931—934. Bocchi, Giancarlo, Carapezza, Marcello, and Leone, Marco, 1969, Studio di alcune magnetiti naturali italiane [Studies of some natural magnetites from Italy]: Mineralogica et Petrographica Acta, v. 14, p. 39-70. Borisenko, L. F., 1968, Certain patterns of trace-element distribu- tion in titanomagnetite and magnetite [in Russian, English summary]: Geokhimiya, no. 7, p. 836-842. Borisenko, L. F., and Zolotarev, B. P., 1969, Trace elements in trap titanomagnetite and in magnomagnetite of deposits in the Tunguska syncline: Geochemistry International, v. 6, no. 6, p. 1129—1137. Boyadzhyan, M. T., and Mkrtchyan, G. M., 1969, Trace elements in magnetite from Razdan [in Russian]: Geologiya Rudnykh Mestorozhdeniy, v. 11, no. 2, p. 77—82. Bragg, W. L., Claringbull, G. F., and Taylor, W. H., 1965, Crystal structures of minerals: Ithaca, N. Y., Cornell University Press, 409 p. Brown, R., Jacobs, M. L., and Taylor, H. E., 1972, A survey of the most recent applications of spark source mass spectrometry: American Laboratory, v. 4, no. 11, p. 29—40. Buddington, A. F., 1929, Geology of Hyder and vicinity, south- eastern Alaska, with a reconnaissance of Chickamin River: U.S. Geological Survey Bulletin 807, 124 p. Callahan, J. E., 1974, A heavy mineral based geochemical explora- tion program: Geological Society of America Abstracts with Programs, v. 6, no. 4, p. 338. Capps, S. R., 1913, The Yentna district, Alaska: U.S. Geological Survey Bulletin 534, 75 p. ___1915, The Willow Creek district, Alaska: U.S. Geological Survey Bulletin 607, 86 p. 1919, The Kantishna region, Alaska: U.S. Geological Survey Bulletin 687, 1 16 p. _1935, The southern Alaska Range: U.S. Geological Survey Bulletin 862, 101 p. Carstens, C. W., 1943, Uber den Vanadiumgehalt norwegischer sediment’a'rer Eisenoxyd- und Eisensulfiderze [On the vanadium content of Norwegian sedimentary iron oxide and iron sulfide ores]: Kongelige Norske Videnskab Selskab Skrift, v. 16, p. 1—4. Cass, J. T., 1959, Reconnaissance geologic map of the Norton Bay quadrangle, Alaska: U.S. Geological Survey Miscellaneous Geologic Investigations Map I—286. Chakraborty, K. L., and Majumdar, Tapan, 1971, Zoning in sedi- mentary magnetite: Geological Society of India Journal, v. 12, no. 4, p. 402-403. 103 Chao, T. T., 1972, Selective dissolution of manganese oxides from soils and sediments with acidified hydroxylamine hydrochlo- ride: Soil Science Society of America Proceedings, v. 36, no. 5, p. 764—768. Chapman, R. M., Coats, R. R., and Payne, T. G., 1963, Placer tin deposits in central Alaska: U.S. Geological Survey open-file report, 52 p. Chistyakov, V. K., and Babanskiy, M. D., 1971, Vein and accessory magnetite in the ore occurrence Yagnyanda and some char- acteristics of its formation [in Russian]: Ministerstvo Vysshego Obrazovaniya SSSR, Izvestiya Vysshikh Uchebnykh Zavadeniy, Geologiya i Razvedka, no. 7, p. 26—33. Clark, A. L., and Cobb, E. H., 197 2, Metallic mineral resources map of the Talkeetna quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—369. Clark, S. H. B., 1972, The Wolverine complex, a newly discovered layered ultramafic body in the western Chugach Mountains, Alaska: U.S. Geological Survey open-file report, 10 p. Clark, S. H. B., and Bartsch, S. R., 1971, Reconnaissance geologic map and geochemical analyses of stream sediment and rock samples of the Anchorage B-7 quadrangle, Alaska: U.S. Geolog- ical Survey open-file report, 70 p. Clark, S. H. B., and Foster, H. L., 1969, Geochemical analyses of stream sediment and rock samples, Tanacross quadrangle, Alaska: U.S. Geological Survey open-file report, 1 2 p. Cobb, E. H., 1972a, Metallic mineral resources map of the Anchor- age quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—409. 1972b, Metallic mineral resources map of the Mount Hayes quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—414. _197 2c, Metallic mineral resources map of the Mount McKin- ley quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—366. 1972d, Metallic mineral resources map of the Talkeetna Mountains quadrangle, Alaska: U.S. Geological Survey Mis- cellaneous Field Studies Map MF—370. 1972e, Metallic mineral resources map of the Bethel quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—455. 1972f, Metallic mineral resources map of the Hagemeister Island quadrangle, Alaska: U.S. Geological Survey Mis- cellaneous Field Studies Map MF—362. 1972g, Metallic mineral resources map of the Lake Clark quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF-378. 1972h, Metallic mineral resources map of the Bendeleben quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—417. 1972i, Metallic mineral resources map of the Iditarod quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—363. 1972j, Metallic mineral resources map of the McGrath quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—379. 1972k, Metallic mineral resources map of the Medfra quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—365. 1972l, Metallic mineral resources map of the Nome quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—463. 1972m, Metallic mineral resources map of the Norton Bay quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—381. 104 1972n, Metallic mineral resources map of the Ruby quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—405. 1972o, Metallic mineral resources map of the Circle quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—391. _1972p, Metallic mineral resources map of the Eagle quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—393. ____1972q, Metallic mineral resources map of the Livengood quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—413. 1972r, Metallic mineral resources map of the Tanacross quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—383. _____19725, Metallic mineral resources map of the Tanana quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—371. 197 2t, Metallic mineral resources map of the Ketchikan quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—420. 1972u, Metallic mineral resources map of the Candle quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—389. _1972v, Metallic mineral resources map of the Solomon quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—445. ___1973, Placer deposits of Alaska: U.S. Geological Survey Bulletin 1374, 213 p. Cobb, E. H., and Condon, W. H., 1972, Metallic mineral resources map of the Goodnews quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—447. Cobb, E. H., and Sainsbury, C. L., 1972, Metallic mineral resources map of the Teller quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—426. Colombo, U. P., Fagherazzi, G., Gazzarrini, F., Lanzavecchia, G., and Sironi, Giuseppe, 1964, Studio sull’ossidazione delle magnetiti [Investigation of the oxidation of magnetite]: La Chimica e l’Industria, v. 46, no. 4. p. 357—363. Crozier, W. D., 1960, Black magnetic spherules in sediments: J our- nal of Geophysical Research, v. 65, no. 9, p. 2971—2977. 1961, Micrometeorite measurements—satellite and ground- level data compared: Journal of Geophysical Research, v. 66, no. 9, p. 2793-2795. 1962, Five years of continuous collection of black, magnetic spherules from the atmosphere: Journal of Geophysical Research, V. 67, no. 6, p. 2543—2548. Damon, P. E., Green, W. D., Halva, C. J., and Sabels, B. E., 1960, Correlation and chronology of ore deposits and volcanic rocks: Arizona University Geochronology Laboratories, Geochemical Section, Annual Progress Report 2 to U.S. Atomic Energy Com- mission, Contract AT(1 1—1)—689, 23 p. Dasgupta, H. C., 1967, Intracrystalline element correlation in mag- netite: Economic Geology, v. 62, p. 487—493. ___1970, Some aspects of gallium, nickel and cobalt distribu- tion in titaniferous magnetite [abs.]: Indian Scientific Congress Association Proceedings, 57th Session, pt. 3, p. 165. Davidson, C. F., 1962, On the cobaltznickel ratio in ore deposits: London, Mining Magazine, v. 106, no. 2, p. 78—85. Deb, S. K., and Banerji, A. K., 1967, Trace element content of some magnetite samples from the Singhbhum shear zone, Bihar [abs.]: Indian Scientific Congress Association Proceedings, 57th Session, pt. 3, p. 233—234. Deb, S. K., and Ray, Uttam., 1971, Semi-quantitative spectro- graphic analyses of chalcopyrite, pyrrhotite, pyrite and magne- EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA tite from Kolihan deposit, J hunjhunu District, Rajasthan: Geological Mining and Metallurgy Society of India Quarterly Journal, v. 43, no. 1, p. 55—56. Deer, W. A., Howie, R. A., and Zussman, J ., 1962, Rock-forming minerals, v. 5, Non-silicates: New York, John Wiley, 371 p. de Grys, Ann, 1970, Copper and zinc in alluvial magnetites from central Ecuador: Economic Geology, v. 65, no. 6, p. 714—717. Detterman, R. L., and Cobb, E. H., 1972, Metallic mineral resources map of the Iliamna quadrangle, Alaska: U.S. Geological Survey Miscellaneous Field Studies Map MF—364. Duchesne, Jean-Clair, 1972, Iron-titanium oxide minerals in the Bjerkrem-Sogndal massif, southwestern Norway: Journal of Petrology, v. 13, no. 1, p. 57—81. Duncan, A. R., and Taylor, S. R., 1968, Trace element analyses of magnetites from andesitic and dacitic lavas from Bay of Plenty, New Zealand: Mineralogy and Petrology Contributions, v. 20, no. 1, p. 30—33. Dutro, J. T., Jr., and Payne, T. G., 1954, Geologic map of Alaska: U.S. Geological Survey, scale 122,500,000. Eakin, H. M., 1914, The Iditarod-Ruby region, Alaska: U.S. Geolog- ical Survey Bulletin 578, 45 p. Elliott, R. L., and Miller, T. P., 1969, Results of stream-sediment sampling in the western Candle and southern Selawik quad- rangles, Alaska: U.S. Geological Survey open-file report, 61 p. Elsdon, R., 1972, Iron-titanium oxide minerals in the upper layered series, Kap Edvard Holm, East Greenland: Mineralogical Maga- zine, v. 38, no. 300, p. 946—956. Finkelman, R. B., 1972, Relationship between manganese nodules and cosmic spherules: Marine Technology Society Journal, v. 6, no. 4, p. 34—39. Fleischer, Michael, 1965, Composition of magnetite as related to type of occurrence, in Geological Survey research 1965: U.S. Geological Survey Professional Paper 525—D, p. D82—D84. Fominykh, V. G., and Yarosh, N. A., 1970, Sostav i raspredeleniye magnetita v razlichnykh tipakh granitoidov Urala [Composition and distribution of magnetite in various types of granitoids in the Urals]: Akademiya Nauk SSSR, Ural’skiy Filial, Trudy In- stituta Geologii i Geokhimii, no. 85, p. 167—202. Foster, H. L., 1970, Reconnaissance geologic map of the Tanacross quadrangle, Alaska: U.S. Geological Survey Miscellaneous Geologic Investigations Map I—593. Foster, R. L., 1969, Nickeliferous serpentinite near Beaver Creek, east-central Alaska, in Some shorter mineral resource investiga- tions in Alaska: U.S. Geological Survey Circular 615, p. 2—4. Fredriksson, Kurt, and Gowdy, Robert, 1963, Meteoric debris from the southern California desert: Geochimica et Cosmochimica Acta, v. 27, no. 3, p. 241-243. Fredriksson, Kurt, and Martin, L. R., 1963, The origin of black spherules found in Pacific islands, deep-sea sediments, and Ant- arctic ice: Geochimica et Cosmochimica Acta, v. 27, no. 3, p. 245—248. Frietsch, Rudyard, 1970, Trace elements in magnetite and hema- tite mainl from northern Sweden: Sveriges Geologiska Under- sokning rsbok, v. 64, no. 3, ser. C, no. 646, 136 p. Fujigaki, Shiyogo, and others, 1967, Adsorption of metal ions (MnII, CuII, PbII, ZnII) on magnetite sand [in Japanese]: In- dustrial Water, no. 108, p. 22—27. Goldschmidt, V. M., 1954, Geochemistry: London, Oxford Univer- sity Press, 730 p. Goni, J ., 1966, Contribution a l’étude de la localization et de la dis- tribution des éléments en traces dans les minéraux et les roches granitiques [Contribution to the study of the localization and of the distribution of trace elements in minerals and granitic rocks]: Mémoires du Bureau de Recherches Géologiques et Miniéres, v. 45, p. 1—68. SELECTED REFERENCES Green, W. D., 1960. Dating of magnetite by the helium method, Appendix A1 of Correlation and chronology of ore deposits and volcanic rocks: Arizona University Geochronology Labora- tories, Geochemical Section, Annual Progress Report 2 to U.S. , Atomic Energy Commission, Contract AT(11—1)—689, 23 p. Green, W. D., and Carpenter, David, 1961, The association of urani- um and thorium in magnetite [abs]: American Association for the Advancement of Science, Southwestern and Rocky Moun- tain Division, 37th Annual Meeting, Tempe, Arizona, April 18—19, 1961, Program, p. 33. Haberlandt, Herbert, 1947, Die Bedeutung der Spurenelemente in der geochemischen Forschung [The significance of trace elements in geochemical research]: Osterreichische Akademie der Wissenschaften Sitzungsberichte, Mathematisch-Natur- wissenschaftliche Klasse, sec. 2b, v. 156, p. 293—323. Hamil, B. M., and Nackowski, M. P., 1971, Trace-element distribu- tion in accessory magnetite from quartz monzonite intrusives and its relation to sulfide mineralization in the Basin and Range province of Utah and Nevada—a preliminary report, in Boyle, R. W., and McGerrigle, J. I., eds., Geochemical exploration: Canadian Institute of Mining and Metallurgy Special Volume 11, p. 331—333. Hamilton, J. C., Boerngen, J. G., Marsh, W. R., and Rosenblum, Sam, 1974, Magnetic tape containing results of semiquanti- tative spectrographic analyses of 1069 Alaskan alluvial concen- trates: U.S. Geological Survey Report USGS-GD—74-002; available from U.S. Department of Commerce National Technical Information Service, Springfield, Va., 22151, as Report PB—231246/AS. Handy, R. L., and Davidson, D. T., 1953, On the curious resem- blance between fly ash and meteoritic dust: Iowa Academy of Sciences Proceedings, v. 60, p. 37 3—379. Hanson, L. G., 1963, Bedrock geology of the Rainbow Mountain area, Alaska Range, Alaska: Alaska Division of Mines and Minerals Geologic Report 2. 82 p. Hawkes, H. E., and Webb, J. S., 1962, Geochemistry in mineral exploration: New York, Harper and Row, 415 p. Hegemann, Friedrich, and Albrecht, F., 1955, Zur Geochemie Oxy- dischereisenerze [On the geochemistry of oxide iron ore]: Chemie der Erde, v. 17, p. 81—103. Herreid, Gordon, 1965, Geology of the Bear Creek area, Seward Peninsula, Candle quadrangle, Alaska: Alaska Division of Mines and Minerals Geologic Report 12, 16 p. Hoare, J. M., and Cobb, E. H., 1972, Metallic mineral resources map of the Russian Mission quadrangle, Alaska: U.S. Geolog- ical Survey Miscellaneous Field Studies Map MF—444. Hoare, J. M., and Coonrad, W. L., 1959, Geology of the Bethel quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Geologic Investigations Map 1—285. 1961a, Geologic map of the Goodnews quadrangle, Alaska: U.S. Geological Survey Miscellaneous Geologic Investigations Map I—339. __1961b, Geologic map of the Hagemeister Island quadrangle, Alaska: U.S. Geological Survey Miscellaneous Geologic Investi- gations Map I—321. Hose, R. K., and Blake, M. 0., Jr., 1970, Geologic map of White Pine County, Nevada (east half): U.S. Geological Survey open-file report, 2 sheets. Howie, R. A., 1955, The geochemistry of the charnockite series of Madras, India: Royal Society of Edinburgh Transactions, v. 62, pt. 3, p. 725—768. Hubert, A. E., and Lakin, H. W., 1973, Atomic absorption deter- mination of thallium and indium in geologic materials, in Jones, M. J ., ed., Geochemical exploration 1972: London, The Institu- tion of Mining and Metallurgy, p. 383—387. 105 James, A. H., and Dennen, W. H., 1962, Trace ferrides in the mag- netite ores of the Mount Hope Mine and the New Jersey High- lands: Economic Geology, v. 57, no. 3, p. 439-449. Jenne, E. A., 1968, Controls on Mn, Fe, Co, Ni, Cu, and Zn concen- trations in soils and water—the significant role of hydrous Mn and Fe oxides, chap. 21 in Trace inorganics in water, American Chemical Society, 153d Annual Meeting, Miami Beach, Florida, 1967, Division of Water, Air and Waste Chemistry Symposium: American Chemical Society (Advances in Chemistry Series 73), p. 337—387. Karkhanavala, M. D., 1958, On radioactive uraniferous iron oxides: Geochimica et Cosmochimica Acta, v. 15, no. 3, p. 229—236. Kaye, C. A., and Mrose, M. E., 1965, Magnetic spherules, colored corundum, and other unusual constituents of a heavy beach sand, Martha's Vineyard, Massachusetts, in Geological Survey research, 1965: U.S. Geological Survey Professional Paper 525-D, p. D37—D43. Kiseleva, I. A., and Matveyev, A. A., 1967, Some patterns of trace- element distribution in magnetites of contact-metamorphic deposits (eastern Sayan) [in Russian]: Geologiya Rudnykh Mestorozhdeniy, v. 9, no. 6, p. 7 6-80. Kisvarsanyi, Geza, and Proctor, P. D., 1967, Trace element con- tent of magnetites and hematites, southeast Missouri iron metallogenic province, U.S.A.: Economic Geology, v. 62, no. 4, p. 449-471. Komov, I. L., 1968, Trace elements in ores and rocks of the Yena- shimo magnetite deposit, Yenisei Ridge [in Russian]: Ministerstvo Vysshego Obrazovaniya SSSR, Izvestiya Vysshikh Uchebnykh Zavadeniy, Geologiya i Razvedka, no. 3, p. 58—60. 1969, Geologic structure, composition, and formation con- ditions of the Yenashimo magnetite deposit, Yenisei Ridge [in Russian]: Ministerstvo Vysshego Obrazovaniya SSSR, Izvestiya Vysshikh Uchebnykh Zavadeniy, Geologiya i Razvedka, no. 3, p. 91—96. Krauskopf, K. B., 1967, Introduction to geochemistry: New York, McGraw-Hill, 721 p. Ksenofontov, O. K., and Davydov, Ye. V., 1971, Petrology, geo- chemistry, and metallogeny of the Barambay pluton, western Turgai [in Russian]: Trudy Vsesoyuznogo Nauchno- Issledovatel'skogo Geologicheskogo Instituta (VSEGEI), no. 169, p. 70—89. Landergren, Sture, 1948, On the geochemistry of Swedish iron ores and associated rocks: Sveriges Geologiska Undersiikning sbok, v. 42, no. 5, 182 p. Landes, K. K., 1927, Geology of the Knik-Matanuska district. Alas- ka: U.S. Geological Survey Bulletin 792, p. 51—72. Langway, C. C., J r., and Marvin, U. B., 1964, Some characteristics of black spherules: New York Academy of Sciences Annals, v. 119, p. 205—223. Lauren, Lennart, 1969, On magnetite-bearing pegmatites in SE Sottunga, Aland Islands: Geological Society of Finland Bulletin, v. 41, p. 107—114. Leblanc, Marc, 1969, Magnétites nickéliferes dans les serpentinites du Graara central (Anti-Atlas, Maroc) [Nickeliferous magnetites in the serpentinites of the central Graara (Anti- Atlas, Morocco)]: Service Géologique du Maroc Notes et Mémoires, no. 213, v. 29, p. 165—171. Lepeltier, Claude, 1969, A simplified statistical treatment of geochemical data by graphical representation: Economic Geology, v. 64, no. 5, p. 538-550. Leung, S. S., 1970, Trace ferride content in magnetites of cores and host rocks of different geologic occurrences [abs.]: Geological Society of America Abstracts with Programs, v. 2, no. 3, p. 227—228. 106 Lewis, J. F., 1970, Chemical composition and physical properties of magnetite from the ejected plutonic blocks of the Soufriére Volcano, St. Vincent, West Indies: American Mineralogist, v. 55, no. 5—6, p. 793-807. Lindsley, D. H., and Smith, Douglas, 1971, Petrography and min- eral chemistry of a differentiated flow of Picture Gorge basalt near Spray, Oregon—geologic setting and mechanism of dif- ferentiation: Carnegie Institution of Washington Year Book 69, p. 264—285. Lipman, P. W., 1971, Iron-titanium oxide phenocrysts in compo- sitionally zoned ash-flow sheets from southern Nevada: Journal of Geology, v. 79, no. 4, p. 438—456. Lopez M., Hector, Castaneda, Abraham, and Salgado U., Benito, 1970, Estudio geologico-magnetometrico del yacimiento fer- rifero La Guyabera, Municipio de Villa Vittoria, Michoacan [Geologic and magnetometric study of La Guayabera iron deposit]: México Consejo de Recursos Naturales no Renovables Bolétin 72, 49 p. Lovering, T. G., 1955, Progress in radioactive iron oxides investiga- tions: Economic Geology, v. 50, no. 2, p. 186—195. Lovering, T. G., and Beroni, E. P., 1959, Preliminary study of radio- active limonite in Colorado, Utah, and Wyoming: US. Geologi- cal Survey Bulletin 1046-N, p. 339—384. McKelvey, V. E., Everhart, D. L. E., and Garrels, R. M., 1956, Sum- mary of hypotheses of genesis of uranium deposits, in Page, L. R., Stocking, H. E., and Smith, H. B., eds., Contributions to the geology of uranium and thorium by the United States Geological Survey and Atomic Energy Commission for the United Nations International Conference on Peaceful Uses of Atomic Energy, Geneva, August 1955: US. Geological Survey Professional Paper 300, p. 41—53. MacKevett, E. M., Jr., 1971, Stratigraphy and general geology of the McCarthy C—5 quadrangle, Alaska: US. Geological Survey Bulletin 1323, 35 p. MacKevett, E. M., Jr., and Cobb, E. H., 1972, Metallic mineral resources map of the McCarthy quadrangle, Alaska: US. Geological Survey Miscellaneous Field Studies Map MF—395. MacKevett, E. M., Jr., and Smith, J. G., 1968, Distribution of gold, copper, and some other metals in the McCarthy B—4 and B—5 quadrangles, Alaska: US. Geological Survey Circular 604, 25 p. ____1972, Geologic map of the McCarthy B—4 quadrangle, Alas- ka: US. Geological Survey Geologic Quadrangle Map GQ—943. McKinstry, H. E., and Kennedy, G. C., 1957, Some suggestions con- cerning the sequence of certain ore minerals: Economic Geology, v. 52, no. 4, p. 379—390. Maloney, R. P., 1962, Investigation of mercury-antimony deposits near Flat. Yukon River region, Alaska: US. Bureau of Mines Report of Investigations 5991, 44 p. Marmo, Vladi, 1959, On the TiO2 content of magnetites as a petro- genetic hint: American Journal of Science, v. 257, no. 2, p. 144—149. Mason, Brian, 1966, Principles of geochemistry: New York, John Wiley, 3rd ed., 329 p. Mertie, J. B., J r., 1919, Platinum-bearing gold placers of the Kahilt- na Valley [Alaska], in Mineral Resources of Alaska, 1917: US. Geological Survey Bulletin 692, p. 233—264. 1936, Mineral deposits of the Ruby-Kuskokwim region, Alas- ka: US. Geological Survey Bulletin 864—C, p. 115—245. __1940a, The Goodnews platinum deposits, Alaska: US. Geo- logical Survey Bulletin 918, 97 p. _1940b, Placer gold in Alaska: Washington Academy of Sci- ence Journal, v. 30, p. 93—124. Mertie, J. B., J r., and Harrington, G. L., 1916, Mineral resources of the Ruby-Kuskokwim region [Alaska], in Brooks, A. H., and EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA others, Mineral Resources of Alaska, 1915: US. Geological Survey Bulletin 642, p. 223—266. Miesch, A. T., 1963, Distribution of elements in Colorado Plateau uranium deposits—a preliminary report: US. Geological Survey Bulletin 1147—E, 57 p. Miller, T. P., 1970, Petrology of the plutonic rocks of west-central Alaska: US. Geological Survey open-file report, 132 p. 1972, Potassium-rich alkaline intrusive rocks of western Alaska: Geological Society of America Bulletin, v. 83, no. 7, p. 2111—2128. Miller, T. P., and Elliott, R. L., 1969, Metalliferous deposits near Granite Mountain, eastern Seward Peninsula, Alaska: US. Geo- logical Survey Circular 614, 19 p. Miller, T. P., Elliott, R. L., Grybeck, D. G., and Hudson, Travis, 1971, Results of geochemical sampling in the northern Darby Mountains, Seward Peninsula, Alaska: US. Geological Survey open-file report, 12 p. Miller, T. P., and Grybeck, D. G., 1973, Geochemical survey of the eastern Solomon and southeastern Bendeleben quadrangles, Seward Peninsula, Alaska: US. Geological Survey open-file report, 115 p. Miller, T. P., Grybeck, D. G., Elliott, R. L., and Hudson, Travis, 1972, Preliminary geologic map of the eastern Solomon and southeastern Bendeleben quadrangles, eastern Seward Penin- sula, Alaska: US. Geological Survey open-file report, 11 p. Moses, L. E., and Oakford, R. V., 1963, Tables of random permuta- tions: Stanford, Calif, Stanford University Press, 195 p. Mulligan, J. J ., 1966, Tin-lode investigations, Cape Mountain area, Seward Peninsula, Alaska; with a section on petrography by W. L. Gnagy: US. Bureau of Mines Report of Investigations 6737, 43 p. Nagaytsev, B. M., 1971, Mestorozhdenie magnetita trubki vzryva “Verkhniy Ollonokon" (basseyn Podkamennoy Tunguski) [Magnetite deposits of the volcanic pipe “Verkhnniy Ollonokon,” Podkamennaya Tunguska River Basin] [in Rus- sian, English summary]: Akademiya Nauk SSSR Sibriskoye Otdeleniye, Geologiya i Geofizika, no. 4. p. 140—143. Nakagawa, H. M., 1975, Atomic absorption determination of silver, bismuth, cadmium, cobalt, copper, nickel, lead, and zinc in calcium- and iron-rich geologic materials, in Ward, F. N., ed., New and refined methods of trace analysis useful in geochemical exploration: US. Geological Survey Bulletin 1408. p. 85—96. Nelson, A. E., West, W. S., and Matzko, J. J ., 1954, Reconnaissance for radioactive deposits in eastern Alaska, 1952: U.S. Geological Survey Circular 348, 21 p. Né'mec, Dusan, 1968, Spurenelemente der Magnetite aus den Skar- nen des Mendenecer Reviers (Erzgebirge): [Trace elements in magnetite from the skarn at Medenecer Reviers (Erzgebirgefl: Chemie der Erde, v. 27, no. 2, p. 151—163. Neuerburg, G. J ., Botinelly, Theodore, and Kulp, K. E., 1971, Pros- pecting significance of hypogene iron oxide distributions in the Montezuma district, central Colorado: Society of Mining Engineers Transactions 250, p. 97—101. Newhouse, W. H., 1936, Opaque oxides and sulfides in common igneous rocks: Geological Society of America Bulletin, v. 47, no. 1, p. 1-52. Nininger, R. D., ed., 1956, Exploration for nuclear raw materials: New York, D. Van Nostrand, 293 p. Nockolds, S. R., 1966, The behavior of some elements during frac- tional crystallization of magma: Geochimica et Cosmochimica Acta, v. 30, p. 267—278. Ogorodova, A. S., 1970, Magnetite and pyrite from hydrothermally altered rocks of Pauzhetka, in Mineralogiya gidrotermal’hykh sistem Kamchatki i Kuril'shikh ostrovov [in Russian]: SELECTED REFERENCES Akademiya Nauk SSSR Sibirskoye Otdeleniye, Institut Vulkanologii Trudy, p. 116—120. Oshima, Osamu, 1971, Compositional variation of magnetite during the eruption and its bearing on the stage of crystallization of magma of Futatsu-dake, Haruna volcano: Journal of Mineralogy, v. 6, no. 4, p. 249—263. __1972, Compositional variation of magnetite in the rocks of the Haruna volcano (2) [abs., in Japanese]: Volcanology Society of Japan Bulletin, v. 17, no. 1, p. 31—32. Overstreet, W. C., Warr, J. J ., and White, A. M., 1970, Influence of grain size on percentages of Th02 and U50, in detrital monazite from North Carolina and South Carolina: U.S. Geological Pro- fessional Paper 700-D, p. D207—D216. Overstreet, W. C., Yates, R. G., and Griffitts, W. R., 1963, Heavy minerals in the saprolite of the crystalline rocks in the Shelby quadrangle, North Carolina: U.S. Geological Survey Bulletin 1162—F, 31 p. Paige, Sidney, and Knopf, Adolph, 1907, Geologic reconnaissance in the Matanuska and Talkeetna basins, Alaska: U.S. Geological Survey Bulletin 327, 71 p. Patton, W. W., J r., 1967, Regional geologic map of the Candle quad- rangle, Alaska: U.S. Geological Survey Miscellaneous Geologic Investigations Map I—492. Pavlov, D. 1., 1969, Chlorine in the formation of Ampalyk magnetite deposit: International Geology Revue, v. 11, no. 8, p. 857—859. Prindle, L. M., 1913, A geologic reconnaissance of the Fairbanks quadrangle, Alaska: U.S. Geological Survey Bulletin 525, 220 p. Prindle, L. M., and Katz, F. J., 1909, The Fairbanks gold placer region, in Brooks, A. H., and others, Investigation of mineral resources of Alaska in 1908: U.S. Geological Survey Bulletin 379, p. 181-200. Ramdohr, Paul, 1940, Die Erzmineralien an gew'dhnlichen magma- tischen Gesteinen [The ore minerals of common magmatic rocks]: Abhandlungen der Akademie der Wissenschaften in Preussische, Mathematisch-Physikalische Klasse, no. 2, 43 p. _1962, Erzmikroskopische Untersuchungen an Magnetiten der Exhalationen im Valley of the 10,000 Smokes [with English abstract]: Neues J ahrbuch fiir Mineralogie Monatschefte, nos. 3—4, p. 49—59. Rankama, K. K., and Sahama, T. G., 1950, Geochemistry: Chicago, The University of Chicago Press, 912 p. Richter, D. H., and Matson, N. A., Jr., 1972, Metallic mineral re- sources map of the Nabesna quadrangle, Alaska: U.S. Geolog- ical Survey Miscellaneous Field Studies Map MF—422. Ringwood, A. E., 1955, The principles governing trace element distributions during magmatic crystallization—Part I. The in- fluence of electronegativity: Geochimica et Cosmochimica Acta, v. 7, p. 189—202. Robinson, G. D., Wedow, Helmuth, Jr., and Lyons, J. B., 1955, Radioactivity investigations in the Cache Creek area, Yentna district, Alaska: U.S. Geological Survey Bulletin 1024—A, 23 p. Rose, A. W., 1966, Geology of the chromite-bearing ultramafic rocks near Eklutna, Anchorage quadrangle, Alaska: Alaska Division of Mines and Minerals Geologic Report 18, 22 p. _1967, Geology of the Upper Chistochina River area, Mount Hayes quadrangle, Alaska: Alaska Division of Mines and Minerals Geologic Report 28, 41 p. Rosenblum, Sam, Day, G. W., Crim, W. B., Overstreet, W. C., and Larson, A. L., 1974, Magnetic tape containing results of semi- quantitative spectrographic analyses of the magnetic fractions of 680 Alaskan alluvial concentrates: U.S. Geological Survey Report USGS—GD—74—001, 7 p.; available from U.S. Depart- ment of Commerce National Technical Information Service, Springfield, Va., 22151, as Report PB—229947/AS. Rost, F., 1940, Spektralanalytische Untersuchungen an sulfi- 107 dischen Erzlagerst'a'tten des ostbayerischen Grenzgebirges, ein Beitrag zur Geochemie von Nickel und Kobalt [Spectrographic analytical investigations of sulfide ore deposits of the east Bavarian Grenzgebirge, a contribution to the geochemistry of nickel and cobalt]: Zeitschrift fiir Angewandte Mineralogie, v. 2, nos. 1—3, p. 1—27. Sainsbury, C. L., 1957, A geochemical exploration for antimony in southeastern Alaska: U.S. Geological Survey Bulletin 1024—H, p. 163—178. ___1963, Beryllium deposits of the western Seward Peninsula, Alaska: U.S. Geological Survey Circular 479, 18 p. 1964, Geology of Lost River Mine area, Alaska: U.S. Geolog- ical Survey Bulletin 1129, 80 p. 1969, Geology and ore deposits of the central York Moun- tains, western Seward Peninsula, Alaska: U.S. Geological Survey Bulletin 1287, 101 p. Sainsbury, C. L., Hamilton, J. C., and Huffman, Claude, J r., 1968, Geochemical cycle of selected trace elements in the tin- tungsten-beryllium district, western Seward Peninsula, Alaska—a reconnaissance study: U.S. Geological Survey Bulletin 1242—F, 42 p. Santos, G. G., J r., and Walters, L. J ., Jr., 1971, Activation analysis of Philippine magnetite sands, in Activation analysis in geochemistry and cosmochemistry: Oslo-Bergen-Tromso, Universitetsforlaget, p. 333—335. Sastry, C. A., and Krishna Rao, J. S. R., 1970, Ore microscopic, X-ray and trace elemental studies of a few iron ores from south India: Geological Society of India Journal, v. 11, no. 3, p. 242—247. Schidlowski, Manfred, and Bitzkowski, Seigfried, 1972, Magentit- kiigelchen aus dem hessischen Teritiaer; Ein Beitrag zur Frage der “kosmischen Kiigelchen” [Magnetite spherules from the Tertiary of Hesse; a contribution to the problem of cosmic spherules]: Neues Jahrbuch fiir Geologic und Pal'aontologie Monatschefte, v. 1972, no. 3. p. 170—182. Sen, N., Nockolds, S. R., and Allen, R., 1959, Trace elements in minerals and rocks of the S. Californian batholith: Geochimica et Cosochimica Acta, v. 16, nos. 1—3, p. 58—78. 7 Sergeyev, V. N., and Tyulyupo, B. M., 1972, O metmorfisme mag- netitov iz mestorozkeniy Tashelgino-Mayzasskoy rudnoy zony Gornoy Shorri [Metamorphism of magnetites in the deposits of the Tashelgino-Mayzas ore zone, Shoriya Mountains]: Akademiya N auk SSSR Sibirskoye Otdeleniye, Geologiya i Geofizika, no. 4, p. 131—137. Shangireyev, Kh. G., 1969, Characteristics of trace element distri- bution in magnetite ores and problems of the genesis of the Sokolovskoye deposit [in Russian]: Akademiya Nauk Kazakhskoy SSR Institut Geologii Trudy, v. 28, p. 96—114. Shcherbak, V. M., 1969, Structural and genetic characteristics of the magnetite deposits in the Yeltai-Kurzhinkol region [in Rus- sian]: Akademiya Nauk Kazakhskoy SSR Institut Geologii Trudy, v. 28, p. 49—65. 1970, Trace elements in magnetites from metasomatic de- posits formed under various conditions, Turgay [in Russian]: Akademiya Nauk Kazakhskoy SSR Izvestiya, Seriya Geo- logicheskaya, no. 2, p. 54—56. Shilin, N. L., 1970, Evaluation of the P-T conditions of formation of the two-phase hypabyssal rock bodies, central Kamchatka, in Magma maloglubinnykh kamer [in Russian]: Akademiya N auk SSSR Sibirskoye Otdeleniye, Institut Vulkanologii Trudy, p. 53—67. Shteynberg, D. S., and Chasechukhin, I. S., 1970, Behavior of tri- valent iron during serpentinization [in Russian]: Zapiski Vsesoyuznogo Mineralogicheskogo Obshchestva, v. 99, no. 5, p. 553—563. 108 Skolnick, Herbert, 1961, Ancient meteoric dust: Geological Society of American Bulletin, v. 72, no. 12, p. 1837—1841. Skylar, V. P., 1972, O megnetite v porodakh Mariupol’skikh zhelez- orudnykh mestorozhdeniy [Magnetite in rocks of the Mariupol (Zhdanov) iron ore deposit]: Geologichniy Zhumal, v. 32, no. 4, p. 98—103. Smith, P. S., and Eakin, H. M., 1911, A geologic reconnaissance in southeastern Seward Peninsula and the Norton Bay-Nulato re- gion, Alaska: US. Geological Survey Bulletin 449, 146 p. Sullivan, E. C., 1907, The interaction between minerals and water solutions with special reference to geologic phenomena: US. Geological Survey Bulletin 312, 69 p. Summers, D. B., 1970, Chemistry Handbook: Boston, Willard Grant Press, 92 p. Taylor, S. R., 1965, The application of trace element data to prob- lems in petrology, in Ahrens, L. H., Press, Frank, and Runcorn, S. K., Physics and chemistry of the Earth, v. 6: Oxford, Pergamon Press, p. 133—214. Theobald, P. K., Jr., Overstreet, W. C., and Thompson, C. E., 1967, Minor elements in alluvial magnetite from the Inner Piedmont Belt, North and South Carolina: US. Geological Survey Pro- fessional Paper 554—A, 34 p. Theobald, P. K., Jr., and Thompson, C. E., 1959, Reconnaissance exploration by analysis of heavy mineral concentrates [abs]: Mining Engineering, v. 11, no. 1, p. 40. __1962, Zinc in magnetite from alluvium and from igneous rocks associated with ore deposits, in Geological Survey re- search 1962: U.S. Geological Survey Professional Paper 450—C, p. C72—C73. Thiel, Edward, and Schmidt, R. A., 1961, Spherules from the Ant- arctic ice cap: Journal of Geophysical Research, v. 66, p. 307—310. Thompson, C. E., Nakagawa, H. M., and Van Sickle, G. H., 1968, Rapid analysis for gold in geologic materials, in Geological Survey research 1968: US. Geological Survey Professional Paper 600—B, p. B130—B132. Unan, C., 1971, The relation between chemical analysis and min- eralogy of oxidized titaniferous magnetites in some Scottish dolerites: Royal Astronomical Society Geophysical Journal, v. 22, no. 3, p. 241—259. US. Army Corps of Engineers, 1961, Nature and distribution of particles of various sizes in fly ash: Vicksburg, Miss., US. Army, Corps of Engineers, Waterways Experiment Station Technical Report 6—583, 21 p. Vartanova, N. S., Zav'yalova, I. V., and Simonova. V. I., 1970, Elementy-primesi aktsessornogo magnetite iz granitoidov vost— ochnogo Zabaykal’ya [Trace elements in accessory magnetite from granitoids of eastern Transbaikalia] [in Russian, English summary], in Problemy petrologii i geneticheskoy mineralogii, v.2: Izdanja Nauka. p. 216—222. Vassileff, L., 1971, Antemagmatic magnetite ores with Mg-skarns from intermediate depths in the Kroumovo deposit, Bulgaria, in International Association of the Genesis of Ore Deposits, EQUIVALENT URANIUM AND SELECTED MINOR ELEMENTS, ALASKA Tokyo-Kyoto Meetings, Papers and Proceedings: Society of Mining Geologists of Japan Special Issue, no. 3, p. 381—386. Vergilov, V., 1969, Chlorite schists containing magnetite, Sakar region [in Bulgarian, English summary]: Bulgarsko Geologique Druzhestvo, Spisaniye, v. 30. no. 1, p. 79—82. Vincent, E. A., and Bilefield, L. I., 1960, Cadmium in rocks and min- erals from the Skaergaard intrusion, East Greenland: Geochimica et Cosmochimica Acta, v. 19, no. 1, p. 63-69. Vincent, E. A., and Crocket, J. H., 1960, The distribution of gold in rocks and minerals of the Skaergaard intrusion, East Greenland, pt. 1 of Studies in the geochemistry of gold: Geochimica et Cosmochimica Acta, v. 18, nos. 1—2, p. 130—142. Vincent, E. A., and Phillips, R., 1954, Iron-titanium oxide minerals in layered gabbros of the Skaergaard intrusion, East Green- land, pt. 1 of Chemistry and ore-microscopy: Geochimica et Cosmochimica Acta, v. 6, no. 1, p. 1—26. Vlasov, K. A. [ed.], 1966, Geochemistry and mineralogy of rare ele- ments and genetic types of their deposits—v. 1. Geochemistry of rare elements: Institute of Mineralogy, Geochemistry and Crystal Chemistry of Rare Elements, Akademie Science USSR, [English translation by Israel Program for Science Translation, Jerusalem, 1968]. Wager, L. R., and Mitchell, R. L., 1951, The distribution of trace elements during strong fractionation of basic magma—a further study of the Skaegaard intrusion, East Greenland: Geochimica et Cosmochimica Acta, v. 1, no. 3, p. 129-208. Waters, A. E., Jr., 1934, Placer concentrates of the Rampart and Hot Springs districts: US. Geological Survey Bulletin 844—D, p. 227—246. Wayland, R. G., 1961, Tofty tin belt, Manley Hot Springs district, Alaska: U.S. Geological Survey Bulletin 1058—1, p. 363—414. Wedow, Helmuth, Jr., Killeen, P. L., and others, 1954, Reconnais- sance for radioactive deposits in eastern interior Alaska, 1946: US. Geological Survey Circular 331, 36 p. Wells, F. G., 1933. Lode deposits of Eureka and vicinity, Kantishna district, Alaska: US. Geological Survey Bulletin 849—F, p. 335—379. West, W. S., 1953, Reconnaissance for radioactive deposits in the Darby Mountains, Seward Peninsula, Alaska. 1948: US. Geo- logical Survey Circular 300, 7 p. West, W. S., and Benson, P. D., 1955, Investigations for radioactive deposits in southeastern Alaska: US. Geological Survey Bulletin 1024—B, p. 25—57. White, M. G., and Stevens, J. M., 1953, Reconnaissance for radio- active deposits in the Ruby-Poorman and Nixon Fork districts, west-central Alaska, 1949: US. Geological Survey Circular 279, 19 p. Whittaker, E. J. W., and Muntus, R., 1970, Ionic radii for use in geochemistry: Geochimica et Cosmochimica Acta, v. 34, p. 945—956. Yamaoka, Kazuo, 1962. Studies on the bedded cupriferous iron sul- fide deposits occurring in the Sambagawa metamorphic zone: Tohoku University Science Reports, 3d ser., v. 8, no. 1, p. 1—68. Page A Ablation products .......................... 88 Abstract ............................ 1 Accessory minerals .............. 22. 45. 49. 83.87. 90, 94, 96. 100, 102 Alaskan placer concentrate file ............ 2, 3. 16. 27 Albert Creek. cobalt ........................ 43 Allanite ........................... 69 H. C. Carstens mine . ............. 27 Portage Creek ...... 27 Alluvium .......................... . 2, 3. 43 Amphibole ................................ 87. 90 Analcime. Jump Creek . . . . 31 Analysis, atomic absorption ............. 2. 16. 19. 23. 65. 83. 89 laser-probe ............................. 88 mineralogical examination ............... 2 procedures ............. 3 radiometric counting .................... 2, 20 semiquantitative spectrographic . . . 2. 65, 89.91.92 Anatose ................................... 87 Anchorage quadrangle .. . 23. 33. 37. 42. 99 Andesite ........................ 58. 59 Bear Creek ............................. 64 porphyritic pyroxene .................... 47 Antimony .......................... 34. 37. 38 Bonanza Creek gold placer ............... 35 Chicken Creek .......................... 40 Flat Creek .......... 35 My Creek .......... 36 Otter Creek ........ 35 Vinasale Mountain ...................... 35 Apatite. H. C. Carstens mine ................ 27 Portage Creek ..................... 27 Appel Mountain . . .. .. ............. 28 Archangel Creek, silver . ............. 37 Arg'illite .......................... 44 Arolik River. cobalt ................ 43 gold .......................... 38 nickel .................................. 43 silver .................................. 38 Arsenopyrite. Fairbanks Creek gold placer . . . . 45 Associations. See Accessory minerals. Atomic absorption analysis ..... 2. I6. 19. 23. 65. 83, 89 Atwater Bar. gold .......................... 39 silver .................................. 39 Azurite .................................... 87. 91 B Barium ..... . . . . 74, 75. 87 Bartels tin mine .. . . . . ............. 40 Basalt .......................... 43. 44. 45. 47. 51. 59 Bear Creek ............................. 64 Base metals ............ 20. 33. 46, 58. 72. 102 Bear Creek .................... 35 Bethel quadrangle . . ........... 99 Candle quadrangle ...................... 58 Livengood quadrangle ................... 36 McGrath quadrangle . . 46 Mount Holmes ......................... 34 Mount McKinley quadrangle ............ 34 Russian Mission quadrangle ............. 35 Talkeetna Mountains ................... 33 Base-metal sulfides. Bendeleben quadrangle . . 74 Basin Province ............................. 2 INDEX [Italic page numbers indicate major references] Page Bear Creek. andesite ........................ 64 basalt .................................. 64 base metals ............................ 35 bismuth ....... i .. .. 59 . . . 53. 58. 64 58 gold ................................... 58 lead ................................... 58. 64 ............ 58 ............ 58 ............ 58 thallium ............................... 64 zinc .................................... 58. 64 35 44 gold ................................... 38 nickel ......................... 44 silver ......................... 38 Bendeleben quadrangle . . . . 23. 30. 38, 65. 80 base-metal sulfides ...................... 74 bismuth ................................ 40 cadmium ............ cobalt ............... copper .............. radioactivity ........................... 25. 99 zinc .................................... 35, 77 Bethel quadrangle ......... 23, 39 base metals ........... 99 copper ................ 34 36. 38 46 43 36 34 ............................... 87 Bismuth ...................... 3. 23. 39. 47. 59. 64. 65. 77. 92. 94. 101, 102 analysis ............................ 2, 16. 18. 20 Bear Creek ............................. 59 Bendeleben quadrangle ..... . . . . 40 Candle quadrangle ......... . 59. 97. 102 Cape Creek ................ . . 40. 100 Circle quadrangle ....................... 40 Clear Creek ............................. 79, 80 Darby Mountains .......... 40. 80 Darby pluton ................ 80. 83 distribution ................. 99 Eagle quadrangle ............ . . 40 east-central Alaska .......... . 40. 100 Fairbanks Creek gold placer .. .. 45 Fish Creek ............................. 4O Fortymile River ........................ 40 Glen Gulch .................. 100 Granite Mountain pluton ....... 59 Greer Gulch ................... . . 40 H. C. Carstens mine ..................... 40 Hagemeister Island quadrangle .......... 39 Hidden Creek .................. . . 35 Humboldt Creek ........................ 40 Hunter Creek pluton .................... 59 Iditarod quadrangle ............ 40. 100 Livengood quadrangle ............ 40 Medfra quadrangle ............... . . 40 Melba Creek ............................ 40 Bismuth—Continued Page Poorman Creek ......................... 100 Portage Creek .......................... 4o. 45 Ruby quadrangle. .. .. .. 40. 78. 100 Seward Peninsula ....... . . ..... 78 Solomon quadrangle .................... 77. 98 Talkeetna Mountains quadrangle ......... 39 Teller quadrangle ..... 'l‘in City .............. Vinasale Mountain ...................... 35 west-central Alaska ..................... 40. 100 Bonanza Creek. gold ............. 38 silver ........................... 38 Bonanza Creek gold placer. elements ..... 35 Bowman Cut ............................... 16 gold ................................... 38 silver ....................... 38 Bradfield Canal quadrangle . . . . ..... 23, 32 C Cache Creek. minerals ....................... 43 Cadmium ........................ 3. 22, 23, 31, 47. 51. 64. 69, 71. 76. 83. 92 analysis ......................... 2. 16. 18. 19, 20 Bendeleben quadrangle ........ . . . 35 Candle Creek gold placer ....... . . . 35 Candle quadrangle ..................... 47, 61. 97 Darby Mountains ....................... 35 east-central Alaska ........... 36 Fish Creek ................... 33 Glacier Peak .................. . 34 Glen Gulch ............................. 100 Iditarod quadrangle ..................... 96. 100 Ketchikan quadrangle ........... . 33 Last Chance Creek ................ . 34 McGrath quadrangle ................ . 35.46.96 Otter Creek tin placer ............. 35 Portage Creek .................... 36 Rainbow Mountain copper deposits . . 34 Roundabout Mountain .................. 35 Ruby quadrangle ....................... 96. 100 Salmon River ..................... 33 Solomon quadrangle ................ 69. 76 southeastern Alaska .......... r ..... . . 32 southern Alaska ........................ 33 southwestern Alaska .................... 34 Spruce Creek ........................... 34 Talkeetna quadrangle ................... 33 west-central Alaska ..................... 35. 100 Calcite ................................ 4'7. 87 Calcium .......................... 22 Candle Creek .............................. 30 Candle Creek placer. elements ................ 35. 44 Candle Creek valley ......................... 44 Candle quadrangle . . . ......... 23. 47. 96. 102 base metals ..................... 58 bismuth ..................... 59, 97. 102 cadmium. . cobalt ........... composition ..... . . copper ............................ 49. 51. 97. 102 correlations ............................ 96 equivalent uranium . 49. 51. 97 gold .......................... 58. 102 hematitic coatings . .......... 47. 97 indium ................................. 46. 64 lead ................................. 49. 51, 102 109 110 Candle quadrangle—Continued Page nickel ................................. 49. 59. 97 radioactivity ..... . 25. 26. 27. 47. 97. 99. 102 sample localities ..... 47 silver ......................... 47. 58. 102 statistical procedures ................... 47 thallium ............................... 46. 64 zinc .............................. 49. 51. 97. 102 Canyon Creek, silver ........................ 38 zinc .................................... 34 Cape Creek, bismuth ................. 40. 100 zinc ........................... 36. 100 Cape Darby ....................... 64 Carbonate minerals ......................... 89 Caribou Creek. nickel ....................... 42 Caribou Creek gold placer. gold ..... 38 Cassiterite ...................... . 87 Cache Creek ................. . . 43 Clear Creek ............................. 80 Fairbanks Creek gold placer ............. 45 Seward Peninsula ............ .. 94 Catawba County ............................ 16 Celsian ........... Chalcopyrite .................. Cheenik Creek. cobalt .......... nickel ..................... . Chemical digestion ......................... 89 Chicken Creek ............................. 39. 40. 44 Chicken district. elements . . . . . . 36 Chlorite ............... 47. 87. 90 Chromite ......................... 101 Granite Creek gold placer ................ 44 Platinum Creek placer ................... 34 Chromium ........................ 89 Chicken Creek ......... 40 Dime Creek ............. 44 Chugach Mountains. cobalt . ........ 42 Cinnabar .......................... 44. 86 Circle quadrangle ................ 23 bismuth ................................ 40 copper ................................. 36 equivalent uranium . . .......... 27 gold .......................... 39, 100 indium ...................... 46 nickel .................................. 45 radioactive magnetic concentrates ........ 25 radioactivity ........................... 25 silver .................................. 39, 100 zinc .................................... Clear Creek ......... bismuth . . . minerals ........ . . . . radioactivity .......................... 74. 76. 99 rare earths ............................. 74. 76 scheelite ...... 80 silver ......... 77 Cleary Summit area. elements ............... 36 Coatings ........................... 28. 30, 47. 86, 89 test for radioactivity .................... 30 See also Hematitic coatings. Cobalt ........................... 3. 22. 40, 47, 49, 59. 64. 74. 75. 80. 83. 90. 92. 95 Albert Creek ........................... 43 analysis ............................ 2, 16. 18. 20 Arolik River ................ 43 Bear Creek gold placer .................. 44 Bendeleben quadrangle .................. 43. 100 Cache Creek ........................ .. 43 Candle Creek .......................... 44 Candle quadrangle . ............... 49. 59. 97 Cheenik Creek .................... 81 Chicken Creek .................... 40 Chugach Mountains ............... . 42 Darby pluton ........................... 72 Devonian limestone ..................... 80 Dime Creek .................... 44. 59 east-central Alaska ............. 45 Fairbanks Creek .............. . 45 Flat Creek .............................. 44 Glen Gulch ............................. 44. 100 INDEX Cobalt—Continued Page Granite Creek gold placer ................ 44 Hagemeister Island quadrangle .......... 43 Hatchet Creek .......................... 43 Iditarod quadrangle . . . ................ 44 Kachauik pluton . . . ................ 77. 80 Kachauik River ......................... 82 Knik Arm .............................. 33 Kwiniuk River .......................... 44, 81 Lake Clark quadrangle .................. 43 McGrath quadrangle .................... 44 Mount Hays quadrangle ................. 42. 99 Norton Bay Indian Reservation .......... 81. 82 Rainbow Mountain ................. .. 42 Ruby quadrangle ....................... 100 Slate Creek ............................. 42 Solomon quadrangle ..... . 72, 77. 80. 97. 98 southern Alaska ........................ 42. 99 southwestern Alaska .............. . 43 Talkeetna Mountains quadrangle ......... 43 Talkeetna quadrangle ................... 43 Tok River ...................... . 45 Tubutulik River ........................ 82 west-central Alaska ............... . 43. 100 Willow Creek ................. 44 Cobalt-nickel ratios ............... 59 Columbium .................... 77 Clear Creek . . . 74 Darby Mountains ....................... 65 Solomon quadrangle . . . . 65. 74. 76 Vulcan Creek ............... 76 Composition. Candle quadrangle . 47 insoluble residues ....................... 89 magnetic concentrates .................. 83 Solomon quadrangle . . . . ..... 64 Composition variation tests . ......... 17 Contaminants. metallic .............. 92 Control samples .................. 3. 16. 18 sources ................................ 16 Copper .......................... 3. 22, 31. a7, 47. 49. 51. 69, 71. 74. 83. 86. 92 analysis ............................ 2. 16. 18. 20 Bear Creek ........... . . 53. 58. 64 Bendeleben quadrangle . . ............. 77 Bethel quadrangle ................... 34 Bonanza Creek gold placer ............... 35 Candle Creek gold placer . . . . ....... 35 Candle quadrangle ...... . . 49. 51 . 97. 102 Circle quadrangle ....................... 36 Clear Creek ............................. 76 Cleary Summit area . ............... 36 Cripple Creek . . , ................. 34. 99 Crooked Creek . . ................. 34 Darby Mountains ....................... 65. 77 Darby pluton ...................... 53, 65. 72. 77 Devonian limestone 72 Dime Creek . . . . 58 Eagle quadrangle ....................... 36 east—central Alaska ..................... 36 Fish Creek ............................. 33 Flat Creek . . , 35 Glacier Peak ............................ 34 Glen Gulch ............................. 100 Granite Mountain pluton ......... Greer Gulch ..................... H. C. Carstens mine .............. . . Hagemeister Island quadrangle .......... 34 Hatchet Creek .......................... 43 Hidden Creek .................. 35 Horn Mountains ........................ 34 Hunter Creek pluton .................... 58. 102 Iditarod quadrangle ..................... 35, 100 Iliamna Lake ........................... 34, 99 Iliamna quadrangle ...... . 34. 99 Kachauik pluton ........................ 53 Ketchikan quadrangle ................... 33. 99 Knik Arm ............... . 33 Kwiktalik Mountains 77 Kwiniuk River ........... 81 Lake Clark quadrangle .................. 34 Copper—Continued Page Last Chance Creek ...................... 34 Little Lake Clark ................. McCarthy quadrangle ............. McGrath quadrangle .............. Marvel Creek .......................... 34 43. 99 Medfra quadrangle ...................... 35 Millets prospect ................ 34, 38 Mount Hays quadrangle... 34 Norton Bay Indian Reservation 77. 82 Norton Bay quadrangle ................ 35 Otter Creek ............................ 35 Platinum Creek placer ......... . 34 Poorman Creek ......................... 100 Portage Creek ......................... 36. 40. 45 Rainbow Mountain copper deposits . . . . 34 Rhode Island Creek ......... 45 Roundabout Mountain ...... .. . 35 Ruby quadrangle ....................... 35. 100 Russian Mission quadrangle ............. 34 Salmon River ............... 32 Slate Creek area ................. 38 Solomon quadrangle ........ . 65 69, 76, 98 South Fork Forty'mile River . ....... 36 southeastern Alaska ........... 32. 99 southern Alaska ............. 33. 99 southwestern Alaska .................... 34, 99 Spruce Creek ........................... 34 Tartana quadrangle ............ 36 Tok River ................... 45 ’l‘ubutulik River ............. 83 Vulcan Creek ........................... 77 west-central Alaska ..................... 35. 100 Willow Creek ......... 35 Copper sulfide minerals . . . ........... 49. 91. 96 Correlations, negative . . . ............. 96 positive ................................ 95 See also Accessory minerals. Cripple Creek. elements ..................... 34. 99 Crooked Creek. elements .................... 34. 46 Crystallization ............................. 102 Crystals, silver chloride . 89 Cuprite .................... 87 D Dahl Creek gold placer. gold ................. 38 Dan Creek, silver ........................... 37 Darby Mountains . bismuth . . . . cadmium ... columbium . . copper ................................. 65. 77 gold deposits ........................... 40 lead ............................ 35. 77 radioactive magnetic concentrates ........ 26 radioactivity ........................... 65 rare earths ...................... 65 tin .............................. 65 tungsten . . . 65 zinc .................................... 35. 77 Darby pluton ............................. 64, 69. 71 bismuth . . ........ . . 80. 83 cobalt . . 72 copper. .. 53. 65, 72. 77 indium ................................. 83 lead .................................. 53. 73. 77 nickel . . . . 72. 80 thallium 83 thorium . 67 zinc ..................... 74, 77 Data handling ............... 3 Deposits. base-metal ......... . 33. 34. 35. 102 cassiterite .............................. 94 contact pneumatobytic .................. 22 fluvioglacial ............... 43 glacial .................... 43 gold, Darby Mountains ..... . . . . 40 Dutch Hills ......................... 34 Deposits—Continued Page gold—Continued Lake Clark ......................... 35 Mount Holmes . 34 Young Creek . . ................. 34 hydrothermal ............ 22 Kennicott copper ...................... 33, 38, 46 Nikolai Butte copper .................... 33. 46 polymetallic mineral .................... 95. 101 Sheep Mountain copper ................. 33 skarn .................................. 22 sulfide ................................. 23. 97 Devonian limestone. cobalt .................. 80 copper ................................. 72 73 80 74 Diadochic substitution ...................... 21. 49 Diadochy, defined .......................... 21 Differentiation. magmatic ................... 22 Digestion, chemical Dime Creek. elements . Diorlte, hybrid ...................... 47. 64. 67. 77. 80 pyroxene ............................... 44 quartz ....................... 42 Distribution. metals . ................ 98 Dry Canyon stock .. . ................ 64 Dry Creek. minerals .................... 30 ' . 42 Dutch Hills. zinc ............................ 34 E Eagle Creek, gold . 58 Eagle quadrangle . . 23 bismuth ................................ 40 copper ................................. 36 gold . . . lead . .. silver . . . . Economics. mineralization .............. Eklutna ............................... Electronegativity, to predict substitution . . Elements. coherent ......................... 95 radioactive ........................... 95. 98, 102 See also specific elements. Epidote ........ Equivalent uranium ................. 27. 49. 51, 66. 97 See also Radioactivity. Uranium. Erosion, glacial ............................. 42 Exploration, geochemical ......... 2, 17. 18, 20. 95. 101 Extraterrestrial material .................... 88 F Fairbanks Creek. cobalt ..................... 45 nickel .................................. 45 Fairbanks Creek gold placer. minerals ........ 45 Fairbanks quadrangle .................. Fairview Mountain, gold ............... Feldspar ................ . .. Fish Creek, elements ....................... 33, 37, 40 Flat ....................................... 35, 40 Flat Creek, elements . . 35. 44 Flint Creek placer ........................... 35 Fluorite, H. C. Carstens mine ................ 27 Portage Creek ..................... 27 Fluvioglacial deposits .................. 43 Fly ash ............................... 88 Forest fires ............................... 88 Fortymile River, bismuth ................... 40 Faster prospect ....................... A 77 Fourth of July Hill .......................... 45 G Gabbro ................................... 42. 43. 44 Galena ................................ 23. 45. 46. 96 Bear Creek ............................. 58 Fairbanks Creek gold placer ............. 45 INDEX Page Garnet ................................... 87. 90. 99 Fairbanks Creek gold placer ............. 45 H. C. Carstens mine ..................... 27 Portage Creek .......................... 27 Geochemical exploration ............ 17. 18, 20, 95. 101 Glacial deposits ..................... 43 Glacial erosion .................... 42 Glacial trough ....................... 43 Glacier Peak, elements ...................... 34 Glen Gulch. elements ....................... 44, 100 Gold ................ . 36‘, 58. 64, 77. 86. 94. 102 analysis ............................... 19. 20, 65 Arolik River ............................ 38 Atwater Bar ................... 39 Bear Creek ...................... 58 Bear Creek gold placer . ............ 38 Bethe] quadrangle ...................... 36. 38 Bonanza Creek ......................... 38 Bowman Cut .................. 38 Cache Creek ................... 43 Candle quadrangle ................. 58, 102 Caribou Creek gold placer ................ 38 chemical digestion ...................... 89 Chicken Creek ..... Chicken district . . . . Circle quadrangle . . Cripple Creek ........................... Dahl Creek gold placer .................. 38 Darby Mountains ...... 40 Dime Creek ............................ 44 Dutch Hills ............................ 34 Eagle Creek ................. 58 Eagle quadrangle . . . . ........ 39. 100 east-central Alaska .............. 39. 100 Fairbanks Creek gold placer ............. 45 Fairview Mountain ..................... 34 Flat Creek ................... 35 Greer Gulch ....................... 40 H. C. Carstens placer gold mine .......... 40 Hidden Creek .................... 35 Humboldt Creek ............. 40 Kachauik pluton ............. 77 Lake Clark ............................. 35 Lake Clark quadrangle .................. 38 Livengood quadrangle . . . . ...... 39 Long Creek placer deposits . ..... 38 McCarthy quadrangle ...... .. . . 99 McGrath quadrangle .................... 38 Marvel Creek ........................... 43. 99 Marvel Creek gold placer . 38 Melba Creek ............... 40 Millets prospect ........... . . .. 33 Mills Creek ............................. 38 Mount Hays quadrangle ................. 38 Mount Holmes ............ .. 34 Myers Fork ............................ 39 native ....................... 22. 86. 90, 91, 94. 96 Otter Creek ................... 35 particulate .................. 92, 94 Platinum Creek placer ........ 34 Portage Creek .......................... 38, 40 Rainbow Ridge ......................... 42 Rhode Island Creek .......... 45 Russian Mission quadrangle . . 38 Slate Creek area ............... .. 38 Snow Gulch gold-platinum placer ......... 38 Solomon Creek ......................... 44, 87 Solomon quadrangle ......... . 65, 77 South Fork Fortymile River ............. 39 southeastern Alaska .................... 36 southern Alaska ............. southwestern Alaska ......... Talkeetna quadrangle ........ .. Tanana quadrangle ..................... 100 Vinasale Mountain ...................... 35 Wade Creek ................. . . 39 Wattamuse Creek gold placer ............ 38 west-central Alaska ..................... 38 Young Creek ........................... 34 111 Page Gold placer districts ........................ 3, 34 Golovnin Bay, minerals ....... . . . 65 Goodnews quadrangle ........ 23. 34 indium ................................. 46 nickel .................................. 43 silver . 38 Gossan .......................... 74. 75 Granite Creek gold placer. cobalt . . . . 44 nickel .................................. 44 Granite Mountain pluton .................... 51, 102 bismuth ...................... copper . equivalent uranium ..................... 51 lead .................... 53, 58. 102, nickel ....................... . 59 radioactivrty ................. Greenstone. gabbroic ............. Greer Gulch. elements ...................... 40 Grouse Creek gold placer. lead ............... 35 H H. C. Carstens mine. elements ............... 40 minerals ............................... 27 Hagemeister Island quadrangle . . . . . 23 bismuth ................................ 39 cobalt .................................. 43 copper ................ 34 nickel ................... 43 Hampton Creek. control sample . . . . 7. 16 Hatchet Creek. elements .................... 43 Hematite ............................. 28. 69. 86, 87. 88, 89, 90. 94 Dry Creek .......................... . A 30 exogenetic ................... 95 Jump Creek ............................ 31 Nixon Fork mining district .............. 29 Hematitic coatings ................. 47, 69. 86. 87. 90. 91. 94, 95. 97. 98 chemical digestion ..................... 89 Hematitic crusts ................... 69 Hidden Creek. elements ............. 35 Hopeite ............................ 87 Horn Mountains, elements .................. 34 Homblendite ............................... 42 Host minerals ..... . 21.23.89 determination .................. 89 Humboldt Creek, elements .......... . 40 Humboldt National Forest. Nevada .......... 16 Hunter Creek pluton ........................ 51, 102 bismuth ........................ . 59 copper ................................. 58. 102 equivalent uranium ..................... 51 indium ......................... 64 lead . . 102 nickel . . 59 silver .................................. 58 zinc .................................... 102 Hyder Quartz Monzonite ............ 27 minerals ....................... 28 Hydrous iron oxides ................ . 95 Hydroxides. iron ........................... 94 I Iditarod quadrangle ............ bismuth .................... 112 Page Iliamna Lake. copper ................. 34. 99 zinc ........................... . 34 lliamna quadrangle ......................... 23. 38 copper ................................. 34. 99 indium . ................ 46 zinc . . . . . . .34 Ilmenite ............... . . 83. 86'. 90 chemical digestion ...................... 89 Fairbanks Creek gold placer ............. 45 Indium ......................... 23. 45. 64. 83. 94 analysis . . ....................... 2. 19. 20. 45 distribution ........................... 99 locations ...................... 46 Innoko district. gold. native . . 23 Ionization potential .......... . 21, 22 Iron ............................. 20. 89 Iron hydroxides . . . Iron meteorites .................. . . Iron ores. magmatic ........................ 42 Iron oxides. radioactive uraniferous .......... 30 Iron sulfate ................................ 87 J. K 30 . . 31 Juneau quadrangle ......................... 23. 32 Kachauik pluton ....................... 64. 67. 73, 77 cobalt. .. 77. 80 copper . . . . .. 53 lead ................................... 53, 73 nickel .................................. 77, 80 Kachauik River. cobalt . ...... 82 lead ............... . r . . 74. 77, 82 nickel ..................... 82 Kennicott copper deposits ................... 33 indium ................................. 46 38 46 Ketchikan quadrangle ...................... 23 cadmium .............. 33 33. 99 equivalent uranium .. . . 27 lead ................................... 33 radioactive magnetic concentrates ........ 25 radioactivity ........................... 25 silver .................................. 37. 99 zinc .................................... 33 Knik Arm. elements ..... 33 Kuskokwim River. zinc . . . 34 Kwiktalik Mountains 69 copper ................................. 77 zinc .................................... 77 Kwiniuk River ..................... 67. 68 cobalt ......................... 44. 81 copper ...................... 81 nickel .................................. 44. 81 74 81 82 L Labrador. exploration, geochemical .......... 2 Lake Clark ................................. 43 copper ................................. 34 gold .......................... 35 Lake Clark quadrangle . .............. 16, 23 .............. 43 34 38 43 34 Laser-probe analysis ........................ 88 Last Chance Creek. elements ................ 34. 42 Lava field .................................. 44 INDEX Page 89 lead ................... 3. 22. 23. 31. 37. 39. 49. 51, 64, 69. 71. 74. 75. 83. 90. .92. 94. 102 analysis ......................... 2. 16. 18. 19. 20 Bear Creek ............................. 58. 64 Bendeleben quadrangle .............. 35. 77 Candle Creek gold placer . 35 Candle quadrangle .................... 49. 51. 102 Chicken district ......................... 36 Cleary Summit area ................. 36 Crooked Creek ...................... 34 Darby Mountains . 35. 77 Darby pluton .......................... 53. 73. 77 Devonian limestone ..................... 73 Dime Creek ........................ 58 Eagle quadrangle ....................... 36 east-central Alaska ..................... 36 Fish Creek ......................... . . 33 . . 35 Flint Creek placer ................... . . 35 Glacier Peak ............................ 34 Glen Gulch ............................. 100 Granite Mountain pluton . . .......... 53. 58. 102 Grouse Creek gold placer ............... 35 Horn Mountains .................... .. 34 Hunter Creek pluton .................... 102 Iditarod quadrangle ..................... 35, 100 Kachauik pluton .................. . 53. 73 Kachauik River ........................ 74. 77. 82 Ketchikan quadrangle ................... 33 Last Chance Creek ................ 34 McGrath quadrangle .............. 35 My Creek ........................ . 36 Otter Creek ............................ 35 Poorman Creek ......................... 100 Portage Creek .................... 36 Portage Creek zinc lode ........ 46 Rainbow Mountain copper deposits . . 34 Rainbow Ridge ......................... 42 Rhode Island Creek ..................... 36. 45 Roundabout Mountains ......... . 35 Ruby quadrangle ....................... 35. 100 Russian Mission quadrangle ............. 34 Salmon River ................... Solomon quadrangle . . . southeastern Alaska . . .. .. . southern Alaska ........................ 33 southwestern Alaska .................... 34 Spruce Peak .................... . 34 Tanana quadrangle ..................... 36 west-central Alaska ..................... 35. 100 Willow Creek ................. 35 Left Fork. thallium ............... 64 Lightning ...................... . 88 Lignite .................................... 43 Limonite ................................... 28. 88 Dry Creek ................ 30 H. C. Carstens mine ....... 27 Portage Creek ............ 27 Little Lake Clark. copper .................... 35 Little Moose Creek. silver ................... 38 Little Susitna River ............. 37 base-metal deposits ............. 33 Livengood quadrangle .............. 23 base metals .................. 36 bismuth ...................... 40 gold . .. .. . 39 nickel .................................. 45. 100 silver .................................. 39 zinc .......................... . 100 Long Creek placer deposits. elements ......... 38 Lyle Creek. North Carolina, control sample . . . . 16. 18 M McCarthy quadrangle—Continued Page silver ...................... 37 zinc ......................... 33. 46. 99 McGrath quadrangle ........................ 23 base metals ............................ 46 cadmium ................... 35, 46. 96 cobalt ................................. 44 copper ................................. 35 gold ....... 38 lead ....... 35 zinc ................................... 35 Magmatic differentiation .................... 22. 41 Magnesium ................................ 89 Magnetic fractions . ............ 2, 3. 18. 30. 69. 83 Malachite .. .. 87. 91 Manganese .. . . l 89 Marble. schistose ........................... 72. 80 Marcasite .................................. 86, 87 Marvel Creek. elements ........ . . 34. 43. 99 Marvel Creek gold placer, gold ............... 38 Medfra quadrangle ......................... 23 bismuth ......................... 40 copper .......................... 35 equivalent uranium .............. . 27 radioactive magnetic concentrates ........ 25 radioactivity ........................... 25 Melba Creek. bismuth ................ 40 gold .. . ................ 40 quartz ................ 40 Mercury ................................... 89 Flat Creek .............................. 35 Otter Creek ................... . . 35 Rhode Island Creek ..................... 36, 45 Metadiorite ................................ 45 Metallic contaminants . .. .. .. 92 Metallic spherules ...... . . . 86. 87. 90. 91 chemical digestion . .. . . . . . . 89 contaminant ........................... 94 sources ................................ 88 Metals. base ............. 20. 33. 34, 35. 46. 58. 72. 102 Metamorphism ...................... 102 Metasiltstone ................... . 71 Meteoric dust .............................. 88 Meteorites. iron ............................ 88 Method. abundance determinations . . . 16 Method of study ............................ 16 Mica ...................................... 87. 90 Millets prospect. elements ...... 34. 38 Mills Creek. gold .............. 38 Mine. H. C. Carstens ........... .. 27. 40 Omilak ................................. 77 placer .................................. 32, 89 Mineralization .............. . 2. 31 base—metal ................. 32 Mineralogical examination analysis . 2 Minerals. accessory ................. 22, 45. 49. 83. 87. 90. 94. 96. 100. 102 basemetal sulfide ...................... 43 carbonate .............................. 89 columbium-bearing, Clear Creek .......... 80 copper carbonate .................. 91 copper sulfide . .. ...... 49. 96 host ........... .. . 21. 23, 89 minor ................................. 22.45.49 radioactive ............................. 77 rare-earth.... ..65.77.80 secondary ............. .. .. .. 87 silicate ................ .. 86.87. 90.91 chemical digestion .................. 89 sulfide ............... 77. 83. 86. 92. 94. 95. 96, 101 chemical digestion ................. 89 thorium ................................ 29 trace ................................... 22, 100 uraniferous titanium niobate ............. 69 See also specific minerals. Monzonite ................................. 47 Hyder Quartz .......................... 27 Mount Hays quadrangle .................... 23 cobalt ........................... 42. 99 copper ................................. 34 Mount Hays quadrangle—Continued Page equivalent uranium ..................... 27 gold ........................... 38 nickel ............................ .. 42. 99 radioactive magnetic concentrates ........ 25 radioactivity .. 25, 30 zinc .................................... 34 Mount Holmes. deposits, base-metal ......... 34 gold ................................... 34 Mount McKinley quadrangle 23. 87 base metals ........ . 34 nickel ....................... 42 silver .................................. 38 Mountain. Appel ........................... 28 Chugach ...... 42 Darby ................... 26. 40. 65. 67. 74. 77. 80 Kwiktalik ............................. 69. 77 Rainbow .............. 34. 42 Talkeetna ............... 33. 37 Vinasale ................ . 35 My Creek, antimony ........................ 36 lead ................................... “ 36 Myers Fork. gold .............. . . 39 silver ....................... . . 39 Multielement highs ......................... 98 N Nabesna quadrangle ........................ 23. 34 Native gold ..................... 22, 86. 90, 91. 94. 96 Nevada. control sample . . . . 3. 16 distribution. magnetites ................. 2 Nickel .............. 3. 22. 40. 49. 59. 64. 74. 75. 79. 80. 83. 87. 90. 92. 94. 95 analysis ......................... 2. 16, 18. 19. 20 Arolik River ............................ 43 Bear Creek gold placer .................. 44 Bendeleben quadrangle .................. 43. 100 Bethe] quadrangle ...................... 43 Cache Creek ... . 43 Candle Creek ..... 44 Candle quadrangle ..................... 49. 59. 97 Caribou Creek .......................... 42 Cheenik Creek ................. 44. 81 Circle quadrangle .. . ......... 45 Darby pluton .................. 72. 80 Devonian limestone ..................... 80 Dime Creek ............................ 44. 59 distribution ............. 99 east-central Alaska . ............. 45. 100 Fairbanks Creek . . . . ............. 45 Flat Creek .............................. 44 Glen Gulch ............................. 44. 100 Goodnews quadrangle . . . ......... 43 Granite Creek gold placer . . . . ......... 44 Granite Mountain pluton ....... . . . 59 Hagemeister Island quadrangle .. ...... 43 Hatchet Creek .......................... 43 Hunter Creek pluton .............. . 59 Iditarod quadrangle ..................... 44. 100 Kachauik pluton ........................ 77. 80 Kachauik River .. 82 Knik Arm ..... 33 Kwiniuk River .......................... 44. 81 Lake Clark quadrangle .................. 43 Last Chance Creek ...................... 42 Livengood quadrangle ................... 45. 100 Marvel Creek ........................... 43 Mount Hayes quadrangle ................ 42. 99 Mount McKinley quadrangle ............ 42 Norton Bay Indian Reservation .......... 81. 82 Poorman Creek ......................... 100 Portage Creek .......................... 45 Rainbow Mountain ..................... 42 Rainbow Ridge .............. 42 Rhode Island Creek ...... . . . . . 45 Ruby quadrangle .......... . 44. 100 Slate Creek ....................... 42 Solomon Creek .. ................. 44 Solomon quadrangle ............. 72. 77. 80. 97. 98 INDEX Nickel—Continued Page southern Alaska ........................ 42. 99 southwestern Alaska .................... 43 Spruce Creek gold placer ......... 44 Talkeetna quadrangle .............. 43 Tanana quadrangle ................ . 45 Tubutulik River ........................ 44. 82 west-central Alaska ..................... 45, 100 Willow Creek ..... 42. 44 Wolverine Creek ........................ 42 Nikolai Butte copper deposit ................. 33 indium .............................. 46 zinc ............................. 46 Nixon Fork mining district . . ........... 28 minerals ............................... 29 Nome quadrangle ........................... 23. 35 North Carolina. control sample ............... 3. 16. 18 Norton Bay Indian Reservation. cobalt ....... 81, 82 copper ................................. 77. 82 nickel .................................. 81. 82 zinc .................................... 77. 82 Norton Bay quadrangle ....... 23. 38 copper ..................... 35 radioactive magnetic concentrates . . . . . 25 radioactivity ........................... 25, 26 O. P Omilak mine ............................... 77 Ophir Creek gold placer. indium ........ 46 Otter Creek ............................ 44 elements ........................... . . 35 Otter Creek tin placer. cadmium ............. 35 zinc .................................... 35 Oxides. hydrous iron . 95 Oxygen ions ................................ 21 Particulate gold ............................ 92. 94 Peridotite ......................... 42 Peters Hills. zinc ................ 34 Phlogopite ................................. 87 Phyllite .................................... 45, 64 Placer concentrate file. Alaskan . . . 2. 3. 16. 27 Placer mines ................... . . . . 32. 83. 89 Platinum. Cache Creek ....... . . . . 43 detrital ................................ 43 Dime Creek ............................ 44 Long Creek placer deposits . . . 38 Platinum Creek placer ........ 34 Platinum Creek placer. minerals . . . . . 34 Platinum placers ........................... 43 Pluton. Darby .................. 53. 64. 65. 67. 69. 71, 72. 73. 74. 77. 80. 83 Granite Mountain ........ 51, 53. 58. 59. 64. 99. 102 Hunter Creek .................. 51, 58. 59. 64. 102 Kachauik .................... 53.64.67. 73. 77.80 Poorman Creek. elements .................... 100 gold and tin placers . . 35 Porphyry .................................. 64 Portage Creek, control sample ............... 16 elements ..................... . 36. 38. 40, 45 H. C. Carstens mine ................... 27 minerals .......................... 27 rare earths ...................... 45 Portage Creek gold placer. gold . . ......... 38. 39 silver ........................... 38, 39 Portage Creek zinc lode, indium .............. 46 lead ................................... 46 Pyrite .............. .. 22, 45, 86.87 Bear Creek ............. . . . . 58 Pyroxenite ..................... . . . 42 Pyrrhotite ................................. 22 Q Quadrangle. Anchorage .............. 23. 33. 37. 42. 99 Bendeleben .............. 23. 25. 26. 30, 35. 38. 40. 43. 46, 65. 74. 77. 80. 99. 100 Bethel ............................ 23. 34. 36. 38. 39. 43. 46. 99 113 Quadrangle—Continued Page Bradfield Canal ......................... 23. 32 Circle .......................... 23, 25. 27. 36. 39. 40. 45. 46, 100 Eagle ......................... 23. 36. 39. 40, 100 Fairbanks ....................... 23. 36 Goodnews ................. 23, 34. 38. 43. 46 Hagemeister Island . . ........... 23. 34. 39, 43 Iditarod ................ 23. 35. 38. 40. 44, 96. 100 Iliamna ........................ 23. 34. 38. 46. 99 Juneau.... ........... 23.32 Ketchikan ........... 23. 25. 27.32. 36, 99 Lake Clark. . . .......... 16. 23. 34. 38. 43 Livengood .................. 23, 36. 39. 40. 45. 100 McCarthy ...................... 23, 33. 37. 46. 99 McGrath . .............. 33, 35, 38. 44. 46. 96 Medfra ...................... 23. 25. 27. 28, 35. 40 Mount Hays .......... 23. 25. 27. 30. 34. 38, 42. 99 Mount McKinley ................ 23. 34. 38. 42. 87 Nabesna . 23. 34 Nome ......................... 23. 35 Norton Bay ................ .. 23. 25. 26,35. 38 Ruby ............. 23. 35. 38. 40. 44. 78. 87. 96. 100 Russian Mission ................ 23. 34. 35. 38. 46 Talkeetna ................... 23. 33. 34. 38. 43. 99 Talkeetna Mountains . . . 23. 34. 39. 43. 46. 99 Tanacross ............................. 23, 36. 45 Tanana ........................ 23. 36. 39. 45. 100 Teller.... .. 23.35.40.100 Valdez ................... 23. 34. 99 Quartz ............... . 86. 87. 90. 91. 94 chemical digesstion ..................... 89 Melba Creek ............................ 40 Quartz diorite ................... 42 Quartz lodes .......................... 42 R Radioactive magnetic concentrates ........... 25. 26 Radioactive minerals. See Equivalent uranium. Radioactivity. Uranium. Radioactivity ................... 16. 29. 47. 49, 51. 67. 69. 80. 94. 97. 98 Bendeleben quadrangle .................. 25. 99 Candle quadrangle ....... 25. 26, 27. 47. 97, 99. 102 Circle quadrangle . . . 25 Clear Creek ........................... 74. 76. 99 Darby Mountains ....................... 65 Golovnin Bay .......... 65 Granite Mountain pluton . . 99 Ketchikan quadrangle . . . 25 Kwiniuk River .......................... 74 Medfra quadrangle ...................... 25 Mount Hays quadrangle . Norton Bay quadrangle ................. 25. 26 Seward Peninsula ....................... 31 Solomon quadrangle ...... 25. 26. 27. 65. 76. 98, 99 source ................................. 90 Vulcan Creek ........................... 76 See also Equivalent uranium. Uranium. Radiometric counting analysis ............... 2, 20 Rainbow Mountain. cobalt ......... . . . 42 nickel .................................. 42 Rainbow Mountain copper deposits. ele- ments .......................... 34 Rainbow Ridge. elements .................... 42 Randomization ....................... 3 Range Province ...................... 2 Rare earths ............................ 77 Clear Creek . . .. .................... 74, 76 Darby Mountains . .................. 65 Golovnin Bay .................. 65 Portage Creek .......................... 45 Rock Creek ............................. 74 Solomon quadrangle . 65. 7 4. 76 Vulcan Creek ................... 76 References .......................... 103 Reproducibility tests ....................... 17 Rhode Island Creek, elements ............... 36. 39. 45 114 Page Rock Creek. indium ......................... 46 rare earths ............... . . . 74 thallium ............................... 46 Rocks. alkaline ............................. 26 chromite-bearing ............ 33 metavolcanic .................. 80 volcanic .................. . . 43. 47. 51. 59 Roundabout Mountain. elements ............. 35 Ruby quadrangle ........................... 23. 87 bismuth ............ . . . . 40. 78. 100 cadmium .................. 96. 100 cobalt ....................... 100 copper ................................. 35. 100 zinc .................................... 35. 100 Russian Mission quadrangle . ......... 23 base metals ................... 35 copper ........................ 34 38 46 34 16.83.819.90 Fairbanks Creek gold placer ............. 45 S Salmon River. minerals .................... 32, 33, 99 Sample localities, Candle quadrangle ......... 47 Solomon quadrangle .................... 65 Scheelite. Cache Creek .. 43 Candle Creek placer . 44 Clear Creek ............................. 80 H. C. Carstene mine ..................... 27 Portage Creek ......................... 27 Semiquantitative spectrographic analysis . . . . 2. 65. 89, 91, 92 Serpentinite .............................. 45. 65. 87 Seward Peninsula ........................... 31 elements ..... 78. 94 minerals ............................... 94 Sheep Mountain copper deposit .............. 33 Silicate minerals ..................... 45. 87. 90. 91 chemical digestion . ............... 89 Silicon .............................. 89 Siltite ..................................... 43 Silver ..................... 3. 22. 36, 46. 58. 64. 77. 83. 86. 90. 92. 94 analysis ...................... 2. 16. 17. 18. 19. 20 Archangel Creek ....................... 37 Aron River . . . . ................. 38 Atwater Bar . . ................. 39 Bear Creek ............................. 58 Bear Creek gold placer .................. 38 Bethel quadrangle ...................... 36 Bonanza Creek area ................. 38 Bowman Cut . .. ................... 38 Candle quadrangle . ........... 47. 58. 102 Canyon Creek .......................... 38 Chicken Creek .......................... 39. 40 Chicken district . . 36 Circle quadrangle ....................... 39, 100 Clear Creek ............................. 77 Cripple Creek . ................ 99 Dan Creek . . .................. 37 Darby pluton ...................... . 77 Eagle quadrangle ....................... 39. 100 east-central Alaska ..................... 39. 100 Fish Creek ...................... 37 35 Glen Gulch ...................... 100 Goodnews quadrangle .............. 38 Greer Gulch ....................... 40 Hunter Creek pluton ............. .. 58 Iditarod quadrangle ..................... 38. 100 Kennicott copper mines ................. 38 Ketchikan quadrangle ............ .. 37. 99 Little Moose Creek ...................... 38 INDEX Silver—Continued Page Livengood quadrangle ................... 39 McCarthy quadrangle ................... 37 Marvel Creek ........................... 43. 99 Millets prospect ........................ 38 Mount McKinley quadrangle ............ 38 Myers Fork .. ................... 39 Otter Creek . . ................... 35 Poorman Creek ......................... 100 Rainbow Ridge ......................... 42 Rhode Island Creek ................ 39 Slate Creek ............................. Snow Gulch gold-platinum placer . . 38 Solomon quadrangle ............. 77 South Fork Fortymile River ...... . . 39 southeastern Alaska .................... 36. 99 southern Alaska ........................ 37. 99 southwestern Alaska ........... . 38. 99 Tanana quadrangle ............ . 39. 100 Tok River ..................... . . 45 Wattamuse Creek gold placer ............ 38 west-central Alaska ..................... 38. 100 Silver chloride crystals . . . . . 89 Slate .................. . . 43, 69. 87 Slate Creek. cobalt ..... 42 copper ...................... 38 gold ........................ 38 nickel .. .. 42 silver .................................. 38 Snake Range. northern ...................... 16 Snow Gulch gold-platinum placer. gold ....... 38 silver .................................. 38 Solomon Creek . . . . 35 gold placer ............................. 44. 87 nickel ............................ 44 Solomon quadrangle ............. 23. 64. 97. 102 atomic absorption analysis . . . 65 bismuth ................... . . 77. 98 cadmium ............................... 69. 76 cobalt .......................... 72. 77. 80. 97. 98 columbium . . . ............ 65. 74. 76 composition ................... 64 copper .............. 65.69. 76. 98 correlations ......... . ...... 96 equivalent uranium ...... 65, 97. 98 gold .......................... 65. 77 indium ................................. 46. 83 lead ............................... 69, 73. 97. 98 nickel ................ 72, 77. 80. 97. 98 radioactivity . . . . . 25. 26. 27. 65. 76. 98. 99 rare earths . . . ................ 65, 74. 76 sample localities ....... 65 silver ................. 77 spectrographic analysis ........ 65 thallium ............................... 46. 83 thorium ................................ 67 threshold determination .. ........ 65 tin ................................. 65.74.76 ........... 65. 74. 76. 81 ...... 67 zinc ............................... 69. 74. 82. 98 Solubility ......... 83 Sorption ................................... 21. 101 Sources .................................... 83 South Fork Fortymile River. elements ........ 36, 39 Sphalerite ......................... 23. 45. 95. 96 Bear Creek 58 Sphene ........... 28. 69 H. C. Carstene mine ..................... 27 Portage Creek .............. 27 Spherules. metallic ..................... 86. 87. 90. 91 chemical digestion .................. 89 contaminant .................... 94 sources .. 88 Spinel ............................... 87 Spruce Creek gold placer. nickel .............. 44 Spruce Peak. elements ...................... 34 Statistical procedures. Candle quadrangle 47 Statistical treatment ........................ 23 Page Stibnite .................................... 42 Substitution. diadochic .. . . . . 21. 22, 49 Substitution. ionic charge ................... 21 Sulfate. iron ................................ 87 Sulfide, base-metal. Darby Mountains . . . . . 74 copper ................................. 91 Sulfide deposits . 23 Sulfide minerals .............. 45. 83. 86. 92. 94. 95. 96 chemical digestion ..... 89 Surface coatings ........... 89 Sweden. magnetite ......................... 59 Sweepstakes Creek. zinc ..................... 58 Syenite ......................... 47. 64. 67. 77. 80. 99 T Talkeetna Mountains ....................... 37 base—metal deposits ..................... 33 Talkeetna Mountains quadrangle ............ 23 bismuth ................................ 39 Talkeetna quadrangle ....................... 23 cadmium ............................... 33 cobalt ........... 43 gold ............ 38 nickel ........... 43 zinc .................................... 34. 99 Tanacross quadrangle ..................... 23. 36. 45 Tanana quadrangle . . 23 copper ........ 36 gold .......................... 100 lead ............................ 36 nickel ........................... 45 silver ......................... 39. 100 Teller quadrangle ........................... 23 bismuth ................................ 40. 100 zinc ........... Tests. reproducibility ....................... variation in composition ................. 17 Texas Creek Granodiorite ................... 27 minerals ............................... 28 Thallium ........... 23. 45. 64. 83. 94. 96. 100 analysis ............................ 2. 19. 20. 45 Bear Creek ............................. 64 Candle quadrangle .................. 46. 64 Darby pluton ....................... 83 distribution ...................... 99 east-central Alaska ............... 46 Granite Mountain pluton .......... 64 Left Fork ...................... 64 Rock Creek ............................. 46 Seward Peninsula ....................... 94 Solomon quadrangle ............ 46. 83 southern Alaska ................ 46 southwestern Alaska ............ 46 west-central Alaska ..................... 46 Thorium ................................... 67 Thorium minerals ...................... 29 Threshold determinations . . ...... 24. 31. 36. 40. 49. 65. 67. 71. 75. 79, 80 Tiktites .................................... 88 Tin ............................. 38, 39. 40, 77.87.94 Chicken district . . . . 36 Clear Creek ............................. 74. 76 Darby Mountains ....................... 65 Fairbanks Creek gold placer . .. . 45 H. C. Carstens placer gold mine . . 40 Humboldt Creek .............. . . . 40 Long Creek placer deposits .............. 38 Portage Creek .......................... 40. 45 Rhode Island Creek ....... . .. 45 Seward Peninsula ....................... 94 Solomon quadrangle ................... 65. 74. 76 Tin City. bismuth ........... 40 zinc .................... 36 Titanium .................. 89 Tofty area. magnetite ....................... 45 \ n Page Tok River. elements ......................... 45 Topaz, H. C. Carstens mine . . . ' 27 Portage Creek ............. 27 Trace element. sorption ........ .. 2i Trace minerals ............................. 22. 100 Tramp iron ................... 44, 45, 49, 59,86, 87, 89 chemical digestion ............. 89 contaminant ........................... 94 sources ................................ 88 Tremolite ....................... 87 Trivalent ions ................... 23 Trough. glacial ...... i ........... . . 43 'l‘ubutulik River ............................ 77 elements .............................. 44, 82, 83 Tungsten ...................... 37. 38. 40, 77. 87, 101 Bonanza Creek gold placer ...... . . 35 Chicken Creek ............... . . 40 Chicken district ......................... 36 Clear Creek .......... . ................... 74, 76 Darby Mountains ............ 65 Fairbanks Creek gold placer ...... 45 Flat Creek ....................... 35 Golovnin Bay ............... 65 H. C. Carstens mine .......... 40 Hidden Creek ................ . . 35 Kwiniuk River .......................... 81 Otter Creek ............................ 35 Portage Creek ....... . 40, 45 Solomon quadrangle . . . 65, 74, 76, 81 Vinasale Mountain ...................... 35 U, V Uraniferous titanium niobate minerals ........ 69 Uranium .......... '. ....................... 27, 28, 67 Uranothorianite. H. C. Carstens mine . . 27 Portage Creek ........... . . .. 27 Utah, distribution, magnetites ............... 2 Valdez quadrangle ......................... 23, 34. 99 Veins. quartz-calcite ........... 58 Vinasale Mountain, elements . . . 35 Volcanic activity .............. .. 88 Volcanic glass .............................. 87 Volcanic rocks ......................... 43. 47. 51. 59 Vulcan Creek ............................... 76, 77 \ \ \ . INDEX Prise W, X, Y Wade Creek. gold ........................... 39 Wattamuse Creek gold placer. gold . . . . . 38 silver ................ , ........... 38 Weathering ................................ 22 Welding spatter, contaminants .......... . . .. 88 Willow Creek, elements .................... 35, 42. 44 Wolframite, Fairbanks Creek gold placer ...... 45 Wolverine Creek. nickel ..................... 42 X-ray diffraction ........................... 31, 83 Young Creek. deposits, base-metal ........... 34 gold ................................... 34 Z Zinc ....................... 3, 22. 23, 31, 37, 49. 51. 69, 71, 75, 86. 87. 90, 92. 94. 95 analysis ......................... 2. 16, 17. 18, 20 Bear Creek . .. . . Bendeleben quadrangle. . . . Bethel quadrangle ...................... Candle Creek gold placer ................. 35 Candle quadrangle .......... . 49, 51. 97, 102 Canyon Creek .......................... 34 Cape Creek ............................. 36. 100 Circle quadrangle ............. 36 Cleary Summit area ........... 36 Cripple Creek ................. . . , 34. 99 Crooked Creek .......................... 34, 46 Darby Mountains ....................... 35, 77 Darby pluton ................. 74, 77 Devonian limestone ............ 74 Dirne Creek .................... . 58 distribution ............................ 99 Dutch Hills ............................ 34 east-central Alaska ............. . Fish Creek ............................. 33 Flat Creek .............................. 35 Glacier Peak . 34 Glen Gulch ....................... 100 Granite Mountain pluton .......... . 58 Hatchet Creek .......................... 43 Horn Mountains ........................ 34 Hunter Creek pluton .................... 58, 102 ’ ’ . 115 / ._ Zinc—Continued Page \ Iditarod quadrangle .................... ~-. 35, T00 Iliamna Lake .................. 34 Iliamna quadrangle . ........... 34 Kennicott copper deposits . ........... 46 Ketchikan quadrangle ................... 33 Knik Arm .............................. 33 Kuskokwim River . . . 34 Kwiktalik Mountains ..... 77 Kwiniuk River .......................... 82 " Lake Clark .................... 34 Lake Clark quadrangle . . ......... 34 Last Chance Creek ............... 34 Livengood quadrangle ................... 100 McCarthy quadrangle .................. 33. 46, 99 McGrath quadrangle . . . . , . . 35 Marvel Creek ........... . . 34, 43. 99 Mount Hays quadrangle ..... . . 34 Nickolai Butte copper deposit ......... -. . .' 46 Norton Bay Indian Reservation .......... 77. 82 Otter Creek tin placer ............ 35 Peters Hills ....................... 34 Portage Creek .......................... ' 36 Rainbow Mountain copper deposits 34 Roundabout Mountain . . .. ..... 35 Ruby quadrangle .......... 35. 100 Salmon River ........................... 33 Solomon quadrangle ................ 69. 74, 82, 98 southeastern Alaska . . . . . . southern Alaska . . . . southwestern Alaska . . Spruce Peak ............................ 34 Sweepstakes Creek ...................... 58 Talkeetna Mountains quadrangle . . . 34. 46. 99 Talkeetna quadrangle ............ Teller quadrangle . .. 'I‘in City ................................ Tok River .............................. 45 'I‘ubutulik River . . ' 83 Vinasale Mountain ...................... 35 Vulcan Creek ........................... 77 west-central Alaska . Willow Creek ........................... Zircon .................................... 69. 87. 90 Dry Creek ........ 30 H. C. Carstens mine . ............... 27 Portage Creek .......................... 27 r.- us, GOVERNMENT PRINTING OFFICE 1980—677.I29/62 ’ GEOLOGICAL SURVEY Eu: at Greenwich 176° 180" UNITED STATES DEPARTMENT OF THE H‘ITERIOR 176° West. cl Greenwich 172" 168° 164“ 180° 158° 152° 148‘ l r II I L! us. DEPOSITORY 9- 51980 n PROFESSIONAL PAPER 1135 PLATE 1 Base from US. Geological Survey, 1946 200 MAP OF ALASKA SHOWING GENERALIZED OUTLINES OF AREAS OF CONCENTRATES AND TOPOGRAPHIC QUADRANGLE MAPS USED IN THIS STUDY SCALE 1:5 000 000 400 I—I 500 600 700 800 900 Kl LOMETERS 500 MILES REPRESENTED BY MAGNETIC FRACTIONS 144° 140° 136° 124° 120° W K 116“ W9; “3., 689 AM ‘v - , “mi-Jr/ \ \ ' A “P 7 l \ “7: r ”0M" “5! ’5? ”’"lldsmz (,u If £25,? I C Ba I. C E V.» EXPLANATION Frow ‘ T .. ~ 17‘, o 3 V. 5». "'"Ska 3 e D 9““, 'fi @721.” R “Mung," b ‘ 3W U“ ° 0% , xv“ Pm H O R T Generalized outlme of area furmshmg 1 g or larger magnetIc fractions of ° ‘_ I' . n V ‘ no D D g ‘3 ' . . ewes 4‘». Is, . . a t, O U . a fig ”(MB 3 MW A U F § .. concentrates for analySIs by atomlc absorptlon spectrophotometry: ) La‘ t‘ a " 0 <7 9 ‘5 hell ” . 1 IrrIson E {g A ‘w : ” \ ,, e 0 Artslfiu a a: r o a Q \ .0 v Q; a a \ D" 'd 0 I I a I Shh “on It on a Bay Ooliktok .\ QM“ ls \ S g g: Q, a 0 . r/ a . V \4 I A 'u/ “ a 0&5.” I “w 939%” $30 0 ° a° ° 0 “0° 9 6 PW” T! ‘ An IE“ '3 9 Ike“ WIN“ n E D ”(5" Q " on Q B ‘3 o ' o a . “ 'Q In‘ 06' ”L .' ., ." ° s b ”a I? e 'u' 2:31 5" -° ".°° I ,‘ ..- Flaxvranl w ” H°°“““’ Times“ '4' 1} ,° °o°°. ‘ a A- AmguyEma /A, o . a Q: ., y a e c: z o i f a . ,° .w ,o u may 90le P M\ g l ‘y ' $0. _I .034 Q I . 03 s u“. ' 7;: Q’ a . ,. ‘v o a g o I I , , A a _ V C H! I n°a° 3: ° II M . I “a ; '9 I u ° . - ° . T. ICY Reel 0 c: \jgscmH yi‘g 5‘ .V‘L‘fgw “ M: a a Q, ' I 3 ., V Index of 1:250,000-sca1e topographrc quadrangle maps of Alaska a v I 9 ”-’ v a .w ,th ‘ 4 ,. >\ Ga -. v o . ' . . ' ”‘ re s ’ £ ° 5 ” ° “4' o ”r ‘ “f .h .. . a a F 2 it D H "’¥“‘“"’\“ , l: 97v "“Cb :03)? I) J l) ” o ’ " . ”a used in this study 5 V C 5 a a o D 0 s . “g; 7‘35 Ira C3 e t 9 undo“ MN V ‘UKEDII ‘ c b w a o ' 3“ u . c 0 - a “In, I mm, C . KL : o a a . o . . I; a a t: g x tn R ° a I , o . u ° ° . 1. Anchorage 18. McGrath air, 84 in! b ‘beau, I ‘(- . u ., ‘98 . 6% i :1} :3 so? 5‘ I H A | BAY '6 ‘ “g a. °ot . . u . 0 , ab”) 38 - . a . ”A: a; 73 % 3‘ “g 2' rv s N ‘1‘ D o 'n‘ a ‘ Wimk “ 1, ' o I 2 Bendeleben 19 Medfra Q0 4,4 lmmto ' } . . . {- II‘ cg Q M :3): \ 5:6 Q N em a I; a . gmdui =9 . u 3 Bethel 20 Mt Hayes + L "o; ,S’ n n ‘ c s .3 R O S o . $01 ake ' -‘ ‘ ' ‘ 4c 0 (z , ~ r. MT \~ I . . . n§ a ’ ”m "’ v °~ou 42., V, . h r 1:. f ‘r " A" °F w s’ ~ “ ~ 7. ° =7 ' <7 % a a 4. Bradfield Canal 21. Mt. McKmley ””1” Tifl RIDG ‘ 0‘" 4k ” Q-' ‘15, 7 TS. nus} b J W) “E '14 0 4 I I; 0" 9 “9” ° E I § 6? a ° “ n N b I "re . .1. 3% MGM. K. . I g 3‘ g "a; 7000 I, | o N S g . é p on P 1 . V 5. Candle 22. a esna . ¢ -x. uK w ‘2. »_ o ' a ‘6 . ‘ M Wampum *1)” K “NE I” (3 g g “‘04v ’4 / ‘ Q’ I‘ a ‘ a n E h (l Q. . 6. Clyde 23. Nome at y u“ =- x E z 4/ u ° 9 ,W 0a,?” LNG M0 8 % 5 w m f a3- 4 S 6 3 ° 619 " . o . °_ I WM I 7. Eagle 24. Norton Bay or , as...” D I 5’5 e m e E o . a - a o " “ g”. 1 “- - 25 R b , ”7a I: I Ngauv :05. "90% § n Sf 0:: A N ‘ m “.2500 I ”a, M e 3 . n a” -, '9 . . 8. Fan-banks , u y 0 e0 v \Q _ 3 , v.05 - 1/ § 4; n w ‘ m, b . . . . r . » . . - x V. 1 ° ”is!” 044] ~52. I‘J'Vulin AOA7Ak ,0”; WWW“ PASS ,3 9: § 7m 0 v w 9900 .39 a , D , \ {QM “ V :0 a . V D a 9, Goodnews 26. RUSSlan MISSlon II n a ° ' 0 ’/ y. :E ChPI‘mI/e l :35“. ‘2” MW “mm ‘ a 0Q“ 9 RR Q \L" '5 “K“ a “ Q ‘1 fl ‘6 I!“ ' D. u ‘ 10’ Hagemerster Bland 27' Solomon 3» - ' u I a u . c,» a v ~ ° 45°00 Q 55"" AN x, M 0 U N T A I N s (a “mo Ar In B ' ' . °§l:. “CNN F““ a“: a v: E: ‘7 - , 156 ‘ ‘ D ° ': 11 Iditarod 28. Talkeema . 0 Autumn 55 v I a ”do n . ' o - . , " 2.3"” )311r‘IIR"\ c 3’ . . v a- a ‘ ,‘ ‘N ‘7 a , l) - 12_ leamna 29. Talkeetna Mounlalns «r \ EAIRDI M “NZ ¢ 5UVEYPA55 .< VngDo m “5"" ° ‘7‘ 9 r1. 0 00 4 “A K II c 0v asao - a “ ‘9 . U o “a . v.9}fi'5/ 13 Juneau 30 Tanacross 0" a a {“300 NS 1; \V\ o\ ‘1 Q L“ “H“ I} 'D o .a an. :0 \‘\ / ,rr . - ..%§> A or , ’1' ,, -o ' o"; 9 «1,0! '1: . Christin“ $5 Baum H .9, ~ , by,’ 5 u a “65“?“ 14. Ketchlkan 31. Tanana ., 5‘62; ' e, 4 I m 1 l 3* 5”” ’9 III/l . K? ' .9, c 0 4/” " , . ‘0‘ " M 15. Lake Clark 32. Teller 0.; . (‘h K 0“ ’“W T) °‘ ° ., ‘ ‘ F 0” mm “L ‘ ° ' 0- 16 Livengood 33 Valdez ”As ‘u w . (In . ° 4‘ > , - ' :2 a V x V: a a o ’,,4 a L v =2 . . . 2 0 /,/’ G efv 34w 5 N K,“ ‘ K ”k \ (06“ 0‘.“ Curr) Venetw 3‘ . o? 4‘ $4» I” ,1 . c . 4)., . A .. 17. McCarthy Vb a v \ V a § 0 - N , ’ , v . , Q v u . ’ ”Wk 6 . . Sin": ' “WWII/or a I" '5: '° 3 (1 ° ,,—"' ~ I r 5 ' a a o e . ”B; is 33/" H. . ' a . o’ mm]: mom“ m a J? . 9 .° F1" ’_,D r a. I’ I \ it § 0 “Q“ 4n,“ . v \ § _ Lair; 3 o a Do .. Ben E l \ I: ., 9 51m Black. 1a b 71‘, ”I” o . o ., M ‘1.“ Q . ¢ _ \§. w 7"“; ..... q. '. -°‘ ' ° OlFlsMug“- Elmo—m"? j (E 5 o o H 1v ~ I: + v 3 9.4% ”v: M ‘9 ° ‘% e , ........ - _«— a m a v a 1 - V --' - ° . “"99 6mm; ° “1““ Sum, m “—x—fl --:"JE‘BEI:et “.9 ,,'I‘ re ______ “1““ a, 11> _ ~, 0“ 3503 w Bw‘ . . ° 2570* wooa N 60 , ““““““““““““ . a ° , ”~1- w F n ' E H I1 R \’\o a ' . “0N ucklamj )1 V0; (mun . a: m “th 5 mo“ Peal R $1 5;: \ If I R \1 a ’- u . 4h . n o o - e9, 3 _ . u 0 R] R . 0,1)! \ 63 Q . a. n ”a ' Huéheg ”‘2‘” D ° ’1‘ ” ° A, . ’p u "I ‘ a; c - \\1 ' . " (I N a . F a , IYCIe 3‘“ I, ’50. Q” \IVO § {3:690'30 a“ .1 . g . R NI \ M A g)” C -(P [3 '3 “,3: ° 'f 0 . 0W e¥j ‘ \ ,9 ‘ \ U x;\§\ ‘., W? ‘5 1’ \ ”*1 \x K I /‘y x I "a ‘5 Q ° ° , ,t Swim" If”, />\’p \\ \ Mi "W6 [If ’) 'VQW l ° «2 °1" ' , I . n ‘545558 \D/ / \x \\ W1 1"“ ‘7, i / a \e“ ° 1 , N no ' , ,c' \ UGI I v bow ‘\ d3 , , . / / / ¢ K H H55 Ix], \ 13°° Q’ \\"’ l0“ a mover r1 / “’4 a "I“ 0 \ W 5 W I‘ ‘ ,- 0’ . 4’ » T" *. \ ‘ . ' o‘s Q4“ mm Rondhuuxh I ‘x w up»; ”sf ““5 ‘f 2, I 3? 00. mm ' yukuk ' I “fir/wan ), 52?.” 758 ET ‘00 I t / -,‘ O"? , / 05' a ‘ A ~ I“ ’ 'I , ’ ‘ $9 . ’MI E {L I, . ““51er 03° :105L \‘6‘ . . ‘0“? a a \ CAMP 5 / to o a \9 / \w‘ a. \NF 62 T a V / k 4A ulm,J /1 /4;8,oo, // Lake 10 .~ 0 M $0“ 0 ‘4 Q coo w a“ . _ ., . , . I o / , / . ' ' . For Y “ L /Q[uk0“Iamp // w j “0‘ -\_‘ /r/ "A A “sf/FEW? Q» Y‘ ‘5in " “k “‘N _ ," I \-/ L ° “5‘ C . o DEW n /// Maw [,mef: c; /§E/LW /’, : Alum“) “‘9“: /‘ 999 a Sululmkul ~'. 62“ L we Lawfiml/ u / :1— //’~ MQL/ l “1 co W o ., 7' w ile we 0er Ljolear o Gov “‘ “ast/—~~ ‘35» / Iv 7/ - "1 ’ I v' u“ ‘ T n I /} 4/ o . r so (1 m 0 “mm \ _ “9,.“ch may" 9 ".a ‘ TN 0 f 0/ I i \ Q ewe“ ‘ 4F , , . / , ;' = .L .l . II . . ‘mm ’1: «M? «W P é. a w ‘ 125°} 4/ .n In 4 l . Q “I u “—1,. g .1. A O“ ‘9 1 9:12“ “r “V 1 6 ,. / is was 0 N h ‘ “p ) I 1"” -% .. Q " 1“ MW“ . L‘ '7 ,; °’ ° ' l . ‘ 9 ‘ “ ann- ‘7‘ MCK‘me’ Pgrk “10° 5‘9“.” 55" A” . $05 £75555” r"‘~,J-'—’ T ‘3 r :> «,l 3‘ 5/ /\ \_ J ‘ l~ & 86°° (_ Flatt R (I? “‘2 7 v‘ / rare ' Creek {a l/ANVIL MOUN7:[N§ 30/ “(g , \\\\ 00:59 Thwlle Crx!551“g ’9 r ‘3 Q9 l9 19‘ a) \ iv t ’14 “y ' GLENLyoN ’ 1’ I V”"30 0K; ., ‘kfi a—\ /,\ A4 a 4),}, , no w n I . I" > v n N we _ ck MIN" Q ‘2'} "I " .‘ ‘ ,/ 1 w \ a a‘ A F U“ SN“ (0 “Q P " If 2- W , a '0 ,ao l3,“ ° \ a . WSO~ R NE: (3).? TRUn‘T 6 I N ‘L()- a A 1 \ I ' . 1 v‘ “ spam R‘azic- 4’ [yr-0" L $33537: M 9500 , ‘ a "5m: ”“6891, {4’ _ l 4‘ m M‘" ”‘2 msmm‘m ‘1 Mo ”‘4 y W” “5.595 L T E E' \ ‘ a I“ ‘ flow; 7'» gg, “ {a ' ,1.) E Q: \ \ ‘ (‘h Nix/1},” 5 a. “mat“ L M o B . gflhfl" -\ 4"“? A, r I . a ‘2)? if” N Bi'al/or R \/~ ‘ 60° ) 5‘ (x B " WWW; ( "meg ‘. “m ° 1 . . e _/ u “I!“ “we“! a a llenn All! n X “mm 3% Law I ”I - ll ‘° moo: ' Ah! \ _ “fiat“ \ ‘lqr R 4,0 ”25 o a ,Ofi ishih‘k 1" ‘ L13“ 65““ C 7‘ wt Lu 0 C ss/ 04v) ’ 3k ‘1‘ C 0“ szul"‘“" _ afizrge e a L 44, ‘ 1,19 “ L oi I ‘70“. F L 1‘“ a“ ‘1 | a Wall ”W 3‘\ I: 1‘ a ‘1“ a ‘_ ”L ‘ “I; fined” it}; {v (1%; 3503" MOW § W“ n D 1 B. an WW 0&3, ‘09 so - / ‘ 4, 8mm." g; . Kluafle a “mo 5 0:, Q I», x , . ‘ ~s ”a, . ~. AL I 4 L] R V‘ \9. I 4 k > , , r . a , 4 - \10‘ ‘9 , 9‘ 0 4' ulil“ 5’0” a: mu ne ,allvon‘ wk : = “. ‘9 , ‘ M“ V; M » ( 440 x I § “ A “\LLS V a a I . 44 *— row“0 a“; 4’ MaoX 109%;qu S” I" 1mm?“ 1 1 Av D 1m m ’ f‘bdd- . .. . , z k its”: 0 3* " a wig/2.x "I a! (13‘1” 1. Ta“ T25 k2 ¢ a“ § 9 "3‘ “Wm I Huh 1’ CW“ “T" “E“ SS N x 9 v “‘a w! ‘7‘ Q ‘51. O (”a hr. Mm Kash In"; "’19 Q. A yunek Tumagag 7 1’ 1," a) ‘ ‘3; a: g 7 , “ ‘ml \5 \ 6 °. 49-. "r $ "Ire $3 I / V I )1 A «In? .,I M “K Vigly$ KN‘WI‘“ 2:: * 0 ¢ 5 whom Hi 3‘. ’ .‘ ‘I :5“ $020502 5 94‘ 7‘ ”‘Wug 5; ”'1‘an # n a J . 1 M . Irma 561 0 P , 9’2”". .. eek engenha’ Kw‘ L 0(me 4 . ‘ 'll (12'3st gm°$fiy 7 , Dawn V e3“ I» w A is R am? U wdww \ Imam.” KID" , . AV ., , 9 ~/ ; h "\ {14% l a GEM“ § 3v . 015:9 t A; III-ch“. \‘ ' Kemp o P0915: u pm fin.- WILIIA W E .13qu J1 A “a 7;“ , - , V "‘7'. Lake. to m'sozmn wilting” - ' tlifl L Tab“ k -.‘ q, t )1; . . ’2 R (IN N; A 1 g V» {"33 ”e am“ ‘, ~ ‘2 er ' Hinchinbmok l “1.26:1 a 4} _ ’7 a 00 I . o . g or‘ {5 Q 2 \‘1 ;/ ‘6 “AD\ a § “no. a: ,3 r . Bay . \c‘l Cm my s) . . “ a ‘er La‘fumr 3" T 0”“ C Sucklmfi \g, ‘11/ ‘27” . A. ,A “j? A ”:00 MW“ ' a Q, ’7’.» e - Kay?!“ WI I a P: 9/“ as: O [,1an R “ “a 1‘ Ever! a6 Yak“ " a k 5 X!“ e” a. Montague! c 3,1Lllus C no ~03 99 N. Anchm I 0w" ' {71) "r. x- ” ’7qu 41¢ / R / hrmrna P. H I C Clear? ‘26?) f “\s Shed“ If laknam'k E , B 0k 0.: 9c) . é a T5 ‘9 kyvok N In: I Kac RY" J Middleton l, 019 PM,“ «L X . {A )9? .353 u o a {)4an ‘ | M I T \A ~ .‘ y . H n :1 Huh , ' . 130116th 501110le 5‘; Pye s e n .weame ,. , / ‘3 (”-3 13"" a nk , E can M / a9? 255‘) p N .1 j, , , ’ Pox-cl ‘5 ,m , ( 2 C ° K . ) J / o ’5 PtGo:e “3 A S K A “09“ kvxfi 4 ./// If} \e A c wk 5. clerks opm- “ 001 H \ m DOUQLA Chunach ls L F O F A L W" M w x / , 3, rt. x- =— o o ' 1v 7 C Umr .5 , a 9 ”If Ir Penn»; 0 8 5 C “51“?! (1 v’ / D .- 4 y n I0 I?" ‘ 25mm “‘90 2° .L k " D 179"“ ‘9 Barron ls G U 0‘1 '5 s g <~~ a $0 5,, a ,, Q . l C Cement." “5%; , °_ 0 0 Tonkl Capo m _ ,3, . I Q 47" W ‘d s ‘“ R; my a: w ‘ 014;, s L “grown = 2;; x) §Marmon + ‘3“ 5"“ «“I ed 0 it 5’13 P C Gler . ” Nut“ ‘ . gee . . \ c on," g 1 4 Km UL,“ ‘7 b '1“ ° 6 .\ MW. ””0“ DIAK 4, ‘ ‘I ‘ 0' . ° ‘ I’yashi/r musk 1‘ Mn air Bay 0 Waggon ,e ‘3” - 5‘3“ 8 L I“ 41km a 1‘ Y C Cw ‘mok O 5" («,ddul‘ -n i' -« I? 1 s T O . . c .. J . '3: Mr .33 C lelIk K- CHIG Shaggy”), pang: ‘93:,“ Q \ 59 .. A04 Aynkuhk - 5..» .h. ‘ Seal I59 .3 I ”l2“ K ' k 9‘ 5m: den ° (:00 SE ‘0 ”EM/n - Sut . N“ \ v w k . , . - u o P‘ ‘ I , 0 CR ' ' Slfkmulr AMI?“ 0 Corofia‘mM “ace “(A V» . *1" “a?“ \ A Chlflnl/r 8 ”gm k l SItkInak I evil“ \ K p ‘ :I - . 3 Av 10"” no!“ a 011M011” 4; 90 " Ik ‘S - a “Z 4 15‘? ”W «035% Cape e TRWTY I harl In " c e" ’ rr . 5 mm 3mm” A “a E . _ P J 1 Sean; "‘- s 6‘ ..e «C W ' om!“ 'iofigw may 9 Anya/“WWW, ”00* P‘ \a 'g 0 m a“ y X ( \‘ u\\\ A 54° .3 <9 \ M "DIEM.“ 1,- — 2-.“ \ ‘3- won $0 F 4’ o g o Ilrovama flv— . ' ,. C \ O UNIM’IK IS t 5’2", ' VLD Vole 5W“. K I a Chinkon \ 4 3181‘ R g Mud” a N 'w o “mum” P? “me O Moi-1“ (I; 3“ “NB“ ' mt. P» V' N 5 $27 Qu’lgald GEQCCQQ" Ofow'n, l CV " P rant‘c’ \ 09;, y 2 S L SM‘" 41:: I. a I" 0v gym 1" 0:09 \l RANCII I” S‘w“i:‘\:"% - a s , e . . Cs . V° ‘ C) “D DUI I w , .Nl \ Dam, f >9 ‘9 0.9., , WI my,“ I 5'8 Konlu” , l \\ DIXOI‘ I ‘ ‘ WWI a BDgowof. A Alran; ”"ha y" 0 . Cos, -‘ - II, I ‘1 \ arm 5 '5‘ IO tum, / ’7! 0 L52. . W , . 4 "9 Kanlu'i : = I. 0‘" rr . rhea \ A N MM‘U‘ Akula” par %{ 010”")? SZFC’I C: an 04 o ' salluk 0"" a f) I I l l I, m: I“ GI “8 \ ) Mn La. ‘ ‘5‘ ll . V «9 / ~ ‘\ 0 , 172- A L E U T C r ”ALAS“ Agog H be it; 9 0.» ‘st :g (was ll 0% B ME x! “9mm” ; 128» or ”,3“ I ‘5 ‘ (a . U I s Krenimn Is I l . ¢ 0 (r, 5 Re My vs NAN I 52",]? 90> “£51K En /// \ 11‘ \s\ ‘0“ ”an” '1 Chc;rl’3/e [EU/m,“ / 3?} a. 9 , ........ T/ \ (£37 ‘ “r Mod 5"“ “Zea/4,. I ”Gib, "was 0 O ‘; 9‘ F omma ‘4 0 y ----- ’ lT/T l fig 36 ’ € 1: > \ n . s z , ., as“- I... .. L A \ 2. ° warm § 4:; n, 5:”08 or W‘ ’ 5k \ 5 \ \ ‘I' w "a; w "k" I F 0 X \ (“(4 LIG‘“ «7 aa‘ h—NH.-._...2“ ..Mr,. "7...”. ,,.m__. L..o———~>~'J“"'"T '7' d \ 4229 3535 Q‘ ”a /;, 188° 164° 180° 156° 152° West of Greenwich 148° 140a 133-: 132.. Data compiled in 1972 (2 us. GOVERNMENT PRINTING OFFICEI I980—-677.I29/62 ‘EoLOGICAL: *«SUCUMEHTS Dammmir NOV 1 6 1979‘ UBRARY WWW 131'" Gil Modeling Highly Transient Flow, Mass, and Heat Transport in the Chattahoochee River near Atlanta, Georgia By HARVEY E. JOBSON and THOMAS N. KEEFER GEOLOGICAL SURVEY PROFESSIONAL PAPER 1136 UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1979 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretmy GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Cataloging in Publication Data Jobson, Harvey E. Modeling highly transient flow, mass, and heat transport in the Chattahoochee River near Atlanta, Georgia. (Geological Survey professional paper ; 1136) Bibliography: p. 41. 1. Stream measurements—Chattahoochee River. 2. Rivers—Georgia— Atlanta region—Temperature. I. Keefer, Thomas N., joint author. II. Title. III. Series: United States. Geological Survey. Professional paper ; 1136. GB1225.G4J6 551.4’83’09758 79-607766 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, DC. 20402 Stock number 024-001-03233-3 CONTENTS Page Page Symbols IV Model calibration and verification—continued Conversion table V Mass transport of a conservative substance .......................... 16 Abstract 1 The temperature model 21 Introduction 1 Discussion of results 26 Model development 2 Flow dynamics 26 Flow model 2 Transport 32 Mass and heat transport models .............................................. 4 Temperature 35 Coupling 10 Summary and conclusions 40 Model calibration and verification ................................................ 11 References 41 Flow model 11 ILLUSTRATIONS Page FIGURE 1. Map of Chattahoochee River showing the data/collection points and tributary measurement sites _________________________________________________ 2 2. Photograph showing view of Chattahoochee River near Settles Bridge (river km 553.0) at low flow .............................................. 2 3. Computation stencil for the linear, implicit finite-difference solution of the flow equation 3 4. Photograph showing Suwanee Creek during a high stage in the Chattahoochee River 4 5. Photograph showing aerial view of Chattahoochee River showing shading conditions 5 6. Schematic of river cross section used to determine the part of the water surface to be shaded by bank vegetation .......................... 6 7. Computation stencil for the finite-difference solution of the transport equation 9 8. Graph showing steady flow depth profile for the Chattahoochee River between Buford Dam and Norcross .................................... 14 9. Graph showing calibration of the Chattahoochee River flow model with the March 1976 stage data ............................................... 15 10. Graph showing comparison of modeled and observed discharge in the Chattahoochee River at the Highway 20 Bridge during the March 1976 calibration period 16 11. Graph showing comparison of modeled and observed discharge in the Chattahoochee River at the Littles Ferry Bridge during the March 1976 calibration period 16 12. Graph showing comparison of modeled and observed discharge in the Chattahoochee River at the Highway 120 Bridge during the March 197 6 calibration period 17 13. Graph showing comparison of modeled and observed discharge in the Chattahoochee River at the Highway 141 Bridge during the March 1976 calibration period 17 14. Graph showing verification of the Chattahoochee River flow model using the October 1975 stage data ......................................... 18 15. Graph showing calibration of the Chattahoochee River flow model using the October 197 5 stage data ........................................... 19 16. Graphs showing comparison of the modeled and observed discharges in the Chattahoochee River during the October calibration run 20 17. Photograph of dye injection site for the March 197 6 dye study, 200 m downstream of Buford Dam 20 18. Graph showing modeled and observed dye concentrations at Littles Ferry Bridge as computed using the conservative transport model with unequal grid spacing 20 19. Graph showing modeled and observed dye concentrations at Highway 141 as computed using the conservative transport model with unequal grid spacing 20 20. Graph showing modeled and observed dye concentrations at Littles Ferry Bridge as computed using the conservative transport model with equal grid spacing 21 21. Graph showing modeled and observed dye concentrations at Highway 141 as computed using the nonconservative transport model with equal grid spacing 21 22. Photograph of meteorologic instrumentation tower of the R. M. Clayton Sewage Treatment Plant showing the anemometer and two psychrometers 23 23. Photograph closeup of the anemometer showing the Chattahoochee River in the background 23 24. Photograph closeup of the pyranometer and pyrgeometers at the R. M. Clayton Sewage Treatment Plant in Atlanta ____________________ 23 25. Photograph closeup of ventilated psychrometer at‘the R. M. Clayton Sewage Treatment Plant in Atlanta 23 26. Photograph closeup of Top Hat psychrometer at the R. M. Clayton Sewage Treatment Plant in Atlanta ........................................ 24 27. Graph showing observed temperature in the Chattahoochee River at the Highway 20 Bridge during the October 1975 Verification period 25 28. Graph showing comparison of the observed and modeled temperature at the Littles Ferry Bridge on the Chattahoochee River during the October 197 5 verification period 25 III IV FIGURE 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. TABLE 1. 2. 3. 4. A ALON = AZS = BW = C = CN = C,0 = Cu = DI = E = e,l = EBH = ELEV = e0 = EW; = g = HA = H;- = HR = Ha) = 1 = K = L = Q = CONTENTS Page Graph showing comparison of the observed and modeled temperatures at the Highway 120 Bridge on the Chattahoochee River during the October 1975 verification period 25 Graph showing comparison of the observed and modeled temperatures at the Highway 141 Bridge on the Chattahoochee River during the October 1975 verification period 26 Graph showing observed temperature in the Chattahoochee River below Buford Dam during the March 1976 verification period __ 27 Graph showing comparison of the observed and modeled temperatures at the Highway 20 Bridge on the Chattahoochee River during the March 1976 verification period 27 Graph showing comparison of the observed and modeled temperatures at the Littles Ferry Bridge on the Chattahoochee River during the March 1976 verification period 28 Graph showing comparison of the observed and modeled temperatures at the Highway 120 Bridge on the Chattahoochee River during the March 1976 verification period 29 Graph showing comparison of the observed and modeled temperatures at the Highway 141 Bridge on the Chattahoochee River during the March 1976 verification period 30 Graph showing loop rating curves for the Chattahoochee River at the Highway 20 Bridge during the March 1976 calibration ________ 31 Graph showing loop rating curves for the Chattahoochee River at the Littles Ferry Bridge during the March 1976 calibration ______ 32 Graph showing loop rating curves for the Chattahoochee River at the Highway 120 Bridge during the March 1976 calibration _____ 33 Graph showing loop rating curves for the Chattahoochee River at the Highway 141 Bridge during the March 1976 calibration _____ 34 Graph showing excess temperature at the Highway 141 Bridge during the March run due to an excess temperature of 1° at Buford 39 Graph showing excess temperature at the Highway 141 Bridge during the October run due to an excess temperature of 1° at Buford 39 TABLES Page Internal reach data for the Chattahoochee River between Buford and Norcross 12 Discharge values for the tributaries and withdrawal points on the Chattahoochee River during the October 20~25, 1975, and the March 21—24, 1976, modeling periods 13 Physical data relevant to the Chattahoochee River between Buford and Norcross 22 Daily average values of meteorologic variables at the R. M. Clayton Sewage Treatment Plant in Atlanta ________________________________ 24 SYMBOLS = cross-sectional area of the channel; longitude of the river (84.2° W); azimuth of the sun; bank width; concentration of dye; Courant number; specific heat of water at constant pressure; heat storage capacity of the bed; longitudinal dispersion coefficient; rate of evaporation; vapor pressure of air; effective barrier height; elevation of the sun in degrees; saturation vapor pressure of air evaluated at a temperature equal to that of the water surface; water-surface elevation at grid point i; acceleration of gravity; hour angle of the sun; sum of last two terms in equation 10 evaluated at grid point i and at time jAt; time of day, in hours; increase in heat content of the slab between time 0 and t; rainfall rate; » kinematic surface exchange coefficient; latent heat of vaporization; latitude of the river (34.0° N); Manning’s roughness coefficient; dimensionless surface exchange number; wetted perimeter of the channel; discharge; lateral inflow per unit length; tributary flow rate; hydraulic radius; part of incoming solar radiation absorbed by the water; = part of the incoming solar radiation which would be absorbed by the water under shade-free conditons; friction slope; cross-sectional average water temperature; time; air temperature; excess water temperature above ambient; final temperature of the water as it leaves the system; initial temperature of the water as it enters the system; temperature at grid point i and at time j At; temperature of tributary inflow; wet-bulb air temperature; meridian 0f the time zone; cross—sectional average velocity; velocity at grid point i and at time jA t; shear velocity; windspeed; = top width of the channel; 8 | CONTENTS — longitudinal distance along the channel; - normal distance from the tops of the trees to the shade point; depth of flow; distance above the insulated bottom of the slab; measured depth at steady low flow; thickness of bottom slab; elevation of the bed above some datum; = acute angle between the azimuth of the sun and the azimuth of the river subreach; — psychrometric constant; declination of the sun; change in tributary storage to occur during a time step; time step in finite-difference solution; temperature rise within the slab; distance step in finite-difference solution; emissivity of water; _ Manning’s roughness at steady low flow; rate of change of Manning roughness with stage; space derivative weighting factor; thermal diffusivity; density of water; = Stefan Boltzman constant for blackbody radiation; = travel time of a water particle through the system; = flux of thermal energy from the bed to the water; = heat flux caused by longwave radiation emitted by the water; = heat utilized by evaporation; heat conducted from the water as sensible heat; II net heat flux caused by incoming radiation from the sun and the sky; heat flux added to the river by tributary inflow; = heat added to the water by rain fall directly on the surface; flux of thermal energy from the air to the water; and empirical wind function. CONVERSION TABLE Multiply metric unit meter (m) kilometer (km) millimeters (mm) meter per second (m/s) cubic meter per second (m3/s) pascal (Pa) watt per square meter By 3.281 0.6214 0.03937 35.31 0.02832 10.00 0.3172 To obtain inch-pound unit foot miles inch foot per second cubic foot per second millibars British thermal units per square foot per hour , w; MODELING HIGHLY TRANSIENT FLOW, MASS, AND HEAT TRANSPORT IN THE CHATTAHOOCHEE RIVER NEAR ATLANTA, GEORGIA By HARVEY E. JOBSON and THOMAS N. KEEFER ABSTRACT A coupled flow-temperature model has been developed and verified for a 27.9-km reach of the Chattahoochee River between Buford Dam and Norcross, Ga. Flow in this reach of the Chattahoochee is con- tinuous but highly regulated by Buford Dam, a flood-control and hydroelectric facility located near Buford, Ga. Calibration and verification utilized two sets of data collected under highly unsteady discharge conditions. Existing solution techniques, with certain minor improvements, were applied to verify the existing technology of flow and transport modeling. The linear, implicit finite-difference flow model was calibrated by use of a depth profile obtained at steady low flow and unsteady flow data obtained in March 1976. During the calibration period, the model was generally able to reproduce observed stages to within 0.15 m and discharges at less than 100 m3/s, to within 5 percent. Peak discharges of about 200 mS/s were under-estimated by about 20 percent. During the verification period, October 1975, the flow model reproduced observed stage changes to within about 0.15 m, and its timing and over-all performance was considered to be very good. Dye was added to the upstream end of the river reach at a constant rate while the river flow was highly unsteady. The numerical solution of either the conservative or nonconservative form of the mass- transport equation did an excellent job of simulating the observed concentrations of dye in the river. The temperature model was capable of predicting temperature changes through this reach of as large as 5.8°C with a RMS (root- mean~square) error of 032°C in October 1975 and 020°C in March 1976. Hydropulsation has a significant effect on the water temperature below Buford Dam. These effects are very complicated because they are quite dependent on the timing of the release with respect to both the time of day and past releases. INTRODUCTION In 1964 the Department of the Interior was desig- nated as the lead agency for coordinating federal activi- ties in water-data acquisition. Responsibility for these activities was assigned to the US Geological Survey. Federal and nonfederal committees were formed to advise the Survey about water-data needs. In the early 1970’s the nonfederal committee recommended that a series of interdisciplinary river-quality assessment studies be performed to (1) define the kinds and amounts of data required to adequately assess various types of river-quality problems, and (2) to develop and document methods for assessing planning alternatives in terms of potential impacts on river quality. Partly because it is in a developed basin with problems at the present time, the Chattahoochee River basin was selected as one site for a river-quality assessment study. The Chattahoochee is one of the largest rivers in the southeastern United States. Its headwaters are in the Blue Ridge province of north Georgia. It flows across the Piedmont and onto the Coastal Plain near Colum- bus, Ga. The Chattahoochee River joins the Flint River at the Georgia-Florida State line to form the Apalachi- cola River which drains into the Gulf of Mexico. The Chattahoochee’s length from headwaters to the conflu- ence with the Flint River is about 710 km, and its drainage area is about 23,000 kmz. On April 1, 1975, a river quality assessment project began addressing problems related to (1) thermal loading and heat dissipation, (2) wastes from concentrated urban- industrial areas, (3) effects on river quality of hydropower pulsation, and (4) sediment sources, trans- port characteristics, and deposition for the Chattahoo- chee River basin below Buford Dam and above West Point Dam. One element of the Chattahoochee River assessment study was to develop and verify coupled flow—tempera- ture models of a 27.9-km reach between Buford and Norcross, Ga., with special emphasis on evaluating the effects of hydropulsation on the flow and temperature regimes. The purpose of this report is to present these models with verification and to identify and specify the procedures necessary for their successful application. Buford Dam, located approximately 65 km northeast of Atlanta, Ga., creates Lake Sidney Lanier. Its hydroelectric units are used for peaking purposes, so the flow in the Chattahoochee River below the dam is highly unsteady. For example, the US Army Corps of Engi- neers maintains a minimum flow of 15.4 m3/s through the dam but typically releases two pulses of water with peak discharges of about 215 m3/s during week days. Just downstream of the dam, the total stage rise associated with the pulses usually occurs in a timespan of 10 to 20 minutes. A linear, implicit finite-difference flow model is coupled with two implicit finite-difference transport models to describe the flow as well as the transport of o 34 2 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA heat and dye in the reach of the Chattahoochee River below Burfod Dam. The study reach is illustrated in figure 1. It extends from the powerhouse at Buford Dam to the Highway 141 Bridge near Norcross. The river flows through rolling hills for most of this distance. The overall slope is 0.00036. The first 3.2 km are steep and rocky with a slope of 0.0011, and supercritical flow is present at several places during low flow. A bedrock control, below Highway 20, ponds the water in about a 6.4-km reach at low flow. The channel ranges from 45 to 65 m wide with a roughly rectangular cross section (see fig. 2). A characteristic cross section has large fallen Lake Sidney Lanier Buford Dam\ \Highway 20 Bridge Littles Ferry Bridge\ Highway 120 Bridge \Highway 141 Bridge, Norcross Chattahoochee River Atlanta metropolitan area 0 25 50 KILOMETERS 0 10 20 30 MILES FIGURE 1.—Chattahoochee River showing the data-collection points and tributary measurement sites (open triangles). FIGURE 2.—View of Chattahoochee River near Settles Bridge (river km 553.0) at low flow. trees protruding 6 to 15 m into the stream at the banks. The trees have fallen in because of bank sloughing which results from the large and rapid stage variations. Along the tips of the fallen trees, the water is somewhat deeper than in the center of the channel. The fallen trees ap- parently act in muct the same way as jetties or groins. The channel bed is primarily coarse sand (1 mm) on bed- rock. At low flow, sand dunes, 1.8 to 3 m in length and about 0.3 m high, cover the bed. A typical Manning roughness coefficient is 0.042. In the following sections, the flow, mass, and tem- perature models are described, and the available data, as well as the calibration and verification of each model, are presented. Two data sets, each containing con- tinuous flow, stream temperature, and meteorologic data, were obtained. One set was for the period of Oc- tober 20 through October 26, 1975, and the other for March 21 through March 24, 1976. In addition, a conser- vative tracer (rhodamine—WT dye) was injected contin- uously at a constant rate during the March run, and samples were collected at 10-minute intervals near the center and at the end of the reach. These data are used to verify the transport model, and in addition, they served as an excellent check for the flow model. After the calibration and verification results have been presented, each model is discussed with special empha- sis being placed on identifying the problems associated with analyzing transport in highly unsteady flows and the potential accuracy of such an analysis. Finally, the effect of the hydropulsation on the flow, transport, and temperature regimes is dealt with in some detail. MODEL DEVELOPMENT FLOW MODEL Techniques available for modeling unsteady open- channel flow have advanced rapidly in the past 10 to 15 years, but almost all models are based on the same basic equations. These are continuity equations describing the conservation of mass 6A 6U Ua—x + AW +a—A-—q=0 6t (1) and the conservation of momentum (9 +gT= flux of thermal energy from the air to the water, W= top width of the channel, Cp=specific heat of water at constant pressure, p = density of water, B = flux of thermal energy from the bed to the water, P=wetted perimeter of the channel, and the other symbols are as previously defined. Assuming water to be incompressible and the product DxA to be independent of x, equation 9 can be simplified to 6T 6T 5 U“ 890 02T ”new CppA C P pA (10) which is called the nonconservative form of the transport equation. If the velocity field satisfies the con— tinuity of flow equation, the exact solutions of equations 9 and 10 are identical. When numerical tehcniques are used, however, equation 9 will provide a more conser— vative solution (Roache, 1972). When the transported substance is dye rather than thermal energy,the values of <1) T and <1) 3 are zero, and T can be replaced by C which represents the concentration of dye. Both equations 9 and 10 were solved when modeling the movement of dye, but only equation 10 was used in the temperature model. The next to the last term in equations 9 and 10 represents the rate of change of water temperature due to exchange of energy between the atmosphere and water. The ratio A/ W is the effective water depth, sometimes called the hydraulic depth. The rate of ex- change of energy between the atmosphere and water has been discussed many times, but one of the first and most complete analyses of the processes involved has been given by Anderson (1954). For the purpose of this study, the net exchange was expressed as the sum cpN—ob—oe—cph+q>R+q>q (11) MODEL DEVELOPMENT 5 in which N = net heat flux caused by incoming radiation from the sun and the sky; <1>b=heat flux caused by longwave radiation emitted by the water; <19, =heat utilized by evaporation; R are positive and the values of (In, che-and (In, are negative if the water is gaining thermal energy as a result of the respective processes. The net flux caused by incoming radiation from the sun and the sky, ‘DNis composed of four components, the incoming atmospheric and solar radiation and the reflected components of each. The incoming com- ponents were measured directly, but the reflected parts had to be estimated. The traditional assumption that 3 percent of the atmospheric radiation is reflected (Ander- son, 1954) was made. The source of all solar radiation is the sun, and if some object is between the water surface and the sun, much of this radiation will be intercepted. The banks of the Chattahoochee River between Buford and Norcross are almost completely tree lined (figs. 2 and 5). The incom- ing solar radiation, however, was measured by use of a pyranometer placed in an unobstructed area. The part of the measured solar radiation actually absorbed by the water is a complex function of the elevation and azimuth of the sun as well as the azimuth of the river reach, the height and density of the trees, and the width of the river. A two-step procedure was used to estimate the part of the measured solar radiation, RS, actually absorbed by the water. The first step was to determine the part of the available solar radiation which would have been ab- sorbed providing n0 shading had occurred. Anderson (1954) has presented a formula for the computation of this shade—free absorption RSM = 1.0 — 1.18 ELEV—O-77 (12) FIGURE 5.—Aerial view of Chattahoochee River showing shading conditions. in which RSM =part of the incoming solar radiation which would be absorbed under shade-free conditions, and ELEV=elevation of the sun in degrees. The sun’s elevation was computed from ELEV = sin 8 sin t + cos 8 cos t cos HA (13) in which 8 =declinati0n 0f the sun, 2 =latitude of the river (N34.0°), and HA =hour angle of the sun which was computed from HA = (180 + ALON - TZM) - 15 HR (14) in which ALON =longitude of the river (84.2°W), TZM =meridian of the time zone, and HR =time of day in hours. The equation of time was ignored and the declination of the sun, obtained from a solar ephemeris, was assumed constant during each run. The second step was to reduce the value of RSM ap- propriately to account for the shading of the water due to trees and other obstructions on the banks. The shaded part of the water surface was assumed to absorb solar radiation at 20 percent of the measured rate, and the clear part of the water surface was assumed to absorb at RSM times the measured rate. The value of RS therefore, was determined as 0.2 times the part of the water surface shaded plus RSM times the part of the water surface exposed to the sun. The part of the water surface in any subreach to be shaded at any time of day was determined from the geometric relation between the elevation and azimuth of the sun, the azimuth of the river subreach, the effective barrier height, EBH, the water-surface width, W, and the bank width, BW. The physical relationship between these terms is illustrated in figure 6. The river cross section was assumed to be symmetric about the centerline, which is a reasonable assumption for the study reach. The normal distance from the tree tops to the shade point, XN, was deter- mined from the expression _ EBH XN — m (15) in which 6=the acute angle between the azimuth of the sun and the azimuth of the river subreach. The azimuth of the sun, AZS, was computed from AZS = arc cos (sin 6 — sin ELEV sin t )+ (cos ELEV cos E) (16) Using the above expressions, the part of the measured solar radiation to be absorbed by the water, RS, was 6 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA Effective barrier heigth Sunlit / Shade point Bank Width width FIGURE 6.—Schematic of river cross section used to determine the part of the water surface to be shaded by bank vegetation. computed once for each subreach and each time step during the day. Analytic expressions must be used to relate all other terms in equation 11 to the water temperature and the meteorologic variables. The expression used to deter- mine each term will be summarized. Longwave radiation emitted by the water surface was computed using the Stefan-Boltzman law for blackbody radiation (1),, = 60 (T + 273.16)4 (17) in which 6 =emissivity of water (0.97), 0: Stefan- Boltzman constant for blackbody radiation (5.67E—8 W/[mZC4]), and 273.16 converts to the Kelvin scale when T is given in degrees Celsius. The energy added by rainfall was determined by (DR = CPpI (Tw — T) (18) in which I =rainfall rate, and TW=wet-bulb air temperature. Equation 18 does not account for the water flowing into the river as sheet flow or from small ditches during or after a rain. In some cases this non- tributary inflow is believed to be a significant, but unac- counted for, item in the energy budget. It is assumed that the rate of evaporation can be estimated by a formula of the Dalton type E = we, — ea) (19) in which E =rate of evaporation in units of length per time, #1 =an empirical coefficient or wind function, eo=saturation vapor pressure of air evaluated at a temperature equal to that of the water surface, and ea=vapor pressure of the air above the water, which is commonly measured at a height equal to the height of the measured wind velocity. The wind function was estimated from ‘1’ = 3.01 + 1.13V (20) in which ll! =wind function that gives the evaporation rate in millimeters per day when the vapor pressure deficit is expressed in kilopascals, and the windspeed, V, is expressed in meters per second. Equation 20 was derived from thermal data collected on the San Diego Aqueduct in southern California (Jobson, 1977), and, to the authors’ knowledge, it is the only wind function ever derived from an energy balance of an open channel. Em- pirical wind functions derived from lake or pan data are numerous (Ryan and Stolzenbach, 1972; Tennessee Valley Authority, 1972; Brutsaert and Yeh, 1970). The thermal energy utilized by evaporation was expressed as (be = pL‘I’ (ea — ea) (21) in which L = latent heat of vaporization. Heat exchange by conduction has received relatively little attention because its magnitude is usually small in comparison to the evaporative heat exchange. Assum- ing that the eddy diffusivities of heat and mass are iden- tical, which leads directly to the Bowen ratio concept, the conduction term can be expressed as (13h = ‘y pL‘I’ (T — Ta) (22) in which 7 =psychr0metric constant (0.598 based on an assumed atmospheric pressure of 98.0 kPa); and Ta = air temperature which should be measured at the same elevation as the vapor pressure. The last term in equations 9 and 10 represents the thermal flux at the bed of the river. Past attempts to model the bed conduction term have estimated the heat flux as the product of the thermal conductivity and the temperature gradient within the bed. This transient temperature gradient was either estimated (Messenger, 1963) or determined from a few measurements within the bed (Brown, 1969; Pluhowski, 1970). Measurements of bed temperatures are difficult and seldom available. In addition, bed conditions are seldom uniform. Even though the bed conduction term has been shown to be significant (Brown, 1969, Pluhowski, 1970), at least for shallow depths, it is usually ignored because of the above difficulties. MODEL DEVELOPMENT The earth under a river can be approximated as an in- finitely thick conducting medium, the thermal proper- ties of which can be estimated, at least approximately. The thermal conductivity of flowing water is much greater, because of turbulence, than that of the soil, so the surface temperature of the bed can be assumed to follow the water temperature very closely. Mathematical expressions for the temperature distribution and heat fluxes within a semi-infinite medium which result from an arbitrary temporal varia- tion in surface temperature are relatively simple (Carslaw and Jaeger, 1959). Unfortunately, these ex- pressions converge slowly, and their use in a thermal model would be expensive. On the other hand, if the earth below the river were considered to be a slab, in— sulated on the bottom and of an arbitrary thickness, Z, the equations are still fairly simple but converge much faster. If the temporal variations in surface temperature are cyclical, the heat fluxes determined by the semi- infinite and finite thickness slab equations become in- distinguishable as the slab thickness increases. In fact,assuming a diurnal water temperature swing of 10°C and thermal properties for saturated sand, the surface heat fluxes for a slab only 25 cm thick are within 6 percent of the values for a semi-infinite medium. The heat exchange between the water and the bed was, therefore, estimated by considering the bed to be a homogeneous slab, insulated on the lower face and with a surface temperature on top equal to that of the overly- ing water. The heat flux into or out of the bed was then determined as a function of the past history of the water temperature. Only the thermal diffusivity and heat- storage capacity of the soil needed to be known. A slab thickness of 100 cm was assumed to be sufficiently thick to give the desired accuracy. The temperature distribution within a slab, initially at constant temperature, for which the surface is subjected to a unit increase in temperature at time zero is given by (Carslaw and Jaeger, 1959) :1. 2 “1)” exp [—K(2n + 1)2772t/4ZZ] ”:0 + 1 cos [2% + 1) Tryb/ZZ] (23) in which ATB = the temperature rise within the slab, K = the thermal diffusivity, Z = thickness of the slab, and yb =the distance above the insulated bottom of the slab. The increase in the heat content of the slab can be evaluated at any time by multiplying equation 23 by the heat-storage capacity, then integrating over the total thickness H(t) = 0,2 1 — exp [—K(2n + 1)27r2t/4Z2] sin [(2% + 1)1'r/2] } (24) in which H(t) =the increase in heat content of the slab between time 0 and t resulting from the unit increase in surface temperature at time zero; and Cu = the heat- storage capacity of the slab which is the product of the density and specific heat. Of course, this heat must have been provided from the overlying water. Equations 9 and 10 are solved by use of a finite- difference approximation that advances in time by discrete steps of duration At. The heat flux to the water AHO) during any time step j At to (j + 1)At which results from a unit increase in temperature at time zero, can be computed as AHQ‘) = HUAt) — H[(j + 1)At] (25) The AHOYS describe the time variation of the response of the system to a unit change in water temperature. Equation 24 is linear with respect to temperature, and since water temperature fluctuations can be represent— ed by a series of step changes, the superposition princi- ple is used to determine the heat flux from the bed to the water for any temperature history by use of the equation (DBQAt) = :AT(kAt)AH(j — k) (26) k=—s in WhiCh (PRU At) is the heat flux to the water from the bed during the time jAt to (7' + 1)At; AT(kAt) is the change in water temperature which occurred at kAt (k = (T, — T)c,pq ‘7 WM (27) in which rl>q=heat flux added to the river by the tributary inflow; Tq=temperature of the tributary in- 8 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA flow; and q =tributary flow rate at the river. The value of (bq is added to the value of d>T for use in the model. The value of q represents the actual interchange of water between the tributary and the river as computed from equation 9. If the flow was from the river to the tributary, the value of Tq was set equal to the river temperature, otherwise it was set equal to the temperature of the water in the tributary storage. The temperature of the water in tributary storage was up- dated each time step by considering the steady tributary flow, QT, and storage volume. No surface exchange was allowed for the water held in tributary storage. The dispersion coefficient used when solving equation 10 was determined from D x = 250 RHU. (28) in which U‘ = shear velocity, and 250 is the approximate average of the indicated ratio for the data summarized by Fischer (1973). Roache (1972) gives a method of de- termining the effective numerical dispersion coefficient for a differencing scheme such as that of Stone and Brian (1963). Under the combination of grid spacing and time steps used for the variable grid model of the Chat- tahoochee (eq. 9), the effective numerical dispersion coefficient often exceeded the value given by equation 28. Although the model coding allowed for its inclusion, the value of D, was assumed equal to zero when solving equation 9. Equations 9 and 10 are Eulerian equations, meaning that they represent a description of the variation of temperature with respect to a fixed coordinate system. Another description, the Lagrangian, considers the variation of the temperature of a given fluid particle or fluid lump as the particle moves through the system. In the Lagrangian framework, one conceptually follows an individual fluid particle while keeping track of the fac- tors which tend to change its temperature. Applying the thermal continuity equation to a unit mass of fluid, one obtains dT (VT (1) W 4) P _ = D __ __T_ 3 dt 9‘ < 0902 > ACpp ACpp (29) Integrating equation 29 during the traveltime, 7 (FT (D W (I) P To T1 f0 < @6902 ACpp ACpp > dt (30) in which T,- = initial temperature of the water particle as it enters; To 2 final temperature of the water particle as it leaves; and T =traveltime of the particle through the system. Expanding the right-hand side of equation 30 yields 1' a2 7' _ W + yL(T Tan—AC? dt (31) Equation 31 cannot be solved before equation 10 because the value of the temperature as a function of distance and time must be known to perform the in- tegration. Equation 31 was solved for each time step in the model, however, because it is of great value in the process of analyzing the results of the model. Its main value lies in evaluating the contribution of each physical process to the total temperature change of a particle of water passing through the system. Solution of the one-dimensional transport equation is a much less formidable task than solving the flow equa- tions. Most numerical efforts have dealt with ways to minimize numerical dispersion of steep concentration fronts. To the writers’ knowledge, numerical dispersion caused by sudden flow changes, such as those that occur on the Chattahoochee, have not previously been analyzed, however. In order to find a stable, accurate solution technique, which would minimize numerical dispersion with large distance steps and widely varying velocities, preliminary experiments were conducted with three types of solu- tions. An explicit scheme was found to be highly disper- sive and its stability is strictly limited to values of At and A90 which satisfy U At .m— 1 (32) The quantity on the left is called the transport Courant number. A finite-element technique with linear basis functions was derived which is more stable and contains less numerical dispersion than the explicit technique, but it still requires the transport Courant number to be less than two. Later research revealed that this centered, implicit 6-point scheme is very similar to the method presented by Price, Cavendish, and Varga (1968). The technique finally selected for use in this study was a slight variation of the implicit scheme of Stone and Brian (1963). This centered 6-point scheme considers all MODEL DEVELOPMENT 9 points in the time derivative as shown in figure 7. The weighting coefficients on the points used to estimate the time derivative, 0, are cyclic functions of time, which ac— cording to Stone and Brian reduce the numerical disper- sion of propagating high frequency harmonics more ef- fectively. With the time derivative weighting factor, 0, greater than 0.5, the method is unconditionally stable; however, it is still desirable to keep the transport Courant number less than 2 when considering high fre— quency transients. The weighting coefficients on the known and unknown concentration values for the space derivatives derived by Stone and Brian (1963) are identical to those for a linear basis function finite-element technique. The authors modified the values of the centering coeffi- cients, 0 and K, from 0.5 used by Stone and Brian, to 0.6. This modification damps the ringing and overshoot of the Stone and Brian scheme while maintaining the numerical dispersion advantage of the finite-element scheme. Application of the finite-difference forms of equations 9 or 10 to the stencil shown in figure 7 results in two less equations than there are unknowns at the new time step. The two additional equations are provided by the upstream and downstream boundary conditions. The upstream condition was simply a known concentration. A zero gradient downstream boundary condition was assumed by computing the new concentration at the downstream boundary from an explicit upwind differen- cing scheme without diffusion. The solution for the re— maining concentrations involved the inversion of a tridiagonal matrix for each time step. At the beginning of the project, a solution code for equations 10 and 31, which had been developed for use on the San Diego Aqueduct (Jobson, 1976), was avail- c(AXi+AXi-1) 0 6f 32f —A ND 8X 3X2 0 i-I i+1 FIGURE 7,—Computation stencil for the finite-difference solution of the transport equation. able. This model contained three simplifications from the modified Stone and Brian technique however. The coefficient, 0, was not time variable, the value of § was 0.5, and only equally spaced grid points were allowed. because of the highly unsteady flow in the Chattahoo- chee, there was some doubt regarding the accuracy of this simplified solution technique to the nonconservative form of the transport equation. To address this concern the transport of dye was modeled by equation 9, which was solved by the modified Stone and Brian technique, and equations 10 and 31, which were solved by the simplified technique. It was found that the dye concen— trations predicted by the two models were essentially equal, so the temperature model was not recoded. The solution to equation 9 will be referred to as the conser- vative model since it solves the conservative form of the transport equation and the simplified solution to equa- tion 10 will be referred to as the nonconservative model. The finite-difference formulation for the nonconser- vative model (eq 10) was [6—1,+%—2%2] Tiil+ [53+2%x _ (1—0)(;fx9;1—U,)+%2] T? [6—it +(—1‘262;]i-1 +§?_A:C2]T{—1+H,.. (33) in which 6 =space derivative weighting factor, (0.60), T],- : temperature at grid point i and time, jAt, U% = velo- city at grid point i and time jAt, and Hi =the sum of the last two terms in equation 10 evaluated at grid point i and time jA t. Little has been written about the numerical simulation of the source-sink terms, H,,in equations 9 or 10 but cer- tain precautions are necessary. In equation 33 the sur- face exchange term is evaluated at the old time step. One consideration is the maximum size of the time step which can be used with this procedure without seriously compromising accuracy. To illustrate the requirements consider a Lagrangian excess temperature model with no bed conduction or dispersion. The governing differ- ential equation simplifies to 10 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA flf:—K_WTe dt A (34) in which Te =the excess water temperature, above am- bient, and K :the kinematic surface exchange coeffi- cient. For steady uniform flow with constant meteoro- logical conditions, the coefficient on the right of equa- tion 34 is constant and an exact solution is easily ob- tained. Evaluating the surface exchange at the old time step, the finite-difference approximation becomes Tej+1 = KWAt > (35) j —_ Te<1 A The dimensionless surface exchange number, NH, as given by KWAt A (36) N H = governs the accuracy of the numerical solution. A few simple calculations will demonstrate that the numerical solution is very accurate for values of N H<0.2 while the numerical solution overshoots and becomes oscillatory for values of N H>1. The surface exchange term, Hj, in equation 10 is nearly a linear function of temperature for temperature differences encountered during any one time step in the model. The surface exchange number, is, therefore, a meaningful criteria for limiting time step size. As shown by Jobson (1973) the value of the kinematic surface exchange coefficient will almost never exceed 8 m/d while the minimum hydraulic depth in the Chattahoochee is always greater than 0.3 m. By use of equation 36, it is easily seen that the numerical scheme should accurately model the surface exchange for values of At less than 10 minutes. A 5-minute time step was used throughout for both the flow and the transport models. Consideration should also be given to the distribution of the surface exchange term between grid points. With a steady uniform condition and no surface exchange ex- cept at grid point k, the water temperature must remain constant both upstream and downstream of grid k and increase by the amount HkAx/Uk between grid point k— 1 and k. Simplifying equation 33 for steady condi- tions by observing that the old and new temperatures must be the same, one obtains Ui—1_Dx (Uta—Ur) was [—2— rxlTi+1+[T+E Ti —[U—i—1+&]Tl—l : 2 M AHiAx (37) in which AH,- represents the distributed part of the sur- face exchange to be applied at grid point 1'. Under the assumed conditions, it is easily seen that AIL-=0 for k— 2 2i2k+ 1. that is, in order to get realistic results from equation 33, a point source of heat must be distributed between two grid points. Applying equation 37 between grid points k+2 and k and simplifying, the amount to be applied at grid k + 1 is obtained Uk—l _ 5 2 Am applying the same equation between grid points k+1 and k + 1, the amount to be applied at grid point It is obtained U k — 1 2 Summing equations 38 and 39, it is seen that the sum of the distributed surface exchanges is equal to the total point source. If a point source is not distributed as in- dicated in equations 38 and 39, the numerical solution will contain errors upstream of the source similar to What is sometimes referred to as ringing. In order to preserve thermal continuity under unsteady, nonuniform conditions, the physical surface exchange terms were distributed between two grid points using Hk Uk = AHk~1 (38) AHk a) fl = (39, AxUk AH, —H,.[0.5 AxU,,1]+H“1[O'5+AxU,~] (40) This procedure was found to work well in both the con- servative and nonconservative models. COUPLING The flow and transport models were run independent- ly. At each time step, five items of information for each of the 48 grid points in the flow model were stored on magnetic disk for use by the transport model. These items include the top width, velocity, cross-sectional area, tributary inflow (at the river), and the tributary flow into the storage volume (beside the river). This ar- rangement saved a significant amount of computer cost because the transport models were not run until the flow model was calibrated. After final calibration, the flow model was not rerun. Likewise, the transport model could be run as many times as necessary without rerun- ning the flow model. MODEL CALIBRATION AND VERIFICATION 11 In using the conservative transport model (eq 9), the coupling was direct, since this solution allowed for une- qual distance steps, and the grid spacings in the flow and the transport models were identical. The nonconservative transport model (eq 10), which also contained the solution code for equation 31, re- quired equal grid spacing. In order to make the output of the flow model compatible with the solution to equation 10, the flow data were interpolated to an equal grid spacing by use of a processor program. The logic of the processor program assumed that the velocity and cross- sectional area of the “true” river varied linearly with distance between the flow model grid points. The cross— sectional area at grid 1' in the equally spaced model was determined by integrating the “true” area from the point x,- — Ago/2 to 001+ Ago/2 and dividing by Ax. This pro- cedure assured that the total instantaneous volume of water within a subreach was the same for both models. The velocity at any grid in the equally spaced model was assumed to be equal to the value at the point in the “true” river. Top widths for the temperature model were deter- mined from the cross-sectional areas provided by the flow model. For each available cross section, the measured relation between width and area was fitted with a third degree polynomial of the form W = a0 + alA + (12A2 + a3A3 (41) in which a0, a1, a2, and as are fitted coefficients and A is the given area. MODEL CALIBRATION AND VERIFICATION FLOW MODEL The most sophisticated mathematical procedures are of little value without adequate data to verify them. Data needed for a flow model include internal reach data, which describe the physical characteristics of the river (geometry, channel elevation, and roughness), and boundary condition data such as flow or stage at each end of the reach and flow in each tributary. For modeling purposes the internal characteristics of the river are discretized at a number of grid points which represent the longitudinal variations of channel geometry, elevation, and roughness. The required number of grid points is a function of the objectives of the study as well as the frequency of temporal variations in the boundary conditions. In this case it was desirable to model extreme flow changes (discharge varying by a factor of 14) which occur in very short time periods (about 10 minutes). Assuming these flow changes are equivalent to a periodic function with a 20—minute period, a 10-minute sampling period should be statisti- cally adequate (Bendat and Piersol, 1966). This provides two samples per cycle of the highest frequency change. A 5-minute time step was used thorughout this study. The magnitude of the time step generally dictates the spacing of the internal grid points. Stability and ac- curacy are related to the Courant number = (U + ng )At Ax (42) CN Explicit models and the method of characteristics become unstable at CN> 1. Even in highly unsteady flow as in the Chattahoochee, implicit models operate satisfactorily for values of CN as large as 15, but ac- curacy decreases as the value of CN departs from unity. The average velocity in the Chattahoochee varied from 0.33 m/s at steady low flow to 0.79 m/s at high flow, while the average depth varied from about 1.1 to 2.2 m. The distance step required for maximum accuracy therefore varied from about 1.1 km at low flow to about 1.6 km at high flow. The actual spacing of the cross- sectional data depended somewhat on field conditions, but the average spacing was 0.7 km (table 1). All hydraulic data were obtained by project personnel of the Chattahoochee River Quality Assessment Project, and a complete description of the data and methods used in its acquisition is in progress (R. E. Faye, US. Geological Survey, written commun., 1978). Briefly, the cross-sectional data were obtained at high flow by use of a sonic depth sounder and a boat. Absolute bed eleva- tions were obtained by referencing the water—surface elevation, at the time the cross section was taken, to references which had previously been set on the bank. The field crew also estimated the flow resistance (Mann— ing n) at the time of the field survey. The cross-sectional data were processed as follows. First the sonic sounder charts were digitized to form coordinate pairs which described the shape and eleva- tion of the cross section. A program was then developed which produced tabular values of area and top width ver- sus maximum depth and fitted these tabular values with an expression of the form A=Ty a m + 1/3(Tby?;n) (43) in which gm 2 maximum water depth in the cross section, Ta=bottom width of the channel, and T1, is the shape factor which determines the rate of increase of width with elevation. The coefficients for equation 43 are shown in table 1 along with the location of the cross section and other 12 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA TABLE 1.—Intemal reach data for the Chattahoochee River between Buford and Norcross Coefficients in Measured Manning’s Section Comments River Bottom equation 43 h draulic roughness number kilometer elevation epth at (m) _1 low flow ‘1 T. (m) Tb (m > (m) n. rum > 1 Buford Gage _______________________________________ 560.20 277.53 76.2 0 0.82 0.060 0.0 2 560.00 277.22 62.2 0.67 .52 .086 .0 3 Interpolated Section ___________________________ 559.76 276.79 54.9 0 1 .55 .054 .0 4 Interpolated Section ___________________________ 559.46 276.19 54.9 0 1 .55 .056 .0 5 559.12 275.47 49.4 1.35 .58 .051 .0 6 558.56 274.67 36.6 .45 1.76 .038 .0 7 557.67 274.18 43.9 .39 1.05 .026 .0 8 556.93 273.85 39.0 .81 1.36 .025 .0 9 Highwa 20 ............... 556.51 273.55 64.0 .44 1.65 .030 —.003 10 Interpo ated Section. 556.30 273.52 48.8 .46 11.68 .041 -.003 11 James Creek ...................................... 556.06 273.00 36.6 .63 2.19 .030 —.003 12 555.26 274.06 46.3 .99 1.11 .021 .0 13 555.01 274.03 36.6 .66 1.12 .021 .0 14 554.72 274.00 34.1 .88 1.14 .018 .0 15 553.87 273.87 41.5 .49 1.24 .018 .0 16 Settles Bridge ____________________________________ 552.97 274.48 63.4 .88 .57 .030 .0 17 552.34 274.20 48.8 .68 .74 .030 .0 18 551.26 273.98 41.5 3.71 .84 .024 —.007 19 Inter Olated Section ___________________________ 550.80 273.55 41.5 .66 11.04 .046 —.007 20 Leve Creek .............. 550.41 272.98 42.7 .58 1.22 .080 —.015 21 Interpolated Section__ 549.86 272.64 46.3 .56 11.23 .092 -.012 22 Dick Creek _______________________________ _ 549.38 272.27 51.2 .64 1.23 .110 .0 23 548.56 271.72 42.7 .21 1.34 .088 .0 24 547.47 271.53 43.9 .23 .80 .058 .0 25 Littles Ferry ______________________________________ 546.95 270.61 39.6 1.02 1.63 .026 .0 26 546.50 270.74 48.8 .97 1.49 .016 .0 27 Interpolated Section ___________________________ 546.02 271.39 47.2 .98 1.82 .019 .0 28 545.54 271.34 46.0 1.75 .77 .023 .0 29 544.97 271.10 51.8 .51 .91 .023 .0 30 Interpolated Section ___________________________ 544.60 271.08 48.8 .52 .88 .026 .0 31 Suwanee Creek _______________ __ 544.26 271.05 47.2 .58 .88 .026 .0 32 Gwinnett County Intake. 544.04 271.03 47.2 .58 1 .88 .029 .0 33 Interpolated Section ___________________________ 543.81 271.09 50.3 .49 1 .76 .027 .0 34 543.59 271.12 57.9 .46 .70 .026 .0 35 543.09 270.79 51.8 .58 .96 .035 .0 36 542.24 270.69 45.1 .0 .84 .041 .0 37 541.24 269.82 40.2 .80 1.12 .051 .0 38 540.16 269.41 40.2 1.15 .95 .061 .0 39 Highway 120 ...................................... 539.55 268.59 51.8 .51 1.01 .050 .0 40 538.78 268.53 42.7 .23 .87 .033 .0 41 537.92 268.04 36.9 1.01 .94 .040 .0 42 536.96 267.68 56.1 .0 1.07 .040 .0 43 536.12 267.46 51.8 .0 1.19 .040 .0 44 535.61 267.31 41.8 .93 1.29 .041 .0 45 535.01 267.10 53.3 .59 1.46 .046 .0 46 534.14 267.10 54.9 .33 1.40 .045 .0 47 533.35 267.46 54.9 .73 .91 .046 .0 48 Highway 141 ...................................... 532.32 266.98 64.0 .91 1.25 .051 .0 Average _____________________________________________ Length of reach is 27.88 km 1.05 0.042 llnterpolated value. pertinent information. The bottom elevation, in table 1, is the bed elevation of a cross section with a shape defined by equation 43 which was judged to best repre- sent actual measured cross section. In general this was the mean elevation of the channel bottom at low flow. In addition to the above internal reach data, which are more or less the standard field data collected for flow modeling purposes, the hydraulic depth at each cross section was measured on July 17, 1976, under conditions of steady low flow. These depths, shown in table 1, were obtained by averaging 3- to 10-point measurements ob- tained at uniform spacing across the river. Boundary condition information consisted of a con- tinuous record of stage at the upstream end of the reach as well as discharge at the four tributaries and at one withdrawal point. A stage discharge rating curve was also available at the upstream end of the reach. This stage-discharge relation was used along with the recorded stage to drive the model. Normally this would be poor practice but was justified here. Unique rating curves apply only to steady flow under constant in- fluence of downstream backwater, if present. Under unsteady flow conditions, a different value of stage will be obtained for the same discharge depending on MODEL CALIBRATION AND VERIFICATION 13 whether the discharge is increasing or decreasing and how fast. The Chattahoochee River below Buford is hydraulically unique. First, the flow is totally governed by releases from the Buford powerplant. These releases occur in fixed increments. Long periods (3 to 5 hours) of steady flow separated by rapid changes (10 to 15 minutes) are the rule. the reach of river from the dam to the Highway 20 bridge is very steep, with supercritical flow at several locations. Virtually no upstream reflec- tion of waves is possible and backwater effects are nonexistent. Thus, when a flow change is made at Buford Dam, the flow at the gage located less than 0.4 km downstream stabilizes rapidly. During periods of unsteady flow, when the hysteresis loop rating should be considered, the time period of the loop is shorter than or equal to the time steps or the resolution of the model. Thus, no great inaccuracy is involved in using the rating curve as a boundary condition. Several runs were made using the stage directly as a boundary condition with no significant change in results. Field reconnaissance and topographic maps indicated that a rock outcrop about 10 km below the Highway 141 Bridge controlled the depth in the lower end of the reach. The downstream boundary conditions were simu- lated, therefore, by assuming that the water between the bridge and this control was ponded to an average depth of about 2 m at low flow. The Manning’s rough- ness coefficient over the control was assumed to be 0.016. Measured stage could have easily been used as a downstream boundary condition, but results were con- sidered good enough without this refinement. In addition to the boundary data, at least partial records of stage and discharge were available at the Highway 20 Bridge, Littles Ferry Bridge, Highway 120 Bridge, and the Highway 141 Bridge. Discharges on tributary streams were virtually con- stant for both periods, so a constant flow was assumed. The observed values are shown in table 2. The calibration of the flow model centered on the March data and involved a two-step process. The first step was to calibrate the model at steady low flow, and the second step involved additional calibration necessary to match the dynamic response of the system. TABLE 2.—Discharge values for the tributaries and withdrawal points on the Chattahoochie River during the October 20-25, 1975, and the March 21-24, 1976 modeling periods October March Tributary flow 0w (m3/s) (ma/s) James Creek 0.5 1.1 Level Creek _. .3 .6 Dick Creek___. .3 .7 Suwanee Creek _____ 1.8 4.0 Gwinnett Co. Intake __________________________________ —.2 —.2 The steady low flow calibration was accomplished as follows. The measured depth was added to the bottom elevation (table 1) and an “observed” water-surface pro- file was plotted. For steady flow the energy equation can be integrated between any two cross sections to give U2 U2. n? UU- Ax _1+EW,= (”n+EW,+1+ ‘2]; (1;; (44) 29 Rm RH(i+1) in which EW, =water-surface elevation at grid point i, and n,=Manning’s roughness coefficient applicable to the subreach between grid points i and i + 1. Equation 44 was easily solved for the unknown roughness coeffi- cient, m, applicable to each subreach since all other terms were known. In a few cases errors in the water- surface elevation were detected. These showed up as subreaches where the water appeared to run uphill or where the computed n value was unrealistically small or large (less than 0.005 or greater than 0.1). When this situation occurred for a subreach, the value of n was set equal to a realistic value, such as the value estimated in the field, and a corrected water-surface elevation for the grid point was computed. The bed elevation at the grid point was then established as the water-surface eleva- tion minus the measured depth. The roughness coefficients computed by use of equa- tion 44 are applicable to the subreach between grid points, and the dynamic model requires roughness coef- ficients applicable to a subreach centered on the grid point. Some judgment was necessary, therefore, in averaging the roughness coefficients computed from equation 44 to obtain values, which were used in the dynamic model. Using the roughness values, tabulated in table 1, the dynamic flow model was run to equilibri- um at steady low flow and the surface profile computed. This computed profile can be compared to the “observed” profile in figure 8. The above procedure assured that the flow model gave realistic depths, volumes, and surface areas, at least at steady low flow. The ability to match the depths at all cross sections under steady low flow conditions does not guarantee that the model will reproduce unsteady flow. In order to match the dynamic response of the system, it was found necessary to vary some of the roughness coefficients with depth. These variations were necessary to make the modeled and observed rises in stage during the hydropulses agree. The roughness was assumed to vary with depth as My) = no + 7210/ - y,,,.,) (45) in which ri(y)=roughness coefficient at depth, y, n0=steady low-flow roughness (table 1), n1=rate of 14 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 275 — Highway 20 Bed elevation 270 — EXPLANATION ELEVATION ABOVE MEAN SEA LEVEL, IN METERS 0 Observed Modeled I I I I | Water-surface elevation Flow = 15.6 m3/s at Buford Littles Ferry Highway 120 Duluth Highway 141 Norcross I | I I I I 265 560 530 FIIVEFI KILOMETER FIGURE 8.—Steady flow depth profile for the Chattahoochee River between Buford Dam and Norcross. change of roughness with stage, also shown in table 1, and ymzmeasured hydraulic depth at low flow. The grid points at which the roughness was to be varied with depth and an approximate value of m was determined from a sensitivity analysis using a simple backwater pro- gram for different flows. Once the model was calibrated at steady low flow, dynamic calibration using the March data only required seven of the roughness values be varied with depth. The variation with depth was not ex- treme. At section 18 the roughness decreased by 85 per- cent as the flow increased to its maximum, but varia- tions in the roughness at other sections were less than 50 percent. One set of adjustments was made to the flow model as a result of observations of the behavior of the transport model. The low-flow traveltime in the upper reach was increased slightly by arbitrarily increasing the low—flow cross-sectional areas, above the values indicated in table 1 at section numbers 7 through 15 in the pool above the control at section 16, by an average of about 36 percent. This adjustment was believed to be justified because of the rather poor quality of the low-flow depth informa- tion in this reach. The adjustment had very little effect on the modeled stage or discharge values but improved the low-flow timing of the transport model at Littles Ferry. The results of the final calibration are illustrated in figure 9 in which the observed and modeled stages at Buford Dam and the four bridges are plotted for the 31/2 —day calibration period. The small rise in the observed stage at Highway 141 on March 21 was caused by a light rain which occurred before 0600 that day. No tributary flow measurements were available during this day, so constant values were assumed in the model. Visually, the stage predictions are good. The cross correlation of observed and modeled stages indicates peak correlation coefficients of 0.997, 0.997, 0.999, and 0.988 at lags in minutes of +20 for Higway 20, —5 for Littles Ferry, —30 for Highway 120, and — 10 for Highway 141, respectively. A positive lag indicates the model lagged behind the observed. At zero lag the correlation coeffi— cients were 0.989, 0.996, 0.979, and 0.987, respectively. During the March run, field crews attempted to con- tinuously measure discharge at the bridges. A detailed description of the manner in which these measurements were obtained is in progress (R. E. Faye, written com- MODEL CALIBRATION AND VERIFICATION 15 280 BT CT 00 OObO AT 0 b0 278 Buford 276 274 Littl F ‘ 272___ as erry Bridge ELEVATION ABOVE MEAN SEA LEVEL, IN METERS 270 Highway 120 Bridge AT 30 O O O O O O O Q\O Highway 141 Bridge 268 I I I I l I I I HT ; dé EXPLANATION O O 0 Observed _ FT g 0 O Modeled O O _ GT 0 OO O . OO O O O O 22 DAY IN MARCH 1976 FIGURE 9.—Calibration of the Chattahoochee River flow model with the March 1976 stage data. The symbols AT through IT represent the time of arrival of specific water particles at the respective locations. mun., 1978), but briefly it involved the periodic measure- ment of velocity and depth at particular transverse sta- tions, plotting the data at each station against time, and interpolating the data to a particular time in order to estimate the instantaneous discharge. A complete tra- verse of the river required about 1 hour. Figures 10 through 13 are presented so that a comparison of the modeled and measured discharges can be made. A rat- ing curve was also available at the Highway 141 Bridge, so that an “observed” discharge could also be deter- mined by use of the table and the observed stage. The agreement of the modeled and measured discharges is excellent for flows as large as 110 m3/s. The differences are less than 5 percent. At higher flows, the model con— sistently predicts lower than observed discharges. The peak modeled discharge on March 23 at Littles Ferry and Highway 141 was 20 percent lower than the meas- ured value. Because discharge measurements under highly unsteady flow conditions are of questionable ac- curacy, the 20-percent difference in results was not con- sidered serious. Overall, the model results were con- sidered to be very good. The consistency with which the transport model reproduced the dye concentrations bears out the accuracy of the flow model. The model was verified by use of the data collected during October 1975. A comparison of the predicted stage values are within 0.15 m of the observed values in most cases, and timing is accurate to within about 30 minutes. Cross correlation of the measured and modeled stage data indicate correlation coefficients of 0.995 and 0.988 when measured values were lagged by +25 and +40 minutes at Georgia Highways 20 and 141 respec- 16 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 250 I I l I I I I I I d I I I EXPLANATION ° 0 Observed (3)0 Modeled o o 3 200 — o — < < 150 — c — 100 DISCHARGE, IN CUBIC METERS PER SECOND 50 21 22 23 24 DAY IN MARCH 1976 FIGURE 10.—Comparison of modeled and observed discharge in the Chattahoochee River at the Highway 20 Bridge during the March 1976 calibration period. tively. Zero lag correlation coefficients were 0.971 and 0.974, respectively. The model does appear to under- estimate the stage at Littles Ferry and Highway 20. The match could be improved by decreasing the roughness, no at sections 16 and 17 to 0.021 from 0.030, setting the value of n, to +0.0033 for sections 25, 29, 35, and 37, and setting the value of 121 to +0.00% and +0.0056 at sections 38 and 39, respectively. The predicted and ob— served stages obtained with these updated resistance coefficients are shown in figure 15 which illustrates that the visual effect of these corrections is small. Use of the updated coefficients with the March data underpre- dicted the high stage at Littles Ferry by about 0.5 m, otherwise, the results were similar. Only a few discharge measurements were available for the October run. These were made on October 23 and 24 and are illustrated in figure 16. The measured peak discharge at Littles Ferry on October 23 was 12 percent higher than the modeled value obtained with the up- dated roughness coefficients. As with the spring run, the agreement between the modeled and observed dis- charge on the rising limb of the hydrograph is good. It was assumed that a slight change in roughness oc- 250 I I T I I I r I I I I I /| EXPLANATION 0 Observed o Modeled D 200 '— o _ Z O 0 In In a: u.I n. g 150 — — E < In E 9 < m 00 D 0 Z 100 — _, '. < u.I 0 t: < I 3 ‘ ° 0 50 — _I o I I I I I I I I I I I I l_ 21 22 23 24 DAY IN MARCH 1976 FIGURE 11.——Comparison of modeled and observed discharge in the Chattahoochee River at the Littles Ferry Bridge during the March 1976 calibration period. curred beween October 1975 and March 1976, so the roughness coefficients given in table 1 were used when generating a flow field for the spring transport model, and the updated coefficients, given above, were used in generating the flow field for the fall transport model. MASS TRANSPORT OF A CONSERVATIVE SUBSTANCE It is generally recognized that neither equation 9 or 10 accurately represent the longitudinal mixing (disper- sion) of a slug injection until considerable mixing has oc— curred. In fact, the criteria given by Fischer (1973) sug- gest that the equations will not accurately represent the dispersion of a slug injection in the Chattahoochee within the first 7.3 km at high flow or within the first 8.6 km at low flow. On the other hand, it has been shown that dispersion plays an almost insignificant role for the case of a steady injection rate (Sayre and Chang, 1968). In consideration of the above, an independent verifica- tion of the transport model was deemed desirable. To generate data against which the transport model could be verified, rhodamine-WT dye was injected into the river just below Buford Dam starting at 1100 hours, March 21, 197 6. The injection rate was held constant for MODEL CALIBRATION AND VERIFICATION 17 25°Ilifilllllllll EXPLANATION 0 Observed Modeled D 200— a _. z o o o a 09’ II Lu 0.. E? 150— _. LIJ ’— LU E 9 m :> o E 100— — LLI w I < I o g o 50— — o I l I I I I I I l I l I 21 22 23 24 DAY IN MARCH 1976 FIGURE 12.—Comparison of modeled and observed discharge in the Chattahoochee River at the Highway 120 Bridge during the March 1976 calibration period. a 3-day period by use of a small positive-displacement pump. Figure 17 contains a photograph of the injection site. The dye was pumped from the barrels through a tub, hung on a cable, to the center of the channel. The dye fell about 3 m before striking the water surface. Three hours after injection began, sampling began at Littles Ferry Bridge, and 6 hours later it began at the Highway 141 Bridge. Dip samples were collected in 25 mL bottles clipped to an angle iron. Samples were taken from the thalweg of the river every 10 minutes, and at 6—hour intervals samples were collected at 6—m interVals across the river. During times of steady flow, the max- imum variation in concentration across the width of the river was about 5 percent. Unfortunately, during unsteady conditions, the traverse results were relatively meaningless because of the time required to obtain the samples. Dye concentrations at Littles Ferry and Highway 141 Bridges were then simulated by both the conservative (eq. 9) and nonconservative (eq. 10) transport models. The results are presented in figures 18, 19, 20, and 21 along with the measured concentration values. The model results in figures 18 and 19 were obtained by use 250 I I | l l I I I l l I l r EXPLANATION 0 Observed (current meter) I] Observed (rating table) Modeled 200 — — 150 100 DISCHARGE, IN CUBIC METERS PER SECOND 01 O 21 22 23 24 DAY IN MARCH 1976 FIGURE 13.—Comparison of modeled and observed discharge in the Chattahoochee River at the Highway 141 Bridge during the March 1976 calibration period. of the conservative model (solution to eq. 9 with une- qually spaced grid points), and the results in figures 20 and 21 were obtained by use of the nonconservative model (solution to eq. 10 with equally spaced grid points). The model results presented in figures 18 and 19 are considered to be excellent. Cross correlation of the measured and modeled concentrations indicated correla- tion coefficients of 0.992 and 0.952 when the measured values were lagged by 15 and 0 minutes, respectively, at the Littles Ferry and Highway 141 Bridges. At Littles Ferry the zero-lag correlation coefficient was 0.986. The results presented in figures 20 and 21 are also considered to be excellent. Comparison of figures 18 and 20 shows that the conservative model maintains slightly better timing of results at Littles Ferry but that its solu- tion is slightly more dispersive even though it was run with the dispersion coefficient set equal to zero. The con— servative model also provides a better simulation of the anomalous rise in concentration beginning about 1000 hours on March 22. On the other hand a comparison of figures 19 and 21 indicates that the non-conservative model gave slightly better timing results at Highway MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 18 £5382 2503mm: 2: an 33: L3 ES,» ”£6me we Etta mo 2:5 2: E333; 28E? SE. .3me wwfim 3.3 52800 2: ME? ESE Boa 32m wwnooozafimzo 2: mo :ocwomcmvxrldfi 5:5; mm mN mhmw mmmOFUO Z. >95sz 00...... \2 amutm our $323... wontm >twu 32:1. omntm om $5591 .225 l wwm Ohm NhN dhm ohm I mum I omN NwN SHELBW NI ”IE/\E‘I VEIS NVEIW 3AOEIV NOILVAEI'IEI 19 MODEL CALIBRATION AND VERIFICATION @5582 338%2 23 as 365 L3 hoax? 258% mo Rita mo 0:5 23 €meth 208:? SE. .Smc owmum mum: $3800 0% ME? ESE Boa 53m emcooosmfiwso ofi Mo seashgmolhfi mchE mhmr ImmOFUO Z_ >LOMDO I o o o a O ZO_._.939: wwutm >tou. 3.33 $25 on >539... Beam 00'. 00‘ \S | wwN mwN CNN NhN VBN ohm whN omN NwN SHEILEW NI "IE/\El'l V35 NVEIW BAOSV NOILVAB‘IE! 20 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 25° ILittlesI Ferry I October 23, 1975 200 _ EXPLANATION _ on... Observed 1‘50 _ Modeled _ D I“ Z 8 LL! 100 In I E 50 (I) CC u.I I- o I l I u.I E 0 6 12 18 24 0 TIME, IN HOURS 3 D 250 I I I I I I 0 Highway 120 Bridge E October 23, 1975 October 24, 1975 u.I 200 " EXPLANATION .' _ 0 n: o o o . o Observed 5 § 150 ._ Modeled : U 0 Q Q 100 50 O | I I I I L I 0 6 12 18 0 6 12 18 24 TIME, IN HOURS FIGURE 16,—Comparison of the modeled and observed discharges in the Chattahoochee River during the October calibration run. FIGURE 17,—Dye injection site for the March 1976 dye study, 200 m downstream of Buford Dam. 141. On the whole, however, the results of both models appear excellent, and their differences appear to be trivial. Because the nonconservative model appeared to satisfactorily represent the transport under the highly 15 I I I I I I I I I I T I I m EXPLANATION II_-I - o . - o Modeled 3 : Observed I 12 — : . _ I: E :- . g E ‘3 5" < ' .' n: ' - 0 ° ' Q 9 ’— : o' - I ' o o ‘ - — I 2 : .' E : 0 ~ 6 D . w — . _ g i ,: < E z 3 — — Lu 0 Z O U o l I l I I I I l I I I 21 22 23 24 DAY IN MARCH 1976 FIGURE 18.—M0deled and observed dye concentrations at Littles Ferry Bridge as computed using the conservative transport model with unequal grid spacing. I I I I I I I I | I ‘5 I I I I: EXPLANATION LII-J o . . o o - observed 3 Modeled a: 12 — . — uI ’ a no a) 0. E < I I: : 0 9 — o —' g : g 5 2 0 Z T. 6 — _.. Z 9 l— < I! E 3 — _ u.I 0 Z 0 U 0 I I I. . . L . . 21 22 23 24 DAY IN MARCH 1976 FIIIIZRE 19.—Modeled and observed dye concentrations at Highway 141 as computed using the conservative transport model with un- equal grid spacing. unsteady conditions in the Chattahoochee and because the existing conservative model did not allow for surface exchange or the solution to equation 31, the tempera- ture model was not recoded to solve the conservative form of the transport equation. MODEL CALIBRATION AND VERIFICATION 21 15 I I I I I I I I ‘I I EXPLANATION E g con-o- Observed I: 0'. Modeled _l :' It 12 '— g: g — LU . l ' m z 5 < a: o 9 — _ O a: 9 E E . 5 — — Z 9 y. < n: '2 u.I 3 _ .— 0 Z O U o l l n I'- 21 22 23 24 DAY IN MARCH 1976 FIGURE 20.—Modeled and observed dye concentrations at Littles Ferry Bridge as computed using the nonconservative transport model with equal grid spacing. 15 I I I ‘l EXPLANATION o 0 I I 0 Observed Modeled 12 — _ ji I I I I | CONCENTRATION, IN MICROGRAMS PER LITEFI 24 DAY IN MARCH 1976 FIGURE 21.—Modeled and observed dye concentrations at Highway 141 as computed using the nonconservative transport model with equal grid spacing. THE TEMPERATU RE MODEL Four types of data are needed for use with a temper- ature model. These are: hydraulic data which describe the flow; physical data which define the geometric rela- tion between the river and its surroundings; meteoro- logic data which help define the heat flux at the air- water interface; and temperature data at the upstream boundary and at all tributaries. Hydraulic data input to the temperature model in- cluded flow velocity and area as well as the tributary flow (both the steady and variable component) at each grid point and time step. These data were passed to the temperature model directly from the flow model. Physical data included the azimuth, effective barrier height, bank width, and the relation between top width and flow area for each grid point. Top widths were determined for each cross section by use of equation 41. Spot checks indicated the relations defined by equation 41 were accurate to within 1.5 percent most of the time for within-bank flow, which is all that is of concern here. The azimuth of each subreach, the effective barrier height, and the coefficients for use in equation 41 are shown in table 3. The azimuth values, centered at the grid point, were determined by use of a 1:24,000 scale topographic map, and the effective barrier heights, figure 6, were estimated in the field at the time the longitudinal depth profile was measured. The banks of the Chattahoochee are very steep and the trees often lean out over the river as shown in figure 5. A constant bank width of 0.3 m was assumed. In so far as the shading computations were concerned, the river top width was assumed to have a value representative of steady low flow at each grid point. The thermal properties of the river bed might also be classified as physical data. The bed of the Chattahoochee River is mostly covered with small sand dunes so the thermal diffusivity and heat storage capacity were assumed to be 0.77 mmZ/s and 0.68 cal/cm°C, respec- tively. Braslavskii and Vikulina (1963) suggest these values are applicable for saturated sand. Stream temperatures in the Chattahoochee River and its tributaries were monitored by personnel of the Chat- tahoochee River Quality Assessment Project. These data and the details of their collection will be reported elsewhere (R. E. Faye, written c0mmum., 1978). The locations of the data collection points, however, are shown in figure 1. During the October run the temperature recorder at the upstream end of the reach malfunctioned, so the upstream boundary of the temperature model was set at the Highway 20 Bridge, which is 3.7 km downstream of the boundary of the flow model. Water temperatures at Highway 20 were available on 15-minute intervals, start- ing at 1300 hours on October 21, 1975, and continuing for the duration of the run. Before 1300 hours on the 21st, the upstream temperatures were estimated. Hour- ly temperatures were available for all tributaries and at the Highway 141 Bridge for the entire Fall run. Inter- 22 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA TABLE 3.—Phystcal data relevant to the Chattahoochee River between Buford and Norcross Temperature Coefficients in equation 41* Azimuth Effective mode barrier grid height 110 a, (12 L13 (m) 1 61.10 39.79 —19.459 2.921 227.3 12 2 60.29 —13.68 13.130 -2.027 214.4 12 3 17.15 80.59 —47.530 8.878 185.7 18 4 35.34 62.55 —48.568 11.658 212.7 12 5 36.28 30.05 -8.855 .948 194.4 12 6 5.29 68.89 —12.038 —5.080 209.9 11 7 12.15 85.74 -—36.101 4.886 208.4 18 8 22.23 60.64 —35.677 6.918 180.2 18 9 20.19 72.47 —34.403 5.035 208.7 18 10 40.48 45.86 —14.771 1.754 212.4 18 11 56.86 7.68 —1.768 .323 192.1 18 12 17.19 137.49 —138.739 42.648 191.3 18 13 40.26 41.25 —26.820 5.800 186.2 18 14 16.29 134.83 —89.536 17.956 198.9 18 15 26.75 63.42 —36.898 6.639 130.8 18 16 36.81 52.39 —35.384 7.307 154.8 18 17 11.39 74.16 —36.025 5.532 113.7 20 18 17.12 93.75 —68.102 15.606 201.5 15 19 35.26 61.05 —43.448 9.683 241.0 18 20 20.45 75.95 -41.749 7.144 217.2 11 21 34.32 42.99 —23.634 4.442 237.3 6 22 40.13 41.08 —27.701 6.164 258.7 12 23 50.82 16.91 —7.528 1.880 254.2 12 24 22.02 78.01 —49.816 10.118 270.6 14 25 10.43 190.32 —281.511 145.283 281.7 18 26 16.54 96.45 —60.425 11.483 246.5 15 27 20.31 85.66 —80.764 23.757 224.3 15 28 14.40 45.43 —22.654 5.028 199.5 12 29 34.30 31.69 —16.888 3.272 195.9 18 30 10.66 150.71 —157.014 48.941 186.3 18 31 25.37 71.79 —61.229 16.904 207.6 18 32 34.72 65.76 —51.248 12.714 234.3 15 33 20.90 102.24 —114.297 41.513 259.9 18 34 18.73 112.03 —77.910 16.383 339.0 18 35 25.77 48.90 -15.380 1.729 359.5 18 *Top widths are given in meters when the area is given in hundreds of square meters. mittent temperature data were available at the Littles Ferry and Highway 120 Bridges. Much more complete temperature data were collected during the March run. Beginning at 0000 hours on March 21 and continuing to 0700 hours on March 24, 1976, 5-minute data were recorded at the upstream boundary as well as at Highway 20, Littles Ferry, Highway 120, and Highway 141. Becasue of a recorder malfunction, approximately 24 hours of data were miss- ing at Highway 120, however. Five-minute data were also recorded on Suwanee Creek from 0000 hours on March 21 until 1800 hours on March 28, 1976. Only a few temperature measurements, obtained at random times, were available for the other tributaries. Using available temperatures, regression equations were derived for each tributary which predicted the instantaneous tributary temperature from the Suwanee Creek tem- perature. After 1800 hours on March 23, all tributary temperatures, including the Suwanee Creek tempera- ture, were estimated from the air temperature using a regression expression which had been derived from the random data. Before 1800 hours on March 23, the pre- dicted temperatures in Level, James, and Dick Creeks were probably accurate to within 12°C and after this time about i3°C. All temperature data were first placed on cards and then plotted to check for keypunching or instrument er- rors. Once the data were verified in this manner, the hourly or 15-minute data were expanded to a 5-minute time base by straight-line interpolation. The expanded data sets were stored on magnetic disk for use with the temperature model. The initial temperature distribution in the river was assumed to vary linearly with distance between points of observation. Meteorologic data needed to drive the temperature model include windspeed, incoming solar radiation, in- coming atmospheric radiation, air temperature, wet- bulb air temperature, and rainfall intensity. All meteorologic data were obtained at the R. M. Clayton Sewage Treatment Plant in Atlanta, Ga. This site was selected primarily because of its security, its nearness to the study headquarters, its proximity to the river, and because it Offered good exposure to the sun and wind. MODEL CALIBRATION AND VERIFICATION 23 The R. M. Clayton plant is about 55 km southwest of Buford Dam and about 85 km southwest of the Highway 141 Bridge. A propeller-type anemometer was used to sense the windspeed. The starting speed of the propeller was about 0.45 m/s with full tracking at about 1.4 m/s. The wind direction was also recorded but not used in the model. The general exposure of the anemometer is shown in figure 22. A closeup of the anemometer with the Chattahoochee River in the background in shown in figure 23. The total incoming solar radiation was determined by use of an Eppleyl precision spectral pyranometer. The instrument is sensitive to radiation with a wavelength between 0.3 and 3 pm. The pyranometer is the instru- ment on the left in figure 24. The incoming atmospheric radiation was determined by use of two Eppley pyrgeometers which are sensitive to radiation in the range of 4 to 50 pm. These instruments are shown to the right of figure 24. Two psychrometers were also used to determine the wet- and dry—bulb air temperatures. These can be seen projecting from the tower in figure 22. A closer View of the ventilated psychrometer is shown in figure 25. The temperatures in this instrument were FIGI‘RH 22.—View of meteorologic instrumentation tower of the R. M. Clayton Sewage Treatment Plant showing the anemometer and two psychrometers. sensed by use of platinum resistance temperature de- vices, and the wet-bulb probe was covered by a wick FIGURE 23.—Closeup of the anemometer showing the Chattahoochee River in the background. FIGURE 24.—Closeup of pyranometer and pyrgeometers at the R. M. Clayton Sewage Treatment Plant in Atlanta, FICI'RI-l 25,—Closeup of ventilated psychrometer at the R. M. Clayton Sewage Treatment Plant in Atlanta. 24 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA which was continually wetted by distilled water. The probes projected across a plastic tube which was shielded from radiation by a curved aluminum sheet and through which air was drawn at a speed of 4.5 m/s by a vane axial fan. A detailed description of this psychro- meter has been given by J obson and Sturrock (1976). A closer view of the nonventilated psychrometer is shown in figure 26. This instrument was originally de- signed for use during the Lake Hefner studies and has been described in detail by Anderson and others (1950). This instrument is generally called a Top Hat psychrom- eter because of the shape of the radiation shield. Tem- peratures are sensed by copper-constant thermocouples which are housed above a distilled water reservoir and within a housing that facilitates natural ventilation. Rainfall was measured by a tipping bucket rain gage. An Esterline Angus D2020 recorder was housed in the tower and used to record all meteorologic data. At a specified time interval, the time as well as the millivolt values of the 10 parameters were printed on a paper tape. No averaging of the readings was possible, so the recorded value represented only an instantaneous reading. A sampling of the 10 channels required about 5 seconds. The recording was at hourly intervals in Oc- tober and at 5-minute intervals in March. Flul'iui 26.—Closeup of Top Hat psychrometer at the R. M. Clayton Sewage Treatment Plant in Atlanta. No special processing of the windspeed and solar radiation data was required for the October period because the records were complete. October 20, 22, and 24 were essentially clear days and the radiation values indicated only a little cloudiness around noon on October 21 and 23. On October 25, it was partly cloudy. Daily average values of all meteorologic data are presented in table 4. The output of the two pyrgeometers agreed to within 3 percent, so their results were averaged for use in the model. Previous experience has indicated that the pyrgeometers indicate too much diurnal variation in the incoming atmospheric radiation, so daily average, rather than instantaneous values, were used in the model. No rainfall occurred during the October run. Both psychrometers worked satisfactorily during the October run. It was believed, however, that the ven- tilated psychrometer was a little more accurate, so its readings were used throughout. Air temperature con- tained a typical diurnal swing of about 19°C and a gradual warming trend (table 4) occurred during the run. Data coverage during the March period was not as complete. N0 meteorologic data were available before 1140 hours on March 21. It was assumed that meteoro- logic conditions between midnight and 1140 hours on March 21 were identical to those which occurred be- tween midnight and 1140 hours on March 23. Although rain is known to have occurred during this period, no record was available, so it was assumed to have been zero. After 1140 hours on the let, complete data were available for the windspeed, solar radiation, and rainfall (which was zero). These values were used directly. The days of March 23 and 24, as well as the afternoon of March 21, were almost free of clouds. It was quite cloudy until about 1500 hours on March 22. The outputs of the two pyrgeometers differed by about 19 percent during the March run. The outputs of the two sensors were averaged, however, to give the values used in the model (table 4). The temperatures from the ventilated psychrometer were again used in the model except for an 8-hour period on March 24. During this period the TABLE 4.—Daily average values of melerologlc variables at the R. M. Clayton Sewage Treatment Plant in A llanta Date Windspeed Solar Atmospheric Air Vapor (in/s) radiation radiation temperature pressure (Watts/1112) (Watts/m2) (° ‘ (kPa) October 20, 1975 1.22 203 315 10.4 0.78 October 21, 1975 .84 189 323 12.6 .86 October 22, 1975 .44 192 335 13.6 .98 October 23, 1975 .51 158 350 15.2 1.25 October 24, 1975 .92 201 353 16.6 126 October 25, 1975 .43 122 368 16.5 1.52 March 21, 1976 1.92 264 346 11.0 .64 March 22, 1976 1.29 171 336 10.7 .61 March 23, 1976 .86 261 348 10.5 .67 March 24, 1976 2.64 263 373 13.6 .95 MODEL CALIBRATION AND VERIFICATION 25 values from the Top Hat psychrometer were used because the wick of the ventilated psychrometer was partly dry. During the March run the diurnal air temperature swing was about 15°C. Like the temperature data, all meteorologic data were placed on punch cards, plotted, and edited. The hourly values, for the fall run, were then expanded to a 5- minute time base by straight-line interpolation. The vapor pressures were then computed and the data sets stored on magnetic disk for use by the mode]. The temperature model contains several physical con- stants and empirical coefficients. The usual modeling procedure is to select a calibration period during which one or several of the empirical coefficients are adjusted until a “good” fit occurs. The calibrated model should then be run with an independent set of data to verify that the coefficients, determined during the calibrating period, are “universally” applicable to the particular river reach. In this case all model coefficients were assumed a pri- ori on the basis of assumptions already given. Because no coefficients were adjusted to fit the measured tem- peratures, both the March and October data should be considered as verifications of the temperature model. The measured and predicted water temperatures for the October verification are shown in figures 27, 28, 29, and 30. The observed temperatures at the Highway 20 Bridge, shown in figure 27, were used as the upstream boundary condition of the temperature model. The water temperature at the Highway 20 Bridge for times before 1130 hours on October 21 had to be assumed because no measured values were available. The as- sumption of these values of course invalidates any verification of the model during these first few hours, but it does allow the starting times of the flow and temperature models to be the same. Very few measured water temperatures were available at Littles Ferry Bridge (fig. 28) or Highway 120 (fig. 29). At Littles Ferry the data on October 22 consisted of three spot measurements which would appear to have been too high by 0.5 to 1°C. It is possible that these spot measurements were taken too near the bank or that poor procedures were used. The accuracy of periodic temperature-measuring techniques used by the US Geological Survey has been estimated to be i0.8°C (Moore, 1969; Rawson, 1970; and Blodgett, 1971). On October 23 the spot field measurements are within 0.5°C of the computed values, and the general shape of the computed and measured curves are alike. Except for the two points on October 23, the general shape and magnitude of the measured temperature distribution at the Highway 120 Bridge (fig. 29) agree closely with the predicted values. A complete temperature record was available at the Highway 141 Bridge, figure 30. The a U1 IIIIIIIIIIIIIIIIII‘IIIII Water temperature at Highway 20 d A I I —| (A) .- .- a O TEMPERATURE, IN DEGREES CELSIUS N I | I I I I I 20 21 22 23 24 25 DAY IN OCTOBER 1975 (D FIGURE 27.—Observed temperature in the Chattahoochee River at the Highway 20 Bridge during the October 1975 verification period. The symbols represent the time of arrival of specific water particles. a 01 IIIIIII‘IIII'IIIIIIIIIII EXPLANATION 14 h 0 Observed _ Modeled TEMPERATURE, IN DEGREES CELSIUS IlllIIIl I J I | I l I I 20 21 22 23 24 25 DAY IN OCTOBER 1975 IIIlII '0 FIGURE 28.—Comparison of the observed and modeled temperatures at the Littles Ferry Bridge on the Chattahoochee River during the October 1975 verification period. 15IIIIIIIIIIIIIIIIIIIIIII EXPLANATION 14 _ 0 Observed (D o — Modeled 13 12 11 10- —‘ TEMPERATURE, IN DEGREES CELSIUS 9 I I 1 I I I I I | | | I l I I I L l l I l J 1 2o 21 22 23 24 25 DAY IN OCTOBER 1975 FIGI'RI—l 29.—Comparison of the observed and modeled temperatures at the Highway 120 Bridge on the Chattahoochee River during the October 1975 verification period. 26 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 15 I I I I I I I I I I EXPLANATION 14 fl - c u o o . Observed Modeled TEMPERATURE, IN DEGREES CELSIUS Water temperature at Highway 141 i I I I I I l I l I I 20 21 22 23 24 25 DAY IN OCTOBER 1975 FIGURE 30.—Comparison 0f the observed and modeled temperatures at the Highway 141 Bridge on the Chattahoochee River during the October 1975 verification period. The symbols represent the time of arrival of specific water particles. agreement between the computed and measured tem- peratures is very good. Ignoring the first 24 hours of record, the RMS (root-mean-square) difference between the measured and computed values is 032°C, and the mean difference is 021°C. The near-perfect fit before noon on October 21 is of course the result of judicious estimates of the upstream temperature and should be disregarded in evaluating the verification of the model. The measured and predicted temperatures for the spring verification are shown in figures 31, 32, 33, 34, and 35. It can be seen from figure 31 that the 113 m3/s discharge pulse on March 23 increased the temperature by almost 2°. The shapes of the modeled and observed temperature curves at Highway 20 (fig. 32) are very similar; however, the modeled temperatures appear to be about 1.5°C higher than the observed values. It is dif- ficult to believe the model could be off this much since very little surface exchange can take place during the short traveltime between Buford Dam and Highway 20. This was particularly true during the high flow pulses around noon on March 22 and 23. At Littles Ferry Bridge the agreement between the modeled and ob- served temperatures is very good (fig. 23). The poorest agreement occurs between noon and 2100 hours on March 21. No meteorologic data were available before 1100 hours on this day, and a light rain had occurred during the night. In order to account for the heat input of the nontributary inflow caused by the rain, 150 Watts/m2 was added to the incoming radiation term for the first 9 hours of March 21. At Highway 120 (fig. 34), the poorest fit occurred between 1500 and midnight on March 21, or about 3 hours later than at Littles Ferry. The errors at Highway 120 are also believed to be the result of the rainfall and lack of meteorologic data dur- ing the first part of March 21. Finally at Highway 141 (fig. 35) the region of poorest fit occurred between 1800 hours on March 21 and 0300 hours on March 22, again about 3 hours later than at Highway 120. It appears that the combination of rainfall and lack of data created an area of poor fit which shows up at each measurement point as the water is convected through the system. The RMS difference between the observed and computed temperatures at the Highway 141 Bridge was 020°C, and the mean difference was + 009°C. These results are considered to be very good. Temperature’s occurring before 2315 hours on March 21 were not considered in these statistics becasue the time required for a particle to traverse the system was 23.24 hours on this day. DISCUSSION OF RESULTS FLOW DYNAMICS The results of the flow model may be summarized as follows. A linear, implicit finite-difference flow model DISCUSSION OF RESULTS 27 10 I I I l I I I l I I T 1 T Water temperature at Buford t. TEMPERATURE, IN DEGREES CELSIUS co O) 21 22 23 24 DAY IN MARCH 1976 FIGURE 31.—Observed temperature in the Chattahoochee River below Buford Dam during the March 1976 verification period. The symbols represent the time of arrival of specific water particles. 1 2 I I l l I I | l l I l I l TEMPERATURE, IN DEGREES CELSIUS 22 DAY IN MARCH 1976 24 FIGURE 32.——Comparison of the observed and modeled temperatures at the Highway 20 Bridge on the Chattahoochee River during the March 1976 verification period. was used to simulate the depth, velocity, and discharge at 48 points in a 27.9-km reach of the Chattahoochee River under highly unsteady flow conditions. The reach is fairly uniform in cross section but varies considerably in local slope (fig. 8). Simulated results can be compared at all points for a single, steady low-flow condition and at four points during unsteady conditons. Generally the results of the flow model are considered to be very good. Visually the comparison of the measured and modeled stages (figs. 9 and 15) are excellent. In all cases the error 28 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 1 4 I I | l I I I l I I I l I TEMPERATURE, IN DEGREES CELSIUS EXPLANATION o n u o o - Observed Modeled - 21 22 23 24 DAY IN MARCH 1976 FIGURE 33,—Comparison of the observed and modeled temperatures at the Littles Ferry Bridge on the Chattahoochee March 1976 verification period. in the modeled stage for the March calibration run was less than 0.16 m, while for flows of less than 100 m3/s the error remained less than 0.1 m. The comparison of the modeled and observed dis- charge values (figs. 10-13), while not as good as the stage comparison, was considered good. Two types of differences between the observed and modeled dis- charges were apparent. These were differences in phas— ing and in peak discharge values. Differences in the peak discharge for the small pulse of March 22 were less than 5 percent, an excellent agreement, but differences as large as 20 percent occurred for the peak discharge of the large pulse of March 23. The percentage error generally seemed to increase in the downstream direc- tion. Although the accuracy of the observed peak discharges, obtained under such highly unsteady condi- tions, may not be very good, they are consistently higher than the peak discharge computed by the model. Thus, it is apparent that the model underestimated the peak discharge. The second type of difference appears in the timing, or phasing, of the discharge curves. The phasing difference appears to increase systematically in the downstream direction, and the worst case appears to occur at the Highway 141 Bridge (fig. 13). Here the modeled and observed discharges agree very well on the rising limb of the hydrograph. The discharge computed by use of the rating table appears to be inadequate on the rising limb of the hydrograph. During a rising stage the hydraulic gradient in a river will be larger than under the steady state conditions upon which a rating curve is based, thus, the velocity and discharge will be larger than the rating table value. On the recession the reverse effect should occur although perhaps to a lesser extent. The observed and rating table values agree very closely on the recessions of figure 13, but the modeled values are much lower. The general phenomenon discussed here is commonly known as a looped rating curve. The reasons this was not significant at the upstream end of the study reach have been discussed earlier under boundary condi- tions. The discharge results for the March run are plot- ted in figures 36 thorugh 39 for the bridges at Highways DISCUSSION OF RESULTS 29 14 | I I I I I I I I I I I I TEMPERATURE, IN DEGREES CELSIUS I I I | | I I I I I I I I I EXPLANATION 0 - 0 ' 0 0 Observed Modeled 21 22 23 24 DAY IN MARCH 1976 FIGURE 34.—Comparison of the observed and modeled temperatues at the Highway 120 Bridge on the Chattahoochee River during the March 1976 verification period. 20, Littles Ferry, 120, and 141, respectively, in a form which demonstrates the looped ratings. Except at High- way 120, the model consistently predicts a more pro- nounced loop than was observed. Both the model and the observations indicate that the stage-discharge loop is affected by the magnitude of the pulse. The results shown in figures 36 through 39 probably present the model results in the worst possible light because for the modeled and observed curves to match, the model must simultaneously reproduce the correct stage, discharge, and rate of change of discharge with stage. The results at Littles Ferry and Highway 120 Bridges are con- sidered very good. The model does appear to have some systematic bias at Highway 141 which may be caused by some bias in the assumed channel characteristics downstream of the bridge, that is, the self-setting downstream boundary condition. The channel charac- teristics in the vicinity of the Highway 20 Bridge were believed to be less accurate than the values for other parts of the reach. The major contribution of this study was to apply ex- isting solution techniques, with certain minor im- provements, to a very comprehensive data set in order to verify the existing technology of flow modeling. There are two questions that should be asked at this point. What can be done to improve the mathematical model, and what additions to the data base or calibration procedures could be incorporated to improve the results? The first question is difficult to answer. In general, the linear-implicit technique is an entirely satisfactory method for routing flow. It is not perfect, however, and an uninitiated user will have problems. These problems will probably be related to boundary or initial conditions at first and later to schematization problems. Contrary to the literature, the linear-implicit scheme is not unconditionally stable. Very good initial condi- tions are required, at least for models configured with nonuniform grid spacing and nonprismatic cross sec- tions. For the Chattahoochee, a compatible backwater program which computed the initial velocities and depths was required. The use of the term compatible is 30 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 14 EXPLANATION o o a 0 Observed TEMPERATURE, IN DEGREES CELSIUS Modeled 6 I I I I I I I I I l I I I 21 22 23 24 DAY IN MARCH 1976 FIGURE 35.——C0mparison of the observed and modeled temperatures at the Highway 141 Bridge on the Chattahoochee River during the March 1976 verification period. The symbols represent the time of arrival of specific water particles. intentional. The friction slope terms in the backwater program must be averaged in the same manner as in the dynamic model. To reduce programming difficulties in the dynamic model, a linear average was used in both the dynamic and backwater programs. Instabilities can also result from boundary conditions. An unrealistic rating curve or a self-setting boundary condition based on discharge only can produce oscilla- tions or waves at the upstream end. Each stream will have its own peculiarities, and experience is the only way to discover them. Schematization problems not related to boundary con- ditions usually take the form of an anomalous grid point depth. For instance, the flow may proceed smoothly down a uniform slope through uniform sections and sud- denly double in depth or produce some peculiar water- surface irregularity. The problem is not an instability, it just doesn’t appear realistic. In the Chattahoochee model, two such places were encountered. The first oc- curred at river-kilometer 551.26 (fig. 8), and the other was at river—kilometer 541.24. These problems can usually be alleviated by adding an interpolated section to shorten the distance step. Apparently the friction and bed slopes are too different, and extrapolating these slopes between grids gives unrealistic conditions. A compatible backwater program is very valuable for spotting and correcting problem areas. Experience with the Chattahoochee model indicates that any change in stage at a given steady discharge, accomplished by changing the roughness coefficients or other parame- ters in the backwater program, will produce an equal change in stage under unsteady discharge in the flow model. The backwater calculations can be performed quickly on a desk-top calculator, so many trials can be evaluated quickly. Efforts which have gone into improv- ing backwater programs, such as expansion loss coeffi- cients, should now be incorporated into unsteady flow models. Another area deserving further study is the method used to represent the cross-sectional shape. Aesthetic appeal is certainly important in some instances. The question can be raised as to whether equation 43 is an DISCUSSION OF RESULTS 31 279 l 1 l EXPLANATION 0—0 Observed for large pulse A——A Observed for small pulse Modeled for large pulse ——-— Modeled for small pulse (/1 E 278 -— __ m A ' E z 5 ii A > ’4 LIJ .J (n ,1 //é" 7 ‘ A Q 2 277 — A; ' V _ < // A ‘ ’4 / u.| / ,, , r4 2 / A 5 [x LU / 0 > / a 2 < / ,1 s ‘ E ’5‘ / I; ,1 , a l- ,, a 4 / 4 g '1 'LV 0 a V B 276 — o a 4 _ u a n ’A / 0" a 2 / 'I / '. / 0 / .. / / / A // 0 l l l | 7 2 50 50 100 150 200 250 DISCHARGE, IN CUBIC METERS PER SECOND FIGURE 36.—Loop rating curves for the Chattahoochee River at the Highway 20 Bridge during the March 1976 calibration. adequate representation of the cross-sectional shape, or should the actual measured shape be used for each sec- tion. Little research has been done on the question, and this study did not address the problem. It is the writers’ opinion that computed results are not very sensitive to the actual cross-sectional shape as long as the correct areas are given as a function of stage. In summary, three areas should be investigated in order to improve the mathematical model. First, more needs to be known about the best way to handle bound- ary and initial conditions. Second, much of the existing knowledge used in sophisticated backwater programs should be incorporated into dynamic models. Finally, the question of how to best represent the cross—sectional shape should be studied. The collection of necessary and sufficient field data as well as the intelligent use of these data appears to be more critical in the development of an accurate flow model than the selection of the particular solution technique. It was found that adequate cross-sectional information could be obtained quickly at high flow by use of a sonic depth-sounder and a boat. Top widths at the time of the traverse were obtained by stadia. Ver- tical control, between the bridges, while desirable, was not critical provided a depth profile at one steady flow conditions was available. The depth profile obtained at steady low flow was perhaps the most useful set of data available in calibrating the flow model. Furthermore these data were fairly easy to obtain. It would have been helpful to check the top widths when this profile was taken, however, so that cross—sectional areas could have been checked more closely. Optimization is currently popular in the model field. Optimization in this case was obtained on a trial and error basis wherein the modeler selected each new trial on the basis of the results of his previous trials and engineering judgment. It would appear that a formal- ized optimization procedure which incorporates all 32 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA 276 l | | | EXPLANATION G—e Observed for large pulse A—A Observed for small pulse Modeled for large pulse — — — Modeled for small pulse 3 275 —— _ w ,— Lu 2 Z J LIJ > LIJ .1 < w w z 274 -— — 4. Lu 2 w > O m < Z 9 l- < i J 273 — —l uJ 272 ,, l J | l 0 50 100 150 200 250 DISCHARGE, IN CUBIC METERS PER SECOND FIGURE 37.—Loop rating curves for the Chattahoochee River at the Littles Ferry Bridge during the March 1976 calibration. factors that were considered in calibrating the Chat- tahoochee flow model would be extremely difficult to design. Furthermore, if all these factors are not con- sidered, there exists a very real possibility that the resulting model would be far from realistic. TRANSPORT Many dye studies have been conducted in rivers of the United States. Virtually all of these, however, were per- formed as nearly as possible under conditions of steady flow and using an instantaneous slug injection. From these studies the steady flow traveltime and dispersion coefficients are inferred. Using the continuous injection procedure, the traveltime and dispersion characteristics can also be inferred from the timing and shape of the first rise in concentration under steady flow conditions. In addition, the response of the system to unsteady flow can also be determined. The continuous injection focuses the concentration changes in the regions of most inter- est, that is, in the regions where the flow is unsteady. The results in figures 18 and 19 or 20 and 21 indicate that the transport response of the system is being reproduced very well by either of the transport models. Any differences in the modeled results in figures 18 and 20 or figures 19 and 21 are due to differences in the solution routines for the conservative and nonconser- vative forms of the transport equation. Although several significant differences existed in the solution schemes, there appears to be little difference in the results. As with the flow model, it appears that the key to successful modeling results is good data and calibration techniques and not the use of highly sophisticated solution schemes. The anomalous rise in concentration on the leading edge of the power wave, shown in figures 18 and 20, is probably the most interesting result to be obtained from the dye study. The plateau concentration of 11 ,ug/L at Littles Ferry, which occurred between about 0600 and 1000 hours on March 22, is the value which would be DISCUSSION OF RESULTS 33 273 l 272 — m E m I- Lu 2 E J m > LIJ .J < u.| (I) Z ._ < 271 m E m > O m < Z 9 .— § EXPLANATION m d o———e Observed for large pulse A—A Observed for small pulse 270 — _ Modeled for large pulse — -— — Modeled for small pulse 269 1 l l O 50 100 150 200 DISCHARGE, IN CUBIC METERS PER. SECOND FIGURE 38.—Loop rating curves for the Chattahoochee River at the Highway 120 Bridge during the March 1976 calibration. computed by dividing the flow rate of dye by the river discharge at Littles Ferry. The observed peak concen- tration which occurred about 1100 hours, on the other hand, is larger than the value which would be predicted by dividing the dye flow rate by the low flow discharge from Buford Dam. The cause of this anomaly served as the subject of many discussions, and full agreement has not been reached. A plausible cause, however, appears to be the interaction of tributary inflow with the river stage. Stage variations in the Chattahoochee River are large and rapid, while considerable storage volume ex- ists in the tributary channels (fig. 4). In fact, backflow up the tributaries was observed in the field during times of rapidly increasing stage. The reduced dilution which occurs as the water wave passes a tributary and reduces or stops the tributary inflow would certainly appear to be at least a partial explanation of the anomaly. Assum- ing complete mixing in the cross section, however, this phenomenon should not raise the peak above 12.7 pg/L. Some of the anomaly may be due to inhibited transverse mixing which perhaps occurs on the leading edge of the power wave. 34 ELEVATION ABOVE MEAN SEA LEVEL, IN METERS 270 269 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA EXPLANATION B—El Rating curve 0—0 Observed for large pulse A—A Observed for small pulse Modeled for large pulse — — —— Modeled for small pulse 268 0 50 100 ' 150 DISCHARGE, IN CUBIC METERS PER SECOND FIGURE 39.—L00p rating curves for the Chattahoochee River at the Highway 141 Bridge during the March 1976 calibration. 200 DISCUSSION OF RESULTS 35 It can be noted that no negative anomaly occurs when the effect of the flow reduction passes Littles Ferry bridge at about 0800 and 2200 hours on March 23. However, the effect of the tributaries at high flow would be expected to be much less. Furthermore, the water stored in the tributaries before the flow reduction probably contains about as much dye as the river water. The flow and transport models were run for one hypothetical case in which the tributary inflow was set equal to zero, but the discharge at Buford Dam was the same as used in figures 18 through 21. An interesting, perhaps disturbing, result was that the anomaly, although reduced in size, remained. TEMPERATURE Both the October and the March runs are considered to be verifications of the temperature model because no calibration was involved. All physical parameters and coefficients were determined by deduction on the basis of other studies. In order to assess the sensitivity of the results to these determinations, however, the parame- ters were arbitrarily changed one at a time and the ef— fect on the predicted temperatures determined. In the interest of economy and simplicity, the sensitivity analyses were restricted to the March data. The sen- sitivity to bed conduction, bank shading, and the two coefficients in the wind function was checked. In addi- tion, the sensitivity of the computed temperatures to the measured values of top width and atmospheric radiation was determined. These two quantities were considered to be of somewhat questionable accuracy. Rerunning the model assuming no heat transfer at the bed (no bed conduction) decreased the mean error in the computed temperature by 0.006°C, yet increased the RMS error by 7 percent. Visual changes in the predicted temperatures were subtle, but generally bed conduction acts as a slight damper to the computed temperatures. The October data were also rerun with no bed con- duction and again the RMS error increased even though the mean error decreased. Clearly, inclusion of the bed conduction term improved the model results. The effect of shading by the banks and trees was thought to be a weakness in the temperature model. Estimates of the effective barrier heights were rather crude, so two cases were rerun. The first case elimi- nated shading entirely, and the second case reduced the effective barrier height by 20 percent. Completely eliminating the shading caused more damping of tem- perature swings at times and less at other times. It increased the mean error by 0.269°C and the RMS error by 86 percent. In general the changes were hard to characterize except that the computed temperatures with shading looked more like the measured values than did the temperatures computed without shading. As- suming that there was a tendency to overestimate the tree height in the field, the effective barrier heights were reduced 20 percent. This change increased the mean error by 0.073°C and the RMS error by 11 per- cent. The visual effect of this change was small. Wind function (eq. 20) was developed from a thermal balance of the San Diego aqueduct which is in the arid region of southern California. Because of the sheltering due to trees and the humid climate in northern Georgia, the applicability of this wind function to the Chattahoo— chee River could be questioned. The constant term, 3.01, and the mass transfer coefficient, 1.13, were varied in- dependently. Reducing the mass transfer coefficient by 25 percent increased the mean error by 0.053°C and the RMS error by 4 percent. Reducing the constant term by 50 percent increased the mean error by 0.118°C and the RMS error by 29 percent. Neither of these changes af- fected the visual fit significantly. The cross-sectional area and flow velocity were believed to be fairly accurate because of the accuracy in timing of the transport model. The top width measure- ments, however, were subject to some doubt because of the dense brush and fallen trees along the banks (see fig. 2). Without changing the flow fields in terms of velocity and area, the top widths were increased by 10 percent and the model rerun. This change, which of course decreased the hydraulic depth by 10 percent, increased the mean error in the computed temperature by 0.055°C and the RMS error by 8 percent. Reducing the atmospheric radiation by 10 percent lowered the mean temperature at the Highway 141 Bridge by 024°C and increased the RMS error by 42 percent. In summary, adjusting each coefficient or question- able measurement by an amount judged to be roughly equal to its maximum probable error always resulted in a poorer fit (in terms of the RMS) of the measured temperatures. Nevertheless, the fit could probably have been improved by adjusting more than one of the param- eters simultaneously. For example, the mean error could have been forced to zero in any of several ways, by reducing the atmospheric radiation by 3 percent, by decreasing the top width by 13 percent, by increasing the effective barrier height by 20 percent, or by some combination of these. Some combination of adjustments could undoubtedly have been found which would have reduced the RMS error significantly. Furthermore, none of the adjustments would be large in comparison with the confidence limits of the original estimate or measurement. For example, a 3-percent error in the at- 36 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA mospheric radiation term seems to be reasonable when it is remembered that the measured atmospheric term represents the average output of two pyrgeometers which differed by 19 percent in March and were located about 30 km from the study reach. No juggling of the terms to improve the fit was attempted, however. Within the limitations discussed above, the tempera- ture model provides a powerful tool for determining the effect of hydropulsations on the temperature regime of the Chattahoochee. The effect on the temperature at the upstream end of the reach will be discussed first, and then the cause of some specific anomalies at the down- stream end of the reach will be investigated. These investigations will also provide some insight into the effects of hydropulsations on other river-quality parameters. Because of the simpler flow release pattern in March, these data will be discussed first. The data presented in figure 31 indicate that the Buford outlet temperature, during times of low flow, was constant at about 7°C. The small temperature rise, of about 025°C, occurring between 0300 and 0800 hours on March 21 is probably due to heating from inflow of the rainfall which occurred during the night. The solar heating, which occurred be- tween 0900 and 1800 hours is obvious. It is perhaps sur- prising that for low flow near noon almost a 1°C temperature rise occurred over the short distance (0.48 km) between the dam and the recorder. At 0700 hours on March 22 (point A in fig. 31), the tur- bines were started to provide the first flow pulse of 110 m3/s. The river temperature increased by 0.6°C almost immediately. At 2200 hours on March 22 the flow was quickly reduced to 15.4 m3/s (point B in fig. 31) and held constant until the start of the second pulse of 220 m3/s which occurred at 0700 hours on March 23 (point C). At 1200 hours on March 23, point D, the flow was quickly returned to 15.4 m3/s and remained constant at this value for the remainder of the study. The small amount of heating after 1200 hours on March 23 is probably due to solar radiation. A gradual decay in temperature oc- curs at the completion of each flow pulse. Considerable water is ponded between the dam and the temperature gage, and this gradual decay is probably the result of the rather slow flushing of this ponded water at low flow. At high flow the traveltime and surface exchange be- tween the dam and the temperature gage is negligible so the immediate temperature rises of 06°C and 18°C, respectively, for the small and large pulses indicate that the turbines withdraw mostly hypolimnion water at low flow but at least a mixture of hypolimnion and epilim- nion water at high flow. The higher the flow the greater proportion of epilimnion water withdrawn. If any water- quality parameter has different values in the hypolim— nion and epilimnion waters of the lake, the temporal variation of the parameter in the river can be easily predicted from the data in figure 31. At steady low flow, the river parameter will be close to the hypolimnion value, but with the starting of the turbines, the value will immediately shift toward the epilimnion value. The magnitude of the shift will be a function of the discharge. After the turbines are shut down, the river value will gradually shift back to the hypolimnion value. During the October verification period, the upstream temperature distribution as well as the inflow hydro- graph is much more complex. In addition, the first temperature measurements were at the Highway 20 Bridge, 4.2 km downstream of the dam. However, the steep gradient in this short reach (fig. 8) precludes much surface exchange, except at steady low flow. Consider October 22 which was a rather typical day. More specifically consider a water particle, which is labeled A in figure 27 and which passed the Highway 20 Bridge at 1000 hours. Since the traveltime for this water (from the upstream end of the model to Highway 20) was 2.75 hours, it had been released from the dam just before 0715 hours, and steady low flow had occurred during its entire passage (fig. 15). The model indicates that this particle cooled by only about 025°C during the passage, so the water released from the dam must have had a temperature of about 98°C during steady low flow. The particle labeled B in figures 15 and 27 was released from the dam a little before 1240 hours and ar— rived at Highway 20 at 1520 hours, just after the tur- bines were started (fig. 15). The entry temperature of this particle would have been 9.8°C, therefore, and the 2.8°C temperature rise between points A and B in figure 27 resulted from surface exchange. Between 1445 and 1500 hours on October 22, the power release at Buford began. Just before this power release started, the particle represented by point C in figures 15 and 27 entered the system with a tem- perature of about 98°C. It did not experience as much surface exchange as particle B, however, because the power pulse, which followed, flushed it through the system more rapidly then the particles which preceded or followed it. Its traveltime from the upstream end of the model was only 0.74 hours as compared to 2.75 hours for particles A and B or 1.01 hours for particle D. The reduction in temperature between points B and C, therefore, is the result of the flushing action of the power pulse. The water particle presented by point D in Figures 15 and 27 entered the system a little before 1700 hours which is about the time of the peak discharge in figure 15. Assuming that surface exchange was negligible for this particle because of the large depth, short travel— time, and the low solar radiation after 1700 hours, its DISCUSSION OF RESULTS 37 entry temperature would have been 12.8°C. As in March the turbines appear to be withdrawing much more epi- limnion water during high flow. A sudden temperatue increase of 3°C (from 9.8 to 128°C) probably occurred at the upstream boundary when the turbines were started. Likewise, a sudden shift in values of other water-quality properties, which differ between the warm surface water and the cooler deep waters, prob- ably occurs each time the turbines are started. Gerald Wiley, US. Army Corps of Engineers (oral commun., 1977), has observed sudden shifts in the dissolved oxy- gen content of the discharge waters upon starting or stopping of the turbines at Buford. Finally, the particle of water labeled E left the dam a little before 2000 hours, which corresponds to the sec- ond peak in discharge in figure 15. This water traversed the 3.7-km reach in 1.07 hours. The minimum tempera- ture between points D and E in figure 27 undoubtedly is the result of a reduced release temperature during the reduced flow occurring at 1900 hours. The gradual flush- ing of the system after the flow shutdown is apparent in the gradual decay in temperature after point E just as in the March run. In light of the above discussion, it is interesting to compare the observed temperatures at Highway 20 on October 24 to those of October 22. Meteorologic condi- tions were very similar on these two days so any tem- perature differences are primarily dependent on dif- ferences in flow scheduling. On Friday, October 24, the first power pulse began 3 hours earlier than on the 22d, figure 15. Moving the pulse ahead to 1200 hours in- creased the flow in the river before the solar heating had time to influence the river temperature. This eliminated the local minimum in temperature which occurred about 1600 hours (point C) on the other days. In summary, hydropulsation had a significant effect on the water temperature below Buford Dam during both the March and October verification periods. At low flow the water temperature remained constant and low, but during the power releases, it almost immediately rose by between 06°C and 18°C in March and by as much as 3°C in October. Furthermore, the flushing ac— tion, illustrated by the temperature changes between points B and D in figure 27, helps accentuate these temperature changes and carry them downstream. At the lower end of the reach, the> effects of hydro- pulsation are still present although somewhat damped. As before, consider the March data first. Before 0700 hours on March 22 (point AT in fig. 35) a steady low flow of 15.4 m3/s existed throughout the reach. The diurnal swing occurring before this time should, therefore, be fairly typical of steady low flow conditions. A light rain occurred during the early morning hours of March 21, and although its exact influence is not known, it is believed to have increased the river temperature by as much as 1.2°C at 0900 hours on March 21 and to have in- creased the peak temperature that day by as much as 06°C. The influence of this warming is not completely dissipated until 0700 hours on March 22. Nevertheless, water particles released from the dam just after mid- night on March 21 with a temperature of 7 .0°C gained 5.8°C during their 23.25-hour traveltime and arrived at Highway 141 at 2315 hours with a temperature of 128°C. The thermal effect of the first power wave is first observed at Highway 141 about 1200 hours on March 22, just 5 hours after the turbines were started. The traveltime of the particles passing the 141 Bridge at 1200 hours on March 22 (labeled ET in figs. 9, 31, and 35) was reduced to 23.16 hours meaning that the flushing had just begun. For reference, the times at which particles left Buford Dam and arrived at Highway 141 are also indicated in figure 9 and figure 31. Stream temperatures decreased rapidly after 1200 hours on March 22 due to the flushing action and reached minimum values at 1630 hours (labeled CT in figs. 9, 31, and 35) on March 22. This water entered the reach only 11.5 hours earlier, at 0500 hours on March 22, 2 hours before the turbines were started (figs. 9 and 31). Even though this water was in the river during the entire heating part of the day (0500—1630 hours), the short traveltime and relatively large depth (1.7 m) dur- ing transit prevented much warming from surface ex- change. The water-temperature increase during transit was 1.6°C, and surface exchange accounted for only 1.0°C of this change. The 06°C occurred due to mixing with the warm water which was both upstream and downstream of the slug. Flushing action decreased the water temperature at Highway 141 by 19°C in 4.5 hours. The water particles that traversed the reach most rapidly (9.75 hours) passed the Highway 141 Bridge at 1730 hours (DT in figs. 9, 31, and 35) on March 22. This water entered the system at 0745 hours, 45 minutes after the turbines were started, with a temperature of 7.4°C. The warming from CT to DT in figure 35 is the result of the arrival of the warm epilimnion water withdrawn by the turbines at high flow. The epilimnion water warmed only 1.2°C during transit, because of the shorter traveltime and higher initial temperature, com- pared to a warming of 1.6°C for the hypolimnion water represented by CT. Some warming occurred even at night, indicating the natural river temperature is higher than the release temperature, even at high flow. Consider the water passing Highway 141 at 0600 hours on March 23 (labeled ET in figs. 9, 31, and 35). This water entered the system 38 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA at 1845 hours on March 22 with a temperature of 7 .3°C (fig. 31). Even though it traversed the system during the cool part of the night, it warmed by 03°C during the transit. The warming of the river between points ET and FT in figure 35 at Highway 141 is caused by natural heating due to surface exchange and longer transit times due to low flow. The water particles reprsented by point FT had the longest traveltime of any water to traverse the system between the two flow pulses (15.75 hours). This water entered the system, during the high flow, on March 22 at 2130 hours (figs. 9 and 31) with nearly the same temperature as the water represented by ET, but it absorbed the morning solar radiation. The drop in temperature from point FT to GT in figure 35 is partly due to flushng action, but it is mainly due to the reduction in the upstream temperature which oc- curred during the low flow between power pulses. The flushing action on March 23 had little effect on the tem- perature because the river was already cold, whereas, it had been warm on March 22 when the flushing action began. The traveltime of the water labeled GT was 14.0 hours, indicating it entered the system at midnight with a temperature of 7.0°C, figure 31. The water particles with shortest traveltime on March 23 (labeled HT) rode the rising stage like the particles labeled DT on March 22. This water entered the system at 0730 hours with a temperature of 8.0°C. Finally, the water leaving the system at 1800 hours on March 23 (IT) traversed the reach at approximately a high steady flow of 225 m3/s. This water had a travel- time of 8.9 hours, a depth of 2.75 m, and temperature rise of 1.2°C. Most of the temperature rise was due to absorbed solar radiation. The effects of hydropulsation during the October verification can be illustrated in the same manner. Con- sider the temperature variations observed to occur at Highway 141 between 1900 hours on October 23 and 2400 hours on October 24. The flow release schedule and meteorologic conditions were similar during most other days. As before, the times for which the temperatures are discussed are labeled in figures 15, 27, and 30. Con- sider first the water particles that passed the Highway 141 Bridge at 1900 hours on October 23. This water, labeled AT, passed the Highway 20 Bridge at midnight with a relatively high temperature of 11.6°C. It traversed the reach during relatively low and receding flow, figure 15, and was in the river during all of the daylight hours. The water temperature increased by about 18°C during its transit. Radiant exchange should have increased its temperature by 1.72°C, so evaporative cooling and other exchange processes near- ly balanced one another. Next consider the minimum water temperature on Oc- tober 23, which occurs at Highway 141 at about mid- night, point ET. This water that left the dam just before the turbines were started was from the hypolimnion and had a low temperature of about 98°C. Furthermore, it was flushed through the system very quickly, and it entered the stream about sunset so surface exchange was also minimal. The model indicates that surface ex- change accounted for only 0.1°C of the 1°C temperature rise actually measured. The remaining 0.9°C temperature rise was the result of dispersive mixing with the warm water ahead and behind it in the river. In summary, three factors combined to give the local temperature minimum at point ET in figure 30. These were flushing, low initial temperature, and small surface exchange because of the nighttime passage. The water passing Highway 141 at 0200 hours on Oc- tober 24 is labeled CT in the figures. The traveltime of this water was only a little longer than that represented by BT (8.9 versus 8.0 hours), but it entered the system after the turbines were started (fig. 15) and, therefore, at a higher temperature (fig. 27). The increase in tem- perature from BT to CT in figure 30 represents the arrival of the warmer epilimnion water being withdrawn at high flow. The net surface exchange for this water was less than 0.1°C, and in fact the total temperature change of the water as it traversed the reach was less than 01°C. The minimum water temperature at Highway 141 on October 24 occurred at 0715 hours and is labeled DT in figures 15, 27, and 30. This water entered the reach at 2000 hours on March 23 between the high flow pulses (fig. 15) with a low temperature of 114°C (fig. 27). Because of nighttime passage and relatively rapid tran- sit speed, little surface exchange occurred. Most of the 0.5°C temperature rise of the water was the result of dispersive mixing with the warm water ahead of and im- mediately behind it in the river. Meteorologic conditions on October 24 were similar to those of the other days, however, the flow release was started about 3 hours earlier. This change in the flow schedule had a significant effect on the temperature of the river. Consider the maximum temperature at High- way 141 on October 24, point ET. The water represent- ed by this point was released from Buford Dam about 2100 hours on October 23 (fig. 15) passing Highway 20 about midnight (fig. 27). This water started out warm (fig. 27), slowly traversed the reach under low-flow con- ditions (fig 15), and was in the river during the entire warm part of the day. As a result it absorbed enough radiation to warm it by 210°C while evaporative cooling lowered its temperature by 03°C. It is interesting that the temperature minimum which occurs at Highway 20 between points AT and BT as well as the temperature maximum which occurs between DT and ET have been DISCUSSION OF RESULTS 39 completely dissipated by the time the water reaches the Highway 141 Bridge. By tracing the route of a few individual water parti- cles through the river reach, the effects of hydropul- sation on the thermal regime of the river have been illustrated. These effects are complicated, however, in that they depend on several factors such as the timing of the release in relation to the time of day and of earlier releases. Rapid temperatures rises of as much as 3°C are common at the upper end of the reach. At the lower end of the reach, these effects are still significant although they are somewhat damped. Rapid tempera- ture decreases caused by flushing are more likely at the lower end of the reach. By studying the travel times and flow release schedules, it is possible to predict the origin of the river water (epilimnetic or hypolimnetic) for any particular time. This could be important if water—quality parameters vary significantly between hypolimnetic and epilimnetic waters. One way to view the effect of the reservoir on the river temperature is to consider an excess temperature, either positive or negative, as the difference between the observed water temperature and the temperature which would have occurred had the reservoir not been present. The question can then be asked: How fast does this excess temperature decay in the downstream direc- tion? To illustrate the rate of decay of a temperature per- turbation downstream of the reservoir, the thermal model was rerun with all conditions identical to those in figures 30 and 35 except that the upstream temperature was increased by 1°C. The difference between the two predicted temperatures is then a direct measure of the percentage of the excess temperature (or effect of the reservoir) which has been dissipated within the reach. Figure 40 is a plot of the excess temperature remaining at Highway 141 for the March run. To facilitate analysis the letters AT through IT shown in the previous figures are also included. Under steady low flow conditions, before point BT, at least 65 percent of any thermal ef- fect caused by Buford Dam should be dissipated by the time the water reaches the Highway 141 Bridge. At a steady flow of about 100 m3/s, points DT to ET, only a little more than 10 percent of the perturbation is dissipated, and at a steady flow of about 200 m3/s, points HT to IT, less than 10 percent of any excess heat would be dissipated by the time the water reaches Highway 141. The percentage of the excess heat to be dissipated within the reach is very dependent on the traveltime. The excess temperatures at the Highway 141 Bridge during the October run are illustrated in figure 41 and again specific points are labeled. In general it appears that a smaller percentage of the excess heat is dissipated during the October run. At high flow, points BT to CT, 1-0 | I I r I I I I I I I I I I I I I 0 _ 0 U) 2 E ‘3 0.8 — it In _ U 3 III _ E LU I 5 o _ LLl _ I- D 0.6 g E n: . 3 “J ‘ I- I n: :> u.I I— n. < 0.4 — u. I 0 LL] 0. Lu 2 _ 0 “J < l- I- a a u] 0.2 - 0 0 n: X Lu w n. O l I I I I I I I I I I I I I I I I I I 100 21 22 23 DAY IN MARCH 1976 FIGURE 40.—Excess temperature at the Highway 141 Bridge during the March run due to an excess temperature of 1° at Buford. 1-0 IIIIIIIII IIIIIIIIIIIIIIIIIII Illo BT _ DT _ 2 CT 2 0.8— ——20 gm ET g 33 »— — '1 :3 AT 20 III-”0.6~— —4OI-'C|u I“(a wI— mw 0-< 2‘” — _ ”-E .932 °$ 00.4— *SOuJ- gm (90 mo __ _ :1 2.2. E w 0.2— —80‘J I: __ _ u: n. oIIIIIIIIIIIIIIIIIIIIIIIIII|IlllIIIl100 20 21 22 23 24 DAY IN OCTOBER 1975 FIGURE 41.——Excess temperature at the Highway 141 Bridge during the October run due to an excess temperature of 1° at Buford. only about 5 percent is dissipated, and at low flow, AT, only slightly over 30 percent is dissipated. There are a number of reasons for the apparent decrease in the rate of return to natural temperature for the October run. First, the traveltime at low flow in October, point AT, is only 19 hours compared to 23.2 hours in March. The frequent power releases in October never allowed the river to reach steady low flow. The release schedule was also different between the two runs. In October the low flow water, AT, spent most of its time in the river at night when meteorologic conditions are not optimal for 40 MODELING FLOW, MASS, AND HEAT TRANSPORT, CHATTAHOOCHEE RIVER, GEORGIA rapid decay. At high flow the March and October results are in bet- ter agreement. The minimum traveltimes in October and March were 8.8 and 7.7 hours, respectively. In March the fast moving water, DT, ET, traversed the reach during the daylight hours, while in October the fast moving water, points BT, CT, traversed the reach at night. It thus appears that the results of the two runs are fairly consistent. At high flow as little as 5 to 10 percent of the excess temperature is dissipated before reaching Highway 141. Unless low flow is maintained for more than 18 hours, seldom would more than 40 percent of the excess temperature be dissipated before reaching Highway 141. SUMMARY AND CONCLUSIONS A coupled flow-temperature model has been developed and verified with data collected on a 27.9-km reach of the Chattahoochee River between Buford and Norcross, Ga. A major contribution of this study has been to apply existing solution techniques, with minor improvements, to a very comprehensive data set in order to verify the existing technology of modeling flow and transport under highly unsteady flow conditions. The modeling analysis has identified transport phenomena unique to unsteady flow. A linear, implicit finite-difference flow model was coupled with implicit finite-difference transport and temperature models. Both the conservative and non— conservative forms of the transport equation were solved, and the differences in the predicted concentra- tions of dye were found to be insignificant. The temperature model, therefore, was based on the simpler nonconservative form of the transport equation. Two extensive data sets were available for calibration and verification of the models under conditions of ex- tremely unsteady flow. Continuous flow, temperature, and meteorologic data were available for the periods of October 20—26, 1975, and March 21—24, 1976. In addi- tion, rhodamine-WT dye was injected to the flow at the upstream end of the reach at a constant rate during the March run, and frequent samples were obtained near the center and at the downstream end of the reach. The dye concentrations were used to verify the transport models. The flow model was calibrated using the depth profile obtained at steady low flow and dynamic stage data col- lected in March 1976. A comparison of the modeled and observed stages at four points in the reach indicated that timing errors were generally less than 15 minutes, and errors in absolute stage were generally less than 0.15 m. Agreement of the modeled and observed dis- charges for flows less than 100 m3/s was excellent, the difference being less than 5 percent. At higher flows the model generally predicted a peak discharge which was about 20 percent less than the observed value. The flow model was verified by use of the data col- lected during October 1975. The modeled values of the stage were generally within 0.15 m of the observed values and timing errors were generally less than 30 minutes. The verification results were considered to be very good. Numerical solution of either the conservative or non- conservative form of the transport equation did an ex- cellent job of simulating the observed concentrations of dye in the river under highly unsteady flow conditions. Timing errors generally appeared to be less than 15 minutes. The agreement between the observed and modeled dye concentrations was considered to be an ex- cellent verification of the transport models. Both the October and March runs were considered as verifications of the temperature model because no calibration was involved. Results were considered to be very good. Temperature change as large as 58°C oc- curred as the water traversed the reach. The model was able to predict the downstream temperature with a RMS error of 032°C in October and 020°C in March. Hydropulsation has a significant effect on the water temperature below Buford Dam. At low flow the release temperature is low and constant. At the beginning of a power release, the water temperature increases by as much as 3°C almost instantaneously because the tur- bines withdraw warm epilimnetic water at high flow. Any water—quality property which differs markedly be- tween the epilimnetic and metilimnetic waters of the reservoir is likely to experience similar sudden changes in the river. The flushing action of the power release often causes a rapid temperature reduction to occur prior to the arrival of the warm epilimnetic water. The effects of hydropulsation are somewhat damped at the lower end of the reach. The thermal effects of hydropul- sation are complicated because they are dependent on the timing of the releases with respect to both the time of day and past releases. At least during the March run, the temperature of the water released from the dam was lower than the natural river temperature, even at high flow. Unless the river is held at steady low flow for longer than 18 hours, at least 60 percent of the thermal effect caused by Buford Dam will remain at the Georgia Highway 141 Bridge. By use of both the Lagrangian and Eulerian forms of the transport equation, it was usually possible to predict the origin of the water in the river, whether from the epilimnion or hypolimnion. This ability could be very helpful in predicting water quality, if the parameter of interest differs significantly between the epilimnetic and hypolimnetic waters of the reservoir. REFERENCES 41 REFERENCES Amein, M. M., and Fang, C. S., 1970, Implicit flood routing in natural channels: American Society of Civil Engineers Proceedings, Journal of the Hydraulics Division, v. 96, no. HYlZ, p. 2481—2500. Anderson, E. R., Anderson, L. J ., and Marciano, J. J ., 1950, A review of evaporation theory and development of instrumentation, Lake Mead water loss investigation: Navy Electronics Laboratory, In- terim Report 159, February 1. Anderson, L. F., 1954, Instrumentation for mass-transfer and energy budget studies, in Water-loss investigation: Lake Hefner studies, technical report: U.S. Geological Survey Professional Paper 269, p. 35—45. Bendat, Julius S., and Piersol, Allan G., 1966, Measurement and analysis of random data: John Wiley and Sons, Inc., New York, 387 p. Blodgett, J. C., 1971, Water temperatures of California streams, Sacramento Basin Subregion: U.S. Geological Survey open-file report, 29 p. Braslavskii, A. P., and Vikulina, Z. A., 1963, Evaporation norms from water reservoirs: Translated from Russian by Israel Program for Scientific Translations, Jerusalem, p 85. Brown, G. W., 1969, Predicting temperatures of small streams: Water Resources Research, v. 5, no. 1, p. 68—75. Brutsaert, W., and Yeh, Gour-Tsyh, 1970, Implication of a type of em- pirical evaporation formula for lakes and pans: Water Resources Research, v. 6, no. 4, p. 1202—1208. Carslaw, H. S., and Jaeger, J. G., 1959, Conduction of heat in solids [2d L ed.]: Oxford University Press, New York, p. 68—75. Fischer, Hugo B., 1973, Longitudinal dispersion and turbulent mixing in open channel flow: Annual Review of Fluid Mechanics, p. 57—98. Fread, D. L., 1974, Numerical properties of implicit four-point finite difference equations of unsteady flow: U.S. Department of Com- merce, National Oceanic and Atmospheric Administration, March, Technical Memorandum NWS HYDRO—18, 38 p. Garrison, J. M., Granju, J. P., and Price, J. T., 1969, Unsteady flow simulation in rivers and reservoirs: American Society of Civil Engineers Proceedings, Journal of the Hydraulics Division, v. 95, no. HY5, p. 1559—1576. Jobson, Harvey E., 1973, The dissipation of excess heat from water systems: American Society of Civil Engineers Proceedings, Jour- nal of the Power Division, v. 99, no. POI, p. 89—103. 1976, Thermal modeling and its relation to canal evaporation: National Conference on Complete Water Reuse Proceedings, 3d, American Institute of Chemical Engineers, Cincinnati, p. 370—379. ___1977, Thermal model for evaporation from open channels: Con- gress of the International Association for Hydraulic Research Proceedings, 17th, Baden Baden, Germany, V. 2, p. 95—102. Jobson,Harvey E., and Sturrock, Alex M., Jr., 1979, Comprehensive monitoring of meteorology, hydraulics and thermal regime of the San Diego aqueduct, California: U.S. Geological Survey Profes- sional Paper 1137, 00 p. Lai, C., 1967, Computation of transient flows in rivers and estuaries by the multiple reach method of characteristics, in Geological Survey research 1967: U.S. Geological Survey Professional Paper 575-D, p. D273—D280. Messenger, Harry, 1963, Dissipation of heat from a thermally loaded system, in Geological Survey research 1963: U.S. Geological Survey Professional Paper 475—C, p. C175—C178. Moore, A. M., 1969, Correlation and analysis of water-temperature data for Oregon streams: U.S. Geological Survey Water-Supply Paper 1819—K, p. 8—14. Pluhowski, E. J., 1970, Urbanization and its effects on the temperature of the streams on Long Island, New York: U.S. Geological Survey Professional Paper 627 —D, p 110. Price, H. S., Cavendish, J. G., and Varga, R. S., 1968, Numerical methods of higher-order accuracy for diffusion-convection equa- tions: Society of Petroleum Engineers Journal, American In— stitute of Mechanical Engineers, v. 8, September, p. 298—303. Rawson, J., 1970, Reconnaissance of water temperature of selected streams in southeastern Texas: U.S. Geological Survey in cooperation with the Texas Water Development Board, p. 2—4. Roache, Patrick J ., 1972, Computational fluid dynamics: Albuquerque, Hermosa Publishers, 434 p. Ryan, P. J ., and Stolzenbach, K. D., 1972, Chapter I of engineering aspects of heat disposal from power generation: Environmental Heat Transfer, MIT Summer Session, June 26—30, Ralph M. Par- sons Laboratory for Water Resources and Hydrodynamics, Cam- bridge, Mass., 75 p. Sayre, W. W., and Chang, F. M., 1968, A laboratory investigation of open channel dispersion for dissolved suspended, and floating dispersants: U.S. Geological Survey Professional Paper 433—E, 71 p. Stone, H. L., and Brian, P. L. T., 1963, Numerical solution of convec- tive transport problems: American Institute of Chemical Engineers Journal, v. 9, no. 5, p. 681—688. Tennessee Valley Authority, 1972, Heat and mass transfer between a water surface and the atmosphere: Water Resources Research Laboratory Report No. 14, Norris, Tenn., April, p. 4.20. Von Rosenberg, D. W., 1969, Methods for the numerical solution of partial differential equations: New York, Elsevier, 128 p. Wylie, E. Benjamin, 1970, Unsteady free-surface flow computations: American Society of Civil Engineers Proceedings, Journal of the Hydraulics Division, v. 96, no. HY11, p. 2241—2251. Yevjevich, V., and Barnes, H. H., 1970, Flood routing through storm drains, Part 1, Solution of problems of unsteady free surface flow in storm drains: Fort Collins, Colorado State University Hydrology Paper No. 43, 108 p. GPO 689-143 Comprehensive Monitoring of Meteorology, Hydraulics, and Thermal Regime of the San Diego Aqueduct, California By HARVEY E. ‘IOBSON and ALEX M. STURROCK, ‘]R. GEOLOGICAL SURVEY PROFESSIONAL PAPER 1137 UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 11179 UNITED STATES DEPARTMENT OF THE INTERIOR CECIL D. ANDRUS, Secretary GEOLOGICAL SURVEY H. William Menard, Director Library of Congress Cataloging in Publication Data Jobson, Harvey E Comprehensive monitoring of meteorology, hydraulics, and thermal regime of the San Diego Aqueduct, California. (Geological Survey Professional Paper; 1137) Includes bibliographical references. Supt. of Docs. No.: I l9.16:1137 1. San Diego Aqueduct, Calif. 2. Meteorology—California, Southern- Observations. 3. Hydraulic me asurements—California, Southern. 1. Sturrock, Alex M.,joint author. [1. Title. III. Series: United States. Geological Survey. Professional Paper; 1137. TC764.J6 628.1’5 79-607792 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 Stock number 024-001-03238-4 CONTENTS Page Conversion table __________________________________________ IV Procedure __________________________________________________ 10 Abstract __________________________________________________ 1 Data collection __________________________________________ 10 Introduction ________________________________________________ 1 Data processing ________________________________________ 10 Acknowledgments __________________________________________ 2 Calibration ____________________________________________ 11 Site description ____________________________________________ 2 Results ____________________________________________________ 13 Instrumentation ____________________________________________ 3 General ________________________________________________ 13 Recording system ______________________________________ 3 Wind __________________________________________________ 22 Wind __________________________________________________ 6 Radiation ______________________________________________ 24 Radiation ______________________________________________ 7 Air temperature ________________________________________ 26 Temperature __________________________________________ 8 Vapor pressure ________________________________________ 26 Stage __________________________________________________ 8 Water temperature ____________________________________ 27 Flow rate ______________________________________________ 9 Discharge ______________________________________________ 28 Other data ____________________________________________ 10 Summary and conclusions ________________________________ ___ 29 References cited ____________________________________________ 29 ILLUSTRATIONS Page FIGURE 1. Map showing the San Diego Aqueduct and data-collection points __________________________________________________ 1 2. Diagram showing typical cross section of the canal ________________________________________________________________ 2 3. Graph showing variation of bank height with distance along the canal ____________________________________________ 2 4. Map showing the San Diego Aqueduct, the diversion points, and locations of siphons and stage recorders ____________ 2 5—20. Photographs showing: 5. Diversion of water from the Casa Loma Canal at the entrance to the San Diego Aqueduct __________________ 3 6. Skinner end of San Diego Aqueduct ______________________________________________________________________ 3 7. Recording system at downstream end of the canal ________________________________________________________ 3 8. Wind exposure at Cottonwood meteorologic station ________________________________________________________ 6 9. Wind exposure at Skinner ______________________________________________________________________________ 6 10. Wind exposure at Skinner ______________________________________________________________________________ 6 11. Psychrometer and supplementary anemometer at Cottonwood ______________________________________________ 6 12. Closeup of supplementary wind sensor and psychrometer at Cottonwood ____________________________________ 7 13. Closup of supplementary wind sensor and psychrometer at Skinner ________________________________________ 7 14. Radiation instrumentation at Cottonwood ________________________________________________________________ 7 15. Closeup of psychrometer ________________________________________________________________________________ 8 16. Recorder and stilling well at the Cottonwood gage ________________________________________________________ 8 17. Canal at the Cottonwood gage __________________________________________________________________________ 9 18. Simpson gage __________________________________________________________________________________________ 9 19. Newport gage __________________________________________________________________________________________ 9 20. So. End gage and diversion ______________________________________________________________________________ 9 21—23. Graphs showing: 21. Monthly mean windspeed ______________________________________________________________________________ 22 22. Frequency of occurrence of daily windspeeds ______________________________________________________________ 22 23. Mean annual diurnal variation in windspeed, averaged by 10-minute time periods for the period July 24, 1973, to July 23, 1974 ______________________________________________________________________________________ 23 24. Diagram showing frequency of occurrence, in percentage of time, of winds from various directions ____________________ 23 25—31. Graphs showing: 25. Daily solar radiation ____________________________________________________________________________________ 24 26. Daily atmospheric radiation ______________________________________________________________________________ 25 27. Typical diurnal variation in atmospheric radiation as measured by the pyrgeometer and by subtracting the measured instantaneous solar radiation from the all—wave radiation measured by the flat-plate radiometer ___25 28. Daily average air temperature __________________________________________________________________________________ 26 29. Daily average vapor pressure ________________________________________________________________________________ 27 30. Daily mean water temperature, averaged over depth __________________________________________________________ 28 31. Mean diurnal variation in water temperature __________________________________________________________________ 28 III IV CONTENTS TABLES Page TABLE 1. Locations of structures on San Diego Aqueduct (information furnished by the Metropolitan Water District of Southern California) __________________________________________________________________________________________ 1 2. Calibration factors (for use in equation 3) of instrumentation used at the San Diego Aqueduct ______________________ 11 3. Profiles of centerline depth for the San Diego Aqueduct under nearly steady flow conditions ________________________ 12 4. Daily averaged values of meteorologic and temperature data for San Diego Aqueduct (July 25, 1973—Ju1y 23, 1974) W 13 5, Daily averaged values of hydraulic and rainfall data for San Diego Aqueduct (July 25, 1973—Ju1y 23, 1974) __________ 17 6. Comparisons of vapor pressures estimated by four data-transfer methods __________________________________________ 27 CONVERSION TABLE [Factors for converting metric (SI) units to inch-pound units are shown to four significant figures] Multiply metric unit By To obtain inch-pound unit meter (In) 3.281 foot (ft) millimeter (mm) 0.03937 inch (in) meter per second (m/s) 35.31 foot per second (ft/s) cubic meter per second (ma/s) 0.02832 cubic foot per second (ft3/s) pascal (Pa) 10.00 millibar (mb) watt per square meter (W/mz) 2065 calorie per square centimeter per day [(Cal/cm2)/d] COMPREHENSIVE MONITORING OF METEOROLOGY, HYDRAULICS, AND THERMAL REGIME OF THE SAN DIEGO AQUEDUCT, CALIFORNIA By HARVEY E. JOBSON and ALEX M. STURROCK, jR. ABSTRACT Water temperature, as well as meteorologic and hydraulic vari- ables which influence the energy budget of the San Diego Aqueduct in southern California, were continuously monitored for a 1-year period beginning July 24, 1973. The incoming solar and atmospheric radiation, windspeed and direction, water temperature, and the wet- and dry-bulb air temperatures were recorded at 10-minute intervals at each end of the 26- kilometer concrete-lined canal, while the flow rates and stages were determined at hourly intervals for five loca- tions. While only daily averaged values are presented in this report, a magnetic tape containing all the data can be obtained from the Automatic Data Processing Unit, US Geological Survey, Water Re- sources Division, Reston, VA 22092. This report presents all infor- mation necessary for the use and interpretation of these data. Windspeeds were typically low during the early morning hours and at maximum during the late afternoon; however, they were var- iable spatially. On the other hand, incoming radiation and absolute vapor pressure appeared to vary little from point to point. At a point where only the air temperature is known, the most accurate method to estimate vapor pressure is to compute it from the wet- and dry- bulb temperatures obtained at a remote site. INTRODUCTION The ability to predict the effects of man-induced ac- tivity on various physical and biological water-quality factors is becoming increasingly important. Water temperature, while being an important water-quality characteristic in itself, affects nearly every physical property of water and influences the rate of nearly all chemical and biological reactions. A major obstacle to accurate temperature prediction in open channels is the estimation of the heat exchange due to evapora- tion. In 1973 a comprehensive study of evaporation from open channels was initiated. The San Diego Aqueduct, owned and operated by the Metropolitan Water Dis- trict of Southern California, was selected as the study site. The San Diego Aqueduct is a concrete-lined, open channel originating near Hemet, Calif. (about 120 km southwest of Los Angeles), and flowing generally south for about 26 km. The canal has a 3.66-m bottom width and side slopes of 1.5 to 1m. At full capacity it will deliver about 28 m3/s, but it generally flows near half capacity. Since it carries water for municipal purposes, the flow rate is steady for long periods of time. Only three diversion points exist. Two of these are seldom used, and the third is insignificant in size. Water temperature, as well as meteorologic and hy- draulic variables which influenced the energy budget of the canal, were continuously recorded for a 1-year period beginning July 24, 1973. The incoming solar and incoming atmospheric radiation, windspeed and direction, water temperature, and wet- and dry-bulb temperatures were recorded at 10-minute intervals for each end of the canal, and the discharge and stage were recorded at hourly intervals for five locations. The purpose of this report is three fold: (a) to present a description of the site, instrumentation, and proce- dures used in the study of the San Diego Aqueduct, (b) to present daily averaged values of the data obtained and provide the information necessary for the use of the complete data set, and (c) to present a general analysis of the temporal and spatial variability of the recorded data. All data obtained during the study are available on magnetic tape and can be obtained from the Automatic Data Processing Unit, US. Geological Survey, Water Resources Division, Reston, VA 22092. The tape con- tains 10-minute values of all meteorologic and temper- ature data, hourly values of hydraulic data, and daily values of rainfall, pan evaporation, and supplementary windspeed measurements. The spatial and temporal variations in windspeed and direction are analyzed in this report. The fre- quency of winds of various speeds and directions is also presented for each end of the canal. In addition to a presentation of the spatial and temporal variations in incoming atmospheric and solar radiation, the meas- urements are compared to computed clear-sky values, and the dependence of the measured diurnal variation in atmospheric radiation on sensor type is illustrated. Because the wet-bulb temperature is such a difficult 2 METEOROLOGY, HYDRAULICS, ANI) THERMAL REGIME, SAN DIEGO AQUEDUC'I‘, CALIFORNIA parameter to measure, an analysis is presented which determines the most accurate method of predicting the vapor pressure using wet— and dry-bulb temperatures from a remote site. The variations in resistance to flow which occurred throughout the year are also presented. ACKNOWLEDGMENTS The authors acknowledge the excellent cooperation extended by the Metropolitan Water District of South- ern California. They granted the US. Geological Sur- vey permission to install instrumentation on their premises, furnished copies of their operating records, and provided valuable assistance in the routine opera- tion of the data-collection program. Especially helpful were Messrs. Paul Singer and Charles Voyles who were instrumental in granting the Survey permission to use the canal and to Mr. Kenneth Gandee who pro- vided for the routine operations of the data collection. SITE DESCRIPTION Evaporation was the process of major concern in the study and thus influenced the selection of a study site. Since the energy exchange due to evaporation is often small in comparison to radiation exchange and energy convected into and out of a stream-flow system, the most desirable site for study would be one where evap- oration would be large and variations in the other ele- ments of the energy budget small. Thackson and Parker (1971) presented monthly averaged meteorologic conditions for 88 locations in the United States. After a thorough study of these conditions, it was decided that the hot, dry climate found in southern California and southern Arizona provided the best general location for the study. Restricting attention to this general region of the country, a canal or other open-channel reach was sought which would have: (a) a traveltime of approximately 12 hours; (b) fairly con- stant and easily defined geometric and hydraulic characteristics; (c) reasonably steady flow rate; ((1) very few tributary inflows of diversions; (e) a maximum depth of about 3 m; and (f) a fairly uniform wind expo- sure along the reach. After considerable reconnais- sance, it was found that the San Diego Aqueduct most nearly met all the requirements. The San Diego Aqueduct carries water from the Casa Loma Canal to Lake Skinner (fig. 1). Water for the San Diego Aqueduct is diverted from the Colorado River, below Parker Dam, and carried by the Colorado River Aqueduct to a point just west of the San Jacinto Mountains (fig. 1, northeast corner). At this point, the water can be diverted to the Casa Loma Canal or directly to the city of San Diego by under- ground pipe lines. Water for the San Diego Aqueduct is diverted from the Casa Loma Canal. The general topography in the region consists of mountain massifs separated by the flat San Jacinto Valley floor. Areas with elevations above 610 m are shaded in figure 1. The elevation of the valley floor is approximately 450 m. The upper 40 percent of the canal traverses the approximate center of the valley, and the lower 60 percent of the canal runs along the extreme eastern edge of the San Jacinto valley; the land generally rises sharply to the east of the canal but is fairly flat to the west. A few short reaches in the lower part of the canal pass through cuts as deep as 18 m with steep side slopes. In general, the hilly areas of the region are used only for grazing, and the valley is grazed or used for the production of hay or other dry- land crops. The National Atlas of the United States (US. Geological Survey, 1970) describes the natural vegetation as Coastal Sagebrush or Chaparral. Vege- tation is of minor importance to the exposure of the water surface to either wind or radiation. Thornthwaite (1931) described the climate as semiarid. The National Atlas (US. Geological Survey, 1970) shows the mean annual precipitation as ranging from 200 to 400 mm, with 30 days per year having more than 0.3 mm. The average annual runoff for the region is low, ranging from 10 to 30 mm; however, up to 130 mm of runoff may be expected in the mountains. Almost no precipitation occurs during the months of May—October. The average solar radiation is 121 W/m2 in January and is 315 W/m2 in July, and the mean temperature ranges from a low of 10°C in January to a high of 24°C in July. The mean dew point temperature is 2°C in January, whereas it is 10°C in July. The San Diego Aqueduct is concrete lined, with a bottom slope of 0.00012. The cross section of the canal is trapezoidal, measuring 3.66 m for bottom width and 1.5 to 1 m for side slopes. The maximum design depth is 3.05 m, and it has a 0.305-m freeboard. Its capacity is about 28 m3/s. The spoil dirt from the canal was piled along the sides so that the typical cross section appears as shown in figure 2. The spoil bank height varies along the length of the canal as shown in figure 3. The three very large bank height values represent places where the canal passes through deep cuts, but for the rest of the canal the bank height is more or less repre- sentative of the height of the spoil bank. The 16 siphons which carry water under roads and drainageways for the San Diego Aqueduct are shown in figure 4. The longest, just south of Holland Road, is 197 m long, and the shortest, under Cottonwood Avenue, is 13 m long. The length, type, and location of all siphons are summarized in table 1. All siphons here have smooth transitions in the channel cross section upstream and downstream. Three points are available for diverting water from INSTRUMENTATION 3 117° [.Akswgvv MOUNTAINS 33°45’ — ©§ San Diego Aqueduct CALIFORNIA Study site EXPLANATION A Gaging station Q Meteorologic station <8) Evaporation pan ———— Underground aqueduct Canal 0 5 l 0 5 MILES 10 KILOMETERS FIGURE 1.—San Diego Aqueduct and data-collection points. Shaded areas are above 610 m. the canal. The first diversion structure, near Simpson Avenue, allows water to be diverted into the under- ground aqueduct shown in figure 1. This structure was in operation during 35 percent of the study and diverted a maximum of 3.02 m3/s. The second diver- sion, known as EM—8, is insignificant in size. The maximum diverted flow was 0.15 m3/s, but it generally diverted only about 0.03 m3/s for 1 or 2 hours during the day. This diversion structure is just below Holland Road and supplies water for a chicken farm. The third diversion structure, the So. End diversion, allows water to be diverted to San Diego without passing through Lake Skinner (fig. 1). The So. End diversion was not used except during the last 71 days (20 per— cent) of the study, when it was in continuous operation. The diversion of water from the Casa Loma Canal at the entrance to the San Diego Aqueduct is shown in figure 5. The Casa Loma flows from east to west just behind the diversion structure. The outlet of the canal into Lake Skinner is shown in figure 6. INSTRUMENTATION RECORDING SYSTEM Three types of recording systems were used for the study. Windspeed, wind direction, solar radiation, at- mospheric radiation, wet- and dry-bulb temperatures, and water temperatures for each end of the canal were 4 METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA Bank height Spoil bank Chain link fence\ Spoil bank BANK HEIGHT, IN METERS | 27 Skinner Cottonwood DISTANCE, IN KILOMETERS FIGURE 3.—Variation of bank height with distance along the canal. recorded digitally at 10-minute intervals. Analog re- corders were used to continuously record the stage at five points along the canal, the discharge of the EM—8 and Simpson road diversions, and after February 28, 1974, the discharge at the lower end of the canal. Analog records were digitized manually at hourly intervals. Rainfall, pan evaporation, pan windspeed, prevailing wind direction, and general weather condi— tions were observed daily. The primary recording system to monitor radiation, wind, and temperatures consisted of Esterline Angus1 D2020 recorders coupled with Pertec magnetic tape recorders. At 10-minute intervals the time, as well as the millivolt values of all 10 parameters, were recorded on magnetic tape. No averaging of the readings was possible, so recorded values were the millivolt readings at the instant they were sampled. A sampling of all 10 channels took only about 5 seconds. At approximately 1—hour intervals, the same information was printed in digital form on a paper tape. The paper tape allowed field monitoring of the system and served as a backup record on a few occasions when the magnetic tape sys- tem failed. At weekly intervals the magnetic and paper tapes were removed and mailed to NSTL Station, Miss., for processing. A total of 12 tapes was sufficient to allow them to be copied, cleaned, and returned to the field for reuse without a shortage occurring. The sys- |The use ofbrand names in this report is for identification purposes only and does not imply endorsement by the US Geological Survey. SIMPSON GAGE A SIMPSON A_VE OLIVEAVE NEWPORT GAGE NEWPORT RD A . EXPLANATION A Stage recorder 9 Siphon O Diversion point RAWSON RD 0 1 2 3 KILOMETERS 0 1 MILES FIGURE 4.—San Diego Aqueduct, the diversion points, and locations of siphons and stage recorders. INSTRUMENTATION 5 TABLE 1.—Locations of structures on San Diego Aqueduct [Information furnished by the Metropolitan Water District of Southern California] Siphon characteristics Distance from canal intake Length Name of structure (m)' (In) Size Cottonwood ,,,,,,,,,,,,,,,,,,,, 297.5 13.0 pipe, 3.96 m ID.2 Cottonwood stage _ 593.7 ”A, ________________ Esplanade __ 1,978.6 75.3 box, 3.81X3.35 m Devonshire 4,453.4 403 box, 3.81 X335 In Florida __ 5,275.8 66.3 box, 3.81X3.35 m Stetson __ 6,982.8 650 box, 3.81x3.35 m Railroad ,, ,,,, _ 8,019.6 1078 box, 3.81X3.35 m Simpson diversion W , 8,793.8 ,_._ ______ .7- Simpson stage ,,,,,, , 9,134.2 ,,,, ,,,,,,,,,,,,,,,, Simpson __________ _ 9,410.9 61.1 box, 3.81X3.35 m Olive ______________ _ 10,1358 142.9 pipe, 3.96 m I.D. Newport stage A”, _ 13,4603 ”W ________________ Newport ________ , 13,6045 55.9 box, 3.81X3.35m Holland Road .1, _ 15,1557 197.5 pipe, 3.96 In 1D. EM—S diversion "W , 15,3532 ____ ________________ South Domanigoni H , 15,9326 133.8 pipe, 3.96 In 1D. Garbani Ranch ,.._ 17,0420 158.5 pipe, 3.96 In 1D. French ,,,,,,,,,, _ 19,5715 36.4 box, 3.81X335 m Rawson Road ______ _ 20,519 8 61.1 box, 3.81 X335 m Bachelor Mountain ,._, . 22,525 7 158.5 pipe, 3.96 in ID. South Bachelor Mountain , 23,852 1 36.4 box, 3.81X3.35 m So. End diversion ,,,,,, 24,978 4 ,_,_ So. End stage ___- I 24,978 4 _ Skinner stage ,_ _ 25,978 1 "W Skinner inlet __________________ 25,9827 ,_,, 'Locations for siphons are determined by location of the siphon intake. 2Inside diameter. FIGURE 5,—Diversion of water from the Casa Loma Canal at the entrance to the San Diego Aqueduct. Photograph taken from the east bank; View is to the north. tem at the Cottonwood (north) end was housed in a plywood shelter (fig. 14). and at the Skinner (south) end it was housed in a concrete building (figs. 6 and 7). The water stage was recorded on vertical-drum graphic recorders. These recorders had an unlimited range in stage because a stylus-reversing device was activated for each 0.3-m change in stage. A distance of 12.7 mm on the chart corresponds to a stage change of 0.305 m, and a distance of 4.6 mm corresponds to a time lapse of 24 hours. The stage records were collected as a part of the Metropolitan Water District’s routine operating data. Charts were changed on a weekly basis, and copies of the charts were furnished to the Survey for analysis. These charts were read manually to determine the stage to the nearest 3 mm each hour. FIGURE 6.—Skinner end of San Diego Aqueduct. Photograph taken from Lake Skinner dam; View is to the west. FIGURE 7 .—Recording system at downstream end of the canal. The head on a Venturi meter at the Simpson road diversion was also determined from records of this type. The recorded head was converted to a discharge by use of a rating curve. The diversion discharge to the chicken farm (EM—8) was recorded on a circular graphic chart which made one revolution per week. The chart was activitated by the head on a Venturi meter and was calibrated to read flow rate directly. The discharge could be easily read to the nearest 0.01 m3/s. After February 28, 1974, the flow into Lake Skinner was recorded on a circular graphic chart which made one revolution per week. The flow rate could be easily read to the nearest 0.3 m3/s. All 6 METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA circular charts were changed weekly; copies were sent to the Survey and digitized on an hourly basis. A sup- plementary windspeed was recorded on a weekly circu- lar chart which contained a pip mark for the passage of each 16 km of wind. These charts were changed weekly, and the mean windspeed for each day was es- timated. WIND Propeller-type anemometers were used as the pri- mary wind-sensing devices. The starting speed of the propeller is about 0.45 m/s, with complete tracking at about 1.40 m/s. The wind direction was measured by the movement of a wiper arm on a potentiometer housed in the main body of the anemometer. The general exposure of the anemometer at Cotton- wood is illustrated in figure 8. The sensor head was located approximately 2 m above the concrete plat- form, approximately 2.4 m above the elevation of the FIGURE 9.—Wind exposure at Skinner. Photograph taken from dam with view generally to the north. asphalt parking lot surrounding the concrete struc— ture, and approximately 3.1 In above the surrounding ground level. The sensor was 3.8 m above the top of the canal. The valley floor is flat in this area, and no major obstructions occur within 1 km. The wind exposure at the Skinner station was quite different from the Cottonwood station. The terrain rises steeply to the north as shown in figure 9, whereas it falls off rather rapidly to the south and west (fig. 10). The exposure of the primary anemometer at Skinner was adequate for south or west winds, poor for north winds, and only fair for east winds. The effective height probably varies with wind direction. The sensor was mounted 0.86 In above the roof of the building shown in figure 6. The railing (which was added after the study began) partly hides the sensor in the figure. The build- FIGURE 10.—Wind exposure at Skinner. Photograph taken with View to the south. FIGURE 11.——Psychrometer and supplementary anemometer at Cot- tonwood. INSTRUMENTATION 7 ing roof is 3.05 In above the level of the parking lot. The parking lot is generally lower than the surrounding terrain but about 0.3 m above the top of the canal. Supplementary Windspeed measurements were ob- tained over the water. At Cottonwood the supplemen- tary anemometer was mounted on a swing-out arm which projected 3.0 m from the canal edge. The anemometer was 3.3 In above the top of the canal lin- ing. The general location of the anemometer is illus- trated in figure 11. Windspeed measurements were ob- tained by use of vertical axis cup anemometers equipped with totalizing dials. An electrical contact was closed upon the passage of each 16 km of wind, and the contact closure times were recorded on a circular chart. A closeup of the anemometer used at Cotton- wood is shown in figure 12. At Skinner the supplemen- tary anemometer was mounted on a short arm project- ing to the west of the bridge over the canal shown in figures 6 and 13. The anemometer was 2.0 m above the top of the canal and was located near the center of the bridge. Average daily windspeeds were obtained at the evaporation pans by reading a totalizing dial on the anemometers at approximately 0900 hours daily. The anemometers were mounted on the northwest corner of the evaporation—pan platform 0.15 to 0.20 m above the lip of the evaporation pan. The location of the evapora- tion pans is shown in figure 1. RADIATION The total incoming solar radiation was determined by use of Eppley precision spectral pyranometers. These instruments are sensitive to radiation with a wavelength between 0.3 and 3 am (micrometers). At Cottonwood the sensor was mounted on top of the plywood shelter housing the recording system (the in- strument to the right in fig. 14). At Skinner it was mounted 1.14 In above the roof of the building (fig. 6), Incoming atmospheric (longwave) radiation was de- termined by use of two types of instruments. From July 24, 1973, until November 27, 1973, atmospheric radia- tion was determined by use of Eppley (longwave) pyrgeometers which are sensitive to radiation in the range of 4 to 50 pm. These instruments were mounted beside the pyranometers (figs. 6 and 14). After November 27, 1974, a flat-plate radiometer of the Gier-Dunkle type was used to determine the incoming atmospheric component at Cottonwood. The flat-plate radiometer is sensitive to radiation of all wavelengths. Accordingly, the solar component, determined by use of the pyranometer, was subtracted from the total incom- if» FIGURE 13.—Closeup of supplementary wind sensor and psychrome- ter at Skinner, View to the south. FIGURE 12.—Closeup of supplementary wind sensor and psychrome- ter at Cottonwood, view to the southeast. ‘ FIGURE 14.—Radiation instrumentation at Cottonwood, view to the south. 8 METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA ing radiation to give the atmospheric component. The flat-plate radiometer was also mounted on top of the plywood shelter. TEMPERATURE All temperatures were determined by use of platinum resistance temperature devices (RTD’S). A 10—volt regulated d-c power supply furnished the re- quired voltage to all the sensors. Capsule-style sensors with a sheath length of 63 mm and diameter of 3.2 mm were used for the dry— and wet—bulb temperatures. For measurement of water temperatures, a sensor with a sheath length of 127 mm and diameter of 6.3 mm was used. Measurements of wet- and dry-bulb temperatures were made with a ventilated psychrometer as shown in figures 12 and 13. To insure proper wet-bulb depres- sion, a vane axial fan with a constant output of 4.5 m/s was used to draw air through the tube. The total length of the tube was 203 mm, and the entire assembly was shielded from radiation by a curved aluminum sheet 380 mm long and curved in a semicircle with a 190-mm radius (fig. 15). The wet- and dry-bulb temperatures were measured by probes projecting across the tube mounted 139 mm and 51 mm behind the tube intake, respectively. The wet-bulb probe was covered by a wick which was continuously wetted by distilled water. The wick was a common white cotton shoelace which had been boiled in detergent to remove sizing and rinsed in distilled water. Water was supplied to the wick by gravity flow from a l-gallon plastic bottle through a capillary tube. The rate of flow in the tube was con- trolled by a needle valve at the base of the bottle. At Cottonwood the psychrometer was mounted under the swing-out arm as shown in figure 12. It was 2.7 m above the top of the canal. The general location of FIGURE 15.—Closeup of the psychrometer. the psychrometer is illustrated in figure 11. At Skinner the psychrometer was mounted on a short arm which projected to the west of the bridge shown in figures 6 and 13. The psychrometer was 1.4 In above the top of the canal and near the center of the bridge. Water-temperature probes were fastened to a wooden plank in such a manner that they would be 0.15, 0.61, 1.1, and 1.5 m above the bottom ofthe canal. At Cottonwood the plank was mounted to the west side of the canal so that it sloped downward with the canal side. The top of the plank can be seen in Figure 11. The temperature probes were held about 44 mm away from the concrete wall. Water temperatures were always found to be uniform, so on November 29, 1973, the bottom probe at Cottonwood was removed and not used during the remainder of the study. At Skinner a plank was mounted to the bridge so that it was held vertically near the center of the canal. The vertical locations of the probes at Skinner were the same as those used at Cottonwood. STAGE The stage in the canal was recorded at the five loca- tions shown in figure 4. In all cases, a float-stilling-well arrangement was used, and zero stage corresponds to zero depth in the canal. The Cottonwood gage was lo- cated on the right bank 0.593 km downstream of the canal entrance. The outlet of the Cottonwood Avenue siphon was 0.28 km upstream of the gage, and the en- trance to Esplanade Avenue siphon was 1.38 km downstream. A gentle curve to the right begins 0.07 km below the gage. A photograph of the recorder and top of the stilling well is shown in figure 16, and the general canal conditions with a flow of about 2.8 m3/s is FIGURE 16.—Recorder and stilling well at the Cottonwood gage, with View upstream. INSTRUMENTATION 9 shown in figure 17. Current meter measurements, made from the bridge shown in figure 17, were used to determine a stage-discharge relationship. The Cotton- wood stage record was used to determine the discharge during times of unsteady flow. The Simpson gage, located 9.13 km downstream from the canal entrance, is mounted over the center of the canal as shown in figure 18. It is 1.01 km downstream of the railroad siphon and 0.28 km upstream of the Simpson Road siphon. The Newport gage, located on the right bank 13.46 km downstream of the entrance, is shown in figure 19. The Newport gage is 3.18 km downstream of the Olive Avenue siphon and 0.14 km upstream of the entrance to the Newport Road siphon. The So. End gage, shown on the left side in figure 20, is located at the entrance to the So. End diversion (fig. 1). Before the construction of Skinner reservoir, the So. FIGURE 17.—Canal at the Cottonwood gage, with View downstream. FIGURE 18.—Simpson gage, view to the south. End diversion was the south end of the open part of the San Diego Aqueduct. The diversion is to the left in figure 20, and the flow into Lake Skinner continues toward the lower right. The Skinner gage was housed in the concrete building shown in figure 6 and is at the downstream end of the canal 25.98 km downstream of the canal entrance. The use of this gage was discon- tinued on February 20, 1974. The discharge at the time the photographs in figures 17, 18, 19, and 20 were taken was about 2.8 m3/s. FLOW RATE Canal and diversion discharges were furnished by the Metropolitan Water District from their operating records during times of steady flow. The flow rate at Cottonwood was determined by the Metropolitan Water District from an analysis of gate openings which were periodically calibrated by current-meter meas- FIGURE 19.—Newport gage. FIGURE 20.—So. End gage and diversion, View to the west. 10 METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQL'EDUC'I‘, CALIFORNIA urements at the Cottonwood gage. The diversion dis- charge at Simpson road is determined by use of a cali- brated weir. At the EM-8 diversion a Venturi meter was used. After February 28, 1974, the flow rate into Lake Skinner was continuously monitored by use of a sonic flow meter. The flow rate at the So. End diversion was not measured but can be obtained from continuity considerations. OTHER DATA The Metropolitan Water District maintains two evaporation pans at the locations indicated in figure 1. - The daily rainfall, pan evaporation, pan windspeed, general weather observations, and wind direction are recorded at about 0900 each day. On October 9, 1973, the observation station near the So. End diversion was moved to a point near the Lake Skinner outlet. Both locations are shown in figure 1. PROCEDURE DATA COLLECTION All meteorologic instrumentation was designed to operate unattended for time periods of up to 1 week. However, on alternate days, the canal patrolman: (a) wiped dust from the radiometer bulbs; (b) checked the wick on the psychrometer (clean or change if neces- sary); (c) filled bottle with distilled water if necessary; (d) checked the printed output tapes for obvious mal- functions; (e) checked blower motors on the psy- chrometers; and (f) reported any malfunction to the Survey for correction. On a weekly basis, Water Dis- trict personnel changed the magnetic and paper tapes as well as all weekly charts. The junior author visited the site on an approximately monthly basis to perform maintenance, check on the overall operation of the data-collection system, and carry out special meas- urements. When the field magnetic tapes were received at the office, they were copied in compacted form on an in- house tape. At the same time, a complete listing was produced as well as a time plot of the millivolt level for each channel. The computer listing and plots were compared with the output on the paper tape and instructions for further processing generated. At this point the major concern was to be sure the correct data were associated with each data block, the recorded times were correct, and that each day contained exactly 144 entries for each parameter. Corrections to the recorded times were frequently necessary because of short-term power outages, and so forth. Power out- ages were easily detected because the recorder clock would automatically start over at 0000 hours (mid— night). All data were referenced to Pacific daylight time. Time checks written on the paper tape served as a reference for time corrections. The hourly stage and discharge values were read from charts and keypunched. Likewise, the daily values of the supplementary windspeed, pan windspeed, pan evaporation, and daily rainfall were computed from recorded parameters and the values added to the data set. The declination of the sun was determined from a solar ephemeris. DATA PROCESSING The next step in the processing of the data was to convert the millivolt values to engineering units and to delete all questionable values from the set. Converting to engineering units was simply a matter of applying the appropriate calibration factors to the millivolt readings, but detection of bad data in the set required considerable patience and systematic sorting. Judgment as to the authenticity of the data was based primarily upon plots of the data as a function of time. Two types of plots were used: the original plots from the field data tapes where each 10-minute value was plotted as a function of time, and daily averaged plots where the entire year’s data for a single parame- ter at each end of the canal could be displayed on one illustration. The recorder at Cottonwood developed a malfunction early in the study which caused the data to drift slowly off the true value. Comparison of the daily average plots from Cottonwood and Skinner on a channel-by-channel basis enabled the processor to determine the time at which a particular channel started drifting. Considerable difficulty was experienced in keeping the wick on the psychrometer sufficiently wet. Time periods for which the wick was not sufficiently wet were obvious when a plot of the wet- and dry-bulb air temperature was inspected. In a like manner, times during which one of the water temperature probes was out of the water, owing to low stage, were also easily detected from a plot of the four water temperatures. Each channel of data was scanned several times, and data which were questionable were deleted. As a final check, after the data sets had been transferred to en- gineering units and questionable data deleted, a pro- gram was written which searched each channel of data for each station and printed the 20 largest values, 20 smallest values, and 20 largest changes to occur in 10 minutes, along with the date and time of occurrence of each event. The data on these listings were then rechecked for consistency and accuracy. As a final step the data were replotted on the 10—minute-interval basis and the plots scanned a final time. Hydraulic data were processed in a similar fashion PROCEDURE 1 1 except that only hourly values and not 10—minute values were involved. The EM—8 diversion discharge was read directly from the circular chart, and no proc- essing was necessary. An analog chart of the head on the weir at the Simpson Road diversion was furnished as well as the diversion discharge during times of steady flow. An analysis of the recorded head and dis- charge indicated that the flow could be computed from the formula Qs = 1.62Y,,.‘-317 (1) in which Y". is the head on the weir in meters, and Q3 is the Simpson Road diversion discharges in cubic meters per second. Equation 1 was used to determine the Simpson Road diversion discharge during time periods when the flow was changing. For each day during which the flow was constant (209 out of 365), the Manning’s roughness coefficient applicable to the Cottonwood stage was determined. The coefficients ranged from 0.0135 to 0.191 and could not be directly correlated with either stage or dis- charge. The variation appeared to result primarily from algal growth on the canal lining and therefore was more related to time of year than anything else. Variable flow, extending over periods of 1—15 days, occurred on 27 occasions. During these time periods, the upstream flow was computed for each hour using the measured Cottonwood stage and Manning’s equa- tion. The roughness coefficients, obtained before and after the period of unsteady flow, were averaged for use in this computation. Between February 28, 1974, and May 15, 1974, the discharge, as determined from the sonic flow meter at Skinner, was used to determine the flow in the canal. This value was read directly from an analog chart. The So. End diversion was in operation after May 15, 1974. Subsequent to that date the stage at Cottonwood was used to determine the flow into the upstream end of the canal. and the sonic flow meter was used to determine the flow at the downstream end. The flow at the So. End diversion could then be determined indirectly. CALIBRATION Table 2 contains a summary of the calibration fac- tors for instruments used in the study. All calibration equations were of the form 0 = a + bE (2) in which 0 is output in the engineering units shown in table 2, E is recorder output in millivolts, and a and b are constants shown in table 2. TABLE 2.—Calibration factors (for use in equation 3) of instrumenta— tion used at the San Diego Aqueduct Parameter a b Units Applicable dates Cottonwood winds eed ______ 0.11 0.1984 m/s 07—25—73—07—23—74 Skinner windspe ,,,,,,,,,, .08 .1997 m/s 07—25—73—07—23—74 Cottonwood wind direction ___- 1.2 2,114 degree 07—25—73—07—23474 Skinner wind direction ...... <7 2089 degree 07—25—73—07—23—74 Cottonwood solar ____________ 0 110.0 W/m2 07-25—73—07—23—74 Skinner solar ________________ 0 103.1 W/mz 07—25—73—07—23—74 All temperatures 7v- 4.995 “C 07—25—73—07—23-74 Cottonwood atmospheri 0 145.2 W/m2 07—25—73—08—28—73 Cottonwood atmospheric ,,,,,, 0 145.6 W/m2 08—29—73—11—27—73 Cottonwood total incoming __ 0 562.4 W/m2 11~28—73—01—31—74 Cottonwood total incoming W 0 355.0 W/m2 03—07—74—04—24—74 Cottonwood total incoming W 0 474.2 W/m2 06—05—74—07—23—74 Skinner atmospheric ,,,,,,,, 0 145.6 W/m2 07—25—73—08—29—73 Skinner atmospheric ________ 0 145.2 W/m2 08—29—73—03—24—74 Skinner atmospheric ________ 0 145.6 W/m2 04—24—74—05—28—74 After the study the anemometers were recalibrated, and their calibration constants had changed by less than 3 percent. Temperature probes were periodically checked against mercury thermometers and always found to be within :0.1°C. The pyranometers were calibrated before and after use, and their calibration coefficients were found to have remained constant to within :1 percent. Considerable difficulty was experienced with the measurement of atmospheric radiation. Both pyrgeometers were calibrated before installation on July 25, 1973. By the end of August 1973, the meas- ured atmospheric radiation at Cottonwood appeared to be drifting to the high side. To see if the difference between Cottonwood and Skinner readings was real or due to a drift in the calibration of pyrgeometers, on August 29, 1973, the sensors were interchanged. No shift in the daily average values at either end of the canal was observed after the sensors were inter- changed, so it was assumed that the sensor calibrations were still valid. Although the recorder at Cottonwood malfunctioned three times between September 1, 1973, and November 28, 1973, the atmospheric radiation values measured at each end of the canal and the com- puted clear-sky values were all in general agreement. On November 28, 1973, the pyrgeometer at Cotton- wood was removed for a calibration check and replaced by a flat-plate radiometer of the Gier-Dunkle type. During December 1973, the measured atmospheric radiation at Skinner was 30 percent higher than that measured at Cottonwood. Unfortunately, it was impos- sible to determine whether one or both sensors were out of calibration. On January 31, 1974, the flat-plate radiometer failed, and it was replaced on March 7, 1974. This second flat-plate radiometer was operated until April 24, 1974, when it also failed. The first flat plate, which had been repaired and recalibrated, was reinstalled at Cottonwood on June 5, 1974. This radiometer operated continuously until the end of the study on July 23, 1974. Meanwhile, difficulty was also being experienced at 12 the other end of the canal. On March 24, 1974, the output of the pyrgeometer at Skinner started dropping sharply, falling essentially to zero by March 31, 1974. Unfortunately, the instrument could not be calibrated, so only the original calibration factor is available. On April 24, 1974, the pyrgeometer removed from Cottonwood on November 20, 1973, was installed at Skinner. Although it had been recalibrated locally, the output using the new calibration factor appeared to be too large, so it was assumed that the original manufac- turer’s calibration factor was still valid. This instru- ment failed in a manner similar to the first pyrgeome- ter on May 28, 1974. After this date, no atmospheric radiation measurements were available at Skinner. The hydraulic data were furnished by the Metropoli- tan Water District, and no extensive check on their accuracy was made by the Survey. However, the accu- racy of the stage recorders at Cottonwood and Simpson Road were checked by measuring the centerline water depth from the bridge and comparing the results to the recorded stage. The centerline depth and stage always agreed to within :6 mm. On January 15—16, 1974, the discharge was measured and the results compared with the value provided by the Metropolitan Water District. The results agreed to within 4 percent, which is consid- ered to be about the accuracy of the flow measure— ments. Personnel of the California District of the Geo- logical Survey made current-meter measurements which were used to calibrate the sonic flow meter at Skinner. The sonic flow meter was installed during the latter part of February 1974 and calibrated in early March. It was realized that the five measured stages may not be representative of depths at all points along the canal. The word “depth” is used to mean the centerline (maximum) depth, which is also the stage (fig. 2) in all cases. An attempt was made to measure a longitudinal depth profile under several flow conditions to deter- mine how to predict mean local depths from the re- corded stages. These profiles were determined in the following manner. First an automobile odometer was used to place reference marks on the canal fence. No reference mark was placed closer than 160 m to either the intake or outlet of a siphon. Water depth was measured at each of these reference points for seven conditions which included flows ranging from 2.8 to 23 m3/s. It was impossible to measure the centerline depth (stage) directly except at a few points along the canal; however, it was fairly easy to measure the slope dis- tance from the top of the canal lining to the water surface. The stage could then be computed from the known geometry of the canal. Comparisons of the stage, as determined from the slope measurements to ME'I‘EOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA the stage as measured directly, where possible, indi- cated that the slope measurements allowed the stage to be estimated within about :9 mm. The results of these measurements and calculations are summarized in table 3, which gives the centerline depth (stage) and the distance from the canal intake for each of seven runs. The flow variations during any run were very small, and the measured stages and dis- charges at the time of the runs will be presented later. Runs 1 and 7 were conducted on March 6 and April 24, 1974, respectively, when the hydraulic conditions were essentially constant. Unfortunately the flow was slightly unsteady during run 2 made on March 24, TABLE 3.~—Profiles of centerline depth for the San Diego Aqueduct under nearly steady flow conditions [All values are in meters] Distance Stage Distance from run from Run Run Run intake 1 intake 2 6 7 593 2.98 148 1.02 1.81 2.47 1,398 2.86 477 1.17 1.98 2.65 2,444 2.84 811 1.12 1.89 2.58 3,249 2.80 1,145 1.03 1.82 2.48 4,054 2.79 1,477 1.03 1.81 2.45 4,897 2.77 1,811 1.03 1.82 2.46 6,147 2.77 2,294 1.06 1.84 2.46 7,450 2.76 2,614 1.04 1.81 2.44 8,529 2.78 2,935 1.03 1.80 2.43 9,135 2.79 3,254 1.01 1.77 2.40 9,794 2.78 3,574 1.03 1.80 2.42 11,083 2.77 3,893 1.02 1.77 2.40 11,888 2.70 4,214 1.01 1.75 2.36 12,692 2.65 4,835 1.02 1.75 2.36 13,497 2.60 5,080 .98 1.71 2.33 14,143 2.64 5,611 .99 1.73 2.33 15,594 2.62 5,927 1.01 1.74 2.33 16,871 2.70 6,266 .98 1.70 2.30 17,602 2.76 6,580 ,,,,,, 1.75 2.25 18,407 2.75 6,715 .95 1.68 2.25 19,212 2.72 7,223 .96 1.70 2.26 20,011 2.72 7,541 .95 1.69 2.26 20,333 2.71 7,780 .98 1.71 2.27 21,385 2.56 8,355 1.01 1.73 2.23 . , .98 1.71 2.16 1.01 1.74 2.16 .95 1.69 2.17 .91 1.67 2.15 1.06 1.81 2.25 1.06 1.78 2.23 1.03 1.77 2.23 1.02 1.75 2.18 1.01 1.74 2.19 .98 1.68 2.05 .96 1.68 2.13 .92 1.64 2.13 .87 1.58 2.11 .84 1.54 2.11 .92 .. . 1.61 2.17 .88 . . 1.55 2.13 .84 . . 1.53 2.13 .77 . . 1.46 2.13 .84 . . 1.57 2.16 .80 . . 1.55 2.16 .92 . . 1.68 2.25 .92 . . 1.69 2.26 1.09 . . 1.85 2.24 1.09 1.05 ,7 2.05 1.84 2.41 1.09 1.05 ,,,,,, 2.05 1.82 2.43 1.12 1.08 ,,,,,, 2.11 1.85 2.40 1.17 1.13 ,,,,,, 2.19 1.89 ,,,,,, 1.20 1.17 ,,,,,, 2.22 1.92 H, H, 1.12 H, 2.19 1.87 2.44 H, 1.19 ,,,,,, 2.25 1.89 2.48 W 1.20 ,,,,,,, 2.26 1.89 2.50 , , 1.22 "H 2.30 1.88 2.48 ,,,,,, 1.20 "A" 2.29 1.84 2.46 1.26 _ 2.37 1.88 ...... 1.33 , 2.43 1.92 2.58 1.37 2.47 1.95 2.60 1.57 2.67 2.11 2.58 ,,,, 1.58 ,,,,,, 2.70 2.09 2.74 ,,,,, 1.61 W , 2.71 2.11 2.75 ,,,,, 1.68 "H 2.78 2.15 2.78 ,,,,, 1.71 , 2.82 2.16 2.80 ,,,,, 1.77 H , 2.87 2.27 2.85 ,,,,,, 1.81 H, 2.91 2.32 2.85 ,,,,,, 1.84 "H 2.94 2.34 2.87 ,,,,,, 1.87 .H. _ 2.98 2.37 2.91 1974, as well as during runs 3 and 4 made on March 25, 1974, and runs 5 and 6 made on March 26, 1974. The effect of the slight unsteadiness was accounted for by noting the time at which each measurement was ob- tained. All measurements were obtained by working in the downstream direction at a very nearly constant rate. The beginning times of runs 2, 3, 4, 5, and 6 were 1354, 0635, 1100, 0632, and 1012 hours, respectively, and the ending times were 1922, 0903, 1325, 0829, and 1705 hours, respectively. RESULTS RESULTS GENERAL 13 All data obtained during this study are available on magnetic tape from the Automatic Data Processing Unit of the US. Geological Survey. The tape contains 10-minute values of all meteorologic and temperature data, hourly values of hydraulic data, and daily values of the supplementary data. Table 4 contains daily averages of all meteorologic TABLE 4.—Daily averaged values of meteorologic and temperature data for San Diego Aqueduct (July 25, 1973—July 23, 1974) Da Wind Wind Solar Atmospheric Air Vapor Water 0 speed azimuth" radiation radiation temperature pressure temperature month (m/s) (degree) (W/m’) (W/m‘z) (°C) (kPa) (°C) COT' SKN2 COT SKN COT SKN COT SKN COT SKN COT SKN COT SKN July 1973 1.31 211 ,,,,,, 332 ...... 394 __________________ 25.4 1.23 192 326 331 390 385 26.2 27.1 25.7 25.5 1.27 222 320 326 386 378 24.5 25.5 25.8 25.7 1 14 203 310 319 391 382 24.3 25.2 25.8 25.7 1.18 204 304 319 396 381 24.4 24.7 26.1 25.8 1.33 215 306 310 398 385 24.8 25.1 26.2 25.9 1 26 ...... 199 310 309 410 401 25.9 25.8 26.0 26.0 August 1973 .18 210 269 246 419 408 26.4 26.4 26.2 26.0 .20 203 288 308 426 413 27.0 27.6 26.5 26.5 .45 220 310 314 421 408 28.0 26.5 26.3 26.6 .36 207 197 208 408 396 23.8 22 7 25.8 25.8 .42 224 297 318 392 386 22.5 22 1 25.9 25.8 .27 238 288 290 391 392 21.7 21.4 26.1 25.9 .06 209 308 306 382 373 22.5 21.4 26.0 25.7 1.07 218 ,,,,,, 329 111111 377 ______ 23.3 ,,,,,, 25.6 1.18 199 300 318 394 384 23.9 23.7 25.9 25.7 1.22 204 317 325 383 369 23.6 22.1 26.0 25.7 1.10 219 302 308 395 395 23.6 22.3 26.1 25.8 1.11 226 289 296 414 402 26.7 25.5 26.3 26.1 1.36 201 239 269 422 409 26.5 24.2 26.5 26.2 1.13 198 267 257 426 413 26.6 25.1 26.5 26.4 1.31 201 288 303 418 406 25.7 25.3 26.4 26.3 1 16 217 280 295 420 407 26.9 26.4 26.7 26.6 1.23 219 260 287 426 411 27.2 26.3 26.8 26.6 1.30 230 276 301 420 413 27.9 29.8 26.5 26.6 1.61 232 291 299 414 411 27.6 29.9 ,,,,,,,,,,,, 26.5 26.5 ...... 248 180 223 431 424 27.8 28.3 111... 1.92 26.4 26.0 1.29 222 ,,,,,, 286 ______ 404 ,,,,,, 27.3 ...... 1.75 ______ 26.3 1.18 184 295 306 385 381 25.2 26.1 1.20 1.28 26.2 25.9 1.24 221 318 323 354 349 22.4 22.6 ,,,,,,,,,,,, 25.9 25.3 1.32 229 313 314 351 355 21.3 21.6 ____________ 25.5 25.1 1.50 206 277 292 364 375 17.6 20.0 1 35 1.54 24.6 24.5 1.71 ,,,,,, 226 287 290 368 370 18.1 20.0 ,,,,,, 1.45 ,,,,,, 24.0 1.44 ______ 244 290 299 383 359 19.8 20.0 ,,,,,,,,,,,, 24.1 1.31 ...... 208 313 300 405 353 22.9 20.9 1 19 122 23.7 1.15 ,,,,,, 227 289 294 ______ 363 24.6 23.0 ...... 1.21 23.9 99 270 207 286 276 413 372 23.8 20.0 ,,,,,, 1.55 24.2 1.25 207 219 282 271 383 376 19.7 18.9 ,,,,,, 1.49 ,,,,,, 24.3 .06 225 200 360 21.3 18.4 ______ 24.2 .31 205 218 366 111111 19.2 24.1 .46 220 231 378 ...... 19.6 24.0 .09 240 209 379 ______ 17.8 23.7 .22 29 249 378 ,,,,,, 19.5 24.2 .00 66 252 366 ,,,,,, 22.4 24.5 .08 211 202 370 ,,,,,, 22,5 ,,,,,, .35 185 219 384 18.9 1 .28 191 211 388 18.5 1 _ .24 221 254 394 18.0 1 1 .32 52 252 385 20.3 1 ,,,,,, .99 261 192 374 19.7 _ 24.4 1.15 259 216 389 18.7 1 24.4 1.11 197 230 379 19.7 1 24.8 102 196 194 368 18.1 1 24.5 1.31 199 206 356 17.7 1 24.2 1.17 206 226 352 18.0 1 23.8 1.23 202 208 339 18.3 1 23.7 1.09 210 209 348 19.5 1 23.7 1.19 235 188 359 18.1 1 23.8 1.08 190 207 354 19.2 1 23.8 1.09 57 202 381 16.7 1 23.4 1.54 216 203 366 17.5 1 23.5 1.08 257 218 354 17.6 _ 23.4 1.68 245 41 367 20.9 1 22.8 1.73 45 231 353 24.6 1 22.4 1.28 70 24 360 25.8 1 22.3 .87 74 212 362 25.2 1 22.1 .85 219 197 358 23.7 1 22.2 1.06 204 197 344 20.8 111111111111111111 22.4 See footnotes at end of table, 14 METEOROLOGY. HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA TABLE 4.—Daily averaged values of meteorologic and temperature data for San Diego Aqueduct ( July 25, 1973—July 23, 1974) —Continued Da Wind- Wind SoIar Atmospheric Air Vapor Water 0 speed azimuth“ radiation radiation temperature pressure temperature month (m/s) (degree) (W/m“) (W/mz) (”C) (kPa) (°C) COT‘ SKN2 COT SKN COT SKN COT SKN COT SKN COT SKN COT SKN October 1973 208 230 358 16,9 16.4 243 189 355 16.4 15,5 192 218 356 17.2 16.9 90 232 341 22,2 21.6 202 233 343 19,4 19.2 197 72 360 14.4 15.5 206 199 357 16.2 16.5 217 ,,,,,, 118 364 ,,,,,, 15,2 184 176 186 345 15.1 15,4 44 224 227 316 14.2 15,6 221 242 218 325 16.2 16,6 62 219 212 339 17.6 19.0 30 218 217 355 21,2 22.7 226 220 220 353 21.5 23.6 239 217 214 358 21.8 23.3 239 199 183 365 23.3 24.2 39 211 1.97 367 24.1 25.3 29 213 205 378 23,8 24.8 197 207 203 368 22.6 23.8 200 211 204 347 19.0 20.1 219 198 203 337 . 16.9 213 191 194 340 14.9 252 151 168 350 16.1 36 198 193 333 17.1 193 191 193 333 18.1 27 161 178 335 18.8 42 204 201 329 21.5 191 200 198 333 21.8 27 191 191 329 20.4 20 197 193 327 20.2 182 191 188 348 340 203 November 1973 220 166 15.9 18.2 225 152 14.4 18.6 180 158 13.4 18.4 202 179 14.0 18.4 205 177 15.2 17.9 257 175 15.1 17.6 192 166 17.2 17.9 184 150 18.6 18.3 19 141 15.2 18.3 39 159 18.7 18.5 20 147 20.9 18.5 200 163 17.9 18.5 181 102 13,7 18.3 237 145 12.2 17.9 221 161 12.0 17.2 187 127 12.4 17.3 188 65 13.7 17.4 225 92 11.5 16.8 232 154 9.2 16.7 189 146 9.1 15.8 198 89 10.5 15.2 189 97 10.0 15.4 86 116 8.2 15.6 193 85 8.0 15.0 200 139 8.2 14.9 41 135 9.7 15.0 29 146 10.8 14.6 8 149 14.4 14.5 1 119 , 14.2 14.6 199 142 301 326 10.1 12.3 14.4 December 1973 105 213 226 41 372 326 7.3 9.2 15.0 13.9 107 201 142 403 312 6.9 8.5 14.9 14.2 84 200 43 148 289 310 7.3 '95 14.0 13.8 57 202 39 84 283 320 7.5 9.6 13.6 12.8 101 3 10 140 314 320 9.4 11.3 13.7 13.2 154 13 49 144 299 325 10.6 13.9 13.8 13.4 1 15 21 4 126 317 332 12.8 14.4 13.7 13.3 98 205 6 142 310 339 12.3 15.3 13.6 13.3 155 203 4 146 299 331 12.3 15.5 13,5 13.2 1 18 5 21 144 299 332 12.6 15.2 13,5 13.1 1 00 204 232 115 274 325 10.7 12.7 13.5 13.1 .82 ______ 200 125 ...... 317 ,,,,, 10,4 ._.. , 13.1 1.05 _ 27 104 325 8.6 ...... 13.1 .84 264 84 339 10.9 , 13.3 1.64 _ 17 132 326 14.1 W 13.2 .65 , 230 127 334 14.4 ,,,,, 13.2 .76 . 207 134 327 13.4 "W , 13.0 1.69 , 33 137 309 , 11.9 W . 12.8 1.55 ,,,,,, 80 130 309 ,,,,,, 12.8 ,,,,, 14.0 211 3 64 141 305 9.1 12.8 12,6 12.3 85 191 241 98 311 8.1 10.6 12.4 12.0 1 01 244 237 90 325 8.4 9.6 12.4 12.2 93 199 7 121 308 6.4 8.4 12.4 12.1 1 70 37 16 132 300 7.4 10.1 12.2 12.0 78 204 188 119 307 6.9 9.5 11.8 11.5 66 229 194 129 313 7.6 9.5 11.6 11.4 82 215 233 38 334 6.7 8.8 11.9 ,. .. , 81 196 202 57 353 10.1 11.2 11.9 102 197 186 118 348 11,3 11.9 12.3 117 184 269 114 337 10.7 12.1 12.3 105 217 200 129 316 8.9 10.2 ,,,,,,,,,,, 12.0 W. , RESULTS 15 TABLE 4.—Daily averaged values of meteorologic and temperature data for San Diego Aqueduct ( July 25, 1973—July 23, 1974)—Continued Da Wind- Wind Solar Atmospheric Air Vapor Water 0 speed azimuth3 radiation radiation temperature pressure temperature month (m/s) (degree) (W/m2) (W/m") (°C) (kPa) (°C) COT‘ SKN2 COT SKN COT SKN COT SKN COT SKN COT SKN COT SKN January 1974 12 40 47 57 10 13 16 80 16 H1 MB NB B5 80 H. m. w. 3 93 H5 no 88 N6 69 80 69 83 80 WA 95 H3 85 98 91 NB 67 78 SJ N6 m4 m5 H1 BS 95 no 13 98 75 ND NA H0 31 no WD me 36 1L5 98 m2 &7 97 84 95 M5 ms m1 U9 92 ms 8A MB 65 82 92 my 85 US 91 H7 NA m3 H2 M3 116 155 no MD NJ NA ,,,,,, H8 ______ 1 A,._,, ,,,,,, __,_,, 1 ,,,,,,,,,,,,,,,,,, 59 13 82 99 H0 N2 95 73 77 mwewoweemw Hmmfimma» HMHHH NNNNFNNNN mHNOQOme 1 1 1 1 1 1 1 1 1 16 METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA TABLE 4.—Daily averaged values of meteorologic and temperature data for San Diego Aqueduct (July 25, 1973—July 23, 1974) —Continued Da Wind- Wind Solar Atmospheric Air Vapor Water 0 speed azimuth“ radiation radiation temperature pressure temperature month (m/s) (degree) (W/m”) (W/mz) (“0) (kPa) (°C) COT‘ SKN2 COT SKN COT SKN COT SKN COT SKN COT SKN COT SKN Aprll 1974 1.78 225 205 139 147 ,,,,,, 2.34 218 209 234 233 13.4 1.16 218 242 287 294 12.5 2.15 46 64 293 300 16.1 1.28 36 254 297 303 1811 1.41 235 184 295 295 16.2 1.16 227 255 291 297 1814 1.15 221 237 296 299 17.4 3135 227 247 258 274 12.0 1.71 267 217 301 305 11.4 1.50 62 53 302 306 14.3 1.19 204 215 302 307 15.9 1.16 17 195 301 298 17.4 1.01 2 64 308 308 18.9 1.09 64- 207 310 314 18.6 1.05 196 208 312 317 17.0 1.19 207 204 318 319 14.9 1.87 226 211 161 232 12.6 1128 212 208 249 251 12.6 199 228 198 239 232 1215 1149 28 56 309 316 17.0 1 02 225 233 284 291 18.1 1 63 217 194 304 277 12.8 1 39 85 219 283 277 13.4 1 53 213 197 312 324 12.2 1 12 232 181 327 323 12.7 1 07 221 216 308 319 1510 1 15 217 214 312 317 16.5 1 69 42 44 326 327 19.5 1.28 36 239 328 328 21.2 .59 1.38 15.2 .75 1.24 16.1 .72 1.37 18.2 .14 1.30 16.0 .15 1.46 16.0 .84 148 14.9 .79 1.53 15.3 .53 1.73 15.7 .06 1.54 15.4 .37 1.63 14.2 .54 2.18 12.8 .39 1.17 17.2 .20 1.37 20.7 .07 1.20 20.6 .47 1.35 16.9 .21 .93 19.7 .31 1.45 26.7 . .96 26.9 1.14 23.0 1.38 18.5 1.34 16.7 1.42 16.2 1.30 162 June 1974 1.34 232 222 266 309 18.2 1.06 220 206 280 313 18.6 1.02 235 225 251 243 16.8 1.21 206 211 290 299 17.9 1.28 211 227 315 317 19.6 1.37 212 204 313 301 18.8 1.51 219 233 74 71 15.3 1.42 213 230 341 343 19.2 1.51 239 238 345 348 20.8 1.05 231 200 333 335 20.0 1.11 51 200 310 324 20.0 1.05 85 207 312 334 21.1 1.04 89 199 323 326 20.6 .91 57 234 349 340 24.1 1.02 53 210 346 341 27.7 1.43 210 212 351 353 24.7 1.19 201 216 351 355 22.0 1.42 215 213 350 357 20.2 1.13 61 217 343 357 20.0 1.10 208 214 353 359 22.7 .96 207 203 354 360 24.0 1.13 60 215 340 355 26,8 .98 56 216 333 343 29.2 1.11 59 259 337 343 30.1 1.24 59 234 332 346 30.4 .92 57 202 334 338 29.2 1.13 248 208 333 342 29.2 1.16 205 199 346 356 29.8 1.29 208 205 348 360 25.1 1.30 217 218 349 362 21.3 RESULTS 17 TABLE 4.—Daily averaged values of meteorologic and temperature data for San Diego Aqueduct (July 25, 1973—July 23, 1974) ——Continued Da Wind- Wind Solar Atmospheric Air Vapor Water ofy speed azimuth3 radiation radiation temperature pressure temperature month (m/s) (degree) (W/m’) (W/mz) (°C) (kPa) (“0) COT‘ SKN2 COT SKN COT SKN COT SKN COT SKN COT SKN COT SKN July 1974 1.41 215 207 330 317 19.6 18.7 24.8 24.7 1.51 217 227 316 337 19.6 19.4 24.8 24.6 1.13 212 218 346 354 22.3 22.5 24.7 24.6 1.07 60 222 331 341 25.4 26.3 24.9 24.8 1.22 216 192 341 346 25.8 25.9 25.2 25.1 1.44 222 214 361 361 22.6 23.1 25.2 25.1 1.52 213 205 349 345 20.1 20.6 24.8 24.8 1.59 216 209 338 344 19.3 19.8 24.4 24.2 1.36 226 219 343 343 20.4 20.3 24.5 24.3 1.67 231 215 306 302 18.8 19.1 24.9 24.5 1.25 226 218 330 344 20.8 20.6 24.7 24.8 1.21 248 222 332 346 22.8 22.6 24.5 24.6 .95 247 214 272 235 25.3 24.5 24.9 24.6 1.12 73 32 135 114 26.4 25.6 25.2 24.9 1.07 39 184 286 269 26.0 24.9 25.1 25.2 1.31 266 194 320 333 26.4 26.8 25.5 25.5 1.14 71 204 321 342 26.8 27.4 25.8 25.8 1.84 223 237 331 343 26.7 27.3 25.9 25.8 1.52 200 216 209 225 26.5 27.1 26.0 25.9 1.38 67 199 318 328 26.9 27.3 25.8 26.0 1.20 229 206 319 322 25.2 25.2 25.7 25.6. 1.35 218 223 274 330 24.8 24.9 26.3 25.9 ,,,,,, 202 HHH 216 HHH 26.0 HHH HHH HHH 26.1 HHH ‘Cottonwood (upstream) end of canal. :ZSkinner (downstream) end of canal. "Direction from which the resultant wind comes measured clockwise from north. and temperature data. Averages are included in table 4 only if sufficient data were present to adequately define the daily mean. The windspeed, solar radiation (short- wave), atmospheric radiation (longwave), and air tem- peratures represent averages of individual meas- urements. The wind azimuth represents the azimuth of the resultant wind vector for the day, and the water temperature represents the average obtained at all levels of measurement. A wind azimuth of zero repre- sents a wind blowing from north to south. The vapor pressure was computed from the wet- and dry-bulb air temperatures for each 10-minute period and the results averaged. The conversion from wet- and dry-bulb tem- peratures to vapor pressures involved a two-step proc- ess. First the saturation vapor pressure of air at the wet-bulb temperature was computed using the formula e... = exp [52.418 — 6788.6/(273.16 + TW) — 5.0016 In (273.16 + TW)] (3) in which e... is saturation vapor pressure in kilopascals, and TW is wet-bulb temperature in degrees Celsius. The vapor pressure was then computed from e... — 96.0(0.00066) (TA — TW) (1 + 0.00115 TW) . .. (4) ea in which ea is vapor pressure of the air in kilopascals, TA is dry-bulb temperature in degrees Celsius, the at- mospheric pressure is assumed to be 96.0 kilopascals, and 0.00066 is the psychrometric constant. Table 5 contains daily averages of all hydraulic data as well as daily rainfall values. The diversion dis- charge at EM—8 is not listed, because the mean daily value was always less than 0.1 m3/s. The diversion discharge at the So. End diversion was not measured but can be determined by continuity considerations, however. This diversion was zero prior to May 15, 1974. The rain gages were serviced at approximately 0900 hours. The observed rainfall was assigned to the day preceding the date of observation. In some cases, the record had the word “trace” written in the rainfall column. In this case, the rainfall was recorded as 0.25 mm. TABLE 5,—Daily averaged values of hydraulic and rainfall data for San Diego Aqueduct (July 25, 1973—July 23, 1974) Day Stage lcm) Discharge (ml/s) Rainfall' (mm) of Cottonwood Simpson Newport South Skinner Cottonwood Simpson Skinner Cottonwood Skinner month gage end inlet inlet wood July 1973 25 , , , , ,241 ,,,,,,,,, 193 190 H. H._.. 13.9 0.0 H H H 0.0 0.0 26 , H H ,240 ,,,,,,,,,, 192 189 H ._H _ 13.7 .0 , H. H. .0 .0 27 , H H, 239 ,,,,,,,,, 191 189 95 13.7 .0 , , H .0 .0 28 ,,,,,,,,,, 239 ,,,,,,,,,, 191 190 170 13.7 .0 , H H, .0 .0 29 H _ , , , 239 ......... 191 190 170 13.7 .0 H , _ .0 .0 30 . H 238 H H. 191 190 170 13.6 .0 , , . .0 .0 31 _ 236 110 191 191 170 13.4 .0 H , v .0 .0 See footnote at end of table. METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA TABLE 5.—Daily averaged values of hydraulic and rainfall data for San Diego Aqueduct (July 25, 1973—July 23, 1974) —Continued Day Stage (cm) Discharge (mu/s) Rainfall‘ (mm) of Cottonwood Simpson Newport South Skinner Cottonwood Simpson Skinner Cottonwood Skinner month gage end inlet inlet w August 1973 187 188 168 12.9 0.0 0.0 0.0 183 184 164 126 .0 .0 .0 183 185 164 126 .0 .0 .0 183 186 164 126 .0 .3 .0 183 186 164 126 .0 .0 .0 183 186 164 126 .0 .0 .0 184 187 165 126 .0 .0 .0 184 187 165 128 .0 .0 .0 185 188 166 129 .0 .0 .0 185 188 166 129 .0 .0 .0 185 188 166 129 .0 .0 .0 185 188 166 129 .0 .3 .5 184 187 166 129 .0 .0 .0 184 187 166 130 .0 .0 .0 184 187 166 130 .0 .0 .0 184 187 166 130 .0 .0 .0 184 186 166 129 .0 .0 .0 184 187 166 129 .O .0 .0 183 187 165 129 .0 .0 .0 183 186 165 128 .0 .0 .0 183 186 165 127 .0 .0 .0 183 186 166 128 .0 .0 .0 183 186 166 129 .0 .0 .0 185 188 167 130 .0 .0 .0 184 188 167 130 .0 .0 .0 184 187 167 130 .0 .0 .0 184 187 166 130 .0 .0 .0 184 187 166 130 .0 .0 .0 184 188 166 130 .0 .0 .0 184 188 167 130 .0 .0 .0 185 210 198 130 .0 .0 .0 September 1973 185 243 241 13.0 0.0 0.0 0.0 240 13.0 .0 .0 .0 239 130 .0 .0 .0 239 13.0 .0 .0 .0 219 .......... .0 .0 .0 166 .......... .0 .0 .0 178 130 .0 .0 .0 235 130 .0 .0 .0 234 130 .0 .0 .0 234 130 .0 .0 .0 237 13.0 .0 .0 .0 259 12.7 .0 .0 .0 .......... 123 .0 .0 .0 .......... 122 0 .0 0 164 120 .0 .0 .0 158 118 .0 .0 .0 158 119 .0 .0 .0 159 118 .0 .0 .0 160 118 .0 .0 .0 160 118 .O .0 .0 160 118 .0 .0 .0 160 118 .0 .0 .0 160 118 .0 .0 .0 160 118 .0 .0 .0 159 118 .0 .0 .0 159 1L8 .0 .0 .0 160 120 .0 .0 .0 163 124 .0 .0 .0 165 127 .0 .0 .0 182 184 165 127 .0 .0 .0 October 1973 181 184 164 12.7 0.0 0.0 0.0 181 185 165 127 .0 .0 .0 182 185 165 128 .0 .0 .0 180 184 164 127 .0 .0 .0 176 182 162 121 .0 .0 .0 174 180 160 1L6 .0 .0 .0 166 175 156 107 .0 .0 .0 160 170 151 100 .0 .0 .8 159 169 150 101 .0 .0 .0 156 167 149 09 .0 .0 .0 161 169 150 107 .0 .0 .0 172 180 158 1L2 .0 .0 .0 173 181 159 1L3 .0 .0 .0 173 181 159 1L3 .0 .0 .0 174 182 159 1L3 .0 .0 .0 174 183 159 1L3 .0 .O .0 188 .... _.. . 166 13.2 .0 .0 .0 208 .... ..... 181 15.6 .0 .0 .0 208 207 181 156 .0 .0 .0 208 207 181 156 .0 .0 .0 208 207 181 156 .0 .0 .0 207 206 181 156 .0 .0 .0 208 206 181 156 .0 .0 .0 206 205 180 156 .6 .0 .0 201 201 177 156 L1 .0 .0 198 198 175 156 L3 .0 .0 198 198 175 156 L3 .0 .0 198 198 175 15.6 1.3 .0 .0 198 199 175 156 L3 .0 .0 199 199 175 156 L3 .0 .0 156 13 .0 0 199 199 175 RESULTS TABLE 5.—Daily averaged values of hydraulic and rainfall data for San Diego Aqueduct (July 25, 1973—July 23, 1974) -—C0ntinued Day Stage (cm) Discharge (m3/s) Rainfall‘ (mm) of Cottonwood Simpson Newport South Skinner Cottonwood Simpson Skinner Cottonwood Skinner month gage end inlet in et wood November 1973 199 175 15.6 0.0 0.0 200 176 15.6 .0 .0 201 178 15.6 .0 .0 202 178 15.6 .0 .0 201 178 15.6 .0 .0 201 177 15.6 .0 .0 200 177 15 6 .0 .0 200 177 15 6 .0 .0 199 176 15 6 .0 .0 198 175 15 6 .0 .0 198 175 15 6 .0 .0 197 174 15 6 .0 .0 195 173 15 7 .0 .0 188 167 14 6 .0 .0 184 163 13 9 .0 .0 180 160 13.6 1.8 1.3 181 160 13.3 19.8 14.0 181 159 12.5 7.1 4.1 ,,,,,,,,,, 154 11.6 .0 .0 ,,,,,,,,,, 145 9.7 .0 .0 148 134 8.5 .3 .0 144 130 8.1 127 17.5 140 127 7.3 .0 .0 131 119 6.9 2.8 .8 131 119 6.9 .0 1.0 131 119 6.9 .0 .0 132 119 6.9 .0 .0 132 118 6.9 .0 .0 132 119 6.9 .0 .0 132 118 6.9 1.3 .0 December 1973 132 119 6.9 .0 0.0 1.3 133 119 6.9 .0 .0 .0 133 119 6.9 .0 .0 .0 134 119 6.9 .0 .0 .0 134 119 6.9 .0 .0 .0 134 120 6.9 .0 .0 .0 135 120 6.9 .0 .0 .0 134 120 6,9 .0 .0 .0 134 119 6.9 .0 .0 .0 142 126 8,7 .0 .0 .0 172 152 11.7 .0 .0 .0 177 157 12.4 .0 .0 .0 172 153 11.9 .7 .0 .0 166 148 11.9 1.8 .0 .0 162 145 11.9 2.1 .0 .0 162 145 11.9 2.1 .0 .0 162 145 11.9 2.1 .0 .0 162 145 11.9 2.1 .0 .0 162 145 11.9 2.1 .0 .0 162 145 11.9 2.1 .0 .0 165 148 11.3 1.2 .5 1.0 168 150 10.5 .0 .0 .0 169 150 10.5 .0 .0 .0 169 150 10.5 .0 .0 .0 169 150 10.5 .0 .0 .0 167 149 10.2 .0 .0 .0 184 166 14.9 .0 .5 .0 231 206 21.5 .0 .0 .0 238 212 22.7 .0 .0 .3 238 212 22.7 .0 .0 .0 238 212 22.7 .0 __________ 11.9 .8 January 1974 238 213 22.7 0.0 0.8 0.8 240 215 23.6 .0 .0 .0 245 221 25.4 .0 .0 .5 250 226 26.6 .0 59.7 53.6 151 227 27.0 .0 2.5 1.8 251 227 27.0 .0 29.5 18.5 249 225 26.6 .0 66.0 67.3 230 208 22.3 .0 5.6 .5 229 206 22.1 .0 1.5 1.0 232 211 ,,,,,,,,,, .0 0 .0 244 223 22.1 .0 0 .0 256 242 22.2 .0 0 .0 267 262 19.6 .0 0 .0 273 277 13.0 .0 0 .0 277 282 9.8 .0 0 .0 280 284 9.1 .0 2 8 1.0 283 287 7.6 .0 .3 .0 281 286 6.5 .0 .0 .0 208 284 6.5 .0 .0 .0 279 283 6.5 .0 4.3 2.3 278 282 5.9 .0 .0 .0 277 281 4.9 .0 0 .0 271 276 4.1 .0 0 .0 265 271 3.7 .0 0 .0 260 265 3.7 .0 3 .0 253 258 3.7 .0 0 .0 248 253 3.7 .0 0 .0 243 247 3.7 .0 0 .0 235 237 3.7 .0 0 .0 224 229 3.7 .0 0 .0 212 218 3.7 .0 0 .0 METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUCT, CALIFORNIA TABLE 5.—Daily averaged values of hydraulic and rainfall data for San Diego Aqueduct (July 25, 1973—July 23, 1974)—Continued Day Stage (cm) Discharge (ms/s) Rainfall‘ (mm) of Cottonwood Simpson Newport South Skinner Cottonwood Simpson Skinner Cottonwood Skinner month gage end inlet inlet wood February 1974 200 205 3, 0.0 0 0 188 193 3. 172 177 152 158 bbbbbbbbbbbbbbbbb'o'o‘ob'o'obb .—‘ 9°'o'o'o'o'o'o'o'o'w'o'o'o‘o'o‘ob‘o'ob'o’o‘o'o'o'o'o'o ‘cn'obb'o'o'o'o‘o'o'o’o'w'o'ob'w'o'o'obb'o'o‘o'o'o'o bb 3" N 0.0 10.7 9.4 7.1 0 12.4 9.9 4.3 0 12.5 .0 4.1 0 15.5 0 .0 0 23.3 0 .0 0 23.5 .0 .3 0 24.1 20.6 13.2 0 24.0 21.6 14.2 0 23.0 0 .0 0 23.2 0 .0 0 23.2 0 .0 0 23.7 0 .0 0 23.5 0 0 0 23.6 0 .0 0 23.5 0 .0 0 23.4 0 .0 0 23.5 0 .0 0 19.0 0 .0 0 12.9 0 .0 0 .......... 3 .0 0 . 0 .0 0 . 0 .0 0 _ 0 .0 0 _ 0 .0 0 _ 3 .8 0 . 3 .8 0 .......... 5 .5 0 . 0 .0 0 . 0 .0 0 . 0 .3 O 0 .0 0.0 20.3 6.4 4.8 0 21.5 .0 .0 0 24.5 .0 .0 0 25,6 .0 .0 0 26,0 .0 .0 0 26.0 .0 .0 0 25.7 .0 .0 0 25.5 .0 .0 0 25.5 .0 .0 0 24.4 .0 .0 0 19.8 .0 .0 0 15.9 .0 .0 O 15.9 .0 .0 0 15.7 .0 .0 0 15.5 .0 .0 .6 13.9 .0 .0 1.7 12.7 .0 .0 2.2 12.9 .0 .0 2.2 12.9 .0 .0 2.2 12.9 .0 .0 2.2 12.9 .0 .0 2.2 12.8 .0 .0 2.2 12.7 .0 .0 2.2 12.7 .0 .0 2.1 12.7 .0 .0 2.1 12.7 .0 .0 1.9 12.8 .0 .0 1.9 12.9 .0 .0 1.9 13.0 .0 .0 1.9 13.0 .0 .0 RESULTS TABLE 5.—Daily averaged values of hydraulic and rainfall data for San Diego Aqueduct (July 25, 1973—July 23, 1974)—Continued Day Stage (cm) Discharge (mS/s) Rainfall' (mm) of Cottonwood Simpson Newport. South Skinner Cottonwood Simpson Skinner Cottonwood Skinner month gage end inlet inlet wood May 1974 0.0 0.0 H D D __ .0 .0 a D 0 H D D .0 .3 .0 .0 18 . D 0 18 . D 0 18 . D 0 13 . D D 13 . D D 13 HA 0 D 13 H3 0 0 1.8 9.9 .0 .0 L8 Tl D 0 18 12 0 0 1.8 7.2 .0 .3 L8 T4 0 0 L8 16 0 0 L8 18 D 0 1.8 7.7 .0 .0 1.8 7.8 .0 .0 L8 TB 0 0 L8 78 D 0 L8 79 D 0 1.8 8.0 .0 .0 25 79 D 0 22 80 0 0 23 85 D 0 23 85 D 0 NNNNNNN?‘E‘E‘F‘E‘P‘!‘!‘!"!"!"!"f"f"l"l“f"f"f"f"l‘3l"l° ApppmmoqquuuuwuwHHHHHHHHHmmpA HFF999EHHHHH99999999999999¢@9@ WHHmthNOHNMmmmwmeme®mQQQHO® bbbbbbbbbbbbbbbbbbbbbbbbbbbbbb t-‘t—‘b—‘D-‘D-‘HHWHHHD—‘O—‘Ht—‘HV-‘t—‘HHHHHD—IHH bbbbbbbbbbbbbbbbbbbbbbbbbbbbbb O HHHHHHHMHHHHHHHHHMHH hbbbbbbbbbbbbbbbbbbbbbb NWNNNNNNNNNNNNNNNWNNNNN mmmmmmmmmmmmmmmmmmmmqpe .f‘T‘FY‘HE‘HHF‘f‘T‘VF‘E‘E‘E‘QE‘E‘!‘ pupwpempappmmwmwmowu >— ,_. 4 N) m2 ubbbbbbbbbbbbbbbbbbbbbb ‘Rainfall represents the accumulated amount between 0900 hours of the indicated day to 0900 hours of the following day. 21 22 IVIE'I‘EOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUEDUC'L CALIFORNIA WIND The average windspeed at Cottonwood during the 12-month period was 1.70 m/s, whereas that at Skinner averaged 1.24 m/s. The lower value at Skinner un- doubtedly reflects the sheltering effect of the hill north of the anemometer. The yearly mean values of the sup- plementary windspeeds were 1.54 m/s at Cottonwood and 0.90 m/s at Skinner. The monthly mean windspeeds at Cottonwood and Skinner are shown in figure 21. A rather pronounced seasonal variation is seen at Cottonwood, but none is evident at Skinner. The frequency of occurrence of various daily average windspeeds is shown in figure 22. The uniform nature of the Skinner windspeed and the absence of high winds in comparison to Cottonwood is apparent. The highest wind gust observed at Cottonwood was 18.5 4 IIIII II Cottonwood _ o A S O N D J F M A M J J 1973 1974 4 I | I I I I I I I I I _ Skinner L WINDSPEED, IN METERS PER SECOND 0 A S O N D J F M A M J J 1973 1974 FIGURE 21.—Monthly mean windspeed. Insufficient data were available for November at Cottonwood and for January and February at Skinner. m/s and occurred at 1640 hours on July 18, 1974. The highest gust observed at Skinner was 9.9 m/s and oc— curred at 1420 hours on January 1, 1974. A pronoumced diurnal variation in windspeed was observed at both stations. In order to illustrate this diurnal effect, the yearly mean windspeed was com- puted for each 10-minute interval of the day. The results of this averaging process are shown in figure 23. It can be seen that the windspeed is usually low at night and in the early morning hours (2200—0800) and usually fairly high in the late afternoon hours (1400- 1800). The daily-average windspeeds at Cottonwood and Skinner were poorly correlated. The correlation coeffi- cient for the entire year was only 0.43. For the months of March and April the correlation coefficient was 0.80, but for the months of May through August it was 0.32, and for the months of September through December it was 0.46. The correlation between the daily—average primary and supplementary windspeeds was also poor. Considering the entire year, the correlation between 0.25 I I Cottonwood U) 0.20 — ~ >— E‘ u. 0.15 ~ a 0 LL! 0 0.10 — — <( '2 Lu 0.05 — — E 7T1 Z 0 1 2 3 4 5 6 Lu. 0 Z Lu I I 0.25 I I I I D 8 Skinner O 0.20 ~ _ u. 0 >. _ .1 0 0.15 Z UJ D 0.10 — — O u.I 0: LL 0.05 — _ o | | I 0 1 2 3 4 5 6 DAILY AVERAGE WINDSPEED, IN METERS PER SECOND FIGURE 22.—Frequency of occurrence of daily windspeeds. RESULTS 23 the primary and supplementary windspeeds was 0.64 at Cottonwood and 0.47 at Skinner. These low correla- tion coefficients highlight the difficulty in attempting to measure a representative windspeed for a reach of any open channel. Surface winds are extremely vari- able from point to point. The wind direction was recorded continuously at both ends of the canal. The percentage of the time dur- ing which the wind was coming from each of eight sec- tors (numbered counterclockwise from the north) was tabulated by month and year. The results of this tabu- lation are shown in figure 24. At Cottonwood the wind is from sector 4 (south-southeast) the largest percent- age of the time, but the percentage is fairly uniformly distributed among all sectors. The two lowest per— centages, for winds from the west, are probably due to a slight sheltering effect of the Lakeview Mountains (fig. Cottonwood D Z O o LLI w n: _ _ $0 .l lilil I. 3:) 0 4 8 12 16 20 24 LL] '— LLI E Z 0‘5 lil’lI | | ICE _ _ e %4— Skinner D _Z_ _ e §3_ _ O 4 8 12 16 20 24 TIME, IN HOURS FIGURE 23.—Mean annual diurnal variation in windspeed, averaged by 10-minute time periods for the period July 24, 1973, to July 23, 1974. 1). The distribution of directions did not appear to vary with time of year. The largest percentage fell in sector 4 for every month of the year. Local topography obvi- ously had a great effect on the measured wind di- rections at Skinner. Notice the sheltering effect of the hill to sectors 1, 2, 7, and 8 and of the dam to sector 4 (figs. 1, 6, 9, and 10). Sector 5 had the highest percent- age for months March through September, and sector 3 had the largest percentage for the months of October through December. The mean windspeed, average speed irrespective of direction, and the resultant windspeed (vectorial aver- age speed) were computed for each day. The ratio of these two windspeeds is a measure of how consistent the wind direction is and is called the wind consistency. Cottonwood Skinner FIGURE 24.—Frequency of occurrence, in percentage of time, of winds from various directions. 24 METI’.()R()I.(,)(;Y, HYDRAULICS, AND THERMAL REGIME. SAN DIF.('-() AQL'I‘IDUC’I‘, (IALIFORN IA The wind consistency was computed for each day and the monthly average determined by averaging the daily values. At Cottonwood the monthly average con— sistency varies from a low of 0.12 in February to a high 0f0.76 in November, but no annual cycle was apparent. At Skinner the consistency values are probably less meaningful, but they varied from a low of 0.15 in De- cember to a high of 0.71 in May. The months of May through August had values greater than 0.60, and Oc- tober, November, and December had values less than 0.25. Other months had values between 0.38 and 0.49. RADIATION The annual pattern of daily solar radiation is shown in figure 25. The yearly mean measured at Cottonwood and Skinner was 238 W/m2 and 242 W/mz, re.- spectively, a difference of less than 2 percent. The daily values at Cottonwood and Skinner were highly corre- lated. The correlation coefficient between daily aver- age values for the entire year was 0.97. The standard deviation of the difference between the daily means at the two ends of the canal was 21.4 W/mz. Also shown in figure 25 is the variation of the esti- mated clear-sky solar radiation. This value was com- puted using the procedure suggested by the Tennessee Valley Authority (1972). The clear-sky solar radiation was determined from ch... = Io_(h,. sin <1) sin 5 + cos <1> cos a sin hss) (5) nrz in which q)“. is clear-sky solar radiation,Io is effective solar constant, r is ratio of actual to mean Earth to Sun distance, h... is hour angle of sunset in radians, d) is latitude, and 5 is declination of the Sun. The effective solar constant was varied by trial and error until the clear—sky curve appeared to form an envelope of the measured data. The value used in figure 25 is 1,046 W/mz. The value ofr was approximated from r = 1 + 0.017 (cos @ (186 — D)) (6) 365 in whichD is Julian date (1 —365). The value ofhxx was determined from h : arc CO-S (sin a... — sin (13 sin 5 S." (7) cos (1) cos 5 in which as, is solar altitude at sunset, which was as- sumed to be zero. The latitude of the Cottonwood and Skinner sensors was N. 33°47’ and N. 33°36’, re- spectively. The declination of the Sun was approxi- mated by use of the expression 360 = .4 — 172 —D . 8 5 23 5cos(365( )) () From equations 5, 6, 7, and 8 the yearly average clear- sky radiation at Cottonwood and Skinner is computed as 268.8 and 269.3 W/mz, respectively, a difference of 0.2 percent. The annual variation of atmospheric radiation is shown in figure 26. The estimated clear—sky atmos- pheric radiation is also shown for reference. The clear- sky value was computed from the formula presented by Idso and Jackson (1969) <12.” = 0' (Ta + 273.16)‘ (9) (1 — 0.261 exp (—0.000777 (Tam) in which (13,.“ is incoming clear-sky atmospheric radia- tion, a is Stefan-Boltzman constant (5.671 X 10‘8 W/mZ), and Ta is air temperature in degrees Celsius. The presence of clouds should increase the incoming 400l||l||||111| Cottonwood 300 Clear sky / 200 f l i D: LIJ 1—Illlllllllll g ASONDJFMAMJJ 'E' 1973 1974 < 3 O (D 0:4oollllllllllll LU o. Skinner 300 DAILY SOLAR RADIATION, IN WATTS Clear sky 200 — / 1111 100# Illl ONDJFMAMJJ A S 1973 1974 FIGURE 25.~Daily solar radiation. RESULTS 25 atmospheric radiation over the clear-sky value by as much as 20 to 25 percent. The measured values would, therefore, be expected to be equal to or greater than the clear-sky values. In general, this appears to be the case as indicated in figure 26. Notable exceptions do occur, particularly at Cottonwood after March 1, 1974. After November 28, 1974, the atmospheric radiation at Cot- tonwood was determined for each 10—minute interval by subtracting the measured solar radiation from the all-wave radiation measured by a flat-plate radiometer. The diurnal variation in atmospheric radiation, measured by the pyrgeometer for relatively clear days, is illustrated in figure 27. This mean diurnal variation was determined by averaging, on a time-period-by- time—period basis, the measured atmospheric radiation for 18 days. These days were selected between July 25, 1973, and December 1, 1973, from both Cottonwood and Skinner, such that the ratio of the measured solar to the computed clear-sky solar radiation was greater 5°°IIIIIIIIIIII 400 MI]! |_ 300; 2” 5M .,- ..-. 3:...:.-.,:'.,:.:3,‘ Clear sky 200 P ‘ 100 — a Cottonwood 0IIIIIIIIIIII ASONDJFMAMJJ 1973 1974 5°°IIIIIIIIIIII PER SQUARE METER 400 :1 _: " Clear sky / l 7 200 — — 300 DAILY ATMOSPHERIC RADIATION, IN WATTS 100 — # Skinner V 4‘ OIIIIIIIIIIII ASONDJFMAMJJ 1973 1974 FIGURE 26,—Daily atmospheric radiation. than 0.97. Also shown in figure 27 is the mean diurnal variation in atmospheric radiation as determined by subtracting the instantaneous value of the measured solar radiation from the instantaneous value of the all-wave radiation measured by the flat-plate radiome— ter. This curve represents an average of the 4 days after November 28, 1973, for which the ratio of the measured solar to computed clear-sky solar radiation was greater than 0.97. Unfortunately these 4 days are not included in the set of 18 days represented by the pyrgeometer, because it was impossible to select any clear-sky days during which both instruments were operating satisfactorily. The diurnal variation in the atmospheric radiation, as measured by the pyrgeome- ter, follows very closely the diurnal variation computed by equation 9. On the other hand, the flat plate, with the solar component deducted, indicates a much 600 I l I l I l I I I I I 400 Pyrgeometer m l- 200 '— E Z - \Flat-plate radiometer ~ — o: _ _ - u.I Z I— 0 I l J_ l I l I l | I I 9 L” o 4 s 12 16 2o 24 l- E <12 LU _ n: 2 < I D 0 O — V’ 600 x m l I l I l l | | T 7* § E ‘ Cloudy ‘ D. — _ 8 g _ Pyrgeometer 7 I— 400 / < 200 * 0 4 8 12 16 20 24 TIME, IN HOURS FIGURE 27.—Typical diurnal variation in atmospheric radia- tion as measured by the pyrgeometer and by subtracting the measured instantaneous solar radiation from the all-wave radiation measured by the flat-plate radiometer. 26 MI‘I'I'I‘ZOROLOGY. HYDRAULICS, AND THERMAI greater night—to-day variation. The mean value of these two curves should not be compared, because each curve represents the average of several but different days. The diurnal variation in the atmospheric radia- tion, as determined by deducting the solar radiation from the flat-plate all-wave radiation, appeared to be a function of the cloudiness. A comparison of the atmos- pheric radiation estimated by use of the pyrgeometer and the flat plate under cloudy to overcast conditions is also illustrated in figure 27. This figure represents the average of all days with a ratio of measured to clear- sky solar radiation ofless than 0.60. Thirteen days are included in the pyrgeometer curve and 20 days in the flat-plate curve. From an inspection of figure 27, it would appear that the flat plate is overly sensitive to the solar component of the radiation spectrum. A. P. Jackman (oral c0mmun., 1975) found flat plates to be overly sensitive to solar radiation. AIR TEMPERATURE The annual pattern of daily average air temperature is shown in figure 28. The mean daily temperatures at Cottonwood and Skinner had a correlation coefficient of 0.99 for the period of record. This correlation coeffi— cient remained stable throughout the year. The mean air temperatures at Cottonwood and Skinner were 17.06 and 17.90°C, respectively. The rather sheltered location of the Skinner station, on a south hillside, is undoubtedly reflected in its higher mean temperature. The standard deviation of the difference between the daily average temperatures at the two ends of the canal was 145°C. VAPOR PRESSURE Other than atmospheric radiation, the vapor pres- sure of air proved to be the most difficult parameter to measure. More specifically the wet-bulb temperature was very difficult to measure continuously because of the problem of keeping the wick saturated. Overall, wet-bulb data are available for 44 and 56 percent of the 10-minute time periods at Cottonwood and Skinner, respectively. The annual pattern of daily averaged vapor pressure values is shown in figure 29. Daily averages are shown in figure 29 only if more than 130 of the 144 possible 10-minute periods contained data. On a daily average basis the correlation between the Cottonwood and Skinner vapor pressures was good. The correlation co— efficient was 0.96, and the standard deviation of the difference between the daily averaged values was 0.15 kPa (kilopascals). The mean value at Skinner was higher than that at Cottonwood by 0.10 kPa. This higher value undoubtedly reflects the nearness of Lake Skinner. The standard deviation among daily values . REGIME, SAN DIEGO AQUEDUUI‘, CALIFORNIA was also higher at Skinner (0.38 kPa) than at Cotton- wood (0.30 kPa). The higher variability at Skinner would also be expected because the presence or absence of the vapor blanket from Lake Skinner would depend on wind direction, which is of course quite variable. Because of the large number of missing wet-bulb data and because most of the time complete data were available at either Cottonwood or Skinner, the transferability between Cottonwood and Skinner on a time-period-by-time-period basis was investigated. An obvious question arises: How can one best estimate the vapor pressure at one point given the dry- and wet-bulb temperatures at a remote site and the dry-bulb air temperature locally? There are four possible ways of estimating the vapor pressure under these conditions: (a) Assume it is the same as at the remote station (this method makes no use of the information known about the local dry-bulb temperature); (b) assume the wet- bulb temperature is the same at both stations (this method makes no use of the dry-bulb temperature at 30} I I I I I I I I I («f Cottonwood 20* m _ 2 , III/W — LIJ ~ _ o - _ U) Lu _ _ Lu _ _ g I I I I I I I I I I I I LU A S O N D J F M A M J J a Z 1973 1974 Lu. 0: B30 < I I I I I I I E Skinner D. 2 Lu l‘20 E < 10 ASONDJFMAMJJ 1973 1974 FIGURE 28,—Daily average air temperature. RESULTS the remote station); (c) assume wet—bulb depression (dry-bulb temperature minus the wet-bulb tempera- ture) is the same at both stations; or (d) assume the relative humidity is the same at both stations. Using the 11,688 10—minute periods for which complete data were available at both ends of the canal, the accuracy of these four transfer methods was checked by assuming one or the other of the wet-bulb temperatures was missing, estimating the vapor pres- sure by all four methods, and comparing the estimated value to the known local value. Morning, afternoon, and nighttime data were grouped separately to deter- mine if time of day had any influence. The results are tabulated in table 6. In every case, transferring the vapor pressure directly resulted in the smallest root- mean-square error in the estimated vapor pressure. In two cases, the correlation ecoefficient resulting from the use of this method was slightly less than that resulting from the transfer of the wet-bulb tempera- Illlllllllll ASONDJFMAMJJ 1973 1974 VAPOR PRESSURE, IN KILOPASCALS llllllllllll ASONDJFMAMJJ 1973 1974 FIGURE 29.—Daily average vapor pressure. 27 TABLE 6.—Comparisons of vapor pressures estimated by four data— transfer methods Transferred Estimating Skinner Estimating Cottonwood quantity vapor pressures vapor pressures Correlation Root-mean- Correlation Root-mean~ coefficient square error coefficient square error All data Vapor pressure ________ 0.837 0.271 0.837 0.271 Wet-bulb temperature __ .825 .327 .774 .326 Wet~bulb depression ,___ .528 .700 .780 .348 Relative humidity ,,,,,, .775 .302 .806 .312 Morning (0600—1200 hours) Vapor pressure ,,,,,,,, 0.875 0.225 A 0.875 0.225 Wet-bulb temperature .. .860 .304 .801 .302 Wettbulb depression .... .631 .509 .833 .283 Relative humidity ,,,,,, .830 .256 .851 .252 Afternoon (1200—1800 hours) Vapor pressure 1..."; _ 0.874 0.227 0.874 0.227 Wet-bulb temperature .. .834 .313 .781 .312 Wet-bulb depression . . .. .581 .606 .823 .289 Relative humidity __._ , .819 .257 .845 .261 Night (1800—0600 hours) Vapor pressure ,,,,,,,, 0.767 0.383 0.767 0.383 Wet-bulb temperature .. .777 .386 .751 .386 Wet-bulb depression .1 ., .500 1.061 .705 .499 Relative humidity W... ._ .723 .415 .734 .447 ture, but no particular significance is attached to this. While transferring the vapor pressure directly was clearly the most accurate procedure, the least accurate procedure is not as well defined. It appears that transfer of the wet-bulb depression is probably the least accurate. Even on a minute-by-minute basis, the vapor pressures are faily well correlated between sites. Very little diurnal variation in vapor pressure was ob- served. It is also interesting that the correlation is poorer between Cottonwood and Skinner and the errors larger at night than during the day. This is probably due to the fact that the lake influenced Skinner to a larger degree at night, when winds were light, than during the day. During the day, the average Skinner vapor pressures were 7.0 percent greater than Cotton- wood values, whereas at night they were 7.7 percent greater. WATER TEMPERATURE The annual pattern of the average water tempera- ture in the canal is shown in figure 30. Daily average water temperatures at Cottonwood and Skinner were highly correlated with a correlation coefficient of 0.999 and a standard deviation of0.28°C. There was always a slight warming trend as the water passed through the canal. The average increase in temperature was 0.02°C during January through April, 011°C during May through August, and 042°C during December. Water temperatures at Cottonwood are not available for Sep— tember through November. The water temperature was always observed to be uniform in the vertical. The diurnal variation in water temperature is illus- trated in figure 31, which shows the annual mean ob- tained for each 10-minute period of the day. These curves were obtained by averaging all available tem- 28 METEOROLOGY, HYDRAULICS, AND THERMAL REGIME, SAN DIEGO AQUE1)UCT, CALIFORNIA perature measurements at each individual time period. The mean annual diurnal range in temperature at Cot- tonwood is only 038°C as opposed to a range of 268°C at Skinner. The small diurnal variation in water tem- perature at Cottonwood results from the fact that the water entering the canal is diverted from the Colorado River Aqueduct, which passes under the San Jacinto Mountains just upstream of the entrance to the San Diego Aqueduct (fig. 1). The phase difference in the two distributions is also interesting. The Cottonwood tem— perature reached its low about midnight and its high at about 1100 hours, but the Skinner distributions reached a low at about 0730 hours and a high at about 1700 hours. DISCHARGE Discharge values were provided by the Metropolitan Water District during days of constant flow. For each of these 209 days the average stage at Cottonwood was 30I I I I I I I I I I I I m Cottonwood W 20— — U) _ W] 1 2 1 / ‘3 l“ L ~ LLI 10* f' “ O (n ”J _ _ LLI I: 8 I I I I I I I I I I I g A S O N D J F M A M J J i 1973 1974 Lu I: Z) :1 30 E fl I I I I I I I I 0- Skinner E N‘\W\ 1 I— I 20— _ E I! < a — r - ,N 10_ J' _ I I I I I I I I I I I I ASONDJFMAMJJ 1973 1974 FIGURE 30.~Daily mean water temperature, averaged over depth. determined and Manning’s n computed. During the summer and early fall, July 24, 1973, to October 9, 1973, and May 1, 1974, to July 23, 1974, the n value remained fairly constant. The mean of 114 computed n values during these times was 0.0175 and the standard deviation was 0.0004. Even the extreme values of 0.0182 obtained on May 8 and 0.0165 obtained on May 22 would result in a variation of less than i 5 percent. Starting about the first of October, the n values started steadily decreasing with time until October 18 when the n value was 0.0151. From October 18, 1973, to De- cember 26, 1973, the values again remained fairly steady. The 53 values available during the latter period averaged 0.0152 and had a standard deviation of 0.0004. During the winter and spring months, the n values varied widely in what appeared to be a random pattern, but a moving average seemed to increase more or less uniformly with time to about 0.0175. The 38 22 I I I I I I I I I ” Cottonwood — 21 * — 20 — — 18 I I I I I I I I I I I 4 8 12 16 20 24 22 I I I I I Skinner WATER TEMPERATURE, IN DEGREES CELSIUS 18 I I I I I I I I I I I 0 4 8 12 16 20 24 TIME, IN HOURS FIGURE 31.—Mean diurnal variation in water temperature. REFERENCES (Il'l‘ED 29 values available for the period of January 16 to May 1, 1973, averaged 0.0170 and had a standard deviation of. 0.0015. Extreme values of 0.0139 occurred on January 19, 1974, and 0.0191 on April 27, 1974. Local operators attribute these shifts in the n values to biological ac- tivity of the water. Apparently algae and scum have some tendency to form at certain times of the year, varying the roughness of the concrete. SUMMARY AND CONCLUSIONS Meteorologic and hydraulic variables which influ— ence the energy budget of the San Diego Aqueduct in southern California were continuously monitored for a 1-year period beginning July 24, 1973. The incoming solar and atmospheric radiation, windspeed and direc- tion, water temperature, and wet- and dry-bulb air temperature were recorded on 10-minute intervals at each end of the 26-km canal, and flow rates and stages were determined on hourly intervals at five locations. These data are available on magnetic tape and can be obtained by contacting the Automatic Data Processing Unit, US. Geological Survey, Water Resources Divi- sion, Reston, VA 22092. A detailed description of the study site, the instrumentation, and the procedures used has been given, as well as all other information necessary for the use of these data by interested per- sons. A general analysis of the spatial and temporal variability of the recorded data, as well as the daily mean values of all data, have been presented. From an analysis of these data, the following conclusions are drawn: 1. A pronounced diurnal variation in windspeed was observed at each end of the canal. Windspeeds were typically quite low during the early morning hours and at a maximum during the late afternoon. 2. Daily average Windspeeds are quite variable from point to point, apparently depending on the local to- pography. 3. Solar radiation measured at Cottonwood was highly correlated to that measured at Skinner. The standard deviation between daily average values, ob- tained 26 km apart, was 21.4 W/mz, and the mean dif- ference was 4 W/mZ. This parameter is relatively easy to measure by use of pyranometers, and an accuracy of i 2 percent can be expected. 4. Atmospheric radiation values are difficult to monitor accurately, and results obtained by use of dif- ferent instrument types are not always comparable. 5. Vapor pressure is another parameter which is dif- ficult to accurately monitor on a continuous basis, but it is fairly uniform spatially. Instantaneous meas- urements taken simultaneously but 26 km apart had a correlation coefficient of 0.837 and a standard devia- tion of 0.27 kPa. The mean difference was 0.10 kPa. 6. At a point where only the dry-bulb temperature is known, the most accurate method to estimate the vapor pressure is to compute it from the wet- and dry- bulb temperatures obtained at a remote site. 7. Flow resistance, as defined by Manning’s n, in the concrete-lined San Diego Aqueduct varied significantly with time of year. Local operators attribute this varia- tion to biological growths. REFERENCES CITED Idso, S. B., and Jackson, R. D., 1969, Thermal radiation from the atmosphere: Journal of Geophysical Research, v. 74, no. 23, p. 5397—5403. Tennessee Valley Authority, 1972, Heat and mass transfer between a water surface and the atmosphere: Norris, Tennessee, Tennes- see Valley Authority, Laboratory Report, no. 14. Thackston, E. L., and Parker, F. L., 1971, Effect of geographical location on cooling pond requirements and performance: Nashville, Tennessee, Vanderbilt University, School of En- gineering, report no. 5, 234 p. Thornthwaite, W. C., 1931, The climate of North America according to a new classification: Geographical Review, v. 21, p. 633—655. US. Geological Survey, 1970, National atlas of the United States of America: Washington, US. Government Printing Office, 417 p: GPO 689-143 RETURN EARTH SCIENEES LIBRARY Tan-b 230 Earth Sciences Bldg. (142-2997 .. . - - ,. ‘ '0, «(Ir - _.