SEDIMENTOLOGIC, GEOCHEMICAL, AND HYDROLOGIC EVOLUTION OF AN INTRACONTINENTAL, CLOSED-BASIN PLAYA (BRISTOL DRY LAKE,CA): AMODELFORPLAYADEVELOPMENTAND ITS IMPLICATIONS FOR PALEOCLIMATE. APPROVED BY SUPERVISORY COMMITTEE: Copyright by Michael Robert Rosen 1989 Tomyparentsforalltheyhavedoneformeandthesupport theyhavegiven throughout my life not just for this project. And, to all the folk musicians of Austin, in particular the O’Briens group (Jon and Nancy, Jim,Norm,Bob, Sharon, DaveandDorothy)andtheColoradoSt.Cafdgroup (Isla, Ed, Richie, Dan and Charlie), these musicians have given me the inspiration to be creative both musically and geologically. SEDIMENTOLOGIC, GEOCHEMICAL, AND HYDROLOGIC EVOLUTION OF AN INTRACONTINENTAL,CLOSED-BASINPLAYA(BRISTOLDRY LAKE,CA); AMODELFORPLAYADEVELOPMENTAND ITS IMPLICATIONSFORPALEOCLIMATE. By Michael Robert Rosen, B.S., M.S. DISSERTATION Presented to the Faculty of the Graduate School of the University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY THE UNIVERSITY OF TEXAS AT AUSTIN May, 1989 ACKNOWLEDGEMENTS I would like tothankfirstofallthethetwosaltcompanies thatoperateonBristol Dry Lake for allowing me on to their property and making special arrangements for my benefit. Inparticular,IwouldliketothankFrankDuke,HerbSantosandLyndon JonesofLeslieSaltCo.foralltheirhelpin thefield,geochemical information,and discussionsintheirair-conditioned offices.Also,IwouldliketothankTomBeegsley of National Chloride Co. for allowing access to National Chloride property. Mr. H. R. "Buster" Burris owner ofRoy's Cafd and motel in Amboy (the last of a dying breed) providedmanyleadsinmyquestforinformationmostofwhichpaidoff. Iwouldlike tothankAnnLiljeandPeteSadlerofU.C.-Riversideforallowing metolookatand Robb Bell and DaveJohnsofSouthernCaliforniaEdisonfinally trackeddownthe sample core from Bristol Dry Lake and eventually take it back to Texas. water datareport fromtheirfileswhichmadeallthecomparisons ofthewaterandisotope datapossible. LyntonLandattheUniversity ofTexas-Austinrantheoxygenandcarbonanalyses on thecarbonateconcretions. JulieKupeczranthestrontiumisotopesonthecelestite, RachelEustice camethrough atthelastminutewiththebromideanalyses, andMike Starcher prepared and ran some of the samples for clay analysis as well as providing variousotherlogisticalsupport. AndreiSama-Wojcicki,CharlesMeyer,andMike HameroftheUSGSinMenloPark, Ca.,provided theagecorrelationsoftephra layers. DiscussionsinthefieldandofficewithJohnWarren,SueHovorka,G.I. Smith, and Will Schweller very helpful in clarifying my thoughts on all aspects of were thisstudy. OtherdiscussionswithRobertHandford,SteveWellsandAltonBrown were also helpful. Iwould like to thank my brother Paul and the Jet Propulsion Laboratory,Pasadena,Ca.,forproviding theSARimageofBristolDryLake.Sally Rothwell provided enthusiastic field assistance despite the heat and conditions. I'd also liketothankDr.CharlesKeransforsome muchneededtimeofftofinishupthe project, for filling in some of the gaps left by John's departure, and for providing me with stimulating employment for my last year at UT. Field work for this project was funded by a University of Texas Research Institute grant, several grants from the R. K. DeFord field research fund from the DepartmentofGeological SciencesatU.T.-Austin, andatravel grantfromChevron, V U.S.A. Laboratory work was funded by ARCO Oil and Gas Co. Reservoir Research Group,theGeologicalSocietyofAmerica, andSigmaXi. Iwouldliketothankall thesegroupsforproviding generous supportforthisproject, especially Dr. W.J. Ebanks, Jr. ofARCO for support for this project (and me) during the summer months. Othersomewhatindirect supportforthisresearchcamefrommybrotherand sister-in-law (Paul and Lois) who provided food, lodging and sight-seeing tours of San FranciscowhileIwasvisitingtheUSGSinMenloPark. WillSchwellerprovidedthe same (except for the sight-seeing tours), as well as "the run of the house", while I was looking at core in Riverside, and Peter Ogilby (Lois' brother) provided food, lodging, sight-seeing toursofAlbuquerque, andneededrest onmywaytoandfromthefield. I'dalsolike tothankColleenBalch(andfriends) foramuchneedreevaluationofmy perspectives onlifewhilevisiting herinYosemiteNationalPark.Withoutallthese havensofrestandfun,notonlywouldthefinancialburdenhavebeen much greater, but the task of doing field work would have been far less enjoyable. I appreciate all you have had to put up with. There are many friends and collegues back in Austin which I would like to thank. I'd like to thank the Geology Librarian Dennis Trombatore; but for his diligent searches for obscure or poorly-cited references, they might never have been found, and thus would have made this work incomplete. Thanks to ToddCounsil for editing the manuscript.Myco-advisors, Dr.JohnWarren,andDr. LyntonLand,andmy committee members Dr. Gary Kocurek, Dr. R. L. (Luigi) Folk and Dr. William Schwellerhaveallbeenextremelyhelpfulinproviding guidancewhensoughtand leaving me alone to plod along as needed. Thanks for the hands-off approach. Special thanks to Dr. Land for stepping in as "co-advisor" at the last minute so that I was able time. tofinishon Finally,I’dliketothankmy"co-advisor".Dr.JohnWarrenwho initiatedtheproject. HehasalwaysbackedmeupwhenIneededhelpandhehasbeen aconstant source of inspiration in this, and other, projects we have worked on together. Through good times and bad he's always come through; I will always appreciate the support, advice, and friendship. Overthe5 yearperiod thatthisdissertationrepresents, I havebeen fortunate to meet some wonderful people who have not only helped me academically but also personally. Inparticular.I'dliketothankMike Sweetforputtingupwithmefor VI almost all 5 years as officemate and good friend. Amy Wilkerson provided a constant sourceofemotional supportforwhichIameternally grateful (why'd you haveto graduate?), and Jay Vogt was always there to lend hand. A short list of additional a friends who have helped along the include; Don Miser, Randy Farr, Wendy way Macpherson, SallyRothwell,JuliaKnight, AliceSpencer,MattandLynnParsley, FranzandJulieHiebert,JulieKupeczandJeffCopley, JanetManchester,Theresa Brown, Mike Starcher, Mike Blum, Patti and Brian Hughes, Jeff Rubin, Robert Single,andJon Blount. Thanksto all. you vii SEDIMENTOLOGIC, GEOCHEMICAL, AND HYDROLOGICEVOLUTIONOF ANINTRACONTINENTAL, CLOSED-BASIN PLAYA(BRISTOLDRY LAKE, CA): AMODELFORPLAYADEVELOPMENTAND ITS IMPLICATIONS FOR PALEOCLIMATE. Publication No., MichaelRobert Rosen, Ph.D. TheUniversityofTexasatAustin, 1989 Supervisoring Professors: John K.Warren and Lynton S. Land Bristol Dry Lake is an intermontaneclosed-basin playa located in the central Mojave DesertofCalifornia,U.S.A.Thebasinhas adrainage areaofover4000km2 andthepresentplayacoversapproximately 150km2 Themodemclimateisarid(ave. . rainfall< 100mm/year). Theplaya-centersedimentsconsistofalternating bedsof halite(up to 10m thick)andcalcareousmudsand sandwithminoramountsofgypsum andanhydrite. Twocontinuouscoresdrilledinthebasinwenttoover500mthrough beddedbrine-panhaliteandsiliciclastic mud-flatfaciesinonecore,andmud-flat - playa margin sediments in the other, and did not reach basement. Correlation of tephra layersinbothofthesecoresindicatethatthebasinisatleast4ma.,and beasold may as 10-12m.a. old. Although sedimentation is episodic and the rates may be highly variable, averagesedimentationrates areslow,generally between50-100 mm/103 years. Structural complications may locally increase the depositional rate by almost an orderofmagnitude to300-800mm/103years. At the surface, the playa exhibits a generalized bulls-eye facies distribution. Low-gradient alluvial fans ring the playa. Extensive calcrete and pedogenic calcite associated with halophyte plants cement the mid to distal fan gravels and sands. Basinwardofthedistalfans,in theplayamarginfacies, anextensive gypsum zone is over 300 m wide. Gypsum and anhydrite fabrics observed in trenches and cores exhibit a vertical alignmentofcrystals similarto thefabric seeninbottom-nucleatedbrine-pond gypsum. However, the gypsum formed in Bristol Dry Lake precipitates displacively within the VIII sedimentwheregroundwater saturatedwithrespect togypsumrecharges aroundthe playamargin. Evidencefordisplacivegrowthofgypsumcomesfrom1)thegeometry of the deposit, and 2) inclusions of matrix which follow twin planes and completely surround crystals as they grow. Within 100 m of the surface, the gypsum dehydrates to anhydrite, although the same vertically aligned fabric is retained through the diagenetic process. Because the fabrics and textures of the gypsum formed by groundwater in Bristol Dry Lake aresimilartothoseofgypsumformedinabrinepond, thecriteriafor distinguishing between subaerial and subaqueous gypsum become more complicated; the particularlyif gypsumhasbeenconvertedtoanhydrite. Whencomparedtothe features observed in a Holocene subaqueous gypsum deposit (Marion Lake, Australia) and aHolocenesubaerialdeposit(AbuDhabiSabkha),theabundanceofmatrix,the bulls-eye geometry of the deposit, the moderate gypsum crystal size and vertical orientationofthegypsumcombinedwith thedistributionofmatrix andfluidinclusions within individual crystals, and the lack of sedimentary structures in the surrounding fabric of the matrix around the gypsumdistinguishes theBristolDry Lakegypsum from theothergypsum-formingenvironments. Within the gypsum zone, celestite forms decimeter-size nodules which may coalesce into meter-size patches. Basinward of the playa margin facies, halite hopper to 0.5mindiameterforminwater-saturatedmudsofthe salinemud-flat crystals up facies. Finally, in the basin-center 0.2 m thick chevron halite crust forms from a evaporation of ponded ephemeral water. Stable carbon andoxygen isotopes ofbasin-centercalcite concretions fromthe surface and core indicate that the isotopic composition of the waters precipitating the concretions have not changed substantially throughout the evolutionof the basin. Sulfur isotopes from surface gypsum, anhydrite and celestite show no significant variation (mean 634 S = +7.7 °/00, std. dev. = ±0.7 °/oo.) However, sulfur values of This anhydrite from core are depleted relative to surface gypsum and anhydrite. isotopic depletion, coupled with petrographic evidence, indicates that bacterial sulfate reductionis involvedin thedehydrationofgypsumtoanhydrite,ratherthanevolving the isotopic composition of the brine. At the surface, the sulfur isotopic composition of gypsum shows no statistically significant trends with depth or lateral distribution. IX Deuterium and oxygen isotope values of surface gypsum hydration water indicate that the gypsum may have precipitated from evaporated groundwater with a Pleistocene isotopic signature. Chemical analyses from groundwater wells and basin-center brine from up to 150 m depth indicate that 1) the brine is dominated by Na, Cl, and Ca, 2) Na, Ca, Mg, K, and Cl increase logarithmically toward the basin-center, whereas, Si, SO4, and HCO3 decreaseslighdy (thispatternisconsistentwiththeobservedevaporitemineral distributioninthebasin), 3)molarNa:ClandCa:Clratiosdemonstratethatthesimple dissolution of previously deposited halite deposits, as has been suggested for other playabasins, cannotaccountfortheproportionsoftheseionspresentinthebrine. However, thelackofBrin thebeddedhaliteindicatesthatthehalitehas undergone significantdissolutionandreprecipitation. Water,stableisotopic,andcalculated chemicalbudget datasuggest thatapossible major sourceofSO4,Cl,HCO3, isfrom atmospheric inputthroughdirectprecipitationintothebasin. ThemajorsourceofMg, K,Mn,Fe,andLi isfromhydrolysisofunstablemineralphasesinthesurrounding bedrock. Calcium and Na probably are derived from a combination of these two sources. However,hydrothermalfluidscannotbecompletelyeliminatedasapossible for Ca, Cl, Li. source or Therepetitious natureofthealternating shallowbrinepondhaliteand siliciclastics, the consistency of the carbonate isotopic data from the surface and core, andtheagreementofwaterchemistrydatawithobservedmineralogies, indicatea relatively stablebrinecompositionformostofthehistoryofBristolDryLake. All indicate that shallow-water sedimentary structures and primary halite fabrics inthe core brine-pond halite alternated with halite-saturated siliciclastic muds in the basin-center. Adelicatebalancebetweensubsidence, andmechanicalandchemicaldepositionof mineralswas necessary tomaintainthelargely ephemeralenvironmentofdeposition through over500mofbasinfill. X The alternating brine pond/saline lake setting in Bristol Dry Lake is not directly relatedtoclimaticinfluences, andthesedimentsdonotrecordmajorclimaticevents demonstrated in other closed-basin lakes. The reason for this insensitivity to climatic the mountains eventsisexplained by theinteriorlocationofthebasin, thelowreliefof surrounding the catchment, the large surface area of the catchment, and the low average sedimentationrates. Itisimportanttoconsidertheinfluenceoftheabovefactorsin determining globalversus localorregional climatecurvesforaparticular basin. xi Table of Contents ACKNOWLEDGEMENTS v ABSTRACT viii CHAPTER 1. Introduction 1 Purpose of the study 1 GeneralGeology andHydrology ofPlayas 5 Geology 6 Hydrology 6 WaterChemistry of Playas 12 Silicate Hydrolysis 13 UptakeofCO2 andSulfate 14 Mineral Precipitation 14 Atmospheric contributions 15 Brine Evolution 15 Silicate precipitation 18 Evaporite Formation 19 Surface morphology 20 Summary 21 CHAPTER 2. General geology of the Bristol Dry Lake basin 23 Structural setting 23 General geology of the drainage area 30 General hydrology of the basin 31 Biota 36 Climate 36 Previous work 37 CHAPTER 3. Methods 40 CHAPTER 4. Depositional Environments 43 Alluvial Fan 43 Proximal fan 43 Mid fan 44 Distal fan 49 Playamargin 53 Gypsum growth and morphology 61 Celestite 66 Surface to core comparisons 68 Saline mud-flat .• 73 Calcite concretions 80 Salt pan 92 Facies distributions and stratigraphic framework 122 Summary 124 CHAPTER 5. Fabric comparison of gypsum depositional environments 125 Gypsum depositional environments 126 Marion Lake (subaqueous) 126 Bristol Dry Lake (groundwater-seepage) 129 AbuDhabiSabkha(subaerial) 131 Criteria for the recognition of gypsum depositional environments 131 XII Conversionofgypsumtoanhydrite 138 Summary 139 CHAPTER6. 140 Tephrachronology Age of the Basin 140 Structural complications 145 Sedimentation rates 146 Summary 149 CHAPTER 7. Geochemical Results 151 Sulfates 151 Carbonates 152 Halite 152 Water data 152 CHAPTER 8. Composition and origin of the brine 155 Major ions 158 167 Minor brine components Trace elements 172 Summary 181 CHAPTER9. Evolutionof the Basin 183 Paleoclimateandpossibleancientdrainagepatterns 186 Summary 195 CHAPTER 10. Conclusions 196 Appendix I Core logs 199 Appendix II Thin section point-count data 220 Appendix IB Isotope data 223 Appendix PV Waterchemistry data 229 Appendix V Tephrachronology data 241 REFERENCES 243 XIII CHAPTER 1. Introduction There are approximately 50,000 playas on Earth (Neal, 1975). Although the individual areaofplayalakesissmall(usually lessthanafewsquarekilometers)and preservation potential is assumed to be low, the study of playas is important for three major reasons. 1) With increasing population and the attraction of more people to the "Sun Belt" areas of the world, the water supply in these regions becomes greatly stressed. Knowledge of playa lake hydrology and how irrigation and other anthropogenic activities may disrupt the balance of the system is important in order to solve these problems. 2) Playa basins are generally considered to contain fairly detailedinformationonpaleoclimatic changes recordedinthesedimentologic, geochemical,andgeomorphicfeaturesofthebasin. Climaticdatawillhelpinthe recognition andreconstructionofbasinevolution. 3)Fromageologicperspective, accuraterecognitionofplayamineralogyandfaciesrelationships mayleadtothe discovery that playa sequences are not as rare in the stratigraphic record as was previously thought. Furthermore, recognition of geometries and facies relations betweenevaporites andsiliciclastic inputmayleadtoabetterunderstandingofoil­ producing areas interpreted to be playa or playa-like deposits such as the Wilkins Peak BasininWyoming(Smoot, 1983)andtheClearforkFormationofWestTexas (Handford, 1982b). Purpose of the study. The approach of this dissertation is to determine the vertical and lateral facies development of a single closed basin and to demonstrate a chemical systemconsistentwiththesedimentrecord.In thisway,therockrecordcan be used to determine the chemical controls on diagenesis and preservation. Bristol Dry Lake, located inthe central Mojave Desert, California (Fig 1.1), was chosen for this studybecause; 1)itistypicalofclosed-basinintermontaneplayas,2)itisaerially extensive(155km2)andrelativelydeep(over500mofsedimentaccumulation) sothat comparisons with ancient sequences can be justified, and 3) there is ready access to hydrologic andchemicaldataandgoodaccesstotheplayasurfaceandsubsurface geology. Several continuous cores drilled over the last 30 years around the basin- center make Bristol Dry Lake an ideal location for studying relatively long-term 1 Figure 1.1. Synthetic aperture radar (SAR) image of Bristol Trough showing present surfacerunoffarea(white oudine) toBristolDryLake(B) andAlkaliLake (A). Amboy Crater and lava flows (L) are visible separating Bristol from Alkali Dry Lake. Cadiz Dry Lake (C), which is approximately the same elevationasBristolDryLake,isalsowithintheBristolTrough. Danby DryLakeisfarthertodiesoutheastoffthepicture(seeFigure 1.2). 3 geologicalfactorswhichcontrolthehydrologicevolutionofthebasin. Inadditionto the data from the cores, surface exposures, trenches, and chemical data provide clues to overallmineralogic evolutionofthebasinwithtime. Recently, emphasis has been placed on geochemical parameters and hydrologic systems in inland evaporite settings in which the variation in evaporite mineralogies in a given basin can be constrained by fluid flow pathways and the availability of ions from the parent solution (Hardie, 1968; Hardie and Eugster, 1970; Hardie etai, 1978; Eugster and Hardie, 1978; Smith et ai, 1987; Schmid, 1988). Other studies, such as Wasson etal. (1984), haveinferredthe geochemistryofthewaterby studying coexistingmineralassemblages. Thepurposeofthisdissertationistodeterminethe basinevolutionofBristolDryLakebydocumentinghowdepositional environments, mineralogies, andisotopicdatarelatetothehydrochemical data.Theprimaryreason for this approach is that in ancient evaporite environments, lithologic data are the only information available to the geologist because hydrologic data generally do not exist. In addition to the primary objective, the availability of continuous for over core 400mprovidesthedatatostudyclimaticvariations relativelylongperiodof over a time. Datable tephra layers from several stratigraphic intervals inthe core provide age constraints on inferred climatic variations and provide a framework for interpreting possible global or regional climatic signatures in the basin. Information on paleoclimate is particularly useful for interpreting Pleistocene pluvial/interpluvial glacial cycles which are prominent depositional packages in otherCalifornia playas. Bristol Dry Lake has lately become one of the most frequently referenced modem evaporite basins from the Southwest United States (Lowenstein and Hardie, 1985; Lowenstein, 1988; Hardie, in Casas and Lowenstein, in Warren, 1989). press; press, This is because its large surface area and thickness permits comparison with large-scale ancient evaporites and its mineral assemblage is similar to marine evaporites (see Handford, 1982a,b). However, no detailedstudies which integrate sedimentologic and geochemical observations support the comparisons that are made and, in fact, the geochemistry anddiagenesisofBristol DryLake arepoorly documented. Abundant data from various sources are publicly available and address various aspects of the sedimentology (Gale, 1951; Bassett etal. 1959; Durrell, 1953 and Handford, 1982a,b) and hydrology (Thompson, 1929; Shafer, 1964; and Calzia, written comm., 1979) of Bristol Dry Lake. Despite this wealth of information, there is no integrated interpretationofthesedata.Thisstudyrepresents thefirstdetailedinterdisciplinary studyoftheBristolDry Lakeareawhichincludes thesedataandaddsnewdetailed sedimentologic, petrographic, geochemical and climatic data to evaluate the hydrologic,geochemical, mineralogic(includingdiagenetic),andclimaticevolutionof the basin through time. GeneralGeology andHydrology of Playas In order to fully appreciate the relationships proposed for the sedimentologic andhydrologic frameworkofBristolDryLake,ageneraldiscussionoftheimportant geochemical and hydrologic features, as they pertain to closed basins in general, is However, before such a discussion can begin, few terms must first be necessary. a defined. Unfortunately, the study ofevaporites, like many subdisciplines of geology, ismiredinpoorlydefinedormisusedterminology. Therefore,thefollowing definitions are given. Thedefinitionofwhatconstitutes playaissomewhatelusive. a The word playa means "beach"inSpanish,but thisdefinitionhasbeenlargely lostdue topoorusagein English with time. According to Neal (1975), playa is a "geological term for the flat andgenerally barrenlowerportions ofaridbasins ofinternaldrainage thatperiodically floodandaccumulatesediment."However, notallplayasorplayalakes(flooded playas)havethesameorigin. Therefore,Neal(1975)suggestedthatplayashould thenbeusedasageneraltermtodescribea"varietyoftopographic depressionsand desiccatedformerlakesthatoccurinthearidzone". In,California, thetermDryLake is usedtodescribethethousandsofintermontanebasins foundintheBasinandRange area (i.e. Bristol Dry Lake). The term is synonymous with playa. In addition to playa, the terms salina and sabkha have also been used to describe evaporitic basins (Warren and Kendall, 1985). Although these latter terms can be restricted to marginal marine settings, recently they have taken on a more general meaning. A thorough discussion ofthesabkhanomenclatureproblemisgivenbyCastens-Seidell(1984). Sheredefined thecriteriausedtorecognize asabkhadepositandcontendedthatthetermsabkha should be used only for marginal marine settings. In addition, she suggested that, becauseoftheshoaling-upward natureofthesetting, halitewouldnotbepartofa normal sabkha sediment package. This definition essentially precludes Bristol Dry Lakeandmostotherintermontaneevaporitic lakes frombeing termedsabkhas. Finally, amorequantitative approach wastakenby Hardieetal.(1978)for perennial intracontinental water bodies. They define saline lake a as a lake with greater than 5000 dissolved salts. ppm In this dissertation, playa, playa lake, and dry lake are used interchangeably for saline mineral-bearing intracontinental hydrologically closed-basins that do not have a perennialwaterbody,andsalinelakeisusedasdefinedbyHardieetal.(1978). The terms sabkha and salina are used solely to describe marginal marine settings such as the sabkhasoftheArabianGulf(Curtisetal.,1963;Kinsman, 1965;Butler,1969;etc.) and the salinas of South Australia (Warren, 1982a,b). Geology. To form a saline or playa lake basin, two conditions must be met. First,evaporationmustexceedevaporationplusinflow,andsecond, thebasinshould be hydrologically closed or greatly restricted (Hardie et al., 1978). Restriction of the basinmaybecausedbymanythings. However,themostfavorableconditionsforthe formationofsalinelakesareinrain-shadowbasins. Suchbasinsprovidethehigh reliefaroundthebasinnecessary totrapprecipitation,whilethebasinfloorremains arid (Eugster and Hardie, 1978). Bristol Dry Lake is located in the east-central Mojave DesertofCaliforniawherethisprocessiscommon. Salinelakesandplayasinthis area are located in the depressions between block-faulted mountainranges and Neogenevolcanics. Thus,intheMojaveDesert,closedbasinswhichreceive relatively little rainfall are the rule, rather than the exception. Playas and saline lakes are common at the deepest portions of the basins, with coalesced alluvial fans ringing the basin margin (Fig. 1.2). Hydrology. Although the interrelationship of siliciclastic inflow from alluvial fans and evaporitic facies is poorly understood, some hydrologic generalities canbemadefromanexaminationoftheliterature.Recently,Duffy andAl-Hassan (1988)modeledthegroundwatercirculationofaclosed-basinplayainNevada. They determinedthatthetopographyofthesurrounding mountainsis amajorinfluenceon In addition, they determined by numerical the hydrology and climate of the basin. Figure 1.2. SAR image of the eastern Bristol Trough (see Chapter 2). Theback and dark areas are coalesced alluvial fans. Areas in white relief are the gray surroundingmountainrangestrendingnorthwesttosoutheast. Thefour playas are isolated by various obstructions. Alkali Dry Lake (A) is separated from Bristol Dry Lake (B) by Amboy Crater lava flows (L). Bristol Dry Lake and Cadiz Dry Lake (C) are separated by a low alluvial divide, and Cadiz Dry Lake and Danby Dry Lake (D) are separated by a projecting mountainrange. 8 modeling (which agrees with field evidence) that a free convection cell within the discharge zoneoftheplayarecirculatesbrinefromundertheplayatothebasin margin. This free convection cell may account for the deficit of precipitated salts insomebasinsbecauserecirculation storesthesaltsinthesubsurface. Thismodelalso has implications for the chemical budget of Bristol Dry Lake and will be discussed later. Figure 1.3,takenfromMifflin(1968),illustratesahypotheticalgroundwater flow system through a typical Great Basin (Nevada area) cross-section. This diagram shows essentially an unconfined "water table" aquifer in the of lateral flow. or zone Thistypeofflownetwouldbetypicalofshallow, relativelypermeable, sequences around the zone of active discharge (the playa basin). In the zone of groundwater artesianwellsresultinNevadabecausethe ofsaturationis discharge, flowing or zone Based observed Sketch of on Figure 1.3. a hypothetical Great Basin flow system. relationships (from Mifflin, 1968). 10 near the surface (Mifflin, 1968). In nature, the artesian character of this region is demonstratedbynaturalseepsandtufaaccumulations(because ofdegassing)around the perimeter of many playa basins; the most spectacular occurrence being the 12 meter hightufamoundsinMonoLake, California. Mifflin (1968) also introduced the concept of terrane flow capacity for Basin and Range type hydrologic systems. The amountof water any particular hydrologic environment could accept and transmit defines the terrane flow capacity. This is a somewhatnebulousconceptuntiltheterranecapacity (Qt)isdirectlyrelatedtoDarcy's Law: = -KA AH Qt (1.1) AL whereQis the terrane K is a capacity measure ofregional permeability in the directionofflow,AH/ALis theregional hydraulic gradient, andAisthecross-sectional areanormaltoflowdirection,which benormaltotheslopeoftheterraneona t, may local basis. Theregional parameters which control Qt are average slope of terrane and average permeability. Should enough moisture for maximum recharge be available, thesetwo variablesdeterminetheamountoftransmittablegroundwaterbeforerejection by discharge from the terrane. Figure 1.4 illustrates capacity ofterrane for three situations typical in Basin and Range settings. This concept may be helpful in determining geochemical sources or gradients for observed precipitated mineral sequences and brine compositions. In other words, does the composition of a particular measuredbrinereflectmajororminorinteractionwiththehostrockor alluvial sediments? Lithology plays an important role in defining the configuration and distribution of flow system boundaries. For example, in an area where there is sufficient precipitation, a high hydraulic gradient may be established in low permeability rocks. High hydraulic gradients result in an intersection of the zone of saturation with the land surface, and local discharge may develop and form perennial streams, seeps, or tufa in the Eocene Green River Formation (Wilkins springs. Thus, thepresenceof PeakMember)mayindicatearelatively highhydraulicgradientinarelatively low­ permeabilitysettingor, asmentionedabove,itmaysimplyindicatetheproximityofthe playa and the zone of discharge. Qmax > Qpresent - Bedrock : Qmax Qpresent Basin : Qmax > Qpresent - Qmax Qpresent Figure 1.4. Diagrammatic sketches of the concept of tenranecapacity for regional groundwater flow (from Mifflin, 1968). Theimportanceofsurfacerunoffcannotbe neglectedwhenconsidering the majorcomponentsofflowtothethe dischargeareaofplayabasins. Rainfallisscarce in playa regions, but when it does rain, it rains hard. Flash flooding and sheet wash are common aspects of playa basin sedimentation. Great influxes ofmeteoric water enter the basin bearing a large siliciclastic suspended load (clay-size particles) which hashadrelatively littletimetoacquire solutesfromthesurrounding alluvium orhost rock. This will in effect halt evaporite precipitation by dilution, and may induce a period of dissolution, diagenesis, and karst within the more soluble playa sediments. Eventually, theplayawillreachthepre-runoffstageofchemicalprecipitation,provided thatnolargeadditionalinfluxofmeteoricwaterhasoccurred. Thus,alternationof chemical precipitates and siliciclastic clays may be common inthe sediment column. Bristol Dry Lake illustrates this alternation in halite that is over 500 - a mud sequence meters thick. Water Chemistry of Playas. The evolution and chemical composition of brine solutions in closed-basin settings are probably the best-studied aspects of saline lakes. Initially,Jones(1966)discussedtheevolutionofclosed-basinbrineswithinthe contextof Western Great Basin (Basin and Range Province) saline lakes. Jones and Van Denburgh (1966), Hardie and Eugster (1970), and Eugster and Hardie (1978) expanded thebasic premiseofJones(1966) andappliedittoallclosed-basinsystems. Harditetal..(1978)summarizedpreviousworkwhileemphasizing asedimentologic framework. ChemistryofInflow Water Jones(1966)wasthefirsttorecognize thatthecompositionofsalinewaterin closed, inlandbasins isinheritedfrom thechemicalweatheringprocessesofthe drainagebasin. Inotherwords,thechemistryofthefluidscomingintothebasinisa product of the original mineralogy of the surrounding rocks. However, significant quantitiesofsolutescan bederivedfrom theoriginalrainwaterinputinto thebasin. Evaporationoftheincomingwaterinthebasin leadstoselectivemineralprecipitation. functionof solute supply and In simple terms, the genesis of saline waters is a evaporative concentration.Theevolutionofthewatercanthenbedescribedby is solution, transport, and mineral precipitation reactions in sequence. This sequence controlledbybasinhydrology. Asmentionedabove,primarysolutecompositionis determinedmainlyfromreactionofnaturalwaterswithlithologies underlyingand surrounding the drainage basin. However, secondary modifications, such as the two mixingof chemically differentinflowwaters,tothesolutecomposition are controlledbyhydrologicsettingandprocess. Classificationofbrinetypewasbasedon recognition of major anions in solution because they were thought to have a more diverse origin than the cations (see Jones, 1966, and Hardie and Eugster, 1970). Intheabsenceofolderchemicallyprecipitated deposits,watercompositions are aproductof: 1)silicatehydrolysis,2)uptakeofC0fromtheatmosphereand/or 2 sulfatefromoxidizedsulfides, 3)precipitationofalkalineearthcompounds,and4)the originalcontributionofsolutesfromrainwaterandaerosols. Ifolderchemical precipitates (evaporites or carbonates) or clays are present in the drainage basin, simple solutionandleaching ofadsorbedions maycontribute to theinitialbrine composition. The following paragraphs summarize the four major processes outline above. Silicate Hydrolysis. The quantitative proportions of igneous and sedimentary rocks incontactwithbasinwatersandthestability ofindividualsilicate phases comprising the rocks determines the effectiveness of hydrolysis. The most commonmethodofhydrolysis involvesthereactionofcarbondioxideandwaterto form carbonic acid. The reaction of cabonic acid with primary silicate minerals providesbicarbonateanionsinproportion tothereleaseofmetalliccationsandthe solutionofsilica. Thisgeneralreactioncanbewrittenas: primarysilicate+H2o(aq)+CC>2(aq)=Clay+cations+HCo3(aq)+H4Sio4(aq) (1.2) - Alkalifeldspars(Na,K feldspars) aregenerallythemostabundantprimarysilicates, and kaolinite a common clay product. In the western Great Basin, andesine (AJI33) is the common feldspar and montmorillonite is the clay weathering product. ThoughmostoftheCO2neededforhydrolysis reactions canbederived within the soil zone (which has a significantly higher PCO2 due to biotic activity), hydrolysisreactionscangetrathercomplicated(see Jones, 1966). Hydrothermal processes or oxidation and hydration of sulfides may become an important factor in areasthatarehighlymineralized. Insummary,silicatehydrolysis byeithercarbonic acid orsulfuric acid(from sulfides) is aformoflowtemperature alterationinvolving the substitution of H+ in silicate phases that are unstable during weathering processes. Uptake of C0 and Sulfate. 2 , An important secondary mechanism for modifyingclosedbasinwatersistheadditionofCO2byorganicrespiration anddecay. Decay at lake bottoms, or interstitially within lake sediments, contributes the most CO2 to the system. This process also gives rise to abundant loss of sulfate through bacterial reductiontohydrogensulfide(H2S). Theeffectofsulfatereductioncanbe in seen trendsofanioncompositionsthatentercertainlakesinthewesternGreatBasin. The tendencyforsulfatepercentageof watertoincreaseisabruptlyhaltedasthe the inflow water enters the standing body of water in the lake proper and sulfate reduction begins. Thismay significantlyalterexpectedaccumulationsofchemicallyprecipitatedsulfates. Basedontheprocessesmentionedabove, asmoresalinelakeswerestudied, and as chemical reactions became better understood, it became possible to use chemical weathering reactionequations todeterminetheorigin ofthemajor ions insolutionin closed basin brines. Eugster and Hardie (1978) presented a review of how the major ions foundin solution are derived. Mineral Precipitation,Precipitation. The final reactive process which modifies closed basinwatersistheprecipitationofsolidmineralphases. Thecompositionofinflow watersintoaclosedbasinisgreatlyinfluencedbymineralprecipitation ascertainions areselectivelyremovedfromsolution. Fourmajorprocessescontrolconcentrationand mineral precipitation in saline lakes: 1) evaporation, 2) loss of gases (especially CO2), 3) mixing of different water compositions, and 4) temperature. Of these four processes,evaporativeconcentrationdominatesinsalinelakes. Asordinarymeteoric water is evaporated and concentrated, alkaline earth carbonates (mainly calcite, aragonite,Mg-calcite) arethefirstmineralsprecipitated. Theextentofcalciumremoval from solution and the point of initial precipitation are controlled by the relative importance oflimestone solution and silicate hydrolysis in the derivation of the primary solute. However, in heavily mineralized areas such as in Saline Valley, California, carbonateprecipitation maybeovershadowedby primarysulfate(gypsum) precipitation. Becausethereis alargedifferenceinsolubilitybetweenalkaliearthcarbonates, sulfates, andalkali salts, changes inanionconcentrationofsalinelakesintermediatein composition tochloridebrinesandmeteoricwater aremainlyattributedtohydrologic andorganiceffects. Netevaporation,depthandtotalareaofthelake,andareal variationwithtimemaycontrol thetotalconcentrationof many closed lakes (Langbein, 1961). The final processtoaffectcompositionofthebasinalbrineis theprecipitation of salines (i.e. halite, sylvite and other bittern salts). This process may occur in standing waterbodiesorinterstitially withinthebottomsediment. Atmosphericcontributions. Thecontributionofionsfromrainwaterand aerosols has generally been considered to be minor in relation to the above mentioned sources. However,ifthesedimentationrateisslow,significantamountsofions, particularly those that are difficult to generate from chemical weathering, such as Cl, may be accounted for by atmospheric input The potential importance of atmospheric input cannot be overemphasized. Recently, hydrothermal sources have been postulated to produce Ca-Na-Cl brines similar to the brine composition of Bristol Dry Lake (Hardie,inpress). Balancingofatmosphericinputmayindicatethattheexistenceof deep hydrothermal brines is not necessary toexplain these "anomalous" brine compositions, (seesectiononChemistryofBristolDryLakeBrine). Brine Evolution. From the above relationships, chemical analyses of lake brines, computer modeling, and simulated reactions, Eugster and Hardie (1978) and Hardieetal.(1978)wereabletoconstructaflowchartbasedon threeelementsofbrine composition which can be used to predict the final brine type of the lake and the minerals it should precipitate (Figure 1.5). Depending on the relative proportions of Mg, Ca, and HCO3 in the initial water, five major brine types can be distinguished. Table1.1lists theassociated mineralsthatshouldprecipitatewithineachbrinetype. Theprecipitationkineticsofalkalineearthcarbonates(calcite, Mg-calcite, aragonite, dolomite, hydromagnesite, magnesite, huntite) are not fully understood because most of these minerals initially form metastably. However, studies of the equilibrium precipitation of calcite and Mg-calcite demonstrate, in a general way, how precipitation of carbonate modifies the potential brine composition. Twoconditionsmustbemetsimultaneously forequilibriumcalcite precipitation (HardieandEugster, 1970): 1)calciumcationsandcarbonateanionsmustbelostfrom must solutioninequalproportions, and2)theionactivityproduct(IAP)ofthesolution be constant at constant total pressure and temperature. The first condition suggests that Figure 1.5. Flow chart for closed basin brine evolution based on computer modelling of brines and field examples. After Eugster and Hardie (1978). theproportionsofCaandCO3insolutionwillchange ascalciteprecipitatesbecauseit isunlikely thatbothspecieswillinitially bepresentinsolutioninthesameabundance. ThesecondconditionsuggeststhatincreaseinCawillleadtoadecreaseinCO3 an because, as the IAP is defined, there must be an antithetic relation between the two ions species. This shows that early calcite precipitation is a critical step in brine evolution because itwill determinewhetherthebrinewill becomecarbonate-richor carbonate- poor. Because all dilute waters (except those derived from ultrabasic rocks) that flow intointracontinentalbasinshaveMg/Caratioslessthan 1,thecalciteformedhaslittle MgCOs (<5 mole percent) in solid solution. As calcite is precipitated, the Mg/Ca will rise, so that the amountofMgincorporated intosubsequently precipitated calcitewill of solution progressively increase. For kinetic reasons, aragonite may precipitate out insteadofcalcite,buttheeffectonbrineevolutionisessentially thesame. At this point, the three paths depicted in Figure 1.5 are the only possibilities for the evolutionofthebrinewater.Ifthewaterinitially hasahigh(>3)HCOs/Ca+Mg Table1.1.ThemajorevaporitemineralsofthedifferentbrinetypesinFigure 1.5.After Eugster and Hardie (1978). Brine type Saline minerals Ca-Mg-Na-(K)-Cl Antarcticite CaCl2-6H20 Bischofite MgCl2-6H2(D Camallite KCl-MgCl2-6H20 Halite NaCl Sylvite KC1 Tachyhydrite CaCI2-2MgCl2­12H20 Na-(Ca)-S04-Cl Gypsum CaS04-2H20 Glauberite CaS04NaS04 Halite NaCl Mirabilite NaSO4 10H2O Thenardite NaSG4 Mg-Na-(Ca)-S04-Cl Bischofite MgCl2-6H2G Bloedite Na2S04-MgS044H20 Epsomite MgS04 -7H20 Glauberite CaS04 NaS04 Gypsum CaS04-2H20 Halite NaCl Hexahydrite MgS04-6H20 Kieserite MgS04-H20 Mirabilite NaSO4 10H2O Thenardite NaSG4 Na-C03-Cl Halite NaCl Nahcolite NaHCG3 Natron Na2CO3 10H2O Thermonatrite Na2C03 H20 Trona NaHC03-Na2C03.2H20 Na-C03-S04-Cl Burkeite Na2C03 -2Na2 S04 Halite NaCl Mirabilite NaSO4-10H2O Nahcohte NaHCG3 Natron Na2CO3 10H2O Thenardite NaS04 Thermonatrite Na2C03-H2G molar ratio (path I), alkaline earth elements are removed from solution rapidly with evaporation, and magnesium enrichment is insufficient to produce significant quantities ofMg-carbonates such asdolomite.ThefinalbrineproductisaNa-C03-CIbrine whichmayormaynotcontainsignificantS04.However,ifthe +Mgmolar ratioislow(pathII),HCO3isremovedrapidly andalkalineearthelements areenriched in solution. Theresult, though, is similar topath I because thebrine produced also is notlikelytoproduceMg-carbonates. ThisisbecauseoftherapidlossofHCO3and thelossofMgfromsolutionbytheprecipitationofMg-silicates. Thefinalbrine produced, then, is a Ca-Na-SCU-Cl solution. Finally, whentheHCOs/Ca+Mgmolarratioisaboutequal (pathID), magnesiumenrichmentissufficienttoproducedolomiteandpossiblymagnesite. From thispoint, thebrinebecomeseitherpoorinalkalineearthelementsand richinHCO3, vice or versa. The alkaline earth-poor solutions will eventually produce path I type brines (aided by sulfate reduction) while, the alkaline-rich solutions will produce various Mg-Cl solutions plus or minus Na-Ca-S04It should be noted that, though . dolomiteshouldprecipitate, kinetic factorsmakethenucleationofdolomitefrom solution difficult. Deep Springs lake, California and Salt FlatPlaya, West Texas, are the only documentedoccurrences ofHoloceneintracontinentalplayadolomite. Dolomiteisgenerallyconsideredasadiageneticmineralandmayreplace acalcite(or aragonite) precursor at any time in the history of a deposit. Thus, it is critical to distinguishbetweenearlypenecontemporaneous dolomiteandlaterreplacement dolomite, but this distinction is not always simple. Silicate precipitation. Because the initial concentrations of A 1 and Fe are lowinthewatersflowing intointracontinentalbasins,onlyMgandNasilicates (sepiolite, talc, magadiite, etc.) needbe considered as significant authigenic phases. Precipitation of silicates is important as a mass balance consideration because they remove both silica and magnesium from solution. Saline Valley, California, for example, precipitates sepiolite as an authigenic mineral(Hardie, 1968). However, precipitationofsilicates cannotaccountforlargesilica deficienciesmeasuredinclosed basin brines (Eugster and Jones, 1977). For example, Jones et al. (1977) calculated thesilicabudgetforLakeMagadi,Kenya. Theydeterminedthatgreaterthan99%of thesilicaislostbetweeninflowandthegroundwaterreservoir. Fromthis,they areas may be an agent ofsolute postulated that wetting and drying in the recharge fractionation. Eugster and Hardie (1978) postulated that silica loss is due to the formationof These a crust (either surface or intrasediment) by capillary evaporation. crusts would be formed by the complete evaporation of interstitial water. Salts, alkali earthcarbonatessulfatesandopalwouldformthecrusts. Subsequentdissolutionof the more soluble constituents (i.e. everything but the opal) during wetting would leave less soluble minerals like opal in theThough this method has been documented crusts. experimentally, natural examples of opal deposits in playa lakes have not been found. Silicawithdrawalbydiatoms seemstobeamorelikelymethodofsequesteringsilica from solution, but, accumulations of diatoms are not always seen in the lake sediments. Selectiveremovalofsilicafromincomingwatershasnotbeenadequately explained, andremains oneofthesignificant problemsinmodelling evaporative processes. Potassium is also removed from solution during evaporative concentration. Removalofpotassium isprobablyduetoexchangereactionsonclayminerals,volcanic glass, and alkali alumina silicate gels (Eugster and Hardie, 1978). Evaporite Formation. Saline minerals can form in all parts of the basin and r recharge area, although the ultimate significant accumulations of saline minerals will be inthedeepestpartofthebasin. Salinemineralsformas1)surfaceefflorescentcrusts, either as the final stage ofevaporation of a standing body of water or by evaporation of capillaryzoneorgroundwaterdischargeon,orneartheplayasurface. Themineralogy and topography of the playa surface reflect the hydrology of the discharge area and can be morphologically complex (see below). 2) Subsurface intrasedimentary growth of salinemineralsfrominterstitialbrines, becauseofevaporativeconcentration, resultsin displaciveandpoikiliticgrowthwhichmayconcentrateintolayersornodules. 3) Single crystals or aggregates may form on the surface of a standing brine body (i.e. halite rafts), or at any depth inthe water column, and settle to the bottom. Subaqueous growth of selenite crystals (gypsum > Icm) on the substrate up into the water column has been documented in coastal salinas (Warren, 1982a; Castens-Seidell, 1984), but not in Basin and Range type playa lakes. 4) Clasts of saline minerals deposited in one mechanical partofthebasinandredepositedinanotherpart(intraclasts)represent a means of accumulating saline minerals. Detrital input, usually carbonates, from the surrounding highlands (extraclasts) may also be important (i.e. Deep Springs Lake, California). 5)Deflationoftheplayasurfaceduringrelativelydryperiodsmay result in theformationofgypsumand/orclaypelletdunes(i.e. SaltFlatPlaya,WestTexasand WillandraLakessystem,Australia). Although thesedunesmaynotbeaspotentially preservable as the other methods ofsaline accumulation, stabilization by plant cover may preserve some dune morphology. Surface morphology. The surface of playa basins can be flat or irregular. Smooth surfaces are usually dry. The flatness of the surface is caused by deflation whicherodesthe sedimentdowntothecapillary zonewherewaterholdstheminerals inplace. Anotherwaytodevelopaflatsurfaceistofloodtheplayawhichdissolves theoldcrustandthendepositclasticmudontop. Aswaterevaporatesanothercrust will form, but, if the water table stays high and keeps the surface wet, a flat surface remain. may Where weak seeps discharge into the playa itself or where evaporating water is drawnupto theplaya surface,thick crustsofsalts andcarbonateforman irregular topography on the playa surface. Pustular algal mats associated with extrusive salt crusts and carbonates also form in the wetter areas of the playa irregularly dispersed over theplaya surface. Extrusionof displacively growing salts can also occur around the margins of desiccation crusts and along tepee structures (lateral growth expansion features). It is obvious, then, that the surface morphology of the playa surface is controlled by the hydrology of groundwater system under the playa. As the groundwater table rises and falls due to availability ofrecharge, new seeps will develop andoldoneswilldryup. Similarly, assedimentsaccumulateinonearea,watermay pond in another Over a significant amount of time, many different areas of the area. playa will have seen different hydrologic conditions. However, it should be stressed thatthe lateralfacies(which arecontrolledby hydrology) maynot bedistributed verticallyinthesame sequence. Theinherentanisotropy ofthesystem shouldbe expected because of the diversity of permeabilities associated with the lateral and verticalalterationofrelativelyinsolublesiliciclasticmudandclay vs.soluble(and permeable) saline minerals. Prediction of a vertical facies model is complicated by the above features. Superimposed on this are diagenetic effects which may further alter the original textures. Playalakeandsalinelakestratigraphyshouldthenbethoughtofintermsof a facies mosaic in which lateral and vertical hydrologic effects must be taken into account. Summary Salinelakesandplayas areformedinhydrologically closedorgreadyrestricted basinswhenevaporation exceedsinflow. Theregional hydrology of can be the basin modelledasanunconfinedaquifersystem. PlayasandSalinelakesoccurinthezone of discharge along with seeps and springs which may mark the edge of the evaporitic basin.Theterranecapacity(Qt),definedtheamountofwateranyparticular as hydrologic environment could accept and transmit, is an extension of Darcy's law to a regionalscale, andiscontrolledbytheslopeandpermeabilityoftheterrane. Lithology isalso importantindefiningtheconfiguration anddistributionofflow system boundaries,though,surfacerunoffmaybeequally asimportantinthesesystemsas groundwater. Thecompositionofclosed-basinsalinewateris inheritedfromtheoriginal solute input from rain water, chemical weathering processes of the drainage basin, and fromthelithologyoftherockssurrounding thebasin. Intheabsenceofolder chemically precipitated deposits, saline water compositions are a product of: 1) silicate hydrolysis, 2) uptake of C02 from the atmosphere and/or sulfate from oxidized sulfides, 3) precipitation of alkaline earth compounds, and 4) atmospheric input. Depending on therelative proportionsofMg, Ca,andHCO3,intheinitial water composition, five major brine types can be distinguished that will control the sequence of minerals that will form in the basin. Initial carbonate reactions will set the compositionofthesolutionsothatthe evaporative pathandthesequenceofmineral arefixedearly inthehistoryofthebrineevolution. precipitation Salinemineralsforminallpartsofthebasinandrecharge area,although significant accumulations of saline minerals will be in the deepest part of the basin. Saline minerals form as 1) surface efflorescent crusts, 2) subsurface intrasedimentary displacive crystals, layers, or nodules, 3) direct precipitates from a standing water body, 4) intra-or extraclasts, and 5) aeolian processes. The surface morphology and subsequent diagenesis ofplaya sediments is directly related to the hydrology of the basin. It should be stressed, then, that facies lesser determinedby relationships in playa lakes (and to a extent saline lakes) are hydrology, and not necessarily by Walther’s Law. Thus, the stratigraphic record of a playa sequence is areflection of the facies mosaic controlled by a dynamic groundwater system. CHAPTER 2. General geology of the Bristol Dry Lake basin Structural setting. Bristol Dry Lake is situated in the Mojave Desert region ofsoutheasternSanBernardinoCounty, California(Fig. 2.1). It is the largest (155 km 2) in a system of three northwest southeast trending dry lakes (playas) located in a - structural troughbetweenmountainrangesintheBasinandRange physiographic province of North America. Bristol Dry Lake is situated on the eastern edge of a structuralprovincewithintheBasinandRange termedtheMojave block(Garfunkel, 1974). TheMojaveblockisdefinedbytheGarlockfaulttothenorth,apartoftheSan AndreasfaultandtheeasternTransverseRanges tothewestandsouth,and asouthern projection of the Soda-Avawatz fault zone to the east (Fig. 2.2), however, this eastern boundaryissomewhatpoorlydefined(Garfunkel, 1974).Thebasinandrange topography of the area was caused by thecounterclockwise rotationoftheMojave block byright-slip movementalong northwesttrendingfaults during LateCenozoic time(Garfunkel, 1974).Strike-slipmovementalongfaultsurfacesintheareawasfirst proposed byDibblee(1961) andlatermodeledby Garfunkel(1974). However, Glazner (1988) observed cross-cutting slickensides in two faults around Ludlow, California.Hesuggested thatapossibleearlyextensionaldip-slipmovementinthe Miocene overprinted the strike-slip movement associated with the San Andreas Fault System andtheopening oftheGulfofCaliforniain theLateMiocene.Recently, Glazneretal.(1989) havefoundnew evidenceeastofBarstowforsignificantearly Miocene (19-18 m.y.a) extension of the Mojave Block. Regardless ofwhether the movementisdip-sliporstrike-slip, theparallel alignmentoftheMojave blockinto northwest-southeast trending basins and mountains is due to rotation of the block as a is wholeandnotduetosimpleextensionalpull-apartintohorstsandgrabens as typically assumedforrift systems(Fig.2.3). Bristol Dry Lake is situated in the southernextensionoftheBarstow-Bristol Trough termed the Bristol-Cadiz-Danby Trough by Bassett and Kupfer (1964, p. 41). TheBarstow-BristolTroughisapproximately 180kmlongandextendsfroman unknownpointnorthofBarstowtobeyondBristolDryLake(Gardner, 1980).The 23 Figure2.1.LocationmapofBristolDryLakeshowing trenchesandwelllocations. Filled circles are water well data. Filled stars are wells with lithic logs and waterdata(1979,U.S.G.S.wells). Thegroundwaterdrainagedivideand saline-fresh water interface are from Shafer (1964). The data for these two lines is sparse and open to interpretation. There is no groundwater drainage divide on the west side of the playa due to lack of data. Circled crosses are lithic logs and/or core data. Numbers on wells correspond to names in appendix IV and in the text. Bristol Dry Lake is located between Wlls° 50'andWll5035'longitude andN34°20'andN34°35'latitude. 25 Figure2.2. RegionalframeworkoftheMojaveBlockwithintheGreatBasin. Bristol DryLake(B)islocatedjustwestoftheSoda-Avawatzfault. TheMojave Blockis definedby theGarlockfaulttothenorth,apartoftheSanAndreas fault and the eastern Transverse Ranges to the west and south, and a southern projection of the Soda-Avawatz fault zone to the east, however, thiseasternboundaryissomewhatpoorly defined(Modifiedfrom Garfunkel, 1974). 27 Holocene strains and slips in the southwest U.S. and Figure 2.3. Major Neogene tonorthwestMexico. Thickarrowsshowmovementsofmajorunitsrelative totheColoradoPlateau. Openarrowsshowregionsanddirectionsof crustal spreading. Circular demonstrate the counterclockwise arrows rotationofblocksorfaults. Smallperpendicularopposingarrowsrepresent areasofanddirectionsofcrustalshortening. Smallparallelopposing arrows illustrate shearing or slip. Bristol Dry Lake (B) is located within the counterclockwise rotation of the Mojave Block (Modified from Garfunkel, 1974). Ref- MainNeogenetoHolocene strainsandslipsinsouthwesternUnitedStares andadjacentregions. = Colorado Plateau. Rigid areas = Colorado Plateau, Sierra Nevada, Pacific Ocean floor southwest of continental borderland. Thick arrows = of major units with to Colorado Plateau. Open arrows erence area movements respect = of blocks arrows regions and directions of crustal spreading. Circular arrows = or faults. Small opposing rotation arrows =regionsanddirectionsofcrustalshortening.Smallparallel =shearingorslip. trough isboundedbytheupliftoftheBristol andCadizMountainsto thenortheastand theNewberryandBullionMountainstothesouthwestalong majorfault zones (Gardner, 1980). Hadley andKanamori(1977) determinedthatthecrusteastofBarstow-Bristol trough hasaseismicvelocityof6.2cm/secandtothewestisdominatedby a6.7 cm/sec seismic velocity layer. They suggested a plate boundary at depth which corresponds withthewesternboundaryoftheBarstow-BristolTrough(Gardner, 1980). HadleyandKanamori(1977)furthersuggestedthattheirdataindicateda possible connectionto theoffshoreMurrayFractureZonewithatrenchtransfer junctionandazoneofdecoupling. Gardner(1980)tooktheserelations one step furtherandsuggestedthatthe Barstow-BristolTroughrepresents amajorbreakinthe earth'scrustandmayextendintotheuppermantle. Suchrelationshipsindicatethatthe trough has had a relatively long history, and that tectonics may play an important role in theboththe short andlong-term evolutionofBristol Dry Lake. Extrusive volcanics alsoplay animportantrolein theevolutionof the Barstow-BristolTrough. Arelativelyyoungcindercone,AmboyCrater,anditsassociated basalt flows, is situated in the northwest portion of the Bristol Dry Lake Basin (Fig 2.1). AmboyCrateristhoughttobelessthan6,000yearsold(Parker,1963), althoughnoradiometricdatinghasbeenconductedtoconfirmthishypothesis. Amboy Crater is one of several similar cinder cones that line up along the northwest-southeast trendofthefaultswithinthebasin.Theserelatively recentbasaltflowsandcinder cones arepresumably relatedtorifting intheBasinandRangeandmayprovide ashort­source. Intrusive igneous rocks also play a role in lived hydrothermal heat may providing hydrothermal fluids to the basin (Mohorich, 1980), however, at this time, there are no subsurface data available on such sources which can be easily evaluated. General geology of the drainage area. The drainage into Bristol Dry This makestheevaluationof Lake passes through a diverse variety of rock types. ultimatesourceofionicspeciesofthebrinemoredifficult. TheBristolandMarble Mountains,which boundBristolDryLaketo thenorthandeast, are composed mostly ofJurassicgranodiorite andquartzmonzonitebutare associatedwithmetamorphosed Paleozoicsedimentaryrocksincludingsomemarbleanddolostones(Gale, 1951). Carbonatesappeartobeabsentelsewhereinthedrainagearea(Bassettetal., 1959). 31 TothenorthwestoftheBristol Dry Lakebasin. Tertiarylavasandpyroclastic sediments are common. Volcanic rocks are also important east of the basin, but are absent to the south where plutonic and metamorphic rocks are most common. AccordingtoBassettetal.(1959) nosaline-bearing sedimentscropoutwithinthe drainageareaofBristol, Cadiz,orDanbyDrylakes. General hydrology of the basin. Bristol Dry Lake is separated from Cadiz of and Danby dry lakes by projecting arms the Marble and Calumet mountainranges thatdefinethetrendofthebasinasawhole. Atpresent,eachofthethreesub-basins have completely separate internal drainage. However, the alluvial divide separating the BristolandCadizbasins isatpresentonlyabout15mhigherthanthelowestpointin BristolDryLake.Ithasbeensuggested thatwaterhasflowedfrom onebasinto the other during pre-Holocene time (Shafer, 1964). This has important implications for the massbalanceofionsreachingBristolDryLake. Recentwaterchemistry dataand observed mineral assemblages support this hypothesis. Unfortunately, at the present time,therearenotenough datatofullyassesstheinputfromtheCadizDryLake drainage basininto theBristol DryLakeoverthegeologicpast A fourth basin. Alkali Dry Lake (see Fig. 1.1), is separated from Bristol Dry Lake by Amboy crater and its associated basalt flows. Before the flows blocked off the northern end of the drainage area possibly less than 6,000 years ago, creating Alkali Dry Lake, this area also drained into Bristol Dry Lake. Therefore, for of the most historyoftheBristolDryLakebasin,thetotaldrainage areaintotheplayaandthebasin itself were significantly greater than they are today. Even withthis reduced drainage area,thepresenttotaldrainageareaintoBristolDryLake,excluding drainagefrom CadizDryLakeandAlkaliDry Lake, isapproximately 4000km2. Shafer(1964)collectedchemical groundwater dataandwatertabledatafroma numberofwellsintheCadiz-BristolBasins(Fig.2.1). Althoughhisdataissparsein crucialplaces, thegeneral surfaceofthewatertablehasbeendefined(Fig.2.4). However, depending on how the data are interpreted, the saline-freshwater interface andtheareasofdrainagedividesmaybesubjectivelyplaced. Shafer(1964)placesa groundwater divide across an alluvial fan surface (Fig 2.1). Looking at the water table dataandtheconfigurationofthebasin,itseemsunlikely thathisdivideisvalid. An alternative groundwater divide and saline-freshwater interface have been plotted in 32 Shafer Figure 2.4 which seem more consistent with playa groundwater systems. (1964) doesnotstatewhetherthegroundwatersystem isconfinedor unconfined, however, based on the drillers logs of the lithologies of the test wells and the shallow depthtostaticwater,itismostprobably anunconfinedsystem. Unconfined groundwater systems are common in arid, porous alluvial fan systems. Thelocationof thesaline-freshwater interface has important implications for modeling brine compositions and flow directions in Bristol Dry Lake. According to Shafer (1964) the wide and random variations in groundwater surface elevations and thevariationsingroundwatersalinity oftheBristolDryLakesideoftheinterfacehave not or beensufficienttoallow thesalinegroundwatertomix,eitherby diffusion advection, acrosstheinterfaceandsignificantly degrade thefreshwaterzone. Furthermore, he stated that the shape and relatively less permeable lake and brine pan deposits are theprobable meansofconfining thesaline watertotheplaya areaand preventing the coalescing of the playa groundwaters to form a more homogeneous body around the margins of the playa. The implication of these statements is thatthe saline-freshwater zones are essentially stationary and by implication the solutes as well. Many studies haveshownthatthisisnotthecase,andthatplaya basinsare exceptionally dynamic (see above; summaries inNeal, 1975; and Duffy and Al-Flassan, 1988). In addition, it seems unlikely that a relatively less permeable lake bed could confine saline water by itself because less permeable muds laterally interfinger with alluvial fan sands, so that, even though vertical flow may be retarded, lateral flow is essentially unconfined. ThesalinecharacteroftheBristolDry Lakebasin groundwaterandtheposition of the saline-freshwater interface may actually be related to the topography and evaporation rate of the basin. Duffy and Al-Hassan (1988) determined by computer simulations thatthereliefof thesurrounding mountainsandtheRayleigh number(the ratioofbuoyancy forcestendingtocauseflowtootherforcestendingtoresistflow; i.e. free convection) will determine the hydrochemical cycling in undrained-closed basins. Forabasinwithhighreliefandthusahighrateofuplandrechargethe a high recharge velocity decreases the Rayleigh numberwouldbe low because Ralyleigh number and so retards free convection (Duffy and Al-Hassan, 1988). On the 33 Figure 2.4. Groundwatercontours for Bristol Dry Lake and Cadiz Dry Lake (modified slightly from Shafer, 1964). The position of the saline water-freshwater interface is fairly well constrained to the north, but to the south the data are not abundant. The groundwater contours suggest that water is draining from the northeastfromCadiz basininto Bristol Dry Lake. v^£-1 1 // - i jr /.1 A s f5 \ 5 .j.S a h- s\n«p 1i / /<¦ \V/UCS «n0 V/ 6/ 1' O X5 Ay u /fl// 1 / n \ ")A* 3/ X/V )|1 /) rA"" 0 Vy .08 -I r~ ,AVi) ' /{_ \T' — r/\ i 1 ,> yr r' S;/ 8| \^-% /( 6£yd-' / — 9/ cr­ \V \-r f V r yi I-° !s\ s *i */ / X X 15J ills ! o­ ~ ¦f. /C V/A ft/ jfV ' < —¦crV ';/ °V r'*'* 0 cO /I/ 7A1 r vj H2 * hi UJ f [/* 2 ro k­ 8 *// 0*/// 0 A 2 jlP 3; °°rVJ S** A-p-0 5 v V 'h; D £MLSr A\ *¦ . « // VN^ o « /i 2/ 1 / I?’ b-.. \ / -70/Wil / \ x d V. \ ? r/|V A f^s' \/J? /~* y J rv a <1 ; -J/ r Sos; / ; S 3j| \t>> 1// d> I5S -O\ 4*35//// 3 ?C1 (U =1 / * 0 o// 3 L1 -s> •* l> s' - \>V/i? 0> -J 8 ?R 1 _» ao-LO / £ / O UJ •> «n 8 dt * « / § / t?<0/ Si 0 0 01 N JiJ*/ 3 JA?-A 9 / si vD o/ v'-"" / j 6 n ' son\ / ’O* ; O Cl? I3 «•* O xS 0 r ? sS >* j ftft " :? s»?S —v. ft-?* J> 0/* J J// * Vi />J* iiljd* 11' is s$ ££ ill! * * Ill] • ¦o ©0O 01 2 otherhand, abasinwithlowreliefandlowprecipitationrateswillhaveahighRayleigh number and so promote free convection. Bristol Dry lake fits the latter description. The relief on the mountains surrounding the basin is only about 1,000 m , the average alluvial fan gradients are low (1.8 m/km), and the precipitation rate is extremely low. Thus, Bristol Dry Lake is a prime candidate for free convection. If free convection is taking place in Bristol Dry Lake the cell pattern may look something like the simulations shown by Duffy and Al-Hassan (1988) (Fig 2.5). The r-+20 a ‘ ¦ / / qz - - 40 80 a « R =4 2,100 K/C 3,156 AP' =0 2 -­ L = 0.74 Ac-0265 0 D/L =0.1 /L b —0.106 h 0010­ — _ 's c > >¦ >>* jr tt >¦ V4.*XX XX f >>>>> >->->• >• >-y>. -f X x rr x* t> >>>>>>> >•yyt X y>Ty XXXX q X v f t-t x XXX XXX t >T>> ¦>TrrT7T* KVYV X X XU V < <>tT *¦ r> -f.Vr XXX XXXX XXi.k 4f ' " <<<<<<<<<<<25 In high laminae. content cm) laminae In part of structure and indurated gravel (1 upper Km horizon. Thin laminae Km horizon Is 0.5-1 m thick. >60 in low may drape over fractured surfaces gravel content v Thick laminae (>l cm) and thin to Indurated dense, strong platy to >50 in high horizon content thick pisolites. Vertical faces tabular structure. Km gravel >75 in low and fractures ate coated with laminated is 1-2 m thick. contentcarbonate (case-hardened surface) gravel VI thick >75 in all ," Multiple generationsof laminae, Indurated and dense, strong y breccia, and pisolites; recemented. tabular structure. Km horizon gravel Many case -hardencd surfaces. is commonly >2 m thick. *HlqhIs more than 50 percent gravel; low is less than 20 percent gravel,2p.tc. nt CaCOj in ithe <2-mm-(tact ion of the soil. Tr, trace of carbonate. Figure4.4. Stagesofcalciumcarbonatemorphologyfromcalcicsoilsandpedogenic calcretesdevelopedinnon-calcereousparentmaterialsunderaridto semi­aridclimates. DataarefromthesouthwestoftheUnitedStates(takenfrom Machette, 1985,whomodifiedtheoriginalworkofGileetal.(1966)and BachmanandMachette (1977). 49 evaporites (celestite and gypsum) formed within the sediments that are now at the surface. Distal fan. The distal fan is dominated by sheet flood deposits mostly composedofsand-sizeandfinerparticles. Athinveneerofthedistalfanoverliesand grades into the playa margin sediments. Pedogenic calcite and gypsum, and small amountsofmeniscushaliteformintheporousdistalfansediments. Thepaleosol that extends fromthelowerproximal fan is exposed in borrowpits excavated by the salt companiesandisgenerally less than0.5 mfromthesurface(Fig4.3). The alluvial fans surrounding Bristol Dry Lake have a low gradient (1.8 m/km), although thenorthernalluvialfanstendtobesteeperthanthefansonthesouthsideof theplaya. Violentstormshavebeenobservedtowashlargevolumesofsand-size particles onto the playa surface. However, these storms do not seem to be powerful enough to move the large particles observed at the surface (Fig. 4.5). Yet surface sediment, eveninthedistalpartsofthefan,rangesuptoboulder-size. Someofthe larger volcanic boulders may be ejecta from nearby Amboy Crater. Others may represent the lag of even larger storm deposits which have since been deflated by wind their destination erosion; or alternatively, they may arrive at a few meters at a time as stormafterstormpushes themtowardstheplayacenter. Where corespenetrated thedistalportionsofthefan, sedimentsconsistofplanar laminated,well-sorted,medium-tofine-grained sandwithabundantmudpartingsin thelowerhalfofcoreCAES#l. Planarlaminaearedefinedbyconcentrationsofheavy minerals, but they are neither inversely graded nor do they show other characteristics of wind-ripple laminations. The planar bedding, fine-scale lamination and mud partings suggest that these sediments were deposited by very shallow sheet-flows in the distal fan subfacies. TheCaliforniaSalt Company's 400mwellNo.2completedin 1961(seeFig. 2.1 and appendix IV for location) was logged as containing 30 m of distal fan deposits underlainby300mofplayaandlacustrinesediments. However,thebottom70mof the the core contains boulders, cobbles, gravel, and sand (Shafer, 1964) that probably representsmidfantodistalfandeposits. Asidefromthesecoarse-grained sediments, which taken along the northern margin of the the facies preserved in this core, was playa, suggest that for much the history of the basin, the playa deposits closely paralleled the margins of the present playa. Although small barchan dunes were observed by Handford (1982a) on the playa margin sand flat, these dunes are no longer present inthe Bristol Dry Lake basin. Smallcoppice dunesarepresentonthedistalalluvialfansandplayamarginwhere plantshavetrappedsufficientsedimenttoraisemounds0.5to 1meterhigh. In Figure 4.5. Large volcanic boulderthe surface of the distal alluvial fan. Pen knife on (=9 cm long) for scale. general, eolian deposits are ephemeral in Bristol Dry Lake because the playa surface lacks sufficient moisture to trap sediment. Although the wind regime in the basin is seasonal,thedominantwinddirectionistothesoutheast. Strongwindshavemoved sand the sides of up onto the divide separating Bristol Dry Lake from Cadiz Dry Lake. Inaddition, alargedunefield,withsomedunesupto15m high,ismigratingto the southeastdirectlydown-windofthedivideintheCadizBasin.Thissuggests thata 51 greatdealofthesandistransported outoftheBristolDryLakebasinandover the divide into the Cadiz basin. in No unequivocally aeolian deposits have been seen core. Paragenesis. Although alluvial fan sediments are not abundantin the cores, particularly theproximalandmidfanareas, theparagenesisofthealluvial fan sequence can be predicted based on the early diagenetic reaction seen at the surface. Themostimportantearly diageneticreactions aretheformationofsoilhorizons and calcrete zones that will significantly occlude porosity by the formation ofclay coats andcalcitecements. However, theearlycementsandcoatingswillhelptopreservethe initialframeworkofthegrainsandpreventcompaction. Ifdiageneticconditions change after some burial, these initially less porous and permeable beds may become highly porous by dissolutionof the cements and the generation of secondary porosity. The preserved initial framework would likely be more porous than many of the uncementedcompactedunits. IntrusionofBristolDryLakebrinesintothealluvialfans the basin. This type ofporosity occlusion be distinguished from the early calcrete cementation by upon burial may also occlude significant porosity later in the history of can thepresenceofunusualsolublecementssuchashalitein theupdippartsofthefans, and the post-compaction grain packing of the cemented unit. An example of this type of post-compaction porosity occlusion was recognized by Mcßride et al. (1987) in halitecementedNorphletFormation(Jurassic) eoliansandsinMississippi and Alabama. Anotherdiagenetic reactionwhich isoccurring near thesurfaceofthedistalfan and into the playa margin sediments is the formation ofopal as pore-filling coatings aroundroots(Fig 4.6). Although theamountofopal generated is small, theirpresence indicatesthatsomeearlymobilizationofsilicaoccurs. Haliteandgypsumalsomay formlocallyinthecapillary zoneofthealluvialfan,however,itisunlikely thatthese early soluble phases will be preserved. The immature character of the alluvial fans indicates that at depth, or with some additional burial, other authigenic phases such as quartz overgrowths, feldspars, chlorite, clays, zeolites etc., may be important late diageneticcements.Thesetypesofcementsandauthigenic phases,alongwith pressuresolution, suturedgrain boundaries, andothercompaction featuresshouldbe expected aslatediageneticoverprintsoftherelatively simpleearlydiagenetichistoryof thealluvial fans. If theselate diagenetic eventscanbeavoidedoroccuronly aslocal 52 minor cements, the fans may provide excellent reservoirs for potential hydrocarbon accumulation. Obviously, sourcing these reservoirs is the major problem. However, Figure4.6.Photomicrographofpore-fillingopal(arrows)coating aholeformedbya rootthathassincedecayed. Longdimensionofphotographisabout1.5 mm. significant organic accumulations may be possible in more pluvial times, at least in basinswithhighhinterlandswhich can trapabundantmoistureandsupport alarge organic community. Playa basins doexist with significant hydrocarbon accumulations inthe alluvial fans such as the Green River Formation (Eocene) of Wyoming (Smoot, 1983). The hydrocarbons appear to be locally sourced from organic oozes infiltrating through mudcracked sediments (Eugster and Hardie, 1975). In Bristol Dry Lake, however, there is no evidence for such oozes ever having existed. 53 Playa margin Playa margin sediments are deposited in transitional zone vegetated by sparsely a populated halophyte shrubs toward the distal alluvial fan side of the facies which then passes gradually into the barren saline mud-flat. The sediments vary from silty sands to sandy muds that contain calcite-cemented nodules (some with minor of opal amounts rimming cement) surrounding root holes of former halophyte shrubs. In addition, mud-filled traces of roots and burrows are also present, and blocky prismatic or columnarfracturesin thesedimentindicate apossiblelowerBtorBnsoilhorizon (Birkeland, 1984, p. 15). Small root traces and possible burrows are present in siliciclastic intervals from the CAES #1 well, as deep as 500 However, calcrete m. or silcreteassociatedwithrootzoneshavenotbeenfoundincore. Thisismostlikelydue to the fact that the cores were taken too far toward the basin-center to penetrate these zones. Numerouswadichannelsanddistributarychannels(Handford, 1982a)dissectthe playamarginandbypass sedimentand organicmatterfrom thealluvialfansdirectly acrosstheplayamargintotheplayacenter. Handford(1982a)recognizedthatthe distributary channels may be an important mechanism for distributing sediment across the playa. Some of these distributary channels are up to 1 m deep and 100 m across. Thechannelsextendwelloutontothesalinemud-flatandcanbe seen onaereal photographs of the playa. Shallow trenches across these wadis and distributary channels revealflaser-beddedsandsandmuds,alsoreportedbyHandford(1982a) from the surface sediments, as well as planar bedded, sheet-flow sands. Basinwardofthevegetation, theplayamarginpassesintoazone,approximately 30 m wide, where large centimeter-size gypsum blades are cemented by fine-grained calcite. Thecalciteinplaces linesvoidsproduced bydecayedrootsordissolutionof This zone represents a mineralogic transition area from calcite- gypsum (Fig. 4.7). dominatedsedimentto dominatedsediment.Theco-occurrenceofcalciteand gypsum-gypsum in this zone is probably due to fluctuating saturation states of the groundwater over time. During periods ofless rain the groundwater would be more concentrated at this point in the basin and so precipitate gypsum. During wetter periods, the groundwater would not reach gypsum saturation until it was farther into the center of thebasin. Thus,thegypsumformedinthesedimentwouldbeoverprintedwithcalcite cements precipitated at a later time. Further towards the basin center, the calcite becomes lessabundantandvertically alignedgypsumandnodularcelestitedominate. Figure 4.7. Secondary electron image of stubby, isopachous calcite lining a pore in calcitecementedsedimentintheplayamargin. Scalebaris100(im. - The gypsum is precipitated in a 150 50 m wide lens-shaped body that encircles the entirebasin(Fig.4.8a,b). Fieldandchemicaldataindicatethatdischarging groundwaterin this zone is saturatedwithrespect togypsumandcelestite. Vertically elongated gypsum crystals (up to 40 mm along the long axis) with inclusions of matrix surrounding growth surfaces demonstrate that the gypsum forms displacively in the 55 Figure 4.8.a)Facies mapofsurface sedimentsshowing locationofgypsum-celestite zone and the "bulls-eye" pattern of the evaporite minerals. The playa margin subenvironmentextendsfromtheedgeofthepresentplaya surface (and under the lava flows) to the edge of the gypsum-celestite zone. Although isolateddisplacive gypsum occursin the salinemudflat, almostallof the gypsumisconcentratedinthegypsum-celestite zone. Notethatatpresent there are two salt pans on the surface, b) Cross-section from A to A' shown in Figure 4.8a. Therelief on this cross section is from topographic maps andthepositionofthedifferentsubenvironmentsatthesurface arebasedon the facies distribution in 4.8a. inferred normal Thedashedlinerepresents an rotationalfault. Thepositionofsaltbedsandfaciesdistributionswithdepth areschematicbasedoncoresandtrenches. Calciteisprecipitated ascalcrete horizonsand nodulesin thealluvial fans andplaya margin sediments. Gypsum and celestite are confined to the playa margin sediments as well. The saline mud flat is dominated by molds of small (5-10 mm) displacive halite which become larger and remain filled towards the basin-center. The modem salt consists of vertically aligned chevron halite. Handford's pan (1982a)crosssectionofBristolDryLake(hisFigure 12),showsgypsum andanhydritein thesalinemudflatallthewayto thebasin-center.Although minoramountsofgypsumandanhydrite arefoundinthesalinemudflat,the bulkofthegypsumisprecipitated inanarrow bandaroundtheplayamargin. This distinction is important in constructing vertical facies models for the evolution of the basin. 56 sediment near the top of the groundwater table where the sediment is still water-saturated and easy to move (Fig. 4.9). Where the groundwater table has been constant for some time, the precipitation of gypsum has been greatest. In this area, gypsum may accountfor upto90%ofthevolumeofsedimentintrenchesthatare2mdeep. The matrixsurroundingthegypsumconsistsofamixtureofdetritalsiliciclastics (quartz, feldspars, clays, and a large suite of heavy minerals) ranging from sand to clay-size particles, and authigenic rhombic calcite less than 1 |im in size (Fig. 4.10). Figure4.9.Fieldphotographoftrenchthroughthegypsum zoneoftheplayamargin facies. Notice the vertically aligned fabric of the individual gypsum crystals. Theverticallyalignedfabricisdescribedinsomedetailinchapter 5. Lens cap is about 5 cm in diameter. Just basinward of the transitional area from calcite to gypsum, there are large blocksof thatarenotsupportedinthetrenchwall.Inspectionofthematerial gypsum of the indicatesdiageneticrecrystallization gypsumintolargepoildliticcrystals overprinting the original fabric and vein of gypsum cements filling voids and oriented perpendiculartotheearlierfabric(Fig.4.1la,b). Thisportionofthetrenchis essentially a gypsum karst area. The present near-surface groundwater system is apparently undersaturated with respect to gypsum at this point in the playa margin sediments and so is dissolving and reprecipitating gypsum with a net loss of gypsum in this area over time. This "cannibalized" calcium sulfate may then be precipitated in a more basinward area of the playa. This gypsum karst is important because it implies that the present chemistry of the groundwater is not in equilibrium with the observed position of the mineral phases in the facies. Figure4.10. Grainmountphotomicrographoffine1-2pmrhombiccalcitecrystals (arrows). Many of the crystals have hollow or fluid inclusion-rich centers. This small rhombic calcite is common in the matrix of the gypsum-celestite zone inthe playa margin facies. Figure 4.11. a) Cross-polarized photomicrograph of large (=2.5 mm) poikilitic gypsum crystal (white area) enclosing the detrital matrix, b) Cross-polarized photomicrograph of large,clear gypsum crystals lining and filling vertical fractures. Original matrix ofsmaller vertically aligned gypsum crystals is still present. Fracture is approximately 2 mm across. 60 61 Thepresenceofverticallyalignedgypsumgrowingdisplacively inthesediment has implications for the criteria used for distinguishing gypsum depositional environments. Thus, it is important to be able to distinguish this fabric from other similarfabricsdepositedindifferentenvironments. Therefore,thedetailsof this fabric are compared to two other modem gypsum environments in a later chapter. However inthe following section, the origin of the gypsum and its morphology will be discussed in some detail. Gypsum growth and morphology. Environmental conditions controlling gypsumnucleationandcrystalmorphology hasbeenstudiedexperimentallyby numerous authors, but most recently by Kushnir (1980), Cody (1976, 1979) and Cody and Cody (1988a,b). A summary of this literature and a classification of gypsum morphologies from field studies is presented in Magee (1988). Gypsumoccursinmanymorphologies, buttherearethreebasicformswhichare used to describe gypsum in the literature: these are 1) prismatic (bladed, spears), 2) pyramidal ( hemi-bipyramidal, lenticular, disk, lozenge, or lens-shaped), and 3) pinacoid. The names in parentheses are common descriptive terms used for these crystal habits,however, thetermsbladedandspearsmayalsobeusedtodescribesome pyramidalforms. Apparently pinacoidalformsarerareinnature(Magee, 1988)and so are not included in the following discussion. Acicular prismatic crystals grow dominantly in pure aqueous solutions supersaturated with respect to gypsum. Edinger (1973) suggested that this is due to the selectiveadsorptionofHandOHontothe 110and 010facesofthecrystal causedby the arrangement of atoms in the crystal. The adsorption on to these faces inhibits growthinthesedirectionsandpromotesgrowthofthe 111faceandelongationinthe c-direction. These prismatic crystals grow mostly in sediment free solutions such as bottom-nucleatedbrinepondgypsum at the sediment-waterinterface and as free crystals at the brine-air interface (Cody and Shanks, 1974; Cody, 1976). Pyramidal forms grow when the 111 face is selectively inhibited. Cody (1976, 1979) and Cody and Cody (1988a) have shown that organic additives commonly presentin terrestrialmudsinhibitgrowthandpromotedevelopmentofthe 111face. Thus, the crystals are flattenedperpendicular to the c-axis. Pyramidal forms are common in muds which are supersaturated with respect to gypsum. These muds 62 generally contain the types oforganic matterwhich inhibit the growth of the 111 face. Thereforemostdisplacive gypsumgrowinginmudispyramidalorlenticularinform. Magee (1988) has used this dataand field observation from an Australian inland lake to concludethat,as aruleofthumb, subaqueously deposited gypsumisdominantly prismatic in form and displacive groundwater seepage gypsum is pyramidal in form. Ingeneral thisconclusionis valid,but althoughBristol DryLakeis dominatedby pyramidal gypsum growing in muds and sands, prismatic gypsum does occur, yet thereisnoevidenceforgypsumsaturatedbrinepondsinthebasin. Theeffectofother factorssuchaspH,temperatureandnucleationratemayinfluencethecrystalform. In addition, diagenetic effectshavenotbeentakenintoaccountinanyofthesestudiesand itnotknownwhatroledissolution-reprecipitation reactionortheconversionofgypsum to anhydrite and back again will have on the final observed crystal form. Although conditions in terrestrial environments such as Bristol Dry Lake vary considerably, the most important parameters which may potentially control gypsum nucleationprocessesare 1)watersalinity,2) temperature,3)typeandamountof organic matter, and 4) pH (Cody and Cody, 1988a). By growing gypsum crystals under controlled laboratory conditions, Cody and Cody (1988a) were able to determine how these parameters affected gypsum morphology, nucleation density, and growth rate in simulated terrestrial environments. The most significant results of their study can be summarized in a diagram relatinggypsum crystalhabitsgrowninamud-richenvironmenttoorganic additivesat moderate(50-200°/oo)salinitiesandpHgreaterthan7.5(Fig.4.12). Although Bristol Dry Lake gypsum exhibits all of these habits in some abundance, the lenticular, lenticulartwin,andlenticulartwincomplex habits arethemostabundantintheplaya marginfacies. IfFigure 4.12isthenapplied, theconclusionwouldbethatalarge amountofthegypsuminBristol DryLakeisformedinmoderatesalinities, relatively hightemperatures (>35°C),moderatetohighconcentrationsoforganicadditives, anda pHgreaterthan7.5. Themoderatesalinitiesandhightemperaturesarereasonablefor BristolDryLake. Thetemperatureoftheplayasandafewmillimetersbelowthe surfaceinOctoberwasmeasuredat44°C whiletheairtemperature wasmeasuredat 36 °C. In the summer months the playa surface would probably be at least 10 °C hotter. the Thesalinity ofthewaterisapproximately 250-300°/ooin thisportionof basin(seeappendixIV).Theamountoforganic mattermayalso bereasonable, howeverithasnotbeenquantifiedinBristolDrylake sediments. Decayingorganic mattercanbe seeninterspersedin thesedimentthroughouttheupper10 m of the playa, and accumulations of halophyte plant leaves and twigs litter the surface of the dried up wadi channels. However, for the organic compound used in the experiment to be an Figure 4.12. Paragenesis of gypsum crystals formed in a muddy media under experimental conditions (from Cody and Cody, 1988a). VL = very low or zeroconcentrationofpolytannate ionconcentration,M=medium concentration, and VH very high. = effective inhibitor, it must be deprotonated. The organic polytannic ion (an analog of terrestrial humic substance) used in the experiments deprotonates at apH greater than 7.5(CodyandCody, 1988a).WateranalysesfromthealluvialfansedimentsofCadiz Basin and Bristol Basin have pH values that from 8-7.2, however, basin-center range brineshavepHvaluesof5.7-6.5 andingeneral thebasincannotbeconsidered alkaline. Unfortunately there are no pH measurements for the water analyses taken in gypsumzoneoftheplayamargin faciesandthereis novalueforanyofthecarbonate speciestodeterminealkalinity. Inaddition,itisnotknownifthepHmeasurements were taken in the field or in the laboratory. However, assuming that the pH measurementsarevalid, thedecrease inpHtowardsthebasin-centercan beattributed, atleastinpart,to thelossofbicarbonateby calciteprecipitation in thealluvial fan sediments. Becausemostofthecalcitehasprecipitatedbythetimethewaterreaches the playa margin facies, it is unlikely that pH will be as high as the 7.5 necessary for the polytannic ion to effective. Cody and Cody (1988a) point out that other threshold-inhibitor functional groups deprotonate above pH of 4-5. However, they do not state whether thesegroupsare abundantinterrestrial environments. Representative samples of gypsum crystals from the trench on the south side of the playa were sent to Dr. R. D. Cody for examination. Although the Bristol Dry Lake crystals areverysimilartocrystalsgrownwithpolytannateions,thegenerallenticular habitdoesnotappeartobediagnostic ofanyparticular organichabitmodifierchemical (Cody, 1989 written comm.). It is possible that the Bristol Dry Lake gypsum crystals undertheinfluenceofsomeorganic habitmodifierthatthat has yettoisolated, or grew thatthereare acomplexassortmentoforganic compounds thatcontrolgypsum crystallization. Infurthersupportofabacterially inducedoriginofthegypsum,Eardleyand Stringham (1952) have postulated that sulfate-reducing bacteria play an important role in the formation of verticallyaligned pocketsofgypsumcrystals forminginthebottom muds in Great Salt Lake, Utah. These deposits are not well described, but they do appearto be similar to the vertically aligned gypsum in Bristol Dry Lake. However, EardleyandStringham(1952)alsorecognized theneedforelevatedtemperaturesand pH to initiate precipitation of gypsum. Although there is good agreement between the experimental and field evidence suggestingorganicmatterandtemperature areimportantinfluencesongypsum morphology, the most important element, the pH of the system, is not well-enough constrained in Bristol Dry Lake to be definitive. It is apparent from other lines of evidencepresented laterinthesectiononcalciteconcretions thatbacteria orotherforms of organic matter are important in the growth of these concretions. Therefore, it is also possiblethatorganicshavehadsomeroleinsulfateformation. Itwouldbeinteresting to conduct isotopic experiments on the organic versus inorganic gypsum produced in thelaboratorytodetermineiftheremaybe somefractionationthatwouldhelp determine the origin of the gypsum. Alternatively, Kushnir (1980) has shown inlaboratory experiments that gypsum growth habits may be controlled by the Ca/SC>4 ratio in the solution. Essentially, Kushnir(1980)hasshownthathighCa/SC>4ratiosproducelenticularcrystals flattened in the c-axis and at low Ca/SC>4 ratios prismatic crystals are formed. Bristol Dry Lake ischaracterizedbyextremely highCa/SC>4ratios (appendixIV)andisdominatedby lenticulargypsumcrystals. Growthoflenticulargypsuminaninorganic systemwould eliminatetheneedforlargeamountsoforganic matterandapHcontrolonthesystem. Withthepresent stateofknowledge, itisnotpossibletodeterminewhetherthe lenticularhabitatBristolDryLakeisorganically orinorganicallycontrolled. However, itmaybethatthereasonnospecific organicsubstancecanbeassignedtotheBristol Dry Lake habit (Cody, 1989; written comm.) is that the habit is inorganically controlled. Another important aspect of the experimental work on gypsum formation is the determinationof the parameters that control the nucleation density of crystals. Cody andCody(1988a)determined thatwithincreasedNaClcontentinthebrinenucleation densitydecreasedproducinglarger,slowergrowingcrystals. Ina5%NaClsolution thenucleationdensitywas 10timeslessthanthedensityinaNaCl-freesolution.The NaCleffectmayexplain whythegypsumcrystals in theplaya marginsedimentsof Bristol Dry Lake become larger towards the basin-center where the NaCl concentrationsincrease. IntheNaCl-dominatedsalinemud-flatgypsumgrowinginthe sediment is rare. This is probably due to a shortage of sulfate ions, but where gypsum does occur in the saline mud-flat, it always forms as large (10-30 cm) widely dispersed crystalaggregates suggesting alownucleationdensityperhapsenhancedbytheamount of NaCl present inthe solution. Celestite.Celestite nodules (up to 0.2 m) of 0.2-0.5 mm-size crystals, are presentbasinwardofthebulkofthegypsum. Celestiteispresentincleansiliciclastic f HOfWI a Binsojog d izoidofoan eiue*lAsuusj am hU*huhi3 Figure 4.13. Handsampleofcelestite nodule.Noticethatthenoduleisrelatively pure except for a thin rind around the outside. Vertically aligned gypsum (present on the left side of the nodule) surrounds the nodule in place. Also notice thatthenoduleislongeralong thevertical axisthanalong the horizontal axis. 67 sands as well as surrounded by vertically elongate gypsum (Fig 4.13). This suggests that the celestite grew before the precipitation of the gypsum surrounding it and that the celestite is a primary diagenetic feature and not the replacement of a previously existing sulfate phase. The nodules may grow as discrete, white, roughly spherical masses of interlocking, randomly oriented,euhedralcelestitecrystals, ortheymay coalesceinto large(up to0.5m) patchesofcelestite(Fig 4.13and4.14).Thenodulesmay be somewhat porous, (approximately 10% ) and may be aligned such that their vertical dimensionislonger thantheirhorizontalwidth, eventhough theindividualcrystals are randomlyoriented. EnergydispersiveX-rayspectraofcelestitecrystalsviewedinthe SEM indicate asmallamountofbariumpresent,andalthough quantitative analyses werenotundertaken, theheightofthepeakssuggest thatitislessthan5molepercent. ThesubstitutionofafewmolepercentBaforSriscommoninsedimentary celestite occurrences(Deereetal., 1978),althoughthepresenceofdiscretebaritecannotbe ruled out. Gale (1951) first described the occurrence of celestite in Bristol Dry Lake and Durrell(1953)provided thestratigraphicframeworkforthecelestitenodules. Durrell (1953) interpreted the nodules as forming by the replacement of matrix sediment by the dissolution of clastic grains. This interpretation was based on the purity of the nodules andthelackofdisruptivebeddingaroundthem. Examinationofcelestitenodulesin trenchesindicatethatwhenthey aresurroundedbysandandmuds,theydodisplace sediment. However, because they formed before the gypsum surrounding them, they do not show displacement of the gypsum. In addition, nodules do include rare clastic grains which show no evidence of dissolution. Finally, the wholesale dissolution of quartz, feldspars, clays and a large numberof heavy minerals in smallnodularpatches is hydrochemically untenable. Thus, the field and petrographic evidence indicate that thecelestiteformeddisplacivelyratherthanreplacively. Inadditiontothefield evidence for a primary diagenetic origin of the nodules, petrographic evidence also suggests that the nodules are primary diagenetic features and are not formed by the Evidenceforthisisdemonstratedby 1)thepurity replacement of other sulfate phases. of the celestite; lacking relict inclusions of anhydrite or gypsum in the nodules, 2) the of the nodules in areas otherwise now barren ofevaporites, and 3) the sharp presence contactbetweencelestitenoduleswithmillimeter-sizecrystals againstgypsumcrystals that are centimeter-size. Figure4.14. Secondaryelectronimageofingot-shapedcelestitecrystalsinrandom orientations. Matrix is smaller celestite crytals, although is some places rhombiccalcitehasprecipitatedinthepores. Scalebarinlowerright comer is 40 |im. Surface to core comparisons. To one degree or another, all fabrics seen atthesurfacehavebeen identifiedincore. IncoreCAES#l,thebasinwardedgeofthe gypsum zone is penetrated at many intervals in the core. Although all the gypsum has beenconvertedtoanhydriteby 100mdepth,thefabricstillretainsthevertical alignment seen at the surface. In thin section, the anhydrite contains calcite and pyrite concentratedaroundtheoutlinesof havebeen the grain boundaries presumed to This mineral originally gypsum which have now been converted to anhydrite. assemblage suggests that bacterial sulfate reduction was involved in the conversion of gypsum to anhydrite (Berner, 1971). The calcite luminesces brightly indicating the incorporationoftraceamountsofreducedironandmanganeseinto thecalcitecrystal lattice. Calcite from the surface does not luminesce. This is consistent with the hypothesis thattheluminescingcalciteprecipitated laterthanthe surfacecalciteandthat theluminescing calciteprecipitated asabyproductofbacterialsulfatereduction.The large celestite nodules found in the surface trenches were not penetrated by the cores. However,small millimeter-sizepatchesofcelestite associatedwithcalcite andanhydrite are present in one sample from 500 m (Fig. 4.15). Figure 4.15. Backscatter electron image of small celestite patches (white areas) from CAES #l. The matrix is mostly calcite, but a few centimeters below core the celestite is a 100 mm thick anhydrite unit. Scale bar is 1000 pm; core depth 481 m. At the surface, interbedded siliciclastic sands and millimeter-laminated calcareous muds are disrupted by bioturbation, displacive evaporite crystal growth, and mechanical compaction of the sediment. Large burrowing spiders which abundant are in newly dug trenches are probably responsible for the burrows. Microfaults and flow structures(Fig.4.l6a) areformedbyacombinationof displacivesulfatemineral growth and unstable density contrasts in the intercalated sands and muds. Disruption of siliciclastic beds by bioturbation, flow structures and microfaulting can be seen in core CAES #l. However, in core the flow structures and microfaulting are not always associatedwiththepresenceofevaporiteminerals(Fig.4.16b). Theabsenceof displacive evaporite minerals may be due to early dissolution and subsequent non-preservation of the evaporite phase. Atypical verticalsequenceofthesurfaceplayamargin/distalalluvialfanfacies (based on the lateral facies positions of the facies) is shown in Figure 4.17 and comparedtoaverticalsequencebasedonthecore(Fig. 4.17).Themaincontrast between the two sequences is the amount of gypsum/anhydrite in the core sequences versus the surface sequence. At the surface over two meters of almost pure gypsum is present, yetincorethereisrarely over 0.3metersofanhydrite. Thereasonforthedifferencemaybeaccountedforbyone(or acombination)of severalhypotheses: 1)thecorewastakenbasinwardofthemainaccumulationofplaya margin sediments, 2) volume loss due to the conversion of gypsum to anhydrite. 3) wholesaledissolutionofcalcium sulfate,4)relativestability ofthepresentgroundwater system over time. If volume loss due to the conversion of gypsum to anhydrite were core andtheretentionof theoriginal important, collapse brecciaswouldbecommonin fabricwouldberare. Thisisnotthecaseinthecore. Fabricsaregenerallyretained andcollapse breccias arerare. Infact,theretentionofthefabricimpliesthatcalcium sulfate was added to the system, not subtracted. The stratigraphy of the core suggests that the lack of calcium sulfate may be due to the position of the core in the facies tract. of athick accumulation of chaoticmud-halitein theupperportions ofthe The presence core and thick accumulations of mud, sand and anhydrite at the base indicates that the core was taken at the transition of the edge of the brine pan/ saline mud-flat facies and the then gradually passes into the playa margin/distal alluvial fan facies at the base of core (see appendix I). However, evidence presented above demonstrates that the present surface of the playa has not accumulated sediment for some time, possibly 0.25 m. y. 71 Figure 4.16. a) Field photo of microfaulted and flowing sand and mud layers at the surface in the playa margin sediments. Displacive gypsum is in the bottom of the photo and a celestite nodule (white patch) can be seen in the upper left, b)Coreslaboffaultedandbrecciatedinterlaminatedmudand carbonatefromCAES#1(depth is290m). Thus,thestabilityofthewatertableoverthattime havecreated may an anomalously high accumulationof gypsum at the surface compared to the core. Besides the sulfate problem, the sequences are remarkably similar (Fig. 4.17). Figure 4.17. Comparison of a vertical cross-section through the distal alluvial fan and playa margin sediments from the core and surface. Notice that the surface sedimentscontainmuchmore gypsumandcelestite thanthecore,but otherwise the similar. sequences are very Eachsequencebeginswithdepositionofasandyunitwhichgradually finesupward intomud. Thispartofthesequencerepresents waningflowdepositionofdistal alluvial fan sediments as the sediment reaches the flat playa surface during storms. In some instances the fan sediments flowed into shallow bodies of standing water. As the reached only after all the water had been waterevaporated, gypsumsaturation was evaporated from any standing water. The sediment, however, was still moist. The vertical formed in the transition area between the completely phreatic zone and gypsum the zone ofcapillary rise. In some sequences at the surface, the gypsum is capped by a reddenedsandy These zone which probably represents a subaerial exposure surface. surfaceshavenotbeenseenincorebecauselocalreducing conditionsassociatedwith the conversion of gypsum to anhydrite have altered the color of the sediment around the anhydrite togreenincore. Repetitionsoftheoverallsequencearecommoninthe lowerhalfofCAES#l. Towardthemiddleandnearer thetopofthecorethesand component tends to be absent, indicating a more basinward position of the anhydrite. Paragenesis. The paragenesis of the playa margin sediments is the most complex ofallthefaciesinthebasin. This isbecause themineralphasewhich isprecipitated depends on thepositionandsalinity ofthedischarge zone.Withtime,thepositionand the salinity of the discharge zone varies depending on both short term and long term variationsinmoisturecontenttothebasin(i.e. changes arebothseasonalandclimatic). These variations will cause contemporaneous overprinting of some diagenetic phases anddissolutionofothersatdifferentendsofthefacies. Ingeneral, thelateralfacies distribution should show a simple evaporative concentration mineral assemblage from calciteintheproximalendofthefaciesthrough vertically elongate gypsumcrystals to celestitenodulesinthedistalendofthefacies. However,overprintingofvertically around the celestite is elongategypsum commonatthedistalendofthefacieswhile contemporaneous karsting of gypsum and subsequent reprecipitation as fracture filling cementisprevalentattheproximal endofthefacies.Small amountsofrhombiccalcite alsoprecipitatethroughoutthefacieswheneversalinitiesdecreasebelow gypsum saturation and sufficient HCO3 is available. By 100 m all of the gypsum has been converted to anhydrite accompanied by bacterial sulfate reduction. This "late" diagenetic reaction is seen in all facies by this depth. Saline mud-flat Thesalinemud-flatcanbesubdividedintoasand-richflatwhich islocatedjust basinwardoftheplayamarginfaciesandatrue salinemud-flatbasinwardofthe sand- flat. InBristolDry Lake,bothofthesesubfaciescontainabundantmudandbothare salineandgenerallywettosomedegree. Inaddition,thecriteriafordeterminingwhere the contact is between these two subfacies is highly variable and subjective. Therefore, in this study all sediments basinward of the playa margin sediments (except for the salt pan) are considered to be in the saline mud-flat. The salinemud-flatis dominatedby detritalmud with minor amountsof siliciclastic sand. The mud and sand are carried into the basin by surface discharge of the alluvial fans. At the surface and in core very few sedimentary structures have been observed in the saline mud-flat. It is, in general, a homogeneous green to red cohesive mudbarrenofflorabutdisruptedbydisplacive haliteandcalcite. At times when the climate is moist enough to allow ephemeral or possibly perennialshallowwaterbodiestoexist, thesedimentcarriedintherunofffromthefans will flow out across the the water body because, even with the included upper part of thebrine in sediment, the density of the incoming fluid will be less than the density of thebasin.Themudwillstayinsuspension untilitflocculatesandthenquickly settleto thebottom. Becausemostofthemudreachesthe bottombysuspension settling, even ina very shallow (<1 m) brine pan, the mud will have few sedimentary structures besideslaminations. However,laminationsarerareinthebasin-centermuds. Thus, thehomogeneous natureofthemudisatleastinpartduetotheoriginaldepositional environmentof the mud. Subsequent homogenization by diagenetic crystal growth is also a factor. Thedominantly grayishgreenmudisoxidizedtoareddishbrowncoloratthe surface.Individualmudpackagesarenotthick,usually lessthan0.5m,andsomeare cappedbymudcrackedsurfaces. Laminationandsubaqueousripples arepresentinthe salinemud-flat,buttheyarenotabundant. Laminationsgenerallyconsistof discontinuous millimeter-thick, clastic mud-carbonate mud couplets. These couplets mayrepresent seasonalchemicalvariationsinthebasinwhenshallowpermanentwater bodies were present, but their discontinuous nature suggests that they are not basinwide events and may represent local chemical perturbations in ephemeral water bodies. Themost common sedimentary structure is lenticular to flaser bedding produced either as the waning flow of the wadis discharged mixed sand and mud onto the dry playa surface, or as the sedimentreworked in shallow standing water was bodies by wind-inducedwaves. Nearthebasincenter,extensivethinlayersofgypsumandanhydrite occurjust below the surface (Fig. 4.18). The gypsum consists of small, millimeter-size disks of lenticular crystals, similar to what is seen in the playa margin sediments. The gypsum isorientedwithitslongaxisparallel to orintheplaneofthebedding. This is the opposite of the gypsum disks in the playa margin sediments which are vertically aligned. Thealignmentofthegypsumparalleltobeddinganditssimilaritytothe gypsum in the playa margin sediments suggest that the gypsum was either blown or washedintoastandingwaterbody ratherthanforminginthesalinemud-flatsediment. Thelenticularmorphologyofthecrystals indicatesthatitisunlikely thatthegypsum precipitated in the water column and fell to the bottom (see above). Discontinuous pods of anhydrite occur along with the gypsum. The anhydrite consists of small (10­20microns)lathsofdenseanhydrite. Itisnoteasytodemonstratewhetherthe anhydritereplaced thegypsumorwhethertheanhydrite is aprimaryprecipitate, however, sulfur isotopic analyses suggest that at least some of the anhydrite may be primary (see chapter 8). The sulfate-laminated zones are generally parallel to bedding, although there are occasional hummocks in the bedding (Fig. 4.18). The sulfate layers thin across the tops of the hummocks indicating that the hummocks are depositional features. In the crests of the hummocks, large v-shaped desiccation cracks are filled with the sediment from above (Fig. 4.19a). Analogous v-shaped cracks are seen in core but are filled withanhydrite insteadofmud(Fig4.19b)Itisnotknownifthesecracks havethesame origin as the desiccation cracks at the surface. Although the laminated gypsum zones appear to be somewhat pervasive at the surface, they are of minor abundance in core. Thismaybeduetodisruptionanddispersalbydisplacivehalitegrowth. Enterolithic structures, contorted layers of anhydrite or gypsum formed by displacive growth, were reportedbyHandford(1982a). Incore,enterolithicanhydriteispresentbutnot Atthesurface,enterolithicstructures areabundantinthenorthwestand common. southwest portion of the saline mud-flat in trenches which have since been filled in (Handford, 1982b;andpers.comm.).Therelativescarcityoftheenterolithic structuresatthesurfaceandincoresuggests thatitis aminorcomponentofthetotal sediment package and should not be used as a diagnostic characteristic of the playa. does Aminoramountofsubvertically elongate gypsumspears occurinthesaline mud flat at the surface; althoough these spears or their equivalent anhydrite This pseudomorphshavenotbeenobservedinthecore. gypsumprobablyhasa formedin theplayamargin sediments similardisplaciveorigintothegypsum exceptin this the gypsum disks derives its sulfate from dissolution of the detrital case, gypsum Figure4.18.Fieldphotographofanexposedtrenchnearthebasin-center. Gypsum occurs aswhiteparallellaminatedstreaksaboveandtotheleftandrightof the hammer. Anhydrite occurs as small discontinuous patch in the sedimentandarenotvisibleinthisphotograph. Noticethatthelaminations benduparoundthefeaturebehind the hammerandthatthelaminations pinch out and thin across the top of the hummock. This morphology indicates that the laminations are forming by the detrital gypsum accumulationofcrystalsacrossapre-existing surfacehigh.Thesurfaceof the high was desiccated and infilled with mud from above (see 4.19 a). The saline mud-flat to alreadypresentinthemuds. canbequitewide,up 5kmacross, and at times in the past extended across the entire basin. The detrital muds are pelleted, probably by brine shrimp which hibernate in the mud when the lake dries out. The muds also are inorganically flocculated as discrete mud particles suspended in the freshwater runoff settle into, or mix with, the underlying saline brine. The flocculation occurs as the ions in the brine adsorb onto the clay surfaces, creating a charge imbalance which attracts nearby clay particles. Figure4.19. a)Close-upviewofthetopofthehummockinFigure4.18.Desiccation crack is outlined by various arrows to show the contact with the underlyingsediment, b)Subsurfaceanhydritefilledcrackfromabout 390minCAES#l. Noticethedisruptionoflaminationaroundthecrack andtotheleftofthecrack. Thescale ofboththesurface andsubsurface cracks are about the same. 78 Thecompositionoftheclays wasdeterminedbyx-raydiffractionofthelessthan2jam sizefractionofairdriedandglycolatedsamplesfromthecoreandthesurface. For 18 samples from the entire length of the cores, the bulk assemblage was constant. The clays in the samples consisted of illite, illite/smectite mixed layer clay, chlorite and some kaolinte. The clays constant relative intensities of were not strict possibly occur in proportions as shown by the the peak heights. However, analytical procedures enough to quantify the relative proportions. Other components in the less than 2 pm sizefractionincludedquartz,calcite,anhydrite,feldspars,andrarely anunidentified zeolite phase.Theconsistency oftheclayassemblagewithdepthindicatesthattheclays are dominantly detrital and that major diagenetic changes within the clays have not occurred. This is not to say that subtle diagenetic alterations have not taken place or thatminoramountsofauthigenicclays havenotformed. Moreworkwouldbe necessarytodeterminethedetailedgenesisoftheclayfraction. Althoughthismaybe importantfordiageneticreactionslaterintheplayas' history,atthispointitwouldonly beconjecture astohowitwouldaffectthefinalrockassemblage. Thesurfaceofthesalinemud-flatis hummocky andinmanyplaces water saturated. Ithasbeenvariously describedintheliteratureas"self-rising ground" and "puffy" (Thompson, 1929; Bassett etai, 1959; Gale, 1951). The "puffiness" is due todisplacivehalitegrowingjustbelowthesurfaceandpushing thesedimentupward Desiccationcracks andmillimeter-thickefflorescenthaliteandcalcium (Fig. 4.20a). chloride crusts are also common on the mud-flat surface (Fig. 4.20b). Neartheplayamargin facies, themudscontainabundantmoldsofwhatwere5­ 10 a mm displacive halite cubes. These molds are now empty but are rimmed with The moldsform anetworkofcubic isolated manganese oxide(?) slain. pores in the mudwhichmaybeupto35%ofagivenvolumeofmud. Handford(1982a) recognized these molds (see his Figure 12), but did not explain their distribution in the text. Moving toward the centerthe playa, the molds get larger and remain filled of withhalite. Inthemostbasinwardareaofthesalinemud-flat,gianthoppercrystalsof halite (Handford, 1982a) can be found, some up to 0.5 m across (Fig. 4.21a). Although the crystals become larger towards the basin-center, they also become less abundant. Large hopper crystals can also be seen in core (Fig. 4.21b), however, the smaller cubic forms only have been found associated with bedded halite and not in marginal salinemud-flatasthey arefoundatthesurface. Acomparisonofatypical sequenceofsaline mud-flatsedimentsfromthecore andsurfacedemonstratethatallofthefeaturesdescribedatthesurface arefoundin core(Fig4.22). Sulfatephases arelessabundantincorethanatthesurface. Thismay beduetothelocationofthecoreorit bethattherewasmoresulfatedeliveredto may thebasininthelast0.5m.y. Becausetheoverallsedimentationratesarelow(see below), its seems unlikely that the present period of inactivity is unique to the history of the basin. Other arid periods must have brought some detrital gypsum into the basin. Therefore,itseemsthelocationofthecoreisamoreplausiblecauseforthelack of sulfate compared with the surface trenches. In support of this hypothesis, deep (10 m)pits excavated byLeslie SaltCompany in the saline mud-flat facieseastof themain trenches have very few sulfate minerals in the muds. Calcite concretions. Small (5-30 mm) calcite concretion form in the center of the basininsalinemud-flatsedimentsatthesurfaceandincore. Exceptforabrief description by Bassett et al. (1959) these concretions have not been previously described or integrated into the stratigraphy of the basin. The concretions are important because theyprovide someinsightinto thediagenetic conditionsunderwhichfluids evolveinthebasin, andtheyprovide furtherevidenceforbacterialorotherorganic involvementintheformationofevaporitemineralsinthebasin. Therefore,because these concretions have not been previously described, some detail is given below. Verylittlehasbeenwrittenonnon-marineconcretions, sothatcomparisonswiththe marine concretion literature are necessary to the discussion. Previous concretion work. In the past, calcite concretions (regular forms) and nodules(irregular forms)havebeen studiedmostlyforestimating compactionand dating of the host rocks. In addition, concretion geochemistry has been studied in relation to the early diagenetic redistribution of sediments (Hudson, 1978; Sass and these studies have been conducted Kolodny, 1972). Unfortunately, almost all of on large(greater than200mm)concretionsformedinmarineenvironments,usuallyina shalehostrock. BristolDryLakecalciteconcretionsareamaximumof20-30mmin diameterandformin anintracontinental playabasinwhichisfilledwith amixtureof Therefore, analogs to marine derived siliciclastic and evaporitic sediments. 81 Figure 4.20. a) Field photograph of the surface of Bristol Dry Lake looking west to AmboyCraterandBristolMountainsinthebackground. Thispicturewas takenafteraminorrainwhichponded somewateronthesurfaceandfilled some wadi channels. Hummocky or "puffy ground" surface is more stable duetocementationbyhaliteandismuch reliableunderfoot.Low more areasinchannels areplaceswherethewatertableintersects the surface.As seenintheshallowtrenchinfrontofthesquatting scientist,thewatertable in the "puffy ground" areas is only centimeters below the surface. The distributionof"puffyground" versus flatareas isvariableandmaybe controlledbyheterogeneous evaportation rates duetovariationsinthe sediement networks, local salinity gradients or slight deviations in the pore elevationoftheplayasurface, b)Close-upviewofflatareathatisinthe of becoming "puffy ground". Initially desiccation cracks form process whichbecomethemostpermeable pathforevaporation oftheunderlying salinefluid. Asevaporationcontinues,saturationwithrespecttohaliteand eventually CaCl2 occurs and is precipitated as the ephemeral white crust seen on the surface. 82 Figure 4.21. a) Photograph of two displacive halite hopper crystals from saline mud flat approximately 5 m below the surface. Competition for space to grow has joined the crystals together, b) Core photograph of a displacive halite hopper crystal from CAES #2. Depth is 357 m. 84 Figure 4.22. Comparison of vertical sequence of saline mud-flat and salt pan sequence from surface and core. Again, one of the main differences is the amount of sulfate minerals present at the surface compared with the core. Besides this difference, all other features are present in both the surface and core stratigraphy except forcomplex polyhedral calcitecrystals which arenot present in the surface salt pan (see below). concretions are best viewed from their fluid flow characteristics and not from their chemical characteristics. Thestudyofcalciteconcretionsandtheirrelationship totheirhostsediments canbetracedbacktoTodd(1896, 1903),andSorby(1908). Bothauthorsnoticedthe relationship between concretion shape and lithology. However, Sorby observed that concretions in sandstones are essentially spherical whereas concretions in shales are elongatealongthebeddingplanesurface. Thishasbeenattributedtorelatively isotropic permeabilities in sandstones versus anisotropic permeabilities in shales in (Raiswell, 1971). Where fluids can flow freely in all directions, as porous sandstones, concretions may grow spherically. Where fluid flow is directed parallel to beddingbyplaty minerals(clays), asinshales,concretionswilltendtogrowin flattenedelliptical shapesparalleltobedding.Evidenceforthistheorycannotbeseen in Bristol Dry Lake. Concretions collected in porous sandy layers are just as elliptical as those collected in clay-rich units. In fact, the growth direction parallel to bedding in thesandyintervalmayevenbemoreexaggerated thanintheclay-richunits. Pantin(1958)renamedRichardson's(1921) threeclassificationsofmarine concretionsfromcontemporaneous, penecontemporaneous, andsubsequentto1) syngenetic; thoseconcretionsformingontheseaflooratthetimeofdepositionofthe enclosing bed, 2) diagenetic; concretions forming below the sediment surface, not long after deposition of the enclosing beds, but in sediments that are unlithified and uncompacted, and 3) epigenetic; concretions formed after compaction and consolidation oftheenclosingsediments. Thisreclassificationisusefulinthatitclarifiesasystemof placing concretions into genetic categories which were formerly becoming vague and misused. In addition, though the classification is genetic, it does place chemical not constraints on concretion formation, only sedimentological constraints. Muchoftherecentliterature onmarineconcretionsdealswithchemical parametersinconcretionformation(e.g.Raiswell, 1971). are Marine concretions commonlythoughttohavesomesortoforganicinfluence. Asearlyas1896,Todd discussed the possibility of an organic influence in concretion formation. Weeks (1953,1957) suggestedthattheanaerobicdecayoforganicsreleasesammoniaand amines diffusing out from the decomposing organic source. This process raises the pH allowing carbonate to precipitate in otherwise carbonate undersaturated What water. little information is available on non-marineconcretions(usually fromglacial settings) typically calls on inorganic groundwater chemicalprecipitates (Dionne andCailleux, 1972). Crystalsizeisrarelymentionedinconcretionliterature,althoughwhenitis,it is generally reported to be between 2-5 fim. SEM photographs ofcrystal morphologies are rarely shown. Observations on hydrochemical conditions necessary for theformationof marine concretions have been studied in marine environments (Dix and Mullen, 1987). However, thesemarinefluidflowmodelsmustbeviewedwithsomecautionwhen applied to non-marine concretions. In marine environments diffusion may be more important than advection due to the fact that concretions are unquestionably forming in a phreatic environment where advective currents are slow. In Bristol Dry Lake, advectionmaybe asimportant,ormoreimportant,thandiffusionduetothenatureof the episodic climate controlled groundwater system influencing fluid movements. In addition, concretion formation maynotnecessarily occurinonlythephreatic environment. Fluctuating groundwater levels, leaving concretions in the capillary or vadose zones, may aid in concretion formation. Occurrenceanddescriptionoftheconcretions. Intheupper10mofsediment in the basin-center, calcite concretions form in both sandy and mud-rich layers. In the core, the concretions are generally restricted to the muddier zones, although in places theconcretions arefoundwithinchaoticmud-halitebeds.Theconcretionsfromthe surface and core may be well lithified, but, the concretions from the core are generally softer than those found at the surface. This most likely is due to the fact that the concretions at the surface have been subjected to surface weathering in the trenches and havedriedout,whereastheconcretionsinthecore arestillwetfromformationwater. Theconcretions fromboththe core andsurface may haveimpressions of cubichalite hopper crystals on the surface and interior of the concretion. In some cases two or three nodules may coalesce into one. Nodules may also have perfectly smooth surfaces withdelicate stellate forms. Examinationofnoduleswithimpressions of haliteshow threetypesof low­minimicrite (Folk; magnesium, non-luminescent, calcite cement(Fig. 4.23): 1) <1 pm 1974)whichcomprises approximately 99%ofthecalcite,2)acicular isopachous rims (c-axis length from 10-50 pm, a-axis length from 5-10 pm) which forms around the voids left by the dissolved halite cubes, and 3) equant and hexagonal(?) crystals up to 40 which also line voids left bythe dissolved halite crystals. The equant and pm hexagonal crystals may simply be the acicular grains viewed looking down the c-axis. However, some of the grains look irregular enough to argue against this hypothesis. Inaddition,if reallytheaciculargrains,theywouldthenbelying the equant grains are Thiswouldbe asomewhatunusual with their c-axis parallel to the surface of the grain. habit for acicular grains. Oftheconcretions examinedpetrographically,only thoseatthesurfacehave acicular and equant spar. The concretions from the core are composed only of minimicrite (micrite < Ipm composed ofMg-calcite). all When viewed at low power the concretions examined to have a clotted fabric which in appear some places is clearly peloidal(0.1-0.2mmlong). Whetherthesepeloidsformabiotically fromflocculation Figure 4.23. Photomicrograph of calcite concretion with abundant inclusions of detrital siliciclastic matrix (small white areas). Dark area is minimicrite and the large cubic white and gray areas are voids which were once halite crystals.Unfilledveinsconnecting halitemoldsthroughthenodulewere the fluid pathways for the halite precipitatiing fluid. This indicates that the nodule was still soft and pliable during halite precipitation. Siliciclastic particles are about 0.1 mm. ofgrainsorfromfecalpelletsisnotcertain. Examinationofsurroundingclaysshows definite pellets most likely produced by brine shrimp when the playa surface was wet. Nobedding,concentriclaminationofthemicrite,or detritalgrainshas orientation of beenseen inthin section orin handsamples. Oneofthemoststriking featuresoftheconcretionsistheabundanceand diversityofsiliciclastic detritalsedimentincludedintheconcretions. Quartz,chert, feldspars (both potassium and sodium), amphiboles and other heavy minerals are all presentintheconcretions. Noneofthedetritalgrainsshowmuchevidenceofchemical weathering (Fig. 4.24). Though the detrital component in the concretions is rarely Figure 4.24. Photomicrograph of a small section of the calcite concretion in Figure 4.23. Dark area is again minimicrite and the small white areas are siliciclastic particles about 0.1 mm in size. Fringing the edge of the concretion and lining halite molds is the isopachous calcite cementwhich is about50 (im in the long dimension.In this photomicrograph, the isopachous cement is fringing the outside edge of concretion. a grain-supported, regularly disseminated detrital siliciclastics vary in the concretions studiedfrom20%to40%ofthefieldofviewinanyparticularplaceontheslide. The abundanceof siliciclastics, particularly in this regularly disseminated fashion, is not welldocumentedintheliterature. Examplesofremnantbeddingandfossils arefound inmarineconcretionsaswellasauthigenic quartzandbarite(SassandKolodny, 1972). However, theabundanceoftheseinclusionsisminorcomparedtothedetrital component in the Bristol Dry Lake concretions. Examples of siliciclastic-rich detrital sediments in concretions are known from both marine and non-marine occurrences. Fossiliferous Miocene concretions from Georges Bank, contain up to 50% framework grains, though they are not grain-supported (Stanley et al., 1967). On the otherhand, DionneandCailleux(1972)reportPleistocene calcareousconcretions(with faulting) thatareonly20%carbonatecement.Thus,thehighsiliciclastic detritalcontentofthe Bristol DryLakeconcretionsis unusualprobably duetolackofdocumentationandnot because it is rare in the rock record. It has been suggested by Folk (1974) that minimicrite requires magnesium-rich chemicalconditionsforittoform. However, theCa:Mgmolarratioinmodempore water samples from Bristol Dry Lake taken by the USGS around 2, and average range from 1.2 to 10. Measured abundances of magnesium average per 0.05 moles liter. Calcium abundances average approximately 0.1 moles per liter. X-ray analyses of two concretions using a silicon standard demonstrate that the calcite contains 1.0 ± 1 mole %MgCOs(i.e.traceamountsofmagnesium,ifany). Therefore,BristolDryLake minimicrite is low in magnesium and may be genetically different from Mg-calcite minimicrite.Folk(1965,1974) alsonoticedabimodaldistributioninmicritegrain sizes with a gap between 2-3 Jim. Bristol Dry Lake micrite is generally less than 1 Jim in size and rarely approaches this "micrite curtain". Certainly, in Bristol Dry Lake the chemical conditions underwhich the concretionsformedwerenottypicalofothercitedexamplesofconcretion occurrence. Thepresenceofhaliteimpressions onthesurfaceoftheconcretions, withanetworkof fracturesconnectingtheimpressionswithintheconcretionssuggests thatthecalcite concretions formed contemporaneously with water saturated with respect to halite. UndertheSEMtheminimicritegrainsexhibitpolyhedralcrystal habitsinaplaty fashion (Fig. 4.25) which is not typical for low-magnesium calcite. Folk (1974) does show similar platy morphologies from freshwater low-Mg calcites from Texas and West Virginia. However, the crystal sizes of these examples are much larger (>5 |im) thanBristolDryLakeminimicriteandmayinfact,bemore analogous tothehexagonal crystals seen in thin section. Complex polyhedral dolomitecrystals have been documentedbyNaimanetal. (1983) fromPermianhalitebedsofTexasandmore recently by Gregg and Hagni (1987) associated with CambrianMississippi Valley-type sulfide ore deposits of Missouri and Gao et al. (in review) from the San Andres the Texas Panhandle (see below). In the first two Formation of the Palo Duro Basin of sulfate-rich brines are invoked to explain the crystal examples, chloride or 91 morphologies. However, little chemical evidence is given is support this hypothesis. Thepossibilityofsulfatepoisoningormorelikelychloridepoisoningnucleation sitesis one possible reason for the strange morphology of the crystals in Bristol Dry Lake, and also their small size. Clearly, the impressions of halite crystals on the surface of the concretions indicates a penecontemporaneous origin with halite saturated waters. Bristol Dry Lake calcite concretion Figure 4.25. Secondary electron image of a showing platy crystal morphology. Grain-size appears larger than minimicrite,but,undercloseexamination,itisevidentthattheplates are madeupofsmallerminimicrite-size subplates. Scalebarinlowerleft comer is 10 |im. The isopachous acicular crystals have typical low-magnesium calcite morphologiesand areobviously alatercementgeneration,growingintothevoidsleft by the dissolved halite. The equant and hexagonal microspar also falls into this category. Formation ofthe concretions. From the description presented above it appears that the calcite concretions form from halite or close to halite-saturated water. The concretions are only moderately displacive in that they incorporate a large portion of the detrital matrix into the concretion. It is difficult to reconcile displacive growth which incorporatedmatrixwithgrowthinunconfinedsediment.The of occurrence concretions along mud partings suggests that they form after at least partial compaction the sedimentbut are best described of as diagenetic rather than epigenetic because the sediment is still unlithified. In some cases the inclusion of matrix be due to the may factthattheconcretions arepartiallydisplaciveandpartiallyvoidfillingcement.The flow along bedding is active during concretion formation. The stable isotopic signature of Bristol Dry Lake concretions supports the hypotheses thatthereis an organicinfluence ontheformationoftheconcretionsand thattheyformbyevaporiticprocesses. Thestableisotopicsignatureoftheconcretions stellate forms and elongation parallel to bedding indicate that some component of isanimportantcluetounderstanding theevolutionofthefluidsinthebasinandwillbe discussed later (see isotope section). Porogenesis.Thesalinemudflathasarelativelysimpleparagenesis. Although therelativetimingofcalcitenoduleandsalthoppercrystal precipitation isnotalways simple, thepresenceofcalcite noduleswithimpressionsofhalitecubesonthesurface indicatesthat,attimes,theyformcontemporaneously witheachother. The consistently heavy oxygen isotopic signature of the concretions (see below) further suggests thatcalcitealwaysprecipitates inverysalinewater,probablynearhalite saturation. Besides the precipitation of halite hoppers and calcite concretions, very tittle occurs in the saline mud-flat. Some gypsum form diagenesis or anhydrite may some enterolithic structures, and some fibrous displacively in the sediment producing halite may cement fractures, but these phases are minor. Salt pan Relative to the other facies common to playa and marine evaporite settings, the salt pan has received the most attention (see Schreiber, 1988; Warren, 1989; and LowensteinandHardie, 1985).Thisisprobablyduetothebeliefthatthesaltpanisthe reveal themostdetailedinformationaboutthe mostdistinctfaciesand socan environment of deposition. Hardie (1984) has shown that the salt pan in itself is not diagnostic of its origin, and subsequent studies have also shown similarities in fabrics andgeochemistrybetweenmarineandnon-marinesaltpans(Hovorka, 1987; Lowenstein, 1988). Therefore, it is important to study the textures, fabrics, and geochemistry ofmodemsaltpanswithinthecontextoftheentirebasinandtocompare the collected evidence with buried samples which have undergone significant diagenetic alteration. Inthiswayimportantdiagnosticfeaturescanbedeterminedforancient examples which have undergone various stages of diagenesis. Bristol Dry Lake presents a unique opportunity to make such a comparison. Although the modem salt pan in Bristol Dry Lake is quite thin (< 300 mm), the fabrics are almost pristine and so provide a good comparison to the halite preserved in the trenches and core. As far as I know this is the only modem halite forming at the surface that has significant accumulations of cored halite at depth from the same basin. This core provides an excellent control on the timing of the diagenetic processes affecting halite and allows direct analysis of the diagenetic processes which alter the pristine depositional fabrics. Many of the features in the salt pan at Bristol Dry Lake have beenrecognized seen inothermodemsaltpans. LowensteinandHardie(1985)describemanyofthe importantevaporite structuresanddissolutionrelationships inmodemsaltpansinboth playaandcoastal salinasettings. Forancientenvironments, Hovorka(1987) demonstratedthatmanyofthedissolutionfeatures seen inmodemsaltpanscanbe recognizedintheancient.Thefollowing descriptionofthemodemBristol DryLake salt pan will be compared with fabrics seen in core. This approach is similar to LowensteinandHardie(1985), exceptthatinthisdiscussion,boththemodemand altered fabrics from the cores are from the same basin. This allows for fewer variables inapplying theinsights gainedfromthecorestotheancientexamples. Theinterestedreadernotfamiliarwiththe halitefabricterms usedbelowwill find excellentexplanations inShearman(1978). at the surface of the basin, one on the east side Currently, there are two salt pans ofthebasinandoneonthewest(seeFig.4.8a). Accordingtolocalresidents,water has been ponding in the east salt pan only since 1982. Both salt pans are composed of almost pure layers of 10-20 mm-size, vertically elongate chevron halite, separated by millimeterbandsofdetritalsiltormud(Fig.4.26a,b). Thehaliteis99%pure, although both pans contain approximately 30% porosity. Haliteformsinthesaltpanaspyramidalhoppercrystalsorraftsthatprecipitate in thebrineandsinktothebottomofthepan, bottom-nucleatedcrystalsofchevron or as halite. Thebottom-nucleatedchevroncrystals produce thevertically aligned crystals which is the dominantfabric in the modem salt pan. It is difficult to the preserve delicateraftsofthepyramidal hoppercrystals because astheyfalltothebottomofthe brine pan they do not pack well and may break become disrupted with burial. They or arethensubjectedtoextensivediagenetic processes.Thus, lessthan1%ofthesaltpan consists of pyramidal hopper crystals. However, Hovorka (1987) has shown that theserafts havebepreserved, incertaincases,inPermiansaltfromwestTexas. Halite reaches about0.4 in thickness in both However, the west salt pan m pans. contains approximately 30% chevrons and 40% clear diagenetic halite, whereas, the east salt pan is the reverse. This suggests that the west salt pan has undergone slightly morediagenesisthantheeastsaltpanandso,maypossiblybeolder. Handford (1982b) statedthattherewas nobrinepanatthetimehedidhis researchatBristol Dry brinepanhalite atthesurfaceoftheplaya isLake in 1979-80. This indicates that no olderthanabout8or9years. Thelackofathicksaltcrustatthesurfacesuggeststhat the present climate is too arid to supply enough water to generate thick accumulations of salt. The implication of this statementis that the thick accumulations of salt seen in climate. This interpretation will be discussed in more detail core must form in a wetter in the paleoclimate section. Themoststriking featuresofthesaltpanare largehalitetepeestructuresupto 0.6 m tall forming in a regular polygon pattern reminiscent of mud crack patterns (Fig.4.27a). Tepees form by the force of crystallization of the halite as it precipitates out of solution and exerts congressional pressure on neighboring crystals ( Lowenstein and Hardie, 1985; Dulhunty, 1987). Tucker (1981) suggested that the initial crack is producedbythermalexpansion, butLowensteinandHardie(1985)demonstratedthat dominantattheveryearlieststagesoftepeedevelopment. In compressional forces are Bristol DryLake, halitetepeesleveledby raininthewintermonthshavebeen observed to reform almost immediately upon desiccation of the pan surface. The daily were not above 30 °C. Although thermal expansion temperatures at this time 95 Figure 4.26. a) Hand sample photo of chevron halite from the modem salt pan. Chevrons are defined by the fluid inclusion-rich bands. Notice the truncation surfaces caused by dissolution across the top of the layers of chevrons, b) Photomicrograph of the same slab showing vertically aligned chevrons. Notice that the fluid inclusion rich bands are in places discontinuous, and clear diagenetic halite has filled the area. Photomicrograph is 3 mm across. 96 may play some role in the development of tepees in the summer months, it seems unlikely that thermalexpansion is necessary for tepee formation. Intheintitialstagesoftepeeformation, thesaltcrustwillatfirstwarporbow upwardandtheneventuallycrack(Fig4.27b). Dulhunty(1987)observedthatinthick salt crusts (> 500 mm) when the halite warped up, the salt would migrate by dissolution-reprecipitationfromthetopsoftheupwarpstothelowpointsproducing a salt layerofirregularthickness. Unfortunatelyhedidnotdescribethefabricofthis reprecipitated salt. Withcontinuedprecipitationofhalite,thesheetswilloverthrust eachothercreatingthetepeestructure. Groundwaterbrineresurgesalongthesecracks and precipitates poorly oriented chevron or clear halite inthe space, helping to displace thesheetsevenfurther(Fig.4.28). Theformationofhalitetepeesissimilartothe processes which form carbonate tepees (Dulhunty, 1987; Kendall and Warren, 1987). However, the fabrics associated with halite tepees are not as varied as those associated with carbonate tepees, and in addition, carbonate tepees are much more likely to be preserved than halite tepees. For example, at Bristol Dry Lake, after large storms, sufficientwaterundersaturatedwithrespecttohalitepondsonthesalt and pan dissolvesthetepeesflatteningthesaltpantoabilliardtablesmoothsurface. Although subsequentevaporationofthewaterandprecipitationofnew halitecreates new tepee structuresalong thesuturesoftheoldtepees,thesucceeding floodsdestroy these not beenobservedin tepees as well. In addition, unequivocal tepee structures have core. One reason that they have not beenobserved is that the lateral relationships for therecognitionofthetepees themselvesaremuchlargerthanthecorediameter. However, tiltedmudpartings inhalitebedsandobliquely orientedhalitechevrons suggest that, in rare cases, some tepees may be preserved. another feature that is Knobs of salt-encrusted organic matter are common on the branchesof thesurface salt pan (Fig 4.29a). The organic matter, usually twigs or blownorwashedinto thebasincenter during storms, forms halophyte bushes that are thenucleusfromwhich thehalitegrows. Thehalitegrowsasaradiating isopachous cement itself to the halite crust. The relief on rimaround the organic matterandmay these knobs may be up to 100 mm and they may be 50-70 mm in diameter. Because theseknobsstandinreliefabovethesaltpansurface, they arelikelytobethefirst filled with There was not enough water from this storm to completely level the tepees, but other storms have completely dissolved die tepees and leveled thesaltpansurface, b)Close-upviewofthebowedupsaltcrust.With continuedprecipitationofsalt,diesheetofsaltwilloverthrustacross the crack propagating along the back of the bow. Figure 4.27. a) Halite tepees in the west salt pan water in betweentepees. 99 Figure 4.28. Field photograph of multiple episodes ofsalt extrusion along an incipient tepee. Groundwaterupwellsandmigratestothesecracksbecausemudis more permeable andthe evaporation rates are greater. features to be dissolved (along with the tepees) with the next influx of fresh water, and soarenotlikelytobepreserved. However,iforganicmatterwereblownintoahalite­saturated brine pond that was 1-2 meters deep, it is possible that these knobs could intohalite"reefs". Relief theorderofmeterswouldbehardertodissolveand on grow bepreserved. Halitereefsforminthecenterofmanyoftheevaporationponds floodedbythesaltcompaniesthatoperateonthelake(Fig4.29b). Althoughtheorigin of these reefs is not known, it is likely that they formed due to some topography on the bottomofthebrinepan,possibly abranchoratwig.Thetwigorotherformofrelief may acted seed for halite nucleation. Because the surface area on the obstruction was as a outward and upward more the smallcrystalsgrew rapidlythanthesmoothbottomof pan, creating the "reef. Itfollowsthatifthesefeaturesexistatthesurfaceandcangrowtorelatively large size,theymaybepreservedintheancientrecord. Todatenohalitereefshavebeen foundinBristolDryLakecore,oranyotherhalitecore However,Idoubtifanyone knew what criteria to usetoidentifythesefeaturesinMorethorough core. examinations ofhalite cores may eventually turn up one of these halite reefs. Dissolution.Thedominantdiagenetic alterationof haliteis dissolution and karsting of the salt pan and the subsequent precipitation of clear diagenetic halite in the void space. Therefore, the important dissolution features seen at the surface will be discussed in somedetailandcompared withtheinterpreted dissolutionfeatures observed in core. Themostabundantdissolution featureofthe saltpanis the creationof secondary porosity between the primary chevron crystals. Because halite grows so rapidly and the lateral compressive force is enough to buckle and break the crust into tepees (as discussed above), very little primary intercrystalline porosity is present. Therefore, the 30% porosity observed in the salt pan must be almost exclusively secondary. Thesecondaryporosityisgeneratedbythedissolutionofchevronhalite aftertheinfluxofhalite-undersaturatedwaterontothepansurface. Theundersaturated water filters downmicroporous openings generally along halite grain boundaries and widens the openings into vugs, deeply embaying the crystals (Fig. 4.30). New poorly­orientedchevrons(i.e.chevronsnotgrowingvertically) maygrowintheporespaceor cleardiagenetic halitemayform. Theformthattheporefillinghalitetakesdependson thenucleationandgrowthrateofthehaliteinsolution. Slowgrowthratesandlow nucleation density promoting large clear crystals of diagenetic halite will be favored in solutions that are only slight supersaturated with respect to halite. Chevron halite will be favored by rapid growth from a highly supersaturated solution. After minor rain which left only a Dissolution may occur on many larger scales. the surface of vertical dissolution tubes millimeters of water ponded on the pan, were developed throughout the salt pan a in approximately 5-10 mm in diameter seemingly random pattern (Fig. 4.31) The tubes generally extend all the way through the salt crust and are not associated with tepees. more water has been ponded on When at the surface. the surface, dissolution pits up to 0.3 meters across form In this case, Figure 4.29. a) Field photograph of the modem salt pan showing knobs of salt that protrudeabovetheflatsurface. asmalldepressionfilledwith To the left is brine. It is doubtful that the two features are genetically related, b) Field photograph of halite "reefs" from one of the salt company evaporation pans. Thelarge"reefinthecenterofthepictureisabout2macross. Although these "reefs" are much larger than the knobs seen on the natural saltpansurface, thewaterismuchdeeperintheevaporation pansthanin the natural salt pan. 103 Figure 4.30. Photomicrograph ofchevron halite growing both vertically and horizontally. Notice the curved deeply embayed edges of the crystals that creates abundantsecondary porosity. This secondary porosity is rapidly filledwithdiagenetichalitewithin2mofburial. Photographisabout3 mm in the long dimension. the halite maybedissolvedfrombelow(fromresurging freshgroundwater) sothatthe surfacehalitecracksandsinksintothemudbelow(Fig.4.32). Thesepitsmay thentrapmudsettling fromthewatercolumnandproducegeopetal structures.The outlineofdissolutionpits(Fig. 4.32) filledwithclearhaliteareabundantinthe core seen andprobablyrepresentthefinalproductofpitssimilartothose inthepresentsalt surface (Fig. 4.31). Wholesale dissolution and brecciation of the salt be presentinthepreserved halitefromcore,butit hasnotbeenobservedon themodem salt pan pan may pan. Although the salt pan halite is only tens ofmillimeters thick at the surface, 3 metersbelowthesurfacethereis a 1m thickhalitebed. Incore,relativelypure(80-90 %) halite may be tens ofmeters thick (Fig. 4.33). The importance of diagenetic dissolution-reprecipitation reactions in the salt pan can be seen in the porosity Figure 4.31. View looking down on a dissolution pit forming in modem salt pan. The pitwillmostlikely fillwithcleardiagenetichaliteandlooklikeFigure 4.32in cross section. distribution with depth. Porosity decreases from 30 to 40% in the modem salt pan to 6%3metersbelowthesurface,adecreaseofapproximately80%. Inthecore,halite porosity averages about 2%. Although chevrons are present in core (Fig. 4.34a) diagenesis has obliterated almost all traces of primary fabric (Fig. 4.34b). Even the layer 3 meters deep has almost no primary halite fabric remaining (Fig.4.35). Complete recrystallization of the halite into clear interlocking cubes is the rule rather thantheexception forallofthecorerecoveredfromBristolDryLakeandtheshallow subsurface halite. Such complete recrystallization of the halite at shallow depths indicates that the haliteremainsin thegroundwater zonefor arelatively longperiodof time. Preservation of halite fabrics is actually better in Permian aged rocks from West Texas (Hovorka, 1987; Lowenstein, 1988) where subsidence rates were perhaps faster or groundwater salinity more stable. Figure 4.32. Core piece from CAES #2 showing large dissolution pit (white area in the middle of the piece) in the light of a window. Slab is about 0.2 m long. Onecommonhalitefabricseen incore thathasnotbeenobservedatthesurfaceis fibrous halite filling vertical fractures (Fig. 4.36a). Hovorka(1987) describes this fabricintheSanAndresFormation(Permian)ofthePaloDuroBasin. Shedoesnot, however, discuss its origin or timing. Because the muds in Bristol Dry Lake are still unlithified, it is obvious that fracturing is not restricted to late post-lithification burial. As long as the muds are competent, halite may fill the fracture. At the surface, similar horizontalvein-filling halite hasbeenobservedassociatedwith desiccationofthesaline mud flat (Fig. 4.36b). This indicates that large stresses are not necessary to form the fibroushalitefabricandthatthefabricmaybeemplacedearly. Itthenfollowsthat vertical fibrous halite should not be used as an indicator oflate halite remobilization. Figure4.33. Corephotographofbeddedhaliteshowingpurityandlackofobvious structure. Clear diagenetic halite (arrows) fills dissolution pits. This m thick. particular bed is almost 4 From CAES #2, 445 m depth. Core box is 0.6 m in the long dimension. CorrelationofsaltbedsincoresfromBristolDryLakeisnearlyimpossible. As an example, CAES #1 and CAES #2 are shown in Figure 4.37. No beds can be correlated, andyetthesecoresare less than3km apart. TheCorrelationsinCAES#2 also be difficult due below 300 m depth may to structural complications (see below). Correlationwiththelithic logsofothercoresshowninFigure4.37(oneofwhichwas same spot as CAES #2) also was unsuccessful. This may be taken virtually in the Figure 4.34. a) Photomicrograph ofprimary vertically aligned halite chevrons from CAES #2, 360 m depth. Photomicrograph is 10 mm across, about the same size as the chevrons from the modem salt pan in Figure 4.26b. b) Secondary electron image of the surface of interlocking halite crytals. Sample is from CAES #2, 415 m. There is less than 1% porosity in this sample and no primary fabric remaining. 109 partly due to different interpretations by the individuals logging the core as to what actually constitutes a halite bed. But even with moderate variations due to logging techniques, the patterns of closely spaced cores are not similar. Amodelofthe 30mofsaltwasconstructedbyLeslieSaltCo.(Fig.4.38). upper In this model, even with closely spaced well control, the correlations are tenuous. The bedsappeartothickenandthinrapidly, arediscontinuous, andtheyallseemtohavea regionaldip. Thediscontinuousnatureofthebedsandthelocalthickeningand thinning indicatethatthe areaofthebrine ponds thatformedthesesaltbeds was relativelysmallandthatthepondwasshallow. Thiswouldexplain thedifficultyin correlationofthesaltbedsfrom fewkilometersapart. cores a below the m Figure4.35. Interlockingcubichalitefromthebeddedhalitelayer3 surfaceoftheplaya. Thebedhasabout6%porosityremaining,and only 2-3% primary fabric is preserved in this bed. Thehydrologyofplayabasins accountsforthelateral variabilityinevaporite beds. Brine pans will form where ever the low point (or points) in the basin may be. Figure 4.36. a) Core photograph of fracture filling halite. This type of halite is common just above or below halite beds, but it may also occur in mud­richzonesunassociatedwithobvioushalitebeds, b)Sametypeof fracture fill halite at the surface as seen in core, except that at the surface it is oriented horizontally rather than vertically. Both the core and surface fracturefillhalitehavethesame internalfibroushalitefabic seeninthis sample. 112 Figure 4.37. Simplified lithic logs of 5 cores showing halite beds (white areas) versus areasdominatedbysiliciclastics(blackareas).Thetop15 20mof - CAES#1isbasaltfromAmboyCrater. Thepointofthisdiagramisto showthecomplete lackofany correlation betweenhalite bedsin the wells.Noneofthesewell of are more m than 5 Km apart. The first 150 CAES #1 (17) and #2 (16) are based on drillers logs of cuttings the rest is logged core. Br-1 ( 13) and Br-2 (14) are based on lithic and geophysicallogslogs form1979USGSwells(Calzia, 1979).Bristol2 (15) is based on 52% core recovery of a USGS well drilled in 1959 (Basset et al, 1959).Numbers in parentheses correspond to numbered locationsinFigure2.1andappendixIV. Unlabelledtickmarks arein 100 ft. intervals. 114 Two salt Thepresent surfaceofBristolDry Lakeis agoodexampleofthisvariability. pans separated by saline mud-flat deposits are forming concurrently on the playa. At other times only one pan may form. Differential sedimentation or compaction could alignbedssuch thatthedifferentconcurrentsalt panscouldbecorrelatedwithpans that formedatadifferenttime.Thevertical sequenceofsalinemud-flatsedimentsandsalt panhaliteillustratestherepetitiousnatureofthebasincentersediments(seecore logsin appendix I).Therepetitionofsaltandsiliciclastics inthebasin-centermaybetermed cyclic aslong astheconnotationofregularityofthecycleisnotincluded. Repetitions inthe cycle may be related to local, regional, or global climatic variations, tectonics or of the halite beds from thebasin-center. Mudisshownasclearareas. Eachpostrepresentsdrill control.Themodelrepresentsabout5kmlaterally. Noticethedipofthe beds, the thickening and thinning of the beds, and the lateral discontinuity of the beds. Permisson to reproduce the model is gratefully acknowledged. Figure4.38. ModelmadebyLeslieSaltCo.oftheupper30m some other unknown source or combination of sources. Therefore, the term cycle is used with some caution until the controls on the cycles are better understood. Chaoticmudsalt.Oneofthemajorunresolvedproblemsinevaporite researchis theoriginofbeddedhalitethathas undergone sufficientdiageneticalteration sothatall primary fabrics have been obliterated (This is also true for gypsum/anhydrite beds). Thequestion is,howmuchsedimentcan halitedisplacebeforeexternalpressures clear that beds (lithostatic, hydrostatic, etc.) prevent additional displacement? It seems of diagenetic halite that are 90-100% halite started out as primary bottom-nucleated chevronsandpyramidalhopperrafts,andbedsofunderabout30%halitestartedout as displacivehalite. Butwhereisthecutoffandhowcanoverprintingbedistinguished? The answers to those questions are not simple, and it is not clear that they can ever be resolvedby sedimentologic orpetrographic criteria. Figure 4.39. Interlocking cubic halite forming chaotic mud salt in CAES #l. Core box is 0.6 m in the long dimension. For example, it is clear from the purity of the halite deposits and rarely preserved chevrons, that the halite beds in CAES #2 where originally deposited in a brinepond. However,inCAES#1thehaliteisroughly a50%mixofmudandsalt which contains noprimaryfabric. Inaddition,insteadofoccurringindiscretebedsas inthecenterofthebasin,allthehaliteinthecoreoccursinasingle 100minterval (appendix I). The crystals are large clear hoppers or cubes that in places interlock and elsewhere aresurroundedmostly bymud(Fig4.39).Thistypeofhalitewouldfallin thechaoticmudstonehaliterock(class D)ofHovorka(1987). She modified the explanationofHandford(1982a)toarriveattheobservedfabric. Smith(1971) originally proposed that the chaotic mudstone fabric seen in boreholes from the English Zechstein Basin (Permian) could be deposited by three possible mechanisms; 1) mudstoneredistributedintohaliteduringrecrystallization ofbrinepondprecipitated halite,2)Simultaneousdepositionofmudandhalite,3)anddisplacive halite introduced into the sediment. Smith (1971) favored the last mechanism to explain the observedfabrics. Handford(1982a)proposed thatacombinationofthefirstandthe lastmechanismsexplain theoriginofthechaoticmudsaltinBristol DryLake. However,hismodelwasdesignedtoexplaintherelativelypure,diagenetically altered, basin-centerbeddedhalite units,andforthosebedshis modelisprobably accurate. ForthechaoticmudsaltinCAES#l,itisdifficulttodefinitively statewhich mechanism(s) is dominant However, acombinationof sedimentologic evidence pointstoadominantlydisplaciveoriginforthisfabric. 1)Thecoreislocatednearthe playa margin so that it is less likely that halite saturated water would pond there, particularly for such a long time. 2) The halite occurs in only one stratigraphic interval that is 100 thick. 3) No primary fabrics have been observed. 4) Abundant over m inclusions of mud and intergrown ash in the halite indicate a displacive origin for much ofthehalite.5)Thehalitedoesnotshowanydepositionalbedding. Separately,these observationsprovidelittle evidenceofadisplacive origin fortheinterval,butin combination, they provide solid criteria for a displacive origin for the observed fabric. of the saline mud flat The thick interval of chaotic fabric probably represents the area justmarginwardofthebrinepan. Here,thegroundwaterbrinewouldbealmost constantly saturatedwithrespecttohaliteandthefinemuddysedimentswouldbeeasy for the halite to push aside. Complex polyhedral calcite. Silt to sand-size multifaceted luminescent calcite crystals form between the crystal boundariesof halite crystals throughout the subsurfacehalite(Fig4.40a)below100m limitof (the upper the core). This type of calcite has not been found at the surface or in the halite layer 1 m below the surface. Theupperlimitis notknownbecauseofthelackofcorefrom theintervalbetween2 metersand100 m. Themultifacetedcalcitemakesupabout1-2%ofsomehalitebeds and is concentrated near mudstone interbeds, particularly inthe chaotic halite fabrics. Themultifacetedcalciteissimilarinmorphology andoccurrence to the multifaceted orcomplexpolyhedral limpiddolomitedescribedbyNaimanetal.(1983) fromthehalitebedsintheUpper ClearForkandGlorietaFormations(Permian)of west Texas and the limpid dolomite described by Gao et al. (in review) in the San AndresFormationofthePaloDuroBasin(Permian)ofwestTexas(Fig4.40b). Based on petrography, stable isotopic analyses, trace element chemistry, and fluid inclusion data,Gaoetal.(inreview)andNaimanetal.(1983) explainedtheformationofthe limpid dolomite as the latest diagenetic phase inthe halite. The dolomite formed by the flushing of meteoric water into the evaporite-precipitating system during deposition and early diagenesis of halite. Dissolution of halite by the meteoric water eventually produced a halite-saturated, but 18G-depleted, brine which was preserved in mudstone interbedswithinthehalite.During burialcompaction, thebrine released from the was mudstones and moved along halite crystal boundaries, the contacts between halite crystals,andbetweenmudstoneinterbedstomixwithbrine trappedinhalite,resulting in the precipitation of the limpid dolomite. They also suggested that bacterial sulfate reduction could provide the bicarbonate necessary for dolomitization. The occurrence thelimpiddolomiteisalsoassociatedwithmudstoneinterbedsandit isalso of luminescent (Gao et al., in review). Thesimilarityinmorphologyandoccurrence betweenthePermianlimpid dolomites and the Bristol Dry Lake multifaceted calcite is remarkable (compare Figs. 4.40a andb). Theonlyrealdifferenceintheirformationisthefactthat intheBristol Dry Lake occurrence, the precipitated carbonate phase is calcite rather than dolomite. The for this is simply because there is little magnesium available in Bristol Dry reason Lakesothattheconditionsfordolomiteformationcannotbeattained(seechapter on limpid dolomite given by Gao et brinechemistry).Theexplanation fortheformationof Figure 4.40. a) Secondary electron image of complex polyhedral calcite in Bristol Dry Lake from about 500 m depth in CAES #2. Scale bar is 100 |im. b) Secondary electron image of complex polyhedral dolomite from the Permian. Scale baris 50 p.m. Photograph, courtesy Dr. R.L. Folk, is partofFigure 6inNaimanetal. (1983). 120 al.(inreview) isequallytenablefortheformationofthemultifacetedcalciteatBristol Dry Lake. Periodic influx of freshwater saturated muds are common at Bristol Dry Lake and evidence for early dissolution and diagenetic replacement of chevron halite is abundantboth at thesurfaceandincore. the calcite has Although the geochemistry of notbeendetermined, itprobably wouldnotbethathelpfulbecause thereare nomarine baseline comparisons touse. Comparison of oxygen and carbon isotopic values of the concretionversus themultifacetedcalcite may provide some insight in the origin of the latter,buttheseanalyseshavenotyetbeenperformed. Inconclusion, theoriginofthe multifaceted calcite is still equivocal until some detailed geochemistry is performed, however,theexplanation offeredbyGaoetal.(inreview)andNaimanetal.(1983)for thePermianlimpiddolomitebestexplains theoriginofthemultifacetedcalcitein Bristol Dry Lake. Theonlymodificationoftheexplanation isthatitseemsprobablethatbacterial sulfatereductionplayed akeyroleintheformationofthemultifacetedcalcite because thebrinechemistryofBristolDry Lakeis bicarbonate-poor(seechapter onbrine chemistry). To achieve calcite saturation, bicarbonate mustbe derivedfrom somewhere. Someofitmaybederivedfromthedewateredmuds,butbecausewe knowcalcite isformingfrombacterial sulfatereductioninotherpartsofthebasin(see playa margin section), there is no reason that it could not be a factor here. Small areencounteredinthebrine facieswhich quantities of anhydrite pan may provided sufficient bicarbonatetomakethereaction go. Paragenesis. Although thediagenetic mineral assemblage is fairly simple in the saltpan,haliteanda amountofanhydriteandcalcite,thediagenesisof minor the halite thatresults in theobservedfabricmaybe so complexthatitmaynotbe possible to reconstructadetailedparagenesis. Diageneticalterationsofthehalite,dissolution­reprecipitation, karsting,andtepeeformationalloccurveryearly,withinthethe top fewmetersofburial. Porosity islostearly,andprimaryfabriccanbetotallyobliterated Mobilization and within vertical open fractures at depth, but similar byclear,slowgrowingdiagenetic halitecementsandreplacements. reprecipitation ofhalite may occur fabricsinhorizontalfracturesatthesurfacesuggest thattheverticalfractures arefilled to tectonics. Some halite-filled fractures relatively early and are not necessarily related that are at 60° to bedding are probably later, tectonically induced, fractures. The latest diagenetic phase istheprecipitationofmultifacetedcalcitearoundhalitegrain boundariesassociatedwith thechaoticmud-salt. Facies distributions and stratigraphic framework The distribution of evaporite minerals into concentric rings around the basin (Fig. 4.8a) implies a simple evaporation path of a relatively homogeneous groundwater as it moves toward the basin-center. The consistent "bulls-eye" pattern distribution of the evaporitemineralsisprobably duetoadiffusegroundwaterinput.Thelackofsprings and tufa deposits also supports a diffuse groundwater input model. The saline mud­flat,betweenthegypsum-celestite zoneandthebrinepan,isdominatedbysiliciclastic muds with only ephemeral displacive halite cubes and hopper crystals present as chemicalminerals. Theseparationoftherelativelythingypsumzonefrom subaqueously deposited halite is a lateral characteristic of Bristol Dry Lake. The cause of this separation is most likely dueto the limited availability of sulfate and the fact that thegroundwatersalinity mustincreaseapproximately 4-foldfromgypsum saturation before it reaches halite saturation (Holser, 1979b). By the time this has occurred the waterhasbeenpondedatthebasin-center. Onlysmallamountsofsulfateminerals (gypsumoranhydrite) arefoundinthesalinemud-flatandbasin-center,andmuchof this is probably wind or water transported from the deflating playa margin sediments. Incore,separation ofthegypsumandhalitezonesis alsoevidentinvertical succession. Inthecorestakeninthebasin-center,brinepananddisplacivehalite alternates with muds from the saline mud-flat deposits for over 500 m. No appreciable accumulationofsulfatemineralsexistsinthesecores. Intheonecoretakennearthe margin of the playa (CAES #1), vertically aligned anhydrite fabrics, pseudomorphing gypsumfabricsatthesurface,alternatewithsandsandmuds. Haliteisonlyfoundin This interval (from 150-260m) this core at one significant interval (appendix I). containssubequal amountsofmudandhalite,mixedintoachaotic texturewhichlacks primary fabric (see above). This halite is interpreted as forming displacively and is part ofthemostbasinward salinemud-flatfaciesandisnotpartofthebrinepanfacies. Therefore,itappearsthatinanygivenverticalsequenceitisunlikely thatgypsum will be overlain by halite (or vice versa) as one might expect in a normal prograding typeofsequence. Figure4.8 b shows anidealizedcross-sectionthroughthemiddleof the playa (see Fig. 4.8 a for location) to illustrate the facies and mineral distributions. AlthoughHandford's (1982a)subenvironmentterminologyisretained, theverticaland lateral facies relationships shown here differ significantly from his reconstruction in that most gypsum is precipitated in a narrow bandin the playa margin sediments and notinthesalinemud-flat. Handford(1982a,b)observedthat andanhydrite gypsum arepresentthroughoutthesaline mudflat,although healsoobservedthatthey areless abundantintheplaya-center.However,theamountofinsitu gypsumformedinthe salinemud-flatisextremelysmall. Mostofthegypsumwaseitherwashedorblown into the playa-center. In addition, the vertical gypsum fabrics observed in the playa margin gypsum-celestite zone are characteristic of this facies. Distinct gypsum fabrics and distribution are important criteria for reconstructing both the hydrology and vertical faciesrelationshipsof BristolDryLaketoother thebasin, inordertocompare evaporite settings. In many ancient evaporite basins interpreted to be marine, the vertical facies distribution usually starts with a basal normal or restricted marine carbonate unit overlainbygypsumoranhydrite andcapped byhaliteandinsomecasesbittern salts. Thiscycleischaracteristicofmarine restrictedlagoonsequencessuch or as parts of the PermianSanAndresFormation(Hovorka, 1987)andSaladoFormation(Lowenstein, 1982) in west Texas. Conversely, in continentally influenced "sabkha" sequences such astheClearForkFormation(Permian)inwestTexas (Handford, 1982c)andin such astheColdLakesubbasin(Devonian)ofAlberta ancient playa sequences (Kendall, 1988),theverticalsequenceconsistsofcarbonate,overlainbyanhydrite whichisinturnoverlainbysiliciclastic mudandthenhalite. Theseparationofthe anhydrite(interpretedtobesubaerialsabkhatypeanhydrite intheClearFork)andthe halite by a mud-rich unit is exactly what is seen in Bristol Dry Lake. In cases where the halite directly overlies the calcium sulfate phase, it is generally interpreted to have precipitated from a brine pan. Therefore, it seems likely that the vertical facies distribution can provide a significant clue as to whether the sequence is a marine or non-marineevaporite whenlittle is knownaboutthe geometry of the basin. Summary "bulls­eye" distributionofevaporite mineralswithinfourbroaddepositionalenvironments. The investigation of surface outcrops from Bristol Dry Lake reveals a Calciteintheformofbox-workcalcreteandpedogenicnodulesis thedominant authigenic mineralphaseinthemidtodistalalluvialfanfaciesandthefanwardsideof theplayamarginfacies. Basinwardofthepedogeniccalcite,theplayamarginfaciesis dominatedby anarrow bandof vertically aligned displacive gypsum andcelestite nodules.Large areasoftheplayabasinwardoftheplayamargininthesalinemud-flat faciesaredominatedbydetritalsiliciclastic mudwhichcontainsabundantdisplacive halite and common calcite concretions. Finally, in the basin-center, subaqueously theformedbrinepanhaliteformsinephemeralbrines. Therecognitionthatthe bulkof isconfinedto anarrow bandaroundtheplaya margin separatedfromthesalt gypsum pan halite by detrital mud is important in reconstructing vertical facies distributions of the basin. Comparison of the surface facies and mineral distributions with features seen incoresdemonstratethat: 1)thebeddedhaliteisneverdirectlyoverlainorunderlainby gypsumbutinsteadalternateswithdetritalmudin thebasin-center.2)almostall features seenatthesurfacecanbe seeninthecore,and3)mostofthediagenetic reactions occurwithinthetop fewmetersofthesurface. environments CHAPTER 5. Fabric comparison of gypsum depositional The classification of gypsum fabrics into either subaerial or subaqueous depositional environments has been discussed by numerous authors. Warren and Kendall (1985) listed criteria for distinguishing subaerial versus subaqueous gypsum which are based primarily on 1) morphology of the crystals, in particular whether the crystals are vertically oriented (subaqueous) or randomly oriented nodules (subaerial sabkha);2)thepurityofthedeposit;and3)thegeometryofthebasin. Whengypsum isconvertedtoanhydritewithburial,criteriasuch asverticallyelongatenodulesand relictinclusions betheonlymorphologicandmineralogicevidenceoftheoriginal may depositionalenvironmentofthegypsum. Examinationoftrenchesanddeepcoresfrom Bristol Dry Lake indicate that these criteria are not as clear cut as previously thought. I, therefore, propose a series of gypsum depositional environments and associated fabrics from subaerial to subaqueous based on the following Recent type localities: Marion Lake, South Australia (subaqueous); Bristol Dry Lake, California (groundwater seepage);andAbuDhabisabkha,U.A.E.(subaerial)(Fig.5.1a-c). Theemphasisof thischapteristobriefly describethe environmentsandthenpresentcriteriafor gypsumdistinguishing thesethreedepositionalenvironmentsbothintheHoloceneandafterthe gypsum has been converted to anhydrite. The gypsum deposits in Marion Lake and Lake Macleod have been well documentedby Warren (1980,1982a,b) and Logan (1987)ashavethedepositsintheArabianGulf(Curtisetal.1963;Kinsman, 1965; Butler, 1969; Evans et al., 1969; Shearman, 1978) and both areas are summarized in Warren and Kendall (1985) and Warren (1989, in press). Therefore, the descriptions of these areas have been held to a minimum and the interested reader is referred to the referencesinTable5.1. ThegeneralsedimentologyofBristolDryLakehasbeen this describedbyHandford(1982a,b),Rosen(inreview),andinprevious chaptersof dissertation, and the surface gypsum deposits have been briefly described by ver However, thedetaileddistributionofgypsummorphologies, their Planck (1952). occurrence, and significance have not been previously described. 125 Table5.1. DocumentedQuaternary sea-margin evaporites (after Warren, 1989). Sabkha Reference Abu Dhabi, Arabian Gulf Butler 1970, Purser 1973 Baja California, Mexico Castens-Siedell, 1984 Bardawil Lagoon, North Sinai coast Levy, 1977 Gulfof Suez coast Gavish, 1980 Laguna Madre, Texas Elliott, 1986 Northwest Australian coast Warren and Kendall, 1985 Sabkha Matti, Arabian Gulf Purser 1973 Spencer Gulf, Australia Ferguson et al., 1982 Tunisian coast Perthuisot et al. 1972 Western Nile delta West et al., 1979 SALINA Reference HuttandLeemanLagoons, WestAustralia Arakel, 1980 Lake MacLeod, West Australia Logan, 1981, 1987 Pekelmeer, Netherlands Antilles Lucia, 1968 Pleistocene coast, Sicily Schreiber and Kinsman, 1975 Ras Muhammad, south Sinai coast Kushnir, 1981 SolarLake, GulfofElat Aharonetal., 1977 Southcoast,Australia Warren,1982a,b Tunisiancoast Perthuisotetal. 1972 Gypsum depositional environments Marion Lake (subaqueous). Much of the sahna gypsum m Marion Lake is coarse grained and bottom-nucleated; pore-infilling gypsum is rare. The gypsum is usually >90% pure and is laid down as a characteristic shallowing-upward evaporite unit. Atthebaseofthesuccession aremassive,poorly layereddomesofcoarse­ grainedgypsum(selenite). Theelongategypsumcrystalsinthiszoneshowlittleorno preferred orientation, and carbonate impurities are distributed randomly through the gypsum. Higher in the section the degree of lamination increases and the dome amplitude decreases until the domes pass into horizontally laminated selenite, where Figure5.1. Generalizedlocationandlateralfaciesmapsofallthreeenvironments; a) Marion Lake South Australia. In subaqueous salinas such as Marion Lake, thevertically aligned gypsumwillbeconcentratedin thecenterof the basin and will be ringed by carbonates, b) Bristol Dry Lake California. In groundwater seepage gypsum, as in Bristol Dry Lake, will beconcentratedaroundthemarginsofthebasinseparated gypsum frombasin-centersubaqueous halitedeposition by athickbandof siliciclastic muds. The arrows point to the trenches through the playa margin gypsum facies, the trench on the south side of the playa is drawn infigure3. Thelocationofthedeepcorethatcontainsplayamargin anhydrite(CAES#1)isshown,c)AbuDhabi,TrucialCoast. Thebulk ofthegypsuminthetypical sabkhafaciesgeometrywill beconcentrated on the most landward up-dip portions of the basin. 128 individual laminae The appeartocrosscutlarge upwardly-aligned gypsumcrystals. gypsum crystals (0.3 to 0.5 m long) are primary, not secondary crystals. They grow withtheirlongaxesperpendiculartobedding,aneffectofcrystal impingementand growth alignment (Fig. 5.2a). Carbonate laminae form by the precipitation and accumulationofaragonitepeloidsduringthespringandearlysummer. Thepeloidsare pellets fromostracodsandbrineshrimpmixedwiththemicritizedremnantsofalgal tubules.Peloids areencasedbytheupwardpoikiliticgrowthofthelargegypsum crystals during the summer and fall. Coarse-grained gypsum, punctuated by carbonate laminae,isinturnoverlainby amm-laminated,sand-sizedgypsareniteaccumulation. The upperpart of this horizontally laminated gypsarenites is often reworked into wave­oscillationripples. Laminatedandrippledgypsumisinturnoverlainbyathin, massive, poorly-bedded gypsarenite unit deposited under seasonally vadose or subaerial conditions. Capping the whole succession is unit of cross-stratified eolian a crust of gypsumand,inareasstabilizedbyvegetation,apedogenic gypsite(silt-sized) has formed atop both lacustrine and eolian sediments. Bristol Dry Lake (groundwater-seepage). A continuous trench 2 m deepand300 mlong,excavatedacross thegypsumfaciesofBristol Dry Lake, reveals bedswithgypsumfabricssimilartothealignedtextureofthesubaqueously formed gypsum. However, the gypsum accumulations in Bristol Dry Lake do not have the same vertical succession seen in Marion Lake. The gypsum forms as stringers in a mudandsandmatrix atthealluvialfanendofthetrench(proximal playamargin facies). Fartherintothelakebasin(medialplayamarginfacies),thestringerscoalesce into a single unit that is over 2 meters thick and composed of stacked beds composed ofalmostpuregypsum(Fig. 5.2b). Fartherbasinward(distalplayamarginfacies), the inthebedagaingraduallyreturnstostringersinamudmatrix,butwith gypsum localizedoccurrencesofcelestitenodulesenclosedbyvertically alignedgypsumor growing displacively in gypsum-free muds. The geometry of the deposit suggests that the gypsum is forming where groundwater supersaturated with respect to gypsum is resurging around the playa margin. Although carbonate is present in the matrix, it is a the sediment. minor constituent of Individual gypsum crystals in Bristol Dry Lake beds are, in general, vertically to 80 mm in the longest dimension, and can form to 90% of the aligned, up up sediment(Fig. 5.2c).Nosedimentarystructuresorevidenceofmechanicalreworking ispresentwithinthegypsum, yetintercalatedbedsofmudandsandexhibitboth organicandinorganicmechanicalreworking.Individualgypsumhorizons beupto may 0.3mthickand areunderlainandoverlainbyadditional gypsum beds. Some horizons haveanundulatoryuppercontactwiththeoverlying gypsum. Crystal sizeisvariable inasinglevertical sequenceandbothupward-finingandupward-coarsening unitsare present. Crystal size increases from the proximal to the distal end of the trench. Morphology of the crystals varies widely from biconcave lens-shaped discs to "swallowtail"twinnedblades,withthebiconcavediscs asthedominantmorphology. Textures, such as matrix inclusions completely outlining all crystal growth surfaces within individual gypsum crystals, indicate the gypsum formed displacively inthe sedimentratherthanasbottom-nucleatedbrinepangypsum(Fig.5.2d). Siliciclastic mudsand sandsintercalatedbetween gypsum accumulations show centimeter laminationsof greenandred burrowedsedimentsindicating alternationsof shallow water and subaerial exposure. Original depositional fabrics within the gypsum accumulations have been completely destroyed by the displacive growth of the gypsum. the Adeepcore takenneartheplayamarginpenetrates thedistalendof groundwatergypsumzone atmanyintervals from270mto500m.All thegypsumhas convertedtoanhydritebythesedepths. Examinationofcoreslabsandthinsections demonstrate the retention of the vertically aligned characteristics observed in the surface trenches, even though bacterial sulfate reduction was involved in the conversion of anhydrite (Fig. 5.2e; Rosen, in review). In thin section, millimeter-size gypsum to felted anhydrite fills the gypsum pseudomorphs (Fig. 5.2f). The anhydrite may also be aligned parallel to the long axis of the pseudomorph. Pseudomorphs are separated by a thin veneer of mud matrix which completely surrounds and outlines former single gypsum crystals. Relatively rare, sand-sized detrital quartz or feldspar grains may be Sand-size calcite crystals includedinthepseudomorphsorinthesurrounding matrix. of calcite and areabundantbothwithinandaroundthepseudomorphs. Thepresence pyrite framboids suggests that bacterial sulfate reduction was involved in the conversionofgypsumtoanhydrite. Thecalciterepresentstheby-productofthe reductionreactionwhichconvertssulfate tohydrogen sulfideandbicarbonate(Berner, 13 1 1971). Thebicarbonatethenreactswiththecalciumreleasedby thedissolutionof gypsum to form calcite (Fig. 5.2f). of Abu Dhabi Sabkha (subaerial). In Abu Dhabi sabkha gypsum, much the sulfate phase grows as displacive gypsum crystals within the capillary zone, creating a penecontemporaneous bedded unit (Fig. 5.2g). Pockets of gypsum crystals dispersed through the near-surface sediments are diagenetically altered to anhydrite nodulesand layers asnewanhydrite alsogrows asnoduleswithin thesediment. Nodulesincreaseinsizeandnumberin landwarddirectionuntilinsomeareasonthe a AbuDhabisabkhatheyeventuallyforminterlockingpolygons. Ongoinggrowthof anhydrite and gypsum layers produces thrust folds, ptygmatic and enterolithic folds, anddiapir-likestructurescomposedofeitheranhydriteorgypsuminthesupratidal unit (Fig. 5.2h). In the more landward sabkha the near-surface anhydrite is overlain by reworked quartzoseeolianite, andtheanhydrite polygons appearfestoonedand truncatedincrosssection.Thetruncationoftheanhydritepolygonsis aresultofthe continued subsurface precipitation of anhydrite lifting the overlying sediments above the capillary zone into the vadose zone, where they dry and deflate. Displacive halite crystals canbefoundintheupperfewcentimetersofsedimentintheuppersupratidal unit, attestingtotheinfrequency offloodinginthisregion. Criteriafor therecognition ofgypsumdepositional environments be Distinguishing betweenthedifferenthydrologic typesofgypsummay difficult,especially afterithasbeenconvertedtoanhydriteandifonlylimitedcoreis available. In particular, distinguishing groundwater gypsum from brine pond gypsum thebasinis not maybeextremelydifficultifthegeometryof known.Ingeneral, will: 1) have a bulls-eye pattern around a more evaporative facies groundwater gypsum such as halite, thus the will not be at the stratigraphically lowest part of the gypsum basin (Fig. 5.1a); 2) consist of vertically oriented individual lens-shaped crystals and have lessabundantpoorly-oriented "swallowtail"twinnedcrystalswhichmay gypsum crystals; 3) be more matrix-rich than overgrown previous generations of subaqueous brinepondgypsum;and4) haveinclusionsofmatrix thatcompletely Figure 5.2. a) Field photograph of vertically aligned selenite crystals from Marion Lake,Australiagrowingfromwell-defined beddedsurfaces(arrows). Noticethatthecrystals coarsen upwardandthatwell-definedcrystal growthlaminationsparallel tobedding can betracedthroughcrystals. Also notice the laminated detrital sediments (DS) between the selenite crystals, b)Weatheredexposureofthetrenchfromthesouthsideof BristolDryLake. Alarge rainstormhaswashedawaytheuncemented siliciclastic matrix between the cemented zones. A sketch of the gypsum gyspum distribution and geometry of this trench is shown in Figure 5.3. Geologic hammerinforeground shadowforscale, c) Close-upofthe vertically aligned gypsum fabric seen in the gypsum cemented zones in 2b. Comparewith2afromMarionLake. Crystalsaresmallerandthereareno well-definedbedding surfaces or growth laminations in the Bristol Dry Lake deposits and there is no bedding in the matrix inclusions, d) Photomicrograph (cross-polarized light) of a lenticular gypsum crystal from Bristol Dty Lake showing matrix inclusions along growth surfaces that all the These inclusions indicate go way around the crystal (arrows). that the gypsum is growing displacively in the sediment. Gypsum crystal is about 2.5 mm long, e) Photograph of core slab from Bristol Dry Lake (depth 350m). Allgypsumhas beenconvertedtoanhydrite, yetthe vertically aligned fabric seen at the surface is retained. Compare with 2c. f) Photomicrograph ofcore slab in 2e. Notice that the matrix outlines the shapeoftheformergypsumcrystals. Pyrite(P)andabundantcalcite(C) canbeseenwithinthecrystals andthematrixindicatingthatbacterial sulfate reduction was involved in the conversion of gypsum to anhydrite. Individual crystals are about 3 mm long, g) Supratidal facies from Abu Dhabi sabkha showing classic enterolithic anhydrite (white areas) in a mixed carbonate-siliciclastic matrix. (Scale is in centimeters), h) Shallow corefromAbuDhabisabkhashowingrandomly oriented gypsum spears (arrows) disrupting thealgal matfacies(ruler is ininches). 133 134 surround growth surfaces of individual crystals. Unfortunately, the preservation of these inclusion surfaces does not generally survive anhydritization. Brine pond gypsum consists of: 1) hydrologically controlled precipitation of gypsum in the stratigraphicallylowestpartofthebasin(Fig.5.1b); 2)vertically-orientedinterlocking "swallowtail"crystalsofgypsumthatmayshowlaminatedmatrixdraping thesurface ofthecrystalsandfillinginasgeopetalstructuresbetweengroupsofcrystals; 3) chevron-like fluid inclusion-rich bands running through the crystals indicating the surfaceofgypsumnucleation; 4)adistinctstratigraphicsequenceofgypsum morphologiesstartingwithpoorlyorienteddomalgypsumpassing upwardinto horizontally laminatedseleniteandgypsarenite, andcapped by cross-stratified gypsum dunes or gypcrete. Distinguishing betweensubaerial andtheothertwo gypsum gypsum types is somewhat easier than distinguishing between subaqueous and groundwater-seepage gypsum.Subaerialdeposits(Fig.5.1c)show: 1)noalignmentofgypsum or anhydrite,2)greatlyfoldedanddisruptedbeddingtruncatedby adeflationsurface,and 3) thin matrix-dominated units relative to subaqueous and groundwater environments. Both the groundwater and subaqueous environments may produce thick and relatively puregypsum beds, but,thelateralextentofthethickestportionsofthegroundwater the gypsum will be confined to a relatively small area (Fig. 5.1a). The geometry of unit as it extends from the thick portion of the deposit will consist of stringers of gypsum in a matrix that increases in abundance, both in the proximal and distal ends, away from the thick portion of the unit (Fig. 5.3). The thick, pure gypsum represents the zone where the groundwater is most supersaturated with respect to gypsum for some period of time (Fig. 5.3). Thelikelihoodofallthreegypsumenvironmentsoccurringin thesamebasin is remote. If the groundwater is supersaturated with respect to gypsum then it will precipitate gypsum before it reaches the brine pond leaving a halite-precipitating brine. Closed basin playas such as Bristol Dry Lake, and others in the Basin and Range of the WesternUnitedStateswillbe dominatedby thistypeofgroundwatergypsum.Other more regionally controlled saline groundwater systems such as some Australian continental salt lakes (Lake Frame, Lake Eyre) may also be dominated by groundwater gypsum (Bowler and Magee, 1988; Magee et al., 1988). Figure5.3. Trenchcross-sectionfromsouthsideofBristolDryLake(seeFigure5.1b for location). Notice the lenticular geometry of the gypsum zones and the almost of the trench. pure area gypsum in the middle of In places, the sedimentisover90%gypsum.Thisthickaccumulationofgypsumis dueto the present stability of the discharge zone. Celestite nodules are concentrated to the basinward portion of the trench, but they are not important to the model for gypsum facies development Vertical exaggeration is 60x. -1 —2m r-TT CI3 BA.SINWARD «¦ 1 X 'Ifw x.^sxn; 30 m i r­—! . * # : ¦ CELESTITE NODULES 1 1 Cl SURFACE VERTICALLY ALIGNED 1 GYPSUM * I r~ MUD n n -—-­— - SAND r oRAvr '­ n Cl SOUTH 1 J 1 H Ifthewaterisnotsaturatedwithrespect togypsum whenitreachesthe brinepondit will havetoevaporate andconcentrateinthebrinepond. Therefore,thegroundwater resurgingaroundthemarginofthebrinepondwill notprecipitate muchgypsum. Coastal salinas such as in South Australia or parts ofBaja, Mexico, demonstrate this type of gypsum deposition. However, changes in the groundwater chemistry may influence the saturation state of gypsum in the groundwater, superimposing groundwater gypsum morphologies around the edges of salinas. Subaerialnodular gypsumandanhydriteformation,asintheAbuDhabisabkha,may occurinanyofthe three settings where enough evaporation occurs in the capillary zone to produce displacive calcium sulfate crystals and aggregates in the sediment. * *' JL — Conversion of gypsum to anhydrite. As mentioned above, the conversion of to anhydrite (or worse, alterations between and gypsum gypsum anhydrite) may theoriginal depositionalenvironmentof makethedeterminationof the difficult gypsum or impossible. Gross morphologic outlines of the original gypsum vertical orientationof the crystals may be preserved as pseudomorphs or simply as a sulfate fabric, but these criteria may not be enough to distinguish groundwater from brine-pangypsum.Ifthegeometryoftheanhydrite depositisknownthenthiscriteria canbehelpfulindistinguishing thetwotypesofdeposit. Ifthegeometryofthedeposit is not known, the vertical relationships which can be seen in a single core may be useful. InBristolDryLakeandotherintermontanebasinplayas,thebrinepanhaliteis extensive saline mud-flat. laterally separated from the groundwater gypsum by an There is little for progradation in these settings because of the geometry of the room basins, so thatbrinepanhaliteis neverassociatedwithextensivegroundwatergypsum andhalite wouldbe in cores. Even if progradation or retrogradation did occur, gypsum separatedby amud-richintervalwhichwouldcontainincreasinglyabundantdisplacive halite towardsthe topof themud-rich intervaljustbelow thebrine pan halite(Rosen, in review). Onthemesoscale, retentionoffluidinclusion-richbandsandgeopetal structurescansurviveconversiontoanhydrite. SomePaloDuroBasincore(Permian) demonstratesthistextureinpreservedgypsumpseudomorphs (S.Hovorka,pers. comm.). However, if gypsum converts to anhydrite in a relatively closed system, the chaotic volume loss associated with anhydritization (Shearman, 1985) may create a fabricinthematrix anddestroyeventheverticalelongationofthethe gypsumcrystals. Summary Depending on the state ofpreservation of the gypsum deposit and/or core and thedegreeofthree-dimensionalcontrol on the basin, therecognition of groundwater versus brine pond or subaerial gypsum may be extremely difficult or impossible. The mostreliablecriteriafordistinguishing thethreeenvironmentsaretheverticaland from lateral facies relationships. Distinguishing groundwater or brine-pond gypsum fabric subaerialdepositsmaybestraightforwardifthevertical alignmentofthegypsum is retained. Even with good preservation, distinguishing between groundwater­seepageandbrine-pond gypsummay vertically oriented bedifficult. Thedominanceof "swallowtail" crystals, laminated gypsum fabrics, fluid inclusion-rich bands, and geopetal fabrics in the matrix are the best mesoscale (i.e. core slab, thin section) A dominanceof criteria for brine-pond gypsum. vertically oriented lens-shaped discs, lackoflamination,andmatrix inclusionswhichsurroundgrowthsurfaces twinplanes and isolate entire crystals, are mesoscale features which suggest groundwater-seepage orallofthefeaturesmentionedabove, in gypsum. Anhydritization may destroy some which case only the facies relationships can to distinguish the gypsum be used environments. exist which have been Ancient examples of groundwater-seepage gypsum may Potentialcandidatesforreinterpretation for interpreted as brine-pond gypsum. groundwater seepageincludepartsofthe JurassicBucknerAnhydrite Member an unnamedPleistocene (Haynesville Formation) inAlabama(Mann, 1988)and formationinPuertocitos,8.C.,Mexico(Holser, 1979b). Carefulreevaluationofthese theabovecriteria addsomenewinsightintotheevolutionof deposits in light of may these basins. CHAPTER 6. Tephrachronology Thecorrelationofthetephra layersinBristolDry Lakewiththetephralayersof known ages from other basins (Fig 6.1) provides the basis for the discussion in this chapter. These correlations, when viewed in light of the sedimentologic and geochemical evidence, provide unique example of how these data can be used not a onlytodeterminedepositional ratesandageofthebasin,butalso todetermine significant structural alteration of the basin that could only have been postulated otherwise. In all, nine tephra layers from two cores (four from CAES #1 and five from CAES#2)havebeencorrelated. Twoofthetephralayerscanbecorrelatedbetween the two cores (Figure 6.2). Details of how these correlations were made are presented in appendix V. AgeoftheBasin. Therearenopreviousestimates ormeasurementsofthe . ageoftheBristolDryLakebasin. Althoughdatafrom36C1measurementsinhalite indicatethatby 150mdepth, thesedimentis approximately 2millionyearsold(Jannik, pers. comm., 1988), this dating technique is still experimental and may have large errors associated with it (Phillips etal., 1983). Nevertheless, there is an indication that sedimentationratesinthebasin arefairly slowandthatthebasinis the average relatively old.Correlationsofthedeepest tephralayersfromCAES#IandCAES#2 indicatethatthebasinisatleast3.7±0.2million yearsoldatthat thoselevelsinthe In CAES #2 the lowest tephra is almost at the base of the core (500 m), but in CAES#l,thereisstillalmost270mofsedimentbelowthedeepestash. Usingthe lowestandhighestcalculatedsedimentationratesforCAES#1(seebelow), thebottom cores. ofCAES#1wouldbesomewherebetween6and10millionyearsold. Although no structural complications are apparent inCAES #1 (see below), faulting, such as seen in CAES #2, could greatly reduce the estimates. GiventhatthebottomofCAES#1isbetween6to 10millionyearsold,thisis stillnotthemaximumagelimitforthebasin. NocorestakeninthecenterofBristol DryLakehavereachedbasementrocks. Therefore,thereisstill anunknowndepthof sedimentbelowthecoredintervalthatisolderthantheestimate. Acrudeestimateof the depth of the basin can be attempted by interpreting the Bouguer gravity map for the Needles quadrangle. Using the equation D 78Agmax /AB (Telford et ai, 1976, = 140 Figure 6.1. Schematic correlations of various marine and non-marine tephra layers throughoutCaliforniaandwesternNevadaintomarineDeep SeaDrilling Project core (no vertical or horizontal scales). Although the tephra layers in Bristol Dry Lake have not been dated, correlation with other tephra layersofknownagemakeageestimatesforBristolDry Lakeacurateto within 0.1-0.3 m.y. 142 Figure 6.2. Cross-plot of tephra age (based on correlations in appendix V) and depth incoreCAES#1andCAES#2. Noticethatbothcoreshavesimilarlow depositional ratesinmostportionsofthecores,except thataround2m.y. thedepositionalrateincreasesby anorderofmagnitudeinCAES#2. Correlations of the same tephra layers betweenshown with cores are arrows. Filled diamonds are tephra layers in CAES #2 and open diamonds showthestratigraphic position ofother tephralayersknownfromother basins that may be present in the basin. Similarly for CAES #l, open squares are tephra layers from Bristol Dry Lake, and the filled squares are tephralayersfromotherbasins thatmaystill befound. 144 is maximum p.87), where D equals the depth in feet of the overburden, Ag max difference in the gravity data, and Afß is the difference in the density of the overburden versusthebedrock, aroughestimateofthethicknessofthebasinfillcanbemade. Assuming adensityofthebedrocktobeabout2.9g/cm3, thedensityofthewetclay andhalitetobeabout2.2g/cm3,and of20milligals(ChapmanandReitman, 1978), the total thickness of the basin-center is approximately 680 meters. If this calculation is correct, bedrock is approximately 100 m belowthebottomofCAES#2. Structural complications. The tephra layer at about 200 m in CAES #1 correlateswith atephralayeratabout450minCAES#2(Fig. 6.2). Ifthiscorrelation iscorrect,theneithertherewasover200mofreliefinthebasinorthereis astructural complication. Because none of the six other tephra correlations in the core violate any stratigraphicprinciples(i.e.adeepertephraisnotcorrelatedtobeyoungera than shallower tephra layer in the same core), the accuracy of the correlation is almost m unquestionable.Thelikelihoodofover200 ofreliefinthebasinisextremelylow. Eventhough thesedimentationrateincreases by anorderofmagnitudeintheinterval from450to285minCAES#2, thetypeofsedimentsseenincore are still thesame alterationofephemeral-lake halite deposits andfine-grained, homogeneous, siliciclastic muds. Displacive halite is also common in the muds (see appendix I). Themostlikely cause ofthis large difference insedimentationratesfromthe of margin core (CAES#I) and the core in the center the playa (CAES #2) is some sort of growth fault or tectonically induced fault striking through the center of the basin. Evidenceofsuch afaultcanbe seenintheintervalfromabout380mto370min over 60° from CAES #2. Both evaporite and mudbeds in this interval are tilted horizontal. Thistiltingcannotbeattributedtodeviationsfromverticalofthecorebarrel or to any diagenetic processes in the core (i.e. solution collapse features are not present Abovethetiltedinterval thebedsreturntohorizontal. in this interval). TiltedbedshavebeenobservedinCadizDryLakestarting atdepthsof80m (Bassett et al., 1959). Below an initial slickensided surface that dips 60°, the beds dip at approximately a 10° angle gradually increasing to 35° by 140 m.Thisfaultis much shallower thanthe occurrence in Bristol Dry Lake. However, this may be due to the relative agesofthesedimentorthelocationofcores relative totheangleofthefault. If therehasbeenrelatively littlesedimentationintheCadizbasin,theshallowerCadizDry Lake sediments may correlate with the deeper Bristol Dry Lake sediments. Alternatively, if the fault dips at some angle, depending on where the core is taken, fault will intersect the basin at different stratigraphic levels. Therefore, it is possible that the faults in the two basins are related and regionally extensive a may bepartof trough-center fault. The nature and timing of these faults may be important not only to the sedimentology and hydrology of the basin, but also to the regional tectonic history of the area. Apparently this fault, or faults, moved slowly enough to allow sedimentationstyletoremainfairlyconstantovertheperiodofmovementon thefault. Althoughsedimentationratesincrease(seebelow), ephemeralplayaandsalinemud-flat facies continue uninterrupted through the tilted intervals. Sedimentationrates.Thesedimentationratescalculatedfromthe tephra correlations yieldaveragesedimentationratesoftheintermediateandlong-termhistory ofthebasin. However,sedimentationinBristolDryLakeishighlyepisodicdueto localweatherpatternsandalsoregionalandglobalclimate. Therefore,thecalculated depositionalratesdonotrepresentthephysical depositionalrate any the of of sedimentary packages. For example some halite beds may actually represent only a few years or possibly days of halite deposition, whereas the mud layers may represent thousandsofyears. Flashfloodsfromonestormmaydepositahalfameterofsandin the basin in an instant, and then no deposition may follow for hundreds or thousands ofyears. Deflationanderosionmayalsodecreasetheaveragesedimentationrate. Thus, the utility of the average sedimentation rates is not to determine physical butrathertodeterminetheoverallrateatwhichthebasinis fillingandto processes, determinetheamountoftimenecessarytodeposit andaccomodatethesediments(i.e. rates of subsidence). sedimentationrateforCAES#1is66mm/103 andthe Theaverage years sedimentation rate for CAES #2 is 135 mm/103 years. These numbers are average derived by taking the total thickness of the sediment above the tephra layer dated at 3.7 If stratigraphic markerin the basin, this exercise wouldillustrate that on the the millionyearsoldanddividingbythenumberofyears. thistephraweretheonly average, basin center sediments accumulate approximately twice as fast as the margin sediments. However,correlationsoftheothertephralayersdramaticallydemonstratethisis a major oversimplification of the sedimentation pattern in the basin. In CAES #2, the interval between 285 m and 450 m is represented by as little as 0.2 million years or as much as 0.5 million basedthe accuracy of the tephra correlations. This makes on years thesedimentationrateforthisintervalbetween33and825 mm/103 forthis years interval. Belowthisintervalthesedimentationrateisbetween52and42mm/103 years which is actually lower than the average playa margin sedimentationrate. However, it isalmostexactly thesame rateforCAES#1 belowthe200m tephralayer(42mm/103 years).From thepresentdowntothefirstcorrelatedtephralayerthesedimentationrate inbothcoresishigherthanatthebase. InCAES#2thesedimentationrateis142 mm/103years,andinCAES#1it isbetween80and90mm/103 years. Ifthe sedimentationratesareclosetotheoverallrates,and average sedimentation is still active, the Holocene should be represented by about 1.4 m of sedimentinthebasincenterandbetween0.8-0.9metersalong theplayamargin. Although there is some question as to the age and correct stratigraphic position of the Plio-Pleistocene boundary (Jenkins, 1987), if an age of 1.8 m.y. is used, the basin-center section should be 252 meters thick. Adding this to the Holocene thickness and addinganother28 mtogetto theageofthefirsttephralayer at285meters,thetotal thickness of section is calculated to be 271.4 m. The close agreement of the position of the tephra layer and calculated thickness of the section (within 15 m) is remarkable considering theuncertaintyinthedatingtechniqueandtheeposidicnatureof sedimentationinthebasin. The calculationsfortheplayamargincore areeven same closerthanthebasin-centercore(within afewmeters).Thesecorrelationsdemonstrate thatthe averageratesare fairlyaccurateindeterminingmoderately longterm(greater than 105 year)sedimentationrecords.However,iftheaverageratefor theentirecore were used,thecorrelationswouldnothavebeen veryclose. The correlations of the tephra layers illustrate three important aspects of sedimentation inthe Bristol Dry Lake basin that will be discussed below. These points are: 1)sedimentationratesforthistypeofenvironmentareslowevenattheirpeakin CAES #2. 2) Before structural complications from about 2.0-2.5 million years ago, thesedimentationratesforthecenteroftheplayaandthemargins wereaboutthesame. 3) Order-of-magnitude variations in sedimentation rates demonstrate that average sedimentationratesbasedononeortwopoints areinadequate todefinedepositional styles in rift basin playas. sedimentation Thefirstpoint,thatthe average ratesareslow,issignificantin determining the chemical budget ofions to the basin and in determining how much tectonicmovementinagivenamountoftimeis necessarytoaccommodatethe sediment. Forcomparison, thesedimentationratesinSearlesLake,California, a closed-basin playa in a similar structural setting to Bristol Dry Lake, average 260 mm/103 for siliciclastic muds and 600 mm/103 years for the saline mineral phases years (Smith,1979). However,theseratesarebasedonarelativelyshortrecordthatis only 25,000 years long. More recently, a 930 m core taken in Searles Lake penetrated tobasement. Althoughthesedimentationratesmeasuredbetweenpaleomagnetic reversalsvariesfrom120-550mm/103 theoverall sedimentationrate years, average for the lacustrine portion of the core (which spans 3.2 m.y.) is approximately 220 mm/103 years (Smith etal., 1983). Given theoverallaveragesedimentationratescalculatedforSearlesLakeand Bristol Dry Lake, the higher rates in Searles Lake be explained by the observation may that Searles Lake derives its water from the steep high mountains of the Sierra Nevada Range. Thesemountainshaveamuchhigherannualprecipitationratethanthe mountainssurroundingtheBristolDryLakebasin, sothatamuchlargervolumeof wateranddetritalsedimentwillreach thebasininagiven yearperunitarea.The increaseinthesedimentationrateinCAES#2fromBristolDryLakeis similarto the sedimentationratesinSearlesLake. Thissuggeststhatreliefonthemountains surrounding BristolDryLake wasperhaps higher atthistimebecause theaddedrelief wouldhaveproducedasteeper slopetothebasinandcapturedmorerainfallresulting in ahighersedimentationrate. Thus,theincreaseinsedimentationratemaybedueto faultingrelatedtorejuvenationofthesource areasaroundthebasin. Incontrasttotherelatively slowaveragesedimentationratesforbothBristol DryLakeandSearlesLake,theSaltonSeageothermal system,atectonically active hydrothermal closed-basin to the southof Bristol Dry lake in Southern California, has sedimentationratesthatareontheorderofmetersperthousandyears. BristolDry Lakeisonlyafewhundredkmtothenortheastofthisareainarelatively similar tectonic setting, and yet its sedimentation rate is at least two orders of magnitude lower. In another tectonically active area of western California, Blatt et al. (1980) calculated a sedimentation rate of 11,000 mm/103 years on datareported by for 11 ma. based Crowell(1974) fortheRidge Basinalong theSanAndreasFaultinCalifornia tothe west of Bristol Dry Lake. TherelativelyslowsedimentationratesinBristolDryLakecompared toallthe basins mentionedabove, allowforthegradual accumulationofionsderivedfrom weathering and direct precipitation into the basin by evaporative concentration. This is importantintermsoftheClmassbalancecalculationsinthebasin.TheCl mass balanceultimatelycontrols theamountofhalitethat canbeprecipitated inthebasinand sodeterminestheoriginofthehalitebeds.Inaddition,slowsedimentationrates are important in determining the preservation of the paleoclimatic record inthe basin. Theobservationthatthenon-tectonicallyinduceddepositional ratesin thecenter ofthebasinandalongthemarginofthebasinaresimilarsuggests thattherelief marginal to the basin when it was not affected by tectonics has, in general, been low. Iftherewereagreatdealofvariabilityindepositional ratescoupledwithachangein sedimentationstyle,variableamountsofreliefmightbeexpected. However, ashas beenshownabove, thevariabilityinBristolDryLakesedimentationratesisdueto tectonically inducedfaultinginthecenterofthebasinandnot becauseofthereliefon the margins of the basin. Finally, the abrupt change in depositional rate in more CAES #2 indicates that detailedinformationisnecessaryinordertoadequatelyconstrainthedepositional styles ofevaporite deposits in tectonically active areas. Even though the sedimentationrate increases by an order of magnitude, the playa still maintains shallow-water brine pan conditions.Itwouldbeinteresting todetermineexactlyhowmuchofanincreasein sedimentationrateisnecessaryinordertochange thedepositional styleofthebasin. Summary indicate that 1) cores Analyses of pristine tephra layers in both CAES #1 and #2 the basin is at least 4 m.y.a. and may be as old as 10 m.y, and 2) correlation of tephra layersbetweencoresispossible. Thesecorrelationscombinedwithsedimentologic evidence reveal that structural complications, possibly a normal or growth fault, are responsible for theabrupt change insedimentationrate inCAES#2from450to285m. Eventhoughtheaveragesedimentationrateincreasesby anorderofmagnitudeinthis interval, the type of sediments seen in core are still characteristic ofephemeral-lake halite deposits and fine-grained, homogeneous, siliciclastic muds. Theaveragesedimentationrate forCAES #1 is66mm/103 years andthe averagesedimentationrateforCAES#2is135mm/103 InCAES#2,from450 years. to 285 m the sedimentation rate is between 330 and 825 mm/103 years, but below this intervalthesedimentationrateis 52and42mm/103yearswhichisactuallylowerthan the averageplaya margin sedimentationrate.Thecorrelationsof the tephralayers illustrates3importantaspectsofsedimentationintheBristolDryLakebasin: 1) sedimentationratesforthistypeofenvironmentare sloweven attheirpeakinCAES#2 comparedtootherbasinsinsimilarsettings, 2)beforestructuralcomplicationsfrom about2.0-2.5million CAES#2),thesedimentationrates yearsago(450 to285min forthecenteroftheplayaandthemargins wereaboutthesame,and3)orderof magnitude variationsinsedimentationratesdemonstratethataveragesedimentation on one or two rates based points are inadequate to define depositional styles in rift basin playas. CHAPTER 7. Geochemical Results Sulfates Sulfurisotopes. The&34Softhesulfateradicalinasolutionis not greatly affected by evaporation or small temperature changes (Holser, 1979a). This is because the sulfur cation is so tightly bound to the four surrounding oxygens that fractionation of free sulfur molecules is greatly limited. Therefore, the of the precipitating sulfatephaseshouldreflecttheratioofthesulfatefromtheparentsolution. Ratiosof from surface samples of gypsum, anhydrite and celestite all fall within a narrow rangeof+6 +9°/ooCDTwithanaverageof7.7andstandarddeviationof±0.7 - regardless of the phase analyzed (appendix III). Over 60 surface samples were analyzed, mostly of gypsum from the 300 m trench, and yet no statistically valid trends could be determined. Anhydrite samples from the cores have tighter values in the range of0 +4o/ooCDTwithanaverageof3.0andastandarddeviationof±1.1. - Deuterium and oxygen of gypsum-crystallization waters. Eighteen samples of the gypsum from the trenches used for sulfur isotope analyses were also analyzedforisotopiccompositionofdeuterium(D)inthecrystallization waterofthe gypsum. Two samples were also analyzed for oxygen (appendix III). In addition, rainwater6Dvalueswereobtainedfor4yearsofprecipitation fromtheUSGS. The range of 6D values for the gypsum crystallization waters is -95 to -67 o/00, the average is -81 °/oo and the standard deviation is ± 8 °/00. The rain water samples for 4 years ofprecipitation arestronglydependentontheseason. Samplescollectedafterwinter storms have an average D of -73 °/oo and samples collected during the summer have an Dof-48o/00. Theoverall oftheDinBristolDryLakeis average average approximately -60 °/00. The two 6 180 values for the gypsum crystallization waters are -0.8 o/oo and -2.6 o/00. Strontium. The87Sr/86Srratioofvariousmineralsmaybeusedtodetermine ofstrontiumin thebasin.Analysesofthe87Sr/86Sr ratioofthecelestite ratio for five samples (87Sr/865r=0.71224 the source (appendix HI) reveals that the average ±0.00003) is constant at the surface regardless of location around the playa margin. 151 Carbonates Oxygen and carbon isotopes. Oxygen, and to a lesser extent, carbon isotopes can be useful in tracing the evaporation path of the water precipitating carbonatemineralphases. Inaddition,carbonisotopesmayindicatetheinfluenceof organisms or organic processes on the solution. Although there is not a great volume of calcite inthe basin-center compared to halite or detrital mud, the calcite concretions are abundant and occur at almost every stratigraphic position in the muds. The concretions alsoabundantin CAES#2which takeninthebasin-center. are core was The thebasin-centerconcretionsfromthesurfaceis - average isotopic composition of 0.4 ± 0.9 o/oo for 6 180 and -11.0 ± 2.2 for 613C. of values is The range oxygen small whereas, the carbon values vary by 6-7 o/oo (Appendix HI). Although slightly lighter valuesarefoundintheoxygenvaluesinthecore(average for6180 is -2.3 ± 1.3o/ooand-11.5±4.0for 613C),thefieldofvaluesforbothcarbonandoxygen overlap. Analysesofafewcalcitesamplestakenfrompedogeniccalciteintheplaya margin sediments demonstrate an opposite trend (average for &18Q is -8.0 ± 0.4 o/oo and -1.1±0.2for613C). Whereasthethe oxygen values in the basin-center where relatively heavy and the carbon values relatively light, the oxygen inthe playa margin calcite is relatively light and the carbon heavy. Halite Bromide, Analyses of Br in halite by XRF show very little Br in not only the halite itself but also the fluid inclusions withinthe halite. Of the seven samples analyzed, onlyonesample wasabovethe5ppmdetectionlimitofthemethod. The concentration of Br in this sample was 5.4 ppm. Water data The chemistry of groundwater from Cadiz and Bristol Dry Lake and brines from Bristol Dry Lake come from a sources from the literature (primarily numberof from Shafer; 1964, and Calzia; 1979), some ranging all the way back to 1910 (appendixIV). Evenso,therearesurprisinglyexcellentcorrelationsbetweendifferent analyses. Figure7.1 demonstratesthatwhenallthenaturallogmolarmajor cationsand anions are plotted against chloride the relative concentrations of the various ions in Figure7.1.Molarconcentrationsofallmajorionsplottedagainst themolar concentrationofCl.LinesforNa,Mg,K, SiC>2,andHCO3representlinear regression lines;whereas,thelinesforSO4andCaareinterpolatedtoshow the trends of the data. All the major cation increase towards the basin- center. Conversely, all the major anions besides chloride decrease towards thebasin-center. Sulfateconcentration(filledsquares) increasestojust fanwardofthegypsum-celestite zone.Precipitationofthesesulfatephases depletes the sulfate concentration to trace quantities in some samples from the basin-center. solutionshowdistinctpathswhichcorrelatewell. Theconcentrationofthemajor cationsNa,Ca,Mg,andKallincreaselogarithmically towardthebasin-center, whereasminorspecies such asHCO3andSiC>2bothdecrease. Sulfateincreasesin concentration toward the basin-center until it reaches the distal alluvial fan where it sharplydecreasesintothecenterofthebasin. Chlorideistheonlymajoranionthatis atitsmostabundantconcentrationinthebasin-center. ThemolarNa/Clratioofthe groundwater averages around2,buttheNa/Clratioofthebasin-centerbrineis approximately0.4 0.7. ThemolarratioofMg/Cainthegroundwaterrangesfrom -- 0.05 to 0.5 and in the brines itis approximately 0.03 0.7. Sodium, calcium, and chlorideare byfarthemostabundantionsinsolutioninBristol DryLake. CHAPTER 8. Composition and origin of the brine Asdescribedinchapter 1,intheabsenceofchemicallyprecipitateddeposits, water compositions are a product of: 1) silicate hydrolysis (weathering reactions), 2) influx of ions (through precipitation or dust) from the atmosphere, 3) uptake of sulfate fromoxidizedsulfides, and3)precipitationofalkalineearthcompounds. Ifolder chemical precipitates (evaporites or carbonates) or clays are present in the drainage basin, simple solutionandleaching ofadsorbedions maycontribute to theinitial brine composition. GroundwaterdatafromCadiz andBristol Dry Lakecorrelatewellwith brine data from the center of Bristol Dry Lake. This suggests that Cadiz and Bristol dry lakes havesimilarsourcesfortheirionicconstituents,andpossibly thatthetwobasins are Na-Ca-Cl brine with hydrologic ally connected. The present brine is essentially a subequal but minor amounts of K and Mg. Chloride is by far the most abundantof all ions present in solution (appendix IV). It is interesting to note that in freshwater, HCO3 is actually 2timemoreabundantthanSO4andClwhichoccurinalmostequal amounts. Butbythetimethewaterreachesthe basin-centerbotharepresentinvery low concentrations. Following thechemicalflowpathmodelofEugsterandHardie(1978) described itbeseenthattheNa-Ca-ClcompositionofBristolDry in chapter 1 (see Fig. 1.5), can Lakebrineis actually oneofthetype basinsforpath HA.TheevolutionofBristolDry Lake brine along this path can be explained by the observed evaporite mineral assemblages in the basin. In the following paragraphs, all the major, and minor ions, and some of the trace ions in solution will be discussed in terms of their importance to theevolutionofthebasinandtheirprobable origin. The following discussion assumes that the present brine composition is in equilibrium with the observed mineral assemblage in the basin. This assumption may notnecessarily becompletely correctbasedonthedeuteriumandoxygen isotopic evidence. Analysesof6Dandvaluesofgypsumwatersofcrystallization average about -80 The 6D value for rain water in the basin is °/oo SMOW (appendix HI). average 155 to thevalues. Thefractionationofdeuteriumfromthewaterprecipiting thegypsum approximately -60 °/oo SMOW, although there is a strong seasonal component to the water included in the crystal structure is between -15 and -20, that is the gypsum crystallization water will be 15 to 20 °/oo lighter than the water from which it precipitates (Fontes and Gonfiantini, 1967). This implies that Bristol Dry Lake gypsumprecipitatedfromwaterwithanisotopic valueofbetween-60and-65°/oo SMOW(Fig.8.1). isthesameasthemodemrainwater but,the This range average, water If must undergo evaporation beforegypsum can precipitate. the starting water has valueof-60°/ooSMOW, beforeitcanprecipitategypsumitwillfractionatedue a toevaporationandbecomeheavier.Thusthegypsumprecipitating from thebrine shouldhaveenricheddeuteriumrelativetotheobservedvalues(Pierre, 1988). The effects of the "loop" trajectory in the evaporation path (Sofer and Gat, 1975) which mayactuallydecrease6Dvalueswithexcessiveevaporation doesnotcomeintoeffect untilthesolutionhasreachedasteady state.InterminallakessuchasBristolDry Lake, a true steady state is not achieved, because evaporation rates change with the activity of the brine and the activity of the brine changes with sporadic meteoric water input as rain. The isotope content changes along with the activity and evolves along a constant slope (Sofer and Gat, 1975). Therefore, the recharge water which was concentratedtoprecipitatethegypsuminBristolDry Lakemusthavehadmuchlighter D valuesthanthepresent rainwater. Similarly, thevaluesforoxygenisotopesofthehydrationwaterareabout-1.5 °/oo SMOW. The fractionation for oxygen is +4 °/oo SMOW (Fontes, 1965) so that thewaterfromwhichthegypsumprecipitated shouldhave a6180ofabout-5.5°/oo SMOW. It is demonstrated below that the 6 18 G of the rain water entering the basin can beconstrainedbetweenapproximately-4to-9. Again,thisvalueisveryclosetothe calculatedvaluefortheoxygeninthewaterofcrystallization ofthegypsum,but aswith the 6D values, the water must undergo evaporation before gypsum can precipitate and so the values should be heavier. The carbonates in the basin definitely show large a the basinthat this increaseinthe6180fromthemarginof tothecenter(seebelow),so increase should be reflected in the 6180 of the waters ofcrystallization in the gypsum. There are three ways of explaining the observed 6D and 5 180 values in the First, the values could be caused by the mixing of the brine with the waters. gypsum freshwaterandthattherelatively lightisotopiccompositions arereflecting amajor contributionoffreshwaterinthe of formation.This alsobe process gypsum may Figure 8.1. Cross-plot of 6D of the surface gypsum hydration water (in °/oo SMOW) versus (in °/oo CDT) from the same crystals. Generally, deuterium is plotted against 6 180 SMOW, but because there are only two data points for6lsO,Sulfurwasusedtographicallydisplaythedata. Average&D values of rain water representing 4 years of data from a station near Amboy are plotted as the box from -60 to -65 °/oo SMOW. The fractionationof6Dfromthegypsumprecipitating watertothehydration waterincludedinthecrystalisbetween-15and-20°/ooSMOW. The datasuggestthatthewaterprecipatingthegypsummay alighter have had 6D signature than the water presently recharging the basin. It is interestingtonotethecorrelationbetweenheavier andlighter6D, but thereason for thiscorrelationis notknown. relatedto diagenetic stabilizationof infresher afterinitial water precipitationinabrine.Anotherpossibilityisthattherelatively depletedisotopic compositions represent fossil recharge water which was more depleted than the present water. Themostlikely timeforamore depleted waterwasduringtheglacial periodsof the gypsum the Pleistocene. Supporting evidence that Bristol Dry Lake is relict Pleistocene a feature can be obtained from the geomorphology of the basin as described above. Finally, the isotopes may reflect the activity effects produced by the brine which suppress heavy isotopic values. This last possibility seems unlikely because the values oxygen are shown to increase substantially in the carbonates. At this point it is not possible to distinguish between the first two possibilities. Deuterium and oxygen analyses of the brine, fluid inclusion in the halite, and the rain water would help to constrain the choices. If the first case is true, then the present brine waters are in equilibrium with the surface evaporite assemblage. If the second case is true, then the present brine may not be in equilibrium with the surface evaporites. It should be pointed out that there is a great deal of uncertainty inthe above discussionbecausemanyofthefactors whichaffectthe6Dand6180 activity such as thesalinity andtypeofionsofthesolution, andhumidityofof theregion(Sofer and Gat,1975),arenotknownexactlyforthetimeofgypsumprecipitation. Thegenerally lowhumidity inthebasin wouldtendtodelay the effects of the"hook"trajectory until higher concentrations were reached, however, the large amounts ofCaCl2 in solution would have the opposite effect (Sofer and Gat, 1975). The most logical explanation, based on the available data, is that the gypsum precipitated from evaporated water initially more depleted than the water which currently recharges the lake. Major ions Chloride.ChlorideisbyfarthemostabundantionintheBristolDry Lakeand CadizDryLakebasins. OneofthereasonsthatClissoabundantisthat,untilwater reacheshalitesaturation.Cl isnotamajorconstituentofchemicallyprecipitated sedimentary rocks. On the other hand, because it is not abundant in rocks that may contributetothesolutesupply, thesourceoftheClisnotclear. Themostobvious sourceofClisthedissolutionofpreviouslydepositedmarineevaporite bedswithinthe surroundingexposedorsubsurfacebedrock. Davis(1981)wasabletodemonstratein ClaytonValley,Nevada,thatpreviously depositedmarineevaporitescouldaccountfor theClandNaconcentrationsinthebasin.Bycalculating themolarNa/Clratiosofthe brine from the basin-center, he was able to show that without exception the ratio was near unity, whereas fresh groundwater ratios were highly variable. Theexcellent correlationofNawithCl inthebasin-centercoupledwiththewidefluctuations in groundwaterratiosindicates thatthesourceofbothoftheseionsmusthavebeen asalt bed. In addition, oldermarine salt beds known toexist in the basin. are IntheBristol-Cadizbasin,thefreshwatermolarNa/Clratiosexhibit amoderate rangeofvaluesbutaregenerallybetween 1and2molar.InboththeBristolandCadiz basin-center brines, the molar Na/Cl ratio are the opposite of the groundwater values andrangefrom0.4to0.7. Itisnotclearhowsodiumisbeinglostbetweenthe groundwater and the basin-center brines because there are no known Na-rich minerals other thanhalite being precipitated. However, the adsorption of Na onto clay surfaces mayaccountforsomesodiumloss. RegardlessofthedifferencesintheamountofNa in solution, the molar Na/Cl ratio is definitely not near unity. It follows then that the simpledissolutionofpreviously existing halitebedscannotaccountforthethe sodium andchlorideinthebrine.Thefinalblowtothishypothesis isthatnoevaporitebedsare known to exist in the basin. If they do exist, they are completely in the subsurface. Withouttheobviouspresenceofolderevaporites asasource ofNaandCl,the otherpossible sourcesofchlorideare: 1)hydrolysisofunstablemineralphases (including fluid inclusions) in the bedrock, 2) hydrothermal fluids, and 3) atmospheric sources (dust and direct precipitation). With the available data, it is impossible to completely dismiss any of these sources as a contributing factor. Evidence for the first of the mid anddistalalluvialfans. Individualcobble-sizepiecescrumbletothetouchfrom hypothesis canbefoundinweatheringofigneousrockfromthesoil zones dissolutionofthelessstablemineralphases. Inthesecondcase,thesettingofBristol Dry Lake in a young rift zone makes hydrothermal sources a likely prospect. Amboy crater and numerous ash layers in the core are testimony to the importance of igneous activityintheregion. Asathirdpossibility, atmosphericdustisattributedtobethe ofCaCOsindesertsoilandcalcreteformations(Gileetai, 1966),and major source certainlycannotbediscounted However,asisthecasefor as a source for chloride. CaCOs,ifatmosphericdustisthemajorsourceofchloride, theNa/Clratioshouldbe near unity (see below). Chloridebudgetcalculations. Because oftheabovesourcesofCl none can be completely eliminated, the next step is to determine, if possible, the relative contributionofeach source. Theeasiest source toconstrain isthecontributionfrom rain water. In addition, because chloride has only one sink, halite, it is a simple processto assess theamountneededtobalancethesources.Chlorideconcentrationsin rainwaterfortheregion areavailableinFeth(1967, 1981)whogivesmaximum(90 ppm),minimum(0.2ppm) andaverage (5.8 ppm)concentrationsfor theMojave Desertregion However, some care mustbeexercisedinevaluating recent . precipitation records for the region because the prevailing winds carry pollutants from the industrialized Los Angeles area into the basin. Sulfate concentrations, and to some extentchloride,mustbeviewedwithsomecaution. Nevertheless,thedatagivesa first approximation of solute concentrations reaching the catchment area. ThedataofFeth(1967) werecollectedasbulkprecipitationwhichincludesdust. However, calculation of Na/Cl molar ratios are never unity in the samples. If all the Cl in thebulkprecipitation came solelyfromdry fallout(halite) thentheNa/Clmolarratios shouldbenear unity. Somedryfalloutofhalitewhich isrecycled fromtheplaya surfaceisexpected, but,becauseNa/Clmolarratios areneverunityinthemeasured samples, dryfalloutofhalitecannotbetheonlysourceofCltothebasin.Itis interesting tonotethatinmostareasaroundtheMojaveregionmolarNa/Clratios are greaterthan1,butaroundBristolDryLake(such asLudlowandDaleDryLake collection sites) Na/Cl ratios are generally less than one. At present there is no explanationforthisphenomenon. Althoughthedatarepresentaveragesovertheentire MojaveDesertregion,examinationofindividualcollectionpoints closesttoBristolDry Lake(i.e.LudlowandDaleDryLakecollectionpointsinTable 1ofFeth(1967)) indicatethattheaveragevalueof5.8 is goodestimateforthe a ppm average concentrationofchlorideintherainwaterentering thecatchmentarea. Inaddition,Feth (1967) points out the remarkable similarity ofaverage Mojave Desert precipitation to bulkprecipitationcollectnearthePacificOceaninMenloPark,Ca. Thisfurther suggests that the chloride content of the Bristol Dry Lake precipitation is high because ofextra-basinalsourcesandnotfromrecyclingofthehalitedustattheplaya surface. Average annual precipitation can be obtained from measured values for Bristol Dry Lake and the surrounding area (see Table 2.1). The total thickness of the halite deposits isknownfromthecores(about 168mofhalite)andtheareaofthedeposits can be estimated fromtheknown extentof thehalite depositjust belowthe surface (Gale, 1951). The size of the area used in the calculation was varied to determine how much variation is possible (Table 8.1). The catchment area is estimated from the present, known catchment area (about 2,000 km 2), and also from the inferred pre- Amboy basalt flow area (about 4,000 km 2). The final necessary variable is the amount of timerepresented bythehalite. Thetephrachronology,presented above,givesa fairlyaccurateassessmentoftheamountoftimethatmustbeaccountedforin the precipitation of the halite beds (3.7 ±O.l m.y.). Using all this information, simple calculationscanbemadeto determinetheamountofchlorideavailableoverthis time period for halite precipitation and how much chloride is needed to precipitate the calculatedvolumesofsaltinthebasin. Theresultsindicate thatfortherangeofareas between9-50km theamountofchlorideneededisroughlybetween5-1013and 2, 3T014molesofchloride(Table 8.1). However, areasonableaveragevalueforthesalt panareaisprobablynogreaterthanabout15km2(9.3 -1013moleofchlorideneeded). Toobtaintheneededchloridetoaccountforthehalitein the 15km2 area an average inputofchloridemustbearound5.8mg/1per100mmofrainperyear,ina2000km 2 basin (Table 8.1). It seems somewhat fortuitous,but those inputs and dimensions are almost exactly what are seen today at Bristol Dry Lake. If the catchment area is doubledto4000km2 thereisamoderatesurplusofchlorideavailable.Basedonthe , above results, it is possible to account for all of the chloride in the basin by the direct precipitationofchlorideinrainwaterintothecatchment area. This result is important because it has recently been suggested that hydrothermal brinesmustbe thecauseoftheunusually highchloridecontentinBristol Dry Lake (Hardie,inpress). Thereasonthattheamountofhaliteaccumulationandthechloride concentrationsappearhighcomparedwithotherareasis thatuntilnow,itwas impossibletotakeintoaccounttheamountoftimethatthesedepositsrepresent. Given therelatively slowaveragedepositionalratesforthebasin,itsproximitytothePacific Ocean, and its large catchmentit is not that surprising that chloride in rain water area, can account for the chloride in the basin. abundant cation in the brine and the freshSodium. Sodium is the most groundwater, even though large quantities of halite have precipitated throughout the Table8.1. CalculationsoftheamountofClandNa(inmoles)neededfromaerosolsto accountforthevolumeofhalitein thebasin. thickness of salt: 163.8NaClm pure estimated area ofhalite: amountofNaor Cl needed: (km2) (moles) 9 5.4561x 13 15.36 9.3117xl0 13 25 1.5156xl0 14 49 2.9705xl014 CHLORIDE drainage basin= 2000km2 annual precip.: 50 mm 100 mm 200 mm 500 mm rain water Cl (mg/liter) 0.2 2.0873xl012 4.1745xl012 8.3491xl012 2.0873xl013 5.8 6.0531xl0 13 1.2106xl014 2.4212x1014 6.0531xl014 90 9.3927xl014 1.8785xl015 3.7571xl015 9.3927xl015 drainage basin=4000km2 0.2 4.1745xl012 8.3491xl012 1.6698xl013 4.1745xl013 5.8 1.2106xl014 2.4212xl014 4.8425xl014 1.2106xl015 90 1.8785xl013 3.7571xl015 7.5142xl015 1.8785xl016 SODIUM drainage basin=2000km2 rain water Na (mg/liter) 0.2 3.2188xl012 6.4376xl012 1.2875xl013 3.2188xl013 1 1.6094xl013 3.2188xl013 6.4376xl013 1.6094x1014 5.8 9.3345xl013 1.8669xl014 3.7338xl014 9.3345xl014 drainage basin=4000km2 0.2 6.4376xl012 1.2875xl013 2.575xl013 6.4376xl013 3.2188xl0 13 6.4376xl013 1.2875x1014 3.2188xl014 1 5.8 1.8669xl0 14 3.7338xl0 14 7.4676xl014 1.8669xl015 history of the basin. Although some sodium may be lost by adsorption on to clay particles, the precipitation ofhalite appears to be the only significant sink for sodium. Its sources, however, are not as well defined. According to Eugster and Hardie, 1978, the major sources of sodium are: 1) dissolution ofpreviously deposited halite, 2) the weathering of feldspars, and 3) the atmosphere. It has been shown above that the dissolution ofpreviously deposited halite cannot account for the Na/Cl ratios in the basin, and there are no halite beds known inthe basin either at the surface or in the subsurface of the surrounding catchment. Therefore, the first source is not likely. To determinetheinfluenceoftheothertwosources, budgetcalculationssimilartothe calculationsusedformeasuringchloridebudgets canbe usedtodeterminetheamount ofsodiumsuppliedfromrainwater. Theamountofsodiumcontributedbyhydrolysis thenbedeterminedbydifference. AswiththecalculationsforCl,usingreasonable estimates of the 3-dimensional size of the salt bodies, estimates of the concentrationof can sodiumfromrainwater,andbracketing theamountofprecipitationintothebasin, almost all of the sodium can be accounted for by rain water input (Table 8.1). This is notto say thatsodium isnotderivedfromhydrolysis reactions; weatheredfeldspars are prevalent in the soil zones. However, based on the calculations, rain water input may be the dominantof sodium to the basin and not hydrolysis reactions. source Calcium..Calcium is the secondmostabundantcationinthebasin.The precipitation of calcite reduces the bicarbonate in solution toward the basin-center, but it does not greatly affect the calcium profile (Fig. 8.2). Because precipitation of calcium carbonate occurs on a 1:1 molar ratio (Hardie and Eugster, 1970), precipitation of carbonates should affect both the calcium and bicarbonate concentrations by the same If mole ratio. This is not the case in Bristol Dry Lake (Fig. 8.2). the only source of calcium andbicarbonatewasfrom thedissolutionofancientcarbonaterocks (evenif some are dolomites), subsequent precipitation of calcite should show similar trends in the concentration profiles. The lack of such a correlation suggests that the dissolution of previously deposited, ancient carbonates in the surrounding drainage area cannot calcium. account for the excess Ifcarbonatescannotaccountfortheexcesscalcium,whatare theotherpossible ofCatothebasin? TheeasiestwaytoaccountforanexcessofCawithoutthe sources Mgwouldbethedissolutionofpreviously deposited gypsumor addition ofHCO3 or anhydrite. Unfortunately, there are no outcrops of evaporites in the entire drainage area tosupportthishypothesis.Unlesssulfatebedsexistatdepthanddonotcropout, the dissolutionofpreviously depositedsulfatescannotaccountfortheexcesscalcium. Another or source of calcium is the hydrolysis of silicates such as plagioclase feldspar pyroxene (Eugster and Hardie, 1978). Complete disintegration of granitic cobble Cl concentration (moles) Figure 8.2. Cross-plot of chloride concentration versus calcium, sufate, and bicarbonate to illustrate how calcite and gypsum precipitation effect their concentration toward the basin-center. The higher the chloride concentration the closer to the basin-center. After calcite precipitation, bicarbonate decreases and calcium and sulfate both decrease slightly as well. After gypsum precipitation sulfate decreases. In the of area gypsum precipitation the slope of the calcium line is less steep, but after gypsum precipitation it increases. clasts in the soil zones indicate that calcium could be derived from this process, is difficult. however, quantification of this process ofcalcium whichEugsterandHardie(1978) considerto A third possible source water. beminoristheinputfromrain However,rainisnottheonlyatmospheric sourceofcalcium. Dustalsocarriesasignificantamountofcalciumassolidcalcium are located in non-carbonate carbonate. Studies of calcretes in alluvial fans that areas indicatethatcalciumderivedfromatmospheric dustmakesanimportantcontributionto the total calcium budget (Gile etal., 1966; Mayer et al. 1988) however, the , introductionofcalcium asCaCOswillproducetwiceasmuchbicarbonatethancalcium onamolarbasis.Therefore, theadditionofcalciumby dustcannotaccountforthe excess calcium. To fully assess the contribution of calcium by rain water, assumptions would have to be made as to the quantity of calcium sequestered by gypsum and calcite precipitation for the history of the basin. Because there is tittle 3-dimensional data that couldevenmoderatelyconstraintheactualvolumesofcalciteand inthebasin, gypsum suchan attempttobalancethecalcium budget overthehistoryofthebasinissubject to errorsthataretoolargetomakeit ausefulexercise. Thefinalpossible sourceforcalciumis ahydrothermallyderivedbrine. At present, a hydrothermal source cannot be discounted as the source for calcium. excess However,ithasbeenshownabovethatitis notnecessarytocall onahydrothermal source fortheabundanceofchlorideinthebasin soit isunlikely thatahydrothermal brine is the cause of the excess calcium. Until more quantitative data are collected, the source(s) of the excess calcium cannotbefullyexplained. Argumentsbasedontheamountofinformationknown, suggest it is likely that the main sources of the excess calcium are rain water input and hydrolysis of silicates. Precipitationofcarbonateminerals. TheexcessCainthebrinesuggeststhatthe limiting factor on both calcite and gypsum precipitation is the availability of the anions. EugsterandHardie(1970) demonstratedbycomputersimulations thattheevolutionof thebrinedependsontherelativeconcentrationsofCa,HCO3, andSO4intheevolving of the brine. Upon evaporation of a brine, gypsum will not precipitate until most HCO3+CO3 have been removed from solution. This type ofevaporation path is not obvious in the groundwater chemistry data, possibly due to the low starting concentrations ofHCO3 in the solution (Fig. 8.2, and see discussion of bicarbonate), butitis certainfrom theabundanceofcalcium insolutionthattheformationofcalcium carbonatemineralsis not limited by the availability ofcalcium. The ofexcessCainthebrinesaccountsfortheabsenceofdolomitein presence thebasin. Ithasbeenwellestablishedthatasaminimumrequirementfordolomite precipitation, the Mg/Ca ratio of the water must be greater than about 1 (Folk and Land, 1975). In Bristol Dry Lake, the basin-center Mg/Ca ratio ranges from 0.03 - 0.7, and - aroundthemarginsofthelakeitrangesfrom0.05 0.5. Howeveronewateranalysis taken near the playa margin sediments has a Mg/Ca ratio of 3.6 (Fig 8.3). Figure8.3.Cross-plotofMgandCaconcentrationsandCa/Mgratios Cl versus concentration.AsinFigure 8.2,thehighertheClconcentrationthecloserto thebasin-center. Noticethatwherecalciteisprecipitating, theCa/Mgratio is greater than 1, but where gypsum is precipitating the ratio is greater than one. See text for explanation. Ifdolomiteis goingtoprecipitate, itshouldbeinornear theplayamargin sediments. This is also a mixing-zone area where the basin-center brines mix with the fresher groundwater. Mixing-zones are thought to be conducive to dolomite precipitation (Hanshaw etal., 1971; Land, 1973; Badiozamani, 1973) However, no dolomitehas beenfoundintheentirebasin. Thelackofdolomiteinthisareamayalsobeduetothe lackofbicarbonate. Calcite hasalready precipitated outascalcretein thealluvial fan little bicarbonate left once the calcium concentrations have been lowered. leaving very In addition, the presence of large amountsof sulfate may inhibit dolomite formation (Baker and Kastner, 1981). Therefore, the possibility of dolomite formation as a primary or early diagenetic phase is remote in Bristol Dry Lake. Minorbrine components Sulfate. There are three major sources for sulfate; 1) rain water, 2) dissolution of gypsumoranhydrite, and3)oxidationofsulfides. EugsterandHardie(1978) consider the oxidation ofsulfides source of sulfate. As shown as the most important above,thedissolutionofgypsumandanhydriteareunlikely aspossiblesourcesof sulfatebecausetherearenoknownevaporitebedsinthebasin. Contributionsfrom rain water can supply up to 30 ppm sulfate (Feth, 1967). If calculations similar to those madeforchlorideandsodiumaremade, theamountofsulfateentering thebasin suggests that there should be significant accumulations of gypsum or anhydrite. The gypsumdeposits atthesurface areextensiveandring theentirebasin, however, sulfate phasesincorearenotabundant.Itispossible thatthecorewastakentoofar basinwardtointersectthebulkofthegypsumphases. However, towardthebaseof CAES #l, the playa margin sediments pass into distal alluvial fan sediments without an increase in sulfate minerals. control to determine At present, there is not enough core thedistributionofsulfatephaseswithdepth. Untilmoredataareavailable,an explanation forthelackofsulfateatdepthcannotbeattempted. Theoxidationofsulfidesisanotherpossible sourceofsulfatewhichcannotbe addresseddirectly. Theamountanddistributionofsulfidesavailableforoxidationand therateofthisprocessarenotknown. However,thestablesulfurisotopedataprovide some indirect evidence for contribution of sulfides. At the surface, the isotopic value ofsulfurin thevarioussulfatemineralsrangesfrom6-9permilCDT.Ifsulfate were derivedprincipally fromoxidationofsulfides,thewidelyvaryingoriginal valuesof sulfurratioswouldthenbe averagedintheevaporatedbrineandgypsumwouldhave a narrow range of sulfur values. However, if the brine did not completely mix before gypsum precipitation, the values should be somewhat variable depending on the location in the basin.. In addition, depending on the source of the sulfur, the values might change over time. Finally, the average homogenized value could be almost any valuedependingonthesources, andthereisnowaytodeterminewhattheaverage valuewill bewithoutknowledge ofthesourceofsulfides. Thevaluesobtainedfromall the sulfates at the surface in Bristol Dry Lake fall within the narrow ofsulfurrange values for rain watersulfate(Kaplan, 1983).Ifthesulfateweredominatedbythe oxidation of sulfide, it fortuitous that the isotopic values obtained fall only seems withinthis narrowrangeofvalues. Although this reasoning is somewhat sketchy and based on flimsy evidence, it not the suggests that the main source of sulfate is derived from the atmosphere and oxidationofsulfides. Inaddition,otherworkershavedeterminedthatsulfatefrom precipitationmaybeimportantinproviding sulfateforgypsumformation. For example,Watson(1985),usingsulfurisotopedataasthemainlineofevidence, also suggested thatthesulfatefromgypsumcrustsforming inthecentralNamibDesertwas derived from meteoric precipitation (in this case fog not rain). Formationofgypsumandanhydrite. Ofthethreesulfatemineralphasesatthe surface,gypsumisbyfarthedominantphase. Thedistributionofcelestiteand anhydrite arenotaswellconstrained,but,fromtheavailabledata,itappearsthat celestiteismoreabundantatthesurfacethananhydrite. Incore,ontheotherhand, anhydrite isthedominantphase, celestiteoccurs inonlytrace amounts,andgypsumis absent. Thewaterchemistryprofiles acrossthebasinshowan increaseinsulfate concentrationfromthealluvial fantotheplayamargin wheregypsumisprecipitating (Fig. 8.2). Basinwardoftheplayamargin sediments, thesulfateconcentrationfallsoff The variability in sulfate concentration in the basin-center brine is probably due to local dissolutionofsulfate minerals.Asdemonstratedabove, thebulkofthesulfateminerals drastically and becomes almost a trace component in some analyses of the brine. are precipitated in the playa margin sediments. This type ofprofile, along with the data from calcium presented above, suggests that sulfate is the limiting ion in sulfate formation. Due to solubility differences, the limited sulfate availability also favors the precipitation ofcelestitewhensulfateis nearly depleted (seecelestiteformationinthe strontium section). The nearly uniform sulfur isotopic values for all sulfate phases at the surface, minimal at are including anhydrite, suggests that bacterial sulfate reduction processes at the surface. Even the anhydrite the surface has sulfur isotope values similar to gypsum and celestite, whereas anhydrite from the cores have values 5-7 °/oo lighter than the surface values. This indicates that bacterial sulfate reduction is not involved in the precipitation of anhydrite at the surface. Furthermore, the similar sulfur isotopic values suggest that the anhydrite either formed directly out of solution or from the inorganic dehydrationofgypsum. Withtheavailabledata,itisnotpossibleto determine which process is forming anhydrite at the surface. Thesulfurisotope valuesfromtheanhydrite in thecore are lighterrelativeto the surface sulfate minerals; but the tight cluster of subsurface values (between 1-3 °/oo), regardless of depth, is about the same as the cluster of values from the surface (Fig. 8.4). In other words. core. Corevaluesseem tohavethesamerangeofvaluesas thesurface values indicating a diagenetic difference rather than a depositional difference. Isotopic values are in °/oo CDT. Figure 8.4. Comparison of surface sulfate values with those of anhydrtite in the depending on the starting depositional value, the anhydrite from core is 4-5 °/oo lighter thanthesurfacesulfurvalue.Thesenarrowrangessuggest thatthevaluesare lighter duetolaterdiagenetdcreactionsratherthandepositionaldifferences. Frompetrographic evidence, it has already been suggested above that bacterial sulfate reduction was involved in the conversion of gypsum to anhydrite. If the anhydrite formed during the processofsulfatereductionratherthanfromtheresidualbrine,lighter sulfurisotopic values wouldbe expected for the precipitating sulfate phase (Kaplan, 1983; Vogt, 1986). Theone forone conversionofgypsumto anhydriteinaclosedsystemwillimpart avolumelossofabout40% (Shearman,1985). depositionalgypsum The retentionof fabrics with no porous areas indicates that calcium sulfate is imported to the site of anhydritization in relatively open hydrologic system. In other areas were the gypsum a fabricsarenotretainedandsomedisruptionofbeddingisevident, thedehydrationof gypsum may have been accompanied by dewatering. The imported calcium carbonate is likely togeneratedlocallyeitherthroughthedissolutionofisolatedgypsumcrystals scatteredthroughoutthe orsimplyfromtheresidualbrinecarriedtodepth.The core, retentionofgypsumfabrics is important,however, becauseit suggests atleasta moderate degreeofopenness to asedimentthatis dominatedby lowpermeability muds. Carbonate.ThefollowingdiscussionismainlyconcernedwithHCO3 asthe ofCO3forcalcite becauseHCO3is thedominantanioninBristolDryLake freshgroundwater. Thisabundancereflectsthegeneralimportanceofcarbonicacidin weatheringreactions(EugsterandHardie, 1978).However,bicarbonateinBristolDry Lakeis aminortotracecomponentofthebrinecompositionandnever reaches a concentrationabove 154ppm. Asstatedabove,HCO3isactually 2timemoreabundantthanSO4andClinthe alluvial fan groundwater, but by the time the water reaches the basin-center there is virtually no bicarbonateremaining insolution(~30ppm).ThelackofHCO3inthe calcrete in the source basin-center can be most easily attributed to the precipitation ofcalcite as alluvialfansandasnodulesinthebasin-center. Hardie(1968)recognizedthe importance of calcite and gypsum precipitation around the margins ofSaline Valley playa as the ultimate controls on the basin-center brine composition. A comparison of calcium, bicarbonate, and sulfate concentrations plotted against chloride (Fig. 8.2) from the alluvial fans to the basin-center illustrates the relationships between brine composition and evaporite mineralogy in the basin. The calcium concentrations plot closelywiththesulfateconcentrations, but,bicarbonateisindependentofbothcalcium andsulfate,anditsconcentrationdeclinestowardthebasin-center. Themost reasonableexplanationforthisindependent trendisthatbicarbonateisthelimitingion in calcite formation. Because calcium is relatively abundant in the groundwater and extremely abundant in the basin-center, as soon as enough bicarbonate is in solution, saturationwithrespecttocalciteis attainedandbicarbonateis removedfrom solution, thusmaintaining theconstantlowconcentrationprofile. The bicarbonate-poor analysis ofthebrinecomposition isthereason why thebrineevolves alongpathninthe Eugster and Hardie (1978) diagram (see above). The stable carbon and oxygen isotopic composition ofpedogenic calcite can be used to infer the origin of the bicarbonate for the precipitation ofcalcite. The 6*Bo values -8.0 ± 0.4 o/oo and -1.1 ± 0.2 for average The precise 6180 values for rainwaterinthebasinhavenotbeenmeasured, but,inregionalreconstructions ofwater analyses for the central Mojave region the 6 180 should be between -4 and -8 °/oo relativetoSMOW(Yurtsever, 1975).Using theequationofFriedmanandO'Neil (1977)fordetermining theoxygenisotopic fractionationbetweeninorganically precipitatedlow-Mgcalciteandwater,thewaterfromwhichthepedogenic calcite precipitatedshouldhaveanisotopic compositionofbetween-7and-9°/ooforthe temperaturerangeof10-20°C. Thiscalculatedvalueiswithintherangeofregional rainwater values suggesting that the bicarbonate in pedogenic calcite is derived mainly from a rainwater source. Furthermore, if the bicarbonate was derived dominantly from thedissolutionofancientmarinecarbonatesequencesfromthesurrounding mountains, theoxygenisotopicvalueswouldmostlikelybeconsiderably heavier,although bicarbonatederivedfromdissolutionofmetamorphosed limestonesanddolomite (Gale, 1951) may have relatively light isotopic values. A summary of solute sources given by Eugster and Hardie (1978) from the data ofHem(1970)andWhiteetal.(1963)indicatesthatallmajorlithologies (carbonates, sandstones, shales, basalts, acid and ultrabasic igneous rocks and metamorphic rocks) shouldyield waterrichinHCO3. This should leadtobicarbonate-richbrines inmost playa basins. Yet, this is clearly not the case in Bristol Dry Lake. Precipitation of calcite reduces the bicarbonate in solution toward the basin-center. It seems likely that therelativepaucityofbicarbonateinsolutionin thebasin-centerindicatesthat significant dissolution of limestones and dolomites and other bicarbonate-rich rocks in thedrainage areacannotaccountforthebicarbonatedistribution. Although some contribution must come from the ancient carbonates and the hydrolysis of silicates, the isotopicevidence suggeststhatthedominantsourceofbicarbonateisfromdirect precipitation into the basin. MagnesiumandPotassium. Magnesiumandpotassium areaboutequally abundantinsolutionin thebasin-centerbrines. Although dataon potassium are lacking in the groundwater, it appears that magnesium is much more abundantin the groundwater. Themainsourcesofbothmagnesiumandpotassiumarefromthe weathering of rocks in the basin (Eugster and Hardie, 1978). Magnesium from comes dissolutionofdolomite,theweatheringofMg-silicates, andlocallyfromweatheringof basic and ultrabasic rocks. Potassium is derived mainly from the weathering of feldspars and micas. There are no known magnesium bearing or potassium bearing minerals precipitating in Bristol Dry Lake. Although rare x-ray patterns contain peaks suspected ofbeingzeolites,they arenotabundantenoughtoaffectthepotassiumconcentrationof thesolution,anditisnotknowniftheparticularzeoliteisK-bearing. Authigenicclays havenotbeenfoundinthebasin,andasdemonstratedabove, theconsistentclay assemblage with depth indicates a dominantly detrital origin for the clays. Therefore, forall practicalpurposesnoMgandK-bearing mineralsformearlyinthehistoryof the basin. TherelativelyminoramountofMgandKwithoutamineralsinksuggests that there is a limited source for these ions. If weathering reactions are the only source for these ions, it seems that weathering by itself is not a dominantprocess in concentrating ionsinthebasin. ToachievetheconcentrationsofsayClorCa,whichdohavelarge mineral sinks, some other must be contributing to the solution. This lends process furthersupporttothehypothesis thatmanyof the dominantionsinsolution inthe basin have come from atmospheric sources. Trace elements; are three possible source for the strontium in the celestite 1) weatheringfromsurroundingrocksinthebasin,2)dissolutionofpreviously deposited celestite, or 3) hydrothermal brines. The dissolution of previously deposited celestite Strontium. There can be discounted for a numberofreasons. First, there is no evidence for the existence ofoldercelestitebedsinthebasin. Thenearestcelestitedepositisapproximately50 km to the west near Ludlow, Ca., on the other side of a drainage divide that has apparentlyexistedsincetheMiocene(Durrell, 1953). Inaddition,thecelestite near Ludlow is also lacustrine in origin. Therefore, even if the celestite in Bristol Dry did come from Ludlow, the problem is not solved. A source for lacustrine Sr is still needed. Thepossibility thatstrontiumiscontributedbyhydrothermalbrinesdoesexist. However, if it is derived from a hydrothermal brine the 87Sr/86Sr should be much lower than is indicated by the analyses of the nodules. Theconcentrationofstrontiumweatheredfromthesurrounding rocksismost consistent with the observed 87Sr/86Sr of the nodulesandtheamountofcelestite forminginthebasin. Theratioisextremelyradiogenic indicating acontinentalcrustal source of the strontium. The weathering of feldspars from the surrounding plutonic igneous rocks is the most likely source of the strontium. Current data are not sufficient todetermineifthesolutionweathering thecrustal rocks is hydrothermal or arelatively can be contstrained to the lower temperature solution, but, the source of strontium weathering of rocks in the basin rather than calling on some external source of strontium. Degradation of feldspars high in Sr can be seen in soil and calcrete zones in the alluvial fans ringing the basin, however, it has been shown that for of the many other ions, hydrolysis reactions only contribute a amount of ions to limited solution. This accountsfor therelatively restricted abundance ofcelestite which only forms as nodulesinanarrow zoneoftheplayamargin sediments. Formationofcelestite. Theconditions necessaryfortheprecipitationofcelestite arediscussedbyWoodandShaw(1976). Forcelestitetoformtheconcentrationsof eitherCa orS04mustbelimitedin thesolutionandtheSr/Caratiomustberelatively high. Wood and Shaw (1976) took the brine dataof Durrell (1953; see appendix IV) the brine 10 times the Sr/Ca of seawater. In addition, they argued that a restriction on the supply of sulfate and showed that in Bristol Dry Lake, the Sr/Ca ratios of are than SrS0because would have a greater limiting effect on the precipitation of CaS04 4 CaS0hasahighersolubilityproductthanSrS0However,theexperimentaldataof . 44MonninandGalinier(1988)indicate thatinaSrSO4-CaCl2-H20solutionat25°C, celestite solubility increases with CaCl2 concentration to the point where SrSC>4 becomes more soluble than gypsum. This leads to the formationof gypsum as a secondary phase by the dissolution-reprecipitation replacement of celestite. The Ca concentrationofthebrinemaythenactasanadditionallimiting factoron the precipitation and preservation of celestite nodules. Brine data from cores takeninthecenterofBristolDry Lakeshowverylittle SO4 inthebasin-centerbrine(r PyuCf*Wx£ •Of . C^- WLi.. COUNT state _ ~ - STRATlGHAPHICINTEftVA. 5 HO C,5> \ !|l! i I j j *)0 —s •:i _ , iiCz ! ;j1 ji I 1 1|j 11 i o Jb-fc”n\!t 1j 1j “¦ tU'-U-U-i 1 ill! C-‘“* ii y i i" •11 £---“--ij j!: ¦!: «#r* *> , _2n |& ]! !i |ji P **"*• |1 J® ii1!i | |if1 : 10-j 1 illMi i!i Ml itf! 1 ¦ii : I iiI ¦ii :'•.<>( III! o ’ (»W vV^f*5{ ! ; ( 111 o :.;} ilA. Ml X/7P !| : 1 ::i Iu ¦ Wi __ ‘ •'i*M M s ! i| 1I| i1 "1 i! ¦!!!i 1 1 •i. ¦ i 16 Li-C. ro ..-­ — > -J :Mr 1 w T Vtr6 :c"3-V) Cr b"[ /»y? "i tf ¦>. u Lc c 9f itvfx -•¦• ­ -20-: V;­ v/,jt/*«*v s ¦o5? 1 T ri /e>n ' ot) , 1V1 u”** A**— •**•¦'•k lf/ 0 iVyH:x‘ieTf.V->-»ev,v/-j_ —' »r — WE.. t,MI £>»Tt C COUNT* “ STFiATlGHAPrilCINTERVA-SVO &*7 Vw DATl 1/25j/9€ LOOGtD B> 1 Gh*ll. H7f ¦ . 1 ¦ ylU>1 :;i ‘""'CTufc.* t »«»»n « COMTUUTIUN f1 rk«tu.U2I r; miuw i (Ir_ ­ 1 ac *,jjLm i »£i: ‘>"o».tT, “ ,s -- /1 .o o :?CrV­ |! ;i ¦iii! _ ¦ «o ; , ftrSL* o ooQc: ¦ !« o_M —5 I *i1 1 )l s | Lvri $ _2a , 0-0 j . i T* 0 ._Sb '-QfiCJ 5ir ->•*. CVvS v. 1 1 s,“ryt *1 1 -v­ | ’ 11 : Mw-i* U^U • a£ ; I ! ! ' ! <(&>** 1 iUQJ sj.' IU j -Cj-N iG­ | j i 1 i i; si i i i i i HIM ¦ iqt ; TV , J 1 . ! 1 Q la *0 :t£p~er &*¥*>. **) i —' i <->«* | 1 1 j T ! I i ! i i i! 11 | li I ! i is «o 60 -Jj 1 i , , t^o _ c-° *1 O* T~n 1 i 'yf? i J i | : ! i I 11 i ! 1 i i 1 °1D LC.C, : ! 6® Cv*\s r~r Cr ?>>Wo-6 fr^+w^r.U vi | , 4& f-,, .. _ -o o /oTR . ¥x. >D1 ¦¦ -AN’ (if/x. _'J ”< /Of i.£c So '" 3 —crto - c”-WL ’ °' /6>fi. ‘ /v\ 0 Hz :rcr U2. *Su\* Q)cvr' SI ATI C.A STMATIGMAPHI1INTERVAL 6°1 O - 0 LOGGEC fc> wn V^/ee M.NIHA. Illf IW g * COM»OS'»iCia onc. »o«osm, C> tlhuduNIt *.* r »»H 5 niluai 1 IbO.UWiI) CNtVIAt fcUl ;l ? s 1 — •->* V *.= ! Mi-U-fiX-A i 1 > «L O - *10 1 i! „• ¦So C*»\ VJ-, U -- t­ . ¦ I i | ‘/o-y D l.n U>~ s • i&l so__£^?! i *! J f'rn J**L ; t| ! i : , 1 1 III i | 1 i I i i 1 1 1 i i !*-£; n ¦ Mh 3 iii i ••• O j-O cc f.W tLi-UU ~~r*7 1 , r -°r | ! 1 , MoI(a*v ! Or fJrrtX^v^) '°K V/C 1 i ! 11 h \ ¦ : I j i 1 _ r-e: :: i! ! o-o 0 o­- i cwtri 5. jS* 2. *> i=I<5fc> . , ': ; ; I * •S5i_ ; ¦J So c&«, •?h-¦ , 1 ¦ Cw«iS 2o j ! r.K !i 1 W­ gL: Df¦ Cr ! to i;: m: ! *° Crv\S r •)© ,!; ;;:I i 2> -) C­ % 'Ui friz* fetJU*-UU-U •j fKi 1 i **fi r o i ;! im 1'i i1 -1 ite -0 il 1 5° —=r—ri A' -s “SOU t­ | I]: •VX/ , 3*.. , . — .... _... «i AVR, V/2_ z» loboc , Sp3 ~ 'S-So 3 i® ­ - *t', C'f.i r C.o — V/ELL _ ft 6DL COUNTV »-»0 ETATt 0 ETRATlGRAPrtlC INTERVAL C? — LOGGEC-BY (A. UATC .11 Hi/ fi com»o iikt ’•OK 11 »YkOCTUAib TilluNl c £ GH*.,.,.« ,DO;OM'Tl CftTllAL Mil « r t: .._. £ - •.°"L''± . S S »o«o __80 CwS < i ’ I 1 ! . : $D 6^3 1 i i i i ; i i i =’-''• ib 'to j G_i O^i .. f-t 1 1 i j i i i ib !\fr *1L v *^D ° 0/fir t D °-S^ Is n?-W?7 i \ El - -«*< m <3P_ _ «o . ! : ! V-r % !=-* j C^S , 1 »?««. ’ 1 ! 1 i ; ! i i ! j p*V5, 0-»5 ;i T i ¦ •>AjW»J i I I I j i i ; < -0> i'cr'V" irV i -* -r-CO b­ !l I 1 1 i ! 1 ;«)> «, l*pL*z i ; : i ! j i HO to-­.’ =o So '-Pfr «=** «u» CC . K> vb£U“ vw­ -i-i >*r~ [ & i %: 1 * 1 : ! i | i *o "cntaq I-CC 5 u j 1 l| i i i 1 i ! i _. 5? 13n?3 -- £}t^ »rt *)0 " 3' S 2A> -• • *> _ £. *v«.V £V& % r.N©/«~i> k(*~a . «t> .-oc -** -<3 -«“ f CT 0­ - ?e> "o> Z-*!>DL "tLL coomty 1*0 situ C.A. STRATIGRAPHIC INTERVAL KA. /oA/ZS^ " LOCGEC-BV DATE A,ULfW \ r .l,(ifi* t/, •*"*»*•*. COIM-OUTiO*. ..NCL *. sTAuciuAcv ;-i jji ,m ;| TI,TW"‘ u E IOO,0*iU « H«. * £ = ¦ ?r ;j‘ ¦ £ c «j • >oiku; i S i u 1 ! I fc­ 1 1 : tiQ. M6ft a 6 «? rf &r ¦5T>-Cwi L LC SI ~ -~/ rp^e! r> 5 1 | l 46 <* C»5 i LC( : £i O » =ri,. <143 1 i “fcl -go * RO feo Ov\S ( AfO *$* *to 3c*. )/ 1 p ( J 1 ! .te -O'"! t-c.c I 1 i R •Ttt.f 1! i . •5C i 4P_ Sop j !•«. ! r : 1 R i , £/on 5Wn»a S-l 7 1 °m£ II <6 £5-^5 S ! -S . 1 i Ioffe. -1 . i 1 — i -:; 1 ; Xl , :ao 4; ;6o *?o Cr\5 y y/<_ i _t . H// 1 ' o •>* i i i cW> i 45*1 /cfti toyb'-L.-M'ii. h. II LLzzb ta Cy*5 5 "* iyJ-»«_­- /t>£­“*o­ cil^ 5K '*y “ 0-0 h 3*> rV>] 1 _~_ --:­ VC i i T-O £, r\_i r.° (F 4 2.2 5 i is 0^3 =3 i (14. .+—t;l\w-\.t V? c iL> 1 v^v-. i NT*** V £2*l :C«s | 1 1 An -6 -o’l - i i | i LKh • « V\ V\ 1 cR i On R o-J i .c* .-*>'-•" io«4/z_ «*•*> 1 i i r 'US C-O i; ' a o- C­P.i 3iraL '6 7*>fe f Q_­ /%^ sU- *3 o-o­ ] • -o o ;oTn>f, ¦ \e> o-% f- J,i 1 1 z"o" a -- WK. ¦A. ASH -O' We, i *¦¦ -"•o-a£ i* "c> a ¦ •. Jb - l' o _-o C£­ —“T -T _ T1J« A1N(103 C2M' - 11V1S 3iHd»Hou»ms i»*»hini -ahbl kg'; a *930901 0 »°31 5>8lhjai~ MIVHO 37IS J1U?Oi . .1.. :2 ““2»| sinnionwuIsi I»U*0*0<¦*• 13WWtt llj *0>l>«0; o5My-t ii 1 _J Q<* cr - - Ul*; *.­ 0^01 wi 3 -n~i 7'~2( oe I 7T jr ¦HW^, '!«• 7-r-Z^ > —«7^t« 5* o°SEEpr ¦ <3 0-0-ft.1 <39 0 ? °L -rr? 1 “1 > b)—*O T*** r 1.“*-¦-Ss ¦ OC* I-SA 5W7 a? 1°o° S ; i1 >y^s 1 •1 11 -«3 Q+. 1 KW«» \ * i ;;>.*’/ • »<•/ 1 oCc* -¦ 'ntjp'i i rw­ ii -9; 1 ofr os: 09 Ofc c ! “¦ otr ; -ait* * ¦' _ 0 3. ot, ZrO~ qS , o& - a H) O' V/ELL £.3 2~ STRATIGRAPHIC INTERVAL J S *? 0 * M.UfBlI ic-tTMUCtUMfS . Irtl COMPOH1IO* V •\\ Ml C\ A-i\ — l&» „„ itfl MNCk vxf woju., g& s .. „„ T p1 i % fK 4Q £ VftX/ Id v id 4o S« or^rAr11 - O & O^WIt £ * ¦¦• WsWctuWuJAx 6-i/fcV.-3,') | Kftta. /fcSo Cn^ * 4o , r? OmS J 0 S * ¦i-=r O H ¦no $ 5^. 'Jw. IG­ A s on r Sc^V\ (— A1 COUNT> o STATE _ *V/O * K*,& DATE 1 LOGGEDB* >0m &>*¦« W| t iU£i.tl»«.l» iOO.OMill H :r8 . « CKTHA. »IZt. > -;s :” y8 u . ; .! ¦ i 111 !i i !1 1 1! ASi l U, 'M* K i j I -3^! L-hp /5 P. Tt-WO y» -dD-| •Itf, i-kU. L l i ; n 1 I 1 T r . < f Cv*3 C?J> Q _ * ~ -C) A . ' !-• ' /lumtw'l AN j 3o -*> -Li - I*** xS v~C .r*> . V° ~ WELL £ STRATIGRAPHIC INTERVAL $T -K u,......vV A |. 1. nhociuHU CO-0.,,,0. ' -v VuU «* -5 1 •f «£° 4 >* Cb 11 n nmT / Uauiiuuu .?> '•v ¦*“ I-td |o7fi^ t_ /•* / VI-ti ! h. i_J i iis!>;«K 0 ....... v. Sda/Tfc. C | 1 COUNTV ki2i cSfer£„ StNW uOGGED Bv c H | i 1 j ! ! ¦ i ; lomifc ; ; I ! i i 1 ; ; « , 1 STATE C­-DATE Cl J 1 A / V /y<^ 1 1 1 i 1 1 , 1 i | 1 1 1 1 I C Ae.<;* / WELL STRATIGRAPHIC INTERVAL 52.0" (5 O /$4? BY 1 DATE LOGGED in CRAIN SI2( MINfRAl u °D»STHOCTUH tsj 71 7 (OOlOMIU It] on cl 1 to Tv,>f5 55"r0 2 ) ROROSITTJ ftp. r° Ite. u b> •" :¦••; :r: tiopbruty fyrte * : -¦ -fe 1 :• Go -— ! h.V S' l 3p — t>0 -\J: -j. : ; ; : 001 5t> : ~ j : ;. -; ....... ... .. £ Vv\ ji« So ; j ¦ 3T> ao ; .'.z;”7. . i; -;: ¦: — '• ; ; -j ; ; ; -J; 50 s* : --::: J ro fro 5 HO so ¦¦¦••¦¦•r::'Q --L SJ.t. 0-5 6\­Sp­:: ¦ ¦ -: :. : . :::. iv . Ho *> ;; \ :­•y-~ ¦-" : “spwli ¦ *­r Am.wi Kw tVs. <1 u_ :. L IS sss • • ; • : | . j “ :r -! ¦.; ¦¦ £ \ . .. . -. ¦ I -_i_ ¦' : : la.— Jp -TS­I:::’ ¦mLvV -: 'k: :. p;;;;; :j .:, S 5r> S -:•••>¦•( l>­h .:: ¦ : rr. :: :: ': -'¦ ': ¦' So ¦; 'TO -ro C 0 . j. .... _ . . :..., \1 Vo v-\ ( 1 ¦j I S*’ HO 0 ) 1 / I 1 | A 1 i 1 1 ! ; | 1 :>( ! r­i j i H £t> $T> bt” 1 => A 1 1 1 ! I 1 1 1 jl1 i j ! 1 1 ’ ! j i ! 1 , 1 ! . 1 ! | i 1 ¦ if! 1 < £­ WELL G“tc< STRAIIGL-PHIC INTERVAL &3sl i>2-& LOGGED BY IV p. DATE 10 I $£ 5> w J -I =*o MlNfHAl (1NCL POAOUTV) P Tt*I I \ 1 j! | GP AIK Mil (OOwOMITI c c 7 s > 4i­fcT) *0 s'* 5b £ it HP S 6 \J 1 «« ¦ /O 5 ..... Ht> <5» So m ........ ._ . -.. _. ¦ : M 1 --­.; < 3t> 5° -4. : : :.:...: . . Ho . ' . ' • ¦ •2a» 5° "bo uo io / i i ..... . ¦ . . . > ¦ r :. r’;: • • ' * v/| ; j.. ¦ . ¦ is : Zc> /o HO ? v iCr ;_ -; -• i: ¦; ¦¦ --... t . ¦ y -— 5o Oi”: Girl: 6o n-o • to :;r: r . Vo / aCr> \ .... ; . -( ¦O'— -k/o 12-0 3® VD UfP . .-¦ £ ... ~v. V* ( .... £ \ -’: ;v " :. ;;:i.:-:; : ¦ • 3-:,,. £ / bo • * 1 ¦ : SO _. I . < \ -­— “ --~ “ ¦ ¦—¦ *¦¦ ¦i :J": 3? ¦ Co 1: * j . 5c 1 to 1 i ¦ ( 1 1!i - 1 4c 'SO ) i \UV Zp j 1 i3o gtn>-:Ho 1 i) 50 1 I G)\ lit,lo-1 30 Z.O­ WELL c^es^/ STRATIGRAPHIC interval -S ;C - °i ?-c LOGGED BY KfaL.1 J2bS»=­ DATE /O (9V 0 m w-JUJi MINIMAL MNCL MOMOSITV) p ’« j|\ TtITUNE ! (.MAIN Ml Z t (DOLOMIK CMTS.IAI. M21) e “m 5 ; « c i c 1 “ 1 a o lo !% (go r­3t< L =^A l\ * s*. * to A . ' D (0 ; * SS» t*> 5o '<< ,pL itTil : *-; ;¦ ¦ t. c 3 l&t «i+­ =iK : *0 3*» 2fe> “TO ¦toO <0 s> C -• ----­i. HS 1 .:.: ;_¦*¦: i~;. ". .. ¦ : . :7 : ;• ._ - - :l.n­ ... —;_.; — --: ¦ — _ -- . - * to_ iw us tintn? ¦J te?*. b&.-\ <­LA : dI.’K.V .1" .: ;.tj.>: — — - ~ P — • ¦ p­•. s¦ “ wi. <* -S'_Oe 0 fcj f 7 I |\ cr H _ O -: 1 — --¦ — ;¦: “ ; ¦ ;.l* .:¦ :r . ** J7 ; ! !7 ! . '.Z. '¦ - 2b, ... Sc.a,~ :: — " : • ~_'l"_ -5i! V :; : :.: :r 7 ¦ • 3* s»­<0­ E X Sfc. 1. •rd*. i f ll * u £. r\ ] •jj'ijuV S dtf/o° — -(3 ­-o~'j 4-?r ' ! * N W*Vrr>*J btA >— S iifc.J.-_ ni : } dT»i -'•’¦( = ' • “ Vb^ci­• 1 ! 1 ! I j p ! i -i, P'viU-S Sftr^ jlll 1 1 i j ji 1 | i i i 1 i ! i * i 1 I 1 i : I j 1 I i j I ¦ • A HMiMmr qfnrr^ x ;Ss fk^-Ayi4 f-io. I 1 1 CAr<^ ; well STRATIGRAPHIC INTERVAL °\ -// Z.O LOGGED BY \\\ . DATE ft> 1‘SC *I s o. MINE COMPOS (INC AL ITY)PONOt f°r,isTHUC LIKES* gsf gS, ""t IU CHAIN SIZE iOOLOMITI CHVSTAL SIZE) 1 . r o * 5 *1 v FOSSILS z i •*¦ --'1 . = M 0=X< —' V %r — 4 (mJuJ i ¦gns*'" a p* \l ‘•*|A l-r_2J dX To * wJer s , • •fU<*y­ LIfi_a"o ¦ ¦ : — >0 It* : ~r :;::: ~ "q <5 r.: !' ;; ... .. urip«iiu4i a .: — It ;-; fl.ai_lttf.W4f CN^v’l. ¦:. ufeV 7^-rr - ¦-r i:: £==r­_i A ¦yj-:' Itv oAvjLi 60 •— »— _.,. : \i~ "5dbo¦ ;¦_¦ ¦; :—. ¦1 . .... ; -jr r:H: ¦::: : ••-¦••¦ -To". - ; U*jfi.»s^W. S( ! #. ' -¦ ;;r ; : J f :• :• mwiHiuti i ,1 l A-J Q~bjo' I ';¦-*•••• ; i| i /'loo ~ *••1 xCsJk£l b*«i< 1 *b ¦V--V • 11 i Villi PlttMS i 1,i1 1 i p-1 -o'cd I i xA ' i ! 1iiiM1i;! /o— 1i j ' (ry-l ! AJ >0­ " rvL—¦ UnW>.»<4<«wA| • Wv-S­ D0*L*,-p o­ . Tv.. i -s^5" Zo~ A A-*L WELL - STRATIGRAPHIC INTERVAL jj ? C <2. BY \\A FlrSC.L^_ DATE toAS-J LOGGED j OH AIN ! MINERAL °C.ISTHUCTU*< ts hlZl IUOLOV.1 ( -i c fOSHLS MNCL =r 5 St| 0 ;gi Tvpt t, r1 POROSITY) ;t,e« ¦ 0 u lo^f>h. aJg'-S"­ t ) 'io iJumirPT | ..*J ¦t u r—_"_­ ¦*K II1U1J Ain 2 ' 1|| jsJ -JIG /I s>.^ jd z*7 La ¦¦ 'tso ;: - - • “«v • v=¦1 • -p* , ::¦ fo ...... 3.15 . .1 . -— 1 ~-C-T! : ;z : ..;¦.; 'I S AJ! ¦_J y - sso <> .... A : 111IWMJ iuA G f2) Uuti7>uiJ 1_!_. i_. _ - :r: :­ ; =L z~l TOf& ¦¦ ih. j :: 1 ' S5*' /iM-;i* liukMii)i -*SwL, WvicSiTZ — ~ ;; ; j- Jl J Pi1*2.‘ wfiyijr ilitS'i * ... /z> V® : i. &:£ .. ; ,; t}D-— ¦ ¦ : i "wA nmfrrr^iv­ iar ¦¦ H*. <^r 1 )1 P-lu.!* 1f) -Jo,; I J»l — i Pj'iocMa* j !' 'F%1 ICr I i|rj |! " 6i >i | U i 11 1 . |A!Un> 1 ! ;i i 11i 1 i: 11 1 : r,• i i , -rkstv . Ij nil"4 1 i5-ho ,'1i iy>ijs ; ] (JO'<( :S 1 1!': 11 j1 i I• '¦! kK?v. ' 1 I •.j:V •4?-. 0-1 - 1 s S(t*r "" 1 “ }D ­ ij UIf*H~ -r«* O4aXo 0 >9 ooo O5 Q4 0 o03 ** -2 « Z33 35 Ei JuOj JJ*i S! -»I»«d »J»mS' 0 ~*2IJiMTil N03,D»i si*n s%iJI »< •bJ iNinjJ «lS >' _-IU3 rJ *o > pp P •’.>(••*iI c4 j ;3 5 £A I , .'li/i'.i'i' u ,;i ?« £aV |2 ,l ji i* r\ 1*r 1 * t ,i:i*’ :.1.; io 1 i 0 II-*|sst i: ;.. 1 ‘ , v : * , l • *i,.;-3,, cr f 3 i o 1 i• ,V i s a,? 1 ;: i , o“pi ­ .. o....a*•' * ««i» v ...,';1: ... ,, r 'iJ--i: Zj­ t * v$i i: o i; ;:i i, i; ss««vH/f ; • .. w ;. * i' . 1 * ||( i 11 .’s 11*1 r ; • i *: ii*i-i * 1:yi nil i*PP-1i 1 'i'l Ur-if« : ¦ ,....., > . •: o /:•;.'y >A! ;i1: j I : ;; m. ;:: it. . iik 'i.aqH:*,,,'f, >*v T/Ij| .• ’‘!¦ ' ¦ •. ¦. • ••' . • ;.I.¦• H ?jl i; ...A -;r¦¦' ¦: ,¦• !.i j i!| : . i(i.li ;. : . Sv_',..__ ¦1¦¦ |1‘i/,.: M•’•' ...a . ;||¦;i; : . . ;¦ 'llSl.'l , S Gi \ 1 " ¦ *1' 1 ='" vx'O ni sb * Ii Q 9'f1,11 i$ <-< *• -T . a -C.V ¦ 'umil'l/A...,,..,. - 2-D WELL C A ?<**¦! STRATIGRAPHIC INTERVA I ‘jzz2_ l5m LOGGED BV IH.Ro DATE lo I hJV » w Vz minima; (INC MOMOS ITV | fcr CTHUCUJMIS -TVMt UKIuNI : CHAIN HZ I (OOlOMIU c i i « ,-1 | -JO lo. > 5 | a eTtT “J=^­ Hp | - (lOJMItPI C­ A ¦••¦¦ ;¦•-> rS CT .­ ._ _ r f'i.'Tuiu' VTuv&uiv. 6t> UU>*>w ; ; :¦: ... — - - UkkilUVt Cuivuto'i • Cr­ ' ¦ fa i k € ‘ - ; - ¦ ¦ ' . ¦ 3­ 5 5^! s. A'JTZillV, 5 .CSP • Sa - ‘ - : - - . • lZ“­ ? A 5 mmw | r- — i$v IH /-f:t ! > r-^A •• <¦ /e •:: ¦ ' -:­ - ¦ ¦¦• T -; - ! . -¦ i 1 ar •". - ¦& ¦¦ ::’ ::::-: i| i Cr \ at.0 : ¦¦ . 3c-¦ -S*i s=; 5 _ _ i 1 H1>U)\U’ D> HII* UN i a?**­ I1 |: A i^\ fgo-Ak% I1 ;| 5 5| ¦ .| 1 - •. M'*’ ’O "s •K. A UJUti/JIIJ f --v O c; I s i1I5;? 7 Cw 3r> ss 1! .. IP Mfont Ttri lit! S17t ii coion CFMFWT r l'' )M *i:'l 'M rr_ O' 11 ; -i £ lllr III .¦•¦•' o> • ; I ; ¦ .'• i'M ; T.V.'-•.¦:> • ; a .-;... 1 ;• r; a : 11 y .VV ;: ; VI. i;!;:: : : Va IaJ­ : . 1 o 1.\» ii ! m 4­ ‘ ¦. • .¦- $ •s.¦.. V« Q/3 ' ¦ ... '•‘'‘ ' •. • . ;.. v. :1 1 . ; : r *: j 1'1s; i . -4­ ; -Li -—-I-—— —— V pju 1 M :• - i. 1 .. || •¦.' k-ii ¦ : ::r » : :i n­ Y0-" /jfa i!3.’* i,|' u ' iffjif' _ IA v ni’j] ;|'i W* V i 3 •• W'’: : . 1 .i - 'k“ .¦';>|1-Wv. '¦; 5 ..u iu r\ u \(A)-s//c-r>\l(\ c* WELl ' * V\ wr 3TO~ dA~f c. & 2. INTFRWA: [ C> / T E J MINU-*.. jo jCTBUCYOBtS TCXIUMt ° ;| "*“¦• E 1 .OHOi'TYl |J 1 w Ty*>| |k TTapjfpi.: “v* ' ® ’ A.< iOeiiiew ' K4.^«*TtA i 'Isr* * * £ I c ! B ' 1 - 0**1* M2( lUOkOMITi CMVblA. U2t; LClCGFt' ,be •-* C. BY FOfcSILb 1 5 t DATE &rfrij*v^«^4ju !«<•* ¦ ' : ¦ ¦ " I I -— ' ‘ ••.... J ..... • Id; • j • ¦ :. . ¦ .. ¦ ; i ; ¦ ' ; . • : . J _ ; II j GEOPHYSICAL LOGS from BRISTOL GEOPHYSICAL LOGS FROM BRISTOL DRY LAKE TEST WELL NO. 1 DRY LAKE TEST WELL NO. 3 Appendix II: Thin section point-count data POINT COUNT DATA BRISTOL DRY LAKE THIN SECTIONS 1= %CHEVRONS,2=%CLEARHALITE,3=%TOTALHALITE, 4=%ANHYDRITE or GYPSUM*, 5=% MUD, 6=% CALCITE, 7=% SILICA, 8=% DETRITALQUARTZ#, 9=% POROSITY, 10=%OTHER 0.01=TRACE *ANHYDRITEDISCRETEFROM MUD #DETRITALQTZDISCRETEFROM MUD 500 PTS. PER THIN SECTION Core CAES #2 HALITE THIN SECTIONS depth (feet) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) 506 0 79 79 0.2 18.2 0.8 0 0.01 1.8 0 902.3 0 93.8 93.8 0 5.6 0.2 0 0.01 0.4 0 1008.6 0.2 86.8 87 0 10.6 0.2 0 0 2.0 0.2 1034.9 0 90.6 90.6 0.2 6.4 0.2 0 0.2 2.4 0 1043 2.8 89 91.8 0.8 4.6 0.01 0 0.2 2.4 0 1144.2 0.01 86 86 0.4 12.0 0.6 0 0.01 1.0 0 1173 7.8 84.8 92.6 0.4 5.8 0.4 0 0.01 0.8 0 1175.3 1.0 84.2 85.2 0.6 9.8 0.6 0 0.6 2.6 0.6 1177 0.2 91.6 91.8 0.6 5.6 1.0 0 0.01 1.0 0 1177.3 3.6 73.8 77.4 1.2 17.8 2.0 0.01 0.2 1.4 0 1178.2 2.2 80.8 83 0.2 14.8 0.4 0 0.01 1.6 0 1315 4.2 89.2 93.4 0.6 2.8 0.2 0 0.01 2.6 0.4 1360 0.6 65.2 65.8 0.2 30.2 0.6 0 0.01 2.8 0.4 1360.1 0 79.6 79.6 0.4 17.8 0.8 0 0.01 1.4 0 1443 0 85 85 0.2 12.0 0.4 0 0.4 2.0 0 1459 1.0 87 88 0.8 9.0 0.4 0 0 1.6 0.2 1459.4 1.0 84.8 85.8 0.01 10.8 0.2 0 0 3.0 0.2 1460 0.4 93 93.4 0.2 3.8 0.01 0 0 2.6 0 1464 0 85.6 85.6 0.2 13.4 0.4 0 0.2 0.2 0 1576.5 0 81.6 81.6 1.4 13.8 0.4 0 0 2.8 0 1579 0 86.8 86.8 2.2 7.4 0.2 0 0 3.4 0 1.1 84.1 85.2 0.5 11.6 0.5 0 0.2 1.9 0.1 averages CAES #1 572.5 0 46.4 46.4 0.01 39.4a 0.4 0 0.2 13.6 573.6 0 41 41 0.01 39.6a 0.2 0 0.8 18.2 0 0 861.1 0.2 87.4 87.6 0.01 5.4 0.4 0 0 2.2 4.4b 863.9 0 39.8 39.8 0.01 0.6 0.01 0 0 10.8 48.8b averages 0.01 53.7 53.7 0.01 21.3 0.3 0 0.3 11.2 13.3 a microporous mud, b ash Surface halite Westpan 24.4 36.6 61 0 0.8 0 0 0 38.2 0 West pan 32.6 41.8 74.4 0 0.2 0 0 0 25.4 0 East pan 40.2 29.6 69.8 0 0.4 0 0 0 29.8 0 32.436 68.4 0 0.5 0 0 0 31.1 0 average Halite bed 2 meters deep 0 87.687.6 0.8 5 0 0 0 6.6 0 1.8 80.4 82.2 3.8 5.6 0 0 0.6 7.6 0.2 SURFACE GYPSUM POINT COUNTS 500 POINTS % Sand in sediment Sample # Depth % Gyp. % Mud % For. % Gypsum (Feet) LSC 1/25/87 8.5 63.4 7.6 6 23 82.3 TRCl 4 52.8 10.8 4.4 32 77.6 TRC9 4.5 53.8 5.2 7.2 33.8 81.3 TR CIO 5 65.6 2.6 2.6 29.2 92.7 58.9 6.55 5.05 29.5 83.5 Averages Appendix III: Isotope data 223 Sulfate data Sample # Min. 634S depth(m) rep. more rep. 6D rep. 6 180 Gyp Mine GYP 8.0 1.22 BC1G GYP 9.4 0.0 9.4 9.5 -67.1 BC2A AN 7 0.0 6.9 C2-C3 GYP 8.3 1.3 TR3 0.2M GYP 7.4 0.2 7.6 -67.1 -74.5 TR3 0.6M GYP 8.4 0.6 -72.8 TR3 1.15M GYP 7.4 1.15 TR3 1.3M CEL 7.9 1.3 7.5 TR3 1.4M S. GYP 8.3 1.4 TR4 0.3M S. GYP 7.5 0.3 TR4NE 0.71 GYP 7.2 0.71 7 TR4NE 0.71 AN 7.7 0.71 TRC1 0.9M GYP 8.8 0.9 -88.0 TRC1 1.8M GYP 8.4 1.8 TRC2 1.2M GYP 8.9 1.2 TRC2 1.9M GYP 7.7 1.9 8.1 TRC50.5M GYP 7.6 0.5 -86.4 TRC5 1.55M GYP 8.2 1.55 8.1 -84.4 -0.8 TRC60.36M GYP 7.3 0.36 -85.0 -82.9 -2.6 TRC6 1.5M GYP 7.8 1.5 -78.3 TRC70.5M GYP 6.8 0.5 -80.5 TRC80.35M GYP 7.0 0.35 7.4 -79.7 TRC8 0.95M GYP 7.4 0.95 -80.3 TRC90.85M GYP 7.5 0.85 -82.4 TRC9 1.3M GYP 7.2 1.3 8.0 8.0 -84.4 TRC9 1.3M CEL 8.2 1.3 8.6 TRC100.3M GYP 7.1 0.3 TRC10 0.65 GYP 6.3 0.65 -67.8 TRC11 0.1M CEL 6.4 0.1 TRC120.4M CEL 7.4 0.4 TRC12 1.7M CEL 7.5 1.7 TRC13 0.6M GYP 6.1 0.6 TRC13 1.9M GYP 6.8 1.9 -71.9 TRB l.OM GYP 7.7 1.0 GYP-C GYP 7 1.8 -74.9 CEL-G CEL 7.6 1.8 PRS-6 GYP 7.9 0.15 Sample # Min. 634S depth(m) rep. more rep. 6D rep. 6 180 TR4 0.3M* GYP 7.3 0.3 7.5 TRC130.6* GYP 6.5 0.6 6.1 TRC8 0.35* GYP 5.3 0.34 7 7.2 BC2A* AN 7.4 0.0 7 TRC11 0.1* CEL 6.3 0.1 6.4 1) 4’ GYP 8.4 1.22 1) 2' GYP 8.1 0.61 1)7" GYP 7.9 0.18 3) 6' GYP 8.4 1.83 8.8 -95.1 3) 5’ GYP 7.8 1.52 3) 4' GYP 8.1 1.22 3) 3' GYP 8.5 0.92 -90.8 3) 2' GYP 8.3 0.61 3) V GYP 8.2 0.32 3)3" GYP 7.7 0.08 7.6 -70.5 5) 5' GYP 7.9 1.52 5) 4' GYP 8.5 1.22 5) 3' GYP 8.4 0.915 5) 2' GYP 7.8 0.61 5) r GYP 8.8 0.315 5) 0-3" GYP 7.9 0.08 CAES1 1694' AN 3.5 516.3 CAES1 1578’ AN 3.6 480.8 3.3 CAES1 1573' AN 4 479.4 CAES1 1457' AN 3.4 444.1 CAES1 1455' AN 3.9 443.5 CAES1 1453’ AN 3.9 442.9 CAES1 1452' AN 3.4 442.4 CAES1 1392' AN 3.3 424.3 CAES1 1117 AN 1.5 340.5 CAES1 1063’ AN 1.5 324.1 CAES1 1047 AN 1.1 319.1 0.7 CAES1 879' AN 2.9 286 2.9 Radiogenic strontium data from surface celestite of Bristol Dry Lake, CA. Location 87Sr/86Sr error 0.712355 0.000035 TrenchC121.7mfrombase 0.712319 0.000025 SW side of playa 0.712132 0.000044 SE side ofplaya 0.712231 0.000028 TrenchCll 0.1 m from base 0.712175 0.000019 North sideofplaya 0.7122424 0.0000302 averages^ std dev= 0.0000941 Calcite Basin center calcite concretions Depth (inches) 6180 error 613C error 29 -0.86 0.07 -15.24 0.09 36 -0.89 0.06 -14.33 0.05 32 -0.53 0.08 -15.36 0.09 73 0.38 0.03 -10.46 0.04 70 0.28 0.04 -10.97 0.02 39 -0.95 0.07 -14.29 0.02 54 -0.57 0.06 -11.12 0.05 38 -0.51 0.04 -13.81 0.05 8 -1.17 0.07 -13.1 0.18 24 -0.89 0.04 -14.19 0.04 70 0.03 0.04 -11.11 0.11 36 -1.13 0.03 -12.27 0.02 25 -0.08 0.05 -10.16 0.02 27 0.03 0.03 -10.08 0.06 34 0.62 0.05 -9.78 0.05 84 0.43 0.06 -8.92 0.09 22 -2.31 0.06 -10.29 0.02 24 -2.16 0.07 -10.66 0.05 32 -2.22 0.03 -8.54 0.07 38 0.41 0.08 -10.33 0.15 40 1.04 0.08 -8.99 0.32 40 0.54 0.04 -10.2 0.08 47 -0.26 0.09 -8.94 0.09 48 -0.58 0.07 -8.21 0.03 51 -0.51 0.06 -8.79 0.06 54 -1.44 0.02 -7.59 0.06 60 0.72 0.04 -10.86 0.02 65 0.48 0.06 -10.46 0.17 69 -0.15 0.04 -9.51 0.04 72 -0.31 0.1 -9.08 0.23 6180 6 13C Depth (inches) error error Core CAES #2 depth (feet) 452.5 -3.90 0.06 -8.23 0.07 453 -3.65 0.06 -7.5 0.07 453 -3.57 0.06 -7.56 0.08 453.75 -3.75 0.06 -10.31 0.10 501.7 -2.81 0.02 -4.01 0.04 599.6 -0.86 0.06 -12.86 0.05 1090 -0.27 0.03 -15.66 0.06 1132.35 -0.64 0.09 -14.75 0.07 1138.5 -1.97 0.06 -11.7 0.05 1315.5 -2.07 0.07 -15.44 0.07 1500 -1.24 0.09 -16.27 0.07 1701 -2.52 0.09 -13.69 0.03 playa margin pedogenic calcite 6 -7.71 0.06 -1.02 0.07 6 -7.77 0.05 -0.92 0.06 6 -8.38 0.08 -1.28 0.08 Appendix IV: Water chemistry data 229 Bristol Dry Lake waterwell data. well; Cadiz No. 2 Cadiz No. 2 number in Figure 2.1 10 10 Source of data: Shafer, 1964 Shafer, 1964 static water surface (feet) max 213.5 min 208 ions (ppm) (moles) (ppm) (moles) Si02 42 0.0007 31 0.0005 Fe Ca 25 0.0012 26 0.0013 Mg 11 0.0009 10 0.0008 Na 55 0.0024 45 0.0020 K 5.3 0.0001 Mn Li Sr U total moles of cations 0.0061 0.0051 co3 0 0 1 0.00003 hco3 134 0.0022 139 0.0023 so4 34 0.0007 31 0.0006 Cl 50 0.0014 46 0.0013 no3 12 0.0002 p I B 0.6 0.00006 F 0.7 0.00004 total moles ofanions 0.0046 0.0043 Na/Cl ratio 1.7 1.5 TDS 332 302 Hardness 105 111 as CaC03 pH 7.8 7.9 Electric Cond. 516 dateof analyses 9/4/58 7/30/59 Bristol Dry Lake water well data com. well: ArchersidingwellNo. 1CadizWellNo.1 numberin Figure 2.1. 12 11 Source of data: Thompson, 1929 Thompson, 1929 static water surface (feet) max 280 220 min 259 208 ions (ppm) (moles) (ppm) (moles) Si02 Fe Ca 64 0.0032 39 0.0019 Mg 7 0.0006 2 0.0002 Na 150 0.0065 65 0.0028 K Mn li Sr U total moles of cations 0.0103 0.0049 co3 hco3 119 0.0020 159 0.0033 73 0.0015 SO4 Cl 166 0.0047 52 0.0015 no3 P I B F totalmolesofanions 0.0080 0.0049 Na/Cl ratio 1.4 1.9 615 289 TDS Hardness 188 as CaC03 Electric Cond. 9/27/10 8/15/10 date of analyses Bristol Dry Lake water well datacont. well: Cadiz Well No. 1 number in Figure 2.1: 11 Source of data: Shafer, 1964 static water surface (feet) max min ions (ppm) (moles) Si02 34 0.0006 Fe Ca 21 0.0010 Mg 12 0.0010 Na 48 0.0021 K Mn U Sr U total moles ofcations 0.0053 C03 0 139 0.0023 HCO3 S04 35 0.0007 Cl 41 0.0012 no3 P B F totalmolesofanions 0.0042 Na/Cl ratio 1.8 TDS 291 Hardness 102 as CaC03 pH 8 Electric Cond. date of analyses 2/23/60 Burris fresh water well 4 Shafer, 1964 206 (ppm) (moles) 47 0.0008 20.8 0.0010 6.3 0.0005 85.6 0.0037 0.0068 0 128.1 0.0021 47.5 0.0010 75.5 0.0021 9.3 0.0002 0.0054 1.7 373.1 78 8.1 585 4/26/62 Bristol DryLakewaterwelldatacont. well: Roy Tull Well No .1 Roy Tull Well No .1 numberin Figure 2.1: 5 5 Source of data: Shafer, 1964 Shafer, 1964 static water surface (feet) max 165 min ions (ppm) (moles) (ppm) (moles) Si02 17 0.0002 23.5 0.0004 Fe Ca 42 0.0021 66.8 0.0033 Mg 19 0.0016 24.3 0.0020 Na 135 0.0059 93.2 0.0041 K Mn li Sr U total moles of cations 0.0101 0.0102 co30 0 hco3 108 0.0018 149.5 0.0025 so4 82 0.0017 193.4 0.0040 Cl 215 0.0061 98.2 0.0028 no3 8.6 0.0001 p I B F totalmolesofanions 0.0095 0.0094 0.97 1.5 Na/Cl ratio IDS 599 634 267Hardness 184 CaC03 as 8 7.7 pH Electric Cond. 1/11/61 11/20/63 dateof analyses Bristol Dry Lake water well data cont. well; Roy Tull Well No.1 Riddle Well numberin Figure 2.1. 5 9 Source of data; Shafer, 1964 Shafer, 1964 static water surface (feet) max 120 min Ms (ppm) (moles) (ppm) (moles) Si02 73.5 0.0012 25 0.0004 Fe Ca 79 0.0039 13 0.0006 Mg 33 0.0027 4 0.0003 Na 97.8 0.0043 88 0.0038 K 4.8 0.0001 Mn U Sr U total moles of cations 0.0134 0.0058 co30 0 hco3 134.2 0.0022 122 0.0020 so4 284.2 0.0059 48 0.0010 Cl 93.3 0.0026 67 0.0019 no3 14 0.0002 1 0.00002 p I B F 4 0.0002 totalmolesofanions 0.0110 0.0051 Na/Cl ratio 1.6 2.0 IDS 735.5 300 Hardness 333 as CaC03 pH 7.7 7.7 Electric Cond. 467 date ofanalyses 2/3/64 5/25/56 Bristol Dry Lake water well data cont. well: Riddle Well Riddle Well numberinFigure2.1: 9 9 Source of data: Shafer, 1964 Shafer, 1964 static water surface (feet) max min Ms (ppm) (moles) (ppm) (moles) Si02 35 0.0006 35 0.0006 Fe 8 0.0004 Ca 18 0.0009 23 0.0011 Mg 7 0.0006 10 0.0008 Na 91 0.0040 71 0.0031 K 5.1 0.0001 Mn Li Sr U totalmolesofcations 0.0067 0.0067 co30 0 hco3 137 0.0022 142 0.0023 so4 53 0.0011 49 0.0010 Cl 69 0.0019 60 0.0017 no3 7.8 0.0001 p B 0.41 0.00004 F 4 0.0002 total moles of anions 0.0057 0.0050 Na/Cl ratio 2.0 1.8 TDS 384 359 Hardness 75 99 CaC03 as 7.3 8.1 pH Electric Cond. 596 dateofanalyses 5/14/57 2/23/60 Bristol Dry Lake water well data cont. well: Trailer Park well Trailer Park well numberin Figure 2.1: 8 8 Source of data: Shafer, 1964 Shafer, 1964 static water surface (feet) max 120 min ions (ppm) (moles) (ppm) (moles) Si02 37 0.0006 26 0.0004 Fe Ca 18 0.0009 19 0.0009 Mg 9 0.0007 12 0.0010 Na 62 0.0027 64 0.0028 K Mn Li Sr U totalmolesofcations 0.0056 0.0056 co30 0 hco137 0.0022 142 0.0023 3 so4 43 0.0009 43 0.0009 Cl 44 0.0012 53 0.0015 no3 p I B F totalmolesofanions 0.0044 0.0047 Na/Cl ratio 2.2 1.9 TDS 321 329 Hardness 82 100 CaC03 as 8.2 8 PH Electric Cond. date of analyses 2/22/60 1/13/61 Bristol Dry Lake water well data cont. well: McConnelRanch Well Easley Well number in Figure 2.1. 7 6 Source of data: Shafer, 1964 Shafer, 1964 static watersurface (feet) max 140 min 90 140 ions (ppm) (moles) (ppm) (moles) Si02 Fe Ca Mg Na K Mn U Sr U 25 26 12 53 0.0004 0.0013 0.0010 0.0023 23 0.001 14 7 61 0.0004 5xl0-8 0.0007 0.0006 0.0027 totalmoles of cations 0.0054 0.0047 C03 hco3 so4 Cl 0 154 39 44 0.0025 0.0008 0.0012 0 130 33 39 0.0021 0.0007 0.0011 no3 p I B F totalmoles ofanions 0.0046 0.0039 Na/Cl ratio IDS Hardness as CaC03 PH Electric Cond. 325 104 8 1.9 265 67 8 2.4 dateof analyses 2/23/60 1/11/61 Bristol Dry Lake water well data cont. well: Ca.Salt Co. Well No. 1 shallow well numberin Figure 2.1. 3 1 Source of data: Shafer, 1964 Durrell, 1953 static water surface (feet) max 82 min ions (ppm) (moles) (ppm) (moles) Si02 13 0.0002 Fe Ca 328 0.0164 17190 0.8578 Mg 718 0.0591 598 0.0492 Na 3145 0.1368 46070 2.0039 K 1479 0.0378 Mn Li Sr 393 0.0101 U total moles ofcations 0.2127 2.9588 C03 0 45 0.0007 HCO3 1064 0.0222 1048 0.0218 SO4 Cl 6709 0.1892 104600 2.9504 NO3 P B F total moles of anions 0.2121 2.9722 Na/Cl ratio 0.7 0.7 IDS 12176 Hardness 3762 as CaCC>3 pH 7.15 Electric Cond. date ofanalyses 1/13/61 1953 Bristol Dry Lake water well data cont. (215 ft.) well: drainage canal Br-2-1 number in Figure 2.1. 2 14 Source of data; Durrell, 1953 Calzia, 1979 static water surface (feet) max min ions (ppm) (moles) (ppm) (moles) Si02 5 0.00008 Fe 2.1 0.00008 Ca 43300 2.1607 24000 1.1976 Mg 1074 0.0884 3500 0.2879 Na 57370 2.4955 84000 3.6538 K 3303 0.0845 2400 0.0614 Mn 110 0.0040 Li no 0.0159 Sr 962 0.0246 860 0.0220 U total moles of cations 4.8536 5.2428 co3 20 0.0007 hco3 24 0.0004 so4 210 0.0044 19 0.0004 Cl 172900 4.8769 220000 6.2054 no3 0.64 0.00001 p 1.9 0.0002 0.09 7x10-7 B 3.2 0.0003 F 1.4 0.00007 totalmolesofanions 4.8813 6.2074 Na/Cl ratio 0.5 0.6 TDS 185000 Hardness CaC03 75000 pH 5.7 Electric Cond. as dateof analyses 1953 5/9/78 Bristol Dry Lake water well data cont. (495 ft.) well: Br-2-2 14 no. in Figure 2.1: Source of data: Calzia, 1979 static water surface (feet) max min kms (ppm) (moles) Si02 3.4 0.00006 Fe 0.33 0.00001 Ca 52000 2.5948 Mg 2300 0.1892 Na 65000 2.8273 K 2300 0.0588 Mn 120 0.0044 Li 96 0.0138 Sr 83 0.0019 U 0.001 totalmolesof cations5.6904 26 0.0009 HCO3 CO3 32 0.0005 S03.1 0.00006 4 Cl 210000 5.9233 N03 0.91 0.00001 P 2.1 0.0002 I 1.3 0.00001 B 3.2 0.0003 F 1 0.00005 total moles of anions 5.9254 Na/Cl ratio 0.5 TDS 357000 Hardness CaC03 140000 pH 5.9 as Electric Cond. dateof analyses (93 ft.) Br-1-1 13 Calzia, 1979 (ppm) (moles) 28 0.0005 3.7 0.0001 15000 0.7485 1200 0.0987 86000 3.7408 3800 0.0972 6.8 0.0002 87 0.0125 650 0.0148 1x10­ •8 4.7139 18 0.0006 22 0.0004 24 0.0005 210000 5.9233 2.7 0.00004 1 0.00001 0.06 5x10-7 8.5 0.0008 1.8 0.00009 5.9258 0.6 236000 43000 6.1 5/9/78 (500 ft.) Br-1-2 13 Calzia, 1979 (ppm) (moles) 14 0.0002 2.9 0.0001 1300 0.0649 920 0.0757 78000 3.3928 3000 0.0767 7.8 0.0003 71 0.0102 500 0.0114 0.0009 1x10-8 3.6326 35 0.0012 43 0.0007 170 0.0035 180000 5.0771 2.3 0.00004 0.85 0.00008 0.08 6x10-7 7.1 0.0007 1 0.00005 5.0834 0.7 285000 7600 6.5 5/2/78 5/2/78 data Appendix V: Tephrachronology 241 Thedatafor thecorrelationoftephralayersinBristol DryLakewereprovided by Andrd M. Sama-Wojcicki of the U.S.G.S. in Menlo Park. Most of the information concerning the methods used to prepare samples and correlate the tephra layers can be foundin Sama-Wojcicki (1976). Atotalofelevensamplesoftephrawere separatedfromthecoreandsentfor analysis. Oftheseelevensamplesnineweresufficientlyunalteredforcorrelation purposes. Between 15 to 20 pieces of volcanic glass shard from each sample were analyzed for Na, Mg, Al, Si, K, Ca, Ti, Mn, and Fe by electron microprobe. Shards withunsualcomposition weredeleted,andtheaveragesoftheremainingshardswere calculated. TheseaverageswerethencomparedbycomputerwiththeU.S.G.S.data baseofanalyzed samplesfromthewesternUnitedStatesandPacificborderland(about 2050 analyses) for a best match (Sama-Wojcicki, 1976). The sample depths, core, and correlatedages are presentedinTableAV.I. TableAV.I. Cores,sampledepths,andagesofcorrelatedtephraunitsin Bristol Dry Lake. CORECAES #2 DEPTH(M) AGE(M.Y.) 285 2.05 424 2.125 450 2.2 2.5 - 501 3.4 513 3.65 CORE CAES #1 649 2.125 809 3.3 829 3.4 857 3.65 REFERENCES AHARON, P., KOLODNY, Y., and SASS, E., 1977, Recent hot brine dolomitization inthe"SolarLake",GulfofElat:Isotopic,chemical,andmineralogical study: JournalofGeology, v. 85, 27-48. p. ARAKEL,A.V.,1980,GenesisanddiagenesisofHoloceneevaporitic sedimentsin HuttandLeemanlagoons. WesternAustralia;Journal ofSedimentary Petrology, 50, 1305-1326. BACHMAN, G.0., & MACHETTE, M.N., 1977, Calcic soils and calcretes in the southwestern United Sates. U.S. Geological Survey Open-File Report 77-794 163 p. BADIOZAMANI, K., 1973, The dorag dolomitization model-application to the MiddleOrdovicianofWisconsin.JournalofSedimentaryPetrology, 43,965­ 984. BAKER, P. & KASTNER, M., 1981, Constraints on the formation of sedimentary dolomite. Science, 213, 214-216. BASSETT, A. M., & KUPFER, D. H., 1964, A geologic reconnaissance in the southeasternMojaveDesert; CalifDivisionofMinesandGeology,Spec. Report 83. 43 pp. BASSETT, A. M„KUPFER, D. H., BARSTOW, F. C., 1959, Core logs from Bristol, Cadiz, and Danby Dry Lakes, San Bernardino County, California: U. S. Geological Survey Bulletin 1045-D, 97-138. 243 O-value BARRERA E. & SAVIN S. M., 1987, Effect of sample preparation on the 6 ls of fine-grained calcite. Chemical Geology, 66, 301-305. BERG, W.W., SPERRY, P.D., RAHN, K.A., & GLADNEY, E.S., 1983, Atmospheric bromine in the Arctic. Journ. Geophysical Research, 88,6719­ 6736. BERGGREN,W.A.,&VANCOUVERING,J.A., 1974,TheLateNeogene: Biostratigraphy, GeochronologyandPaleoclimatologyofthelast15million yearsinmarineandcontinentalsequences. DevelopmentsinSedimentology 2, Elsevier Scientific Pub. Co., Amsterdam, 216 pp. BERNER, R.A., 1971, Principles ofchemical sedimentology. New York, McGraw-Hill, 240 p. BIRKELAND, P.W., 1984, Soils and Geomorphology , Oxford University Press, New York, 372 p. BLACKWELDER, E., 1933, Lake Manly, an extinct lake of Death Valley. Geog. Rev., 13, 471 pp. 1954, Pleistocene lakes and drainage in the Mojave Region, southernCalifornia:In R.H. Jahns(ed.)GeologyofsouthernCalifornia. chapter 5: Geomorphology, 35-40. the Alton BLACKWELDER, E. & ELLSWORTH, E.W., 1936, Pleistocene lakes of Basin, California. American Journalof Science, Fifth Series, 31, 453-463. BLATT,H,MIDDLETON,G.,&MURRAY,R.,1980,Originofsedimentary rock. 2nd edition, Prentice-Hall Inc., Englewood Cliff, 782 pp. BLISSENBACH, E., 1954, Geology of alluvial fans in semi arid regions. Geol. Soc. America Bulletin, 65, 175-190. BOWLER, J.M., & MAGEE, J.W., 1988, Lake Frome, stratigraphy, facies analysis andevolutionofamajorplaya. Abstract fromSLEADSconference'BB, 4-6. Australian National University, p. BROWN, W.J. & ROSEN, M.R., 1989, An attempt to destroy the myth of a mid-late PleistoceneconnectionbetweenDeathValleyandtheColoradoRiver: Evidence from Silver-Soda and Bristol Dry Lake Basins, Ca. (Abst.) Cordilleranlßocky Mountaincombinedsectionmeeting, v.21. BULL, W.8., 1972,Recognitionofalluvialfandepositsinthestratigraphicrecord.In J.K.Rigby, andW.K. Hamblin(eds.), RecognitionofAncientSedimentary Environments. Spec. Pub. Soc. Econ. Paleont. Mineralogist, Tulsa, 16, 63-83. BUTLER,G.P.,1969,Modemevaporite depositionandgeochemistryofcoexisting brines, thesabkha,TrucialCoast,ArabianGulf: JournalofSedimentary Petrology, 39, 70-89. 1970,Holocene oftheAbuDhabisabkha,TrucialCoast: an gypsum alternativeexplanationoforigin: 3rd SymposiumonSalt,NorthernOhio Geological Society, 1, 120-152. CAT .71A, J.P., 1979, Geophysical, lithologic, and water quality data from Bristol Dry Lake, San Bernardino County, California: U.S. Geol. Survey, Menlo Park, Ca., Unpublished written communication. CASAS, E. & LOWENSTEIN, T.K., in press, Diagenesis of saline pan halite: Comparison of petrographic features of Modem, Quaternary and Permian halites. Sedimentology. CASTENS-SEIDELL,8., 1984,Theanatomyofamodernmarinesiliciclastic sahkha in a rift valley setting: Northwest Gulfof California tidalflats, Baja California, Mexico: Unpub. Ph.D. dissertation, Johns Hopkins University, 386 p. CHAPMAN,R.H.&REITMAN, J.D.,1978,BouguergravitymapofCalifornia, Needlessheet. CaliforniaDivisionofMinesandGeology, Sacramento, California. CICERONE, R. J., 1981, Halogens in the atmosphere. Reviews of Geophysics and Space Physics, 19, 123-139. 1984,Halogens.mGlobalTroposphericChemistry: aPlanforAction, National Academy Press, Washington, 128-132. CODY, R.D., 1976, Growth and early diagenetic changes in artificial gypsum crystals grownwithinbentonitemudsand gels. GeologicalSocietyofAmerica Bulletin, 87, 1163-1168. 1979, Lenticular occurrencesinnature,andexperimental gypsum; determinationsofeffectsofsolublegreenplantmaterial onitsformation. Journal of Sedimentary Petrology, 49, 1015-1028. CODY,R.D.,&CODY,A.M., 1988a,Gypsumnucleationandcrystalmorphologyin analogsalineterrestrialenvironments.JournalofSedimentaryPetrology, 58, 247-255. CODY,R.D.,&CODY, A.M., 1988b,Tannicacidinducedpenetration twinningin gel-grown calcium sulfate dihydrate. Crystal Research Technology, 23, 1437­1441. CODY, R.D., & SHANKS, H.R., 1974, A comparison of calcium sulphate dihydrate growninclaygelsandinsodiumsilicate gels.JournalofCrystalGrowth,23, 275-281. COHMAP MEMBERS, 1988, Climatic changes of the last 18,000 years: observations and model simulations: Science, 213, 1043-1052. CROWELL, J.C., 1974, Sedimentation along the San Andreas Fault, California, in Dott,Jr.,R.H.,& Shaver,R.H.(eds.),Modernandancient geosynclinal sedimentation Spec. Pub. Soc. Econ. Paleont. Mineralogist, Tulsa, 19, 292­ 303. CURTIS, R. EVANS, G., KINSMAN, D.J.J., & SHEARMAN, D.J., 1963, Associationofdolomiteandanhydrite intherecentsedimentsofthePersian Gulf: Nature, 197, 6779-6800. DARTON,N.H.,1915, GuidebookofWesternUnitedStates;partC,TheSantaFe Route: U.S. Geol. Survey Bulletin 613, p. 152-155. DAVIS, J.R., 1976,Theinfluenceofdrainageareauponthedistributionoflithiumin playa sediments, in Vine, J.D. (ed.), Lithium resources and requirements by the 2000. United States Geological Survey Professional Paper 1005, 105­ year 109. l979,Stratigraphic and tectonic setting of the lithium brine field, Clayton Valley, Nevada,in Newman,G.W.andGoode,H.D.(eds.) 1979Basinand Range Symposium. Rocky Mountain Association of Geologists, 421-430. l9Bl,LateCenozoicGeologyofClaytonValley, Nevadaandthegenesis of lithium-enriched brine. Unpub. Ph.D. Dissertation University of Texas- a Austin, Austin, 233 pp. DEERE, W.A, HOWIE, R.A. & ZUSSMAN, J., 1978, An introduction to the rock forming minerals. Longman Group Limited, London, 11th impression, 528 p. DIBBLEE,T.W.,JR., 1961,Evidenceofstrike-slipmovementonnorthwest-trending faultsinthewesternMojave Desert,California.UnitedStatesGeological Survey Professional Paper 424-B, 8197-8199. DIONNE, J. & CAILLEUX, A., 1972, Faulted calcareous concretions in Pleistocene sediments. Journalof Geology, 80, 744-748. DDC, G.R. & MULLINS, H.T., 1987, Shallow, subsurface growth and burial alterationofMiddleDevoniancalciteconcretions: JournalofSedimentary Petrology, 57, 140-152. DORN, R.1., 1988, A rock varnish interpretation of Alluvial-fan development in Death Valley,California. NationalGeographicResearch,4,56-73. DORN, R.1.,&OBERLANDER, T.M.,1982,Rockvarnish.ProgressinPhysical Geography, 6, 317-367. DUCE, R.A, WINCHESTER, J.W. & VAN NAHL, 1965, lodine, bromine, and chlorineintheHawaiianmarineatmosphere. Journ.GeophysicalResearch, 70, 1775-1799. DUFFY,C.J.,&AL-HASSAN, S., 1988,Groundwatercirculationin acloseddesert basin:topographicscalingandclimaticforcing.WaterResourcesResearch, 24, 1675-1688. DULHUNTY, J.A., 1987, Salina bed instability and geodetic studies at Lake Eyre, SouthAustralia.TransactionsoftheRoyalSocietyofSouthAustralia, 111, 183-188. DURRELL, C, 1953, Celestite deposits at Bristol Dry Lake, near Amboy San Bernardino County, California: in Geological investigation of strontium depositsinSouthernCalifornia: Spec.Rep.CaliforniaStateMiningBureau 32, 37-48. EARDLEY, A.J., & STRINGHAM, 8., 1952, Selenite crystals in the clays of Great Salt Lake. JournalofSedimentaryPetrology, 22,234-238. EDINGER, S.E., 1973, An investigation of the factors which affect the size and growth rates of the habit faces of gypsum. JournalofCrystal Growth, 18, 217-224. ELLIOTT, T., 1986, Siliciclastic shorelines, in Reading, H.G., (ed.), Sedimentary environmentsandfacies, 2ndEdition, Elsevier, NewYork, p. 155-188. EPSTEIN, S., GRAF, D.A. & DEGENS, E.T., 1964, Oxygen isotope studies on the origin of dolomite, in Craig, H., Miller, S.L. & Wasserburg, G.P. (Eds.), Isotopic and cosmic Chemistry, North Holland, Amsterdam, p. 169-180. EUGSTER, H.P., AND HARDEE, L.A., 1975, Sedimentation in an ancient playa­lakecomplex: theWilkinsPeakMemberoftheGreenRiverFormationof Wyoming.GeologicalSocietyofAmericaBulletin, 86,319-334. EUGSTER, H.P., AND HARDIE, L.A., 1978, Saline lakes: in, A. Lerman (ed.). Lakes: Chemistry, Geology, Physics, Springer-Verlag, New York, 237-293. EUGSTER,H.P.,ANDJONES,B.F., 1977,Thebehaviorofpotassiumandsilica duringclosedbasinevaporation: 2ndlAGCSymp.Water-RockInteraction, Strasbourg, 2, 1-12. EVANS, G., SCHMIDT, V. BUSH, P., & NELSON, H., 1969, Stratigraphy and geologic history of the sabkha, Abu Dhabi, Persian Gulf: Sedimentology, v. 12, p. 145-159. FERGUSON, J„ BURNE, R.V., & CHAMBERS, L.A., 1982, Lithification of peritidal carbonates by continental brines at Fisherman Bay, South Australia, to formamegapolygon/splelean limestoneassociation;JournalofSedimentary Petrography, 52, 1127-1147. FETH,J.H., 1967,ChemicalcharacteristicsofbulkprecipitationintheMojaveDesert Region, California. In Geological Survey Research, 1967, United States Geological Survey Professional Paper 575-C, C222-C227. .1981,Chlorideinnaturalcontinentalwaters-areviewUnitedStates Geological Survey Water-supply Paper 2176, 30 pp. FOLK,R.L.,1965,Someaspectsofrecrystallizationinancientlimestones: Societyof Economic Mineralogists andPaleontologists Spec. Pub., 13, 14-48. .1974,Thenaturalhistoryofcrystalline calciumcarbonatecarbonate:effectsof magnesiumcontentandsalinity; JournalofSedimentaryPetrology,44,40-53. FOLKR.L.andLANDL.S., 1975,Mg/Caratioandsalinity:twocontrolsover crystallization of dolomite. American Association ofPetroleum Geologists Bulletin, 59, 60-69. FONTES, J.C., 1965, Fractionnement isotopique dans I’eau de cristallisation du sulfate de calcium. Geol. Rundschau, 55, 172-178. FONTES, J.C. & GONFIANTINI, R., 1967, Fractionnement isotopique de I'hydrogene dans I'eau de cristallisation du gypse. C.R. Acad. Sci. (Paris), 265, 4-6. FRIEDMAN, 1., & O'NEIL, J.R., 1977, Compilation of stable isotope fractionation factorsofgeochemical interest, in Fleischer,M.,ed..DataofGeochemistry, 6th ed.. United States Geological Survey Professional Paper, 440 KK. GALE, G. K., 1951, Geology of the saline deposits, Bristol Dry Lake, San Bernardino County, California: Spec. Rep. California State Mining Bureau 13, 22 pp. GAO, G., HOVORKA, S.D., & POSEY, H.H., in review, Limpid dolomite in Permian San Andres halite rocks, Palo Duro Basin, Texas Panhandle: Characteristicsandpossibleorigin. SubmittedtoJournalofSedimentary Petrology. GARDNER,D.L., 1980,TheBarstow-BristolTroughcentralMojaveDesert, California:In D.L.Fife andA.R.Brown(eds.).Geologyandmineralwealth oftheCalifornia Desert.. SouthCoastGeological Society, SantaAna, 204-214. GARFUNKEL, Z., 1974, Model for Late Cenozoic tectonic history of the Mojave Desert, California, and for its relation to adjacent regions: Geological Society of America Bull., 85, 1931-1944. GAVISH, E., 1980, Recent Sabkhas marginal to the southern coasts of the Sinai, Red Sea, in Nissenbaum, A. (ed.). Hypersaline brines and evaporitic environments, Elsevier, Amsterdam, p. 233-251. GILE, L.H., PETERSON, F.F. & GROSSMAN, R.8., 1966, Morphological and ofcarbonateaccumulationindesertsoils: SoilScience, 101, genetic sequences 347-360. GLANZMAN, R.K. & MEIER, A.L., 1976, Lithium brines associated with nonmarine evaporites. in Vine, J.D. (ed.), Lithium resources andrequirements by the year 2000. United States Geological Survey Professional Paper 1005, 88-92. GLAZNER, A.F.,1988,Stratigraphy,structure,andpotassicalterationofMiocene volcanic rocks in the Sleeping Beauty area, central Mojave Desert, California: Geological Society ofAmerica Bull., 100, 424-435. GLAZNER, A.F., BARTLEY, J.M. & WAUKER, J.D., 1989, Magnitude and significance ofMiocene crustal extension in the central Mojave Desert, California. Geology, 17, 50-53. GREGG, J.M. & HAGNI, R.D., 1987, Irregular cathodoluminescent banding in late dolomite cements: Evidence for complex faceting and metalliferous brines: Geological Society ofAmerica Bulletin, 98, 86-91. HADLEY, D. and KANAMORI, H., 1977, Seismic structures of the Transverse Ranges,California:Geological Society ofAmericaBull., 88, 1469-1478. HALE, G.R.,1984,ReassessmentoftheDeathValley-ColoradoRiveroverflow evidence. Unpub. Ph.D. dissertation, University of hypothesis inlightofnew Califomia-Berkeley, 291 p. _1985,Mid-PleistoceneoverflowofDeathValley towardtheColoradoRiver. In Hale,G.R.(ed.)QuaternaryLakesoftheeasternMojaveDesert,California. Field trip Guidebook, Friends of the Pleistocene, Pacific Cell, 36-81. HANDFORD, C. R., 1982a, Sedimentology andevaporite genesis in aHolocene continental-sabkha playa basin-Bristol Dry Lake, California: Sedimentology, 29, 239-253. 1982b, Terrigenous clastic and evaporite sedimentation in a Recent continental-sabkha playa basin, Bristol Dry Lake, California: in C. R. , Handford, R.G. Loucks, and G.R. Davies (eds.) Depositional andDiagenetic Spectra ofEvaporites A core workshop. Soc. Econ. Paleont. Mineralogists - core workshop No. 3, Calgary Canada, 65-74. 1982c,Permianfaciessequencesandevaporitedepositional styles,Texas panhandle, C. R. Handford, R.G. Loucks, and G.R. Davies (eds.) in , Soc. - Depositional andDiagenetic Spectra ofEvaporites A core workshop. Econ. Paleont. Mineralogists core workshop No. 3, Calgary Canada, 210-237. HANSHAW, 8.8., BACK, W. & DEIKE, R.G., 1971, A geochemical hypothesis for dolomitization by groundwater. Economic Geology, 66,71-724. HARDIE,L.A., 1968,TheoriginoftheRecentnon-marineevaporitedepositof SalineValley,InyoCounty,California. GeochimicaetCosmochimicaActa, 32, 1279-1301. non-marine?AmericanJournalofScience, 284, 1984, Evaporites; marine or 193-240. in press. Potash evaporites, rifting, and the role of hydrothermal brines. JournalofSedimentary Petrology. HARDIE,L.A.,ANDEUGSTER,H.P., 1970,Theevolutionofclosed-basinbrines: Spec. Pap. Miner. Soc. Amer. 3, 273-290. HARDIE, L. A., SMOOT, J. P., AND EUGSTER, H. P., 1978, Saline lakes and their deposits: a sedimentological approach; in, A. Matter and M. E. Tucker (eds.), Modern andAncient Lake Sediments, Int. Assoc. Sediment. Special Pub., 2, 7-41. HAY, R.L., PEXTON, R.E., TEAGUE, T.T. & KYSER, T.K., 1986, Spring-related carbonaterocks, Mgclay, andassociatedmineralsinPliocene depositsofthe Aramgosa Desert,NevadaandCalifornia.Geological Society ofAmerica Bulletin, 97, 1488-1503. the chemical characteristics of natural water. U.S.Geological Survey water supply paper 1473 269 pp. HEM, J.D., 1970, Study andinterpretation of , HIGGINS, C.T., 1980, Geothermal resources of California. California Geological Map Series,MapNo.4,DivisionofMinesandGeology, California DepartmentofConservation., Sacramento,California. HOLSER, W.T., 1966, Bromide geochemistry of salt rocks, in Rau, J.L. (ed.) SecondSymposiumonSaltVolumeI, TheNorthernOhioGeological Society, Inc., Cleveland, 248-275. 1970, Bromide geochemistry of some non-marine salt deposits in the southernGreatBasin.Mineralogical SocietyofAmericaSpecialPaperNumber Three, 307-319. 1979a, Trace elements and isotopes in evaporites. in R. G. Bums (ed.) Marine Minerals Reviews in Mineralogy Vol. 6, Min. Soc. America 295-346. , 1979b, Mineralogy of evaporites. in R. G. Bums (ed.) Marine Minerals Reviews in Mineralogy Vol. 6, Min. Soc. America, 211-294. HOLSER, W.T., WARDLAW, N.C. & WATSON, D.W., 1972, Bromide in salt rocks:extraordinarily lowcontentintheLowerElkPointsalt,Canada, in Richter-Bernburg, G. (ed.) Geology ofSaline Deposits, Proceedings ofthe Hanover Symposium, May 1968. Unesco, Paris, 63-75. HOVORKA, S., 1987, Depositional environments of marine-dominatedbedded halite, Permian San Andres Formation, Texas. Sedimentology, 34, 1029-1054. HUBBS, C.L. & MILLER, R.R., 1948, The zoological evidence: Correlation between fish distribution and hydrographic history in the desert basins of western UnitedStates. InTheGreatBasin,withEmphasis onGlacialandPostglacial Times,University ofUtahBulletin 38. HUBER,N.K.,1981, AmountandtimingofLateCenozoicupliftandtiltonthe central Sierra Nevada, California—Evidence from the upper San Joaquin River Basin. U.S.Geological Survey professional paper 1197, 28 pp. HUDSON, J.D., 1978, Concretions, isotopes, and the diagenetic history of the Oxford Clay (Jurassic) of central England. Sedimentology, 25, 339-370. JENKINS,D.G., 1987,WasthePliocene-Pleistoceneboundaryplacedatthewrong stratigraphic level? Quaternary Science Reviews, 6, 41-42. JONES,B.F., 1966,Geochemicalevolutionofclosedbasinwatersinthewestern Great basin: Ohio Geol. Soc. Symposium on Salt, 2nd, 1, 181-200. JONES, B. F., and VAN DENBURGH, A. S., 1966, Geochemical influences on the chemicalcharacterofclosedlakes: Int.Assoc.Sci.Hydrol.Symposium Garda, 70, 435-446. JONES, B. F., EUGSTER, H. P., AND RETTIG, S. L., 1977, Hydrochemistry of the Lake Magadi Basin, Kenya: Geochim. Cosmochim. Acta, 41, 53-72. KAPLAN, 1.R., 1983, Stable isotopes of sulfur, nitrogen and deuterium in recent marine environments, in. Stable isotopes in sedimentary environments. Soc. Econ. Paleont. Mineralogists short course No. 10., Dallas, 2-1-2-108. KENDALL, A.C., 1988, Aspects of evaporite basin stratigraphy, in Schreiber, 8.C., (ed.) Evaporites andHydrocarbons, ColumbiaUniversity Press, New York, 11-65. KENDALL, G. ST. C, & WARREN, J.K., 1987, A review of the origin and setting of tepees and their associated fabrics. Sedimentology, 34, 1007-1027. KINSMAN, D.J.J, 1965, Gypsum and anhydrite of recent age, Trucial Coast, Persian Gulf: in Rau, J.L. (ed.), 2nd Symposium on Salt, Northern Ohio Geological Society, 1, 302-306. KUSHNIR, J.,1980,EffectoftheCa/SO4ratioonthegrowthrateandcrystalhabitof gypsum. Internal. Assoc. Sedimentologists Ist Europ. Mtg., Bochum (Abstracts), 239-241. l9Bl, Formation and early diagenesis of varved evaporitic sediments in a coastal hypersaline pool: Journal of Sedimentary Petrology, 51, 209-214. LAND,L.S., 1973,Contemporaneous dolomitizationofMiddlePleistocenereefsby meteoric water, North Jamaica. Bulletin of Marine Science, 23, 64-92. LANGBEIN, W. 8., 1961, Salinity and hydrology of closed lakes: U. S. Geol. Survey Prof. Paper 412, 1-20. LEVY,Y., 1977,Descriptionandmodeofformationofthesupratidalevaporite facies innorthernSinaicoastalplain:JournalofSedimentaryPetrology, 47,463­ 474. LOGAN,8.W., 1981,Modemcarbonateandevaporite sedimentsofSharkBayand Lake Macleod, Western Australia: Geological Society of Australia Field Excursion Guidebook, sth Australia Convention, 64 p. 1987, The MacleodEvaporite Basin, Western Australia; American _, Association of Petroleum Geologists Memoir 44, 140 p. LONG, J.V.P. & AGRELL, 5.0., 1965, Thecathodoluminescenceof minerals in thin section. Mineral. Magazine, 34, 318-326. LOWENSTEIN, T. K., 1982, Primary features in a potash evaporite deposit, the Permian Salado formation of west Texas and New Mexico, in C. R. , Handford, R.G. Loucks, and G.R. Davies (eds.) Depositional andDiagenetic - Spectra ofEvaporites A core workshop. Soc. Econ. Paleont. Mineralogists core workshop No. 3, Calgary Canada, 276-304. l9BB, Origin of cycles in aPermian "saline giant": the Salado (McNutt zone) evaporites of New Mexico and Texas. Geol Soc. America Bulletin, 100, 592-608. LOWENSTEIN, T.K.,&HARDIE,L.A.,1985,Criteriafortherecognitionofsalt­pan evaporites. Sedimentology, 32, 627-644. LUCIA, F.J., 1968, Recent sediments and diagenesis of South Bonaire, Netherlands Antilles; Journal ofSedimentary Petrology, 38, 845-858. MACHETTE,M.N.,1985,CalcicsoilsofthesouthwesternUnitedStates, inD.L. Wielde(ed.)SoilsandQuaternary GeologyoftheSouthwesternUnitedStates. Geological Society ofAmerica Special Paper 203, 1-21. MAGEE,J.W., 1988,ChemicalandclasticsedimentsandLateQuaternaryhistory, Prungle Lakes, NewSouth Wales. Unpub. Masters thesis Australian National University, Canberra, Australia. MAGEE, J.W., BOWLER, J.M., LOCK, D.E., & WILLIAMS, D.L.G., 1988, Lake Eyre: stratigraphy, sedimentology and Late Quaternary history: Abstract fromSLEADSconference'BB,AustralianNationalUniversity, p.61-64. MANN, S.D., 1988, Subaqueous evaporites of the Buckner Member, Haynesville Formation,NortheasternMobileCounty,Alabama. Transactions—GulfCoast Association of Geological Societies, XXXVIII, 187-196. MAYER, L., McFADDEN, L.D. & HARDEN, J.W., 1988, Distribution of calcium carbonateindesertsoils: amodel. Geology. 16,303-306. McBRIDE,E.F.,LAND,L.S.&MACK,L.E., 1987,Diagenesis ofeolianandfluvial feldspathic sandstones, Norphlet Formation, (Upper Jurassic), Rankin County, Mississippi, andMobileCounty,Alabama.AmericanAssociationofPetroleum Geologists Bulletin, 71, 1019-1034. McGOWEN, J.H., & GROAT, C.G, 1971, Van Horn Sandstone, West Texas: an alluvialfanmodelformineralexploration. TexasBureauofEcon.Geology, Report of investigations 72, 57 pp. MIFFLIN,M.D., 1968,Delineationofground-waterflowsystems inNevada: Tech. Report SeriesH-W,Univ.Nevada, DesertResearchInstitute, 4,54p. MOHORICH,L.M.,1980,GeothermalresourcesoftheCaliforniaDesert asummary - review:In D.L.FifeandA.R.Brown(eds.),Geologyandmineralwealthof theCaliforniaDesert.. SouthCoastGeological Society,SantaAna,171-189. MONNIN, C. & GALINIER, C, 1988, The solubility of celestite and barite in electrolyte solutions and natural waters at 25 °C: a thermodynamic study. Chemical Geology, 71, 284-296. MOYER, J.L. & DUCE, R.A., 1972, Gaseous and particulate bromine ion the marine atmosphere. Journ. Geophysical Research, 77, 5330-5338. NAIMAN, E.R., BEIN, A. & FOLK, R.L., 1983, Complex polyhedral crystals of limpid dolomite associated with halite, Permian Upper Clear Fork and Glorieta Formations, Texas. JournalofSedimentary Petrology, 53, 549-555. NEAL,J.T., 1975,Introduction:in:J.T.Neal(ed.),PlayasandDriedLakes, BenchmarkPapersinGeology,Dowden, Hutchinson,andRoss, Inc..Stroudsburg, 1-5. PANTIN,H.M., 1958,Rateofformationofadiageneticcalcareousconcretion. Journalof SedimentaryPetrology, 28,366-371. PARKER,R.8., 1963,RecentvolcanismatAmboyCrater,SanBernardinoCounty, California. SpecialReport76,CaliforniaDiv.MinesandGeology.21p. PERTHUISOT, J.P.,FLORIDA,S„&JAUZEIN,A.,1972,Unmodulerecentde bassincoderasedimentationsaline:laSebkhaelMelah(Zarzis, Tunisie); Revue de Geologic Dynamique et de Geographic Physique14, 67-83. , PHILLIPS, F.M., SMITH, G.1., BENTLEY, H.W., ELMORE, D. &GROVE, H.E., 1983, Chlorine-36 dating of saline sediments: preliminary results from Searles Lake, California. Science, 222, 925-927. PIERRE, C, 1988, Applications of stable isotope geochemistry to the study of evaporites. In Schreiber, B.C. (ed.), Evaporites and hydrocarbons, Columbia University Press, New York, 300-344. POTTER, R.M., & ROSSMAN, G.R., 1977, Desert varnish: the importance of clay minerals. Science, 196, 1446-1448. PURSER, 8.H., 1973, Sedimentation around bathymetric highs in the southern PersianGulf:in Purser,B.H.(ed.),ThePersianGulf:Holocenecarbonate a sedimentation and diagenesis in shallow epicontinental sea, Springer-Verlag, New York, p. 157-177. RAISWELL,R., 1971,ThegrowthofCambrianandLiassicconcretions. Sedimentology, 17, 147-171. RANCHER, J. & KRITZ, M.A., 1980, Diurnal fluctuations of Br and I in the tropical marine atmosphere. Journ. Geophysical Research, 85, 5581-5587. RICHARDSON,W.A., 1921,Therelativeageofconcretions.Geo/.Mag.,63,114­ 124. ROSEN, M.R., in review, Mineralogic and stable isotopic constraints on the hydrologicevolutionofBristolDry Lake,California, U.S.A.: Palaeogeography, Palaeoclimatology, Palaeoecology. SARNA-WOJCICKI,A.M., 1976,CorrelationofLateCenozoictuffsintheCentral CoastRanges ofCalifornia by meansof trace-andminor-elementchemistry. UnitedStatesGeologicalSurveyProfessionalPaper972, 30pp. SASS, E. & KOLODNY, Y., 1972, Stable isotopes, chemistry, and petrology of carbonateconcretions(MishashFormation,Israel). ChemicalGeology, 10, 261-286. SCHMID, R.M., 1988, Lake Torrens halite accumulation (South Australia). Z. dt. geol. Ges., 139, 289-296. SCHREIBER,8.C.,1988,Evaporitesandhydrocarbons. ColumbiaUniversity Press, New York, 475 p. SCHREIBER, 8.C., & KINSMAN, D.J.J., 1975, New observations on the Pleistocene evaporitesofMontallegro, Sicily andamodemanalog:Journalof Sedimentary Petrology, 45, 469-479. SHAFER, R. A., 1964, Report on investigations of conditions which determine the potentials for development in the desert valleys ofeastern San Bernardino County, California. Unpublished Engineering Department Report, Southern California Edison Co., 172 pp. SHEARMAN, D.J., 1978, Evaporites of coastal sabkhas: in Dean, W.E. and Schreiber, B.C. (eds.). Marine Evaporites, Society ofEconomic Paleontologists andMineralogists ShortCourseNo. 4, p.6-42. 1985,Syndepositionalandlate diageneticalterationofprimary gypsum to anhydrite in Schreiber, B.C. (ed.), Sixth international symposium on salt, vol. 1, Salt Institute, Alexandria,Va., U.S.A., p. 41-50. SMITH,C.L., 1976,Useoflithiumandchlorideconcentrationsingroundwaterfor lithium exploration, in Vine, J.D. (ed.), Lithium resources andrequirements by the year 2000. United States Geological Survey Professional Paper 1005, 103-105. SMITH, D.8., 1971, Possible displacive halite in the Permian Upper Evaporite Group of Northeast Yorkshire. Sedimentology, 17, 221-232. SMITH, G.1.,1976,OriginoflithiumandothercomponentsintheSearlesLake evaporites, California, in Vine, J.D. (ed.), Lithium resources andrequirements 2000. United States Geological Survey Professional Paper 1005, by the year 92-103. .1979, Subsurface stratigraphy and geochemistry of Late Quaternary evaporites, Searles Lake, California. U.S.Geological Survey professional paper 1043, 130 pp. .1984, Paleohydrologic regimes in the Southwestern Great Basin, 0-3.2 my ago, Compared with other long records of "global" climate. Quaternary Research, 22, 1-17. SMITH, G.1., & STREET-PERROTT, F.A., 1983, Pluvial lakes of the Western UnitedStates: inPorter, S.C.(ed.) Late-QuatemaryenvironmentsoftheUnited States, VolumeI, theLatePleistocene. University ofMinnesotaPress, Minneapolis, 190-212. SMITH, G.1., FRIEDMAN, I. & McLAUGHLIN, R.J., 1987, Studies of Quaternary salinelakes-Hl.Mineral,chemical, andisotopicevidenceofsaltsolutionand crystallization processesinOwensLake,California, 1969-1971.Geochimicaet Cosmochimica Acta, 51,811-828. SMITH,G.1.,BARCZAK,V.J.,MOULTON,G.F.&LIDDICOAT,J.C., 1983, Core KM-3, asurface-to-bedrockrecordofLateCenozoicSedimentationin Searles Valley, California. U.S.Geological Survey professional paper 1256, 24 pp. SMOOT, J. P., 1983, Depositional subenvironments in arid closed basin; Wilkins an PeakMemberoftheGreenRiver Formation(Eocene),Wyoming, USA: Sedimentology, 30, 801-827. SOFER, Z. & GAT, J.R., 1975, The isotope composition of evaporating brines: effect oftheisotopic activity ratioin salinesolutions.EarthandPlanetary Science Letters, 26, 179-186. SORBY, H.C., 1908, On the application of quantitative methods to the study of the structure and history of rocks. Quarterly Jour. Geol. Soc. London, 64, 171­ 233. STANLEY, D.J., SWIFT, D.J. & RICHARDS, H.G., 1967, Fossiliferous concretions of Georges Bank: JournalofSedimentary Petrology, 37, 1070­1083. STARKY, H.C., BLACKMAN, P.D., and HAUFF, P.L., 1984, The routine mineralogical analysis of clay bearing samples. UnitedStates Geological Survey, Bulletin 1563, 32 p. TELFORD, W.M., GELDART, L.P., SHERIFF, R.E. & KEYS, D.A., 1976, Applied Geophysics, Cambridge University Press, Cambridge, 860 p. TEN HAVE, T. & HEUNEN, W., 1985, Cathodoluminescence activation and zonation in carbonate rocks: an experimental approach. Geologic en Mijnbouw, 64, 297-310. THOMPSON,D.G.,1929,TheMojaveDesertregion,California: U.S.Geol. Survey Water-supply Paper 578, 759 pp. TODD, J. E., 1896, Log-like concretions and fossil shores. American Geologist, 17, 347-349. .1903, Concretions and their geological effects. Geological Society ofAmerica Bulletin, 14, 353-368. TUCKER,R.M., 1981,GiantpolygonsintheTriassicsaltofCheshire,England:a thermalcontractionmodelfortheirorigin. JournalofSedimentaryPetrology, 51, 779-786. VAN DEVENDER, T.R., & SPAULDING, W.G., 1979, Development of vegetation and climate in the Southwestern United States: Science, 204, 701-710. VERPLANCK,W.E., 1952,GypsuminCalifornia:CaliforniaDivisionofMines Bulletin 163, p.47. 1958,SaltinCalifornia;CaliforniaDivisionofMinesBulletin 175,168 P­ search. JournalofResearchJJnited States Geological Survey, 3, 479-485. VINE,J.D.,1975,Lithiuminsedimentsandbrines-how,why, andwhereto 1976, Nonpegmatite lithium resource potential, in Vine, J.D. (ed.). Lithium resourcesandrequirementsbytheyear2000.UnitedStatesGeological Survey Professional Paper 1005, 54-58. ~ VOGT, J.N., 1986,DolomitizationandanhydritediagenesisoftheSanAndres (Permian) Formation, Gaines County, Texas. Unpub. M. A. Thesis, University of Texas-Austin, Austin, 203 pp. WARREN, J.K., 1980, Late Quaternary evaporite-carbonate depositional environmentsinSouthAustralia-Theirpaleohydrological implications. Unpubl. Ph.D. dissertation, Flinders University of South Australia, Adelaide, S.A., 242 pp. WARREN, J. K., 1982a, The hydrologic setting, occurrence and significance of in late Quaternary salt lakes in South Australia: Sedimentology, 29, gypsum 609-637. __l9B2b, Thehydrological significance of Holocenetepees, stromatolites and boxworklimestonesincoastalsalinasinSouthAustralia:Journal of 1171-1201. Sedimentary Petrology, 52, 1989, Evaporite sedimentology: importance in hydrocarbon accumulation. Prentice Hall, Englewood Cliffs, 285 p. Seamarginalevaporites. AmericanAssociationofPetroleum ; in press. Geologists Memoir. formed in marine sabkha (subaerial) and salina (subaqueous) settings—modem and ancient: Amer. Assoc. Petroleum Geologists Bulletin, 69, 1013-1023. WARREN, J. K., & KENDALL, G. ST. C.,1985, Comparison of sequences WASSON, R.J., SMITH, G.1., & AGRAWAL, D.P., 1984, Late Quaternary sediments, minerals,andinferredgeochemicalhistoryofDidwanaLake,Thar Desert, India. Palaeogeo., Palaeoclim., Palaeoecol, 46, 345-372. WATSON, A.W., 1985, Structure, chemistry and origins of gypsum crusts in southernTunisiaandthecentralNamibDesert. Sedimentology,32,855-875. WEEKS,L.G., 1953,Environmentandmodeoforiginandfaciesrelationshipsof carbonateconcretionsinshales. JournalofSedimentaryPetrology, 23, 162­ 173. _1957,Originofcarbonateconcretionsinshales, Magdalena Valley,Columbia. Geological SocietyofAmericaBulletin, 68,95-102. WELLS, S.G., & McFADDEN, L.D., 1987, Comment on "Isotopic evidence for Climatic influence on alluvial-fan development in Death Valley California". Geology, 15, 1178. WEST, 1.M., ALI, Y.A., & HILMY, M.E., 1979, Primary gypsum nodules in a modem sabkha on the Mediterraneanof coast of Egypt: Geology, 7, 354­ 358. WHITE, D.E., HEM, J.D. & WARING, G.A., 1963, Chemical composition of surface waters. U.S.Geological Survey professional paper 440-F, 67 pp. WHITE, D.E., THOMPSON, J.M. & FOURNIER, 1976, Lithium contents of thermal mineral waters, in Vine, J.D. (ed.), Lithium resources and requirements by the year 2000. United States Geological Survey Professional Paper 1005, 105 109. WINOGRAD, 1.J., SZABO, B„ COPLEN, T.B. & DOTY, G.C., 1983, Pliocene andPleistocenecalcitic veinsasindicatorsofpaleohydrologic, paleoclimatologic, and neotectonic events, southern Great Basin: an intitial appraisal, in Paleoclimate and Mineral Deposits, United States Geological Survey Circular 822, 8-9. WOOD, M.W., & SHAW, H.F., 1976, The geochemistry of celestites from the Yate area near Bristol (U.K.). Chemical Geology, 17, 179-193. YURTSEVER, Y., 1975, Worldwide survey of stable isotopes in precipitation. Kept. Sect. Isotope Hydrology, lAEA, November, 1975, 40 p. The vita has been removed from the digitized version of this document.