https://doi.org/10.1130/G46419.1

Manuscript received 19 April 2019 Revised manuscript received 26 July 2019 Manuscript accepted 14 August 2019

© 2019 Geological Society of America. For permission to copy, contact [email protected]. Published online 6 September 2019

The Aspen paleoriver: Linking Eocene magmatism to the world’s largest Na-carbonate evaporite (Wyoming, USA) Alexander P. Hammond1, Alan R. Carroll1, Ethan C. Parrish1, M. Elliot Smith2 and Tim K. Lowenstein3 1Department of Geoscience, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA 2School of Earth Sciences and Environmental Sustainability, Northern Arizona University, Flagstaff, Arizona 86011, USA 3Department of Geological Sciences and Environmental Studies, Binghamton University, Binghamton, New York 13902, USA

ABSTRACT (Smith et al., 2008b; Chetel et al., 2011). A po- Deposition of trona, nahcolite, and other Na-carbonate evaporite minerals in lakes is com- tential resolution of this paradox would be that monly closely associated with active volcanism, suggesting that the excess alkalinity required the Green River Formation lakes were indirectly for their formation may arise from fuid-rock interactions involving hydrothermal waters infuenced by river transport of alkalinity from a

that contain magmatic CO2. Paradoxically, the world’s largest Na-carbonate occurrence, distant magmatic center. Trona deposits at Sear- contained within the Eocene Green River Formation in Wyoming, USA, was not associated les Lake in California (United States) offer a po- with nearby active magmatism. Magmatism was active ∼200 km southeast in the Colorado tentially useful Quaternary analogue. Fluid-rock Mineral Belt, however, suggesting that a river draining this area could have supplied excess interactions occur during hydrothermal circula- alkalinity to Eocene lakes. Sedimentologic studies in southwestern Wyoming, along the course tion through igneous and metamorphic rocks in of the hypothesized Aspen paleoriver, document fuvial and deltaic sandstone with generally the western Long Valley caldera, ∼300 km to the northwest-directed paleocurrent indicators. Sandstone framework grain compositions and north. These waters emerge as thermal springs + – detrital zircon ages are consistent with derivation from the Colorado Mineral Belt and its enriched in Na and HCO3 (Sorey, 1985; Farrar host rocks. These results provide the frst confrmation of a fuvial connection to downstream et al., 1987), which were transported to Searles Eocene lakes, and indicate that lake deposits may offer a unique perspective on upstream Lake via surface drainage (Earman et al., 2005; magmatic and hydrothermal histories. Lowenstein et al., 2016). The Absaroka and Challis volcanic provinces INTRODUCTION subject to debate. Waters with elevated Na+ and in Wyoming and Idaho, respectively, lie to the – Large endorheic lakes commonly serve as HCO3 can originate from weathering of volca- north of the Green River Formation, but neither terminal sinks for regional- to subcontinen- nic rocks or volcaniclastic sediments via silicate appears to have drained into Eocene lakes dur- tal-scale rivers. The Great Salt Lake in Utah hydrolysis (Jones, 1966), but many evaporative ing evaporite deposition (Chetel et al., 2011). (United States), for example, receives its larg- lakes associated with such rocks are not alkaline Lowenstein et al. (2017) instead hypothesized est surface infow from the Bear River, which and do not precipitate Na-carbonate minerals that the Colorado Mineral Belt (CMB), a large

has headwaters ∼560 km upstream in the Uinta (Earman et al., 2005). CO2 produced through magmatic and hydrothermal province ∼200 km Mountains. The Aral Sea in central Asia receives microbial or thermogenic degradation of sedi- southeast of Eocene Lake Gosiute (Wyoming; drainage from the Amu Darya and Syr Darya mentary organic matter could also increase al- Fig. 1), could have supplied excess alkalinity. rivers, both of which fow >2000 km from their kalinity, but such mechanisms have not created Smith et al. (2014) previously proposed such a headwaters in the Pamir Mountains and Tian hyperalkalinity in modern lakes (Lowenstein paleoriver and named it after the town of As- Shan, respectively. Lake deposits therefore offer et al., 2017). pen, Colorado, but its existence has remained unique, high-resolution stratigraphic archives Earman et al. (2005) argued that subsurface conjectural until now. Our study presents new of past tectonic, magmatic, climatic, and biotic fuid-rock interaction caused by injection of sedimentological and provenance data that test

processes, integrated across larger watersheds. magmatic CO2 into groundwater is critical for the hypothesized fuvial connection between the Lacustrine evaporite facies can be especially generating the alkalinity required to precipitate CMB and Lake Gosiute. instructive because they convolve information Na-carbonate evaporite. However, the largest about drainage networks, climate, and infuent known deposit superfcially appears to contradict GEOLOGIC SETTING water chemistry. Sodium-carbonate evaporite this hypothesis. Trona (a non-marine evaporite, The Green River Formation was deposited

occurrences are comparatively rare (Warren, Na2CO3·NaHCO3·2H2O) in the Eocene Green by freshwater to hypersaline lakes that occu- 2010), in part because they evolve from un- River Formation in Wyoming (United States) pied the segmented Laramide foreland east of – 2– usual parent waters in which [HCO3 + OC 3 ] accounts for ∼90% of global commercial soda the Sevier fold-and-thrust belt, in what is now initially exceeds [Ca2+ + Mg2+] (Hardie and Eug- ash reserves (Bolen, 2018), but is not associ- Wyoming, Colorado, and Utah (Dickinson et al., ster, 1970). The origin of this excess alkalinity is ated with evidence for nearby, coeval volcanism 1988; Smith et al., 2008b; Fig. 1). Recent studies­

CITATION: Hammond, A.P., et al., 2019, The Aspen paleoriver: Linking Eocene magmatism to the world’s largest Na-carbonate evaporate (Wyoming, USA): Geology, v. 47, p. 1020–1024, https://doi.org/10.1130/G46419.1

1020 www.gsapubs.org | Volume 47 | Number 11 | GEOLOGY | Geological Society of America

Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/47/11/1020/4852043/1020.pdf by University of Wisconsin Madison user on 23 October 2019 111° W 110°W 109°W 108°W 107°W 106°W 105°W the northeastern segment of the CMB (Klein 43°N et al., 2010; Fig. 1). Eocene Wind River Range Wind River BIghorn evaporite Basin Basin PALEORIVER DEPOSITS IN THE Eocene Granite Range Laramie Range lake WASHAKIE BASIN Eocene Cathedral Bluffs Tongue sedimentary facies river in the southern Washakie Basin (Fig. 2) may 42°N be broadly subdivided into three associations. ID Lake UT (1) Carbonate-rich, kerogenous mudstone fa- Gosiute Great Divide Rock Basin Archean cies containing ostracods occur near the base Bridger Springs Basin FC of the Cathedral Bluffs Tongue, and are inter- arch Figure 2

Fossil Basin Washakie Sierra preted to record profundal to sublittoral lacus- Madre Proterozoic Basin Medicine thrust belt Bow Range trine deposition during expansions of Eocene Sevier fold-and- WY 41°N Lake Gosiute. (2) In contrast, silty red and green CO Sand Wash Park Range mudstone facies are carbonate poor and contain Uinta

Basin Front Range North & paleosols, desiccation cracks, and carbonaceous Middle debris. They are interpreted as foodplain de- Park Basins posits. (3) Sandstone facies encompass amal- gamated beds 1–17 m thick, which are laterally Lake 40°N discontinuous over scales of hundreds of meters. Uinta Gore range Grain size ranges from fne to very coarse sand, Uinta White River Basin Denver with occasional layers of pebbles up to 1 cm Piceance uplift Creek in diameter. Bed bases in some cases contain Basin Aspen Leadville mudstone intraclasts, which are associated with UT CO channel scour. Sedimentary structures include RangeSouth Paleogene basin Park 39°N Sawatch planar-parallel lamination, climbing ripples, and Basin trough crossbeds with amplitudes up to 15 cm. Cretaceous Uncom- pahgre Tabular crossbeds reach amplitudes up to 2 m, Cambrian–Jurassic uplift and shingled sigmoidal bedforms reach 1–3 m. Precambrian-cored California paleoriiver The sandstone facies association is interpreted uplift Colorado 100 km Mineral Belt to represent fuvial to deltaic deposition. The sigmoidal bedforms are interpreted as later- Figure 1. Schematic paleogeography of the western United States at ca. 51 Ma (modifed from ally accreting bars, and 2-m-amplitude tabular reconstructions of Smith et al., 2008b, 2014). Dashed line denotes approximate boundary between the Archean Wyoming province and Proterozoic terranes to the south. Red circles crossbeds at one site (Sand Creek delta) are in- are areas with reported radioisotopic age determinations falling within analytical uncertainty terpreted as downstream-accreting delta fore- of 51 Ma (Klein et al., 2010). FC—Firehole Canyon locality; ID—Idaho; UT—Utah; WY—Wyo- sets. Measured paleocurrent directions based ming; CO—Colorado. on trough and tabular crossbeds are bimodal in the Wasatch Formation main body (Red Creek have demonstrated that the lakes received drain- thins, becomes fner-grained, and eventually site), with modes indicating southeast and north- age from regional paleorivers as long as 700 km disappears going from south to north. west transport. The signifcance of this bimo- (Davis et al., 2010; Chetel et al., 2011; Dick- The CMB is an ∼500-km-long southwest- dality is unclear, but it may refect deposition inson et al., 2012). Intercalated volcanic tuffs northeast–trending mosaic of stocks, laccoliths, by higher-sinuosity streams. In contrast, vector establish a detailed chronology of evaporite sills, dikes, and hydrothermal mineral deposits mean paleocurrent directions are consistently deposition (Smith et al., 2008b). The giant Na- emplaced from ca. 75 to 0 Ma (Fig. 1; Book- to the west or northwest in the Cathedral Bluffs carbonate evaporite deposits in Wyoming occur strom, 1990; Klein et al., 2010). Laramide plu- Tongue (Fig. 2). principally within the Wilkins Peak Member in tonic rocks range in composition from alkaline Sandstone consists of subangular, poorly the Bridger Basin, which was deposited by Lake monzonites and quartz monzonite in the north- sorted arkose to lithic arkose, dominated by Gosiute between ca. 51.6 and 49.8 Ma. east, to calc-alkaline granodiorite in the central­ monocrystalline quartz and potassium feld- Wilkins Peak Member facies initially ex- CMB. The CMB crosscuts older southeast- spar, which average 50% and 29% of framework tended across the Bridger Basin and into parts northwest–oriented structural trends, and its grains, respectively. Lithic fragments average of the Great Divide, Washakie, and Sand Wash origin is enigmatic. Chapin (2012) attributed 13% of the framework grains, of which 65% Basins. Alluvial facies of the Cathedral Bluffs it to dilation of a segment boundary within the are sedimentary, 21% are volcanic, and 14% Tongue of the Wasatch Formation subsequent- underlying, subhorizontally subducting Faral- metamorphic. Detrital zircon U-Pb age distribu- ly flled the eastern basins (Smith et al., 2014; lon plate. Igneous ages generally young to the tions are broadly similar in all six samples in this Fig. 1). Arkosic alluvial sediment was also de- southwest, but few radioisotopic data have been study. Relatively few Archean grains are present posited episodically across the Bridger Basin reported for hydrothermal minerals themselves. in any of the samples; they compose 3.5% of the during major lake lowstands that recurred on The detailed timing of hydrothermal activity is Wasatch Formation main body sample and <2% an ∼100 k.y. eccentricity time scale (Pietras therefore based largely on cross-cutting geolog- of all other samples. All samples exhibit age and Carroll, 2006; Aswasereelert et al., 2013). ic relationships, and in some cases on fssion- peaks at 1760–1770 Ma, 1670–1680 Ma, and ca. The resultant deposits are informally designated track annealing ages of apatite and zircon in 1430 Ma, although their relative magnitude var- “marker beds” A–I. They typically range from host rocks (e.g., Bryant et al., 1990). Ages that ies. Grains with ages of 900–1200 Ma are also medium-grained sandstone to siliciclastic mud- potentially overlap that of the deposition of the present in all samples, but are minor in the Ca- stone. Each of these marker beds progressively Wilkins Peak Member appear to be limited to thedral Bluffs Tongue. All samples also ­contain

Geological Society of America | GEOLOGY | Volume 47 | Number 11 | www.gsapubs.org 1021

Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/47/11/1020/4852043/1020.pdf by University of Wisconsin Madison user on 23 October 2019 108°00’W 107°57’W 107°54’W 107°51’W 2012). Some or all of the Archean grains likely share the same origin. Detrital zircon grains with Quaternary 55–66 Ma ages are consistent with igneous ages in the northeast segment of the CMB (Book- Miocene strom, 1990; Klein et al., 2010). Magmatic ages of 60–66 Ma are widely reported from the CMB Laney Member Hangout Wash between the cities of Denver and Aspen. CMB ages of 55–60 Ma are less common, but have N = 58 Hangout 41°6’N been reported between Denver and Leadville, of Wasatch Formation Sand Creek delta Ridge Colorado. The variation in Paleogene detrital Tipton Shale Member of N= 94 N = 98 zircon age distributions reported here may re- Green River Formation fect changes in the Aspen paleoriver headwaters McPMcPhersonherson Springs through time, or alternatively could refect grain Wasatch Formation N = 98 Hawk Ridge main body size–related differences in hydrodynamic sort- N = 58 ing of zircon grains of different ages. Fault Cowfeed Ridge Taken together, the results of this study N = 69 Red Creek confrm the existence a fuvial connection be- 41°3’N tween the CMB and Lake Gosiute. Riverine N = 93 transport of excess alkalinity to Lake Gosiute is more diffcult to prove, primarily because Cannonball Ridge the detailed history of the CMB is much less N = 74 well known than that of the Green River Forma- tion. The youngest detrital zircon ages reported Cherokee Creek here predate the onset of evaporite deposition N = 64 CR-148 5 km by ∼3–4 m.y. (Fig. 3), which on its face might suggest that hydrothermal activity ceased before N evaporite deposition began. Several other fac- tors could also account for this temporal gap, Figure 2. Geologic map of the study area showing geologic formations, faults, paleocurrent however, including inheritance of older zircon rosettes, and mean vector paleocurrent direction at nine locations in the southern Washakie Basin (Wyoming, USA; see Fig. 1 for location). Formations and faults are adapted from Love cores, low zircon fertility in exposed igneous and Christiansen (1985); see Figure DR1 in the Data Repository1 for measured sections. Total rocks, and the time required to unroof newly of 707 paleocurrent measurements were made from planar and trough crossbeds. CR-148— emplaced intrusive bodies. More signifcantly, County Road 148 locality. the timing of Laramide hydrothermal activity is only loosely known. Bryant et al. (1990) pos- subordinate 75–500 Ma grains, in approxi- consistent directions (Pietras and Carroll, 2006). tulated that a hydrothermal event occurred at mately equal proportion to the 900–1200 Ma However, the pronounced northward fning and 52.1 ± 4.4 Ma near Aspen, based on localized grains. Finally, all samples exhibit multiple thinning trends observed within the sandstone fssion-track annealing of zircon in a miner- peaks <75 Ma. These grains compose ∼5% of marker intervals A–I (Pietras and Carroll, 2006; alized porphyry, but other similar studies are the Wasatch Formation main body sample and Smith et al., 2015) strongly imply overall south- scarce. The late Cenozoic history of the CMB 14%–15% of the Cathedral Bluffs Tongue and to-north transport. clearly demonstrates that hydrothermal activity Wilkins Peak Member samples. Their detailed The composition and textural immaturity of need not be associated with coeval detrital zir- distribution varies, but the most prominent peak sandstone in the Washakie and Bridger Basins con. No intrusive rocks younger than 5 Ma are in all samples occurs at 56–59 Ma. are consistent with frst-cycle derivation from exposed anywhere in the CMB today. Modern Proterozoic crystalline rocks in the Front, Gore, hydrothermal “soda springs” are nonetheless DISCUSSION AND CONCLUSIONS and/or Sawatch Ranges of Colorado. The pres- common, and have chemistries similar to the Fluvial facies with west- to northwest-direct- ence of sedimentary lithic grains likely indicates waters that provided excess alkalinity to Sear- ed paleocurrents clearly attest to a paleoriver (or recycling of Paleozoic–Mesozoic strata fank- les Lake (Evans et al., 1986; Lowenstein et al., rivers) that entered the Washakie Basin from the ing Precambrian-cored uplifts. Volcanic lithic 2016). Alkali-olivine basalt <5 Ma does occur southeast. We infer that river transport expand- grains may have been derived from CMB volca- in within the CMB (Klein et al., 2010). Detrital ed further west during ∼100 k.y. lowstands of nic rocks. Preserved extrusive rocks of Laramide zircon grains <30 Ma are absent from Holocene Lake Gosiute, crossing the southern end of the age are generally scarce within the CMB, how- river samples collected downstream, however Rock Springs arch and depositing fner-grained ever, either because they were never widespread (Kimbrough et al., 2015), refecting the limited siliciclastic sediment downstream in the Bridger or because of later erosion. areal extent of basalt fows, their low zircon fer- Basin. The local paleofow pattern within the The scarcity of Archean detrital zircon grains tility, or both. Bridger Basin is more ambiguous; measured pa- clearly demonstrates that few if any grains were Based on the above arguments, the CMB leocurrent indicators generally have not yielded derived directly from Wyoming province base- cannot be excluded as the source of excess al- ment. Conversely, detrital peaks with Yavapai- kalinity that enabled downstream Na-carbonate Mazatzal province and “anorogenic” granite evaporite deposition. A simple numerical model 1GSA Data Repository item 2019361, Figure DR1 ages are consistent with derivation from Pro- shows that thermal spring discharge equal to (measured sections), Table DR1 (sandstone petrog- terozoic basement in central Colorado. The 1%–10% of the total runoff to Lake Gosiute raphy), and Table DR2 (detrital zircon age data), is available online at http://www.geosociety.org/datare- 900–1200 Ma and 75–500 Ma zircons are in- could account for the observed mass of trona pository/2019/, or on request from editing@geoso- ferred to derive primarily from reworking of Pa- (Fig. 4). Additional studies are needed to ful- ciety.org. leozoic–Mesozoic strata (cf. Dickinson et al., ly elucidate the detailed history of Laramide

1022 www.gsapubs.org | Volume 47 | Number 11 | GEOLOGY | Geological Society of America

Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/47/11/1020/4852043/1020.pdf by University of Wisconsin Madison user on 23 October 2019 Yavapai- REFERENCES CITED 56 Ma 64 Ma Figure 3. Kernel density (n = 43) Mazatzal estimates based on detri- Aswasereelert, W., Meyers, S.R., Carroll, A.R., 59 Ma Anorogenic Green River Fm. tal zircon ages reported in ­Peters, S.E., Smith, M.E., and Feigl, K.L., 2013, granite Wilkins Peak this study (cf. Vermeesch, ­Basin-scale cyclostratigraphy of the Green Riv- Member 2012). A total of 861 detri- er Formation, Wyoming: Geological Society of n = 294 tal zircon U-Pb ages were America Bulletin, v. 125, p. 216–228, https://doi​ measured at the Ari- .org/10.1130/B30541.1. zona LaserChron Center Bolen, W.P., 2018, Soda ash, in Mineral Commodity (Tucson, Arizona, USA), Summaries 2018: Reston, Virginia, U.S. Geologi- 58 Ma (n = 65) Grenville Wasatch Fm. based on five samples cal Survey, p. 152–153. from Washakie Basin and Bookstrom, A.A., 1990, Igneous rocks and carbon- Tongue one sample from Firehole ate-hosted ore deposits of the central Colo- n = 449 Canyon in Bridger Basin, rado ­mineral belt, in Beaty, D.W., et al., eds., 62 Ma 66 Ma (4 samples) Wyoming (Fig. 1; see Table ­Carbonate-Hosted Sulfde Deposits of the Colo- Wilkins Peak Member Wilkins Peak DR3 [see footnote 1] for rado Mineral Belt: Society of Economic Geolo- full data). Vertical scale on gists Economic Geology Monograph 7, p. 45–65. 59 Ma Wasatch Fm. each plot is an expression Bryant, B., Naeser, C.W., and Stegen, R.J., 1990, (n = 8) Colorado Mineral main body of the relative probability Reconnaissance fssion-track geochronology of Belt n = 118 of encountering a par- the Aspen mining district, central Colorado, in ticular zircon age in each Beaty, D.W., et al., eds., Carbonate-Hosted Sul- 66 Ma 76 Ma sample. Circles beneath fde Deposits of the Central Colorado Mineral each plot indicate the Belt: Society of Economic Geologists Economic measured ages of indi- Geology Monograph 7, p. 301–307. vidual zircon grains. Plots Chapin, C.E., 2012, Origin of the Colorado Min- on the left were calculated eral Belt: Geosphere, v. 8, p. 28–43, https://doi​ 50 60Ma 70 80 010002000 3000 Ma with bandwidth = 1 m.y., .org/10.1130/GES00694.1. plots on the right with Chetel, L.M., Janecke, S.U., Carroll, A.R., Beard, bandwidth = 0 1 m.y. Four Cathedral Bluffs Tongue (Wasatch Formation) sandstone samples B.L., Johnson, C.M., and Singer, B.S., 2011, Pa- were combined in one plot due to their geographic proximity, similar petrographic composi- leogeographic reconstruction of the Eocene Idaho tion, and similar detrital zircon age distributions. River, North American Cordillera: ­Geological Society of America Bulletin, v. 123, p. 71–88, https://doi​.org/10.1130/B30213.1. 12 Figure 4. Thermal spring Davis, S.J., Dickinson, W.R., Gehrels, G.E., Spencer, alkalinity versus dis- J.E., Lawton, T.F., and Carroll, A.R., 2010, Pa- charge rate required to leogene California River: Evidence of Mojave- 10 Assumptions: account for trona depos- Uinta paleodrainage from U-Pb ages of detrital 14 its of the Wilkins Peak zircons: Geology, v. 38, p. 931–934, https://doi​ mg/yr .org/10.1130/G31250.1. 8 Member of the Green 10 L/yr River Formation, Wyo- Dickinson, W.R., Klute, M.A., Hayes, M.J., Janeck­ e, ming, USA. Based on this S.U., Lundin, E.R., McKittrick, M.A., and 6 analysis, spring discharge ­Olivares, M.D., 1988, Paleogeographic and representing ∼1%–10% of paleotectonic setting of Laramide sedimentary total runoff to the Eocene basins in the central Rocky Mountain region: 4 Lake Gosiute (Wyoming) Geological Society of America Bulletin, v. 100, could have supplied the p. 1023–1039, https://doi​.org/10.1130/0016- 7606(1988)100≤1023:PAPSOL≥2.3.CO;2. 2 excess alkalinity required to precipitate the pre- Dickinson, W.R., Lawton, T.F., Pecha, M., Davis, served mass of trona. S.T., Gehrels, G.E., and Young, R.A., 2012, 0 Total runoff to the lake Provenance of the Paleogene Colton Forma- 0 1000 2000 3000 4000 5000 was estimated by Doe- tion (Uinta Basin) and Cretaceous–Paleogene Spring alkalinity (mg/L) bbert et al. (2010, 2014) provenance evolution in the Utah foreland: based on mass balance ­Evidence from U-Pb ages of detrital zircons, Range of modern Colorado soda springs modeling of Lake Gosiute paleocurrent trends, and sandstone petrofa- (Utah). Alkalinity fux is cies: Geosphere, v. 8, p. 854–880, https://doi​ based on the Smith et al. .org/10.1130/GES00763.1. (2008a) calculation of trona accumulation rate in the lower Wilkins Peak Member (subject Doebbert, A.C., Carroll, A.R., Mulch, A., Chetel, L.M., and Chamberlain, C.P., 2010, Geomorphic to ± 21% uncertainty). Data for modern CO2-rich soda springs in Colorado were reported by Evans et al. (1986). controls on lacustrine isotopic compositions: Evidence from the Laney Member, Green Riv- er Formation, Wyoming: Geological Society of ­hydrothermal activity in time and space. Ironi- magmatic histories and the broader evolution of America Bulletin, v. 122, p. 236–252, https://doi​ cally, the best evidence for this history may continental landscapes. .org/10.1130/B26522.1. come not from the CMB itself, but from fur- Doebbert, A.C., Johnson, C.M., Carroll, A.R., Beard, B.L., Pietras, J.T., Rhodes Carson, M., Norsted, ther investigation of its downstream infuence ACKNOWLEDGMENTS B., and Throckmorton, L.A., 2014, Controls on as recorded stratigraphically within the Green We thank S.P. Peters, S.R. Meyers, and P.E. Brown Sr isotopic evolution in lacustrine systems: Eo- River Formation. For example, 3He/4He isotope for their helpful comments and insights. I. ­Dammann cene Green River Formation, Wyoming: Chemi- ratios within fuid inclusions of evaporite de- provided invaluable assistance in the field, and cal Geology, v. 380, p. 172–189, https://doi​ I. McBrearty provided computational support. posits might provide evidence for mantle input .org/10.1016/​j.chemgeo.2014.04.008. M. Pecha helped obtain data at the University of Ari- Earman, S., Phillips, F.M., and McPherson, B.J.O.L., to magmatic fuids. The potential relationship zona LaserChron Center (Tucson, Arizona, USA). We 2005, The role of “excess” CO2 in the forma- of CMB magmatism to Na-carbonate evaporite also thank the anonymous reviewers of this paper and tion of trona deposits: Applied Geochemistry, deposits in the Piceance Creek and Uinta Ba- of an earlier version. Funding was provided by National v. 20, p. 2217–2232, https://doi​.org/10.1016/​ Science Foundation grants EAR-1812741, EAR- sins (Fig. 1) also needs further investigation. j.apgeochem.2005.08.007. 1813278, and EAR-1813350, a Geological Society of Evans, W.C., Presser, T.S., and Barnes, I., 1986, Se- More generally, this study points to the utility America Student Research Grant, and the Department lected soda springs of Colorado and their ori- of lake deposits in helping to interpret complex of Geoscience, University of Wisconsin–Madison. gin, in Subitzky, S., ed., Selected Papers in the

Geological Society of America | GEOLOGY | Volume 47 | Number 11 | www.gsapubs.org 1023

Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/47/11/1020/4852043/1020.pdf by University of Wisconsin Madison user on 23 October 2019 ­Hydrologic Sciences 1986: U.S. Geological Sur- Love, J.D., and Christiansen, A.C., compilers, 1985, America Bulletin, v. 120, p. 54–84, https://doi​ vey Water-Supply Paper 2310, p. 45–52. Geologic map of Wyoming: Reston, Virginia, .org/10.1130/B26073.1. Farrar, C.D., Sorey, M.L., Rojstaczer, S.A., Janik, C.J., U.S. Geological Survey, 3 sheets, scale 1:500,000, Smith, M.E., Carroll, A.R., Scott, J.J., and Singer, B.S., Winnett, T.L., and Clark, M.D., 1987, Hydrologic https://doi​.org/10.3133/70046739. 2014, Early Eocene carbon isotope excursions and geochemical monitoring in Long Valley Calde- Lowenstein, T.K., Dolginko, L.A., and García-Veigas, and landscape destabilization at eccentricity min- ra, Mono County, California, 1985: U.S. Geologi- J., 2016, Infuence of magmatic-hydrothermal ima: Green River Formation of Wyoming: Earth cal Survey Water-Resources Investigations Report activity on brine evolution in closed basins: and Planetary Science Letters, v. 403, p. 393–406, 87-4090, 71 p., https://doi.org/10.3133/wri874090​ . Searles Lake, California: Geological Society of https://doi​.org/10.1016/​j.epsl.2014.06.024. Hardie, L.A., and Eugster, H.P., 1970, The evolution of America Bulletin, v. 128, p. 1555–1568, https:// Smith, M.E., Carroll, A.R., and Scott, J.J., 2015, closed-basin brines, in Morgan, B.A., et al., eds., doi​.org/10.1130/B31398.1. Stratigraphic expression of climate, tectonism, Fiftieth Anniversary Symposia: Mineralogy and Lowenstein, T.K., Jagniecki, E.A., Carroll, A.R., Smith, and geomorphic forcing in an underflled lake Petrology of the Upper Mantle; Sulfdes; Miner- M.E., Renaut, R.W., and Owen, R.B., 2017, The basin: Wilkins Peak Member of the Green River alogy and Geochemistry of Non-Marine Evapo- Green River salt mystery: What was the source of Formation, in Smith, M.E. and Carroll, A.R., eds., rites: Mineralogical Society of America Special the hyperalkaline lake waters?: Earth-Science Re- Stratigraphy and Paleolimnology of the Eocene Paper 3, p. 273–290. views, v. 173, p. 295–306, https://doi.org/10.1016/​ ​ Green River Formation, Western U.S.: Dordrecht, Jones, B.F., 1966, Geochemical evolution of closed j.earscirev.2017.07.014. Springer, p. 61–102, https://doi.org/10.1007/978-​ basin water in the western Great Basin, in Ran, Pietras, J.T., and Carroll, A.R., 2006, High-resolution 94-017-9906-5_4. J.L., ed., Proceedings of the Second Symposium sequence stratigraphy of an underflled lacustrine Sorey, M.L., 1985, Evolution and present state of on Salt: Cleveland, Northern Ohio Geological basin: Evidence for tectonic controls on the saline the hydrothermal system in Long Valley cal- Society, p. 181–200. Wilkins Peak Member of the Green River Forma- dera: Journal of Geophysical Research, v. 90, Kimbrough, D.L., Grove, M., Gehrels, G.E., Dorsey, tion, Wyoming: Journal of Sedimentary Research, p. 11,219–11,228, https://doi​.org/10.1029/ R.J., Howard, K.A., Lovera, A.A., House, P.K., v. 76, p. 1197–1214, https://doi​.org/10.2110/ JB090iB13p11219. and Pearthree, P.A., 2015, Detrital zircon U-Pb jsr.2006.096. Vermeesch, P., 2012, On the visualization of detrital age provenance of the Colorado River: A 5 m.y. Smith, M.E., Carroll, A.R., and Mueller, E.R., 2008a, distributions: Chemical Geology, v. 312, p. 190–194, record of incision into cover strata overlying Elevated weathering rates in the Rocky Moun- https://doi​.org/10.1016/​j.chemgeo.2012.04.021. the Colorado Plateau and adjacent regions: tains during the Early Eocene Climatic Optimum: Warren, J.K., 2010, Evaporites through time: Tec- Geosphere, v. 11, p. 1719–1748, https://doi​ Nature Geoscience, v. 1, p. 370–374, https://doi​ tonic, climatic and eustatic controls in marine .org/10.1130/GES00982.1. .org/10.1038/ngeo205. and nonmarine deposits: Earth-Science Re- Klein, T.L., Evans, K.V., and DeWitt, E.H., 2010, Geo- Smith, M.E., Carroll, A.R., and Singer, B.S., 2008b, views, v. 98, p. 217–268, https://doi.org/10.1016/​ ​ chronology database for central Colorado: U.S. Synoptic reconstruction of a major ancient j.earscirev.2009.11.004. Geological Survey Data Series 489, 13 p., https:// lake system: Eocene Green River Formation, doi​.org/10.3133/ds489. western United States: Geological Society of Printed in USA

1024 www.gsapubs.org | Volume 47 | Number 11 | GEOLOGY | Geological Society of America

Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/47/11/1020/4852043/1020.pdf by University of Wisconsin Madison user on 23 October 2019