SEARCH · Geology Reports
Results for “GSA Special Papers”
Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Introduction in New perspectives on Rio Grande rift basins: from tectonics to groundwater
Basins of the Rio Grande rift have long been studied both for their record of rift development and for their potential as host of natural resources. Early workers described the basin geomorphology and the character of infilling sediments (e.g. Siebenthal, 1910; Bryan, 1938; Speigel and Baldwin, 1963), and subsequent research compilations provided general stratigraphic and tectonic overviews of rift basins and described their geophysical characteristics within the crust (Hawley, 1978; Riecker, 1979; Baldridge et al., 1984; Keller, 1986). Subsurface knowledge gained from hydrocarbon exploration activities coupled with detailed surface studies of basins and their flanking uplifts were presented in Geological Society of America (GSA) Special Paper 291, edited by Keller and Cather (1994a).
Allogenic and autogenic controls on sedimentation in the central Sumatra basin as an analogue for Pennsylvanian coal-bearing strata in the Appalachian basin
Recent sedimentation patterns in the central Sumatra basin, Republic of Indonesia, may help to explain the cyclic stratigraphy of the Pennsylvanian System of the eastern United States. Modern influx of fluvial siliciclastic sediment to the epeiric seas of the Sunda shelf, including the Strait of Malacca, appears to be highly restricted by rain forest cover within the ever-wet climate belt of equatorial Sumatra. As a result, much of the marine and estuarine environments appear to be erosional or nondepositional except for localized deposition of sediment in slack water areas, such as the down-stream end of islands. Contemporaneously, thick (>13 m), laterally extensive (>70,000 km 2 ), peat deposits are forming on poorly drained coastal lowlands. Modern peat formation in this study, therefore, is not coeval with aggrading fluvial siliciclastic systems, a situation that commonly is assumed in many depositional models of coal formation. The stratigraphy of Pleistocene and Holocene sediments on the Sunda shelf, as well as those of the Pennsylvanian System, appears to be better explained by the allocyclic controls of climate and sea-level change on sediment flux rather than by depositional models that are based on autocyclic processes. The objective of this paper is to evaluate allocyclic and autocyclic controls on sedimentation in an epeiric setting in a humid (ever-wet) tropical region. Of particular interest are the factors that control peat formation and siliciclastic sediment flux in rivers, estuaries, and open marine environments.
Overcoming the momentum of anachronism: American geologic mapping in a twenty-first-century world
The practice of geologic mapping is undergoing conceptual and methodological transformation. Profound changes in digital technology in the past 10 yr have potential to impact all aspects of geologic mapping. The future of geologic mapping as a relevant scientific enterprise depends on widespread adoption of new technology and ideas about the collection, meaning, and utility of geologic map data. It is critical that the geologic community redefine the primary elements of the traditional paper geologic map and improve the integration of the practice of making maps in the field and office with the new ways to record, manage, share, and visualize their underlying data. A modern digital geologic mapping model will enhance scientific discovery, meet elevated expectations of modern geologic map users, and accommodate inevitable future changes in technology.
The 2008 U.S. Geological Survey national seismic hazard models and maps for the central and eastern United States
In this paper, we describe the scientific basis for the source and ground-motion models applied in the 2008 National Seismic Hazard Maps, the development of new products that are used for building design and risk analyses, relationships between the hazard maps and design maps used in building codes, and potential future improvements to the hazard maps.
High-precision 40Ar/39Ar age spectrum dating of sanidine from the Middle Pennsylvanian Fire Clay tonstein of the Appalachian basin
40 Ar/ 39 Ar plateau age spectra of seven sanidine samples from the Fire Clay tonstein (Middle Pennsylvanian), collected along a 300-km traverse in the Appalachian basin, range from 310.3 to 311.4 Ma. All plateau ages agree, within the limits of analytical precision, with their respective total gas ages. This agreement, together with the reproducibility between samples, suggests the analyzed samples did not contain any significant contaminant feldspar. The mean of these seven plateau ages, 310.9 ± 0.8 Ma, is interpreted to represent a precise numerical estimate of time of eruption and deposition of this tonstein and the coal bed in which it is found. The lack of any discernible difference between the age of two samples of the Fire Clay tonstein collected from east of the Pine Mountain thrust fault, along with the age of five samples from west of this fault, suggests that the Fire Clay tonstein has been reliably correlated with a tonstein on the Cumberland overthrust sheet. This correlation, together with the age data presented in this paper, indicates that the Pine Mountain thrust fault must be younger than the 310.9-Ma age obtained for the Fire Clay tonstein. The Fire Clay tonstein is biostratigraphically correlated with the Trace Creek Shale Member of the Atoka Formation in the Midcontinent of North America and with a position near the Westphalian B-C boundary in Western Europe. Our age of 310.9 ± 0.8 Ma for the Westphalian B-C boundary represents a well-constrained point, useful for the numerical refinement of the geologic time scale.
Geologic map of Lake Mead and surrounding regions, southern Nevada, southwestern Utah, and northwestern Arizona
Regional stratigraphic units and structural features of the Lake Mead region are presented as a 1:250,000 scale map, and as a Geographic Information System database. The map, which was compiled from existing geologic maps of various scales, depicts geologic units, bedding and foliation attitudes, faults and folds. Units and structural features were generalized to highlight the regional stratigraphic and tectonic aspects of the geology of the Lake Mead region. This map was prepared in support of the papers presented in this volume, Special Paper 463, as well as to facilitate future investigations in the region. Stratigraphic units exposed within the area record 1800 million years of geologic history and include Proterozoic crystalline rocks, Paleozoic and Mesozoic sedimentary rocks, Mesozoic plutonic rocks, Cenozoic volcanic and intrusive rocks, sedimentary rocks and surficial deposits. Following passive margin sedimentation in the Paleozoic and Mesozoic, late Mesozoic (Sevier) thrusting and Late Cretaceous and early Tertiary compression produced major folding, reverse faulting, and thrust faulting in the Basin and Range, and resulted in regional uplift and monoclinal folding in the Colorado Plateau. Cenozoic extensional deformation, accompanied by sedimentation and volcanism, resulted in large-magnitude high- and low-angle normal faulting and strike-slip faulting in the Basin and Range; on the Colorado Plateau, extension produced north-trending high-angle normal faults. The latest history includes integration of the Colorado River system, dissection, development of alluvial fans, extensive pediment surfaces, and young faulting. Stratigraphic units exposed within the area record 1800 million years of geologic history and include Proterozoic crystalline rocks, Paleozoic and Mesozoic sedimentary rocks, Mesozoic plutonic rocks, Cenozoic volcanic and intrusive rocks, sedimentary rocks and surfi cial deposits. Following passive margin sedimentation in the Paleozoic and Mesozoic, late Mesozoic (Sevier) thrusting and Late Cretaceous and early Tertiary compression produced major folding, reverse faulting, and thrust faulting in the Basin and Range, and resulted in regional uplift and monoclinal folding in the Colorado Plateau. Cenozoic extensional deformation, accompanied by sedimentation and volcanism, resulted in large-magnitude high- and low-angle normal faulting and strike-slip faulting in the Basin and Range; on the Colorado Plateau, extension produced north-trending high-angle normal faults. The latest history includes integration of the Colorado River system, dissection, development of alluvial fans, extensive pediment surfaces, and young faulting.
Estimates of the abundances of some chemical elements and their reliability
During the past thirty years the development of improved methods of analysis, especially optical spectrography, x-ray spectrography, and colorimetric methods, has resulted in a wealth of new data on the abundances of many elements. Nevertheless, present estimates differ greatly in reliability; some, such as the estimate for bismuth, are based on very few actual determinations, whereas others, such as that for gallium, are based on analyses of many samples. This paper reviews the estimates of abundance for the elements bismuth, germanium, gallium, and yttrium, taken as examples to illustrate the wide variation in reliability. For gallium and yttrium, a study has been made of spectrographic analyses by four laboratories of more than 600 chemically analyzed rocks from many areas of the world. Comparison shows that these furnish valuable data on abundance, although attention must be directed to regional variation. The comparison also shows the need for continued interlaboratory standardization. The estimate for bismuth is based on so few determinations that no appraisal of its accuracy can be made. The estimate for germanium is based on very few determinations but is almost certainly of the right order of magnitude, though perhaps slightly too high. The estimates for gallium and yttrium, based on many determinations, are almost certainly of the right order of magnitude, though that for gallium may be slightly low and that for yttrium may be substantially low.
Correlation of late Cenozoic basaltic lava flows in the Carbondale and Eagle collapse centers in west-central Colorado based on geochemical, isotopic, age, and petrographic data
Major-, minor-, and trace-element abundance data on 220, late Cenozoic, basaltic rocks in and around the Carbondale and Eagle collapse centers in west-central Colorado are combined with isotopic, age, and petrographic data to correlate lava flows and establish the timing and minimum areal extent of collapse events associated with removal of Pennsylvanian evaporite. On the basis of these data, 46 distinct compositional groups of volcanic rocks were identified. The rocks within each group, which are represented by at least two samples, have compositions and ages that are indistinguishable from each other and are either (1) undifferentiated samples from the same eruption and possibly from outcrops of the same flow, (2) differentiated from the same magma batch erupted at different times or (3) related to each other by very small differences in the degree of partial melting. The areal extent of chemically correlated and dated volcanic flows in the region was established and these results were used to recognize and understand many of the collapse events as described in companion papers in this volume. Compositional data are also used to infer the petrogenetic processes that generated the parental magmas. Subtle but significant differences among rocks that are broadly similar in geochemical and isotopic composition and were erupted over a small time interval (<0.5 m.y.) suggests that the mantle source region of these magmas is quite heterogeneous. Over the past 11 m.y., the lavas became less mafic and more enriched in incompatible trace elements. This heterogeneity is attributed to variable contributions of subducted material in the lithosphere during the melting processes. To account for its isotopic features, the source material must be at least mid-Proterozoic in age. A melt contribution from underlying asthenospheric mantle can not be ruled out but none of the volcanic rocks have clear characteristics of oceanic-island basalts or mid-oceanic-ridge basalts. The trace-element compositions of rocks that were sampled from multiple stacked flows are indistinguishable from one another.This suggests that lower and/or upper crustal assimilation had a minor effect on their overall chemistry. However, Pb isotopic data suggest that crustal contamination of at least some samples did occur.
Late Holocene alluvial geomorphology of the Virgin River in the Zion National Park area, southwest Utah
This study traces the geomorphic development of the alluvial valley of the Virgin River in the Zion National Park region of southwest Utah. The purpose is to identify, date, and interpret the patterns of erosion and deposition that formed the alluvial valley over the past 1,000 years. This information is a basis for understanding how the geomorphology of the alluvial valley changes under essentially natural flow conditions. Sediment in the alluvial valley is classified as mainstem or tributary in origin. Mainstem alluvium is the largest volumetrically; it is mainly light-colored sand derived from upstream sources that accumulated on now abandoned flood plains (or terraces) by overbank deposition. Tributary sediment is typically dark colored, coarse-grained sand or gravel transported to the alluvial valley by streamflow or debris flow. Tributary and mainstem sediment are interbedded near the margin of the valley, and the older deposits are truncated parallel to the river, suggesting that in time the river removes both its own deposits and those of tributaries. In the natural flow regimen, the river probably maintains a balance between erosion and deposition of mainstem and tributary deposits. Four terraces and the active channel and flood plain are widespread along the Virgin River; from oldest to youngest these are the prehistoric, settlement, historic, and modern terraces. Dating was done by archeologic context, tree-ring methods, historic documents, relocation of early photographs, and correlation with other streams on the southern Colorado Plateau. Results indicate that prehistoric deposition ended by about A.D. 1100–1200, deposition of the settlement alluvium was from about A.D. 1400–1880, development of the historic terrace was from after 1883 until 1926, deposition of the modern alluvium was from 1940–1980, and development of the active channel and flood plain was after about 1980. The principal deposits (prehistoric, settlement, and modern alluviums) are separated by two periods of stream entrenchment and channel widening referred to as the prehistoric and historic arroyo cutting, respectively. Erosional activity of roughly similar age occurred in most southern Colorado Plateau streams. The early erosion is not well dated in the study area, although regional relations suggest A.D. 1200–1400, if not somewhat earlier. Historic arroyo cutting began after 1883 in the study area and continued until around 1940 when deposition of the modern alluvium began. Both erosions affected the human population of the region. Although the dating is imprecise, the Anasazi probably abandoned the region during prehistoric arroyo cutting, partly because of adverse environmental conditions. Likewise, historic arroyo cutting caused major losses of property and economic hardship among Anglo settlers. Erosion and deposition were largely contemporaneous with variations in streamflow. Long-term streamflow of the Virgin River was estimated from calibration of annual tree growth with measured streamflow. Results indicate that erosion was during periods of unusually high streamflow and that deposition was during periods of relatively low streamflow. These relations are best illustrated by historic arroyo cutting and subsequent deposition of the modern alluvium. Precipitation and runoff immediately before and during historic arroyo cutting were the most unusual of the past 300 years; they varied from the driest immediately preceding erosion to the wettest during erosion. Deposition of the modern alluvium was during relatively low runoff after 1940. High runoff destabilizes the channel, enhancing flood erosion. Conversely, relatively low runoff increases channel stability, reducing the erosional effect of floods and enhancing flood plain deposition. Adjustments in the width and depth of the channel are frequent, even after relatively small, short-term variations of streamflow.
Cenozoic basins of the Death Valley extended terrane as reflected in regional-scale gravity anomalies
The contrast in density between basin-filling deposits and underlying basement rocks in the Basin and Range province produces a distinctive pattern of gravity anomalies that roughly correlates with basin-and-range topography. This is especially apparent in the Death Valley extended terrane, where minima in residual gravity closely correspond to well-known extensional basins, such as Owens Valley, Saline Valley, Pahrump Valley, and Death Valley itself. A three-dimensional calculation of the thickness and shape of Cenozoic deposits throughout the Basin and Range based on digital gravity, geologic, and topographic information is here interpreted in terms of the depths and shapes of extensional basins in the Death Valley region. In most basins, the gravity inversion predicts a much more complex basement than might be assumed from the surface topography. Death Valley itself is, to first order, a two-dimensional feature; as reflected by the topography of the bounding ranges down to 400–800 m below the valley floor, it is a long, narrow, and continuous valley along its entire length of 200 km. This apparent uniformity, however, belies a complex basement surface beneath the Cenozoic cover as revealed by the gravity inversion. The floor of Death Valley is pockmarked with four deep (3–5 km), steep-sided depressions that may have formed as relatively small pull-apart structures superimposed on the more uniform extension that created the twodimensional aspects of Death Valley. Lateral dimensions of these subbasins suggest a minimum of 10 km of northwest extension across Death Valley. If a regionally continuous detachment surface lies beneath Death Valley, it must lie deeper than the subbasins, that is, below a depth of 3–5 km. Similar deep, steep-sided subbasins are seen elsewhere in the Death Valley extended terrane, notably at Saline Valley, Panamint Valley, Pahrump Valley, the Amargosa Desert, and Fremont Valley, but are noticeably absent from the Mojave Desert block south of the Garlock fault zone where extensional rates are lower. The floor of the caldera complex at the southwest Nevada volcanic field is predicted by the gravity inversion to be extraordinarily deep (>8 km) and rugged. Deep, linear basins that branch out of the caldera complex may reflect modification of extensional strain due to massive injections of magma into the middle and upper crust.
Late Cenozoic stratigraphy and tephrochronology of the western Black Mountains piedmont, Death Valley, California: Implications for the tectonic development of Death Valley
Geologic mapping combined with the tephrochronology of spatially isolated sedimentary sections along the western Black Mountains piedmont adjacent the Death Valley fault zone (DVFZ) improves the late Cenozoic stratigraphy from relative age to correlated age. Pliocene tephra layers identified in Funeral Formation conglomerates at Artist Drive and Copper Canyon include a “Nomlaki-like” tephra bed (ca. 3.4 Ma), the tuffs of Mesquite Spring (3.1–3.3 Ma), and a tuff of the lower Glass Mountain family (1.86–1.92 Ma). We informally name the early(?) to middle Pleistocene Mormon Point formation1, which contains tephra layers correlated with the upper Glass Mountain/Bishop family of tephra layers (0.76–1.2 Ma), the Lava Creek B ash bed (ca. 0.66 Ma), and the Dibekulewe ash bed (ca. 0.51 Ma). Identification of these tephra layers indicates that the maximum age of the overlying and inset lacustrine gravel and alluvial fan deposits is 0.51 Ma. The correlated age stratigraphy indicates that the dextral-oblique DVFZ has stepped basinward at Mormon Point and Copper Canyon since the late Pliocene. In contrast, during that same time the DVFZ at Artist Drive has not stepped basinward, but developed into a graben. The age of faulting on the low-angle (~19°–40°) Mormon Point turtleback fault is bracketed between 0.76 and 0.18 Ma, and the overlying Mormon Point formation shows no evidence of tilting, indicating slip on the turtleback fault was at a low-angle. Early Quaternary slip on the low-angle turtleback fault conflicts with the present versions of the pure shear, rolling-hinge, and detachment/rift models for Death Valley extension. Early Quaternary slip is most compatible with turtleback faults as folded or warped detachment fault. We propose that the warping is thermally driven and related to the Black Mountains igneous complex.
Overview: Ancient Lake Creede
Lake Creede was moderately saline closed-basin lake that developed in the 26.9 Ma Creede caldera in the San Juan Mountains in the southwest Colorado. The volcaniclastic sediments deposited within the late Oligocene lake were first described and named as the Creede Formation by Emmons and Larsen (1923). The lake and its sedimentary fill are of interest first as representatives of a caldera-hosted lake in a silicic volcanic terrane, and second because of the likely involvement of lake fluids or related pore waters in the deposition of the 25 Ma silver and base-metal ores of the Creede mining district north of the Creede caldera (Fig. 1), as proposed Bethke and Rye (1979). Much of the material presented in this volume is based on observation of core samples and on downhole geophysical measurements obtained as part of a U.S. Continental Scientific Drilling Program in the moat of the Creede caldera. These core and downhole studies are supplemented by outcrop studies, some initiated in support of the drilling program (Bethke and Lipman, 1987), and by conceptual studies of the evolution of the Creede caldera and its surrounding landscape. Not surprisingly, not all authors agree on all interpretation. Most disagreements are pointed out in this overview chapter, and may present opportunities for future study.
Hydrologic budget of the late Oligocene Lake Creede and the evolution of the upper Rio Grande drainage system
The filling history, hydrologic budget, and geomorphic development of ancient Lake Creede and its tributary basin are evaluated to determine the factors that controlled its character. The lake filled the Creede caldera that formed in the late Oligocene as a consequence of the eruption of the Snowshoe Mountain Tuff. The caldera's sedimentary fill accumlated to a depth of about 1.26 km and had a volume of about 89 km 3 . The highest lake level was ~3300 m (10,800 ft) present altitude before it drained eastward across a broad volcanic plateau as the ancestral Rio Grande. A tributary canyon several hundred meters deep was cut into hard rhyolite in the north wall of the caldera before the lake was more than half full; its presence demonstrates that ancient Lake Creede filled slowly and thus occupied a long-lived, closed basin. The slow filling rate is incompatible with the present water flux through the Creede caldera basin, because such a flow would fill the basin geologically instantaneously. This mismatch, together with the recognition that the Oligocene climate was similar to that of today, forces the reexamination of the hydrologic and geomorphic history of the caldera. That appraisal shows that the caldera cannot have resurged rapidly immediately after caldera collapse, and that ancient watershed must have been lass than half as large as the present upper Rio Grande basin. The ancient lake had a more or less constant surface area of about 200 km 2 that approximated a steady-state condition between inflow and evaporation. Although the lake level fluctuated with climatic variations, its surface elevation steadily climbed as sediment accumulated, accelerating as resurgance and dome growth usurped spacewithin the basin. It could have had one playa stage early in its development and another after the basin had nearly filled with sediment, but there is no direct evidence for either. At least the lower half of the sedimentary column (the part sampled by the scientific drilling) formed in an euxinic environment. This argues against a persistent early playa, although evaporative accumulation of brine was inevitable. When the rate of resurgance was rapid relative to sedimentary infilling, the lake would have been deep (i.e., bordered by bedrock rather than sedimentary fans). The geomorphic evolution of the Creede caldera and its watershed tracks a two-phase topographic history, the first the Oligocene through Miocene, and the second for Pliocene to the recent. In Oligocene time, the San Juan volcanic field was a hydrologically immature, gently undulating, and outward sloping, constructional volcanic plateau straddling the ancient Continental Divide. West of the Creede caldera, a dendritic drainage discharged northeastward into ancestral Cebolla Creek (a tributary of the ancestral Gunnison River) through an early stage of the Clear Creek graben in the vicinity of Spring Creek Pass. Miocene basalt choked, but did not reconstruct, the drainage. By the end of Miocene time a mature topography of moderate relief developed, exposing some of the higher ores in the Creede district to weathering. In the late Miocene-early Pliocene time the San Juan Mountains were uplifted and titled eastward; the ancestral Rio Grande was revitalized and cut deeply into the older terrain, excavating much of the accessible sediment from the moat of the Creede caldera and exposing successively lowe levels in the Creede district to oxidation. Simultaneously, the southeast end of the Clear Creek graben was reactivated and breached the southwest wall of the Creede caldera. The rejuvenated Rio Grande captured the formerly northeast-directed headwaters of ancestral Cebolla Creek, shifting more than 1000 km 2 from the Pacific-directed drainage to the Atlantic. The water budget for ancient Lake Creede was strictly limited by the early stages of the fist geomorphic cycle; the modern water budget is the product of the second cycle.
Stratigraphy, correlation, depositional setting, and geophysical characteristics of the Oligocene Snowshoe Mountain Tuff and Creede Formation in two cored boreholes
Core descriptions and geophysical logs from two boreholes (CCM-1 and CCM-2) in the Oligocene Snowshoe Mountain Tuff and Creede Formation, south-central Colorado, are used to interpret sedimentary and volcanic facies associations and their physical properties. The seven facies association include a mixed sequence of intracaldera ash-flow tuffs and breccias, alluvial and lake margin deposits, and tuffaceous lake beds. These deposits represent volcanic units related to caldera collapse and emplacement of the Snowshoe Mountain Tuff, and sediments and pyroclastic material deposited in the newly formed caldera basin, Early sedimentation is interpreted to have been rapid, and to have occurred in volcaniclastic fan environments at CCM-1 and in a variery of volcaniclastic fan, braided stream shallow lacustrine, and mudflat environments at CCM-2. After an initial period of lake-level rise, suspension settling, turbidite, and debris-flow sedimentation occurred in lacustrine slope and basin environments below wave base. Carbonate sedimentation was initially sporadic, but more continuous in the latter part of the recorded lake history (after the H fallout tuff). Sublacustrine-fan deposition occurred at CCM-1 after a pronounced lake-level fall and subsequent rise that preceded the H tuff. Variations in density, neutron, gamma-ray, sonic, and electrical properties of deposits penetrated oin the two holes reflect variations in lithology, porosity, and alteration. Trends in the geophysical properties of the lacustrine strata are linked to downhole changes in authigenic mineralology and a decrease in porosity interpreted to have resulted primarily from diagenesis. Lithological and geophysical characteristics provide a basis for correlation of the cores; however, mineralogical methods of correlation are hampered by the degree of diagenesis and alteration.
Stable isotope evolution and paleolimnology of ancient Lake Creede
The lacustrine carbonate and travertine (tufa) deposits of ancient Lake Creede preserve a remarkable record of the isotopic evolution of the lake. That record indicates that the δ18O of the lake water, and by analogy its salinity, evolved through evaporation. Limited and less reliable data on hydrous minerals and fluid inclusions in early diagenetic carbonates indicate that the δD of the lake waters also evolved through evaporation. The isotope data place restrictions on models of the physical limnology of the lake and its evolution. The closed-basin Lake Creede formed shortly after collapse of the 26.9 Ma Creede caldera. Throughout most of its history it occupied the northern three quarters of the moat between the resurgent dome and wall of the caldera. The Creede Formation was deposited in the basin, dominantly as lacustrine sediments. Travertine mounds interfinger with Creede Formation sediments along the inner and outer margins of the lake basin. An estimated one-half of the original thickness of the Creede Formation has been lost mainly to erosion although scattered remnants of the upper portion remain on the caldera walls. Two diamond core holes (CCM-1 and CCM-2) sampled the uneroded portion of the Creede Formation as part of the U.S. Continental Drilling Program. Volcaniclastic material, including tuff units deposited directly into the lake and ash washed in from the watershed, compose the main lithologies of the Creede Formation. These volcaniclastic strata were produced by episodic ring-fracture volcanism. Lacustrine carbonates make up about 15% of the section sampled by drill core. They occur as 1 mm to 2 cm low-Mg calcite laminae alternating with siliciclastic laminae in scattered intervals throughout the preserved section. The carbonate laminae are accumulations of 5–20 µm crystallites (microsparites) and brine shrimp fecal pellets (peloids) composed mainly of microsparite particles. Low-Mg calcite also occurs as an early diagenetic replacement of gypsum or ikaite (CaCO3 ·6H2O) crystals grown displacively in the muds and silts near the water-sediment interface (rice grains). Other studies indicate that aragonite was the original CaCO3 precipitate forming the microsparite and peloidal laminae and that it converted to calcite during burial diagenesis. Samples from CCM-2 and nearby outcrop do not appear to have undergone significant isotope exchange during recrystallization. Samples from CCM-1 and nearby outcrop, however, appear to have undergone extensive oxygen isotope exchange with meteoric water-dominated fluids possibly during a local 17.6 Ma hydrothermal event. The δ18O-δ13C data set produced by microsampling of individual carbonate lamellae and rice grains is exceptional in several aspects and provides important clues concerning the evolution of limnologic structure of the lake and its chemical and isotopic composition. Travertine and ikaite pseudomorphs in travertine deposits extend the record an additional 330 m above the collar of CCM-2. The δ18O values on CCM-2 samples range from 10.4‰ to 37.3‰ and δ13C values range from –10.8‰ to 9.6‰. The data fall into two distinct groups, a covariant group and an invariant group. The covariant group shows a strong negative covariance and a large range of δ18O and δ13C values. The negative covariance is opposite that normally reported for lacustrine carbonates. The large range of δ18O and δ13C values requires that the carbonates precipitated from waters have a large range of temperature and carbon and oxygen isotopic composition. The invariant group has a narrow range of large δ18O values (35‰ ± 2‰) and a wide range of δ13C values (–10.8‰ to 9.6‰), indicating precipitation from waters with a narrow range of temperature and δ18O but a wide range in δ13C of aqueous carbon. The ranges of isotope values for microsparite and peloid samples are virtually identical; two-thirds are in the covariant group. By contrast, the values for almost all rice grain samples are in the invariant group. The range in δ18O for all samples reflects precipitation from waters having varying proportions of deep, cold evaporated lake water and shallow, warmer meteoric water. The range for δ13C reflects varying proportions of organic carbon and carbon of volcanic or atmospheric origin, probably dominantly volcanic, in the aqueous carbon. Changes in the detailed carbon-oxygen isotope systematics with stratigraphic position define three periods of isotopic evolution of Lake Creede. Period I is represented by the lowest ~200 m of Creede Formation core in CCM-2. Analyses of individual microsparite and peloidal carbonate laminae within single thin sections of samples from that interval are tightly grouped. The data set as a whole shows a negative covariance. Rice grains are not found in this interval. Period II is represented by the succeeding 120 m of core in CCM-2. In that interval, δ13C-δ18O values for individual microsparite and peloidal carbonate laminae within single thin sections show strong negative covariance, and the set of values for the entire interval also shows strong negative covariance. Rice grains occur near the top of the interval. Period III is represented by the upper 225 m of CCM-2 core. In this interval, rice grains are abundant and δ13C-δ18O values for microsparite and peloidal laminae as well as rice grains fall in the invariant group. During Period I the lake was well mixed and the oxygen isotopic composition of the lake in the productive zone was only slightly influenced by short-term (e.g., annual) variations in the water budget of the lake. In Period II the lake was stratified, possibly with annual overturn. The productive zone included the mixolimnion and the isotopic composition of the microsparites and peloids reflected mixtures of shallow surface (meteoric) water containing volcanic or atmospheric CO2 (epilimnion) and cold underlying waters, the oxygen isotopic compositions of which had evolved through evaporation and were dominated by CO2 produced by the oxidation of organic matter (hypolimnion). During Period III the lake remained stratified. The productive zone was in the hypolimnion, probably due to a thinning of the epilimnion resulting from an increase in the surface area of the lake or a decrease in input waters reflecting a climate change. An upsection increase in values of δ18O for the heaviest samples during Periods I and II indicates evaporative concentration of 18O and, by analogy, salinity in the hypolimnion. The δD-δ18O evolution of the lake is inferred on theoretical evaporation trends, comparison to Mono Lake, and measurement of the δD in fluid inclusions in a calcite pseudomorph after ikaite. The δD-δ18O composition of the lake water followed a curved path that eventually hooked over at a nearly constant δ18O value for the lake of 2‰ ± 2‰ Travertine (tufa) mounds formed along the inner and outer margins of the lake in a zone of mixing of warm, volcanic CO2-bearing, meteoric waters and lake water. Ikaite crystals formed on the mounds from unmixed saline lake water, probably below the thermocline. As the position of the thermocline deepened, likely following the spring runoff, the ikaite was replaced by calcite and the resulting “pearls” were covered with travertine deposited from mixed meteoric and lake waters. The upsection increase in δ18O values of the carbonates, the long period of invariance of large δ18OH2O values, the presence of brine shrimp fecal pellets, and the inferred hooked δD-δ18O path are consistent with evidence from other studies that Lake Creede obtained significant salinity rather early in its history and certainly by the time the lake became permanently stratified.