Most scientists have concluded previously that the west coast of Spitsbergen, Svalbard, remained ice-free during the late Weichselian, between 25,000 and 10,000 yr B.P. We conclude that the glaciation was more extensive. Terraces that were postulated to have been ice-free are covered by a thin, late Weichselian till. Sudden drop in the marine limit and basal radiocarbon dates of raised glaciomarine sediments demonstrates that the glaciers in the main fjords, Isfjorden and Van Mijenfjorden, terminated west (outside) of the fjord mouths. Basal radiocarbon dates from glaciomarine clay above till in cores from the continental shelf west of Spitsbergen yielded ages of about 12,500 yr B.P., from which we conclude that the ice extended to the shelf edge. Based on the extent of amino acid diagenesis in radiocarbon-dated molluscs, the duration of the maximum extension of the late Weichselian glaciation was short, certainly less than 10,000 years. During the ice-free period preceding that glaciation, at least back to 40,000 yr B.P., the glaciers on Svalbard were not significantly larger than at present, as shown by marine deposits close to the glacier snouts. Many radiocarbon dates place deglaciation of the outer coast at about 12,500 yr B.P. At about 10,000 yr B.P., the rest of the archipelago rapidly became ice-free.

Loess deposits of late Quaternary age in the Channeled Scabland and western Palouse regions of Washington state contain distinctive sequences of calcic buried soils, loess layers, tephra layers, and stratigraphic unconformities caused by cataclysmic flooding. Four new soil-stratigraphic units, the Sand Hills Coulee Soil, Washtucna Soil, Old Maid Coulee Soil, and the Devils Canyon Soil, are recognized across a 100-km transect. The pedostratigraphic correlations are strengthened by two tephra layers found at several sites that are correlative with the Mount St. Helens set S (radiocarbon ages about 13,000 yr B.P.), Mount St. Helens set C (radiocarbon ages about 36,000 yr B.P.), and by a third one that may be from an earlier (pre-set C) Mount St. Helens eruption. The set S tephra layer separates the Sand Hills Coulee and Washtucna Soils; the set C tephra layer separates the Old Maid Coulee and Devils Canyon Soils; and the pre-set C tephra layer underlies the Devils Canyon Soil. The well-known late Wisconsin episode of Scabland flooding and two older episodes of flooding are marked by regional stratigraphic unconformities that lie between pedostratigraphic units. One unconformity that underlies the Washtucna and Old Maid Coulee Soils was caused by a major episode of Scabland flooding that occurred before 35,000 yr B.P., probably during the early Wisconsin (80,000 to 65,000 yr B.P.). A possible second and older episode of flooding, recognized by an unconformity that underlies the Devils Canyon Soil, may be Illinoian (ca. 130,000 to 200,000 yr B.P.) in age.

Aminostratigraphic correlations of emergent Quaternary deposits along the U.S. Atlantic Coastal Plain have employed independent radiometric data, regional temperature history models, and assumptions regarding the nature of the preserved late Quaternary sea-level record on this passive margin. A substantial “aminostratigraphic offset” is required if regional aminozones are rigorously constrained by all available Th/U data. New insights regarding the relation of this offset to subsurface stratigraphy in the Cape Fear region of southeastern North Carolina can explain these conflicts as consequences of the highly incomplete post-Cretaceous depositional record of the region. Southward projection of theoretical aminostratigraphic correlation trends suggests that stage 5 correlative marine units are rarely preserved on the emergent portion of the Coastal Plain between Cape Lookout and central South Carolina and that samples of this age would be most frequently found in this region only as fragmentary (and/or reworked) deposits on the inner shelf or in the subsurface of modern barrier islands. If this hypothesis is correct, then the accuracy of several Th/U coral dates from the South Carolina Coastal Plain must be questioned, along with sea-level, tectonic, and paleoclimatic conclusions derived from these dates.

The Dead Sea is surrounded by chemical and detrital sediments that were deposited in its larger precursor lakes, Lake Samra and Lake Lisan. The sedimentary history of these lakes was recon-structed by means of 230Th/234U ages of 30 samples, mostly of argonite laminae, from 8 columnar sections up to 110 km apart. The general validity of the ages was demonstrated by subjecting them to tests of internal isotopic consistency, agreement with stratigraphic order, and concordance with 14C ages. In the south, only the part of the Samra Formation older than 170,000 yr is exposed, while the aragonite-detritus rhythmites found in the central and northern region are generally younger than 120,000 yr. The Lisan Formation started accumulating about 63,000 yr B.P., with the clay and aragonite beds in the south-central area reflecting a rise in water level to at least −280 m. The upper part of the Lisan Formation, the aragonite-rich White Cliff Member, started accumulating about 36,000 yr B.P. The lake probably reached its highest level sometime after this, based on the ages of Lisan sediments preserved in the southernmost reaches of the basin.

A 52-m core from Qidong at the tip of the Yangtze River (Chang Jiang) delta provides a history of sea-level change, deltaic development, and vegetational and climatic changes during the last 12,000 yr. About 12,000 yr ago, when sea level was about 60 m below the present level, the coring site was situated in the innermost part of the exposed continental shelf. The late-Pleistocene vegetation on the uplands of the Lower Yangtze River valley was a mixed forest of deciduous and broadleaved evergreen trees in which Betula, Ulmus, Tsuga, and Cupressaceae were slightly more abundant than at present. Abies and Picea were probably present as relict populations on mountains bordering the region. Rapidly rising sea level converted the Qidong area to a coastal or estuarine environment between 11,000 and 10,800 yr B.P., leading to widespread development of wetlands dominated by Gramineae and Cyperaceae. During the next 2500 yr coastal erosion due to marine transgression obliterated the sedimentary record. Deltaic sedimentation predominated from 8300 to 3800 yr B.P., as the Yangtze River delta prograded by successively building a series of estuarine sand bars. During the mid-Holocene, the climate was slightly warmer and more humid than at present, allowing the subtropical broadleaved evergreen trees to increase their populations. Pinus and Quercus became more abundant after 3800 yr B.P. in response to climatic cooling. The present deltaic plain in Qidong formed less than 200 yr ago.

Pleistocene alluvial terraces of the nonglaciated Ba River drainage basin on the north flank of the Qinling Shan are capped by a succession of loess units and paleosols that correlate with the standard marine isotope chronology and are used to date the subjacent alluvial gravels. Alluvial fills were deposited during isotope stages 2, 6, 8, 12, and 16, whereas terracing occurred during interglacial stages 1, 5, 7, 11, and 15. The apparent absence of terraces dating to stage 14 and stage 4 may be due to the lesser intensity of these glaciations compared to that of stage 2, although disruption of the alluvial regime by local tectonism is a likely alternative for the lack of a stage 4 terrace. A stage 10 terrace was not positively identified from available exposures; its possible absence could be related to post-stage 12 uplift. Aggradational episodes correlate with glaciations and loess deposition, whereas degradational episodes correlate with interglaciations or interstades and soil formation, implying that climate is the primary control on Quaternary paleohydrology. This in turn points to variations in the Earth's orbital geometry as the major factor that modulates both climate and, ultimately, the fluvial system in the Qinling Shan. In this region, glaciations were dominated by a cold, dry winter monsoon climate, whereas during interstades and interglaciations a warmer and wetter climate prevailed, implying strengthening of the summer monsoon. Both the loess/paleosol and the alluvial records are consistent with climate-model simulations spanning the last 18,000 yr that show a change from cold, dry conditions during the last glacial maximum to a climate warmer and wetter than present during the first half of the Holocene.

In central New York, deformation of salt of the Salina Formation by the Laurentide Ice Sheet is shown by a thinning of the salt stratum in and near the outcrop zone where it was overridden by the ice sheet and by a thickening down-dip near the glacial limit. Less definitive suggestions of deformation by glacial overriding of the Salina outcrop zone in northeastern Ohio are seen in salt-cored anticlines in the vicinity of the glacial limit. In the lower peninsula of Michigan, isopachs of salt curve around Saginaw Bay, which suggests that the salt was displaced southwestward, away from the bay, by pressure of the Saginaw lobe of the ice sheet. All but one of the major Paleozoic salt deposits of North America lie in a zone that girdles two-thirds of the Canadian Shield from where glacial erosion removed all but occasional outliers of Phanerozoic strata.

Paleomagnetic records are used to correlate sedimentary sequences from pluvial Lakes Chewaucan and Russell in the western Great Basin. This correlation is the basis for age control in the relatively poorly dated sequence from Lake Chewaucan. The resulting chronology supports a lack of sedimentation in Lake Chewaucan during the interval 27,400 to 23,200 yr B.P., an assertion supported by the presence of a lag deposit at the corresponding stratigraphic horizon. Because the Lake Chewaucan outcrop (near Summer Lake, Oregon) is near the bottom of the lake basin, we conclude that Lake Chewaucan was at a lowstand during this time interval. The Chewaucan lowstand is coeval with the lowstand accompanying the Wizard's Beach Recession (isotope stage 3) previously seen in the geologic record from nearby pluvial Lake Lahontan. The ages of six tephra layers, including the Trego Hot Springs tephra, were also estimated using the paleomagnetic correlation. Together, the new age of the Trego Hot Springs tephra (21,800 yr B.P.) and the lake surface level prehistory of Lake Chewaucan imply a revised model for the lake surface level prehistory of Lake Lahontan. The revised model includes a longer duration for the Wizard's Beach Recession and the occurrence of a younger lowstand of short duration soon after the lowstand corresponding to the Wizard's Beach Recession.

On western North Island, New Zealand, a record of climatic change during the last glaciation is preserved in a terrestrial coverbed sequence of dominantly andesitic provenance. Here, a succession of five loess-like Andisol units postdates the global high sea-level stand of oxygen isotope substage 5e (<125,000 yr). Tephra erupted from western and central North Island allow precise chronological correlation of the loess-like units. Aerosolic quartz additions determined by quantitative X-ray diffraction (XRD) record two major peaks that correlate with oxygen isotope stages 2 and 4. The most likely source of quartz-rich dust at these times is the surrounding continental shelf, then exposed by low sea level; however, quartz of Australian provenance may also be represented. This study provides the first confirmation from the terrestrial New Zealand record that rates of atmospherically transported particles increase during glacial stages.

The radial-growth patterns of white spruce were studied on a number of trees growing in subarctic dunes along the eastern coast of Hudson Bay to calculate the rates of accumulation, erosion, and migration of cold-climate sand dunes. The average rate of sand accumulation in sheltered dunes (forest sites) was 2.5 to 3.3 cm/yr, which is two to three times lower than in highly exposed dunes with a rate of sedimentation of 7.65 cm/yr. The average erosion rate was 1.4–1.7 cm/yr, about two times lower than the accumulation rate. The migration rate of sheltered dunes was 18 to 30 cm/yr, three to five times lower than for an exposed dune which advanced at a speed of 74 cm/yr. This migration rate established for highly exposed dunes in the Subarctic with tree-ring methods is about 10 times lower than that established for a barchan in the Sahara with other methods.

Stratigraphic, sedimentologic, and pedologic studies of beach ridge and lacustrine deposits indicate that up to five times during the Holocene, shallow lakes covered Silver Lake playa in southeastern California for periods of years to decades. The two youngest lacustrine events (at about 390 ± 90 yr B. P. and 3620 ± 70 yr B. P.) coincide with the early and late Neoglacial episodes of North America. Increasing evidence in recent years from other nonglaciated areas leads us to conclude that the effects of these climatic episodes were much more widespread than previously thought. The climate during these episodes was characterized by an increased frequency of winter storms in the southwestern United States, causing wetter conditions that affected diverse, hyperarid environments in the Mojave Desert and adjacent regions. We propose that this wide areal coverage was caused by large-scale, winter atmospheric circulation patterns, which are probably related to changes in sea-surface temperatures and oceanic circulation in the eastern North Pacific Ocean.

This study evaluates obsidian-hydration dating in postglacial fluvial terraces cut into an outwash plain near West Yellowstone, Montana. Fluvial transport fractures obsidian grains. However, some old hydration rinds may be preserved, thus, a grain may record several fracturing events. The most recent fracturing event at West Yellowstone is recorded in surface sediments from all of the terraces, which were cut in a shorter period of time than the technique can discern. They formed about 19,000 ± 1000 yr ago, using published hydration-rate estimates and a mean rind thickness of 6.34 ± 0.14 μm (1 SE). Alternatively, the application of published hydration-rate constants for the Obsidian Cliff flow with an estimated effective hydration temperature of 1.4°C yield an age of 24,400 ± 1100 yr (1 SE). Thicker rinds record fracturing during Bull Lake glaciation and cooling cracks from the emplacement of several source flows. Much of the observed spread in rind thicknesses (6.34 ± 1.69 μm: 1 SD) is probably the result of chemically induced variations in hydration rate. Terrace ages based on a single rind would range from 13,000 to 39,000 yr (±1 SD). Therefore, it is inappropriate to (1) use a set of hydration-rate constants determined from a single sample to calculate ages for multiple artifacts or geological samples, (2) date an archaeological or geological event on the basis of a single artifact, or (3) generate a chronostratigraphy on the basis of individual dates as a function of depth. Multiple evaluations of source chemistry and hydration rates and multiple rind measurements are required to date fracturing events.

Peaty, tidal-marsh soils interbedded with estuarine mud in late Holocene stratigraphic sequences near Coos Bay, Oregon, may have been submerged and buried during great (M > 8) subduction earthquakes, smaller localized earthquakes, or by nontectonic processes. Radiocarbon dating might help distinguish among these alternatives by showing that soils at different sites were submerged at different times along this part of the Cascadia subduction zone. But comparison of conventional 14C ages for different materials from the same buried soils shows that they contain materials that differ in age by many hundreds of years. Errors in calibrated soil ages represent about the same length of time as recurrence times for submergence events (150–500 yr)—this similarity precludes using conventional 14C ages to distinguish buried soils along the southern Oregon coast. Accelerator mass spectrometer 14C ages of carefully selected macrofossils from the tops of peaty soils should provide more precise estimates of the times of submergence events.

High-resolution seismic profiles, as well as sedimentological and micropaleontological analyses of three cores, are used to reconstruct the environmental and sedimentological evolution of Preservation Inlet, the southernmost New Zealand fjord. Toward the end of the last glaciation, a series of deep oligotrophic lakes, impounded by shallow sills, occupied Preservation Inlet. Glaciers filled the headwater valleys and the vegetation consisted of a sparse cover of grass, scrub, and shrubs. The principal rivers discharged into the head of these lakes forming large sandy deltas, while finely laminated clays were deposited in the distal basins. As the climate started warming ca. 18,000 yr B.P., the snowline rose and glaciers retreated. Developing forests were dominated by Metrosideros and Cyathea fringed by coastal shrubland. In the now more productive lakes, a rich freshwater fauna developed, sedimentation rates increased, and organic mud accumulated. Under rapidly rising sea level, between 9500 and ca. 8000 yr B.P., the sills enclosing the lakes were successively overtopped. Marine water intruded into the fjord basins and flooded the deltas at the head of the lakes. By 6500 yr B.P. sea level had stabilized and the fjord assumed its present condition. Shrubs decreased in abundance and forests dominated by Weinmannia and Dacrydium cupressinum then developed. A forest dominated by Nothofagus fusca spread between 2000 and 1500 yr B.P., indicating a cooler climate. In Preservation Inlet and other New Zealand fjords, eustatic sea-level rise has been greater than isostatic rebound or tectonic uplift. Coastal inundation has resulted in a transgressional sequence from a limnic to marine environment. This contrasts with fjord coasts of the northern hemisphere where isostatic rebound has produced coastal emergence, or coastal emergence followed by submergence.

The accuracy and precision of 137Cs and 14C for dating post-1950 alluvial deposits were evaluated for deposits from known floods on two rivers in Arizona. The presence of 137Cs reliably indicates that deposition occurred after intensive above-ground nuclear testing was initiated around 1950. There was a positive correlation between the measured level of 137Cs activity and the clay content of the sediments, although 137Cs was detected even in sandy flood sediments with low clay content. 137Cs is a valuable dating tool in arid environments where organic materials for 14C or tree-ring dating are scarce and observational records are limited. The 14C activity measured in different types of fine organic detritus yielded dates within 1 to 8 yr of a 1980 flood deposit, and the accuracy was species-dependent. However, undifferentiated mixtures of fine organic materials from several post-bomb deposits of various ages repeatedly yielded dates between 1958 and 1962, and detrital charcoal yielded a date range of 1676–1939. In semiarid environments, the residence time of most types of organic debris precludes accurate annual resolution of post-bomb 14C dates.

Superimposed phases of stromatolite buildup are observed on the eastern margin of Lake Manyara and depict a paleoshoreline at about 20 m above modern lake level. Radiocarbon and Th/U measurements permitted the dating of the last two phases of stromatolite formation at ca. 90,000 yr and between 27,000 and 23,000 yr B.P., respectively. The Th/U chronology is based on the decay of a strong 230Th-excess (over 234U) inherited with the detrital particles cemented into the stromatolites. The various generations of stromatolites show comparable stable carbon and oxygen isotope contents and are located at the same paleolake stabilization levels. This indicates that stringent hydrological conditions are necessary for the development of the encrusting benthic microbial communities responsible for stromatolite formation. A comparison with similar stromatolitic units from the nearby Lake Natron-Lake Magadi basin shows that such conditions occurred during only a few of the late Quaternary humid episodes known in eastern Africa and that they are different in each basin. Stromatolites do not necessarily represent all high lake levels that Lake Manyara experienced during the late Pleistocene and Holocene.

A 7.86-m sediment core from Swamp Lake in Yosemite National Park, California, provides a continuous record of environmental change over the last ca. 16,000 yr, as inferred from pollen, macrofossil, and microscopic charcoal analyses. The core stratigraphy documents late Wisconsin (Tioga stage) deglaciation between >16,000 and 13,700 yr B.P., approximately 6000-3500 yr earlier than higher-elevation Sierra Nevada records. The core includes five volcanic ash layers, chemically identified as four Mono Craters ashes and the Tsoyawata ash (Mt. Mazama, Oregon). The fossil record shows that herbs and sagebrush dominated the glacial environment at Swamp Lake. By 12,000 yr B.P., a mixed conifer forest composed of high- and mid-elevation conifers grew around the lake, suggesting a cool, wet late-glacial environment. The modern Sierra montane forest did not become established until ca. 10,400 yr B.P., when maximum charcoal concentrations and minimum fir pollen percentages indicate an early Holocene xeric period. The record suggests that a cooling trend began ca. 6500 yr B.P. and persisted until ca. 3700 yr B.P. when the modern climatic regime was established.

The Richmond River Valley of northern N.S.W. contains a late Pleistocene succession dating back to approximately 250,000 yr B.P. Dinoflagellate and spore-pollen assemblages from the lowest interval, the lower “Dungarubba Clay” of Drury (1982), indicate deposition in a restricted estuarine environment at approximately 250,000 yr. Deposition in the overlying interval, the upper “Dungarubba Clay” and “Gundurimba Clay”, at approximately 120,000 yr B.P., began in a restricted estuary, but rising sea level caused inundation and deposition in a more open, marine-dominated environment. Dinoflagellate cyst assemblages from the last interglaciation (stage 5) are interpreted by analogy with those from the morphologically similar, modern Broken Bay, N.S.W. They are indicative of an open, marine-dominated environment and imply that barrier formation in the Richmond River Valley, and possibly elsewhere in northern N.S.W., did not commence until after the initial postglacial transgression. Synchronous changes in sea level and rainforest development suggest that there was no significant time lag between climate and sea-level change.

High-resolution shallow seismic profiles from the Ventura shelf reveal two contrasting, structurally controlled styles of deposition. On the northern shelf, folded Plio-Pleistocene strata comprising the south limb of the Ventura Avenue anticline are unconformably overlain by undeformed late Quaternary strata. In contrast, a thicker section of undeformed, prograding deltaic sequences preserved on the subsiding southern shelf contains a detailed record of at least five middle and late Quaternary glacioeustatic sea-level oscillations. A transgressive surface at −140 m in upper Pleistocene Saugus deposits is overlain by transgressive and highstand systems tracts. A shelf-margin systems tract subsequently developed as the sea level lowered, marking the end of Saugus deposition. Thereafter, over 50 m of post-Saugus upper Pleistocene sediment accumulated during a period of smaller-scale sea-level oscillations, followed by the development of a pair of marine terraces, presently at −50 and −33 m, and a large fluvial channel, now buried at −113 m. Correlations of Saugus and terrace marker horizons with the middle to late Quaternary paleosea-level curve suggest development of the Saugus marker horizons during oxygen isotopic stages 7-5e and formation of the terraces during Wisconsinan high sea levels. Maximum rates of vertical separation on the Oak Ridge and Pitas Point faults are estimated to be 0.13 and 0.30 m/1000 yr, respectively.

A core 120 km off the coast of southern Oregon was examined for changes in lithology, diatoms, and pollen over the past 30,000 yr. Primary production during the late Pleistocene was about half that of the Holocene. Evidence from diatoms and pollen indicates that summer upwelling was much weaker, implying an absence of strong northerly winds. Early Pliocene diatoms found throughout the late Pleistocene section were probably derived from diatomites east of the Cascades and provide evidence for strong easterly winds over a dry continental interior. The findings verify predictions of a climate model based on glacial maximum conditions. There is no compelling evidence for a climatic reversal corresponding to the European Younger Dryas chron. During the early Holocene (9000−7000 yr B.P.) there may have been years when winds were insufficiently strong to support upwelling, so that warm stratified waters lay closer to the coast.