A numerical ice-sheet model was used to reconstruct the Late Weichselian glaciation of the Eurasian High Arctic, between Franz Josef Land and Severnaya Zemlya. An ice sheet was developed over the entire Eurasian High Arctic so that ice flow from the central Barents and Kara seas toward the northern Russian Arctic could be accounted for. An inverse approach to modeling was utilized, where ice-sheet results were forced to be compatible with geological information indicating ice-free conditions over the Taymyr Peninsula during the Late Weichselian. The model indicates complete glaciation of the Barents and Kara seas and predicts a “maximum-sized” ice sheet for the Late Weichselian Russian High Arctic. In this scenario, full-glacial conditions are characterized by a 1500-m-thick ice mass over the Barents Sea, from which ice flowed to the north and west within several bathymetric troughs as large ice streams. In contrast to this reconstruction, a “minimum” model of glaciation involves restricted glaciation in the Kara Sea, where the ice thickness is only 300 m in the south and which is free of ice in the north across Severnaya Zemlya. Our maximum reconstruction is compatible with geological information that indicates complete glaciation of the Barents Sea. However, geological data from Severnaya Zemlya suggest our minimum model is more relevant further east. This, in turn, implies a strong paleoclimatic gradient to colder and drier conditions eastward across the Eurasian Arctic during the Late Weichselian.

Integrated sedimentological and micropaleontological (foraminifera, ostracods, pollens) investigations of a 173-m-long continuous core in the subsurface of the SE Po Plain reveal a cyclic facies distribution of continental to shallow-marine deposits. Radiometric and pollen data enable an age attribution of the study units to the late Middle Pleistocene–Holocene. Glacioeustasy appears to have exerted a major control on facies development, as documented by the close relationships between cyclic lithofacies patterns and the curves of eustatic fluctuation. Particularly, relative sea-level falls and lowstand phases were characterized by the development of alluvial sedimentation, whereas major transgressive pulses and subsequent highstands are reflected by abrupt shifts to paralic and shallow-marine deposits. This facies architecture is paralleled by a distinctive pollen distribution, reflecting the alternation of a glacial type of vegetation, dominated by Pinus and nonarboreal pollen types, with interglacial mixed deciduous broad-leaved forests.

We revise substantially the regional chronology of lake-level fluctuations from the late-glacial/early Holocene humid phase along a high altitude transect (3500 to 4500 m) between 18°S and 28°S in the Southwestern Altiplano of Northern Chile. Radiocarbon dates and 210Pb profiles for limnic and terrestrial materials allow us to estimate and justify reservoir correction values for conventional 14C dates. Our chronology suggests that the latest Pleistocene/early Holocene humid phase started between 13,000 and 12,000 14C yr B.P., and that maximum lake levels were reached between 10,800 and 9200 14C yr B.P. This is significantly younger than what has been established so far for the Titicaca–Uyuni Basin in Bolivia. The paleolakes disappeared sometime between 8400 and 8000 14C yr B.P. Our revised chronology agrees with the regional history of human occupation, and is broadly synchronous with vegetation changes in subtropical continental South America, and with the onset of wetland expansion in the northern hemisphere tropics.

Molluscan changes that occurred in China over the S2–L2 upper middle Pleistocene series may be caused by monsoon variation. Study of terrestrial mollusks from the loess sequence in Luochuan, gathered in ecological groups according to the moisture and temperature requirements of the identified species, indicates alternating strengthened summer and winter paleomonsoons between 130,000 and 244,000 yr. The four occurrences of species, currently distributed in S.E. China in the sequence, indicate that the climate conditions were warmer and wetter than today between about 242,000 and 233,000 yr, at about 210,000, 164,000, and 140,000 yr. The main occurrence of xerophilous taxa at about 180,000, 154,000, and 138,000 yr is interpreted as indicating a drier environment than today. Such suggestions are in agreement with other proxy data such as grain size distribution. Higher numbers of individuals in the different ecological groups from the S2–L2 sequence indicate more favorable general environmental conditions than in the youngest S1–L1. This may agree with an increase in the regional aridity, since 500,000 yr, deduced from the study of the eolian flux in the northwestern Pacific downwind from China.

High-resolution (∼≤100 yr sampling interval) Chilean pollen data between ∼10,000 and 60,000 cal yr B.P. exhibit systematic fluctuations in Subantarctic Parkland development. These variations are dominated by a 30,000–40,000-yr cycle similar to that in the Northern Hemisphere GISP2 δ18O data and other climatic records. Both Chilean and GISP2 data show oscillations in the 5000–12,000 and 1000–3000-yr period bands. The coherence is, however, generally low, and distribution of power spectra differs, with the dynamics of the pollen assemblage dominated by a combination of 12,000- and 5000-yr cycles. We suggest a preferential nonlinear response of Chilean vegetation to climatic forcing and interhemispheric differences as possible mechanisms underlying the resemblances and dissimilarities between both records.

Holocene glacial deposits in Sredniya Avacha headwaters are subdivided into three age groups (events A, B, and C) based upon geomorphic features, tephrochronology, and lichenometry. Tephras of Opala volcano (1400–1500 yr B.P.), Ksudach volcano (1700–1800 and 6000 yr B.P.), and Zavaritskiy volcano (2800 yr B.P.) are used as stratigraphic markers. Rhizocarpon geographicum (L.) DC and Rhizocarpon section Alpicola growth curves are established using tephrochronologically dated and historical surfaces. The age of event A (pre-Hypsitermal?) moraines is constrained by an age of 6000 yr B.P. for overlying Ksudach-2 tephra and an age of 7200 (?) lichenometric (L) yr B.P. Event B (Neoglaciation) had a multiple nature, with the most prominent advances at 4300–3500, 3300–2800, 2600–2100, 1800–1400, and 1300–1100 (L) yr B.P.; the culmination occurred before 2800 yr B.P., as suggested by a date for Zavaritskiy volcano tephra found on the glacial and outwash deposits. Less-extensive glacier advances of event C (“Little Ice Age”), occurred at 800–600, 500–200, 180–110, and 90–40 (L) yr B.P. The ELA depression was 200–250, 100–150, 30–70 m during the culmination of events A, B, and C, respectively.

New stable isotope analyses on molluscan shells from a long core drilled in the crater lake of Valle di Castiglione, near Rome, extended the investigated portion of the core to 37 m. The succession of δ18O‰ values in the core interval 37–2.3 m ranges from −2.8 to +6.9‰ with only six samples below 0‰ (PDB). These results point to arid climatic phases coupled with the high measured δ18O values of the biogenic carbonate. In contrast, depleted 18O samples correspond to wet climatic periods, in agreement with a strong evaporative control on the lake water isotopic composition. The 13C content of the shells shows sharp changes controlled by the dissolved inorganic carbon isotope budget. Isotopic data suggest that the whole body of water behaved as a closed system, thus resembling lacustrine systems located in arid and semiarid regions where hydrological control dominates the geochemical parameters.

Pleistocene and Holocene vegetation dynamics in the American tropics are inferred largely from pollen in continental lake sediments. Maritime influences may have moderated climate and vegetation changes on Caribbean islands. Stable isotope (δ18O) study of a 7.6-m core from Lake Miragoane, Haiti, provided a high-resolution record of changing evaporation/precipitation (E/P) since ∼10,300 14C yr B.P. The Miragoane pollen record documents climate influences and human impacts on vegetation in Hispaniola. The δ18O and pollen data near the base of the core indicate cool, dry conditions before ∼10,000 14C yr B.P. Lake Miragoane filled with water in the early Holocene as E/P declined and the freshwater aquifer rose. Despite increasing early Holocene moisture, shrubby, xeric vegetation persisted. Forest expanded ∼7000 14C yr B.P. in response to greater effective moisture and warming. The middle Holocene (∼7000–3200 14C yr B.P.) was characterized by high lake levels and greatest relative abundance of pollen from moist forest taxa. Climatic drying that began ∼3200 14C yr B.P. may have driven some mesophilic animal species to extinction. The pollen record of the last millennium reflects pre-Columbian (Taino) and European deforestation. Long-term, Holocene vegetation trends in southern Haiti are comparable to trends from continental, lowland circum-Caribbean sites, suggesting a common response to regional climate change.

Ice-sheet modeling tightly constrained by empirical studies provides an effective framework to reconstruct past climatic and environmental conditions. Scotland was severely affected by the abrupt climate change associated with the Younger Dryas Stade, during which an extensive ice sheet formed across the west highlands after a period of ice-free conditions. Here, a quasi-three-dimensional, time-dependent ice flow/mass-balance model is developed and applied to Scotland at 1 km resolution. The flow model is based on the driving stress approximation with an additional longitudinal correction term, essential at this scale of operation. Surface mass balance is driven by temperature and precipitation changes and further mass wastage is achieved through an empirically defined calving term. The ice dynamics and mass-balance components are coupled through the equation for mass continuity, which is integrated through time over a finite-difference grid which yields the geometric evolution of the ice sheet. Initial experiments reveal the model to be relatively insensitive to internal parameters but highly sensitive to mass balance. Furthermore, these experiments indicate that Scotland is readily susceptible to glaciation with large glaciers building up on the flanks of Ben Nevis after a temperature depression of 2.5°C, under present-day precipitation.

The Younger Dryas is modeled using a GRIP temperature series locally adjusted for amplitude and a systematic series of runs enables the isolation of the climate which best matches mapped ice limits. This “optimum-fit” configuration requires an annual temperature cooling of 8°C and the introduction of substantial west–east and south–north precipitation gradients of 40 and 50%, respectively, to the present-day regime. Under these conditions, a series of substantial independent regional ice centers develop in agreement with trimline studies and after 550 year the modeled ice sheet closely resembles the maximum limits as indicated by field mapping. However, modeled ice continues to expand beyond 550 yr, in conflict with the mapped ice limits which suggest a prolonged period of stability. This discrepancy may be explained by the onset of extreme aridity ca. 400 yr into the Stade associated with a southern migration of the Polar Front, leading to a reduction in atmospheric circulation which effectively starved the ice sheet of its moisture source, preventing further expansion. Introduction of an additional 20% reduction in precipitation to the “optimum-fit” regime after 350 yr brings the modeled ice sheet to equilibrium, substantiating this conclusion.

The Northern Hemisphere ice sheets decayed rapidly during deglacial phases of the ice-age cycle, producing meltwater fluxes that may have been of sufficient magnitude to perturb oceanic circulation. The continental record of ice-sheet history is more obscured during the growth and advance of the last great ice sheets, ca. 120,000–20,000 yr B.P., but ice cores tell of high-amplitude, millennial-scale climate fluctuations that prevailed throughout this period. These climatic excursions would have provoked significant fluctuation of ice-sheet margins and runoff variability whenever ice sheets extended to mid-latitudes, giving a complex pattern of freshwater delivery to the oceans. A model of continental surface hydrology is coupled with an ice-dynamics model simulating the last glacial cycle in North America. Meltwater discharged from ice sheets is either channeled down continental drainage pathways or stored temporarily in large systems of proglacial lakes that border the retreating ice-sheet margin. The coupled treatment provides quantitative estimates of the spatial and temporal patterns of freshwater flux to the continental margins. Results imply an intensified surface hydrological environment when ice sheets are present, despite a net decrease in precipitation during glacial periods. Diminished continental evaporation and high levels of meltwater production combine to give mid-latitude runoff values that are highly variable through the glacial cycle, but are two to three times in excess of modern river fluxes; drainage to the North Atlantic via the St. Lawrence, Hudson, and Mississippi River catchments averages 0.356 Sv for the period 60,000–10,000 yr B.P., compared to 0.122 Sv for the past 10,000 yr. High-amplitude meltwater pulses to the Gulf of Mexico, North Atlantic, and North Pacific occur throughout the glacial period, with ice-sheet geometry controlling intricate patterns of freshwater routing variability. Runoff from North America is staged in the final deglaciation, with a stepped sequence of pulses through the Mississippi, St. Lawrence, Arctic, and Hudson Strait drainages.

Optical luminescence dates for 19 samples from the degraded linear dune field of western Zambia indicate multiple periods of regional dune building for the late Quaternary, 32,000–27,000, 16,000–13,000, 10,000–8000, and 5000–4000 yr ago. These dates show that the last glacial maximum was not the only time when dune construction, commonly linked to marked aridity, occurred in central-southern Africa during the late Quaternary. Whereas rainfall significantly less than today's ca. 1400 mm yr−1 is a prerequisite for dune construction in the area, adequate sediment supply also determines dune construction and preservation, so that dune building cannot be simply and singularly linked to marked aridity. The Zambezi River system is proposed as an important source of dune sediments, with the nature of linear dune activity explaining why stacked sediments preserve several phases of dune formation. Chronologies of dune construction in western Zimbabwe and the southwest Kalahari are in broad agreement with our Zambian chronology and support a model of rainfall shifts along a SW–NE gradient, with some notable disparities. These are probably a function of interregional sediment supply differences, the number of samples used to delimit constructional periods, and the multicausal nature of forcing mechanisms.

Fourteen organic-rich sedimentary layers in the deposits at Quebrada de los Burros, in coastal southern Peru (Tacna department), lie between two debris-flow units, interpreted to result from El Niño events, at 8980 cal yr B.P. and after 3380 cal yr B.P., respectively. The accumulation of the fine-grained and low-energy sediments of this deposit during the mid-Holocene is incompatible with the occurrence of El Niño events in this region, as these would produce catastrophic flood deposits. The occurrence of organic-rich sediments and evidence of an enhancement of upwelling strength at this time imply the existence of a permanent water supply resulting from an increased condensation of fog at mid-altitudes. These results suggest a lower intensity and perhaps, a lower frequency of occurrence of the El Niño phenomenon during the mid-Holocene. It is precisely during this period that the most important human settlements are found at this site, probably indicating the presence of reliable supply of fresh water. The chronologies for wetlands in the central south altiplano are out of phase with those indicating increased soil moisture episodes on the coast, implying a long-term difference in climate between these two regions.

Initial interpretation of the sediments from the Burmester core (Eardley et al. (1973). Geological Society of America Bulletin 84, 211–216) indicated that 17 deep-lake cycles, separated by shallow-lake and soil-forming intervals, occurred in the Bonneville basin during the Brunhes Chron (the last 780 × 103 yr). Our re-examination of the core, along with new sedimentological, geochronological, and paleontological data, indicate that only four deep-lake cycles occurred during this period, apparently correlative with marine oxygen-isotope stages 2, 6, 12, and 16. This interpretation suggests that large lakes formed in the Bonneville basin only during the most extensive of the Northern Hemisphere glaciations.

Geological records have shown that the deserts east of the Helan Mountains in northern China were covered by grass during the Holocene Optimum, whereas during marine oxygen isotope stages 2 and 4 distribution of the deserts was almost the same as at present. The wide advance–retreat cycles of the deserts may have exerted an important control on grain-size changes in the loess of the Loess Plateau by altering the distance between the source and the accumulation zone of the loess. This challenges the widely accepted model that winter monsoon winds were the sole factor responsible for spatial and temporal changes in loess texture. To observe spatial changes in sedimentological characteristics of loess during the last glacial–interglacial cycle, the texture of loess was measured along a north–south transect of the Loess Plateau. This transect consists of nine loess sections, starting at Yulin in the transitional region between the Loess Plateau and the Mu Us Desert and ending at Weinan in the southernmost part of the Loess Plateau. Southward changes in sand (>63 μm) content along the transect suggest that variations in desert extent have indeed played a significant role in loess grain-size distributions, particularly in the northern part of the Loess Plateau. It is proposed that sand content (>63 μm%) of loess in the loess–desert transitional zone may be used as a proxy indicator for proximity to the desert margin.

Deposits of a transgressive-phase Lake Bonneville stillstand or oscillation are found just below the elevation of the regressive-phase Provo shoreline at numerous exposures throughout the Bonneville basin. Existence of these subProvo shoreline deposits provides a new explanation for the massive size of Provo depositional and erosional landforms, which can no longer be explained by a long stillstand at the Provo shoreline. Provo coastal landforms are large because they are superimposed on subProvo landforms. Results also help to clarify divergent interpretations regarding the relative age of the Provo shoreline and the number of times it was occupied by the water plane. Occupation of approximately the same level during both the transgressive and the regressive phase of Lake Bonneville may be coincidental, or it may indicate that a bedrock sill controlled outflow at subProvo as well as Provo time. Rise to the Bonneville level could have occurred after massive slope failure plugged the outlet pass.

Fluctuations of the piedmont lobe of Bering Glacier and its sublobe Steller Glacier over the past two millennia are reconstructed using 34 radiocarbon dates and tree-ring data from 16 sites across the glaciers' forelands. The general sequence of glacial activity is consistent with well-dated fluctuations of tidewater and land-terminating glaciers elsewhere along the Gulf of Alaska. Extensive forested areas along 25 km of the Bering ice margin were inundated by glacio-lacustrine and glacio-fluvial sediments during a probable ice advance shortly before 500 cal yr A.D. Regrowth of forests followed the retreating ice as early as the 7th century A.D., with frequent interruptions of tree growth due to outwash aggradation. Forests overrun by ice and buried in outwash indicate readvance about 1080 cal yr A.D. Retreat followed, with ice-free conditions maintained along the distal portions of the forefield until the early 17th century after which the ice advanced to within a few kilometers of its outer Neoglacial moraine. Ice reached this position after the mid-17th century and prior to 200 yr ago. Since the early 20th century, glacial retreat has been punctuated by periodic surges. The record from forests overrun by the nonsurging Steller Lobe shows that this western ice margin was advancing by 1250 A.D., reaching near its outer moraine after 1420 cal yr A.D. Since the late 19th century, the lobe has dominantly retreated.

The diatom stratigraphy of a 23-m sediment core from Lake Tilo, a maar lake in the Ethiopian Rift Valley, provides a 10,000-yr record of lake salinity and trophic status. Until 5500 14C yr B.P., the phytoplankton was dominated by Aulacoseira granulata, with only minor changes in the abundance of other diatoms; the lake was over 50 m deep, eutrophic, and oligosaline. At 5500 yr B.P., geothermal groundwater inflow, inferred from calcite and silica deposition rates, declined abruptly, and the lake became more oligotrophic, as indicated by a rapid rise in Cyclotella stelligera. About 4500 yr B.P., lake salinity began to increase, reaching approximately its present state ca. 2500 yr B.P., but with a temporary reversal to lower salinity at 4000–3500 yr B.P. The record shows no evidence of salinity increases equivalent to early Holocene low stands of the larger river-fed Rift Valley lakes, probably because of high rates of geothermal groundwater influx. It responded to reduced rainfall at 4500 yr B.P., when levels of the larger lakes also fell, because geothermal groundwater flux had diminished 1000 year earlier, independently of climate. Because geothermal groundwaters can form a significant proportion of a crater lake's hydrological budget and dominate its salinity budget, these results show that variable geothermal groundwater flux can override lake hydrochemical response to climate change. Palaeoclimatic interpretation of palaeosalinity proxies from the sediments of volcanic crater lakes should be approached with caution.

Near Mesquite Spring on the southern edge of the Soda Lake basin in the Mojave Desert, there is a shoreline of an ancient lake at an elevation of 340 m above sea level. At present, Soda Lake would overflow at 280 m; a lake surface at 340 m would extend ∼240 km northward, to the northern end of Death Valley. Shorelines and lacustrine deposits near the Salt Spring and Saddle Peak Hills, 75 km north of Mesquite Spring, are at ∼180 m; a lake surface at this elevation today would also extend to the northern end of Death Valley. The most prominent shoreline of the pluvial lake that occupied Death Valley during the Pleistocene, Lake Manly, is that of the Blackwelder stand which ended ∼120,000 yr ago. This shoreline is ∼90 m above sea level. The Mesquite Spring and Salt Spring Hills shorelines were probably formed by the Blackwelder stand and subsequently displaced with respect to one another, tectonically, due to transpression in the northeastern Mojave Desert and NW–SE extension across Death Valley. This tectonic regime would result in subsidence of Death Valley and the Salt Spring Hills relative to Mesquite Spring. A reconstruction suggests that the topography at the time of the Blackwelder stand would have had a sill near the level of the highest lake, and also one ∼20 m lower, corresponding to the next most prominent shoreline in Death Valley. Expansion of the lake over these sills would have increased evaporation, thus possibly stabilizing the lake level.

The oxygen isotope record of planktonic foraminifera from tropical core MD77194 (Eastern Arabian Sea) exhibits a clear two-step deglaciation with a brief δ18O transient event. In the tropics, this δ18O maximum could correspond to a cooling or to a change in the δ18O content of sea water. In this study, past sea-surface temperature (SST) and primary production (PP) are reconstructed from foraminiferal transfer functions and compared to values estimated from alkenone measurements. SST and PP records from both proxies indicate a 1.5–2.5°C deglacial warming, coupled with a PP decrease, and a 0.5–1°C cooling during the Younger Dryas (YD). A detailed comparison between independent micropaleontological and geochemical proxies helps us identify potential biases and thus strengthen the paleo-reconstructions.

Shoreline altitudes of several pluvial lakes in the western Great Basin of North America record successively smaller lakes from the early to the late Pleistocene. This decrease in lake size indicates a long-term drying trend in the regional climate that is not seen in global marine oxygen-isotope records. At +70 m above its late Pleistocene shoreline, Lake Lahontan in the early middle Pleistocene submerged some basins previously thought to have been isolated. Other basins known to contain records of older pluvial lakes that exceeded late Pleistocene levels include Columbus-Fish Lake (Lake Columbus-Rennie), Kobeh-Diamond (Lakes Jonathan and Diamond), Newark, Long (Lake Hubbs), and Clover. Very high stands of some of these lakes probably triggered overflows of previously internally drained basins, adding to the size of Lake Lahontan. Simple calculations based on differences in lake area suggest that the highest levels of these pluvial lakes required a regional increase in effective moisture by a factor of 1.2 to 3 relative to late Pleistocene pluvial amounts (assuming that effective moisture is directly proportional to the hydrologic index, or lake area/tributary basin area). These previously unknown lake levels reflect significant changes in climate, tectonics, and (or) drainage-basin configurations, and could have facilitated migration of aquatic species in the Great Basin.