The chronology of Late Weichselian to Holocene deglaciation of Kongsfjorden, west Spitsbergen has been reconstructed based on the geomorphic and stratigraphic record of ice retreat, relative sea-level relationships, and 14C dating of associated marine organic materials. The seaward extent of glacial drift and fjord bathymetry constrain a secure reconstruction for the ice sheet near the mouth of the fjord at ca. 13,000 yr B.P., but do not rule out the possibility that more extensive glaciation was achieved earlier during the Late Weichselian. Regional shoreline relations, rates of emergence, and radiocarbon dating of foraminifera deposited just above till indicate that deglaciation occurred in two steps: one beginning during or just prior to the Late Weichselian Marine Limit phase at 13,000–12,000 yr B.P. and another beginning at 10,000–9500 yr B.P. The fjord was completely deglaciated by 9440 ± 130 yr B.P. A period of stable relative sea level began 10,700 yr B.P. and ended between 10,000 and 9500 yr B.P., which we take to indicate renewed glacial loading during the Younger Dryas. Glacier readvance within Kongsfjorden at this time was diminutive, suggesting that most of the Younger Dryas ice-sheet growth was confined to the eastern part of the archipelago and/or to the Barents Shelf. The two-step deglaciation of Kongsfjorden occurred during intervals of accelerated global ice-sheet melting and rapid oceanic and atmospheric warming in more temperate latitudes of the circum-Atlantic region. This coincidence most likely resulted from contemporaneous increases in the poleward transport of oceanic heat.

The origins of the magnetic susceptibility variations of the Chinese loess and paleosols are explored by scanning and transmission electron microscopy of magnetic extracts, and by magnetic modeling of magnetic hysteresis data, to provide quantified estimates of the major magnetic components. Microscopy identifies several distinct size and shape characteristics in the magnetic carriers. Lithogenic magnetites, intact and abraded, dominate the coarse-grained magnetic fraction. The smallest of the coarse grains is ∼ 2 μm. The remaining magnetic materal is ultrafine in size, with two types of magnetite particles present. Type A particles strongly resemble soil magnetites produced by inorganic precipitation. Type B particles, which occur rarely, are probably bacterial in origin. Quantitative modeling of these magnetic assemblages shows that over 90% of the susceptibility variations is accounted for by the superparamagnetic magnetite component. Compared to the loess units, the paleosols are richer in magnetite, particularly of superparamagnetic size, and have a threefold higher ratio of magnetite to hematite. We identify pedogenic formation of magnetite as the major contributor to the loess magnetic record. Matching this record against other paleoclimatic records, we find an extremely high correlation with the standard 18O record. The Chinese loess sequences record a very high resolution magnetic stratigraphy directly related to changing climate.

Magnetic, hydrological, and geochemical data for 11 Scottish peat cores confirm previous reports that anthropogenic pollutants constitute the principal source of ferrimagnets to such environments, but indicate that postdepositional Fe3O4 persistence may be more limited than formerly envisaged. In all cores, the lower stratigraphic limit of magnetic enrichment coincides with the water table, below which negative mV values and enhanced ratios of Fe/Mn and Cu/Zn depict an abrupt transition to a strongly anoxic regime. Dissolution of Fe3O4 in the reducing, inundated zone is also signified by the more rapid down-core loss of magnetic properties in cores from topographically depressed “Sphagnum lawns” than in those from more elevated “Calluna hummocks”. As a consequence, the peak and cumulative SIRM values for cores from the former settings appear disproportionately low, and the magnetostratigraphic records from adjacent, but topographically variant sites are frequently asynchronous. In contrast to most marine and lacustrine sediments, dissolution of Fe3O4 in peat induces a refinement of magnetic mineral assemblages, highlighted by declining SIRM/ARM ratios with depth and a “hardening” of remanence near the base of the magnetically enhanced strata. SEM micrographs of magnetic extracts from one Scottish core indicate that this refinement results from preferential dissolution along crystallographic weaknesses within multidomain phases, ultimately causing their disintegration and the adoption of “single domain-type” magnetic behavior. Rates of dissolution are considerably greater than those reported for marine sediments, probably due to the unusually high dissolved sulfide concentrations in mire waters.

Late Pleistocene tephra deposits found from Sitka to Juneau and Lituya Bay are assigned to a source at the Mount Edgecumbe volcanic field, based on similarity of glass compositions to nearvent deposits and on thinning away from Kruzof Island. The sequence of near-vent layers is basaltic andesite and andesite at the base, rhyolite, and mixed dacite and rhyolite on top. The only breaks in the tephra sequence are two 1-mm-thick silt partings in a lake-sediment core, indicating a depositional interval from basaltic andesite to dacite of no more than about a millennium. Tephra deposits at sites >30 km from the vent are solely dacite and rhyolite and are 10,600 to 11,400 14C yr old based on interpretation of 18 radiocarbon ages, including 5 by accelerator mass spectrometry (AMS). Basaltic andesite and andesite deposits nearer the vent are as much as 12,000 yr old.

Discrepancy among radiocarbon ages of upland tephra deposits provisionally correlated as the same grainfall is resolvable within ±2 σ of analytical uncertainty. Comparison of bulk and AMS ages in one sediment core indicates a systematic bias of +600 to +1100 yr for the bulk ages; correlation of tephra deposits among upland and lacustrine sites implies an additional discrepancy of 200–400 yr between upland (relatively too young) and lacustrine ages. In any case, the Mount Edgecumbe tephra deposits are a widespread, latest Pleistocene stratigraphic marker that serves to emphasize the uncertainty in dating biogenic material from southeastern Alaska.

An oxygen isotope temperature record over a large part of the past 30,000 yr has been obtained for the southern Cape Province of South Africa by combining data on the isotopic composition of a stalagmite from a deep cave with that of a confined groundwater aquifer in the same region. Results show that temperatures during the last glacial maximum were on average about 6°C lower than those today, with peaks up to 7°C lower. A detailed analysis of the past 5000 yr suggests multiple fluctuations, with generally lower temperatures (1–2°C) around 4500 and 3000 yr B.P. The carbon isotopic composition of the stalagmite indicates significant vegetation changes between the late Pleistocene and today, and also during the second half of the Holocene.

The Colorado Creek mammoth locality in west-central Alaska contains the remains of two mammoths that were scavenged by carnivores. Sedimentologic interpretations of the reworked eolian deposits surrounding the bones, supplemented by 10 radiocarbon dates, indicate that the lower and upper mammoths died and were buried within separate, but superimposed, thaw gullies about 23,000 and 16,000 yr ago, respectively. From our results, we propose a polycyclic taphonomic model for thaw gullies governed largely by slope aspect, rather than regional climate, and in which mixing between faunal horizons is more likely than not. Variations in the rate of silt influx and the position of the permafrost table provide a paleoclimatic proxy record that can be correlated to other records in eastern Beringia.

Teeth of northern bog lemming, Mictomys borealis, are reported from Cathedral and Smith Creek caves and represent the first Wisconsin remains of the genus from the Great Basin. Specimens from Cathedral Cave, Snake Range, are associated with U-series ages of 24,000 to 15,000 yr B.P. Previous work with pollen and packrat middens, dating to the same age as the Mictomys, indicate that Smith Creek Canyon contained a riparian, locally mesic community, including Picea engelmannii (spruce), Betula sp. (birch), Cercocarpus sp. (mountain mahogany), and Artemisia sp. (sagebrush) among other species. Exposed canyon slopes and the adjacent valley apparently contained a more xeric steppe community including sagebrush and Chenopodiineae species; rocky outcrop permitted Pinus flexilis (limber pine) and P. longaeva (bristlecone pine) to grow adjacent to Lake Bonneville or low in the canyon. The region apparently experienced a dry climate (not necessarily drier than today); however, Smith Creek Canyon was fed by glacial meltwater from Mt. Moriah. The northern bog lemming probably lived only in the riparian community and possibly on the north-facing slope below Cathedral Cave. Few canyons of the Snake Range would have had the unusually mesic conditions found in Smith Creek Canyon.

Late Quaternary paleoenvironmental data for East Africa are derived primarily from montane sources and thus we know little about the changing composition of East African savannas. Four archaeological sites at Lukenya Hill in the savanna of the Athi-Kapiti Plains of Kenya that date to the last 40,000 yr preserve a large mammalian fauna. The prehistoric hunters concentrated on migratory ungulates and virtually ignored the resident inselberg ungulates throughout the occupation. Faunas of the last glacial maximum are dominated by an extinct small alcelaphine antelope. Arid-adapted ungulates are present that are regionally absent historically, and Pelorovis is present as well. The small alcelaphine and arid-adapted ungulates are absent in the Holocene deposits. This suggests that there was an expansion of dry savannas during the last glacial maximum. The last glacial maximum aridity, combined with a lack of pastoral-set fires, would have resulted in a vegetative mosaic distinct from the present. Dry woody growth and dry and/or tall grass, all of which are poor forage for ungulates, would have been common where seasonally moist short grasslands are presently extant. These conditions favored the large-bodied, highly hypsodont species in Africa that became extinct with the onset of wet conditions during the early Holocene.

The composite frequency of radiocarbon ages for Holocene marine fossils from beaches around Antarctica is significantly different than random. Variations in the frequency of fossil ages coincide with the timing of Holocene climate changes inferred from Antarctic ice cores, sub-Antarctic lakes, polar and alpine moraines, and sea level. Extant Antarctic marine species that occur as fossils in beaches may reflect coastal meltwater impacts associated with ice sheet marginal fluctuations that were circumpolar during the Holocene.