Late Pleistocene and Holocene glacial history of James Ross Island, Antarctic Peninsula

Studies of Quaternary glacial stratigraphy and morphology around the Antarctic Peninsula have shown that James Ross Island in the western Weddell Sea probably has the best occurrences of stratigraphic sections with dateable material in the region. The stratigraphy includes sections with indefinite radiocarbon age, and three separate aminozones can be recognized. Except for indications of an early deglaciation around c . 10,000 BP, the field evidence from northern James Ross Island suggests a glacial readvance around 7000 BP. It is concluded that the readvance reflects the combined effects of eustatic sea level rise and Holocene warming, leading to increased precipitation and a positive mass balance. The most recent large-scale deglaciation in the area took place around 6000–5000 BP. This confirms the evidence from lake sediments and moss banks in other parts of the Antarctic Peninsula region, which shows that, in most cases, the initiation of organic deposition took place after c . 6000 BP. The literature on the Holocene glacial and environmental history of the region is reviewed in light of the new field evidence.     

Ice-flow evolution of the Labrador Sector of the Laurentide Ice Sheet: a review,with new evidence from northern Quebec

Outcrops bearing stained, striated facets indicative of north–northeastward moving ice, truncated by unstained, striated facets indicative of various younger flows occur in the Caniapiscau area of north–central Quebec. This is the first report of differential staining of striated facets in the region. We propose that the staining occurred in an ice-free interval of probable interglacial age. This early ice flow probably occurred during ice retreat toward the Quebec highlands. Ice flow and glacial transport data from the southern Hudson Bay and James Bay basins indicate that the next major regional ice flow was toward the northwest and resulted from the expansion of an Early Wisconsinan glacier in the Quebec highlands. The northern part of this flow was diverted northwestward through Hudson Bay, and the southern part southwestward across James Bay, following a progressive counterclockwise rotation of flow. A zone of intersection (ZI) of two major glacier bedform systems, often referred to as the horseshoe-shaped Labrador Ice Divide, represents the head of a large northward convergent ice-flow system that extended to Ungava Bay and beyond. The Ungava flow propagated southward and captured the head of the opposing flow from an outflow centre located east of Caniaspiscau reservoir. We propose that this capture event correlates with the Gold Cove Advance in Ungava Bay and on Baffin Island at about 9900 14C yr BP. It is the largest advance of Quebec-Labrador ice yet proposed for the region. This correlation is based on the relative ice-flow chronology, accommodation of glacial lakes Naskaupi and McLean in the deglaciation sequence, the constraints placed on Last Glacial Maximum (LGM) ice configuration by the postglacial uplift pattern and events in the deep-sea record. Therefore, the Ungava ice-flow pattern is not a relict pre-Wisconsinan glacial landscape as recently proposed.     

A relative sea-level history for Arviat,Nunavut, and implications for Laurentide Ice Sheet thickness west of Hudson Bay

Thirty-six new and previously published radiocarbon dates constrain the relative sea-level history of Arviat on the west coast of Hudson Bay. As a result of glacial isostatic adjustment (GIA) following deglaciation, sea level fell rapidly from a high-stand of nearly 170 m elevation just after 8000 cal yr BP to 60 m elevation by the mid Holocene (~ 5200 cal yr BP). The rate of sea-level fall decreased in the mid and late Holocene, with sea level falling 30 m since 3000 cal yr BP. Several late Holocene sea-level measurements are interpreted to originate from the upper end of the tidal range and place tight constraints on sea level. A preliminary measurement of present-day vertical land motion obtained by repeat Global Positioning System (GPS) occupations indicates ongoing crustal uplift at Arviat of 9.3 ± 1.5 mm/yr, in close agreement with the crustal uplift rate inferred from the inferred sea-level curve. Predictions of numerical GIA models indicate that the new sea-level curve is best fit by a Laurentide Ice Sheet reconstruction with a last glacial maximum peak thickness of ~ 3.4 km. This is a 30–35% thickness reduction of the ICE-5G ice-sheet history west of Hudson Bay.     

Timing of the Northern Prince Gustav Ice Stream retreat and the deglaciation of northern James Ross Island,Antarctic Peninsula during the last glacial–interglacial transition

The Northern Prince Gustav Ice Stream located in Prince Gustav Channel, drained the northeastern portion of the Antarctic Peninsula Ice Sheet during the last glacial maximum. Here we present a chronology of its retreat based on in situ produced cosmogenic ¹⁰Be from erratic boulders at Cape Lachman, northern James Ross Island. Schmidt hammer testing was adopted to assess the weathering state of erratic boulders in order to better interpret excess cosmogenic ¹⁰Be from cumulative periods of pre-exposure or earlier release from the glacier. The weighted mean exposure age of five boulders based on Schmidt hammer data is 12.9 ± 1.2 ka representing the beginning of the deglaciation of lower-lying areas (< 60 m a.s.l.) of the northern James Ross Island, when Northern Prince Gustav Ice Stream split from the remaining James Ross Island ice cover. This age represents the minimum age of the transition from grounded ice stream to floating ice shelf in the middle continental shelf areas of the northern Prince Gustav Channel. The remaining ice cover located at higher elevations of northern James Ross Island retreated during the early Holocene due to gradual decay of terrestrial ice and increase of equilibrium line altitude. Schmidt hammer R-values are inversely correlated with ¹⁰Be exposure ages and could be used as a proxy for exposure history of individual granite boulders in this region and favour the hypothesis of earlier release of boulders with excessive ¹⁰Be concentrations from glacier directly at this site. These data provide evidences for an earlier deglaciation of northern James Ross Island when compared with other recently presented cosmogenic nuclide based deglaciation chronologies, but this timing coincides with rapid increase of atmospheric temperature in this marginal part of Antarctica.     

Holocene relative sea-level changes in the Qaqortoq area, southern Greenland

We present results from an investigation of relative sea-level changes in the Qaqortoq area in south Greenland from c. 11 000 cal. yr BP to the present. Isolation and transgression sequences from six lakes and two tidal basins have been identified using stratigraphical analyses, magnetic susceptibility, XRF and macrofossil analyses. Macrofossils and bulk sediments have been dated by AMS radiocarbon dating. Maximum and minimum altitudes for relative sea level are provided from two deglaciation and marine lagoon sequences. Initially, relative sea level fell rapidly and reached present-day level at ∼9000 cal. yr BP and continued falling until at least 8800 cal. yr BP. Between 8000 and 6000 cal. yr BP, sea level reached its lowest level of around 6-8 m below highest astronomical tide (h.a.t.). At around 3750 cal. yr BP, sea level has reached above 2.7 m below h.a.t. and continued to rise slowly, reaching the present-day level between ∼2000 cal. yr BP and the present. As in the Nanortalik area further south, initial isostatic rebound caused rapid isolation of low elevation basins in the Qaqortoq area. Distinct isolation contacts in the sediments are observed. The late Holocene transgression is less well defined and occurred over a longer time interval. The late Holocene sea-level rise implies reloading by advancing glaciers superimposed on the isostatic signal from the North American Ice Sheet. One consequence of this transgression is that settlements of Palaeo-Eskimo cultures from ∼4000 cal. yr BP may have been transgressed by the sea.     

A new Holocene relative sea‐level curve for Terra Nova Bay,Victoria Land,Antarctica

More than 100 radiocarbon dates of penguin guano and remains, shells and seal skin afford ages for raised beaches adjacent to Terra Nova Bay, Antarctica. These dates permit construction of a new relative sea‐level curve that bears on the timing of deglaciation. Recession of the Ross Sea ice‐sheet grounding line from Terra Nova Bay occurred no earlier than 7200 ¹⁴C yr (8000 cal. yr) BP. Retreat along the Victoria Land coast may have been rapid, possibly contributing to eustatic sea‐level rise centred at ca. 7600 cal. yr BP. The presence of a significant amount of ice remaining in the Ross Sea Embayment in Holocene time lessens the chance that Antarctica contributed significantly to meltwater pulse 1A several thousand years earlier. Copyright © 2004 John Wiley & Sons, Ltd.     

New relative sea-level curves for the southern Scott Coast,Antarctica: evidence for Holocene deglaciation of the western Ross Sea

Here we present new relative sea-level (RSL) curves developed from Holocene-aged raised beaches along the southern Scott Coast of the western Ross Sea, Antarctica. Fifty-four dates of marine shells, seal skin and elephant seal remains incorporated within raised beaches during storms afford a chronology for these curves. All of the curves show the same pattern and timing of RSL change within a small range of error. The best-dated curve suggests that final unloading of grounded Ross Sea ice from the southern Scott Coast and McMurdo Sound region occurred shortly before 6500 ¹⁴C yr BP. This age is consistent with glacial geological evidence that places deglaciation between 5730 and 8340 ¹⁴C yr BP. Our data strongly suggest that grounding-line retreat of the Ross Sea ice sheet southward through the McMurdo Sound region occurred in mid- and late Holocene time. If this is correct, then rising sea level could not have driven ice recession to the present-day grounding line on the Siple Coast, because global deglacial sea-level rise was essentially accomplished by mid-Holocene time. Copyright © 1999 John Wiley & Sons, Ltd.     

Volume of Antarctic Ice at the Last Glacial Maximum, and its impact on global sea level change

The volume of Antarctic ice at the Last Glacial Maximum is a key factor for calculating the past contribution of melting ice sheets to Late Pleistocene global sea level change. At present, there are large uncertainties in our knowledge of the extent and thickness of the formerly expanded Antarctic ice sheets, and in the timing of their release as meltwater into the world’s oceans. This paper reviews the four main approaches to determining former Antarctic ice volume, namely glacial geology, glacio-isostatic studies, glaciological modelling, and ice core analysis and attempts to reconcile these to give a ‘best estimate’ for ice volume. In the Ross Sea there was a major expansion of grounded ice at the Last Glacial Maximum, accounting for 2.3–3.2 m of global sea level. At some time in the Weddell Sea a large grounded ice sheet corresponding to c. 2.7 m of global sea level extended to the shelf break. However, this ice expansion has not yet been confidently dated and may not relate to the Last Glacial Maximum. Around East Antarctica there was thickening and advance offshore of ice in coastal regions. Ice core evidence suggests that the interior of East Antarctica was either close to its present elevation or thinner during the last glacial so the effect of East Antarctica on sea level depends on the net balance between marginal thickening and interior thinning. Suggested East Antarctic contributions vary from a 3–5.5 m lowering to a 0.64 m rise in global sea level. The Antarctic Peninsula ice sheet thickened and extended offshore at the Last Glacial Maximum, with a sea level equivalent contribution of c. 1.7 m. Thus, the Antarctic ice sheets accounted for between 6.1 and 13.1 m of global sea level fall at the Last Glacial Maximum. This is substantially less than has been suggested by most previous studies but the maximum figure matches well with one modelling estimate. The timing of Antarctic deglaciation is not well known. In the Ross Sea, terrestrial evidence suggests deglaciation may have begun at c. 13,000 yr BP¹ but that grounded ice persisted until c. 6,500 yr BP. Marine evidence suggests the western Ross Sea was deglaciated by c. 11,500 yr BP. Deglaciation of the Weddell Sea is poorly constrained. Grounded ice in the northern Antarctic Peninsula had retreated by c. 13,000 yr BP, and further south deglaciation occurred sometime prior to c. 6,000 yr BP. Many parts of coastal East Antarctica apparently escaped glaciation at the LGM, but in those areas that were ice-covered deglaciation was underway by 10,000 yr BP. With existing data, the timing of deglaciation shows no firm relation to northern hemisphere-driven sea level rise. This is probably due partly to lack of Antarctic dating evidence but also to the combined influence of several forcing mechanisms acting during deglaciation.     

Neogene glacigenic debris flows on James Ross Island,northern Antarctic Peninsula,and their implications for regional climate history

Detailed sedimentological and microtextural analyses of newly-discovered late Neogene diamictites and other coarse-grained facies, mostly sandwiched between hyaloclastite of the James Ross Island Volcanic Group and Cretaceous sandstone and mudstone, indicate deposition mainly by glacigenic debris flows. The deposits on James Ross Island (northern Antarctic Peninsula) constrain the depositional setting, ice–bed dynamics and regional palaeoclimate. The sequences on James Ross Island vary in age but are mainly late Miocene and Pliocene. Unlike Neogene sedimentary sequences elsewhere in Antarctica, those on James Ross Island are unusually well-dated by a combination of ⁴⁰Ar/³⁹Ar and ⁸⁷Sr/⁸⁶Sr analyses on fresh interbedded lavas and pristine bivalve molluscs, respectively. The Sr isotopic ages of the debris flows cluster around 4.74, 4.89, 5.44, 5.78, and 6.31 Ma and probably date relatively warm periods in the northern Antarctic Peninsula region, when the bivalves lived under ice-poor or seasonally ice-free conditions. The bivalves are often remarkably well-preserved, lack adhering lithified sediment and, in at least two locations, are large, mainly unfragmented and sometimes articulated, suggesting that they were alive immediately prior to their incorporation in subaqueous debris flows at the margins of an advancing glacier. These fossiliferous glacigenic debris flows signify episodes of ice expansion during relatively warm periods, or “interglacials”, of the late Miocene and Pliocene. The James Ross Island glacigenic sedimentary successions attain thicknesses of up to 150 m and extend over 4 km laterally. The high volume of glacigenic sediment delivery implicit in the James Ross Island successions indicates that a series of dynamic ice fronts crossed the region during the late Miocene and Pliocene epochs. Associated evidence, in the form of clast abrasion (including striations and faceting) and bedrock erosion, is indicative of basal sliding and subglacial sediment deformation active at the ice–bed interface and wet-based temperate or polythermal regimes, prior to remobilisation. The evidence further suggests two local ice caps on James Ross Island during the warm periods, as well as ice-overriding by the Antarctic Peninsula Ice Sheet from the west and northwest.     

Rapid isostatic rebound in southwestern Iceland at the end of the last glaciation

In connection with a new deglaciation concept for Iceland, implying an extensive glaciation during the Younger Dryas and the decay of the Icelandic inland ice sheet during the Preboreal, the history of relative sea-level changes on Iceland has been re-evaluated. New field data from the Reykjavik area, in Faxaflói Bay southwestern Iceland, were obtained in order to construct the first stratigraphically controlled curve of relative sea-level displacements for Iceland. The curve is constructed on the basis of radiocarbon-dated shells in raised marine deposits and on tephrostratigraphically controlled and radiocarbon-dated, submerged peat deposits. The curve suggests that a post-glacial relative sea-level change of about 45 m, from + 43 m a.s.1. to — 2 m a.s.l, occurred over a period of 900 ¹⁴C-years in the Reykjavik area between 10 300 BP and 9400 BP. The sea-level curve shows a shoreline displacement of c . 5 cm ¹⁴Cyr^-1 for that period. The mean absolute uplift rate is calculated to be 6.9 cm ¹⁴C yr^-1, which is about double the fastest rate reported from any other coastal North Atlantic site. Although this rapid uplift can probably be partly explained by a ¹⁴C plateau around the termination of the Pleistocene, it is more than likely controlled by rapid Preboreal deglaciation, together with low asthenosphere viscosities below Iceland and the release of hydroisostatic stresses in connection with the deglaciation.     

East Asian monsoon variation and climate changes in Jeju Island, Korea, during the latest Pleistocene to early Holocene

A 4.96-m-long sediment core from the Hanon paleo-maar in Jeju Island, Korea was studied to investigate the paleoclimatic change and East Asian monsoon variations during the latest Pleistocene to early Holocene (23,000-9000 cal yr BP). High-resolution TOC content, magnetic susceptibility, and major element composition data indicate that Jeju Island experienced the coldest climate around 18,000 cal yr BP, which corresponds to the last glacial maximum (LGM). Further, these multi-proxy data show an abrupt shift in climatic regime from cold and arid to warm and humid conditions at around 14,000 cal yr BP, which represents the commencement of the last major deglaciation. After the last major deglaciation, the TOC content decreased from 13,300 to 12,000 cal yr BP and from 11,500 to 9800 cal yr BP, thereby reflecting the weakening of the summer monsoon. The LGM in Jeju Island occurred later in comparison with the Chinese Loess Plateau. Such a disparity in climatic change events between central China and Jeju Island appears to be caused by the asynchrony between the coldest temperature event and the minimum precipitation event in central China and by the buffering effect of the Pacific Ocean.     

Saalian to Weichselian stratigraphy and sedimentation along the Lågjæren – Høgjæren escarpment,southwest Norway

Quaternary sediments along a profile crossing the southern part of the Jæren escarpment, southwestern Norway, have been investigated with regard to their glacial history and sea-level variations. Deposits from at least three glaciations and two ice-free periods between Oxygen Isotope Stage 6 and the Late Weichselian have been identified. Subglacial till directly overlain by a glaciomarine regressional succession indicates a deglaciation, and amino acid ratios in Elphidium excavatum between 0.083 and 0.118 date this event to Oxygen Isotope Stage 6. Sea-level dropped from 130 to below 110 m a.s.l. Subsequently, a short-lived ice advance deposited a marginal moraine and a sandur locally on the escarpment. Stratigraphical position and luminescence dates around 148 ka BP suggest deposition during the final stage 6 deglaciation. A Late Weichselian till covers most of the surface of Jæren. In addition to a well documented westerly ice flow, glaciotectonic indications of ice flow towards the north have been found. Ice flow directions and a hiatus between Oxygen Isotope Stage 6 and the Weichselian indicate enhanced erosion along the escarpment and the influence of a Norwegian Channel ice-stream. Copyright © 1999 John Wiley & Sons, Ltd.     

Reconstruction of late Quaternary sea-level change in southwestern British Columbia from sediments in isolation basins

Bracketing ages on marine—freshwater transitions in isolation basins extending from sea level to 100 m elevation on Lasqueti Island, and data from shallow marine cores and outcrops on eastern Vancouver Island, constrain late Pleistocene and Holocene sea-level change in the central Strait of Georgia. Relative sea level fell from 150 m elevation to about —15 m from 14000 cal. yr BP to 11 500 cal. yr BP. Basins at higher elevations exhibit abrupt changes in diatom assemblages at the marine-freshwater transition. At lower elevations an intervening brackish phase suggests slower rates of uplift. Relative sea level rose to about +1 m about 9000 cal. yr BP to 8500 cal. yr BP, and then slowly fell to the modern datum. The mean rate of glacio-isostatic rebound in the first millennium after deglaciation was about 0.11 in a ^-1, similar to the peak rate at the centres of the former Laurentide and Fennoscandian ice complexes. The latter feature smooth, exponential-style declines in sea level up to the present day, whereas in the study area the uplift rate dropped to less than one-tenth of its initial value in only about 2500 years. Slower, more deeply seated isostatic recovery generated residual uplift rates of <0.01 m a^-1 in the early Holocene after the late-Pleistocene wasting of the Cordilleran ice sheet.     

Rates of Holocene isostatic uplift and relative sea-level lowering of the Baltic in SW Finland based on studies of isolation contacts

Southwestern Finland was covered by the Weichselian ice sheet and experienced rapid glacio-isostatic rebound after early Holocene deglaciation. The present mean overall apparent uplift rate is of the order of 4-5 mm/yr, but immediately after deglaciation the rate of crustal rebound was several times higher. Concurrently with land uplift, relative sea level in the Baltic basin during the past more than 8000 years was also strongly affected by the eustatic changes in sea level. There is ample evidence from earlier studies that during the early Litorina Sea stage on the southeastern coast of Finland around 7000 yr BP (7800 cal. yr BP), the rise in sea level exceeded the rate of land uplift, resulting in a short-lived transgression. Because of a higher rate of uplift, the transgression was even more short-lived or of negligible magnitude in the southwestern part of coastal Finland, but even in this latter case a slowing down in the rate of regression can still be detected. We used evidence from isolation basins to obtain a set of 71 ¹⁴C dates, and over 30 new sea-level index points. The age-elevation data, obtained from lakes in two different areas and located between c. 64 m and 1.5 m above present sea level, display a high degree of internal consistency. This suggests that the dates are reliable, even though most of them were based on bulk sediment samples. The two relative sea-level curves confirm the established model of relatively gradually decreasing rates of relative sea-level lowering since c. 6100 yr BP (7000 cal. yr BP) and clearly indicate that the more northerly of the two study areas experienced the higher rate of glacio-isostatic recovery. In the southerly study area, changes in diatom assemblages and lithostratigraphy suggest that during the early Litorina Sea stage (8300-7600 cal. yr BP) eustatic sea-level rise exceeded land uplift for hundreds of years. Evidence for this transgression was discovered in a lake with a basin threshold at an elevation of 41 m above sea level, which is markedly higher than any previously known site with evidence for the Litorina transgression in Finland. We also discuss evidence for subsequent short-term fluctuations superimposed on the main trends of relative sea-level changes.     

Late Quaternary history of Washington Land, North Greenland

During the last glacial stage, Washington Land in western North Greenland was probably completely inundated by the Greenland Ice Sheet. The oldest shell dates from raised marine deposits that provide minimum ages for the last deglaciation are 9300 cal. yr BP (northern Washington Land) and 7600 cal. yr BP (SW Washington Land). These dates indicate that Washington Land, which borders the central part of Nares Strait separating Greenland from Ellesmere Island in Canada, did not become free of glacier ice until well into the Holocene. The elevation of the marine limit falls from 110 m a.s.l. in the north to 60 m a.s.l. in the southwest. The recession was followed by readvance of glaciers in the late Holocene, and the youngest shell date from Neoglacial lateral moraines north of Humboldt Gletscher is 600 cal. yr BP. Since the Neoglacial maximum, probably around 100 years ago, glaciers have receded. The Holocene marine assemblages comprise a few southern extralimital records, notably of Chlamys islandica dated to 7300 cal. yr BP. Musk ox and reindeer disappeared from Washington Land recently, perhaps in connection with the cold period that culminated about 100 years ago.     

The mid-Holocene sea-level highstand at Bogenfels Pan on the southwest coast of Namibia

The radiocarbon ages of mollusc shells from the Bogenfels Pan on the hyper arid southern coast of Namibia provide constraints on the Holocene evolution of sea level and, in particular, the mid-Holocene highstand. The Bogenfels Pan was flooded to depths of 3 m above mean sea level (amsl) to form a large subtidal lagoon from 7300 to 6500 calibrated radiocarbon years before present (cal yr BP). The mollusc assemblage of the wave sheltered lagoon includes Nassarius plicatellus, Lutraria lutraria, and the bivalves Solen capensis and Gastrana matadoa, both of which no longer live along the wave-dominated southern Namibian coast. The radiocarbon ages of mollusc shell from a gravely beach deposit exposed in a diamond exploration trench indicate that sea level fell to near or 1 m below its present-day position between 6500 and 4900 cal yr BP. The rapid emergence of the pan between 6500 and 4900 cal yr BP exceeds that predicted by glacio-isostatic models and may indicate a 3-m eustatic lowering of sea level. The beach deposits at Bogenfels indicate that sea level rose to 1 m amsl between 4800 and 4600 cal yr BP and then fell briefly between 4600 and 4200 cal yr BP before returning to 1 m amsl. Since 4200 cal yr BP sea level has remained within one meter of the present-day level and the beach at Bogenfels has prograded seaward from the delayed arrival of sand by longshore drift from the Orange River. A 6200 cal yr BP coastal midden and a 600 cal yr BP midden 1.7 km from the coast indicate sporadic human utilization of the area. The results of this study are consistent with previous studies and help to refine the Holocene sea-level record for southern Africa.     

Late Weichselian deglacial history of Disko Bugt, West Greenland, and the dynamics of the Jakobshavns Isbrae ice stream

New relative sea-level (RSL) data from Disko Bugt, a large marine embayment in West Greenland, are used to examine the deglacial history of the Jakobshavns Isbrae ice stream. RSL data show rapid deglaciation after 10.3 ka cal. yr BP. Once deglaciation began, a bedrock high in the west of the bay exerted no discernible influence on the deglacial chronology. Following initial rapid retreat, ice stream recession slowed as it approached the eastern shores of the bay. Seabed elevations increase here and the ice stream terminus lingered for several thousand years before retreating into the narrow bedrock-confined Jakobshavns Isfjord. The seabed topography of Disko Bugt includes several deep channels which probably record the former course of the ice stream. Using a simple water depth/calving velocity relationship it is estimated that the maximum calving velocity on deglaciation was c. 4.8 km a^-1. This is less than the present rate (6–7 km a^-1), although ice discharge was two to four times that observed today. Initiation of rapid ice stream retreat was probably caused by ice stream thinning and increased surface melting. A critical point in time was the retreat of the ice stream from shallow continental shelf waters ( c. 400 m) into the deep bedrock trough (>800 m) which marks the entrance to Disko Bugt.     

Post-glacial relative sea level, isostasy, and glacial history in Icy Strait, Southeast Alaska, USA

We use the radiocarbon ages of marine shells and terrestrial vegetation to reconstruct relative sea level (RSL) history in northern Southeast Alaska. RSL fell below its present level around 13,900 cal yr BP, suggesting regional deglaciation was complete by then. RSL stayed at least several meters below modern levels until the mid-Holocene, when it began a fluctuating rise that probably tracked isostatic depression and rebound caused by varying ice loads in nearby Glacier Bay. This fluctuating RSL rise likely reflects the episodic but progressive advance of ice in Glacier Bay that started around 6000 cal yr BP. After that time, RSL low stands probably signaled minor episodes of glacier retreat/thinning that triggered isostatic rebound and land uplift. Progressive, down-fjord advance of the Glacier Bay glacier during the late Holocene is consistent with the main driver of this glacial system being the dynamics of its terminus rather than climate change directly. Only after the glacier reached an exposed position protruding into Icy Strait ca. AD 1750, did its terminus succumb - a century before the climate changes that marked the end of the Little Ice Age - to the catastrophic retreat that triggered the rapid isostatic rebound and RSL fall occurring today in Icy Strait.     

An early Holocene Greenland whale from Melville Bugt, Greenland

Radiocarbon age determination of a Greenland whale (Balaena mysticetus) vertebra from Melville Bugt in northwestern Greenland yields an age of 9259-8989 cal yr BP. The margin of the Greenland Ice Sheet in Melville Bugt was situated behind its AD 1950-2000 position in the early Holocene, at a similar position to that being reached following rapid retreat in recent years. Such an early deglaciation of areas close to the Greenland Ice Sheet is unusual. This probably reflects the unique glaciological setting resulting from the narrow fringe of ice-free islands and peninsulas and offshore waters with deep areas that characterize this part of Greenland. The timing of Greenland Ice Sheet retreat to its present margin varies significantly around Greenland.     

Global glacial ice volume and Last Glacial Maximum duration from an extended Barbados sea level record

Fundamental characteristics of the climate system during the most recent precessional cycle of the Earth's orbit around the Sun consist of the final expansion of land ice to its maximum extent, the subsequent episode of deglaciation, and the variations of global sea level that accompanied these events. In order to address the important issue of the variation of continental ice volume and related changes in global sea level through the late glacial period, we employ an extended set of observations of the pre-glacial and postglacial history of sea-level rise at the island of Barbados, together with a refined model of continental deglaciation and an accurate methodology for the prediction of postglacial sea-level change. Although our results provide unambiguous evidence that the post LGM rise of eustatic sea-level was very close to the widely supported estimate of 120 m, the data also provide evidence that LGM must have occurred 26,000 years ago, approximately 5000 yr earlier than the usually assumed age.