Modelling the influence of glacial isostasy on Late Weichselian ice‐sheet growth in the Barents Sea

A fully integrated ice‐sheet and glacio‐isostatic numerical model was run in order to investigate the crustal response to ice loading during the Late Weichselian glaciation of the Barents Sea. The model was used to examine the hypothesis that relative reductions in water depth, caused by glacio‐isostatic uplift, may have aided ice growth from Scandinavia and High Arctic island archipelagos into the Barents Sea during the last glacial. Two experiments were designed in which the bedrock response to ice loading was examined: (i) complete and rapid glaciation of the Barents Sea when iceberg calving is curtailed except at the continental margin, and (ii) staged growth of ice in which ice sheets are allowed to ground at different water depths. Model results predict that glacially generated isostatic uplift, caused by an isostatic forebulge from loads on Scandinavia, Svalbard and other island archipelagos, affected the central Barents Sea during the early phase of glaciation. Isostatic uplift, combined with global sea‐level fall, is predicted to have reduced sea level in parts of the central Barents Sea by up to 200 m. This reduction would have been sufficient to raise the sea floor of the Central Bank into a subaerial position. Such sea‐floor emergence is conducive to the initiation of grounded ice growth in the central Barents Sea. The model indicates that, prior to its glaciation, the depth of the Central Deep would have been reduced from around 400 m to 200 m. Such uplift aided the migration of grounded ice from the central Barents Sea and Scandinavia into the Central Deep. We conclude that ice loading over Scandinavia and Arctic island archipelagos during the first stages of the Late Weichselian may have caused uplift within the central Barents Sea and aided the growth of ice across the entire Barents Shelf. Copyright © 2000 John Wiley & Sons, Ltd.     

Tunnel channels in southeast Alberta, Canada: : evidence for catastrophic channelized drainage

Tunnel channels in southeast Alberta are attributed to erosion by channelized, subglacial meltwater flows. An anabranching tunnel channel network dissects the preglacial drainage divide of the ancestral Milk River. Channel morphology and landform associations are used to evaluate competing hypotheses of tunnel channel formation. Mechanisms that invoke subaerial channel incision, direct glacial erosion or steady state, time-transgressive erosion at the ice margin cannot explain convex-up longitudinal channel profiles, anabranching channel networks or confinement to the preglacial drainage divide. Results conclude that the tunnel channel network in southeast Alberta represents late-stage erosion by a channelised subglacial flow of catastrophic dimensions. Interpretations for this tunnel channel network are in agreement with conclusions obtained for the regional subglacial landform assemblage.     

Salt-induced crestal faults control the formation of Quaternary tunnel valleys in the southern North Sea

Tunnel valleys are major features of glaciated margins and they enable meltwater expulsion from underneath a thick ice cover. Their formation is related to the erosion of subglacial sediments by overpressured meltwater and direct glacial erosion. Yet, the impact of pre-existing structures on their formation and morphology remains poorly known. High-quality 3D seismic data allowed the mapping of a large tunnel valley that eroded underlying preglacial delta deposits in the southern North Sea. The valley follows the N–S strike of crestal faults related to a Zechstein salt wall. A change in downstream tunnel valley orientation towards the SE accompanies a change in the strike direction of salt-induced faults. Fault offsets indicate important activity of crestal faults during the deposition of preglacial deltaic sediments. We propose that crestal faults facilitated tunnel valley erosion by acting as high-permeability pathways and allowing subglacial meltwater to reach low-permeability sediments in the underlying Neogene deltaic sequences, ultimately resulting in meltwater overpressure build-up and tunnel valley excavation. Active faults probably also weakened the near-surface sediment to allow a more efficient erosion of the glacial substrate. This control of substrate structures on tunnel valley morphology is considered as a primary factor in subglacial drainage pattern development in the study area.     

Constraints on the Late Weichselian ice sheet over the Barents Sea from observations of raised shorelines

Direct evidence for Late Weichselian grounded glacier ice over extensive areas of the Barents Sea is based largely on indirect observations, including elevations of old shorelines on Svalbard and arguments of isostatic rebound. Such isostatic models are discussed here for two cases representing maximum and minimum ice-sheet reconstructions. In the former model the ice extends over the Kara Sea, whereas in the latter the ice is limited to the Barents Sea and island archipelagos. Comparisons of predictions with observations from a number of areas, including Spitsbergen, Nordaustlandet, Edgeøya, Kong Karls Land, Franz Josef Land, Novaya Zemlya and Finnmark, support arguments for the existence of a large ice sheet over the region at the time of the last glacial maximum. This ice sheet is likely to have had the following characteristics, conclusions that are independent of assumptions made about the Earth's rheological parameters. (i) The maximum thickness of this ice was about 1500–2000 m with the centre of the load occurring to the south and east of Kong Karls Land. (ii) The ice sheet extended out to the western edge of the continental shelf and its maximum thickness over western Spitsbergen was about 800 m. (iii) To the north of Svalberg and Frans Josef Land the ice sheet extended out to the northern shelf edge. (iv) Retreat of the grounded ice across the southern Barents Sea occurred relatively early such that this region was largely ice free by about 15,000 BP. (v) By 12,000 BP the grounded ice had retreated to the northern archipelagos and was largely gone by 10,000 BP. (vi) The ice sheet may have extended to the Kara Sea but ice thicknesses were only a fraction of those proposed in those reconstructions where the maximum ice thickness is centered on Novaya Zemlya. Models for the palaeobathymetry for the Barents Sea at the time of the last glacial maximum indicate that large parts of the Barents Sea were either very shallow or above sea level, providing the opportunity for ice growth on the emerged plateaux, as well as on the islands, but only towards the end of the period of Fennoscandian ice sheet build-up.     

The Plio-Pleistocene glaciation of the Barents Sea–Svalbard region: a new model based on revised chronostratigraphy

Based on a revised chronostratigraphy, and compilation of borehole data from the Barents Sea continental margin, a coherent glaciation model is proposed for the Barents Sea ice sheet over the past 3.5 million years (Ma). Three phases of ice growth are suggested: (1) The initial build-up phase, covering mountainous regions and reaching the coastline/shelf edge in the northern Barents Sea during short-term glacial intensification, is concomitant with the onset of the Northern Hemisphere Glaciation (3.6–2.4 Ma). (2) A transitional growth phase (2.4–1.0 Ma), during which the ice sheet expanded towards the southern Barents Sea and reached the northwestern Kara Sea. This is inferred from step-wise decrease of Siberian river-supplied smectite-rich sediments, likely caused by ice sheet blockade and possibly reduced sea ice formation in the Kara Sea as well as glacigenic wedge growth along the northwestern Barents Sea margin hampering entrainment and transport of sea ice sediments to the Arctic–Atlantic gateway. (3) Finally, large-scale glaciation in the Barents Sea occurred after 1 Ma with repeated advances to the shelf edge. The timing is inferred from ice grounding on the Yermak Plateau at about 0.95 Ma, and higher frequencies of gravity-driven mass movements along the western Barents Sea margin associated with expansive glacial growth.     

Depositional environments and sea-level changes deduced from Middle Weichselian tidally influenced sediments, Arkhangelsk region, northwestern Russia

Deposits from a Middle Weichselian transgression, the Mezen Transgression, are found in coastal sections in the Mezen and Chyorskaya Bays, northwestern Russia. The marine event is bracketed between two ice advances from the Barents and Kara Sea shelves and dated by Optically Stimulated Luminescence (OSL) to around 60 kyr BP. The deposits represent a shallowing upward succession from offshore marine to intertidal coastal environments. Relative sea-level maximum was at least 40 m above the present owing to significant isostatic subsidence. The sedimentary record is dominated by shallow-marine, subtidal deposits bounded below by an erosion surface representing a downward shift in facies and above by subaerial exposure. The succession reflects deposition during forced regression due to isostatic uplift. A rapidly aggrading succession of subtidal deposits at one site suggests a relative sea-level rise or stillstand superimposed on the isostatically controlled sea-level fall. The rhythmic tidal deposits allow identification of semi-monthly to yearly cycles, providing an estimate of the sedimentation rate of 39 cm/year. This implies a high sediment yield and a rapid relative sea-level rise. We correlate this signal with the rapid eustatic sea-level rise at the end of OIS 4 known from deep-sea records.     

Modelling ice-sheet sensitivity to late weichselian environments in the svalbard-barents sea region

Ice-proximal sedimentological features from the northwestern Barents Sea suggest that this region was covered by a grounded ice sheet during the Late Weichselian. However, there is debate as to whether these sediments were deposited by the ice sheet at its maximum or a retreating ice sheet that had covered the whole Barents Sea. To examine the likelihood of total glaciation of the Late Weichselian Barents Sea, a numerical ice-sheet model was run using a range of environmental conditions. Total glaciation of the Barents Sea, originating solely from Svalbard and the northwestern Barents Sea, was not predicted even under extreme environmental conditions. Therefore, if the Barents Sea was completely covered by a grounded Late Weichselian ice sheet, then a mechanism (not accounted for within the glaciological model) by which grounded ice could have formed rapidly within the central Barents Sea, may have been active during the last glaciation. Such mechanisms include (i) grounded ice migration from nearby ice sheets in Scandinavia and the central Barents Sea, (ii) the processes of sea-ice-induced ice-shelf thickening and (iii) isostatic uplift of the central Barents Sea floor.     

Weichselian stratigraphy and palaeoenvironments at Bellsund, western Svalbard

A coastal cliff facing the ocean at the west coast of Spitsbergen has been studied, and seven formations of Weichselian and Holocene age have been identified. A reconstruction of the palaeoenvironment and glacial history shows that most of the sediments cover isotope stage 5. From the base of the section, the formation 1 and 2 tills show a regional glaciation that reached the continental shelf shortly after the Eemian. Formation 3 consists of glacimarine to marine sediments dated to 105,000–90,000 BP. Amino acid diagenesis indicates that they were deposited during a c . 10,000-year period of continuous isostatic depression, which indicates contemporaneous glacial loading in the Barents Sea. Foraminifera and molluscs show influx of Atlantic water masses along the west coast of Svalbard at the same time. Local glaciers advanced during the latter part of this period, probably due to the penetration of moist air masses, and deposited formation 4. A widespread weathering horizon shows that the glacial retreat was succeeded by subaerial conditions during the Middle Weichselian. Formation 5 is a till deposited during the Late Weichselian glacial maximum in this area. The glaciation was dominated by ice streams from a dome over southern Spitsbergen, and the last deglaciation of the outer coast is dated to 13,000 BP. A correlation of the events with other areas on Svalbard is discussed, and at least two periods of glaciation in the Barents Sea during the Weichselian are suggested.     

Glacigenic deposits of the Central Deep: a key to the Late Quaternary evolution of the eastern Barents Sea

Sparker and shallow drilling data indicate that the Quaternary deposits in the Central Deep of the Barents Sea are mainly composed of glacigenic sediments. They comprise basal till and proximal and distal glaciomarine sediments deposited during the last glacial cycle. Apparent glaciotectonic features imply strong glacial erosion of Mesozoic bedrock. The general ice movement is assumed to have been from off Novaya Zemlya and it is concluded that the whole eastern Barents Sea was covered by the Late Weichselian ice-sheet.     

Quaternary geology in Zanskar,NW Indian Himalaya: evidence for restricted glaciation and preglacial topography

Research into the Quaternary geology of the NW Himalaya has concentrated on the elucidation of the glacial sequence. However, whilst the main ranges of the Himalaya have been subjected to numerous glaciations and are now an obvious alpine glaciated terrain, much of the landscape in Zanskar and Ladakh is more equivocal and does not appear to have been glaciated during this time. These landscape facets may therefore have a much older origin and relate to preglacial events.In Zanskar, the main ice source in all glaciations was the strongly glaciated and still glacierized north slope of the main Himalaya. This ice then flowed generally northwards in the valleys of the Zanskar river and its tributaries leaving between them a landscape supporting only a few and scattered minor local glaciers. Evidence of early glaciation has been found on isolated valley-side remnants >200 m above the present rivers. Reconstruction of these preglacial valley cross profiles show them to be generally broad and shallow, with gentle slopes. This is in distinct contrast to the present major valley systems which can usually be divided into two parts—a lower unglaciated fluvially eroded section, such as the Lungnak (Tsarap Lingti Chu) Gorge and an upper broad glacial section, such as the Stod (Doda) valley.Down-valley extent of glaciation is defined by the upper ends of unglaciated fluvial gorges. Laterally, the glaciers were confined progressively to their valleys. Inevitably there is only evidence of successively smaller subsequent glaciations, but the tectonic uplift of the southern ranges may have been a factor in this forming an increasing barrier to the snow-bearing monsoon winds.     

Tracking the isostatic and eustatic signals in an end Ordovician–early Silurian deglacial sedimentary record (Algeria – Libya border)

To elucidate the signature of isostatic and eustatic signals during a deglaciation period in pre‐Pleistocene times is made difficult because very little dating can be done, and also because glacial erosion surfaces, subaerial unconformities and subsequent regressive or transgressive marine ravinement surfaces tend to amalgamate or erode the deglacial deposits. How and in what way can the rebound be interpreted from the stratigraphic record? This study proposes to examine deglacial deposits from Late‐Ordovician to Silurian outcrops at the Algeria–Libya border, in order to define the glacio–isostatic rebound and relative sea‐level changes during a deglaciation period. The studied succession developed at the edge and over a positive palaeo‐relief inherited from a prograding proglacial delta that forms a depocentre of glaciogenic deposits. The succession is divided into five subzones, which depend on the topography of this depocentre. Six facies associations were determined: restricted marine (Facies Association 1); tidal channels (Facies Association 2); tidal sand dunes (Facies Association 3); foreshore to upper shoreface (Facies Association 4); lower shoreface (Facies Association 5); and offshore shales (Facies Association 6). Stratigraphic correlations over the subzones support the understanding of the depositional chronology and associated sea‐level changes. Deepest marine domains record a forced regression of 40 m of sea‐level fall resulting from an uplift caused by a glacio‐isostatic rebound that outpaces the early transgression. The rebound is interpreted to result in a multi‐type surface, which is interpreted as a regressive surface of marine erosion in initially marine domains and as a subaerial unconformity surface in an initially subaerial domain. The transgressive deposits have developed above this surface, during the progressive flooding of the palaeo‐relief. Sedimentology and high‐resolution sequence stratigraphy allowed the delineation of a deglacial sequence and associated sea‐level changes curve for the studied succession. Estimates suggest a relatively short (<10 kyr) duration for the glacio‐isostatic uplift and a subsequent longer duration transgression (4 to 5 Myr).     

Rethinking Late Weichselian ice-sheet dynamics in coastal NW Svalbard

New marine geological evidence provides a better understanding of ice-sheet dynamics along the western margin of the last Svalbard/Barents Sea Ice Sheet. A suite of glacial sediments in the Kongsfjordrenna cross-shelf trough can be traced southwards to the shelf west of Prins Karls Forland. A prominent moraine system on the shelf shows minimum Late Weichselian ice extent, indicating that glacial ice also covered the coastal lowlands of northwest Svalbard. Our results suggest that the cross-shelf trough was filled by a fast-flowing ice stream, with sharp boundaries to dynamically less active ice on the adjacent shelves and strandflats. The latter glacial mode favoured the preservation of older geological records adjacent to the main pathway of the Kongsfjorden glacial system. We suggest that the same model may apply to the Late Weichselian glacier drainage along other fjords of northwest Svalbard, as well as the western margin of the Barents Ice Sheet. Such differences in glacier regime may explain the apparent contradictions between the marine and land geological record, and may also serve as a model for glaciation dynamics in other fjord regions.     

Pliocene and Quaternary regional uplift in western Turkey: the Gediz River terrace staircase and the volcanism at Kula

Along the upper reaches of the Gediz River in western Turkey, in the eastern part of the Aegean extensional province, the land surface has uplifted by 400 m since the Middle Pliocene. This uplift is revealed by progressive gorge incision, and its rate can be established because river terraces are capped by basalt flows that have been K–Ar and Ar–Ar dated. At present, the local uplift rate is 0.2 mm a⁻¹. Uplift at this rate began around the start of the Middle Pleistocene, following a span of time when the uplift was much slower. This was itself preceded by an earlier uplift phase, apparently in the late Late Pliocene and early Early Pleistocene, when the uplift rate was comparable to the present. The resulting regional uplift history resembles what is observed in other regions and is analogously interpreted as the isostatic response to changing rates of surface processes linked to global environmental change. We suggest that this present phase of surface uplift, amounting so far to 150 m, is being caused by the nonsteady-state thermal and isostatic response of the crust to erosion, following an increase in erosion rates in the late Early Pleistocene, most likely as a result of the first large northern-hemisphere glaciation during oxygen isotope stage 22 at 870 ka. We suggest that the earlier uplift phase, responsible for the initial 250 m of uplift, resulted from a similar increase in erosion rates caused by the deterioration in local climate at 3.1 Ma. This uplift thus has no direct relationship to the crustal extension occurring in western Turkey, the rate and sense of which are thought not to have changed significantly on this time scale. Our results thus suggest that the present, often deeply incised, landscape of western Turkey has largely developed from the Middle Pleistocene onwards, for reasons not directly related to the active normal faulting that is also occurring. The local isostatic consequences of this active faulting are instead superimposed onto this “background” of regional surface uplift. Modelling of this surface uplift indicates that the effective viscosity of the lower continental crust beneath this part of Turkey is of the order of 10¹⁹ Pa s, similar to a recent estimate for beneath central Greece. The lower uplift rates observed in western Turkey, compared with central Greece, result from the longer typical distances of fluvial sediment transport, which cause weaker coupling by lower-crustal flow between offshore depocentres and eroding onshore regions that provide the sediment source.     

Erosion-driven uplift of the modern Central Alps

We present a compilation of data of modern tectono-geomorphic processes in the Central European Alps which suggest that observed rock uplift is a response to climate-driven denudation. This interpretation is predominantly based on the recent quantification of basin-averaged Late Holocene denudation rates that are so similar to the pattern and rates of rock uplift rates as determined by geodetic leveling. Furthermore, a GPS data-based synthesis of Adriatic microplate kinematics suggests that the Central Alps are currently not in a state of active convergence. Finally, we illustrate that the Central Alps have acted as a closed system for Holocene redistribution of sediment in which the peri-Alpine lakes have operated as a sink for the erosional products of the inner Central Alps.While various hypotheses have been put forward to explain Central Alpine rock uplift (e.g. lithospheric forcing by convergence, mantle processes, or ice melting) we show with an elastic model of lithospheric deformation, that the correlation between erosion and rock uplift rates reflects a positive feedback between denudation and the associated isostatic response to unloading. Thus, erosion does not passively respond to advection of crustal material as might be the case in actively converging orogens. Rather, we suggest that the geomorphic response of the Alpine topography to glacial and fluvial erosion and the resulting disequilibrium for modern channelized and associated hillslope processes explains much of the pattern of modern denudation and hence rock uplift. Therefore, in a non-convergent orogen such as the Central European Alps, the observed vertical rock uplift is primarily a consequence of passive unloading due to erosion.     

The Barents Sea Ice Sheet — A model of its growth and decay during the last ice maximum

On the basis of geomorphological and sedimentological data, we believe that the entire Barents Sea was covered by grounded ice during the last glacial maximum. ¹⁴C dates on shells embedded in tills suggest marine conditions in the Barents Sea as late as 22 ka BP; and models of the deglaciation history based on uplift data from the northern Norwegian coast suggest that significant parts of the Barents Sea Ice Sheet calved off as early as 15 ka BP. The growth of the ice sheet is related to glacioeustatic fall and the exposure of shallow banks in the central Barents Sea, where ice caps may develop and expand to finally coalesce with the expanding ice masses from Svalbard and Fennoscandia.The outlined model for growth and decay of the Barents Sea Ice Sheet suggests a system which developed and existed under periods of maximum climatic deterioration, and where its growth and decay were strongly related to the fall and rise of sea level.     

Land uplift effects on the phosphorus cycle of the Baltic Sea

The isostatic land uplift after the latest glaciation period in northern Europe means that the descending wave base in the eutrophicated Baltic Sea continuously exposes new bottom areas to increasing wind and wave-induced erosion. Erosion adds considerable amounts of phosphorus (P) and clay particles to the water column. This study has used a dynamic mass-balance model to investigate how land uplift affects the whole P cycle in the five major subbasins of the Baltic Sea. The model uses a unitary set of variables and constants for all subbasins with the exception of measurable, basin-specific driving variables. Differences in P concentrations between the subbasins could be quite accurately quantified only when the land uplift gradient was used as a driving variable. The clarifying effect from clay particles was found to be a major reason why those subbasins with the most intensive land uplift rates were also the ones with the lowest P concentrations. Without using the land uplift gradient as a model input, concentration differences could not be quantitatively explained in a meaningful way. Furthermore, simulations showed that clay particle erosion from land uplift has a substantial impact on all major internal P fluxes of the Baltic Sea. At the turn of the millennium, one of the subbasins (the Bothnian Bay) was oligotrophic, whilst the other four major subbasins were mesotrophic. Without the clarifying effect from the clay particles added to the water column during erosion of the rising seafloor, all five major subbasins of the Baltic Sea would probably be substantially more eutrophic.     

龙门山均衡重力异常及其对青藏高原东缘山脉地壳隆升的约束

青藏高原东缘处于不均衡状态，自西而东可分为青藏高原弱负均衡重力异常区、龙门山正均衡重力异常区和四川盆地负均衡重力异常区，表明该区的不均衡状态并未导致Airy均衡运动的产生，即龙门山没有均衡下降，而处于不断的隆升状态，显示该地区反均衡运动的构造抬升是导致龙门山隆升的主因。本次采用似三度体重力异常计算方法对该区的正均衡重力异常进行模拟和反演，研究了大尺度地貌分异与均衡重力异常分区之间的相互关系，结果表明，龙门山的下地壳顶面抬升了11．2～12．6km，造成了龙门山的正均衡异常，揭示了构造抬升和剥蚀作用在相似的时间尺度上和空间尺度上控制着龙门山地貌的形成，龙门山的表面隆升是构造隆升和剥蚀作用相叠加的产物。     

Be data from meltwater channels suggest that Jameson Land, east Greenland, was ice-covered during the last glacial maximum

Along the northeast Greenland continental margin, bedrock on interfjord plateaus is highly weathered, whereas rock surfaces in fjord troughs are characterized by glacial scour. Based on the intense bedrock weathering and lack of glacial deposits from the last glaciation, interfjord plateaus have long been thought to be ice-free throughout the last glacial maximum (LGM). In recent years there is growing evidence from shelf and fjord settings that the northeast Greenland continental margin was more extensively glaciated during the LGM than previously thought. However, little is still known from interfjord settings. We present cosmogenic ¹⁰Be data from meltwater channels and weathered sandstone outcrops on Jameson Land, an interfjord highland north of Scoresby Sund. The mean exposure age of samples from channel beds (n = 3) constrains on the onset of deglaciation on interior Jameson Land to 18.5 ± 1.3–21.4 ± 1.9 ka (for erosion conditions of 0–10 mm/ka, respectively). This finding adds to growing evidence that the northeast Greenland continental margin was more heavily glaciated during the LGM than previously thought.     

Geomorphological evidence from the Lleyn Peninsula constraining models of the magnitude and rate of isostatic rebound during deglaciation of the Irish Sea Basin

A stepped series of sand and gravel terraces on the Lleyn peninsula of North Wales is used to test the magnitude and rate of isostatic depression required by the recently proposed glacimarine model of deglaciation of the Irish Sea Basin. A relative sea-level fall of 70 m is required while the ice remained pinned at the north Lleyn coast. Even taking the maximum known rate of isostatic uplift, the margin would have to remain stationary for 1400 years. It seems more reasonable to interpret the Lleyn terraces, and similar features around the Irish Sea Basin, as glacifluvial and glacilacustrine.     

Course and origin of the Fennoscandian uplift: the case for two separate mechanisms

The classical glacial isostatic uplift paradigm of Fennoscandia is in need of a thorough revision as to its course and origin. The total uplift amounts to 830 m (not 300 m). The centre of uplift started to rise 13,000 BP (not 9300 BP). The uplift appears to be composed of two separate mechanisms; one true glacial isostatic rebound, and one linear uplift factor possibly representing a phase boundary adjustment. Calculations and modelling must treat these two factors independently (which has not been the case).