Senglaciala ackumulationsformer och glaciationsförhållanden i Narvik — Skjomenområdet,Norge

Abstract:

DAHL, RAGNAR. Late-glacial accumulation forms and glaciation in the Narvik-Skjomen district, Norway. Norsk geogr. Tidsskr. 1967, 21, 157–241.

Late-glacial accumulation forms at corresponding levels are described and discussed. Most of the accumulations are moraine ridges deposited at the lateral margins of the ice flowing in the main valleys or at the lateral and frontal margins of ice tongues branching out into tributaries.

The frontal moraine ridges appear only in small tributary valleys or similar depressions in the terrain. They have an arched form and an advanced position in the depressions and are broken by melt-water channels. Together with the different indications of pressure in the material, this shows that the moraines correspond to phases of the deglaciation period in which the ice masses were temporarily advancing. The calculated gradients of the ice surface and radio-carbon datings indicate that these phases correspond to the two glacial phases of the Tromsø-Lyngen substage, which, according to ANDERSEN (1965), approximately correspond to the Early Dryas and Late Dryas periods.     

LANDSLIDE‐GLACIER INTERACTION IN A NEOPARAGLACIAL SETTING AT TVERRBYTNEDE,JOTUNHEIMEN, SOUTHERN NORWAY

A tongue‐like, boulder‐dominated deposit in Tverrbytnede, upper Visdalen, Jotunheimen, southern Norway, is interpreted as the product of a rock avalanche (landslide) due to its angular to subangular boulders, surface morphology with longitudinal ridges, down‐feature coarsening, and cross‐cutting relationship to ‘Little Ice Age’ moraines. The rock avalanche fell onto glacier ice, probably channelled along a furrow between two glaciers, and stopped on the glacier foreland, resulting in its elongated shape and long runout distance. Its distal margin may have become remobilized as a rock glacier, but a rock glacier origin for the entire landform is discounted due to lack of source debris, presence of matrix, lack of transverse ridges, and sparcity of melt‐out collapse pits. Lichenometric dating of the deposit indicates an approximate emplacement age of ad 1900. Analysis highlights the interaction of rock‐slope failures and glaciers during deglacierization in a neoparaglacial setting, with reduced slope stability due to debuttressing and permafrost degradation, and enhanced landslide mobility due to flow over a glacier and topographic channelling. Implications for the differentiation of relict landslides, moraines and rock glaciers are discussed and interrelationships between these landforms are considered in terms of an ice‐debris process continuum.     

DEPOSITS FROM LANDSLIDES AND AVALANCHES TRIGGERED BY SEISMIC ACTIVITY IN SWEDISH LAPLAND

In the mountain area of Swedish Lapland, landforms with semicircular or horse‐shoe ridges encircling hummocky ground occur. They are interpreted as landslide or avalanche deposits directly upon, from or adjacent to down‐wasting ice. Since they are restricted to the vicinity of fault scarps in the Pärvie fault system the releasing factor is suggested to be displacement of bedrock blocks or related seismic activity.     

Columbia Mountain landslide: late-glacial emplacement and indications of future failure, Northwestern Montana, USA

The well preserved and undissected Columbia Mountain landslide, which is undergoing suburban development, was studied to estimate the timing and processes of emplacement. The landslide moved westward from a bedrock interfluve of the northern Swan Range in Montana, USA onto the deglaciated floor of the Flathead Valley. The landslide covers an area of about 2 km², has a toe-to-crown height of 1100 m, a total length of 3430 m, a thickness of between 3 and 75 m, and an approximate volume of 40 million m³. Deposits and landforms define three portions of the landslide; from the toe to the head they are: (i) clast-rich diamictons made up of gravel-sized angular rock fragments with arcuate transverse ridges at the surface; (ii) silty and sandy deposits resting on diamictons in an internally drained depression behind the ridges; and (iii) diamictons containing angular and subangular pebble-to block-sized clasts (some of which are glacially striated) in an area of lumpy topography between the depression and the head of the landslide. Drilling data suggest the diamictons cover block-to-slab-sized bedrock clasts that resulted from an initial stage of the failure.The landslide moved along a surface that developed at a high angle to the NE-dipping, thinly bedded metasediments of the Proterozoic Belt Supergroup. The exposed slope of the main scarp dips 30–37°W. A hypothetical initial rotational failure of the lower part of a bedrock interfluve may have transported bedrock clasts into the valley. The morphology and deposits at the surface of the landslide indicate deposition by a rock avalanche (sturzstrom) derived from a second stage of failure along the upper part of the scarp.The toe of the Columbia Mountain landslide is convex-west in planview, except where it was deflected around areas now occupied by glacial kettles on the north and south margins. Landsliding, therefore, occurred during deglaciation of the valley while ice still filled the present-day kettles. Available chronostratigraphy suggests that the ˜1-km thick glacier in the region melted before 12,000 ¹⁴C years BP—within 3000 years of the last glacial maximum. Deglaciation and hillslope failure are likely causally linked. Failure of the faceted interfluve was likely due tensile fracturing of bedrock along a bedding-normal joint set shortly after glacial retreat from the hillslope.Open surficial tension fractures and grabens in the Swan Range are limited to an area above the crown of the landslide. Movement across these features suggests that extensional flow of bedrock (sackung) is occurring in what remains of the ridge that failed in the Columbia Mountain landslide. The fractures and grabens likely were initiated during failure, but their morphologies suggest active extension across some grabens. Continued movement of bedrock above the crown may result in future mass movements from above the previous landslide scarp. Landslides sourced from bedrock above the scarp of the late-glacial Columbia Mountain landslide, which could potentially be triggered by earthquakes, are geologic hazards in the region.     

The genesis of the northern Kettle Moraine, Wisconsin

Interpreting past glacial dynamics from the glacial record requires that the depositional environments of glacial sediments and landforms be understood. In the case of interlobate deposits, models that incorporate various components of pro, supra and subglacial deposition have been developed and tested in the northern Kettle Moraine (nKM), Wisconsin; a large interlobate deposit that formed between the Green Bay and Lake Michigan lobes of the Laurentide Ice Sheet during the last deglaciation. In this paper, we interpret a new genesis for the nKM using sediment analysis and distribution along with landform distribution. In Sheboygan County, the nKM consists of two steep-sided, high-relief, hummocky ridges separated by a low elevation and low-relief central axis. Gravel in the bounding hummocky ridges is well-sorted and well-rounded. Some bedding is collapsed. Large, isolated moulin kames are restricted to the axis area and composed of relatively poorly sorted, more angular gravel and diamicton. The distribution of these different sediments and landforms are explained by the accumulation of supraglacial debris that insulated the ice below the axis of the nKM, while the melting of cleaner ice on either side formed channels on the ice surface. As deglaciation proceeded, a substantial thickness of well-rounded, stream-deposited sand and gravel accumulated on ice in the bounding channels. Eventual collapse of this sediment formed the two hummocky ridges. Poorly sorted debris along the axis fell and slid into moulins and larger collapse areas in the ice. Thus, differential debris insulation and ice ablation controlled the mainly supraglacial deposition of this part of the nKM.     

POST-DRAINAGE ICE DAM RESPONSE AT LAKE MERZBACHER, INYLCHEK GLACIER, KYRGYZSTAN

The Inylchek glacier system in Central Tian Shan, Kyrgyzstan, comprises a large glacier‐dammed lake which usually drains once a year through a subglacial drainage system. Detailed GPS measurements on the ice dam and the analysis of Aster scenes from several subsequent years provide insight into the post‐drainage dam response and the changed ice dynamic conditions. We demonstrate that during high water levels in the lake a large part of the ice dam is afloat, lifting the ice surface up to almost 20 m in the central dam region. During this phase of extensive flotation strong calving is facilitated, which is supported by the high density of ice debris in the lake. In general, surface ice velocities are about 1.5‐2 times higher during summer than winter. Closer to the lake, however, ice velocities increase considerably after the drainage event, showing values more than three times the annual mean. The increased mass flux during the phase of high lake level needs to be compensated by replenishment of the lost ice from the dam. Therefore the ice velocities show compressive flow during the remaining part of the year. These results show that Southern Inylchek glacier is strongly influenced by the existence of the lake.     

The Origin and Significance of Debris-charged Ridges at the Surface of Storglaciären, Northern Sweden

Storglaciären is a 3.2 km long polythermal valley glacier in northern Sweden. Since 1994 a number of small (1–2 m high) transverse debris‐charged ridges have emerged at the ice surface in the terminal zone of the glacier. This paper presents the results of a combined structural glaciological, isotopic, sedimentological and ground‐penetrating radar (GPR) study of the terminal area of the glacier with the aim of understanding the evolution of these debris‐charged ridges, features which are typical of many polythermal glaciers. The ridges originate from steeply dipping (50–70°) curvilinear fractures on the glacier surface. Here, the fractures contain bands of sediment‐rich ice between 0.2 and 0.4 m thick composed of sandy gravel and diamicton, interpreted as glaciofluvial and basal glacial material, respectively. Structural mapping of the glacier from aerial photography demonstrates that the curvilinear fractures cannot be traced up‐glacier into pre‐existing structures visible at the glacier surface such as crevasses or crevasse traces. These curvilinear fractures are therefore interpreted as new features formed near the glacier snout. Ice adjacent to these fractures shows complex folding, partly defined by variations in ice facies, and partly by disseminated sediment. The isotopic composition (δ¹⁸O) of both coarse‐clear and coarse‐bubbly glacier ice facies is similar to the isotopic composition of the interstitial ice in debris layers that forms the debris‐charged ridges, implying that none of these facies have undergone any significant isotopic fractionation by the incomplete freezing of available water. The GPR survey shows strong internal reflections within the ice beneath the debris‐charged ridges, interpreted as debris layers within the glacier. Overall, the morphology and distribution of the fractures indicate an origin by compressional glaciotectonics near the snout, either at the thermal boundary, where active temperate glacier ice is being thrust over cold stagnant ice near the snout, or as a result of large‐scale recumbent folding in the glacier. Further work is required to elucidate the precise role of each of these mechanisms in elevating the basal glacial and glaciofluvial material to the ice surface.     

Changing characteristics of arctic pressure ridges

The advent of multibeam sonar permits us to obtain full three-dimensional maps of the underside of sea ice. In particular this enables us to distinguish the morphological characteristics of first-year (FY) and multi-year (MY) pressure ridges in a statistically valid way, whereas in the past only a small number of ridges could be mapped laboriously by drilling. In this study pressure ridge distributions from two parts of the Arctic Ocean are compared, in both the cases using mainly data collected by the submarine “Tireless” in March 2007 during two specific grid surveys, in the Beaufort Sea at about 75° N, 140° W (N of Prudhoe Bay), and north of Ellesmere Island at about 83° 20′ N, 64° W. In the Beaufort Sea the ice was mainly FY, and later melted or broke up as this area became ice-free during the subsequent summer. N of Ellesmere Island the ice was mainly MY. Ridge depth and spacing distributions were derived for each region using the boat's upward looking sonar, combined with distributions of shapes of the ridges encountered, using the Kongsberg EM3002 multibeam sonar. The differing shapes of FY and MY ridges are consistent with two later high-resolution multibeam studies of specific ridges by AUV. FY ridges are found to fit the normal triangular shape template in cross-section (with a range of slope angles averaging 27°) with a relatively constant along-crest depth, and often a structure of small ice blocks can be distinguished. MY ridges, however, are often split into a number of independent solid, smooth blocks of large size, giving an irregular ridge profile which may be seemingly without linearity.Our hypothesis for this difference is that during its long lifetime an MY ridge is subjected to several episodes of crack opening; new cracks in the Arctic pack often run in straight lines across the ridges and undeformed ice alike. Such a crack will open somewhat before refreezing, interpolating a stretch of thin ice into the structure, and breaking up the continuity and linearity of the ridge crest. Many such episodes over a number of years can cause the ridge to become simply a series of blocks. This has implications for ridge strength and for permeability to spilled oil. As the percentage of MY ice in the Arctic diminishes, Arctic ridging will be more and more dominated by FY ridges, and we discuss the implications of this change of character of the ice underside in the light of the statistics that we have generated for the two types of ridge.     

Dynamics of a complex mass movement triggered by heavy rainfall: a case study from NW Turkey

Following a period of heavy precipitation, a large and complex mass movement, namely the Dagkoy landslide, occurred in the West Black Sea Region of Turkey on May 21, 1998. This paper describes the conditioning factors of the landslide and interprets the mass transport processes in terms of a movement scenario. Geology, geomorphology and vegetation cover were considered as the conditioning factors of the failure. Observations showed that the gently sloping (about 10°) area is mostly covered by dense forest trees at the crown where the motion initiated. Significant intersection of the collapsed slope with dip of the local marls seems to have contributed to the formation and geometry of the landslide. The distance from the crown down to the toe of the landslide measured more than 600 m, with about 0.6 km³ total earth material displaced. The landslide has both a block sliding characteristics in the upper portions and a debris flow/soil flow component around the margins of the sliding blocks in the middle parts and at the toe. The proposed scenario for the landslide reveals that the movement was initiated near crown as a result of the excess water content in the marls at the end of 3 days of heavy rainfall. The early perturbations (transverse cracks, ridges, etc.) lasted for 6–7 h, after which the central part of the zone started to move as a soil flow in which very large intact blocks were transported. Even though the movement was very rapid (1.2 m/min), there was no loss of life. However, the movement destroyed 38 houses, one mosque and a considerable amount of farmland.     

乔治王岛菲尔德斯半岛周围海岸地貌研究

本文对南设得兰群岛乔治王岛菲尔德斯半岛周围现代和上升海岸地貌进行了研究。认为现代海岸地貌有三类:碎屑海岸、基岩海岸和冰崖海岸。对碎屑海岸受负载浮冰的波浪作用造成的各种现象进行了成因和类型研究。系统研究上升海岸之后得出结论,该区上升海岸地貌以海拔20米为界,上下分别属较老组和较新组。海岸平均上升速度为10毫米/年。     

Glacially Folded Outwash Near Lago Llanquihue, Southern Lake District, Chile

Folded outwash occurs in four distinct clusters in an arcuate arrangement just west of the terminal Llanquihue moraines deposited by the Lago Llanquihue piedmont ice lobe at the last glacial maximum. These clusters are physically connected along the eastern side to the Llanquihue terminal moraines, and along the western side to the Llanquihue outwash plain. Each cluster consists of three to eleven elongated ridges. The maximum height of individual ridges varies from cluster to cluster beween 18 and 28 m; the maximum length of individual ridges is between 93 and 1074 m. The average orientation of the ridges ranges over a 60° sector relative to former ice-flow direction. The folded out-wash sediments are cut by two distinct internal fault systems with only a faint surface expression below the Holocene top soil.
The folded outwash ridges are interpreted as a push moraine system produced by the same mechanical forces that act in a critically tapered wedge. The folded sediment is a sandy gravel with an angle of friction on the order of φ = 40°. Interpretations of structural data and of mechanical comparisons point to a basal thrust plane in a sand unit with φ between 24° and 30° and with a pore water pressure index of l = 0.7.
It is very unlikely that the observed and analyzed features wereformed under permafrost conditions.     

佛山市顺德区飞鹅山Ⅲ号滑坡形成机理与防治技术

采用工程地质钻探、物探、地质测绘及室内试验等技术方法探讨飞鹅山Ⅲ号滑坡形成机理与防治技术。结果表明：1）滑坡体主要岩性为泥质粉砂岩,飞鹅山滑坡属于新形成的深层中型牵引式滑坡,在平面上呈圈椅状。2）滑坡属于双层滑面滑坡,主滑面以中型深层滑坡为主,主滑体上部发育中型中厚层滑坡。3）滑坡产生的原因为：①泥质粉砂岩倾向与坡向基本一致,且岩层倾角为中等倾角;②人工开挖使坡脚形成高陡临空面,抗滑力大为降低;③雨水沿层面及节理裂隙入渗至坡体深部,大大增加岩土体容重,同时泥质粉砂岩遇水软化,抗剪强度显著降低。4）结合该滑坡区地质环境条件,采用坡面削坡+锚杆（索）+格构梁+双排预应力锚拉抗滑桩+三维网植草绿化+截排水+毛石挡墙的综合治理方法进行防治,监测结果显示该滑坡变形及位移已得到有效控制,整治效果良好。     

Spatial distribution of surface conditions and runoff generation in small arid watersheds, Zin Valley Badlands, Israel

High rill density may be regarded as indicative of frequent and integrated runoff along a drainage network. However, field observations of soil development and rill geometry in small, first-order catchments (0.1 to 1 ha) of the Zin Valley Badlands, northern Negev, Israel, suggest a pattern of partial area contribution and frequent flow discontinuities along hillslopes and channels. Changing soil properties, associated with an increase of slope angle and slope length, appear to be responsible for high infiltration on the slopes, leading to nonuniform runoff generation within small drainage basins. Runoff observation, sprinkling tests, and soil analysis along ridges and sideslopes were carried out to test this hypothesis. The results confirmed that infiltration capacity on the sideslopes is significantly higher than on the ridges. Under current rainfall conditions, only extreme rainfall intensities are sufficient to generate runoff along entire slopes. The discontinuous nature of most runoff events causes erosion on ridges and deposition on slopes, which enhances soil development on the valley sideslopes, creating a positive feedback on infiltration rate and depth. This demonstrates that the links between within-storm rainfall conditions and spatial distribution of soil characteristics have to be incorporated into our understanding of landscape development in badlands.     

Analysis of LiDAR-derived topographic information for characterizing and differentiating landslide morphology and activity

This study used airborne laser altimetry (LiDAR) to examine the surface morphology of two canyon-rim landslides in southern Idaho. The high resolution topographic data were used to calculate surface roughness, slope, semivariance, and fractal dimension. These data were combined with historical movement data (Global Positioning Systems (GPS) and laser theodolite) and field observations for the currently active landslide, and the results suggest that topographic elements are related to the material types and the type of local motion of the landslide. Weak, unconsolidated materials comprising the toe of the slide, which were heavily fractured and locally thrust upward, had relatively high surface roughness, high fractal dimension, and high vertical and lateral movement. The body of the slide, which predominantly moved laterally and consists mainly of undisturbed, older canyon floor materials, had relatively lower surface roughness than the toe. The upper block, consisting of a down-dropped section of the canyon rim that has remained largely intact, had a low surface roughness on its upper surface and high surface roughness along fractures and on its west face (unrelated to landslide motion). The upper block also had a higher semivariance than the toe and body. The topographic data for a neighboring, older and larger landslide complex, which failed in 1937, are similarly used to understand surface morphology, as well as to compare to the morphology of the active landslide and to understand scale-dependent processes. The morphometric analyses demonstrate that the active landslide has a similar failure mechanism and is topographically more variable than the 1937 landslide, especially at scales > 20 m. Weathering and the larger scale processes of the 1937 slide are hypothesized to cause the lower semivariance values of the 1937 slide. At smaller scales (< 10 m) the topographic components of the two landslides have similar roughness and semivariance. Results demonstrate that high resolution topographic data have the potential to differentiate morphological components within a landslide and provide insight into the material type and activity of the slide. The analyses and results in this study would not have been possible with coarser scale digital elevation models (10-m DEM). This methodology is directly applicable to analyzing other geomorphic surfaces at appropriate scales, including glacial deposits and stream beds.     

Deglaciation pattern in sub-aquatic — supra-aquatic transitional environment illustrated by the Klarälven valley system, Värmland, western Sweden

The downwasting of the Weichselian ice sheet in the Klarälven valley and its continuation southwards has been studied. The meltwater drainage below the highest coastline (HCL) on the plain of Lake Vänern took place subglacially causing breaking‐up and calving around tunnel mouths in the ice margin. When the ice margin reached into the extension of the Klarälven valley, the drainage pattern changed to lateral. The ice disappeared as a narrow ice tongue later than in the surrounding highland. When the tongue receded northwards it left dead‐ice bodies behind, that remained in the valley after the general retreat. Along the tongue and dead‐ice bodies glaciofluvial sediments accumulated as kame terraces, deriving from esker trains in tributary valleys. At the mouths of the tributaries, deltas were deposited. Halts in the retreat occurred at a few places marked by thresholds of till. According to the present study the HCL rises northwards, from 180 m a.s.l. at Brattforsheden to 228 m at Höljes 170 km to the north. At halts in the ice recession it dropped a few metres. A tentative dating based on the level of the HCL, the tilt of raised shorelines and the rate of shoreline regression has given the time 10900 to 10360 calibrated years BP for the retreat of the ice margin from Brattforsheden to Höljes.     

ICE‐MARGINAL SEDIMENT DELIVERY TO THE SURFACE OF A HIGH‐ARCTIC GLACIER: AUSTRE BRØGGERBREEN,SVALBARD

Enhanced delivery of water‐saturated, ice‐marginal sediments to the glacier surface is a response to glacier thinning that has the potential to increase both levels of sediment transfer through the glacier hydrological system and total basin sediment yields. Preliminary observations made during summer 2007 at Austre Brøggerbreen, Svalbard, confirm that ice‐marginal debris flows in the upper reaches of the glacier are actively delivering sediments to the glacier surface, which may then be flushed into the glacier's hydrological system. During a four‐day observation period, several stochastic pulses in water turbidity were observed at a single portal where solely supra‐ and englacial drainage emerge at the glacier margin. The erratic suspended sediment fluxes were hypothesized to originate from ice‐marginal sources. Quantitative analysis of continuous turbidity and discharge data confirm that discharge is not driving these turbidity pulses and, combined with observational data, that the most likely origin is the delivery of water‐saturated sediments to the glacier surface from ice‐marginal, debris flows with subsequent transfer to the portal via the glacial drainage system. These observations illustrate the potential importance of the paraglacial component to the overall sediment cascade of deglaciating basins and highlight the need for careful interpretation of turbidity records, where stochastic pulses in turbidity may be attributed to sources and processes other than ice‐marginal sediment inputs.     

Analysis of flute forming conditions using ice sheet reconstructions and field techniques

Basal shear stress and sediment strength associated with the development of glacial flutes exposed during the 20th century in the Saskatchewan Glacier Valley Alberta, Canada, were determined by comparing reconstructed ice thicknesses, basal shear stresses, and field properties of sediments with the morphologically similar Kiwa Glacier Valley, British Columbia, Canada, where flutes are absent. Reconstructed subglacial conditions in these two valleys were compared to understand why flutes were developed in the former and not the latter. Using an existing topographic map of each glacier, equations for a series of longitudinal profile lines were determined to represent the existing ice surface. A previous ice surface, identified by trimlines along the valley walls, was reconstructed by applying the equations of longitudinal profile lines from the existing ice surface to a previous terminus between 5 and 10 km downvalley. After subtracting the elevation of the land surface (determined from topographic maps) from the reconstructed glacier surface, and calculating former ice surface slope, ice thickness and basal shear stress distributions were determined. Sediment texture and the location of flutes on a morainal topographic high, downglacier from a proglacial lake basin, allowed high porewater pressures to develop as glaciers extended to terminus positions in the Saskatchewan Glacier Valley. Sediment strength was reduced sufficiently below values of reconstructed shear stress plots to allow deformation creating flutes. The absence of a similar topographic high and different sediment textural characteristics in the Kiwa Glacier Valley resulted in lower porewater pressures and consequently less reduction in sediment strength preventing the development of glacial flutes despite higher shear stress values here. Results indicate that the degree to which sediment characteristics and porewater pressure allow reduction of subglacial sediment strength relative to basal shear stress is important in determining conditions when flutes may develop.     

Glacier Dammed Lakes in Norway

Glacier dammed lakes may have their outflow over a bedrock threshold like other lakes, or they may flow on the surface or below the damming ice. In the former case the lake, over some considerable period, has a constant maximum level with marked shore lines corresponding to the threshold. But when the level is determined by the ice dam, the conditions are unstable. The level of the lake will then to a certain extent vary in step with the glacier, and never remain at a certain level for any length of time. The normal behaviour of such a lake will be a moderately fast rise of the water level, dependent upon the supply, until a certain critical level is reached, whereupon drainage is rapid. Such is also to some extent the case with the type of lake first mentioned, if the height of the pass is situated near to the critical level for drainage.     

Reconstruction of the Ross Ice Drainage System, Antarctica, at the Last Glacial Maximum

We present here a revised reconstruction of the Ross ice drainage system of Antarctica at the last glacial maximum (LGM) based on a recent convergence of terrestrial and marine data. The Ross drainage system includes all ice flowlines that enter the marine Ross Embayment. Today, it encompasses one-fourth of the ice-sheet surface, extending far inland into both East and West Antarctica. Grounding lines now situated in the inner Ross Embayment advanced seaward at the LGM (radiocarbon chronology in Denton and Marchant 2000 and in Hall and Denton 2000a, b), resulting in a thick grounded ice sheet across the Ross continental shelf. In response to this grounding in the Ross (and Weddell) Embayment, ice-surface elevations of the marine-based West Antarctic Ice Sheet were somewhat higher at the LGM than at present (Steig and White 1997; Borns et al. 1998; Ackert et al. 1999). At the same time, surface elevations of the East Antarctic Ice Sheet inland of the Transantarctic Mountains were slightly lower than now, except near outlet glaciers that were dammed by grounded ice in the Ross Embayment. The probable reason for this contrasting behavior is that lowered global sea level at the LGM, from growth of Northern Hemisphere ice sheets, caused widespread grounding of the marine portion of the Antarctic Ice Sheet, whereas decreased LGM accumulation led to slight surface lowering of the interior terrestrial ice sheet in East Antarctica. Rising sea level after the LGM tripped grounding-line recession in the Ross Embayment, which has probably continued to the present day (Conway et al. 1999). Hence, gravitational collapse of the grounded ice sheet from the Ross Embayment, accompanied by lowering of the interior West Antarctic ice surface and of outlet glaciers in the Transantarctic Mountains, occurred largely during the Holocene. At the same time, increased Holocene accumulation caused a slight rise of the inland East Antarctic ice surface.     

Mantle convection under simulated plates: effects of heating modes and ridge and trench migration, and implications for the core—mantle boundary, bathymetry, the geoid and Benioff zones

Summary. Numerical convection models are presented in which plates are simulated by imposing piecewise constant horizontal velocities on the upper boundary. A 4 × 1 box of constant viscosity fluid and two-dimensional (2-D) flow is assumed. Four heating modes are compared: the four combinations of internal or bottom heating and prescribed bottom temperature or heat flux. The case with internal heating and an isothermal base is relevant to lower mantle or whole mantle convection, and it yields a lower thermal boundary layer which is laterally variable and can be locally reversed, corresponding to heat flowing back into the core locally. When scaled to the whole mantle, the surface deflections and gravity and geoid perturbations calculated from the models are comparable to those observed at the Earth's surface. For models with migrating ridges and trenches, the flow structure lags well behind the changing surface 'plate'configurations. This may help to explain the poor correlation between the main geoid features and plate boundaries. Trench migration substantially affects the dip of the cool descending fluid because of induced horizontal shear in the vicinity of the trench. Such shear is small for whole mantle convection, but is large for upper mantle convection, and would probably result in the Tonga Benioff zone dipping to the SE, opposite to the observed dip, for the case of upper mantle convection.