[1] The basal mass balance of the Amery Ice Shelf (AIS) in East Antarctica is investigated using a numerical ocean model. The main improvements of this model over previous studies are the inclusion of frazil formation and dynamics, tides and the use of the latest estimate of the sub-ice-shelf cavity geometry. The model produces a net basal melt rate of 45.6 Gt year À1 (0.74 m ice year À1 ) which is in good agreement with reviewed observations. The melting at the base of the ice shelf is primarily due to interaction with High Salinity Shelf Water created from the surface sea-ice formation in winter. The temperature difference between the coldest waters created in the open ocean and the in situ freezing point of ocean water in contact with the deepest part of the AIS drives a melt rate that can exceed 30 m of ice year À1. The inclusion of frazil dynamics is shown to be important for both melting and marine ice accretion (refreezing). Frazil initially forms in the supercooled water layer adjacent to the base of the ice shelf. The net accretion of marine ice is 5.3 Gt year À1 , comprised of 3.7 Gt year À1 of frazil accretion and 1.6 Gt year À1 of direct basal refreezing.

Totten Glacier, the primary outlet of the Aurora Subglacial Basin, has the largest thinning rate in East Antarctica 1,2 . Thinning may be driven by enhanced basal melting due to ocean processes 3 , modulated by polynya activity 4,5 . Warm modified Circumpolar Deep Water, which has been linked to glacier retreat in West Antarctica 6 , has been observed in summer and winter on the nearby continental shelf beneath 400 to 500 m of cool Antarctic Surface Water 7,8 . Here we derive the bathymetry of the sea floor in the region from gravity 9 and magnetics 10 data as well as ice-thickness measurements 11 . We identify entrances to the ice-shelf cavity below depths of 400 to 500 m that could allow intrusions of warm water if the vertical structure of inflow is similar to nearby observations. Radar sounding reveals a previously unknown inland trough that connects the main ice-shelf cavity to the ocean. If thinning trends continue, a larger water body over the trough could potentially allow more warm water into the cavity, which may, eventually, lead to destabilization of the low-lying region between Totten Glacier and the similarly deep glacier flowing into the Reynolds Trough. We estimate that at least 3.5 m of eustatic sea level potential drains through Totten Glacier, so coastal processes in this area could have global consequences.The Totten Glacier drains into the Sabrina Coast in an area where we find coastal ice grounded below sea level and the potential for local marine ice sheet instability 12 upstream of the grounding line ( Fig. 1b). We infer the bathymetry seaward of the grounding line using inversions of gravity data 9 informed by magnetics data 10 and ice-thickness measurements 11 . The inversion reveals the southwest area of the Totten Glacier Ice Shelf (TGIS) cavity is the deepest, reaching 2.7 ± 0.19 km below sea level ( Fig. 2), comparable to the grounding line depths of Amery Ice Shelf 13 and the segment of the Moscow University Ice Shelf (MUIS) overlying the Reynolds Trough 11 . The shallowest area of the cavity (∼300 mbsl) is found beneath the calving front of the ice shelf where a large coastparallel ridge connects Law Dome with a peninsula of grounded ice protruding from the east side of the cavity (Fig. 2). The ridge extends 40 km seaward of the calving front and would have been a source of backstress on the Totten Glacier as recently as 1996 when ice rises were last detected 14 . The inversion reveals depressions located near the centre of the ridge (650 ± 190 mbsl) and to the east of the grounded ice peninsula (860 ± 190 mbsl) (Fig. 2, Profile A-A ). Looking along the long axis of the full Totten cavity we see it is an average of 500 m deeper along the western (Law Dome) side. We infer two basins on the long axis reaching depths of 2.7 ± 0.19 km and 2.0 ± 0.19 km (SW and NE, respectively; Fig. 2) separated by a narrow ridge causing an ice rise near the middle of the ice shelf (the left-hand panel in Fig. 2) 14 .Published grounding lines 14,15 indicate an area of grounded ice bounded by the MUIS t...

No abstract

The first Cenozoic ice sheets initiated in Antarctica from the Gamburtsev Subglacial Mountains and other highlands as a result of rapid global cooling ∼34 million years ago. In the subsequent 20 million years, at a time of declining atmospheric carbon dioxide concentrations and an evolving Antarctic circumpolar current, sedimentary sequence interpretation and numerical modelling suggest that cyclical periods of ice-sheet expansion to the continental margin, followed by retreat to the subglacial highlands, occurred up to thirty times. These fluctuations were paced by orbital changes and were a major influence on global sea levels. Ice-sheet models show that the nature of such oscillations is critically dependent on the pattern and extent of Antarctic topographic lowlands. Here we show that the basal topography of the Aurora Subglacial Basin of East Antarctica, at present overlain by 2-4.5 km of ice, is characterized by a series of well-defined topographic channels within a mountain block landscape. The identification of this fjord landscape, based on new data from ice-penetrating radar, provides an improved understanding of the topography of the Aurora Subglacial Basin and its surroundings, and reveals a complex surface sculpted by a succession of ice-sheet configurations substantially different from today's. At different stages during its fluctuations, the edge of the East Antarctic Ice Sheet lay pinned along the margins of the Aurora Subglacial Basin, the upland boundaries of which are currently above sea level and the deepest parts of which are more than 1 km below sea level. Although the timing of the channel incision remains uncertain, our results suggest that the fjord landscape was carved by at least two iceflow regimes of different scales and directions, each of which would have over-deepened existing topographic depressions, reversing valley floor slopes.

A simple computer scheme developed by Budd and Smith (1985) and modified by D. Jenssen has been further developed to provide a rapid computation of steady-state balance fluxes over arbitrary ice masses, given the surface elevations and net accumulation distribution. The scheme provides a powerful diagnostic tool to examine the flux and state of balance over whole ice masses or limited regions to interpret field observations for dynamics or the state of balance.In many cases the uncertainty in the state of balance may be much less than the uncertainty in the deformation and sliding properties of the ice and so the flux and velocities derived from balance could provide a useful guide for the dynamics where direct observations are sparse.The scheme assumes that, on a horizontal scale of many ice thicknesses, the ice-flow direction is approximately down the steepest surface slope. The continuity equation is used to compute steady-state implied downslope fluxes at each grid point from integrations of the net accumulation over the area from the summits to the edges. The algorithm ensures the exact integral balance of the surface net flux over the area with flow through boundaries.Applications are demonstrated for the whole of Antarctica and for regional areas. Comparisons are made between fluxes computed from observed ice thicknesses and velocities and those computed from balance. The observed ice thicknesses can also be used to compute surface velocities from assumed column-to-surface velocity ratios. The combined fluxes from observations and balance can be used to compute rates of change of elevation with time.

ABSTRACT. The generalized (Glen) flow relation for ice, involving the second invariants of the stress deviator and strain-rate tensors, is only expected to hold for isotropic polycrystalline ice. Previous single-stress experiments have shown that for the steady-state flow, which develops at large strains, the tertiary strain rate is greater than the minimum (secondary creep) value by an enhancement factor which is larger for shear than compression. Previous experiments combining shear with compression normal to the shear plane have shown that enhancement of the tertiary octahedral strain rate increases monotonically from compression alone to shear alone. Additional experiments and analyses presented here were conducted to further investigate how the separate tertiary shear and compression strain-rate components are related in combined stress situations. It is found that tertiary compression rates are more strongly influenced by the addition of shear than is given by a Glen-type flow relation, whereas shear is less influenced by additional compression. A scalar function formulation of the flow relation is proposed, which fits the tertiary creep data well and is readily adapted to a generalized form that can be extended to other stress configurations and applied in ice mass modelling. BACKGROUNDIn natural ice masses the most important and common state of deformation is arguably a combination of approximately bed-parallel shear and vertical compression. For deformational flow with a stationary boundary, a region of simple shear is associated in an essential way with bulk transport of ice in glaciers, ice sheets and ice shelves, and this is generally accompanied by normal deformations associated with increasing velocities along the flow and divergence or convergence transverse to the flow.For a coordinate system with x and y horizontal and z vertical, and corresponding component velocities (u, v, w), simple shear deformation in the x direction can be characterized by du/dz = c where we note that the horizontal planes on which the forces generating shear deformation act do not rotate, while compression normal to these planes is described by dw/dz = k, where c/2 and k are the respective shear and vertical compressive strain rates. The compressive flow may be confined or unconfined, and quite generally the accompanying horizontal normal strain rates are du dx ¼ ð À 1Þk and dv dy ¼ Àk where the factors involving indicate the proportions of the deformations in the horizontal directions, relative to the rate of vertical compression. Note that = 1/2 corresponds to uniaxial compression in the z direction, while = 1 corresponds to longitudinally confined compression in the experiments reported here (Fig. 1).The generalized flow relation for ice involving the second invariants of the stress deviator and strain-rate tensors (Nye, 1953;Glen, 1958) provides a useful formulation for the interactions between the individual stress and strain-rate components for isotropic ice. This relation is not expected to apply for anisotropic i...

Laboratory creep deformation experiments have been conducted on initially isotropic laboratory-made samples of polycrystalline ice. Steady-state tertiary creep rates, , were determined at strains exceeding 10% in either uniaxial-compression or simple-shear experiments. Isotropic minimum strain rates, , determined at ˜1 % strain, provide a reference for comparing the relative magnitude of tertiary creep rates in shear and compression through the use of strain-rate enhancement factors, E, defined as the ratio of corresponding tertiary and isotropic minimum creep rates, i.e. . The magnitude of strain-rate enhancement in simple shear was found to exceed that in uniaxial compression by a constant factor of 2.3. Results of experiments conducted at octahedral shear stresses of to = 0.040.80 MPa indicate a creep power-law stress exponent of n = 3 for isotropic minimum creep rates and n = 3.5 for tertiary creep rates. The difference in stress exponents for minimum and tertiary creep regimes can be interpreted as a t0 stress-dependent level of strain-rate enhancement, i.e. .The implications of these results for deformation in complex multicomponent stress configurations and at stresses below those used in the current experiments are discussed.

Subglacial hydrology in East Antarctica is poorly understood, yet may be critical to the manner in which ice flows. Data from a new regional airborne geophysical survey (ICECAP) have transformed our understanding of the topography and glaciology associated with the 287,000 km2 Aurora Subglacial Basin in East Antarctica. Using these data, in conjunction with numerical ice sheet modeling, we present a suite of analyses that demonstrate the potential of the 1000 km‐long basin as a route for subglacial water drainage from the ice sheet interior to the ice sheet margin. We present results from our analysis of basal topography, bed roughness and radar power reflectance and from our modeling of ice sheet flow and basal ice temperatures. Although no clear‐cut subglacial lakes are found within the Aurora Basin itself, dozens of lake‐like reflectors are observed that, in conjunction with other results reported here, support the hypothesis that the basin acts as a pathway allowing discharge from subglacial lakes near the Dome C ice divide to reach the coast via the Totten Glacier.