Glaciology and geological signature of the Last Glacial Maximum Antarctic ice sheet

Highlights

• We describe the glaciology of a modelled LGM Antarctic ice sheet model.

• We compare its likely geological signature to inferences from empirical records.

• We show that greatest subglacial erosion coincides with climate transitions.

• We argue that geological records integrate time-transgressive ice-sheet behaviour.

Abstract

Dynamical changes in contemporary ice sheets account for significant proportions of their current rates of mass loss, but assessing whether or not these processes are a natural part of ice-sheet evolution requires inference from palaeo-glaciological records. However, a robust mechanism for translating sparse geological data into meaningful interpretations of past glacier dynamics at the continental scale is lacking, since geological archives can be ambiguous, and often their chronology is only poorly constrained. To address this, we combine the interpretation of high-resolution Antarctic ice sheet model results with continent-wide geological evidence pertinent to the dynamical configuration of the ice sheet during the last, and possibly preceding, glacial maxima. We first focus on the thermal regime of the ice sheet, its pattern and velocity of flow, variability in likely subglacial erosion and sediment transport, and how these characteristics evolve during glacial transitions. We show that rapid basal sliding was restricted to discrete outlets that eroded and advected sediment toward and across the continental shelf primarily during the early stages of advance and retreat of the ice sheet, highlighting the need to consider time-transgressive behaviour in the interpretation of geological archives. Secondly, we present new modelling that attempts to improve the fit of our numerical model to geologically-based reconstructions in the Ross Sea. By accounting for locally-enhanced ablation in McMurdo Sound, our new simulation achieves a much closer fit to empirically-derived flow patterns than previously. Growth of the modelled Last Glacial Maximum ice sheet takes place primarily by marine ice accretion in the major embayments, as a consequence of cooler ocean temperatures and reduced sub-ice-shelf melting, and at its maximal extent represents a grounded ice volume excess above present of approximately 8.3 m sea-level equivalent. This figure thus provides an upper bound on the possible Antarctic contribution to deglacial meltwater pulses.

Introduction

The contiguous Antarctic ice sheets (AIS) reached their maximum size relatively late in the last glacial cycle, at around 18 ka, somewhat later than elsewhere in the Southern Hemisphere (e.g. Patagonia, New Zealand). This globally diachronous event therefore represents the last major change in volume and extent of the AIS (e.g. Clark et al., 2009, Denton et al., 2010), but, given the cyclicity apparent in glacial–interglacial oceanic and atmospheric conditions (e.g. Petit et al., 1999, EPICA community members, 2004, Lisiecki and Raymo, 2005, Elderfield et al., 2012), the Last Glacial Maximum (LGM) ice-sheet was most likely similar in geometry and dynamics to other Plio-Pleistocene glacial maxima AIS. These expanded AIS removed significant volumes of water from global oceans (Huybrechts, 2002, Pollard and DeConto, 2009), and profoundly affected interhemispheric oceanic circulation (Anderson et al., 2009, Denton et al., 2010). Given this important role, it is valuable to establish as accurately as possible the geometry of these glacial maxima ice sheets. Indeed, this has been attempted by studies of varying complexity and geographic scope (e.g. Drewry, 1979, Bentley and Anderson, 1998, Denton and Hughes, 2000, Anderson et al., 2002, Denton and Hughes, 2002, Bentley et al., 2006, Le Brocq et al., 2011, Whitehouse et al., 2012), and has led to sea-level-equivalent (s.l.e.) volume change estimates that have decreased over recent decades from ca 30 m s.l.e. to less than 10 m s.l.e. (Bentley, 1999, Huybrechts, 2002, Golledge et al., 2012). Surface exposure dating and interpretations from ice-cores increasingly help constrain surface elevations of the Antarctic ice sheets, particularly through the Pleistocene. Combined with a growing spatial coverage of marine geophysical surveys the three-dimensional shape of the LGM AIS is now relatively well-constrained (Ivins and James, 2005, Livingstone et al., 2012, Whitehouse et al., 2012), although there are still some sectors with little or no geological evidence of LGM ice surface heights (e.g. Enderby Land). Ages from glacial deposits in terrestrial and marine settings are increasing in number, but their unambiguous interpretation is still hampered by complexities such as uncertain environmental conditions at the time of either deposition or initial exposure (e.g. cosmogenic nuclide production rates, environment of sediment deposition, marine radiocarbon reservoir ages) and post-depositional environmental changes (rate and magnitude of uplift of rock surfaces, erosion and reworking of sediments) (cf. Andrews et al., 1999, Hillenbrand et al., 2010b). However, one of the areas of greatest uncertainty surrounding glacial maxima AIS remains the quantification of former ice sheet dynamics, and how they might affect the interpretation of geological archives. Given that recent studies increasingly recognise the contribution of dynamic effects (rather than purely mass balance-driven changes) to the changes currently being observed on contemporary ice sheets (Pritchard et al., 2009, Kjær et al., 2012, Shepherd et al., 2012), it is clear that a more comprehensive understanding of palaeo-dynamic behaviour is of value not just to the geological community, but also to glaciologists attempting to interpret present-day observations in the context of longer-term ice-sheet trajectories, to climate modellers who require accurate ice geometries for their global climate models, and to those interested in all aspects of sea-level change prediction.

Here we present new data from a 5 km-resolution whole-continent numerical ice-sheet model of the LGM AIS (Golledge et al., 2012) in an attempt to better quantify spatial variations in its glaciological character, and by extension that of the AIS during other Plio-Pleistocene (5.3–0.11 Ma) glacial maxima. Specifically we describe: 1) ice-sheet thermal regime at the LGM; 2) ice-sheet dynamics (velocities and pattern of flow); 3) erosion potential and patterns of likely sediment flux of the ice sheet at equilibrium; and 4) how changes in dynamics arise under perturbed environmental forcings and their implications for changes in erosion potential during glacial cycles. We discuss these results in the context of both marine and terrestrial geological studies, and assess the degree to which our numerical simulations agree or conflict with the available empirical evidence. We present new model simulations that attempt to resolve model–data mismatches, but our principal aim is to provide a robust glaciological framework for the interpretation of sedimentological and geomorphological records.