Rupert Gladstone
Arctic Centre, University of Lapland
Finland
The Antarctic Ice Sheet (AIS) provides currently the greatest source of uncertainty in future sea levels on a century timescale. It comprises kilometers thick land ice, some of it resting on bedrock well below sea level (termed "marine ice sheet"), flowing to the ocean through a combination of internal deformation and sliding at the bed. Interactions between the ice sheet and its bed, and between the floating part of the ice sheet ("ice shelves") and the ocean, are complex and crucial to the dynamical behaviour of the AIS.
A number of challenges face computer models attempting to simulate the behaviour of this system. At one point it was proposed in the glaciological community that real ice sheets may exhibit neutral equilibrium behaviour in relation to attempts to simulate grounding line movement in computer models. This was later modified to the suggestion that neutral equilibrium occurred only as a numerical artifact in computer models and not in the real system. I will demonstrate that this resolution dependent behaviour in models is not in fact neutral equilibrium, and I will point out some consequences of this.
I will show that this resolution dependence in models can be mitigated by orders of magnitude through choice of "sliding relation", which governs AIS basal resistance and sliding velocity. I will present preliminary outputs from an AIS inversion for basal resistance, in which a cost function based on observed-model upper surface velocity discrepancy is minimised. These outputs support a more model-friendly choice of sliding relation than is typically used, crucially featuring a dependence on basal water pressure.
Finally I will introduce a new coupling system designed to simulate ice-ocean interactions, which are crucial components in determining the stability of marine ice sheets due to the evolution of ocean-induced melt rate distribution on the underside of the AIS floating ice shelves.