Models of ice-sheet flow must be initialised by the specification of the topography of the bed beneath the ice. When the model is run, this topography will ensure that the model reproduces the correct pattern of ice-flow, the distribution of glaciers and ice shelves. Most modellers do not produce their own bed topographies but rely on products (such as the one for Greenland produced by ice2sea) produced using compilations of data collected over many decades and many expeditions. For some time the resolution of basal topographies has exceeded the demands of most model, but the development of a new-generation of models in recent years has called for a close examination of the impact of uncertainties on the results from models. Led by a group from the Laboratoire de Glaciologie and Géophysique de l’Environment in Grenoble, France; ice2sea has undertaken such a study using one of the most complete and sophisticated model of ice-flow yet devised. Their results make clear that observational glaciologists need to concentrate their efforts to collect new bed data in the coastal margin of the ice sheets, and that to ensure reliable results, considerably more data needs to be collected. Indeed, the target they set, if for the bed topography in coastal margins to be resolved to 1-km resolution. Only in a very few areas is this much data available, and to achieve this resolution over the entire Greenland and Antarctic ice sheets will probably require new technologies to be developed. For the present, resources must be focussed, and 1-km sub-glacial topographies will be acquired over specific target glaciers where change is most rapid and significant.
Publication: Durand, G., O. Gagliardini, L. Favier, T. Zwinger, and E. Le Meur (2011), Impact of bedrock description on modeling ice sheet dynamics, Geophysical Research Letters, 38, L20501-20507.