Ice Sheet Modelling

Recent interest in the response of the Antarctic Ice Sheet to changing climates has been an important factor in driving the development of improved glaciological ice-sheet models. As well as three-dimensional ice flow by internal deformation, these models now incorporate a temperature calculation throughout the ice sheet. Thermo-mechanical coupling allows a more realistic representation of ice flow with the additional benefit that basal sliding can be parameterised over areas of the bed that reach the pressure melting point. Ice-sheet models require inputs of bedrock elevation, initial ice thickness, sea level and climate. They can be used to provide quantitative information about ice flow and the response of an ice sheet to changes in the model parameter values or to a change in external forcing. Models are best used along with field data - both present day ice-sheet extent and past extents inferred from geological evidence. The ability of models to make predictions in areas of sparse data and for time periods which cannot be constrained by geological evidence means they are considered a valuable tool in reconstructing past ice-sheet configurations.

The first comprehensive glaciological modelling of the Antarctic Ice Sheet over a full glacial-interglacial cycle was published by Huybrechts (1990). This work has now been updated with several improvements to both the model and the climatic input (Huybrechts, 2002). An independent study with an alternative model formulation by Ritz et al. (2001) provides a useful comparison for the case of reconstructing the extent and thickness of Antarctica at the LGM. This model has also recently been forced with output from a Global Climate Model providing a further comparison (Philippon et al., 2006).

Both Huybrechts (2002) and Ritz et al. (2001) use time-dependent climatic input to force their models. In both studies a reference simulation is produced using paleoclimatic information derived from the Vostok ice core (Petit et al., 1999) to run the model through four glacial-interglacial cycles (~400ka). Free parameters in each model are then tuned to match the results to observations of the present day ice-sheet. The ice-sheet produced at the end of the reference run is then used as the initial condition for a range of further simulations to assess the model sensitivity to changes in parameter values and forcings.

0 0

Post a comment