| Summary: | Ice sheet models are the only physically-based tools that allow us to simulate the future evolution of the AntarcticIce Sheet, including its contribution towards changes in global sea level. However, due to limitations in ourunderstanding of ice sheet dynamics, modelling is an inherently uncertain exercise. A typical approach towardsoptimising ice sheet models is to "tune" key physical parameters by finding the values that give the most realisticsimulations of the present-day ice sheet, based on criteria such as ice sheet geometry or ice velocity. However, thisapproach assumes that there are no errors in the boundary conditions being used to drive the models. Here, we use the Parallel Ice Sheet Model to explore the sensitivity of the simulated Antarctic Ice Sheet tothe available geothermal heat flux (GHF) datasets. We find that the choice of GHF is a significant source ofuncertainty, leading to basin-wide differences in excess of 1000m in the simulated ice thickness. Using differentGHF datasets to drive the model, we then "tune" it by determining the optimal values of key physical parameters.We show that the parameter combinations obtained are sensitive to the choice of GHF. Our results highlight the importance of GHF in ice sheet modelling. Reliable GHF estimates are critical tooptimising numerical models of the Antarctic Ice Sheet and, therefore, to reducing uncertainty in projections offuture global sea level rise.
|
|---|