My research is exploring the sensitivity of the Greenland Ice sheet models with respect to their boundary conditions. Because climate models do not take the contribution of sea level rise from ice sheet mass loss, ice sheet models are needed and used to estimate this. The sensitivity of ice sheets varies with the different boundary conditions which are applied to the model equation, and so yield different amounts of mass loss, and hence sea level rise.
Differences in surface and bed geometry, as well as atmospheric forcing boundary lead to the differences observed in the mass balance (accumulation – ablation) of the ice sheet. The fundamental equations used in models are the ice thickness and temperature equations. The EISMINT benchmark was set up to compare models using different geometries and accumulation rates (Huybrechts 1996). Because changes in boundary conditions give different results, this will be done to investigate the sensitivity of models.
Heimbach (2009) employed an adjoint model of the SICOPOLIS ice sheet model to improve its sensitivity. Adjoint modelling is an alternative to finite differences to solve the equations by getting their derivatives, with respect to their boundary conditions. This method will be adapted and used in the 3 Dimensional thermomechanical Glimmer ice sheet model.
A part of the project will involve using the BLUE ice cluster to perform parallel programming and run Glimmer.
The overall aim of the project will be to address the spatially-distributed precision and coverage needed datasets, particularly those of basal topography, to get stable results and to advise on future data collection requirements for the improvement of model sensitivity.
- Heimbach, P. a. B., V. (2009). "Greenland ice sheet volume sensitivity to basal, surface, and initial conditions, derived from an adjoint model."
- Huybrechts, P., Payne, T. (1996). "The EISMINT benchmarks for testing ice-sheet models." Annals of Glaciology 23: 1-12.
Supervisors: Prof. Tavi Murray, Dr. Ian Rutt