Master thesis defense by Signe Gevik

Title: Greenland Ice Sheet modelling: A study of calving parametrisations

 
Abstract: Climate models are important tools for predicting future changes of the global climate. Recently, coupled climate models have been developed to capture the feedback processes between the climate system components. The coupling of ice sheet- and atmosphere/ocean models is in an early stage, and optimization of the setup in such models is still a topic of research. This thesis investigates the Greenland Ice Sheet with the Parallel Ice Sheet Model (PISM), and its sensitivity to internal and external parameters in a coupled ice sheet-atmosphere setup with the global climate model EC-EARTH.

The project is divided into three parts. First, a sensitivity study is done on the ice dynamical parameters and model resolution. Parameters of the ice softness and basal conditions are examined, and a set of parameters is chosen for the remaining studies. A study on the importance of the climate boundary conditions is performed. Climate data from ERA-interim, RACMO and EC-EARTH is investigated, and the effect of data implementation method is studied. The model simulations experience a volume variation of 10-13% from the different forcings and implementations. The interpretation of climate forcings and treatment of surface melt in areas with temperature around the melting point, especially in the coastal areas, is the key variable determining the size of the ice sheet.

An initialization is performed to capture the historical temperature fluctuations in the ice sheet interior and obtain an accurate ice softness. The initialised ice sheet is adjusted to achieve a stable equilibrium with the EC-EARTH climate forcing, which results in an initial state 8.4% larger than the observed ice sheet.

Finally, the model sensitivity to marine dynamics is investigated. The five calving parametrisations implemented in the PISM model are studied, along with an additional frontal melt which is added to the surface mass balance of the model with a simple method. It is diffcult capturing dynamics of the GrIS outlet glaciers in a coarse resolution. Only at extreme parametrisations does the calving have a significant impact on the simulated ice sheet, with a thinning of the major marine glaciers Jakobshavn and Petermann exceeding 100 m. The frontal melt parametrisation provides additional ablation, especially in outlet glaciers where the dynamical calving schemes have little effect. The melt is added with a simplified method, and needs more investigation to quantify its effect in a large scale GrIS model. The results show that frontal melt is an important factor in the simulation of marine ablation, and can contribute to the further improvement of Greenland Ice Sheet models.