| Summary: | The surface mass balance is the main connection between the atmospheric climate change and the evolution of Greenland ice sheet over the next century. This thesis focuses on the development of surface mass and energy balance model for simulations of timescales above a century. The Bergen Snow SImulator (BESSI) needs to compromise between the necessary complexity to resolve the relevant physical processes and the computational costs of the model. There were three main studies published for the PhD focusing on model sensitivity, uncertainty assessment, transferability in space and time, and the modeling of the surface mass balance until the end of the current century. BESSI is an energy balance model that accounts for snow albedo decay, vapor fluxes and sub-surface water percolation and refreezing. The sensitivity of the surface mass balance towards the individual free model parameters was assessed for a cold and a warm period for the first publication of this thesis. The dominant factor during the warm period are uncertainties associated with the long-wave radiation and clouds, while during the cold climate of the last glacial maximum, sublimation and deposition cannot be neglected. BESSI provides useful SMB simulations over the entire Greenland ice sheet, but the uncertainties associated with the long-wave radiation are better reduced by relying on climate data input. The influence of the boundary climate conditions was studied next. The ice sheet is relatively stable to temporal variability changes, if the absolute range of change stays the same. Nevertheless, simulations based on a climatology instead of variable climate lead to a drastic overestimation of the surface mass balance. Climatologies have small amounts of daily snowfall, which lead to an increased snow albedo. A possible solution to obtain a good forcing for BESSI, and likely other surface mass balance models, is by distributing the precipitation based on the real temporal and spatial precipitation patterns. After the thorough sensitivity and ...
|
|---|