Master´s thesis defense by Pia Nielsen-Englyst

During the 21st century, the Greenland Ice Sheet has started to lose more mass than it gains due to increasing temperatures. Being able to quantify this mass loss from the Greenland Ice Sheet is essential in order to provide projections for the potential environmental, economic, and social repercussions, as well as implications of the mass loss around the world.

The primary energy source for melting snow and ice is absorption of shortwave radiation. The amount of shortwave radiation, received at Earth's surface, depends on the incident radiation and the albedo of the surface. The albedo is the ratio of reflected to incoming solar radiation at the surface and is one of the most crucial climate parameters. There are large differences between the albedo of fresh snow, old snow, melting snow and bare ice, and to get the timing, magnitude and spatial distribution of surface melting in a climate model right, it is vital to capture these differences. The main motivation behind this thesis is to capture these spatial and temporal differences in the surface albedo to create a better and more physically based albedo parameterization for the Greenland Ice Sheet. The focus has been on the snow albedo and there has been no attempt to parameterize the differences in the ice albedo.

In the new albedo parameterization the snow albedo decays exponentially over time with a decay rate dependent on the temperature. The albedo is only allowed to increase following a snowfall event. The new choice of albedo formulation has been tested and fitted against in-situ data provided by PROMICE weather stations. The optimal albedo formulation has been implemented and tested in a stand-alone version of the surface mass balance scheme of the Danish regional climate model HIRHAM5 covering all of Greenland with 5 km resolution.

In order to validate the new albedo parameterization a comparison of variables at temporal and spatial scale against observations has been made. Comparisons have been made with MODIS satellite derived albedo data and in-situ observations by PROMICE. As a third evaluation method, NSIDC (NASA National Snow and Ice Data Centre) data has been used to compare the simulated melt against melt data based on satellite measurements. The sensitivity to the albedo parameterization has been tested in detail by running the same albedo scheme with different control parameters to investigate the influence on e.g. surface mass balance, melt extent, and duration. Finally, a number of output parameters have been compared to outputs from the previous albedo parameterization, already implemented in HIRHAM5, in order to determine the overall performance of the new albedo parameterization.