Effects of digital elevation model spatial resolution on distributed calculations of solar radiation loading on a High Arctic glacier

High-resolution airborne lidar data are used to produce digital elevation models (DEMs) of an arctic valley glacier (midre Lovenbreen, Svalbard) at resolutions of 2.5-2000 m, using three different interpolation schemes. These data are used in a distributed model of solar radiation loading for glaciers. When the mean of all lidar measurements within a DEM cell is used to calculate cell height, the differences between the finest- (2.5 m) and coarsest-resolution (2000 m) DEMs for the calculated annual whole-glacier spatial means of total potential direct-beam solar radiation, potential duration of direct-beam solar radiation, and intensity of potential direct-beam solar radiation are 20%, 56% and 23% of the 2.5 m DEM values respectively. A resolution change from 2.5 m to 200 m affects the whole-glacier spatial mean summer net solar radiative flux by an average of 5%, and the summer melt production from the glacier by an average of 3% compared with the 2.5 m DEM values, for the years 2001-03. These changes are largely driven by underestimation of shading by surrounding topography at coarser DEM resolutions. This dependency is reduced in the second and third interpolation schemes, especially at resolutions finer than 50m, which use the maximum lidar height measurement in some or all DEM cells. These results suggest that resolutions of 50 m are the coarsest that should be adopted in high-resolution glacier surface energy-balance models for glaciers of similar size and in similar topographic situations to midre Lovenbreen, and that the impact of DEM resolution on calculated solar radiation receipts can be reduced by an appropriate choice of DEM interpolation scheme.