The Effect of Topographic Shadowing by Ice on Irradiance in the Greenland Ice Sheet Ablation Zone

Accurately predicting runoff contributions to global sea level rise requires more refined surface mass balance (SMB) models of the Greenland Ice Sheet (GrIS). Topographic shadowing has shown to be important in the SMB of snow-covered regions, yet SMB models for the GrIS generally ignore how surface topography affects spatial variability of incoming solar radiation on a surface. In the ablation zone of Southwest Greenland, deeply incised supraglacial drainage features, fracturing, and large-scale bed deformation result in extensive areas of rough surface topography. This topography blocks direct radiation such that shadowed areas receive less energy for melting while other topographic features such as peaks recieve more energy. In this study, we quantify how shadowing from local topography features changes incoming solar radiation. We apply the ArcGIS Pro Solar Radiation Toolset to calculate the direct and diffuse irradiance in sunlit and shadowed areas by determining the sun's movement for every half hour increment of 2016. Multiple digital elevation models (DEMs) with spatial resolutions ranging from 0.06 to 5m were derived from fixed wing and quadcopter UAV imagery collected in summer 2016 and the ArcticDEM dataset. Our findings show that shadowing significantly decreases irradiance compared to smoothed surfaces where local topography is removed. This decrease is exponentially proportional to the DEM pixel sized with 5m DEMs only able to capture a small percentage of the effect. Applying these calculations to the ArcticDEM to cover a larger study area indicates that decreases in irradiance are nonlinearly proportional to elevation with highly crevassed areas showing a larger effect from shadowing. Even so, shading at higher elevations reduces irradiance enough to result in several centimeters snow water equivalence (SWE) per year of over-prediction of runoff in SMB models. Furthermore, analysis of solar radiation products shows that shadowing predicts albedo variability far better than a range of variables derived from UAV imagery mosaics including slope, aspect, elevation, or the distance to dark surface features. In summary, implementation of the effect of shadowing on irradiance should therefore be considered for accurate surface mass balance calculations for the Greenland ice sheet.