Characterizing the spatial snow depth distribution in mountainous terrain

Spatial snow depth information is an essential component for many applications such as the computation of snow melt rates in hydrologic applications or the energy balance in meteorological or climate models. Due to computational constraints, small-scale distributed modeling is rarely feasible for large regions. This is especially true over mountainous, complex terrain. Sub grid parameterizations for spatial snow depth distributions are therefore particularly valuable in the context of larger-scale applications. Past research has shown that for rather homogeneous landscape units the pre-melt spatial distribution of snow depth can be approximated by a log-normal distribution. Furthermore, seasonally recurring snow accumulation patterns have been reported, mostly shaped by terrain and wind. We focus on heterogeneous landscape units in order to investigate the impact of topographic parameters on sub grid snow depth distribution at peak of winter. Here we analyze a new data set with a particular view on developing a sub grid parameterization of snow depth for operational hydrologic applications and verifying previously developed models. Snow depth data obtained from an airborne laser scanning campaign in the Wannengrat Alpine catchment as well as from a unique opto-electronic scanning data set (Sensor ADS 80, Leica Geosystems) in the larger Dischma catchment (40km2) are used. Both areas are located above Davos, in the eastern Swiss Alps. A sub grid parameterization of snow depths variability at peak of winter will allow to compute snow covered area as a function of spatial melt rates during the ablation period, an issue which is also relevant for the parameterization of sub grid albedos in other large-scale models