Assessing effects of the choice of meteorological forcing datasets and downscaling methods on distributed snow simulations in a mountainous catchment

High resolution distributed snow models provide insights into the spatiotemporal variability of snow-related quantities, which is essential for understanding climate effects on the hydrologic cycle. However, the fidelity of these models is subject to meteorological forcing uncertainty induced by the selection of forcing data and the spatial downscaling method. In this study we examine the Utah Energy Balance model's sensitivity to meteorological forcing datasets and downscaling methods for a mountainous catchment, where complex terrain affects fine-scale processes and weather stations are scarce. The model is forced by multiple sets of meteorological input data derived from reanalysis products and observations at weather stations, differing in radiative energy flux estimation, downscaling method and orographic adjustment. Comparisons of model simulations reveal substantial difference in snow water equivalent and surface water input rate, particularly at high elevation. The results suggest that input uncertainty arising from the choice of forcing dataset and downsampling method is nonnegligible and thus needs to be accounted for in model calibration and subsequent applications for water resources management.