Evaluating Wind Fields for Use in Distributed Snow Models

Abstract

Mountain winds are the driving force behind snow accumulation patterns in mountainous catchments, making accurate wind fields a prerequisite to accurate simulations of snow depth for water resource forecasting. Here we evaluate calculated wind fields using a new method for inferring the wind direction from snow depth patterns. This method leverages established relationships between snow depth and wind direction and provides the dominant wind direction at 260 locations in the Tuolumne River Watershed, CA. We compare these inferred wind directions to wind fields derived from combinations of coarse data and downscaling schemes. Coarse data come from meteorological towers, NLDAS, or HRRR data, and downscaling schemes tested include MicroMet, WindNinja, and bilinear interpolation. All downscaled wind fields replicate the south-west winds suggested by snow-depth-inferred wind directions. We then use these wind fields to force SnowModel, which contains the wind-transport scheme SnowTran. We find that wind fields derived from meteorological observations rarely achieve high wind speeds necessary for wind redistribution of snow. NLDAS data are derived from a 32 km DEM which smooths the Sierra Nevada range and results in low wind speeds over isolated peaks. Wind fields created by bilinearly interpolating coarse data do not contain discontinuities in wind speed necessary for deposition of snow. Two wind fields derived from 3 km HRRR data and downscaled with respect to terrain produced snow depth maps that best matched observations of snow depth from airborne LiDAR. We recommend 3 km resolution gridded wind data downscaled with respect to terrain as input to distributed snow models.