Application and Evaluation of a Snow Energy and Mass Balance Distributed Model in the Merced and Tuolumne River Watersheds of the Sierra Nevada, California

Abstract:

Characterization of snow accumulation and melt in forested environments is a key component of the hydrological cycle in mountainous regions, yet is often over-simplified in hydroecological watershed modeling due to computational limits and paucity of input data. A key first step in addressing this deficiency is to assess potential improvements in predicting snow water equivalent using a land surface model (full energy and mass balance of snow pack) over a sufficiently large geographic area to address societal questions centered on forest change and water yield of river basins. The snow energy and mass balance model ISNOBAL was applied over a 14,300 km² region encompassing the Merced and Tuolumne River watersheds in the Sierra Nevada at a 1-hour time step and 100-meter spatial resolution for water years 2010-2014. Results show a bias toward a slight over-estimation of snow water equivalent when compared to manual snow course measurements, though cumulative melt was consistent with summed river gage data and independent estimates of evapotranspiration. Ratios of gaged runoff to water available for runoff from the base of the snowpack ranged from 0.45 to 0.59 for three different gages. Modeled results were compared to independently-derived estimates of snow water equivalent from MODIS Snow Covered Area reconstruction and airborne LiDAR estimates of snow depth, highlighting potential deficiency in SCA reconstructions, and the use of LiDAR derived snow surface maps providing snow distribution patterns in physically based models. Further, ISNOBAL highlighted trends in snow distribution patterns and snowmelt timing in an average (2010), above average (2011), and below average (2014) year.