Strategies to Improve WEPP Snowmelt Simulations in Mountainous Terrain

Abstract

The ability to predict snow accumulation and melt in complex vegetated terrain is critical for effective water supply, water quality, and forest land management. Accurate numerical simulation of these processes is important for erosion and sedimentation modeling in the mountainous western U.S., where snow processes dominate the winter phase of the hydrologic cycle. This work assessed approaches to adjust meteorological data and canopy parameterizations to improve Water Erosion Prediction Project (WEPP) snowmelt model predictions and to identify areas in which to focus future model development efforts. Four data modification approaches were used that represented a range of operator cost and effort. Approaches included direct application of: (1) nearby climate data, (2) empirically derived precipitation adjustment factors based on snowpack measurements, (3) custom climate data collected at the site of interest, and (4) adjusted canopy cover factors. Site-specific climate and calibrated canopy parameters produced the best simulation of snowcover duration and melt rate but underestimated peak snow water equivalent (SWE) by approximately 1%, 34%, and 40% in clearcut, partially cut, and fully forested sites, respectively. The best simulation produced time to snowpack depletion from peak SWE of 11, 15, and 45 days earlier than observed in clearcut, partial cut, and full forest, respectively. Results illustrate the difficulties of simulating snowpack dynamics with WEPP in forested mountainous systems of the interior Pacific Northwest where air temperatures are close to 0C during snow deposition. Although improvements in simulation accuracy are needed in forests, the proposed methods can improve confidence in WEPP snowpack simulations or be applied to other models to improve predictions in complex snow-dominated systems.