Simulating snow distribution using the Water Erosion Prediction Project (WEPP) model

Dongjie Shen

Snow distribution is an important phenomenon in winter hydrology. Successfully modeling snow distribution will help facilitate sound water resources management and planning, especially in cold regions where snow is the main available water resource. However, many existing watershed hydrologic models do not consider the process of snow drift, and those that do include this process generally require extensive input data and are often too complicated to apply. The Water Erosion Protection Project (WEPP) model is a physically-based, continuoussimulation, distributed-parameter model for watershed hydrology and erosion. WEPP includes a snow drift subroutine in its winter hydrology component. However, preliminary assessment revealed inadequacies of the subroutine. In this study the snow drift subroutine in the WEPP model was tested and modified through applications to two south-facing hillslopes under different tillage practices: the no-till (NT) slope within the Cook Agronomy Farm and the conventional tillage (CT) slope within the Clark Farm, five miles northeast of Pullman, WA. The original subroutine led to a substantial overestimate of snow depths for both the NT and CT slopes, and, in addition, a greater snow depth for the CT than for the NT, inconsistent with field observations in a previous study. Three major errors were found in the snow drift subroutine of the WEPP model; one was the miscoding of the input variable, the slope azimuth, as a constant (360°), the second was a mistake in coding the formula for calculating threshold velocity of ground snow drift, and the third was the use of an inadequate formula for calculating snow storage capacity. The modified subroutine with the coding errors corrected yielded adequate simulation results for snow depth that are agreeable with the field measurements as corroborated by both the Wilconxon tests (p-value of 0.697 for the NT hillslope and 0.486 for the CT hillslope) and Nash-Sutcliffe model efficiencies (0.91 for the NT and 0.31 for the CT). To further improve the snow drift subroutine in the WEPP model, future efforts should be devoted to applying and testing the model for other study sites under different hydrological and landuse and management conditions.

Simulating snow distribution using the Water Erosion Prediction Project (WEPP) model

Files and links (1)

Abstract

Metrics

Details