Does WEPP meet the specificity of soil erosion in steep mountain regions?

Abstract

We chose the WEPP model (Water Erosion Prediction Project) to describesoil erosion in the Urseren Valley (central Switzerland) as it seemsto be one of the most promising models for steep mountain environments.Crucial model parameters were determined in the field (slope, plantspecies, fractional vegetation cover, initial saturation level),by laboratory analyses (grain size, organic matter) or by the WEPPmanual (rill- and interrill erodibility, effective hydraulic conductivity,cation exchange capacity). The quantification of soil erosion wasperformed on hill slope scale for three different land use types:meadows, pastures with dwarf shrubs and pastures without dwarf shrubs.Erosion rates for the vegetation period were measured with sedimenttraps between June 2006 and November 2007. Long-term soil erosionrates were estimated by measuring Cs-137 redistribution, depositedafter the Chernobyl accident. In addition to the erosion rates, soilmoisture and surface flow was additionally measured during the vegetationperiod in the field and compared to model output. Short-term erosionrates are simulated well whereas long term erosion rates were underestimatedby the model. Simulated soil moisture has a parallel developmentcompared to measured data from April onwards but a converse dynamicin early spring (simulated increase and measured decrease in Marchand April). The discrepancy in soil water during springtime was explainedby delayed simulated snow cover melting. The underestimation of simulatedlong term erosion rates is attributed to alpine processes other thanoverland flow and splash. Snow gliding processes might dominate erosionprocesses during winter time. We assume that these differences leadto the general simulated underestimation of erosion rates. Thus,forcing erosion processes which dominate erosion rates in mountainousregions have to be implemented to WEPP for a successful applicationin the future.