Evaluating the WEPP Rangeland Hillslope Model Using Cesium‐137 Estimated Spatial Erosion Data

Abstract

Core Ideas WEPP rangeland model was evaluated using spatial erosion data for the first time. WEPP and 137Cs methods predicted similar low soil erosion rates. Spatial erosion patterns predicted by WEPP and 137Cs methods agreed well. WEPP predicted soil erosion patterns well along rangeland hillslopes. Lack of spatially distributed soil erosion data has greatly hampered the development and improvement of process‐based soil erosion models. The 137Cs method has been widely used to estimate soil erosion and redistribution but has seldom been used to evaluate process‐based erosion models. This study evaluated spatial erosion patterns predicted by the Water Erosion Prediction Project (WEPP) model using 137Cs‐estimated spatially distributed erosion data and quantified the effects of slope length on erosion along rangeland hillslopes. Eight rangeland hillslopes in midwestern Oklahoma were selected. The WEPP rangeland model was parameterized with observed soil and vegetation data. One downslope transect was sampled for each hillslope. At each sampling position, seven samples along a contour line were composited to represent the 137Cs inventory for that slope position to reduce random spatial variation. Both methods predicted low soil erosion rates, with the maximum erosion rates being near 2 Mg ha−1 yr−1. The erosion rates predicted by the two methods were significantly correlated at P = 0.01, and the medians were not significantly different; however, a paired t‐test showed that WEPP predicted significantly lower erosion rates than the 137Cs method. More importantly, the WEPP‐predicted soil erosion patterns in response to slope length agreed very well with those predicted by the 137Cs method. The loci of the maximum and minimum erosion rates were similar between the two methods for seven transects. The WEPP rangeland model predictions for the spatial patterns of relative soil erosion intensity along a hillslope compared favorably with 137Cs method estimates.