Comparison of Measured and WEPP Predicted Runoff and Soil Loss for Midwest Claypan Soil

Abstract

Runoff and soil loss predictions from the Water Erosion Prediction Project (WEPP) Hillslope model were compared to measured losses from continuous cultivated fallow and continuous corn and soybean cropping systems under conventional, chisel, and no-till tillage methods. Runoff and soil loss data were collected over an 11-yr period (1983-1993) from the study plots on Mexico claypan soil near Kingdom City, Missouri. For all treatments, the model predicted runoff and soil loss reasonably well for the wet years with annual runoff >200 mm but greatly underpredicted for the dry years with annual runoff <100 mm. During the dry seasons, most of the rainfall events occurred after a long dry period and the model overestimated runoff from several events when little or no runoff was measured. When averaged by tillage treatment, mean annual runoff predictions for continuous corn and continuous soybean were within 15% of those measured. The model overpredicted soil losses from continuous corn and continuous soybean by 22 and 87%, respectively. Both measured and predicted runoff and soil losses from continuous soybean were higher than from continuous corn. When averaged over crop, the model slightly overpredicted mean annual runoff for conventional, chisel, and no-till systems. The model did not show appreciable difference in annual runoff prediction for the different tillage systems. Predicted soil losses for conventional, chisel, and no-till tillage systems were 50, 67, and 29% higher, respectively, than those measured. Runoff and soil loss data were also analyzed by cropstage periods. A tillage year was divided into rough fallow (F), seedbed (SB), rapid growth (P12), reproduction and maturation (P3), and residue (P4) periods. Runoff predictions for the F and SB periods were reasonable. The model overpredicted runoff for the P12 and P3 periods and underpredicted for the P4 period. The model overpredicted soil loss from all periods except for the F period. For continuous corn and soybean cropping systems, more than 50% of the total measured and predicted tillage year soil loss occurred during the SB period. The model was also tested for seven large events that occurred during the study period. Predicted runoff and soil loss for these events compared well to those measured.