Evaluation of interrill component of the WEPP model for three contrasting soil types in Iran

Abstract

Measured interrill erosion rates were compared to soil loss predicted by the proposed equations in WEPP (Water Erosion Prediction Project) in a laboratory study for three contrasting soil types from Iran. The soils used in this study were: a calcareous clay loam from a cultivated field, a well aggregated clay loam from a rainforest, and a dispersive sandy loam soil without any defined stable aggregates. Mean weight diameters (MWD) of the aggregates were 0.53, 2.04 and 0.20 mm for the three soils respectively. The soils were subjected to simulated rainfall at different rates using a 1ױװ.1 m drainable flume at different slopes. The interrill component of the WEPP model was evaluated with and without calibration. Baseline interrill erodibility of the soil was calculated using WEPP-recommended equations and measured soil properties, when the model was evaluated without calibration. A Jack knifing method was used for calibrating the model. The coefficients of efficiency were respectively -14.51, 0.45 and -1.49 without calibration (using soil property-based erodibility values) and 0.89, 0.94 and 0.83 using calibrated erodibility values for the three soil used. These results indicated that WEPP-recommended equations for predicting interrill erodibility using soil properties are not applicable outside their US database and therefore the model calibration is essential. Results also showed that the model tends to under-predict the higher values of interrill erosion. Contribution of flow driven erosion processes in interrill areas especially at steep slopes and high rainfall rates in conjunction with eventual interaction between flow and rain driven processes could be the reason of this bias. Ignoring the effect of water depth on interrill erosion is probably one of the most important problems with WEPP model.