Comparing WEPP with USLE based models: The role of bare fallow runoff and soil loss plots

Abstract

There are many soil erosion models and model applications. However, as a general rule, models of rainfall erosion cannot fully model the complexity of the detachment and transport processes involved in soil erosion by rain so that it is crucial that soil erosion models are tested against experimental data. In developing the USLE, the designers recognised that the fundamental ability of a model to predict erosion in croplands began with its ability to account for soil losses from bare fallow areas under natural rainfall. Given this, any event-based model perceived to be a replacement for USLE-based models should be first tested for its ability to account for event soil losses from bare fallow areas under natural rain. Comparisons between the abilities of WEPP, RUSLE2 and the USLE-M to account for event soil loss on bare fallow runoff and soil loss plots leads to questions about the capacity of WEPP to model erosion on areas where some storms produce rills but others do not. One reason for this may lie in the fact that, in WEPP, sediment produced by raindrop-driven erosion is moved by flow-driven sediment transport to the outlet in situations where flow-driven sediment transport in channels does not occur. The modelling approach adopted by the designers of the USLE requires the veracity of any alternative erosion model to be established on bare fallow runoff and soil loss plots before focusing on erosion on vegetated areas.