『Summary
　The Universal Soil Loss Equation (USLE) is the most widely used and misused prediction equation in the world. although it was designed to predict long-term average annual soil loss, it has the capacity to predict event soil losses reasonably well at some geographic locations and not well at others. Its lack of capacity to predict event erosion is highly influenced by the fact the event rainfall-runoff factor used in the USLE and its revisions (RUSLE, RUSLE2) does not consider runoff explicitly. While including direct consideration of runoff in the event rainfall-runoff factor improves the capacity to predict event erosion when runoff is measured, that capacity is reduced by inaccurate runoff prediction methods. Even so, the predictions may be better than when the traditional event rainfall-runoff factor is used if the rainfall-runoff model used to predict runoff works reasonably well. Direct consideration of runoff in the rainfall-runoff factor may improve the ability of the model to account for seasonal effects. It also enhances the ability of the model to account for the spatial variations in soil loss on hillslopes which result from spatial variations in soil and vegetation. However, the USLE model will not provide a capacity to account for deposition taking place on concave hillslopes unless it is coupled with an appropriate sediment transport model, as in done in RUSLE2. Changing the basis of the event rainfall-runoff factor has consequences on a number of the other factors used in the model, in particular new values of the soil erodibility factor need to be determined. Using runoff values from cropped areas is necessary to account for differences in infiltration capacities between vegetated and tilled bare fallow areas, but requires re-evaluation of the crop factors.

Keywords: Rainfall erosion; Hillslopes; Spatial and temporal variation』

Contents
Introduction
Runoff and event erosivity
　The impact of erosivity indices on other factors in the USLE model
　The K factor
　The C factor
　The P factor
　The L and S factors
　　Non uniform runoff production
　　Hillslope form
　Event erosion and the unit plot approach
　Model predictive capacity
　　Determining EI30 and the R factor
　Accuracy of runoff predictions
　Accuracy of predictions of soil loss by more process-based models
Conclusion
References