The Universal Soil Loss Equation (USLE) is presently one of the most widely used models for evaluating soil erosion. The rainfall amount is often used to estimate the R factor for Universal Soil Loss Equation(USLE) modeling. Daily, monthly, and annual rainfall data are commonly available. The purpose of the present study is to establish three models for estimating rainfall erosivity based on daily, monthly, and annual precipitation, respectively,and to identify the most suitable model in the Ailiao River watershed located in southern Taiwan. The data set consists of 2266 storm events, monitored by 6 rainfall stations there. The results show that the 6 stations indicate strong positive relationship (r2 > 0.75) between annual rainfall erosivity and annual rainfall with a 99% confidence level, which means the simplified estimation methods based on annual precipitation are useful for predicting long-term annual rainfall erosivity in most of locations in the Ailiao River watershed. The results also show that annual and monthly precipitation correlate with erosivity better than does daily precipitation based on the analysis results of root mean square error (RMSE) and mean absolute percentage error (MAPE) in the Ailiao River watershed.
Key Words : Annual precipitation, Rainfall erosivity index, Ailiao River watershed, USLE.