Rainfall
erosivity shows the potential ability of the soil loss caused by rainfall and it is very important for predicting
soil loss quantitatively. Five models to estimate
rainfall erosivity using average annual rainfall、average monthly rainfall、annual rainfall、monthly rainfall or daily rainfall were compared from the data of 66 weather stations in China. The daily rainfall
Erosivity model was obviously better than the other four models by using annual or monthly rainfall. For daily rainfall erosivity model,the average relative error of estimating the average annual rainfall erosivity was 4 2% and the maximal relative error was 24 1%. With the set of recommended parameter values of the model,the average relative error using estimated parameter values increased to 17 9% and the maximal relative error decreased to 87 1%. For four models using annual or monthly rainfall, the average relative error estimating average annual rainfall erosivity was from 32 1% to 46 3% and the maximal relative error was from 235 4% to 569 2%. Moreover, the average monthly rainfall model was best among the 4 models. The performance of each model was better in the region where the rainfall was abundant than where not.