Abstract:Daily rainfall data of 1951to 2010obtained from Shapingba weather station was used to build simple algorithm of rainfall erosivity.Accordingly,the prediction accuracy of the simple algorithm was analyzed,the daily rainfall corresponding to the highest prediction accuracy was determined,and the temporalcharacteristics of annual rainfall erosivity and rainfall parameters were explored.The main results are as follows:(1)There is an exponential relationship between annual rainfall erosivity and annual rainfall.The annual rainfall with daily rainfall of≥25mm presents the closest relation with annual rainfall erosivity.Additionally,the accuracy for annual rainfall algorithm using daily rainfall of≥25mm in the prediction of rainfallerosivity is better than the algorithms using other levels of daily rainfall.(2)The changes of annual rainfallerosivity and rainfall parameters with time in the year from 1951to 2010are not significant.But for a certainperiod of time,the correlation coefficient between time and the annual rainfall with daily rainfall of≥25mm is similar to that between time and annual rainfall erosivity.The responses of soil erosion to climate changecan be analyzed by using the annual rainfall with daily rainfall of≥25mm as an alternative indicator.The results from the research can be used for reference in enhancing the prediction accuracy of the simplified modelfor annual rainfall erosivity and understanding the responses rainfall erosivity to climate change and regionalsoil and water loss control.