Regression models for the evaluation of the rainfall factor with regard to climate change on the basis of monthly values

The R-factor as a measure of the erosivity of precipitation events is used to quantify soil losses using the Universal Soil Loss Equation (USLE). The calculation of the exact R-factor requires precipitation data with a high temporal resolution, which are usually not available widely. Due to this, regression models, such as the German Federal State related equation of DIN 19708:2017-08 -"Soil quality - Predicting soil erosion by water by means of ABAG" or spatially high-resolution radar rain data are used. In this study, two simple and practical, regional regression models for the Mansfeld-Sudharz district for the calculation of R-factors are presented. Regression equation 1 calculates the R-factor at the county level similar to the state equation, but based on 6 monthly precipitation factors. The results show that the equation has a significantly higher accuracy than the DIN equation for Saxony-Anhalt. Regression equation 2 is used to determine the changes in the R-factor on the basis of monthly precipitation change factors. The equation is specifically developed for the consideration of climate change issues. Validation using change signals from a regional climate model ensemble of the RCP8.5 scenario shows high model quality with a mean deviation of the DIN-R factor in the Near and Far Future of about 1 % and a correlation coefficient of greater than 0.9.