Spatial and temporal variability of rainfall erosivity factor for Switzerland

Rainfall erosivity, considering rainfall amount and intensity, is an important parameter for soil erosion risk assessment under future land use and climate change. Despite its importance, rainfall erosivity is usually implemented in models with a low spatial and temporal resolution. The purpose of this study is to assess the temporal- and spatial distribution of rainfall erosivity in form of the (Revised) Universal Soil Loss Equation R-factor for Switzerland. Time series of 22 yr for rainfall (10 min resolution) and temperature (1 h resolution) data were analysed for 71 automatic gauging stations distributed throughout Switzerland. Regression-kriging was used to interpolate the rainfall erosivity values of single stations and to generate a map for Switzerland. Latitude, longitude, average annual precipitation, biogeographic units (Jura, Midland, etc.), aspect and elevation were used as covariates, of which average annual precipitation, elevation and the biographic unit (Western Central Alps) were significant (p < 0.01) predictors. The mean value of long-term rainfall erosivity is 1330 MJ mm ha(-1) h(-1) yr(-1) with a range of lowest values of 124 MJ mm ha(-1) h(-1) yr(-1) at an elevated station in Grisons to highest values of 5611 MJ mm ha(-1) h(-1) yr(-1) in Ticino. All stations have highest erosivity values from July to August and lowest values in the winter months. Swiss-wide the month May to October show significantly increasing trends of rainfall erosivity for the observed period (p < 0.005). Only in February a significantly decreasing trend of rainfall erosivity is found (p < 0.01). The increasing trends of rainfall erosivity in May, September and October when vegetation cover is scarce are likely to enhance soil erosion risk for certain agricultural crops and alpine grasslands in Switzerland.