Estimation of rainfall erosivity in Piedmont (Northwestern Italy) by using 10-minute fixed-interval rainfall data

Rainfall erosivity index (EI30) is widely used in soil erosion models for predicting soil loss. This index consists in the product between the maximum intensity of 30-min rainfall and the total kinetic energy of a precipitation event. The main goal of this study was to characterize the soil erosion in Piedmont (Northwestern Italy), studying the magnitude, frequency, and trends of rainfall erosivity. Rainfall erosivity for twelve stations well distributed over the whole region were firstly computed on the basis of 10-min time-resolution rainfall data using a continuous 17-year series of daily rainfall events. For each station the equation to predict EI30 from daily rainfall data was calculated, and, using the Nash and Sutcliffe (1970) model-efficiency, the relationships between real EI30 and modeled EI30 was validated. The rainfall erosivity model was applied to the long term daily rainfall series of the selected stations, to create annual and seasonal erosivity time series for the climate normal period 1986–2015. Afterwards, the Mann–Kendall non-parametric test statistic to detect time trends in the rainfall erosivity time series was applied. The results have led to the conclusion that the annual rainfall erosivity should have experienced mixed trends in most of the study area, although more than half of the stations did not show a statistical trend.