Evaluating the performance of different empirical rainfall erosivity (R) factor formulas using sediment yield measurements

Abstract

The study aims to evaluate the performance of nine empirical rainfall erosivity (R) factor formulas at the Venetikos River catchment, Northwestern Greece. The goal is to select the most appropriate one, for the accurate estimation of soil erosion. Since high temporal resolution precipitation data (required for the analytical calculation of the R-factor, based on the EI30 index) are only available at the Spilaio station, and for a shorter time period (1973–82) than the one the study focuses on (1965–82), a direct validation of R-factor (thus of the empirical equation used for its estimation) can't be applied. Consequently, an indirect validation is attempted considering the catchment's sediment yield, given the convergence between the respective observed and modelled time series. The former was based on field measurements provided by the Greek PPC (Public Power Corporation), while the latter on the (annual; multi-annual) implementation of RUSLE, once for every different R-factor approximation. Provided that all other factors (K, LS, C, P) remain unchanged, the model's relative results (of each application against all others) are only depended on the alternative R-factor values. Regarding the latter, the approximation (thus R-factor equation) that performed best was the one using the Renard and Freimund (1994) [F-based] formula, by displaying the smallest deviations (%) against the observed measurements. All approaches allowed identification of the most susceptible to erosion areas.