3.2 Estimated soil erosion rates under climate scenarios

The simulated SE rates were averaged for the whole watershed over a period of 20 years providing a value of 3.3, 2.9, and 3.0 t/ha/year for the HIST, RCP 4.5, and RCP 8.5 scenarios, respectively. Maximum SE rates were also estimated providing a value of 25.3, 23.1 and 26.1 t/ha/year for the HIST, RCP 4.5, and RCP 8.5, respectively. The average value of SE for HIST, RCP 4.5 and RCP 8.5 in each HRU was used as a single point estimate to obtain a spatial distribution of the SE phenomenon over the entire basin, applying the kriging interpolation method. The spatialized data were grouped in 5 classes of SE susceptibility following the classification proposed by Panagos et al. (2015a): very low (SE ≤ 1.0 t/ha/year), low (1.0 <SE≤ 2.5 t/ha/year), medium (2.5 <SE≤ 4.0 t/ha/year), high (4.0 <SE≤ 5.5 t/ha/year), and very high (SE > 5.5 t/ha/year). The susceptibility maps of the basin are showed in Figure 6. The SE susceptibility of the area during the reference period (HIST) shows the highest values in the central and southern parts of the study site, while the lowest values are estimated for the eastern border of the basin and in the northern area (Fig. 5a). In future conditions (RCP 4.5 in Fig. 5b and RCP 8.5 in Fig. 5c) the distribution of the SE susceptibility remains similar to the one of the HIST scenario, in terms of spatial distribution of the areas most prone to SE risk (Fig. 5a). In all cases, the portions of the basins most susceptible to SE are located in correspondence to Leptosols (Fig. 2c). The most interesting results were observed calculating the patterns of SE change between the future projections (RCP 4.5 and 8.5) and the HIST scenario (Fig. 6), obtained using a spatial difference through raster calculator in GIS environment.