4.2. Comparison with global and local studies on soil erosion estimation
Using RUSLE, the global scale produced annual soil erosion maps (2001 and 2012) at the 25 km spatial resolution were produced and reported by Borrelli et al. (2017). At the national scale and seven administrative units of Pakistan; the globally produced 2012-annual soil erosion was compared with the 2015-annual soil erosion developed in the study. For the spatial and statistical comparison, the produced soil erosion 2015 map was resampled at 25 km. On the national scale as well as in each administrative unit, the global study reported higher values and spatial variability as compared to national produced soil erosion estimations and maps (Table 7 & Figure 9). In few areas of Punjab, Sindh, Balochistan, and Gilgit-Baltistan, in the global soil erosion map, no data observed, even where this study reported low (<1 ton/ha/year) annual soil erosion, the global study reported medium (1-5 ton/ha/year) annual soil erosion. For the annual soil erosion, within Pakistan at the watershed level and basin-scale, using the RUSLE, four studies over the five sites were reported, which were compared with this study by comparing mean soil erosion rate (ton/ha/yr) (Figure S1). Over the Rawal watershed and the Simly watershed we observed the finding of this study are matching well while in the Ghabbir watershed, Potohar region and Soan river basin, this study reported very low mean annual soil erosion values (ton/ha/year) as compared to sub-national scale studies (Table S5). This might be because of input data sources, spatial resolutions, variation in factor equations, and study scales.