2.4.4. Land Use and Land Cover: A Particularly Rich Natural Limnological Setting

The land occupation map of the Claise was obtained by on-screen digitizing of aerial photography at 0.5 m resolution. Such fine scale was used in order to ensure a detailed representation of the study area, particularly to account as accurately as possible for the scale of ponds.

For the CORINE model, the obtained land occupation map was reclassified into fully protected (forest, permanent pasture, and scrublands) and not-fully protected (cultivated or bare land) areas as demonstrated in Table 3.


**Table 3.** Numerical distribution of the land occupation setting of the Claise basin.

In the case of SWAT, land use and land cover classes were reclassified into SWAT's classes. Particular attention was given to the water (in SWAT terms WATR) classes. As reported by Jalowska and Yuan (2019) [38], the reason for this is that though SWAT allows the creation of HRUs with WATR, water bodies should be modeled either as reservoirs or ponds. Almendinger et al. (2014) and Jalowska and Yuan (2019) [38,48] have shown that an accurate representation of basin processes requires

the integration of the SWAT model's impoundments function. Furthermore, Wang et al. (2008) [77] highlighted the importance of considering impoundments, such as ponds, by testing scenarios of impoundment integration versus impoundment disregarding in a basin covered by only 3% of impoundments. Their results confirmed that simulations were considerably affected even with this small cover. In the same manner, Jalowska and Yuan (2019) [38] simulated different scenarios following integration or absence of impoundments like reservoirs rather than just a normal "water" land occupation class. They noted that disregarding impoundments leads to a series of uncertainties starting by an inaccurate SWAT performance, which in turn leads to inefficient calibration efforts, and overall inaccurate model performance.

After computing the slope, erosivity and soil erodibility indices, the potential soil erosion risk map was constructed. By combining the potential soil erosion risk map to the vegetation cover layer in the "raster calculator" tool and crossing each ones indices, the actual soil erosion risk map was obtained.
