Spatial Modeling of the K factor for two sub-catchments with different tillage and grazing Case study: loessial paired sub-catchments in the north-east of IRAN

Soil erosion is the most extensive process of land degradation worldwide that has negative impacts on sustainable development. In many hilly cultivated areas, erosion caused by tillage is at least as important as water induced soil erosion. In particular, the effect of tillage erosion on fertile loess soils is of high importance. We examined the effect of tillage on intensively cultivated and hilly loessial sub-catchments in the north-east of the Golestan Province, Iran. In this study, different strategies to assess the soil erodibility factor in two sub-catchments with different management policies, which include tillage and grazing, are considered. Different interpolation approaches are compared using cross-validation. For the first sub-catchment, all spatial modelling strategies for erodibility in terms of the K-equation result in the same mean value. However, this is not the case for the second sub-catchment. Tillage and grazing have a substantial effect on soil erodibility, but different strategies of implementing the K-equation may not give this result. Both sub-catchments have the same range of local uncertainty. We conclude that in case of mapping and assessing uncertainty, two different modelling strategies should be used.