Assessing soil erosion rates in vineyards using the RUSLE model: an insight on the role of the topographical factor (LS)

Soil erosion is an important aspect, especially in sloping areas such as hilly and mountain regions. Here agriculture often relies on row-crops, represented by orchards and vineyards. In this context, certain ground cover types can contribute to reduce soil losses by erosion. The Revised Universal Soil Loss Equation (RUSLE) is one of the most widely used models for soil erosion rates prediction. It can be successfully applied at both plot and wide area level (e.g., catchment or region). RUSLE considers different factors to describe erosion risk: rainfall erosivity (R), soil erodibility (K), cover and management (C), slope length and steepness (L and S, often considered jointly), and support practices (P). Among those, the topographical LS factor is one of the most difficult to define; numerous algorithms have been developed and applied in various studies that possibly rely on GIS tools involving Digital Elevation Models (DEMs). In this work, operator-measured LS values were compared with those obtainable using different algorithms available in SAGA GIS. Comparison was achieved with reference to a hilly study area located in the Alto Monferrato (province of Alessandria, NW Italy). DEMs having different geometrical resolution (5m, 10m and 25m) were considered while testing algorithms performances. Algorithms proved to generate significantly different LS values from the operator-measured ones. Nevertheless, the adoption of an automated procedure allows to reduce computational times and to minimize human error. A linear regression-based approach was used to test model performances depending on algorithm settings and DEM resolution. The best results (R2 >= 0.8) were obtained using the DEM with a higher resolution, independently from the adopted algorithm. Additionally, this made possible to map the spatial variability of LS factor within the field.