Estimating Soil Surface Roughness by proximal sensing for soil erosion modelling implementation at field scale

Soil Surface Roughness (SSR) is a physical feature of soil microtopography, which is strongly influenced by tillage practices and plays a key role in hydrological and soil erosion processes. Therefore, surface roughness indices are required when using models to estimate soil erosion rates, where tabular values or direct measurements are typically used. Field measurements often imply out-of-date and time-consuming methods, such as the pin meter and the roller chain, providing inaccurate indices. A novel technique for SSR measurement has been adopted, employing an RGB-Depth camera to produce a small-scale Digital Elevation Model of the soil surface, in order to extrapolate roughness indices. The values obtained for SSR indices were implemented in the MMF (Morgan-Morgan-Finney) and ORUSCAL (Orchard RUSLE CALibration) models, to validate the reliability of the proposed methodology by comparing the models' results for sediment yields with long term soil erosion measurements in sloping vineyards in NW Italy.