Assessment of soil erosion areas using a process-based overland flow modelling approach

Soil erosion is a complex process driven by the interaction between climatic factors, soil properties, topography, vegetation, and land use. It involves detachment, transportation, and deposition of soil particles due to surface runoff and wind, causing severe environmental and economic challenges. To manage erosion, several models ranging from empirical to process-based and hybrid approaches have been developed. For example, the most widely used empirical models is the Revised Universal Soil Loss Equation (RUSLE) which estimates long-term erosion rates but not reliable in short-term assessments. Process-based models, such as the Water Erosion Prediction Project (WEPP) and the European Soil Erosion Model (EUROSEM), simulate physical erosion mechanisms but require extensive data. Hybrid models like the Sediment Delivery Distributed (SEDD) and the Limburg Soil Erosion Model (LISEM) balance usability and mechanistic accuracy but face challenges in data-scarce or complex landscapes.

This study applied a hydraulic Overland Flow (OF) model to the Oltrepò Pavese region in north-western Italy, a geologically and hydrologically diverse area influenced by natural processes and human activities. In particular, the model was applied for a rainfall event with a return period of two year in three representative mountain catchments of the region: Scuropasso, Versa, and Ardivestra, characterised by mild to steep slopes, forested areas, rural settlements and vineyards. The OF model, based on the resolution of the Shallow Water Equations (2D-SWEs) calculates hydrodynamic variables such as flow depth and velocity. Erosion-prone areas were identified through literature empirical equations employed by using the OF model output, incorporating shear stress, stream power, and sediment transport capacity. To validate the OF model, the results were compared to those generated by the RUSLE. The comparative analysis was conducted to assess spatial overlap in erosion-prone areas between the two models. To ensure consistency, minimum erosion thresholds were applied to exclude areas non-relevant to erosion, such as water bodies, rocky areas, infrastructures, and forested zones which showed negligible erosion. The thresholds optimized the alignment of erosion-prone area estimations between the two models, revealing a significant degree of overlap and demonstrating the reliability of the OF model in determine prone-erosion areas. In addition, despite uncertainties in empirical formulations, the hydraulic OF model provided prone-erosion areas by using less input information than RUSLE. This study highlights the potential of integrating hydrodynamic modelling and empirical approaches to improve soil erosion assessments. Future advancements in model dynamics, land-use representation, and climate impact analysis are essential for addressing soil conservation challenges in diverse landscapes.

Acknowledgement: This study is part of the project NODES which has received funding from the Italian Ministry of University and Research (MUR) – M4C2 1.5 of PNRR funded by the European Union - NextGenerationEU (Grant agreement no. ECS00000036).