Coupling soil erosion sensibility and runoff connectivity to assess slope contributions to watershed sediment budget

International audienceStream water quality in grassland region is currently deteriorated by suspended sediment coming from soil erosion on cultivated areas located in the upstream parts. In the same way, landscape connectivity, spatial organization of hedges and the distance from sediment fluxes sources to points of measurement influence the global sediment budget. Therefore, well understanding the sedimentary transport of a watershed requires an accurate identification of the relationships between runoff capacity, soil erosion sensibility and runoff connectivity from local to global scales. In order to address this objective, a coupled modeling approach using SCALE (a large-scale assessment and mapping model of soil erosion hazard) and RuiCells (Runoff simulation on Cellular Mesh) is proposed. The SCALE model allows mapping sensitive areas based on soil characteristics and agricultural practices collected at parcel scale. This model guarantees a large accuracy on the local level of soil erosion sensibility. Runoff flow paths and connected areas, observed at local scale and validated in reality, are after integrated in the Geographical Cellular Automata RuiCells which takes into account topography and land use organisation. The impacts of man-made linear structures like hedges, roads and ditches networks are also combined to assess their influence on sediment transport. This study is specifically focused on three watersheds of similar size (~15km²) located in the northern-western part of France, in Europe. Results underline that cultivated areas indirectly connected by man-made structures are the most important while directly connected areas have a minor influence. The simulations obtained by RuiCells permit us to map flow pathways and to quantify sediment budget at the final watershed outlets. Variability between the three watersheds is discussed in relationships with their morphological structure, the spatial distribution of land use and with the efficiency of man-made structures