Impact of model simplifications on soil erosion predictions: application of the GLUE methodology to a distributed event-based model at the hillslope scale

International audience; In this paper, we analyse how the performance and calibration of a distributed event-based soil erosion model at the hillslopescale is affected by different simplifications on the parameterizations used to compute the production of suspended sediment byrainfall and runoff. Six modelling scenarios of different complexity are used to evaluate the temporal variability of thesedimentograph at the outlet of a 60 m long cultivated hillslope. The six scenarios are calibrated within the generalized likelihooduncertainty estimation framework in order to account for parameter uncertainty, and their performance is evaluated againstexperimental data registered during five storm events. The Nash–Sutcliffe efficiency, percent bias and coverage performanceratios show that the sedimentary response of the hillslope in terms of mass flux of eroded soil can be efficiently captured by amodel structure including only two soil erodibility parameters, which control the rainfall and runoff production of suspendedsediment. Increasing the number of parameters makes the calibration process more complex without increasing in a noticeablemanner the predictive capability of the model.