Estimation du taux de production sédimentaire spécifique des sédiments en suspension par échantillonnage discret dans différentes régions de Belgique

peer reviewedPurpose Suspended sediment transport, which represents the majority of the sediment load, has been studied across very different scales and in a wide variety of regions and climates. Despite numerous studies, data for European watersheds are generally limited and correspond to large rivers systems. Especially, in Belgium, little data is available outside the Belgian loess belt. Moreover, the high heterogeneity of soil erosion and sediment transport makes it difficult to measure or model at watershed scale. The purpose of this research is to estimate the median sediment yields in different geographical regions and to detect their explanatory variables. Materials and methods Gathering data from 1994 to 2016, more than 2,000 measurements of suspended sediment concentration at 72 river stations mainly located in South-Belgium were sampled according to a flood-event-based manual methodology. This allowed fast acquisition of data in watersheds ranging from 7 to 3,600 km² in different geographical regions. Median area-specific sediment yields (SSY) were calculated at watershed scale while looking for regional differences. Results and discussion Median area-specific sediment yields computed for the period 1996-2018 show regional differences: 19.2 t km-2 yr-1 on sandy substrate (Lorraine), 24.9 t km-2 yr-1 on schisto-sandstone substratum (Ardenne) and up to 119 t km-2 yr-1 in the loamy Brabant plateau, with a link to the agricultural land cover and, to a lesser extent, to the watershed slope. The high temporal and spatial variability of rainfall has great effects on the SSY, necessitating the gathering of more than 20 years of data to smooth the high variability of SSY. A multiple correlation of land cover variables and the average slope of the watershed with SSY managed to explain 48% of the variance within the SSY observations. Conclusions The agricultural land cover has an important effect on median SSY values. While the regionalization of Belgium is largely based on lithology, soils and altitude, the land use resulting from these physical and climatic characteristics explains the differences in SSY. Field values of clogged dams and waterways confirm the matching of the SSY computation from discrete samples, despite the high temporal variability of sediment transport