A RUSLE approach to model suspended sediment load in the Lo river (Vietnam): effect of reservoirs and land use changes

This study addresses the problem of modelling sediment erosion at the catchment scale, in order to predict the possible impact of reservoirs and land use changes on sediment load in South East Asia. The investigated basin is the Lo river basin, (38,165 km2 in Viet Tri), a left tributary of the Red River, where the Thac Ba and the recently built Tuyen Quang reservoirs have already been changing downstream sediment load since 1971 and 2005 when they were, respectively, in operations. The RUSLE equation is adopted in a distributed GIS framework to assess catchment erosion, and is coupled with a sediment accumulation and routing scheme to model suspended sediment load in the Lo basin at a monthly scale. Monthly precipitation were collected and used as input to the model. Suspended sediment load data, measured at eight gauging stations in Vietnam, from 1959 to 2007, were compared with the model’s simulated sediment yield. Resulting average Nash–Sutcliffe efficiency is 0.45, ranging from 0.33 to 0.62 and bias is 5% at Vu Quang, close to the basin outlet, and results 1.4% by averaging biases at the eight stations, thus confirming that the model is adequate. Effects of reservoirs were analysed by modelling erosion and sediment yield passing from natural to impounded conditions and resulted in a suspended sediment load reduction of about 95% and 75% downstream, respectively, the Thac Ba reservoir in the Chay river and the Tuyen Quang reservoir in the Gam river. Land use change scenarios, parameterized on the basis of observed land use changes in the impounded basin and assuming that 20% of forest area is converted into rice and agricultural crops and 15% into bushes, shrubs and meadows, are expected to induce a 28% increase of suspended sediment load which can compensate, at least in part, sedimentation in reservoirs. Also agricultural and hillslope maintenance practices can modify sediment erosion in the basin.