Prediction of sediment yield in an ungauged basin under the impact of cascade dam-reservoirs development

Abstract

Using the sediment rating curve (SRC) model, the regression-based regionalization approach was evaluated in 16 catchments. It is relied upon multiple regression equation in which the fitted model parameters are the dependent variables and catchment descriptors are the independent variables. Based on the jack-knife validation, the overall predictive accuracy is reasonably good and satisfactory result was obtained in large majority of the catchments. The validated model parameter-catchment descriptor relationship was then applied to parameterize SRC in the upper Sre Pok river basin (Vietnam) facing development of cascade dam-reservoirs, for subsequent sediment yield prediction. This ungauged basin has potential to release suspended sediment about 2.09 Mt/yr to downstream. This amount could be reduced to 0.31 Mt/yr if the development does not provide appropriate sediment flushing facilities.