Modelling of man-made flood routing in the lower Yellow River, China

Large-scale prototype experiments of regulating flow and sediment have been conducted annually in the Yellow River, China since 2002, with large amounts of observed hydrological data being collected during the experiments. In the present study, a two-dimensional hydrodynamic model based on the finite-volume method has been refined, with an improvement being made in the spatial reconstruction method of the state variables. The enhanced spatial reconstruction method can account for the effect of a wet or dry interface on the calculation of numerical fluxes, especially for low discharges, which can avoid the occurrence of unreal flow fluxes. The refined model was applied to simulate flood routing in the 2004 and 2006 experiments in the lower Yellow River and relatively close agreement between the model predictions and observed data was obtained. The effect of different bed roughness values on the predicted results was investigated using a sensitivity test, and various factors influencing the accuracy of the predicted discharges are presented. The simulated results also show that the refined spatial reconstruction method results in higher predictive accuracy for low discharges, different bed roughness coefficients have only a slight effect on the predicted discharges and the predictive accuracy of the model is also influenced by the resolution of bed topography and the way in which the open boundary conditions are specified.