Search

High aggradation and degradation in a river induced by last flood event will respectively raise and decrease the risk of an upcoming flood event. However, this point is likely to be ignored or is not fully considered in some 2D fluvial flood models. To address this problem, this work develops a 2D high-resolution hydrodynamic model coupled with the sediment transport and the river bed evolution models. The modelling system is within the framework of a Godunov-type finite volume scheme. GPU technique is applied to accelerate the computation by more than 10 times, comparing to the CPU counterpart. After being validated against an experimental benchmark test, the model is applied to simulate the effects of the morphological change on flood patterns for the Bayangaole Reach of Yellow River, China. The results indicate that the effect of perturbed bed could be of significance for the fluvial flood over movable bed. It is therefore suggested to take into account when evaluating the flood risk in this case. [ABSTRACT FROM AUTHOR]