A Well-Balanced Numerical Scheme for Shallow Water Simulation on Adaptive Grids
H. J. Zhang,J. Z. Zhou,S. Bi,Q. Q. Li,Y. Fan
DOI: https://doi.org/10.1088/1742-6596/495/1/012049
2014-01-01
Journal of Physics Conference Series
Abstract:The efficiency of solving two-dimensional shallow-water equations (SWEs) is vital for simulation of large-scale flood inundation. For flood flows over real topography, local high-resolution method, which uses adaptable grids, is required in order to prevent the loss of accuracy of the flow pattern while saving computational cost. This paper introduces an adaptive grid model, which uses an adaptive criterion calculated on the basis of the water lever. The grid adaption is performed by manipulating subdivision levels of the computation grids. As the flow feature varies during the shallow wave propagation, the local grid density changes adaptively and the stored information of neighbor relationship updates correspondingly, achieving a balance between the model accuracy and running efficiency. In this work, a well-balanced (WB) scheme for solving SWEs is introduced. In reconstructions of Riemann state, the definition of the unique bottom elevation on grid interfaces is modified, and the numerical scheme is pre-balanced automatically. By the validation against two idealist test cases, the proposed model is applied to simulate flood inundation due to a dam-break of Zhanghe Reservoir, Hubei province, China. The results show that the presented model is robust and well-balanced, has nice computational efficiency and numerical stability, and thus has bright application prospects.