Modeling Stream Bank Erosion: Practical Stream Results and Future Needs

Abstract

Coupled two-dimensional (2D) morphodynamic and bank erosion models are gaining attentions in recent years. It has been shown that such models have advantages over the one-dimensional (1D) modeling approaches. In this paper, a previous 2D bank erosion model with the moving mesh method is extended to include the fixed mesh approach. Further, two practical streams with differing complexity are simulated to demonstrate the extended model. Both the moving mesh and fixed mesh methods are used in the modeling. The model consists of two components: a 2D flow and mobile-bed model for vertical bed changes and hydraulic forces acting on a bank and a lateral bank retreat model. The 2D vertical model and the lateral bank erosion model are coupled together spatially and temporally through a special procedure and a common mesh. With the experiences gained with practical stream modeling, the modeling procedure and key model input parameters are described. The study shows that the moving and fixed mesh methods together make the extended bank erosion model numerically robust and capable of predicting both the vertical bed changes and the lateral stream bank erosion for complex streams. Each individual method, however, has its own limitations in terms of model accuracy and efficiency. The moving mesh works well if bank retreat is relatively small, e.g., less than one channel width, and produces more accurate results than the fixed mesh method. The fixed mesh may be needed for ensuring numerical stability if a bank may be subject to significant retreat (e.g., more than one channel width). The fixed mesh method, however, is less accurate than the moving mesh method and a much refined mesh may be needed. Both methods need future research and improvements in terms of their model accuracy.