SEARCH

SEARCH BY CITATION

Keywords:

  • model benchmarking;
  • 2D hydraulic models;
  • flood inundation modelling

Abstract

For a nationwide flood risk assessment in Germany, simulations of inundation depth and extent for all major catchments are required. Therefore, a fast two-dimensional (2D) hydraulic model is needed. From the range of existing methods, two storage cell models are evaluated to find an appropriate method for large-scale applications. The Dynamic Rapid Flood Spreading Model (Dynamic RFSM) based on irregular storage cells, and a raster-based model with inertia formulation of momentum equation are compared. Simulation performed with the fully dynamic shallow water model InfoWorks RS 2D served as a reference. The hydraulic models are applied to a test area having a very flat topography adjacent to the river Elbe. As a benchmark scenario, the outflow through a hypothetical dike breach was chosen. To investigate the impact of the grid resolution on run time and model performance, the simulation with the raster model is carried out with different grid sizes. Furthermore, the sensitivity of the Dynamic RFSM to the choice of time step was analysed. Both models were able to simulate the final inundation extent and depths with a reasonable accuracy. However, the Dynamic RFSM showed some weakness in simulating inundation extent over the flat test area. Coarsening the grid resolution reduced the run time of the raster-based model considerably and can be regarded as a promising strategy to constrain the computational efforts for a large-scale application, although the model accuracy gradually deteriorated. With similar run time, the raster-based model performed better than the Dynamic RFSM in terms of inundation extent and comparable regarding maximum inundation depth. Generally, an application at national scale appears feasible with both hydraulic modelling schemes. Copyright © 2012 John Wiley & Sons, Ltd.