Continuous 1D-2D flood routing for regional-scale flood risk assessment
Abstract
The regional flood risk model (RFM) is a chain of process-based models developed for regional-scale flood risk assessment to serve national policy development, insurance appraisal and large-scale disaster management planning. It includes a multi-variate multi-site weather generator, the distributed rainfall-runoff model SWIM, 1D river routing, 2D hinterland inundation model and the multi-parameter damage estimation model FLEMO. The entire simulation chain accounts for the heterogeneous patterns of precipitation, discharge, inundation and damage and represents the catchment and floodplain processes in a holistic way. A multisite multi-variate weather generator followed by SWIM generate consistent continuous flow hydrographs at sub-basin outlets. The river hydraulics is simulated by the 1D diffusive wave routing model which uses the SWIM output as boundary conditions. A 2D hinterland inundation (raster based) model with the inertia implementation is coupled to the 1D model and is activated in case of overland flow. The 2D code is implemented in the CUDA Fortran environment to enable the application of model on highly parallelized NIVIDIA Graphical Processor Units (GPU). For each flood event where dike overtopping and hinterland inundation occur, maximum water levels at each cell are extracted and used for flood loss calculation.
The previously used sub-basin wise 1D routing scheme is substituted by a continuous routing approach using the diffusive wave approximation. The new model also integrates 8-point river cross-sections derived from a 10x10m DEM and incorporates statistically derived bankfull area and depth. The new approach fully accounts for the flood wave attenuation, the superposition of waves at confluences and channel-floodplain interactions in case of dike overflow and subsequent inundation. The new scheme is benchmarked against the previous sub-basin routing and the implications for flood hazard and risk are analyzed.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H13L1878F
- Keywords:
-
- 1805 Computational hydrology;
- HYDROLOGY;
- 1849 Numerical approximations and analysis;
- HYDROLOGY;
- 1906 Computational models;
- algorithms;
- INFORMATICS;
- 1932 High-performance computing;
- INFORMATICS