A GIS-based DEM Correction Method to Improve Urban Flood Simulation Results

Abstract:

Flood modelling provides useful information to support flood risk assessment and management and reduce flood impacts. The accuracy of flood simulation results is highly dependent on the quality of input data. In particular, terrain data may directly influence the accuracy of flood predictions and improper representation of certain urban terrain features including road-crossing and river-crossing bridges may lead to incorrect prediction of flood paths and extents, and thus misleading flood risk. In this work, a GIS-based correction method is proposed to automatically identify and correct the representation of bridge structures in high-resolution Digital Elevation Models (DEMs), which will improve flow connectivity for urban roads and river channels and better represent flood paths in reality, leading to more accurate flood simulation results. The proposed DEM correction method is combined with a 2D High-Performance Integrated hydrodynamic Modelling System (HiPIMS) and applied to support large-scale urban flood modelling in the City of Fuzhou, China. Comparisons are made between the simulation results with and without implemented the DEM correct method. The results demonstrate that correct representation of the bridge-like structures in the DEM may have a significant influence on the flood simulation results. The correction method has a great potential in improving the digital terrain data to support more reliable urban flood modelling and can be readily applied to other cities.