Improvement of Flood Risk Analysis Via Downscaling of Hazard and Vulnerability Maps

Abstract

Flood risk maps are important references for disaster prevention and emergency response. The accuracy of risk maps is greatly affected by the resolutions of hazard and vulnerability maps. To determine the impact of map resolutions on flood risk analysis, a total of 12 risk maps were generated for Shanhua District, Taiwan, via the integration of hazard and vulnerability maps under two resolutions and three return periods. The hazard, vulnerability, and risk maps were classified into five levels according to flood depth, socio-economic indicators, and their products, respectively. The results show that the downscaling of hazard maps greatly increases the hit rate by 28% and decreases the false alarm rate by 53% in the flood risk analyses of households. In contrast, the downscaling of vulnerability maps only slightly increases the hit rate without an obvious decrease in the false alarm rate. To improve flood risk analysis under time and budget limitations, numerical downscaling of hazard maps should be given higher priority because it reduces the structural errors in hydraulic simulations that cannot be compensated for by the statistical downscaling of vulnerability maps.