Evaluating the Impact of Digital Elevation Models on Urban Flood Modeling: A Comprehensive Analysis of Flood Inundation, Hazard Mapping, and Damage Estimation

Digital Elevation Models (DEMs) play a crucial role in flood management. This study aims to assess the effect of various global DEMs (GDEMs), including ALOS-12.5 m, ALOS-30 m, SRTM-30 m, SRTM-90 m, and NASADEM-30 m, on flood risk modeling in a densely urban area. The 1D-2D MIKE FLOOD hydraulic model was employed for the flood modeling. The process involved using a high-resolution DEM (Pleiades-1A 1 m) as the reference map (RM1), along with other GDEMs, to simulate a 50-year return period flood. The performance of GDEMs was then assessed in terms of flood inundation extent, flood hazard, and flood damage estimation, assessing their accuracy against the RM1. The study also explored the trade-offs between accuracy and efficiency by examining the effects of substituting the high-resolution map with a 5-m resolution map (Res_5 m) created through resampling. Results revealed that GDEMs tend to overestimate flood extent and underestimate depth, leading to inaccurate flood risk assessments. Among the GDEMs, NASADEM-30 and SRTM-30 outperformed others in simulating flood inundation extent but resulted in a more uniform flood depth distribution; approximately 70% of the flood extent was categorized with depths of less than 0.3 m, nearly double that of the RM1. This discrepancy led to an underestimation and overestimation of higher (H3-H6) and lower (H1) hazard levels by approximately 50%, respectively. Furthermore, GDEMs significantly overestimated flood damages, with NASADEM-30 showing a 161% overestimation compared to the RM1. Ultimately, the Res_5 m was a viable alternative for urban flood simulations as it led only to a modest 6% decrease in the flood zone area.