Elevation data in the form of Digital Elevation Models (DEMs) has been recognised as a basic piece of information for the accurate representation of topographic controls exerted in hydrologic and hydraulic models. Yet many practitioners rely on open-access global datasets usually obtained from space-borne survey due to the cost and sparse coverage of sources of higher resolution. In may 2016 the Japanese Aerospace eXploration Agency (JAXA) publicly released an open-access global DEM at an horizontal resolution of 30 m, the ALOS World 3D-30m (AW3D30). So far no published study assessed the flood modelling capabilities of this new product. The purpose of this investigation is twofold. Firstly, to present an assessment of the capacity of the AW3D30 DEM for flood modelling purposes and secondly, to compare its performance with regards to computed water levels and flood extent maps calculated using other freely available 30 m DEMs for model setup (e.g. SRTM and ASTER GDEM). For this comparison, the reference to reality is given by the water levels and flood extent maps computed with the same numerical model but using a LiDAR based DEM (5 m of spatial resolution re-sampled to 30 m). The numerical model employed in this investigation is based on a damped partial inertia approximation of the Saint-Venant equations on a regular raster grid, which is forced with a simple and synthetic rainfall storm event. Numerical results using different elevation data in model setup are compared for two regions with contrasting topographic gradients (steep and smooth). Results with regards to water depth and flood extent show that AW3D30 DEM performs better than the SRTM DEM. Notably, in the case of mountainous regions results derived with the AW3D30 DEM are comparable in skill to those obtained with a LiDAR derived DSM, suggesting its suitability in the numerical reproduction of flood events. This encouraging performance paves the way to more accurate modelling for both data-scarce regions and global flood models.

A newer version of this preprint is available on EarthArxiv: https://doi.org/10.31223/osf.io/vqgx4