Abstract. Digital elevation models (DEMs) are essential datasets, particularly for flood inundation mapping in one-dimensional (1D) to two-dimensional (2D) flood models. Given the significant uncertainties associated with DEMs that can affect flood modelling accuracy, minimizing these inaccuracies is essential. This study aims to improve the performance of 1D–2D flood models using satellite Earth observation (EO) data, focusing on the lower Chao Phraya (CPY) basin. Two workflows are proposed: DEM analysis and flood map analysis. The DEM analysis evaluates 10 DEM products, including three local DEMs provided by Thai agencies (LDD, JICA, and a merged LDD-JICA DEM) and seven global DEMs derived from EO data (ASTER GDEM V3, SRTM V3, MERIT, GLO30, FABDEM V1–2, TanDEM-X, and TanDEM-EDEM). The evaluation process uses ICESat-2 ATL08 data processing, vertical datum reference processing, and evaluation of DEMs using ICESat-2 ATL08 benchmark processing. The DEMs are assessed using satellite laser altimetry data from the Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) as the benchmark. The evaluation employs standardized metrics, including point-wise, grid-wise, and track-wise comparisons, to identify the most suitable DEM for integration in the flood model. Results indicate that the merged LDD-JICA DEM and FABDEM V1–2 DEM exhibit the highest accuracy among local and global products, respectively, with root mean square errors (RMSEs) of 1.93 and 1.95 m, and percentage biases (PBIASs) of −15.38 % and 4.59 %. The flood map analysis workflow involves comparing flood extent maps derived from multi-mission satellite datasets and simulated flood maps generated from 1D–2D flood models using the best available DEMs. This workflow utilizes surface water extent (SWE) maps from the WorldWater project, obtained from the Sentinel-1 and Sentinel-2 imaging satellites, and flood maps from the Geo-Informatics and Space Technology Development Agency (GISTDA) in Thailand to validate flood maps produced by the 1D–2D flood model based on the merged LDD-JICA DEM and FABDEM V1–2 DEM. The results reveal that flood maps based on the FABDEM V1-2 DEM slightly outperform those based on the merged LDD-JICA DEM, with an improvement of approximately 13.55 %–25.56 % in the critical success index (CSI). This study highlights the potential of leveraging satellite EO data to enhance the accuracy and reliability of 1D–2D flood models, thereby improving flood inundation predictions for effective flood management.
Charoensuk et al. (Thu,) studied this question.