Assessment of open-source and proprietary generated digital elevation models and building footprints for urban flood modelling in Cartagena De Indias, Colombia

Abstract

Urban hydrological models are critical for flood risk management. However, the availability of high-resolution topographic data for reliable outputs remains challenging in data-scarce cities. Therefore, determining data quality in producing reliable information is fundamental. We evaluated open-source and proprietary topographic data for use in the two-dimensional hydrological model City Catchment Analysis Tool (CityCAT). We modelled 12 scenarios using combinations of open-source and light detection and ranging (LiDAR)-generated datasets, validating the results with community-generated flood risk maps. The findings show high agreement with scenarios using LiDAR-derived digital elevation models (DEMs) (bootstrapped Spearman’s ρ ≈ 0.90). However, open-source building footprints performed better, demonstrating that both are necessary for reliable urban flood risk mapping. As LiDAR is costly with limited access, we urge for publicly available high-resolution datasets for low- and middle-income countries (LMICs) disproportionately impacted by climate change. Therefore, we address this gap by focusing on a Latin American context with Cartagena de Indias (Colombia) as a case study.