Modelling the influence of future climate and land-use change on fluvial flood inundation in data-scarce regions using CMIP6 ensembles, FABDEM and a validated 2D hydrodynamic model

Accurate prediction of future flood inundation in data-scarce regions is severely constrained by lack of reliable in situ observed hydro-meteorological and post-flooding data. In this study, a comprehensive flood modelling approach that combines statistical downscaling of CMIP6 global climate model data, cellular automata–Markov chain land-use change modelling and hydrological and hydrodynamic modelling was applied to quantify future fluvial flood inundation in the Semliki River catchment. The developed 2D hydrodynamic model was calibrated and validated using Google Earth Engine cloud computing and Sentinel-1 imagery and applied to simulate flooding impacts for the baseline and 18 future climate and land-use change scenarios. The study findings suggest that by 2080, climate change could lead to a 142% increase in average river flows (from 46.6 to 112.7 m3/s), 73% increase in flooding extent (from 136.7 to 235.6 km2) and an increase in maximum flood depths by 34% (from 3.2 to 4.3 m). In contrast, land-use changes will cause smaller increases: 25% in peak flows, 7% in flooding extent and 3% in flood depths. The presented approach provides a replicable methodology that can be used to reliably estimate future flooding impacts and to facilitate comparative evaluation of potential flood resilience strategies in other data-scarce regions.