A mountain basin in southern High Mountain Asia (HMA). HMA is susceptible to flash flood hazards, while the difficulty of topographic data acquisition challenges its flash flood modelling. Little attention has been paid to the influence of open-source topographic data on flash flood modelling in the region. In this study, the influences of four open-source topographic data used widely on flash flood modelling were explored. Results indicated significant influences of these open-source topographic data. The 30-m ASTER GDEM data with unreasonable results were not recommended. At the basin scale, results from 30-m SRTM, 12.5-m ALOS PALSAR and 30-m COP DEM data underestimated the maximum flood depth by −16.04% ∼ −5.18%, and 30-m COP DEM data underestimated the maximum flood extent by −3.51% ∼ −4.63%. The simulated maximum flood velocity and hydrodynamic force, as important disaster-causing mechanisms of flash floods, exhibited biases of −16.38% ∼ 11.92% and −99.46% ∼ −5.87%, respectively. At three infrastructure sections concerned, different qualities of three open-source topographic data caused inconsistent biases. The simulated temporal characteristics of the above indicators were consistently delayed by 1.23∼180.22%. Such biases have significant influences on data-scarce basins in HMA, causing underestimated flash flood risk and delayed early warning. In-depth studies of topographic data fusion and bias correction are suggested to improve flash flood modelling. • Influence of four open-source DEM data on flash flood modelling is investigated. • Open-source DEM data cause large underestimation of hydrodynamic characters of flash flood. • Temporal characters of flash flood are consistently delayed by open-source DEM data. • 30-m ASTER GDEM data with wrong result is not recommended for flash flood modelling.
Li et al. (2026) studied this question.