Integrating GIS and catchment hydrology to assess flash floods in Ha Giang Province, Vietnam

Flash floods are major natural disasters in mountainous area such as Ha Giang Province, Vietnam, causing severe loss of life and property. Prediction of flash floods is challenging due to complex interactions among terrain, geology, soil, land cover, and rainfall. Existing studies lack a comprehensive, data-driven approach integrating geospatial analysis and hydrological modelling. This study addresses that gap using an advanced GIS-based methodology combined with remote sensing data and catchment-scale hydrology. A multi-source dataset—including flash flood inventory, rainfall records, topographic, geological, and soil maps, and Landsat-8 imagery—was used to identify key contributing factors. Parameters such as lithology, soil, rainfall, elevation, slope, land use, NDVI, and catchment characteristics were analysed. Results show that soil and lithology are the most critical factors, followed by rainfall, elevation, and slope. Catchment size and shape also influence susceptibility, while forest cover may be insufficient to mitigate floods. The findings support improved early warning, land-use planning, and disaster management. The framework is adaptable for flash flood risk assessment in similar regions worldwide.