Application of geospatial technologies in constructing a flash flood warning model in northern mountainous regions of Vietnam: a case study at TrinhTuong commune, Bat Xat district, LaoCai province

Abstract

The model was constructed based on GIS spatial analyses, combined with Analytic Hierarchy Process (AHP) and Multi-Criterion Analysis method (MCA). The data gathered for the study were mainly from remote-sensing images, statistical data and surveys. Field experiments were conducted in Trinh Tuong Commune, Bat Xat District, Lao Cai province. This is a typical remote mountainous region of Vietnam in which flash floods often occur. The study analyzes and evaluates six primary factors that incite flash flood, namely: geomorphological characteristics, soil properties, forest and fractional vegetation cover types, local drainage basin slopes, maximum average rainfall of various years, and the river/stream density of the region. The zoning map showing flash flood potentials has determined that 19.91% of the area had an extremely high risk of flash flood occurrence, 64.92% of the area had a medium risk, and 15.17% had a low or very low risk. Based on the employment of daily maximum rainfalls as the primary factor, an online flash flood warning model was constructed for areas with a "high"or "very high"risk of flash flood occurrence.