Assessing drought vulnerability and water allocation in a monsoonal river basin using integrated modeling

In monsoonal river basins, seasonal drought risk depends on the timing of runoff, infrastructure operation, and competing sectoral demands as well as on annual water availability. Earlier provincial applications of related models did not represent the full Ba cascade-irrigation-coastal allocation network or combine projected industrial-demand growth with hectare-scale drought-hotspot mapping. A sequential coupling of the MIKE NAM rainfall-runoff model with MIKE HYDRO Basin across 23 sub-basins represents reservoir operation, environmental-flow requirements, sectoral priorities, and infrastructure connectivity. Simulations compared the 2017 baseline scenario (SC1) with a 2030 planning scenario based on Representative Concentration Pathway 4.5 (RCP4.5; SC2) under four inflow exceedance probabilities (PoE = 25%, 50%, 75%, and 90%). Approximately 68.0% of modeled annual inflow occurred during September-December, whereas irrigation demand was concentrated during the following eight-month dry season. Operational verification of reservoir storage produced correlation coefficients of 0.9944 and 0.9939 and Nash-Sutcliffe efficiency values of 0.9599 and 0.9632 for the Winter-Spring and Summer-Autumn seasons, respectively. Under SC2, total annual water demand increased by 17.04%, and industrial demand increased nearly twentyfold relative to SC1. The Upper Dong Cam annual deficit increased from 37.63 to 55.79 × 10⁶ m³, while the province-wide area affected by extreme drought at PoE = 90% increased from 4016.4 to 6300 ha. Regulated sub-basins retained greater dry-season supply reliability, whereas isolated coastal sub-basins and small storage systems showed the greatest modeled vulnerability. Across the 23 sub-basins, drought severity and spatial extent varied with hydrological conditions and with differences in reservoir operation, local demand, and infrastructure.