Improved comprehensive flood risk index method and multi-scenario flood simulation for downstream area of the reservoir

Climate change-induced extreme rainfall amplifies flood risks to reservoir safety, underscoring the need for advanced risk assessment integrating socioeconomic vulnerability and multi-scenario modeling. An improved comprehensive flood risk index framework was developed by integrating hazard and vulnerability indices, with flood probability incorporated, to evaluate flood risk downstream of Fenglongwan Reservoir. The Integrated Flood Modeling System (IFMS) models were established and flood losses were quantified for six scenarios including emergency flood discharge and dam-break events under design, check, and overtopping flood magnitudes. The results show: (1) The validated IFMS model accurately simulate flood dynamics, with a 6.1% peak discharge deviation at Qixingqiao Station and 7~14 cm water level errors at three control sections, supported by high hydrograph agreement (R² = 0.86–0.92) and low bias (PBIAS = 0.17–3.70%) in the flood process. Dam-break scenarios show larger peak discharges and shorter flood arrival times than emergency flood discharges. (2) Inundation areas ranged from 8.22 to 36.01km², mean velocities from 1.39 to 4.44 m/s, and durations from 32.20 to 49.96 hours across six scenarios. Dam-break scenarios incurred 4–10 times higher economic losses than emergency flood discharge scenarios, with agricultural loss comprising over 85%. (3) Low-risk areas dominated the risk distribution in both events (56.55% for dam-break and 63.07% for emergency flood discharge), while high-risk zones clustered in river-adjacent agricultural and densely populated regions. Jiangwai Village and Lawei Village exhibit high risk in the dam-break event. This research provides a robust framework for reservoir safety management and targeted flood control planning.