A hybrid hydrological-hydrodynamic method and its application in the “23·7” catastrophic flood of Yongding River, China

Flash flood is one of the most destructive natural disasters, and with the increasing frequency of extreme weather events, there is a growing demand for more accurate forecasting methods. To address this demand, a novel hybrid hydrological-hydrodynamic method is provided. This method employs a kilometer-scale distributed hydrological model to capture rainfall-runoff data, with the coupling process considering the slope runoff. Concurrently, a multi-GPU accelerated hydrodynamic model is utilized to simulate flood routing, enabling prediction of flood discharge, inundation scenarios and inundation water depth. In July 2023, the Yongding River basin in Beijing experienced the “23·7” catastrophic flood, triggered by prolonged extreme rainfall. This flood is the most severe flash flood in nearly a century, causing widespread destruction and leaving a profound impact. Using the “23·7” catastrophic flood as a case study, this paper employs the proposed method to simulate the flood routing and inundation scenarios. The results are compared with the field investigations, which show that the provided hybrid hydrologic-hydrodynamic method can effectively simulates flood routing and inundation, and has high computational efficiency. This method provides significant reference value for disaster prevention and mitigation departments, water resource management agencies, as well as researchers engaged in river basin flood forecasting and emergency response. It can offer scientific support for tasks such as flash flood early warning, emergency scheduling, and post-disaster evaluation.