Evaluating flood responses with and without a reservoir module using the distributed Liuxihe hydrological model

Reservoirs influence runoff processes and flood routing, affecting flood simulation accuracy in distributed hydrological models. This study incorporated a reservoir module into the Liuxihe model to evaluate the impact of reservoir module on flood simulation. The improved model explicitly represents reservoir-covered areas as reservoir units within the distributed model. The Baipenzhu and Xinfengjiang reservoir basins in Guangdong Province, China, were selected as study areas. Results showed that the reservoir module improved flood simulation performance, with mean Nash coefficient, correlation coefficient, process relative error, flood peak error, and peak time error of 0.924, 0.850, 0.400, 0.090, and 1.21 h, respectively. Compared with the model without the reservoir representation, the improved model reduced errors in flood peak and peak time. These results highlight the importance of reservoir representation in distributed hydrological modelling and provide a practical approach for flood simulation in data-limited basins.