| Abstract: | River–lake water exchange reflects hydrological connectivity and the dynamic relationship between the river and the lake. The water exchange is crucial for lake level variation, downstream river discharge and the ecosystem. To figure out the water exchange between the Yangtze River and Poyang Lake, a data‐driven model was established based on the support vector regression and genetic algorithm technique. Nine scenarios were set with different river–lake hydrological conditions, divided into two categories: single‐element change scenarios, where only the river conditions or only the lake conditions changed, and combined scenarios, where both elements changed simultaneously. The model could accurately simulate the river–lake water exchange variations. Scenario simulation results show that increasing the river flow or lowering the lake level could cause a decrease in the lake outflow. Conversely, decreasing river flow or raising the lake level could cause an increase in lake outflow. Changing lake conditions have a stronger impact on the water exchange variation than changing river conditions if the change percentages of the situation indicator values are the same. Similarly, lake level increase has a stronger impact on the water exchange variation than lake level decrease. The combined scenarios indicate the additive effect of the corresponding single‐element change scenarios, with a clear linear relationship between their lake outflow changes. This study provides an efficient model for simulating complex hydrological flow relationships in river–lake systems, and supports the management of the Yangtze River and Poyang Lake by providing the characteristics and causes of the river–lake water exchange. |
