Abstract
Abstract Interbasin water transfer projects inevitably change downstream hydrologic conditions and pose serious threats to the reproduction of endemic drifting-egg fish in the Han River, including the four major Chinese carps. To overcome the limits of traditional assessments that separate hydrologic and ecological processes and rely on single optimization goals, this study builds a Hydrodynamic–Ecological Collaborative Optimization (HECO) model. The model develops an automated Python-based data exchange interface, enabling dynamic runtime coupling between MIKE 11 and Physical Habitat Simulation, and integrates an improved Non-dominated Sorting Genetic Algorithm II (INSGA-II) to solve high-dimensional nonlinear scheduling strategies. Crucially, to bridge the gap between mathematical optimization and biological reality, the framework embeds an empirical temporal-thermal gating mechanism and a transient hydrologic pulse trigger directly into the objective function, successfully circumventing the 'isothermal trap'. Using the middle and lower reaches of the Han River as an example, the model's ecological predictive validity was first corroborated against historical field observations. The results show that the model reaches high hydrodynamic simulation accuracy, with the Nash–Sutcliffe Efficiency of discharge above 0.80. The intelligent optimization scheme maintains water supply benefits while increasing the average Habitat Suitability Index to 0.89 and the Connectivity Index to 0.91. The mechanism analysis shows that the optimized scheme rebuilds a pulse-like flow pattern mathematically congruent with natural hydrology, with R2 increasing to 0.926. It breaks the ecological signal loss caused by conventional constant-reduction scheduling and achieves maximum ecological benefits under limited water availability. This study provides a robust theoretical foundation and precise technical guidance for the ecohydraulic fine-tuning of large-scale water conservancy operations.