Determining the limiting water level of early flood season by combining multiobjective optimization scheduling and copula joint distribution function: A case study of three gorges reservoir

摘要

During the early flood season of the Yangtze River, which encompasses the spawning season of four major Chinese carps, reservoirs tend to remain at a low limiting water level in case of large floods. However, a low limiting water level, which is the initial water level of the spawning season, decreases the reservoir volume, causing insufficient outflow for downstream fish spawning. Therefore, the Limiting Water Level of Early Flood Season (LWLEFS) must be optimized to achieve a comprehensive benefit considering fish spawning, apart from hydro-power generation and flood control. Using the Three Gorges Reservoir (TGR) as a case study, this study aims to combine multiobjective optimization scheduling using the non-dominated sorting genetic algorithm and the copula joint distribution function, and propose an optimizing framework to determine the LWLEFS. The optimizing framework is as follows: first, multiple inflow time series during the spawning season are randomly simulated based on historical hydrological wavelet analysis, to support optimization scheduling. Next, considering three objectives, i.e., hydropower generation, flood control, and fish spawning, multiobjective optimization scheduling is conducted, and the optimal individual for each inflow time series is obtained at different LWLEFSs. Subsequently, the copula joint distribution function of the three objectives is established, and the Probabilities Achieving the Established Objective Values (PAEOVs) are calculated. Additionally, the relationships between different LWLEFSs and PAEOVs are obtained, and the LWLEFS that leads to the maximum PAEOV is identified. For the TGR, the optimal LWLEFSs are 154,149,148,147, and 145 m for extremely dry, generally dry, normal, generally wet, and extremely wet hydrological conditions, respectively. The optimizing framework can be extended to other reservoirs with local targets, and it offers technical support for ecological and environmental management. Plain language summary: The fish in Yangtze River, China often spawns during June 15th to July 20th. which is located in the early flood season. In these days, the basin reservoir such as Three Gorges Reservoir always keeps a low water level for flood control. However, under the low operation water level, the reservoir volume is small and cannot release adequate water for the downstream fish spawning. The above water level needs be changed considering downstream fish spawning. Therefore, we have analyzed the multi objectives of reservoir operation, including hydropower generation, flood control and fish spawning. And we propose an optimizing framework to determine the operation water level based on their trade-off relationship. The optimizing framework is practical for ecological and environmental management, and can be applied for other reservoirs with different operation targets.