A hybrid modeling approach for analyzing the risk propagation of water quality deterioration associated with water transfer during flood season

Study region The Nansi Lake is a key storage reservoir for the Eastern Route of the South-to-North Water Diversion Project in China. Study focus As a new practice in existing water diversion projects, inter-basin water transfer during flood season poses new challenges to water quality deterioration risk management. Risk propagation analysis is crucial for risk management, which constitutes a critical determinant for the success of flood-season water transfer operations. Commonly used risk propagation analysis models in water quality deterioration, such as the hydrodynamic-water quality model or data-driven models, either fail to quantify the probability and magnitude of risk propagation, or are limited by insufficient risk associated records. Therefore, a new hybrid modeling approach is designed to overcome the limitations, coupling a Hydrodynamic-water quality model, a C-Vine Copula function, Sine functions, and a Markov chain. It can depict the physical and stochastic processes of water quality deterioration risk, further deciphering its spatial evolution pattern. New hydrological insights for the region The results show that 33∼67% of the lake area is highly likely (97.5%) to suffer water quality deterioration when water transfer scheduled during flood season, and the middle and northern regions of the lake are the hotspots of high risk. Moreover, the risk propagation route is from rivers to the lake and from the south to the north in the lake. From the spatial risk evolution perspective, it maintains low-risk or medium-risk propagation in the south, while exhibiting risk-enhanced propagation in the middle, and maintaining high-risk propagation in the north. Thus, rivers flowing into the north and middle of the lake are the priorities for the risk management. Overall, the study quantitatively reveal the spatiotemporal evolution patterns of water quality deterioration risk, identifying risk hotspots in the lake and the priority areas for risk management. These results provide decision-makers with a scientific basis for optimizing water environment management strategies.