Water quality plays a pivotal role in human health and environmental sustainability. However, traditional water quality prediction models are limited by high model complexity and long computation time, whereas AI models often struggle with high-dimensional time series and lack physical interpretability. This paper proposes a two-dimensional water quality surrogate model that couples physical numerical models and AI. The model employs physical simulation results as input, applies spectral proper orthogonal decomposition to reduce the dimensionality of the simulation results, utilizes a long short-term memory neural network for matrix forecasting, and reconstructs the two-dimensional concentration field. The simulation and predictive performance of the surrogate model were systematically evaluated through four design scenarios and three sampling dataset lengths, with a particular focus on the convection–diffusion zone and high-concentration zone. The results indicated that the model achieves high prediction accuracy for up to 7 h into the future, with sampling dataset lengths ranging from 20 to 80 h. Specifically, the model achieved an average R² of 0.92, a MAE of 0.38, and a MAPE of 1.77%, demonstrating its suitability for short-term water quality predictions. The methodology and findings of this study demonstrate the significant potential of integrating spectral proper orthogonal decomposition and deep learning for water quality prediction. By overcoming the limitations of traditional models, the proposed surrogate model provides high-accuracy predictions with enhanced physical interpretability, even in complex, dynamic environments. This work offers a practical tool for rapid responses to water pollution incidents and supports improved watershed water quality management by effectively capturing pollutant diffusion dynamics. Furthermore, the model’s scalability and adaptability make it a valuable resource for addressing intelligent management in environmental science.

This study investigates emergency response strategies for sudden pollution incidents in long-distance water diversion tunnels. The tunnel section of the Yin Chao Ji Liao Project in Inner Mongolia is used as a case study. A one-dimensional hydrodynamic water quality model was developed in Storm Water Management Model (SWMM) to analyze pollutant transport characteristics in the tunnel under various operating conditions. Based on the actual engineering conditions, control scenarios with multiple flow rates and multiple gate combinations were set up. Emergency control strategies for sudden pollution events were developed to address extreme pollution scenarios. The feasibility of scheduling gate operations according to pollutant transport response time for effective pollution mitigation was evaluated. On this basis, an expression for calculating gate-operation timings for emergency pollution control was derived. The results indicate that the peak concentration in the tunnel shows a decreasing trend as the flow rate increases, and the change process shows a stage-by-stage characteristic. Accounting for the response time of water discharge can improve pollution disposal efficiency by 4.34–52.14%. The efficiency gains become increasingly pronounced at higher flow rates, indicating that this strategy can effectively enhance water discharge efficiency. Installing water quality monitoring instruments near the drainage gate section helps improve the precision of regulation and effectively enhances the timeliness and accuracy of operations, and provides a theoretical reference for on-site emergency regulation and control.

To investigate the movement characteristics of fish eggs and to clarify their movement behavior as flow conditions transition from slow to rapid, a hydrodynamics-based fish egg movement model was proposed. Indoor flume experiments under slow-flow conditions conducted previously by the research group were used as a basis. Twenty-three operating conditions with different water depths and discharges were designed, including fifteen rapid-flow conditions and eight slow-flow conditions. Numerical simulations were performed to examine the influence of flow velocity on fish egg movement under different flow conditions. The results show that fish eggs drift with the flow under different flow conditions, and their longitudinal velocity lags behind the flow velocity. At low flow velocities, the vertical velocity distribution of fish eggs is relatively concentrated. With increasing flow velocity, the vertical velocity becomes more dispersed in the high-velocity range. A power–law relationship exists between flow velocity and the trajectory slope of fish egg movement. When the flow velocity is lower than 0.5 m/s, the trajectory slope varies significantly with flow velocity; when it exceeds 1.2 m/s, the slope approaches a constant value. Water depth has a limited influence on fish egg velocity and trajectory slope under both slow-flow and rapid-flow conditions. By combining the relationships among flow velocity, trajectory slope, and suspension rate, a flow velocity of 0.3 m/s is identified as the critical flow velocity for maintaining the safe drifting of fish eggs. The findings provide technical support for ecological operation strategies aimed at fishery resource conservation.