Parameter optimization method for the water quality dynamic model based on data-driven theory.

Area of Science:

• Environmental Science
• Computational Fluid Dynamics
• Ecological Modeling

Background:

• Accurate parameterization is crucial for developing reliable dynamic water quality models.
• Complex three-dimensional models require efficient methods for parameter selection and optimization.
• Data-driven approaches offer a promising avenue for improving model calibration.

Purpose of the Study:

• To apply a data-driven method for selecting and optimizing parameters in a complex 3D water quality model.
• To couple a data-driven model with a physical model for enhanced simulation accuracy.
• To utilize the Particle Swarm Optimization (PSO) algorithm for efficient parameter optimization.

Main Methods:

• Developed a data-driven model to establish relationships between phytoplankton and environmental factors.
• Established and coupled an eight-variable water quality dynamic model with a physical model.
• Performed parameter sensitivity analysis and employed PSO for control parameter optimization.

Main Results:

• The data-driven model successfully trained the response relationship between phytoplankton and environmental factors.
• Parameter sensitivity analysis provided guidelines for control parameter selection.
• The PSO algorithm effectively optimized control parameters for the water quality model in Xiangshan Bay.

Conclusions:

• The integrated data-driven and PSO approach is effective for parameter optimization in complex water quality models.
• This methodology enhances the accuracy and reliability of dynamic water quality simulations.
• The optimized model provides a valuable tool for water quality management in Xiangshan Bay.