An integrated model for simulating and diagnosing the water quality based on the system dynamics and Bayesian network

Area of Science:

• Environmental Science
• Water Resource Management
• Computational Modeling

Background:

• Effective water quality management requires integrating real-time monitoring with upstream control strategies.
• Existing models often lack the ability to both simulate water quality dynamics and diagnose pollution sources effectively.

Purpose of the Study:

• To develop and validate an integrated model for simulating water quality and diagnosing its deterioration causes.
• To provide a quantitative basis for developing efficient outlet management strategies in river basins.

Main Methods:

• Developed an integrated model combining system dynamics for water quality simulation and Bayesian networks (BN) for diagnostic analysis.
• Applied the integrated model to a case study on the Songhua River, from Baiqi to Songlin sections.

Main Results:

• The integrated model accurately simulated water quality trends, with an average relative error below 10%.
• The diagnostic function identified upstream water quality and reservoir discharge as primary causes of deterioration in the Songlin section.
• The model demonstrated its reasonability and accuracy in a real-world river basin case study.

Conclusions:

• The integrated system dynamics and BN model offers a robust tool for water quality assessment.
• The model provides valuable, quantitative support for basin managers in formulating effective water quality control strategies.
• This approach enhances the ability to prevent further water quality degradation through informed management decisions.