New approach for point pollution source identification in rivers based on the backward probability method.

Area of Science:

• Environmental Science
• Water Resource Management
• Chemical Engineering

Background:

• River pollution from industrial discharge and spills threatens water security.
• Rapid identification of pollution sources is crucial for effective emergency response and pollutant management.

Purpose of the Study:

• To develop and validate a novel approach for identifying the source, timing, and quantity of sudden water pollution events in rivers.
• To improve the accuracy and efficiency of pollution source identification compared to existing methods.

Main Methods:

• The study proposes the Linear Regression-Backward Probability Method (LR-BPM), integrating Linear Regression (LR) and Backward Probability Method (BPM).
• The ill-posed source identification problem is transformed into an optimization model solved by the Differential Evolution Algorithm (DEA).
• Decoupled released mass parameters do not require prior information, enhancing identification efficiency.

Main Results:

• Hypothetical case studies demonstrated relative errors below 10% for identified location, release time, and released mass.
• Model equifinality was identified as a source of uncertainty, but averaging results from repeated tests significantly reduced errors.
• Increasing the number of gauging sections further improved identification accuracy.

Conclusions:

• The LR-BPM offers a robust and efficient solution for point source identification of sudden river pollution.
• Real-world application confirmed the LR-BPM's superior accuracy and time-saving benefits over Bayesian-MCMC and basic DEA methods.
• The method provides critical data for timely emergency disposal and effective river pollution management.