A DiffeRential Evolution Adaptive Metropolis (DREAM)-based inverse model for continuous release source identification...

Area of Science:

• Environmental Science
• Water Resource Management
• Computational Modeling

Background:

• Continuous river pollution poses significant risks to water ecosystems and requires effective emergency response strategies.
• Existing research primarily addresses instantaneous pollution sources, highlighting a gap in managing continuous discharge events.
• Accurate identification of continuous pollution sources is crucial for mitigating ecological damage and informing response actions.

Purpose of the Study:

• To propose a novel inverse model for identifying continuous point sources in river pollution incidents.
• To estimate key source parameters: strength, location, release time, and spill duration.
• To quantify uncertainties in source parameter estimation.

Main Methods:

• Development of an inverse model integrating the Differential Evolution Adaptive Metropolis (DREAM) algorithm with a forward transport advection-dispersion equation.
• Inference of the posterior probability distribution of source parameters to quantify uncertainties.
• Comparative performance analysis against Metropolis-Hastings (MH) and Genetic Algorithm (GA) based models.

Main Results:

• The DREAM-based model demonstrated high accuracy in identifying continuous pollution sources for both hypothetical and field tracer scenarios.
• Comparative analysis revealed the DREAM model's superiority in computational efficiency, result accuracy, and pollutant concentration reconstruction.
• Observation errors were found to significantly impact the accuracy of the identification results.

Conclusions:

• The proposed DREAM-based inverse model offers a robust and efficient solution for identifying continuous river pollution sources.
• Model performance is sensitive to the dispersion coefficient and river velocity, underscoring the importance of accurate hydrological data.
• Enhancing monitoring network density and proximity to spill sites can further improve the accuracy of pollution source identification.