Line source estimation of environmental pollutants using super-Gaussian geometry model and bayesian inference

Area of Science:

• Environmental Science
• Atmospheric Chemistry
• Geophysics

Background:

• Pollution sources often have linear geometries, such as roadways.
• Current source term estimation (STE) methods typically model sources as points, neglecting geometric details.
• This simplification can lead to inaccuracies in STE.

Purpose of the Study:

• To develop and validate a novel line source estimation method.
• To incorporate geometric information into source term estimation.
• To address the limitations of point-source assumptions in STE.

Main Methods:

• A Bayesian inference framework combined with a super-Gaussian function was employed.
• The super-Gaussian function allows for intuitive approximation of source shapes.
• The method was tested via simulation in an urban boundary layer and experimentally in a wind tunnel.

Main Results:

• The proposed method successfully estimated line source information without prior geometric data.
• Simulations and wind tunnel experiments confirmed the method's effectiveness.
• Conventional point-based methods failed to accurately estimate the line source.

Conclusions:

• Geometric information is crucial for accurate source term estimation.
• The developed line source estimation method offers a significant improvement over point-source models.
• This approach is essential for reliable pollution source characterization.