Fugitive greenhouse gas (GHG) emissions in industrial parks are characterized by high opacity and spatial dispersion. Existing localization and quantification methods often rely on idealized meteorological assumptions and low-precision mobile monitoring data, making it difficult to achieve accurate source characterization. This study focuses on the Nansha Economic and Technological Development Zone in Guangzhou—one of the first pilot zones for synergistic pollution and carbon reduction in China—to develop an atmospheric inversion model based on multi-site fixed monitoring. By integrating GHG concentrations with multi-dimensional meteorological parameters, the model couples an atmospheric dispersion framework with a Bayesian inversion algorithm. Specifically, site-specific conditions and high-frequency meteorological data are utilized to constrain dispersion parameters, effectively reducing model uncertainty driven by meteorological variability. Within the Bayesian framework, the model enables the simultaneous inversion of both the locations and emission strengths of multiple sources. Results identified three distinct fugitive emission sources: one primary source in the International Auto Industrial Park with a CO2 emission intensity of 103.15 g/s and two sources in the Western Industrial Park with intensities of 0.051 g/s and 0.26 g/s, respectively. Overall, this research framework significantly enhances the accuracy and spatial resolution of emission inversion, providing robust technical support for precision carbon management and the development of targeted mitigation strategies for key industrial processes.
Hu et al. (Thu,) studied this question.