As urbanization intensifies, the quantification of methane (CH4) emissions at city scales faces unprecedented challenges due to spatial heterogeneities from industrial and transportation activities and land use changes. This paper provides a review of the current state of top-down atmospheric CH4 emission inversion at the city scale, with a focus on CH4 emission inventories, CH4 observations, atmospheric transport models, and data assimilation methods. The Bayesian method excels in capturing spatial variability and managing posterior uncertainty at the kilometer-scale resolution, while the hybrid method of variational and ensemble Kalman approaches has the potential to balance computational efficiency in complex urban environments. This review highlights the significant discrepancy between top-down inversion results and bottom-up inventory estimates at the city scale, with inversion uncertainties ranging from 11% to 28%. This indicates the need for further efforts in CH4 inversion at the city level. A framework is proposed to fundamentally shape city-scale CH4 emission inversion by four synergistic advancements: developing high-resolution prior emission inventories at the city scale, acquiring observational data through coordinated satellite–ground systems, enhancing computational efficiency using artificial intelligence techniques, and applying isotopic analysis to distinguish CH4 sources.
Li et al. (Thu,) studied this question.