Constraining Anthropogenic CO2 Emissions using XCO2 Observations from OCO-3 over Xiamen-Zhangzhou-Quanzhou Metropolitan Area

Accurate quantification and monitoring of urban fossil-fuel CO2 (FFCO2) emissions at improved spatial granularity are critical to emission control and climate change mitigation policies. In this work, we use a top-down Bayesian inversion method and Eulerian transport modeling to constrain FFCO2 emissions from the Xiamen-Zhangzhou-Quanzhou metropolitan area, China, based on high-resolution areal snapshots of total column CO2 (XCO2) from OCO-3 Snapshot Area Maps (SAMs) during September 2019 to July 2023. The emissions from point sources and different areas are constrianed simultaneously, including the area sources Xiamen, local power plants in Xiamen, and other adjacent urban areas. Observed XCO2 enhancements range from 0.70±0.53 ppm to 2.29±1.16 ppm, indicating potential capability of OCO-3 SAMs on detecting emission signatures.  We show the mean posterior emission from Xiamen of about 7.79 ± 0.92 MtC/yr, being within the spread of different inventories and 34 % higher than their average. Several challenges hampering the inversion performance are revealed, including the spatial displacements of the modeled and observed enhancements, the limited representation of local power plants, and data availability. The results provide insights on inversely disentangling imprints of emission sources based on dense space-borne observations, facilitating applications of future missions with improved XCO2 mapping coverage and frequency.