Quantifying urban fossil fuel CO2 emissions through OCO-2 XCO2 analysis: Impacts of satellite overpass geometry and prior inventory selection

Cities, as the primary source of fossil fuel carbon dioxide (ffCO 2 ) emissions, are responsible for approximately 70% of global ffCO 2 emissions, making precise monitoring of urban ffCO 2 emissions essential for achieving global emission reduction targets. This study selected Ordos, a high-emission city, as the research area. We used three mainstream fossil fuel carbon emission inventories : ODIAC, EDGAR, and MEIC , for cross-validation and comparative analysis. We integrated column-averaged carbon dioxide (XCO 2 ) data from OCO-2 satellite observations and used a Bayesian inversion method to optimize emission factors. After fully considering satellite observation errors and atmospheric transport model uncertainties, we systematically analyzed the sources of uncertainty in the inversion results and investigated the influence of orbit type and emission inventory selection on the accuracy of urban ffCO 2 inversion. The results show that, for the selected overpasses, MEIC-based simulations showed relatively closer agreement with observed winter enhancements, whereas ODIAC-based simulations were more consistent with the summer cases analyzed in this study. The average posterior uncertainty for the three emission inventories was reduced by 50%, 55%, and 50.63%, respectively. When wind speeds were between 2 and 3.5 m s −1 , the consistency between the simulated increments from optimized transit overpasses and the observed increments significantly improved. By systematically analyzing the effects of inventory type, prior uncertainty, overpasses and environmental factors on the inversion results, we validated the feasibility of this method for effectively estimating the impact of urban fossil fuel carbon emissions.