Satellite observations are expected to play an important role in studying carbon fluxes. However, it is necessary to properly remove spatiotemporal bias from these observations. In this study, we estimated the spatiotemporal bias in satellite XCO2 data by making an inversion of in-situ observations. Compared to the GOSAT XCO2 concentrations derived from this analysis, the global average difference is −1.26 ppm, with a difference of −0.76 ppm over the land and −1.54 ppm over the ocean. We then developed a method to correct the bias between the satellite XCO2 and the in-situ inversed XCO2 data. Using this bias correction method, we performed an inverse analysis using in-situ and satellite observations and found that the error relative to the independent observations decreased mainly in the middle and upper troposphere compared to the in-situ inversion. Compared to previous studies, the introduction of satellite observations also suppressed large variations in the regional CO2 flux, except in regions with few observations. After correction using the proposed method, the estimated CO2 flux can be used to provide improved estimates of regional CO2 fluxes; however, the results also show a need for increasing the CO2 observation sites and improving the analysis system.