We compared ground-based (Bruker 125 HR) and satellite (OCO-2) datasets of simultaneous CO2 measurements in the vicinity of St. Petersburg. It is shown that correcting the ground-based XCO2 values by 2.5% and choosing an optimal setup during the spectra analysis allowed us to reach good agreement between the satellite and ground-based measurements. The bias between the two datasets is -0.01-0.16 ppm (-0.00-0.04%) for the means, and the standard deviation of the means is 1.42-1.49 ppm (0.35-0.37%) with a spatial mismatch of XCO2 data pairs of 100-300 km. Such small disagreement between the two types of measurements permits to use both methods for solving the inverse task of atmospheric transfer - the estimation of anthropogenic emissions of CO2.