Abstract: Rapid increases in Carbon dioxide (CO₂) levels could trigger unpredictable climate change. The assessment of spatiotemporal variation and influencing factors of CO₂ concentration are helpful in understanding the source/sink balance and supporting the formulation of climate policy. In this study, Greenhouse Gases Observing Satellite (GOSAT) data were used to explore the variability of CO₂ concentrations in China from 2009 to 2020. Meteorological parameters, vegetation cover, and anthropogenic activities were combined to explain the increase in CO₂ concentration, using pixel-based correlations and Covariance Based Structural Equation Modeling (CB-SEM) analysis. The results showed that the influence of vertical CO₂ transport diminished with altitude, with a distinct inter-annual increase in CO₂ concentrations at 17 vertical levels. Spatially, the highest values were observed in East China, whereas the lowest were observed in Northwest China. There were significant seasonal variations in CO₂ concentration, with maximum and minimum values in spring (April) and summer (August), respectively. According to the pixel-based correlation analysis, the near-surface CO₂ concentration was positively correlated with population (r = 0.99, P < 0.001), Leaf Area Index (LAI, r = 0.95, P < 0.001), emissions (r = 0.91, P < 0.001), temperature (r = 0.60, P < 0.05), precipitation (r = 0.34, P > 0.05), soil water (r = 0.29, P > 0.05), nightlight (r = 0.28, P > 0.05); and negatively correlated with wind speed (r = −0.58, P < 0.05). CB-SEM analysis revealed that LAI was the most important controlling factor explaining CO₂ concentration variation (total effect of 0.66), followed by emissions (0.58), temperature (0.45), precipitation (0.30), wind speed (−0.28), and soil water (−0.07). The model explained 93% of the increase in CO₂ concentration. Our results provide crucial information on the patterns of CO₂ concentrations and their driving mechanisms, which are particularly significant in the context of climate change.