Satellite remote sensing is the most efficient way to monitor global CO₂ flux. ‘Full physics’ retrieval algorithms applied to Greenhouse Gases Observing Satellite (GOSAT) observations are introduced in this paper, and the differences among the algorithms are briefly summarized. The quantity of retrieval data from each algorithm is analyzed, and the spatial coverage indicates that use of only one dataset is insufficient for the study of XCO₂(column-averaged CO₂ dry-air mixing ratio). Therefore, an ensemble average method that fuses four datasets is applied, which aims to increase the data spatial coverage indirectly. Using the ensemble average results, the spatial and temporal distribution of XCO₂ over China is studied. The results indicate strong variation of XCO₂, both spatially and temporally. A seasonal trend is identified, with the maximum and minimum appearing in spring and summer, respectively, over the whole of China, and most of the area shows large XCO₂ values[>380 ppm (×10^-6)]. However, there is a significant difference between east and west. In the east of China, strong CO₂ sources due to high levels of human activity, and sinks due to large areas of vegetation cover, lead to large variation in XCO₂(8 ppm). Whereas, in the west of China, the relatively sparse human population and vegetation cover lead to small variation in XCO₂(5 ppm).