Specific patterns of XCO 2 observed by GOSAT during 2009–2016 and assessed with model simulations over China

Spatiotemporal patterns of column-averaged dry air mole fraction of CO 2 (XCO 2 ) have not been well characterized on a regional scale due to limitations in data availability and precision. This paper addresses these issues by examining such patterns in China using the long-term mapping XCO 2 dataset (2009–2016) derived from the Greenhouse gases Observing SATellite (GOSAT). XCO 2 simulations are also constructed using the high-resolution nested-grid GEOS-Chem model. The following results are found: Firstly, the correlation coefficient between the anthropogenic emissions and XCO 2 spatial distribution is nearly zero in summer but up to 0.32 in autumn. Secondly, on average, XCO 2 increases by 2.08 ppm every year from 2010 to 2015, with a sharp increase of 2.6 ppm in 2013. Lastly, in the analysis of three typical regions, the GOSAT XCO 2 time series is in better agreement with the GEOS-Chem simulation of XCO 2 in the Taklimakan Desert region (the least difference with bias 0.65±0.78 ppm), compared with the northern urban agglomeration region (−1.3±1.2 ppm) and the northeastern forest region (−1.4±1.4 ppm). The results are likely attributable to uncertainty in both the satellite-retrieved XCO 2 data and the model simulation data.