Spatio-temporal consistency evaluation of XCO2 retrievals from GOSAT and OCO-2 based on TCCON and model data for joint utilization in carbon cycle research

Abstract

The global carbon cycle research requires precise and sufficient observations of the column-averaged dry-air mole fraction of CO2 (XCO2) in addition to conventional surface mole fraction observations. In addition, assessing the consistency of multi-satellite data are crucial for joint utilization to better infer information about CO2 sources and sinks. In this work, we evaluate the consistency of long-term XCO2 retrievals from the Greenhouse Gases Observing Satellite (GOSAT), Orbiting Carbon Observatory 2 (OCO-2) in comparison with Total Carbon Column Observing Network (TCCON) and the 3D model of CO2 mole fractions data from CarbonTracker 2017 (CT2017). We create a consistent joint dataset and compare it with the long-term model data to assess their abilities to characterize the carbon cycle climate. The results show that, although slight increasing differences are found between the GOSAT and TCCON XCO2 in the northern temperate latitudes, the GOSAT and OCO-2 XCO2 retrievals agree well in general, with a mean bias ± standard deviation of differences of 0.21 ± 1.3 ppm. The differences are almost within ± 2 ppm and are independent of time, indicating that they are well calibrated. The differences between OCO-2 and CT2017 XCO2 are much larger than those between GOSAT and CT XCO2, which can be attributed to the significantly different spatial representatives of OCO-2 and the CT-transport model 5 (TM5). The time series of the combined OCO-2/GOSAT dataset and the modeled XCO2 agree well, and both can characterize significantly increasing atmospheric CO2 under the impact of a large El Niño during 2015 and 2016. The trend calculated from the dataset using the seasonal Kendall (S-K) method indicates that atmospheric CO2 is increasing by 2-2.6 ppm per year.