Global gridded time-series of XCO2 concentrations derived from satellite observations and mathematical interpolations

A global product of XCO2 timeseries based on an enhanced version of the XCO2SAT+ method, which incorporates mathematical interpolation techniques in cases with too few or missing satellite observations per location and month. The XCO2SAT+ method uses L2 EMMA (v4.5, https://doi.org/10.24381/cds.f74805c8) as an input. The resulting dataset spans 20 years, from January 2003 to December 2022, and provides monthly averaged XCO2 concentrations. The dataset includes these concentrations for global land locations and adjacent coastal strips for 23958 uniformly sized polygons with an area of each 7774km2 and an approximate distance of 95km between adjacent polygon centers. The XCO2SAT+ method was validated using data from 28 globally distributed sites within the Total Carbon Column Observing Network (TCCON). Analysis revealed a mean absolute error (MAE) of 0.76ppm between the XCO2SAT+ dataset and observations from the TCCON sites over the available overlapping time intervals.