Performance Assessment of Spatial Interpolation Methods for the Estimation of Atmospheric Carbon Dioxide in the Wider Geographic Extent

Spatially continuous observed atmospheric carbon dioxide (CO2) is necessary to validate spatially resolved emission inventory and measure policy effectiveness in reducing greenhouse gas emissions. However, currently available ground-based observation networks are insufficient to meet this purpose. Moreover, the atmospheric transport model-based CO2 concentration data are mostly criticized for variable emission amounts as model input. The study presents a methodological approach to resolve this limitation by interpolating the satellite-based observed column-averaged carbon dioxide (XCO2) database that offers the advantage of dense spatial coverage. Three spatial interpolation methods (SIMs), Inverse distance weighting (IDW), Spline, and Ordinary Kriging (OK), were used for this purpose. The SIMs’ performances were evaluated and compared based on three statistical indices. Besides, we used ground-based station observed data in original and bootstrapped samples with replacement to evaluate the variability of the interpolated XCO2 with the observed station data. The study concludes that the Spline method interpolates XCO2 better than IDW and OK to a wider spatial extent. On the other hand, the OK shows better results in explaining the ground-based station data. This study approach is limited to capturing the annual variability of XCO2 in the interpolation methods due to the unavailability of consecutive periods’ satellite-based quality data points in a wider spatial range.