Publications

Li, J; Jia, K; Wei, XQ; Xia, M; Chen, ZL; Yao, YJ; Zhang, XT; Jiang, HY; Yuan, B; Tao, GF; Zhao, LL (2022). High-spatiotemporal resolution mapping of spatiotemporally continuous atmospheric CO2 concentrations over the global continent. INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION, 108, 102743.

Abstract
Carbon dioxide (CO2) in the atmosphere is an important variable that connects the atmosphere and terrestrial ecosystems. However, satellite-observed atmospheric CO2 concentrations are always spatially discrete, and spatiotemporally continuous atmospheric CO2 concentration maps with fine resolution are scarce at the global scale. Therefore, a spatiotemporally continuous CO2 concentration dataset with high-spatiotemporal resolution was generated based on satellite CO2 observations and environmental factors in this study. First, the environmental factors affecting atmospheric CO2 concentrations were selected, and by integrating the atmospheric CO2 concentration data observed by the OCO-2 satellite, a sample dataset was created. The extreme random tree (ERT) model was then trained to determine the relationship between environmental factors and atmospheric CO2 concentrations. Finally, spatiotemporally continuous atmospheric CO2 concentration data (0.01 degrees and 8-day resolution) over the global continent were generated based on the established ERT model and environmental data. Validation results based on ground atmospheric CO2 concentration observations indicated that the estimated atmospheric CO2 concentration achieved satisfactory performance with coefficients of determination and root mean square errors of 0.83 and 1.79 ppm, respectively. The spatiotemporal analysis found that the distribution of atmospheric CO2 concentrations had obvious seasonal patterns and spatial heterogeneity across different climatic zones. The generated atmospheric CO2 concentration data have the potential for subsequent studies of global climate change and carbon cycling.

DOI:
10.1016/j.jag.2022.102743

ISSN:
1872-826X