High-Resolution Daily CO₂ Dataset Illuminates Spatiotemporal Dynamics in China
09 Feb 2026
Researchers from the Aerospace Information Research Institute of the Chinese Academy of Sciences (AIRCAS), in collaboration with Chongqing University of Posts and Telecommunications (CQUPT), have developed a high-resolution, daily atmospheric CO₂ dataset for China spanning 2016–2020. The dataset provides unprecedented insights into the spatiotemporal variations of column-averaged dry-air CO₂ mole fraction (XCO₂) and offers a robust data foundation for refined carbon monitoring and management.
Led by Associate Researcher YANG Aixia and Senior Engineer WANG Dacheng of AIRCAS, together with Prof. YUAN Zhengwu of CQUPT, the research team constructed the dataset using a novel XGBoost–Bayesian Optimization (XGBoost-BO) framework. The approach effectively addresses key limitations of existing satellite-based carbon monitoring systems, including spatial imbalance, temporal discontinuity, and sensitivity to meteorological conditions. To enhance model interpretability, the team incorporated SHapley Additive exPlanations (SHAP), enabling quantitative assessment of the relative contributions of climate factors, vegetation dynamics, and human activities to XCO₂ variability.
With a spatial resolution of 0.1° × 0.1°, the dataset delivers continuous daily XCO₂ coverage across China, significantly filling gaps in satellite-retrieved observations. It integrates multi-source data, including satellite measurements from OCO-2 and GOSAT, ground-based observations from TCCON, vegetation indices (NDVI, EVI), meteorological variables from ERA5-Land, anthropogenic emissions data (ODIAC), nighttime light data from VIIRS, and global fire emissions data (GFED), ensuring comprehensive representation of both natural and anthropogenic CO₂ sources and sinks.
Validation results demonstrate the dataset's high accuracy and reliability. Comparisons with OCO-2 observations yield an R² of 0.98, RMSE of 0.58, and MAPE of 0.07%. Independent evaluation using TCCON observations at the Hefei and Xianghe sites further confirms its superior performance, consistently outperforming the CAMS global greenhouse gas reanalysis. At the Hefei site, for example, the dataset achieves an R² of 0.92, RMSE of 1.16, and MAPE of 0.2%, compared with CAMS values of 0.88, 1.39, and 0.3%, respectively.
The study was published in Scientific Data. The dataset is openly available on Zenodo, and the model code has been released on GitHub, facilitating broad application in climate change research, carbon emission assessment, and policy analysis. By providing high-precision, continuous XCO₂ data, the dataset enables more accurate characterization of carbon cycle processes and supports evidence-based environmental governance.
This research was supported by the National Key Research and Development Program of China. The dataset is expected to serve as a fundamental resource for refined carbon management, contributing to China's efforts to address climate change.
Research News
High-Resolution Daily CO₂ Dataset Illuminates Spatiotemporal Dynamics in China
Researchers from the Aerospace Information Research Institute of the Chinese Academy of Sciences (AIRCAS), in collaboration with Chongqing University of Posts and Telecommunications (CQUPT), have developed a high-resolution, daily atmospheric CO₂ dataset for China spanning 2016–2020. The dataset provides unprecedented insights into the spatiotemporal variations of column-averaged dry-air CO₂ mole fraction (XCO₂) and offers a robust data foundation for refined carbon monitoring and management.
Led by Associate Researcher YANG Aixia and Senior Engineer WANG Dacheng of AIRCAS, together with Prof. YUAN Zhengwu of CQUPT, the research team constructed the dataset using a novel XGBoost–Bayesian Optimization (XGBoost-BO) framework. The approach effectively addresses key limitations of existing satellite-based carbon monitoring systems, including spatial imbalance, temporal discontinuity, and sensitivity to meteorological conditions. To enhance model interpretability, the team incorporated SHapley Additive exPlanations (SHAP), enabling quantitative assessment of the relative contributions of climate factors, vegetation dynamics, and human activities to XCO₂ variability.
With a spatial resolution of 0.1° × 0.1°, the dataset delivers continuous daily XCO₂ coverage across China, significantly filling gaps in satellite-retrieved observations. It integrates multi-source data, including satellite measurements from OCO-2 and GOSAT, ground-based observations from TCCON, vegetation indices (NDVI, EVI), meteorological variables from ERA5-Land, anthropogenic emissions data (ODIAC), nighttime light data from VIIRS, and global fire emissions data (GFED), ensuring comprehensive representation of both natural and anthropogenic CO₂ sources and sinks.
Validation results demonstrate the dataset's high accuracy and reliability. Comparisons with OCO-2 observations yield an R² of 0.98, RMSE of 0.58, and MAPE of 0.07%. Independent evaluation using TCCON observations at the Hefei and Xianghe sites further confirms its superior performance, consistently outperforming the CAMS global greenhouse gas reanalysis. At the Hefei site, for example, the dataset achieves an R² of 0.92, RMSE of 1.16, and MAPE of 0.2%, compared with CAMS values of 0.88, 1.39, and 0.3%, respectively.
The study was published in Scientific Data. The dataset is openly available on Zenodo, and the model code has been released on GitHub, facilitating broad application in climate change research, carbon emission assessment, and policy analysis. By providing high-precision, continuous XCO₂ data, the dataset enables more accurate characterization of carbon cycle processes and supports evidence-based environmental governance.
This research was supported by the National Key Research and Development Program of China. The dataset is expected to serve as a fundamental resource for refined carbon management, contributing to China's efforts to address climate change.