ChinaHighXCO₂: A Daily Seamless 10 km Column-Averaged Dry-Air Carbon Dioxide Dataset for China (2014–2023)

ChinaHighXCO2 is a high-spatiotemporal-resolution, seamless, and high-quality column-averaged dry-air carbon dioxide concentration dataset for China. It is generated by fusing multisource satellite observations (i.e., GOSAT, GOSAT-2, OCO-2, and OCO-3) and multiple auxiliary covariates (e.g., meteorological factors, land-surface characteristics, and human activity indicators) using a spatiotemporal enhanced XGBoost model with Bayesian optimization, taking into account the spatiotemporal heterogeneity of XCO2. This dataset provides daily, monthly, and yearly 10 km XCO₂ data for China from 2014 to 2023 and the dataset exhibits high quality, with a 10-fold cross-validation R2 of 0.9704, an RMSE of 1.1772 ppm, and an MAE of 0.7519 ppm.Independent validation against TCCON ground-based observations further demonstrates good agreement, with an R2 of 0.891 at the Hefei site, and an R2 of 0.780 at the Xianghe site..This dataset can effectively characterize the spatiotemporal characteristics of XCO2 across China.