Abstract: With rapid increase in atmospheric carbon dioxide (CO₂) volume fractions, global climate change brings about unprecedented challenges. In 2020, China sets “dual carbon” goals of achieving carbon peak and carbon neutrality. In this context, the development of long-term, high-resolution datasets for atmospheric CO₂ distribution is of significant practical importance for mitigating the greenhouse effect and promoting low-carbon circular development. The atmospheric carbon dioxide column-average dry air mole fraction (XCO₂) is a critical indicator for characterizing the volume fraction of CO₂ in atmosphere. Although satellite missions specifically designed for XCO₂ observation provide extensive spatial coverage, the data collected often exhibit significant gaps due to interference from clouds, aerosols, and other environmental factors. To address this issue, the study uses XCO₂ data from Orbiting Carbon Observatory-2 (OCO-2) satellite, combined with meteorological, vegetation, and other multi-source information, and employs machine learning techniques to construct a high-precision, seamless daily XCO₂ dataset for China region at 0.05° resolution from January 2015 to March 2024. Among various modeling methods, the XGBoost algorithm exhibits the best performance in cross-validation (determination coefficient is 0.984, root mean square error is 0.894×10^-6) and ground station validation (determination coefficient is 0.912, root mean square error is 1.720×10^-6). Results indicate that XCO₂ in China exhibits a distinct spatial gradient, with higher levels in the eastern regions and lower levels in the western regions. High XCO₂ areas are primarily located in economically developed regions, such as Beijing-Tianjin-Hebei Region, the Yangtze River Delta, and Guangdong-Hongkong-Macao Region. In terms of temporal variation, the national annual average XCO₂ has continuously increased, rising from 401.00×10^-6 in 2015 to 419.91×10^-6 in 2023, with an annual growth rate of 2.36×10^-6. This trend aligns closely with the average annual growth rate of 2.30×10^-6 over the past decade reported in 2023 China Greenhouse Gas Bulletin by China Meteorological Administration. Although the concentration of XCO₂ has continued to rise, the rate of growth has gradually slowed, decreasing from 3.17×10^-6 between 2015 and 2016 to 2.20×10^-6 between 2022 and 2023. Additionally, XCO₂ exhibits distinct seasonal variation, with higher levels in spring and winter, and lower levels in summer and autumn. This seasonal variation may be linked to carbon sequestration effects of vegetation and is consistent with observations from Waliguan Atmospheric Baseline Station. These results may contribute to a deeper scientific understanding of the distribution characteristics of CO₂ and its role in global climate change.