In recent years, the concentration of atmospheric carbon dioxide in China has been increasing year by year. Satellite observation is the main means of obtaining atmospheric carbon dioxide concentration. However, currently, spaceborne sensors used to measure atmospheric carbon dioxide have a narrow observation range and cannot obtain spatially and temporally continuous atmospheric carbon dioxide concentrations. Therefore, this dataset proposes a daily full coverage XCO2dataset generation method based on the DSC-DF-LGB (Deep Separable Convolutional Neural Network and Deep Forest concatenated with LightGBM) model to obtain the spatiotemporal distribution of atmospheric carbon dioxide in China. The purpose of establishing the DSC-DF-LGB model is to train the mapping relationship between OCO-2 XCO2retrieval and related variables (reanalysis of XCO2, vegetation parameters, human factors, altitude, and meteorological parameters). The model was used to generate a daily 0.1 ° full coverage XCO2 dataset in China from 2015 to 2020. The XCO2dataset with full coverage and high resolution can provide data support for carbon source and sink research
Including estimated XCO2data and on-site measured XCO2data
method for generating a daily full coverage XCO dataset based on the DSC-DF-LGB (Deep Separable Convolutional Neural Network and Deep Forest concatenated with LightGBM) model was proposed to obtain the spatiotemporal distribution of atmospheric carbon dioxide in China
The cross validation (CV) results of XCO2show that the model has strong performance in estimating XCO2, with R2and RMSE of 0.9633 and 0.9761 ppm, respectively. The independent on-site verification results of TCCON indicate that the estimated XCO2 is highly consistent with the measured values on site, with R2and RMSE of 0.8786 and 1.5452 ppm, respectively
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
CSTR:11738.11.NCDC.ZENODO.DB6751.2025
10.5281/zenodo.13352556