Global 0.1° Monthly Anthropogenic CO₂ Emissions Dataset with Near-Real-Time Updates

This dataset provides global monthly anthropogenic CO₂ emission estimates at a spatial resolution of 0.1° × 0.1°, with the latest update extending through September 2026. The dataset is designed to provide a valuable resource for assessing the global carbon budget and the effectiveness of climate mitigation efforts. The estimates are generated using a deep learning–based framework that employs a convolutional neural network as its backbone and incorporates a mask fusion module to address spatially missing input data. The framework integrates satellite observations, meteorological reanalysis data, and temporal and spatial encoding features to estimate anthropogenic CO₂ emissions at a global scale. The data are provided in NetCDF format on a global 0.1° × 0.1° grid (1800 × 3600 cells) using the WGS84 geographic coordinate system. Each file contains monthly anthropogenic CO₂ emission estimates for all 12 months of a given year, expressed in tonnes month⁻¹ cell⁻¹. Files follow the naming convention ACO2_01deg_YYYY.nc, where YYYY denotes the corresponding year. The dataset is also updated regularly at doi: 10.12237/casearth.6a01788fe8a21d1686165027 A manuscript describing the methodology and dataset is in preparation.