A high-resolution hourly XCO ₂ dataset for eastern China (2016–2020) based on WRF-Chem simulations and multi-source validation

We present a 5-year (2016–2020), hourly XCO2 dataset at 12 km resolution over eastern China, generated using the WRF-Chem model with selected physical parameterizations and state-of-the-art anthropogenic emission inventories. A comprehensive sensitivity analysis was conducted to identify the best-performing model configuration, which was subsequently validated against multiple independent sources, including OCO-2 satellite data, TCCON ground-based measurements, ERA5 reanalysis data, GOSIF data, and NIES data. The simulation results achieved strong agreement with observations, with RMSE = 1.64 ppm and [Formula: see text] = 0.85 against OCO-2. The dataset captures seasonal CO2 variations, spatial emission gradients, and long-term growth trends across diverse climate and urban conditions. The high spatiotemporal resolution of this dataset enables its application in atmospheric CO[Formula: see text] modeling studies. It can provide reference fields for data assimilation systems and support the development of data-driven approaches that combine numerical simulations with observational constraints. This may be particularly valuable for carbon cycle research in regions with limited observational coverage, such as eastern China, where understanding emission patterns and atmospheric transport is challenging. The dataset is publicly available at https://doi.org/10.5281/zenodo.15704323.