Satellite data-driven estimates of NO x and co-emitted CO 2 from coal-fired power plants in China (2021–2024)

Abstract Coal-fired power generation dominates China’s electricity mix and accounts for over one third of fossil CO 2 emissions, with substantial air pollutants. We estimate nitrogen oxides (NO x ) and co-emitted CO 2 emissions from more than 1000 coal-fired power plants in China during 2021–2024 using a lightweight top-down framework. The method combines TROPOMI NO 2 observations, ERA5 wind fields, and variable NO x /NO 2 ratios from Goddard earth observing system composition forecast within an improved directional derivative approach to derive gridded anthropogenic NO x flux. Plant-level NO x emissions are obtained by integrating weighted contributions from multiple plants within a 15 km radius at monthly temporal resolution, facility-specific CO 2 /NO x emission ratios are applied to infer CO 2 . Estimated annual NO x emissions decrease from 5.49 to 4.78 Mt during 2021–2024, while CO 2 emissions decrease from 4.61 to 4.10 Gt. Corresponding uncertainties are 30.9%–31.5% for NO x and 31.3%–32.0% for CO 2 . A confidence assessment based on internal statistical consistency and external physical plausibility constraints classifies 85.1% of plant-level estimates as high and medium confidence, showing good agreement between observational indicators and derived emissions and supporting the robustness of the framework. The approach performs less reliably for low-capacity facilities and for regions with sparse satellite coverage or relatively high wind speeds.