First Top-Down Diurnal Updates to NOx Emissions Inventory in Asia Informed by the Geostationary Environment Monitoring Spectrometer (GEMS) Tropospheric NO2 Columns

Abstract Pioneering the use of the Geostationary Environment Monitoring Spectrometer’s (GEMS) observation data in air quality modeling, we updated Asia’s NOx emissions inventory by leveraging its unprecedented sampling frequency. GEMS tropospheric NO₂ columns served as top-down constraints, guiding our Bayesian inversion to hourly update NOx emissions in Asia during spring 2022. This effectively remedied the prior underrepresentation of daytime NOx emissions, significantly improving simulation accuracy. The GEMS-informed update reduced the extent of model underestimation of surface NO₂ concentrations from 19.23–11.36% in Korea and from 12.85–4.42% in China, showing about 6% greater improvement compared to the update based on the sun-synchronous low earth orbit observation proxy. Improvements were more pronounced when larger amounts of observation data were available each hour. Our findings highlight the utility of geostationary observation data in fine-tuning the emissions inventory with fewer temporal constraints, thereby more effectively improving the accuracy of air quality simulations.