NOx emissions constraints from GEMS NO2 retrievals: Inversion methodology and air quality model evaluation in Bangkok using ASIA-AQ multi-platform observations

Abstract

Nitrogen Dioxide (NO2) is a key component of tropospheric chemistry and air quality, yet large uncertainties persist in regional NOx emissions across rapidly developing megacities in Southeast Asia. Observations from the Geostationary Emissions Monitoring Spectrometer (GEMS) provide new constraints on anthropogenic NO2 variability, while the 2024 NASA Airborne and Satellite Investigation of Asian Air Quality (ASIA-AQ) campaign, offers an extensive, independent dataset for model evaluation. Here, we examine air quality in Bangkok using coarse (20 km) and high-resolution (4 km) WRF-Chem simulations during ASIA-AQ. We develop a top-down framework that uses hourly GEMS NO2 columns to derive constraints on the daytime cycle of NOx emissions. Emissions are first estimated from GEMS using a Cross-Sectional Flux (CSF) inversion and then incorporated into WRF-Chem through a novel optimization that reshapes the magnitude and daytime structure of NOx while accounting for lifetime and satellite vertical sensitivity. GEMS-constrained NOx emissions for March 2024 are estimated to range from 2.7 to 4.3 kt month-1 after accounting for known low biases in the GEMS retrievals. Re-running WRF-Chem with the updated emissions leads to substantial improvements in modeled NO2 magnitude and temporal variability when evaluated against independent ground-based, Pandora, and airborne measurements. Remaining negative biases are consistent with a systematic low bias in the GEMS v3 NO2 product, highlighting the importance of multi-platform evaluation using independent observations. Together, these results demonstrate the value of hourly geostationary observations combined with high-resolution modeling as a scalable pathway for improving urban NOx emissions estimates and air quality simulations in Southeast Asia.