Refining NOₓ Emissions using Satellite Observations: Inverse Modeling through DART–CHIMERE Data Assimilation of S5P/TROPOMI NO₂ Retrievals

Emission inventories data used in chemical transport models (CTMs) are subject to uncertainties that propagate into air quality simulations. Air quality data from satellite observations can provide additional constraints on emissions, enabling a top-down approach that complements conventional bottom-up inventories. 

In this work, we performed an inverse modeling within the framework of the DART–CHIMERE data assimilation system. A state vector augmentation method is applied to NOₓ emission fields, allowing emissions to be adjusted along with initial chemical concentrations. This approach aims to mitigate the limited persistence of corrections obtained through initial-condition-only assimilation, which are often damped by CTM dynamics.

The methodology is tested over the European domain for S5P/TROPOMI NO₂ total column retrievals, and the impact of emission adjustments is evaluated using independent surface NO₂ measurements from ground-based monitoring stations. First results based on a test case are presented to illustrate the potential of the approach. While the approach does not aim to replace established bottom-up inventories, the results indicate that satellite-informed emission corrections can provide additional, dynamically consistent constraints, supporting their use as a complementary component in CTM-based air quality analyses.