Influence of uncertainties on the inverse modeling of Texas NOx emissions using satellite NO2 observations

Abstract This work explores the influence of uncertainties on performing inverse modeling combined with satellite NO2 observations to develop a top-down NOx (NO + NO2) emission inventory for Texas air quality modeling. Applying inverse modeling techniques with NO2 measurements to constrain NOx emissions can help indicate possible biases in current NOx emission inventories, which may have a significant impact on model performance and mislead control strategies in air quality management. However, the inverse modeling results mainly rely on the performance of the modeling system and the quality of the measurement data. This study first applies the Discrete Kalman filter (DKF) inversion technique combined with Ozone Monitoring Instrument (OMI) satellite NO2 measurements to the regional Comprehensive Air Quality Model with extensions (CAMx) model, to create a top-down NOx emission inventory for Texas ozone abatement modeling. Secondly, the influence of model uncertainties that may impact modeled NO2 vertical columns such as choosing different vertical mixing schemes, photolysis rates, and chemical mechanisms, and that of using different OMI retrieval products which may affect measured NO2 vertical columns have been tested separately on the inverse modeling system, to investigate the robustness of the DKF inversed NOx emissions.