Inverse Modeling of Texas NOx Emissions using OMI NO2 Observations

Uncertain emissions inventories for nitrogen oxides (NOx) are among the leading causes of uncertainty in simulating ozone concentrations and their sensitivities to emissions. Emission inventories from regulatory modeling are typically derived from bottom-up approaches. Satellite observations of NO2 column densities can provide top-down estimates of NOx emissions. Most of the major metropolitan regions in Texas exceed the proposed ozone standards, so accurate emissions inventories to inform attainment efforts are a high priority. Here, we apply NO2 column densities observed by OMI to estimate NOx emission rates over Texas and surrounding states. Modeling episodes developed for recent ozone attainment planning efforts are used to simulate relationships between NOx emissions and NO2 concentrations, using the Decoupled Direct Method (DDM) in CAMx. Inverse modeling uses these sensitivity relationships to estimate a posteriori emissions. Results are compared across alternate inverse modeling approaches that differ in the extent to which they retain the spatial structure of the a priori inventory. We also test the impact of including lightning NO emissions in the a priori inventory, and of incorporating ground-based measurements of NOx along with satellite data in the inversions. The resulting a posteriori emissions estimates are applied in the attainment modeling episodes, to evaluate their impact on model estimates of ozone concentrations and sensitivities to control strategies. At the same time, collaborators on this NASA-funded project are developing GOES satellite-based photolysis rates for these episodes. The relative impacts of satellite-based NOx emissions and photolysis rates will be compared.