Impacts of global NO x inversions on NO 2 and ozone simulations

Abstract. Tropospheric NO2 and ozone simulations have large uncertainties, but their biases, seasonality, and trends can be improved with NO2 assimilations. We perform global top-down estimates of monthly NOx emissions using two Ozone Monitoring Instrument (OMI) NO2 retrievals (NASAv3 and DOMINOv2) from 2005 to 2016 through a hybrid 4D-Var/mass balance inversion. Discrepancy in NO2 retrieval products is a major source of uncertainties in the top-down NOx emission estimates. The different vertical sensitivities in the two NO2 retrievals affect both magnitude and seasonal variations of top-down NOx emissions. The 12-year averages of regional NOx budgets from the NASA posterior emissions are 37 % to 53 % smaller than the DOMINO posterior emissions. Consequently, the DOMINO posterior surface NO2 simulations greatly reduced the negative biases in China (by 15 %) and the US (by 22 %) compared to surface NO2 measurements. Posterior NOx emissions show consistent trends over China, the US, India, and Mexico constrained by the two retrievals. Emission trends are less robust over South America, Australia, western Europe, and Africa, where the two retrievals show less consistency. NO2 trends have more consistent decreases (by 26 %) with the measurements (by 32 %) in the US from 2006 to 2016 when using the NASA posterior emissions. The performance of posterior ozone simulations has spatial heterogeneities from region to region. On a global scale, ozone simulations using NASA-based emissions alleviate the double peak in the prior simulation of global ozone seasonality. The higher abundances of NO2 from the DOMINO posterior simulations increase the global background ozone concentrations and therefore reduce the negative biases more than the NASA posterior simulations using GEOS-Chem v12 at remote sites. Compared to surface ozone measurements, posterior simulations have more consistent magnitude and interannual variations than the prior estimates, but the performance from the NASA-based and DOMINO-based emissions varies across ozone metrics. The limited availability of remote-sensing data and the use of prior NOx diurnal variations hinder improvement of ozone diurnal variations from the assimilation, and therefore have mixed performance on improving different ozone metrics. Additional improvements in posterior NO2 and ozone simulations require more precise and consistent NO2 retrieval products, more accurate diurnal variations of NOx and VOC emissions, and improved simulations of ozone chemistry and depositions.