Evaluation of high resolution simulated and OMI retrieved tropospheric NO2 column densities over Southeastern Europe

High resolution model estimates (10 × 10 km2) of tropospheric NO2 column amounts from the Comprehensive Air Quality Model (CAMx) for the Balkan Peninsula are compared with OMI/Aura measurements (13 × 24 km2 at nadir) for the year April 2009 to March 2010. The Balkan area contributes significantly to the NO2 burden in European air and so numerous urban, industrial and rural regions are studied aiming to investigate the consistency of both satellite retrievals and model predictions at high spatial resolution. It has already been shown that OMI can detect the tropospheric column of NO2 over polluted Balkan cities due to its fine horizontal resolution and instrument sensitivity (Zyrichidou et al., 2009). In this study the improved OMI DOMINO v2.0 satellite retrievals showed that over South-Eastern Europe the monthly mean NO2 tropospheric column density fluctuated between 2.0 and 5.7 ± 1.1 × 1015 molecules/cm2 over urban areas, 1.6-5.0 ± 0.7 × 1015 molecules/cm2 over large industrial complexes and 1.1-2.2 ± 0.4 × 1015 molecules/cm2 over rural areas for the year studied. The Comprehensive Air Quality Model with extensions (CAMx) version 4.40 is a publicly available open-source computer modeling system for the integrated assessment of gaseous and particulate air pollution. The anthropogenic emissions used in CAMx for the Greek domain being studied were compiled employing bottom-up approaches (road transport sector, off-road machinery, etc.) as well as other national registries and international databases. The rest of the Balkan domain has natural and anthropogenic emissions based on the TNO emission inventory of 2003. The high-resolution CAMx simulations reveal consistent spatial and temporal patterns with the OMI/Aura data. The annual spatial correlation coefficient between OMI and CAMx computed in this high spatial resolution analysis is of the order of 0.6, somewhat improved over those estimated in Zyrichidou et al. (2009) (R ≈ 0.5). However, in such a validation study it is important to take into account the averaging kernel (AK) information in order to achieve the creation of comparable data sets. Minor differences are found for area-averaged model columns with and without applying the kernel, which shows that the impact of limiting the effect of the a priori profile on the comparison is on average small. The main aim of the paper, which was to evaluate OMI retrieved and high resolution simulated tropospheric NO2 column densities over South-Eastern Europe and to assess the use of the averaging kernels, is achieved and the two data sources are being employed further in an inverse emission inventory creation study (Zyrichidou et al., in preparation).

► Similar spatiotemporal patterns (R > 0.5) and seasonality between OMI and CAMx. ► The main patterns of tropospheric NO2 are discerned in satellite and model plots. ► Simulated tropospheric NO2 columns are overestimated over Athens and Thessaloniki. ► CAMx shows a slight underestimation over the domain the summer except for urban regions when averaging kernels are used. ► CAMx with the averaging kernels information and OMI DOMINO v2.0 are better correlated (R = 0.64) over the urban sites.