Investigating NO2 distributions from satellite, airborne and ground-based measurements: spatiotemporal variability of NOx emissions and validation of the TROPOMI NO2 product

Abstract

Tropospheric columns of nitrogen dioxide, NO2, a key air pollutant, can be retrieved by differential optical absorption spectroscopy (DOAS) measurements. These measurements can be performed from various observation platforms, including satellites, aircraft, cars, and stationary ground-based sites. Satellite-based measurements provide a global data set of NO2 pollution on a daily basis. With the high spatial resolution TROPOspheric Monitoring Instrument (TROPOMI) on Sentinel-5 Precursor, small-scale emission sources like individual cities and isolated power plants can be probed. This thesis uses TROPOMI tropospheric NO2 columns to quantify the variability of NOx emissions and lifetimes for 50 sources distributed around the world. The retrieved NOx emissions reproduce the variability seen in power plant stack measurements reasonably well but are generally lower than emission inventory data. The NOx emission estimates show a clear seasonality, depending on the dominating source type and location. NOx lifetimes show only a weak seasonal variation but a systematic latitudinal dependence. Except for source regions dominated by industry or power plant emissions, NOx emissions are found to be reduced on weekends compared to working days but with high variability for the analyzed source regions. Strong short-term reductions in NOx emissions were attributable to the COVID-19 containment measures. During the S5P-VAL-DE-Ruhr campaign, airborne imaging, ground-based stationary, and mobile car DOAS measurements were conducted in the Rhine-Ruhr region, one of the NO2 pollution hotspots in Europe. This data set is used to validate TROPOMI’s tropospheric NO2 vertical column density (VCD) product and investigate the known underestimation. Ground-based stationary and car DOAS measurements are used to evaluate the airborne tropospheric NO2 VCDs, showing a reasonably good agreement. The airborne data set is compared to the operational (V01.03.02), a modified reprocessed (V02.03.01), and scientific TROPOMI NO2 products. It is demonstrated that the underestimation of the TROPOMI tropospheric NO2 VCD has been significantly improved by modifications in the cloud product introduced in the V02.03.01 NO2 retrieval. The comparison can be further improved with an additional cloud treatment. Minor improvements are achieved by spatially higher-resolved a priori NO2 profiles and surface reflectivity data. Mobile DOAS measurements are an excellent option to determine the spatial distribution of NO2 or other trace gases but are mainly performed on a campaign basis. To perform daily mobile DOAS measurements, a robust small DOAS instrument was developed and installed on a tram in Bremen. The instrument is introduced, and comparisons to measurements from existing instruments are analyzed, which show good agreement. After installation on the tram, the instrument performed measurements all over the Bremen tram network. These measurements are investigated regarding their spatial distribution of NO2 pollution and are compared to the TROPOMI tropospheric NO2 VCDs, showing similar NO2 distribution patterns.