First estimate of NO2 in the upper troposphere from TROPOMI

<p>Nitrogen oxides (NO<sub> x</sub> = NO + NO<sub>2</sub>) in the upper troposphere (~10-12 km) are effective at producing ozone in the upper troposphere where ozone is a potent greenhouse gas. Observations of NO<sub>x</sub> in the upper troposphere are limited in time to a few intensive research aircraft campaigns and in space to commercial aircraft campaigns. There are satellite-derived observations of NO<sub>2</sub> in the upper troposphere from the Ozone Monitoring Instrument (OMI), but these are at very coarse resolutions (seasonal, > 2,000 km). The high-resolution Sentinel-5P/TROPOMI instrument offers higher spatial resolution and better cloud-resolving capability than OMI. Here we use synthetic columns of NO<sub>2</sub> from the GEOS-Chem chemical transport model to assess feasibility of deriving NO<sub>2</sub> in the upper troposphere using partial columns of NO<sub>2</sub> above cloudy scenes (the so-called cloud-slicing technique). The model is also used to quantify errors induced by uncertainties in cloud-top height and to determine whether NO<sub>2</sub> over cloudy scenes is representative of all-sky conditions (the "truth"). We find that the cloud-slicing approach is spatially consistent (R =0.5) with the "truth", but with a small (10 pptv) bias in background NO<sub>2</sub>. Cloud-slicing is then applied to TROPOMI total columns of NO<sub>2</sub> to derive near-global observations of NO<sub>2</sub> in the upper troposphere and assessed against the existing OMI products and aircraft observations from NASA DC8 aircraft campaigns.</p>