Tropospheric NO2 inference from airborne APEX hyperspectral data Spreaker: Frederik Tack, Aeronomie Belgium

Nitrogen dioxide (NO2) is a key pollutant with strong harmful effects on human health and with highly variable concentrations in space and time. NO2 is currently inferred based on ground stations and satellite observations. On the one hand, the ground stations measure NO2 with high precision, but only deliver punctual values which cannot be easily extrapolated. On the other hand, satellite observations ensure a large spatial coverage, but have a relative coarse spatial resolution. Quantitative information about the NO2 horizontal variability at high spatial resolution is currently scarce, but very valuable for a better assessment of the risks induced by NO2 emissions on the human population.

APEX is an airborne pushbroom hyperspectral imager with high spatial (60 by 80 m2) and spectral (2.8 - 3.3 nm) resolution. APEX flights were conducted over several cities in Belgium (Antwerp, Brussels) and Germany (Berlin) in 2015 and 2016, in the framework of the BUMBA, AROMAT and AROMAPEX projects. NO2 vertical column densities (VCDs) are retrieved based on the DOAS analysis of the observed spectra in the visible region (470 nm - 510 nm), and air mass factor calculations with the radiative transfer model VLIDORT 2.6. Results show that APEX is suitable (1) to detect the fast varying spectral signatures of a trace gas like NO2 and (2) to identify small scale gradients in the NO2 field and to resolve individual emission sources. Both the NO2 retrieval scheme and campaign results will be presented.

Frederik Tack (1982) is a Geographer and obtained a PhD in Remote Sensing in 2012 (UGent, Belgium). Since then he is working as an atmospheric remote sensing scientist in the UV-Vis DOAS group at the Belgian Institute for Space Aeronomy (BIRA) with main expertise in the retrieval of trace gases, such as NO2 and SO2, from airborne sensors. Additionally, he gained a strong expertise in mobile-DOAS applications (car, bike), in order to obtain cross-validation horizontal distribution data.