Burning season emissions of reactive nitrogen from fires in subtropical southern Africa determined with TROPOMI and IASI

Widespread and intense dry season burning in subtropical southern Africa (2-20°S), peaking in June to September, results from ignition by humans for agricultural practices and is propagated by a continuous dry savanna landscape. These fires produce large quantities of the reactive nitrogen compounds nitrogen oxides (NO_x≡NO+NO₂) and ammonia (NH₃), influencing tropospheric ozone and aerosol budgets and so affecting climate. Here we use observations of NO₂ from the TROPOspheric Monitoring Instrument (TROPOMI) and NH₃ from the Infrared Atmospheric Sounding Interferometer (IASI) to evaluate NO₂ and NH₃ simulated by the GEOS-Chem model driven with 3 distinct biomass burning inventories. These inventories differ in their use of satellite data products to constrain the timing, extent, severity and longevity of fires. The Global Fire Emissions Database version 4 with small fires inventory (GFEDv4s) uses burned area, Fire INventory from NCAR version 2.5 (FINNv2.5) uses fire counts detected as thermal anomalies, and the Global Fire Assimilation System version 1.2 (GFASv1.2) uses fire radiative power. All use similar landcover-specific, temporally static emission factors. The emissions from these inventories for the 2019 burning season are the same for NO_x from GFEDv4s and FINNv2.5 (4.3 Tg as NO), peaking in July for GFEDv4s and August for FINNv2.5, and much less for GFASv1.2 (1.5 Tg), peaking in August. For NH₃, emissions from GFEDv4s (0.68 Tg) and GFASv1.2 (0.52 Tg) are about half that from FINNv2.5 (1.3 Tg) and peak emission months are the same as NO_x for GFEDv4s and FINNv2.5, but a month earlier (July) for GFASv1.2. Averaging kernels from the satellite products are used to mitigate influence of the vertical sensitivity of the instrument and a priori assumption of the vertical profile of NH₃ when comparing to GEOS-Chem. We apply TROPOMI NO₂ averaging kernels to GEOS-Chem for comparison to TROPOMI and re-retrieve IASI NH₃ columns using GEOS-Chem as a priori. In our comparison to TROPOMI, GEOS-Chem monthly mean NO₂ driven with GFASv1.2 and GFEDv4s is more spatially consistent (R ≥ 0.9) than FINNv2.5 (R = 0.4-0.6) and is typically just 12-27% less than TROPOMI using GFASv1.2, compared to up to 80% more for GFEDv4s and 55% more for FINNv2.5. The much greater NO_x emissions from GFEDv4s and FINNv2.5 contribute to tropospheric ozone chemical production that totals 38-49 Tg in June-September compared to 29 Tg for GFASv1.2, though differences in volatile organic compound emissions will also influence these production rates. Assessment against IASI NH₃ is underway.