Evaluation of soil NO emissions in the tropics using field data and TROPOMI NO2 columns. 

Abstract

<p>Nitrogen oxides (NO<sub>x </sub>= NO + NO<sub>2</sub>) play a major role in tropospheric chemistry through their impact on ozone and hydroxyl radical (OH) distributions, and therefore on the oxidizing capacity of the atmosphere. Whereas anthropogenic NO<sub>x</sub> emissions are dominant globally, natural sources are responsible for ca. 30 % of the total emissions into the atmosphere. These sources include soil emissions (due to microbial nitrification and denitrification) and lightning (due to thermal dissociation of O<sub>2</sub> followed by recombination with N<sub>2</sub>). Chemistry-transport models (CTMs) rely on bottom-up (BU) inventories, the uncertainty of which is acknowledged, especially for natural sources. Soil NO<sub>x</sub> is mainly emitted as nitric oxide (NO) and current global BU estimates range from 4 to 15 Tg N yr<sup>-1</sup> with nearly 70% occurring in the tropics. Satellite retrievals of tropospheric NO<sub>2</sub> columns are used as top-down constraints in CTMs to derive NO<sub>x</sub> emissions from various sources (anthropogenic, biomass burning, soil, lightning) such as illustrated in Martin et al. (2003), Jaeglé et al. (2005), Müller et Stavrakou (2005), Stavrakou et al. (2008) or Vinken et al. (2014). This is realized through the method of source inverse modelling, which consists in the optimization of emissions in a CTM in order to minimize the discrepancy between observed and simulated NO<sub>2</sub> columns.</p><p>In this study, we present top-down monthly soil NO<sub></sub>emissions at 0.5° resolution over Africa and South America for 2019 based on spaceborne tropospheric NO<sub>2</sub> columns from the TROPOspheric Monitoring Instrument (TROPOMI). In a first step, we evaluate three global BU inventories against each other and against flux observations over the Tropics. The following BU estimates are considered: (1) YL-MAG, based on the Yienger and Levy parameterization (1995), (2) CAMS, provided by the Copernicus Atmosphere Monitoring Service (Granier et al., 2019; Simpson and Darras, 2021), and (3) HEMCO, calculated using Harvard–NASA Emission Component software (Weng et al., 2020). The last two estimates rely on the parameterization of Hudman et al. (2012). We assess YL-MAG, CAMS and HEMCO inventories against in situ measurements of biogenic soil NO fluxes compiled from literature distinguishing between seasons (dry/wet) and biomes. Based on this evaluation, the best BU inventory is selected and further used as a priori information in the regional MAGRITTE CTM (Müller et al., 2019) run at 0.5°×0.5° resolution for the year 2019. Monthly top-down NO<sub>x</sub> fluxes (from the anthropogenic, biomass burning, soil and lightning categories) are inferred from TROPOMI NO<sub>2</sub> columns using an inversion framework based on the adjoint of MAGRITTE. The top-down soil NO fluxes and NO<sub>x</sub> abundances are subsequently validated against in situ measurements over the two tropical regions.</p>