Articles | Volume 14, issue 18
Atmos. Chem. Phys., 14, 10363–10381, 2014
Atmos. Chem. Phys., 14, 10363–10381, 2014

Research article 30 Sep 2014

Research article | 30 Sep 2014

Worldwide biogenic soil NOx emissions inferred from OMI NO2 observations

G. C. M. Vinken1, K. F. Boersma2,3, J. D. Maasakkers*,1, M. Adon4,5, and R. V. Martin6,7 G. C. M. Vinken et al.
  • 1Department of Applied Physics, Eindhoven University of Technology, Eindhoven, the Netherlands
  • 2Department of Meteorology and Air Quality, Wageningen University, Wageningen, the Netherlands
  • 3Climate Observations Department, Royal Netherlands Meteorological Institute, De Bilt, the Netherlands
  • 4Laboratoire d'Aérologie, UMR CNRS/UPS 5560, Toulouse, France
  • 5Laboratoire de Physique de l'Atmosphère, Université Félix Houphouët-Boigny, Abidjan, Côte d'Ivoire
  • 6Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, Canada
  • 7Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA
  • *now at: Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA, USA

Abstract. Biogenic NOx emissions from soils are a large natural source with substantial uncertainties in global bottom-up estimates (ranging from 4 to 15 Tg N yr−1). We reduce this range in emission estimates, and present a top-down soil NOx emission inventory for 2005 based on retrieved tropospheric NO2 columns from the Ozone Monitoring Instrument (OMI). We use a state-of-science soil NOx emission inventory (Hudman et al., 2012) as a priori in the GEOS-Chem chemistry transport model to identify 11 regions where tropospheric NO2 columns are dominated by soil NOx emissions. Strong correlations between soil NOx emissions and simulated NO2 columns indicate that spatial patterns in simulated NO2 columns in these regions indeed reflect the underlying soil NOx emissions. Subsequently, we use a mass-balance approach to constrain emissions for these 11 regions on all major continents using OMI observed and GEOS-Chem simulated tropospheric NO2 columns. We find that responses of simulated NO2 columns to changing NOx emissions are suppressed over low NOx regions, and account for these non-linearities in our inversion approach. In general, our approach suggests that emissions need to be increased in most regions. Our OMI top-down soil NOx inventory amounts to 10.0 Tg N for 2005 when only constraining the 11 regions, and 12.9 Tg N when extrapolating the constraints globally. Substantial regional differences exist (ranging from −40% to +90%), and globally our top-down inventory is 4–35% higher than the GEOS-Chem a priori (9.6 Tg N yr−1). We evaluate NO2 concentrations simulated with our new OMI top-down inventory against surface NO2 measurements from monitoring stations in Africa, the USA and Europe. Although this comparison is complicated by several factors, we find an encouraging improved agreement when using the OMI top-down inventory compared to using the a priori inventory. To our knowledge, this study provides, for the first time, specific constraints on soil NOx emissions on all major continents using OMI NO2 columns. Our results rule out the low end of reported soil NOx emission estimates, and suggest that global emissions are most likely around 12.9 ± 3.9 Tg N yr−1.

Final-revised paper