Articles | Volume 16, issue 18
Atmos. Chem. Phys., 16, 12305–12328, 2016
Atmos. Chem. Phys., 16, 12305–12328, 2016

Research article 29 Sep 2016

Research article | 29 Sep 2016

Interannual variability of ammonia concentrations over the United States: sources and implications

Luke D. Schiferl1, Colette L. Heald1,2, Martin Van Damme3, Lieven Clarisse3, Cathy Clerbaux3,4, Pierre-François Coheur3, John B. Nowak5, J. Andrew Neuman6,7, Scott C. Herndon5, Joseph R. Roscioli5, and Scott J. Eilerman6,7 Luke D. Schiferl et al.
  • 1Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
  • 2Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
  • 3Spectroscopie de l'atmosphère, Chimie Quantique et Photophysique, Université Libre de Bruxelles, Brussels (ULB), Belgium
  • 4LATMOS/IPSL, UPMC Université Paris 06 Sorbonne Universités, UVSQ, CNRS, Paris, France
  • 5Aerodyne Research, Inc., Billerica, Massachusetts, USA
  • 6Cooperative Institute for Research in Environmental Sciences, University of Colorado Boulder, Boulder, Colorado, USA
  • 7Chemical Sciences Division, Earth System Research Laboratory, NOAA, Boulder, Colorado, USA

Abstract. The variability of atmospheric ammonia (NH3), emitted largely from agricultural sources, is an important factor when considering how inorganic fine particulate matter (PM2.5) concentrations and nitrogen cycling are changing over the United States. This study combines new observations of ammonia concentration from the surface, aboard aircraft, and retrieved by satellite to both evaluate the simulation of ammonia in a chemical transport model (GEOS-Chem) and identify which processes control the variability of these concentrations over a 5-year period (2008–2012). We find that the model generally underrepresents the ammonia concentration near large source regions (by 26 % at surface sites) and fails to reproduce the extent of interannual variability observed at the surface during the summer (JJA). Variability in the base simulation surface ammonia concentration is dominated by meteorology (64 %) as compared to reductions in SO2 and NOx emissions imposed by regulation (32 %) over this period. Introduction of year-to-year varying ammonia emissions based on animal population, fertilizer application, and meteorologically driven volatilization does not substantially improve the model comparison with observed ammonia concentrations, and these ammonia emissions changes have little effect on the simulated ammonia concentration variability compared to those caused by the variability of meteorology and acid-precursor emissions. There is also little effect on the PM2.5 concentration due to ammonia emissions variability in the summer when gas-phase changes are favored, but variability in wintertime emissions, as well as in early spring and late fall, will have a larger impact on PM2.5 formation. This work highlights the need for continued improvement in both satellite-based and in situ ammonia measurements to better constrain the magnitude and impacts of spatial and temporal variability in ammonia concentrations.

Short summary
This study combines new observations and a simulation to assess the interannual variability of atmospheric ammonia concentrations over the United States. The model generally underrepresents the observed variability. Nearly two-thirds of the simulated variability is caused by meteorology, twice that caused by regulations on fossil fuel combustion emissions. Adding ammonia emissions variability does not substantially improve the simulation and has little impact on summer particle concentrations.
Final-revised paper