Articles | Volume 12, issue 19
Atmos. Chem. Phys., 12, 8993–9011, 2012
https://doi.org/10.5194/acp-12-8993-2012
Atmos. Chem. Phys., 12, 8993–9011, 2012
https://doi.org/10.5194/acp-12-8993-2012

Research article 02 Oct 2012

Research article | 02 Oct 2012

Estimation of mercury emissions from forest fires, lakes, regional and local sources using measurements in Milwaukee and an inverse method

B. de Foy1, C. Wiedinmyer2, and J. J. Schauer3 B. de Foy et al.
  • 1Department of Earth and Atmospheric Sciences, Saint Louis University, MO, USA
  • 2National Center of Atmospheric Research, Boulder, CO, USA
  • 3Civil and Environmental Engineering, University of Wisconsin – Madison, WI, USA

Abstract. Gaseous elemental mercury is a global pollutant that can lead to serious health concerns via deposition to the biosphere and bio-accumulation in the food chain. Hourly measurements between June 2004 and May 2005 in an urban site (Milwaukee, WI) show elevated levels of mercury in the atmosphere with numerous short-lived peaks as well as longer-lived episodes. The measurements are analyzed with an inverse model to obtain information about mercury emissions. The model is based on high resolution meteorological simulations (WRF), hourly back-trajectories (WRF-FLEXPART) and a chemical transport model (CAMx). The hybrid formulation combining back-trajectories and Eulerian simulations is used to identify potential source regions as well as the impacts of forest fires and lake surface emissions. Uncertainty bounds are estimated using a bootstrap method on the inversions. Comparison with the US Environmental Protection Agency's National Emission Inventory (NEI) and Toxic Release Inventory (TRI) shows that emissions from coal-fired power plants are properly characterized, but emissions from local urban sources, waste incineration and metal processing could be significantly under-estimated. Emissions from the lake surface and from forest fires were found to have significant impacts on mercury levels in Milwaukee, and to be underestimated by a factor of two or more.

Download
Altmetrics
Final-revised paper
Preprint