[1] We describe the development of a model for transport and photochemistry of atmospheric mercury at the regional scale, along with an application to the eastern United States and adjacent Atlantic Ocean and Gulf of Mexico, and comparison with aircraftbased measurements in Florida. The model is the Community Multiscale Air Quality model (CMAQ) with modifications to include an integrated solution for gas phase and aqueous photochemistry. The expanded chemistry includes O 3 , NO x , organics, sulfur, halogens and mercury. Divalent reactive gaseous mercury (RGM) is formed slowly through gas phase reactions and removed rapidly by aqueous reactions in cloud water. Model results show that elevated RGM (up to 260 pg m À3 ) forms intermittently over the Atlantic Ocean in air masses that have a cloud-free history. Aircraft measurements in Florida show RGM varying between 10 and 250 pg m À3 and increasing with altitude, a pattern that is consistent with model results. Ambient RGM would increase by 50% if aqueous reduction reactions were omitted. The model predicts that ambient elemental mercury and RGM anticorrelate in regions where RGM is produced photochemically and correlate in regions dominated by direct emissions. Model results also suggest positive correlations between RGM and SO 2 , reactive nitrogen and H 2 O 2 , which may be used to identify photochemically produced versus directly emitted RGM. RGM in the model is strongly correlated with O 3 during pollution events, and ozone formation from anthropogenic precursors is predicted to cause a significant increase in RGM.

Section 112(m) of the 1990 Clean Air Act Amendments, referred to as the Great Waters Program, mandated an assessment of atmospheric deposition of hazardous air pollutants (HAPs) to Lake Champlain. Mercury (Hg) was listed as a priority HAP and has continued to be a high priority for a number of national and international programs. An assessment of the magnitude and seasonal variation of atmospheric Hg levels and deposition in the Lake Champlain basin was initiated in December 1992 which included event precipitation collection, as well as collection of vapor and particle phase Hg in ambient air. Sampling was performed at the Proctor Maple Research Center in Underhill Center, VT. The range in the annual volume-weighted mean concentration for Hg in precipitation was 7.8-10.5 ng/l for the 11-year sampling period and the average amount of Hg deposited with each precipitation event was 0.10 microg/m2. The average amount of Hg deposited through precipitation each year from 1993 to 2003 was 9.7 microg/m2/yr. A seasonal pattern for Hg in precipitation is clearly evident, with increased Hg concentrations and deposition observed during spring and summer months. While a clear trend in the 11-year event deposition record at Underhill was not observed, a significant decrease in the event max-to-monthly ratio was observed suggesting that a major source influence was controlled over time. Discrete precipitation events were responsible for significant fractions of the monthly and annual loading of Hg to the forested ecosystem in Vermont. Monthly-averaged temperatures were found to be moderately correlated with monthly volume-weighted mean Hg concentrations (r2 = 0.61) and Hg deposition (r2 = 0.67) recorded at the Vermont site. Meteorological analysis indicated the highest levels of Hg in precipitation were associated with regional transport from the west, southwest, and south during the warmer months.

No abstract

No abstract