Atmospheric speciation of Southeastern mercury in two contrasting US airsheds

Simultaneous measurement of gaseous elemental, reactive gaseous, and fine particulate mercury took place in Tuscaloosa AL, (urban airshed) and Cove Mountain, TN (non-urban airshed) during the summers of 2002 and 2003. The objective of this research was to (1) summarize the temporal distribution of each mercury specie at each site and compare to other speciation data sets developed by other researchers and (2) provide insight into urban and non-urban mercury speciation effects using various statistical methods. Average specie concentrations were as follows: 4.05 ng m(-3) (GEM), 13.6 pg m(-3) (RGM), 16.4 pg m(-3) (Hg-p) for Tuscaloosa; 3.20 ng m(-3) (GEM), 13.6 pg m(-3) (RGM), 9.73 pg m(-3) (Hg-p) for Cove Mountain. As a result of urban airshed impacts, short periods of high concentration for all mercury species was common in Tuscaloosa. At Cove Mountain a consistent mid-day rise and evening drop for mercury species was found. This pattern was primarily the result of un-impacted physical boundary layer movement, although, other potential impacts were ambient photochemistry and air-surface exchange of mercury. Meteorological parameters that are known to heavily impact mercury speciation were similar for the study period for Tuscaloosa and Cove Mountain except for wind speed (m s(-1)), which was higher at Cove Mountain. For both sites statistically significant (p < 0.0001), inverse relationships existed between wind speed and Hg-O concentration. A weaker windspeed-Hg-O correlation existed for Tuscaloosa. By analyzing Hg concentration-wind speed magnitude change at both sites it was found that wind speed at Cove Mountain had a greater influence on Hg-O concentration variability than Tuscaloosa by a factor of 3. Using various statistical tests, we concluded that the nature of Tuscaloosa's atmospheric mercury speciation was the result of typical urban airshed impacts. Cove Mountain showed atmospheric mercury speciation characteristics indicative of a non-urban area along with potential influence from steady regional input of mercury pollution from larger sources. (c) 2005 Elsevier Ltd. All rights reserved.