Atmospheric mercury concentration and dry deposition measurements using active and passive systems

This thesis focuses on mercury (Hg) sources, fate and transport in the atmosphere. Hg concentrations were measured and its source identified in an urban area of upstate New York. Hg concentrations in Rochester were between those typically measured in rural and in suburban sites. The snow-melting factor was correlated with elemental Hg (Hg0) concentrations and temperature, and was always observed in winter. The coal combustion factor composed by reactive gaseous mercury (RGM) and sulfur dioxide (SO2) was found during the entire campaign, and was associated with a local coal-fired power plant and regional transport from the Ohio valley using CPF. High ozone (O3) mixing ratios synchronous with RGM was classified gas phase oxidation in the warmer seasons. Mobile sources were found to be a potential source based on the positive correlation between Hg0 and carbon monoxide (CO) and highly negative relationship with O3 mixing ratios. Non-coal combustion was verified as a major source with high Hg, PM2.5 concentrations and CO mixing ratios. There are three minor Hg sources based on PCA results, including wet deposition and clean air (both negative), and Hg emission from soil. The concentrations of the three mercury species from the direction of the coal-fired power plant (CFPP) were significantly reduced following the shutdown of this source.In the investigation of two commonly used surrogate surfaces to directly measure Hg dry deposition, both plates were found to create a well defined laminar boundary layer near their centers. Hg dry deposition velocity (V d) was determined as a function of position on the surrogate surface and the bulk wind speed using a simplified knife-edge surrogate surface (KSS). Increasing Vd was observed with increasing wind speed for both plates however, Vd was found to be approximately 30% higher to the frisbee-shape surrogate surface (FSS) than to the KSS. The higher V d was caused by a thinner boundary layer. CFD results suggest that the Vd to the FSS is less variable than to the KSS for different plate angles with respect to the wind however, experimental results suggest the opposite.CFD simulations coupled with wind tunnel experiments were used to determine the sampling rate of a widely used passive sampler. In the wind tunnel experiments, water evaporation rates were determined by weight-change of a water-saturated PUF disk installed in the sampler at various wind speeds and sampler angles. These rates were converted to polychlorinated biphenyl (PCB) sampling rates using ratios of air diffusivities and compared with field measurements. Overall, modeled (3D CFD) and measured sampling rates agree well and are similar to rates found in field studies. Sampling rates shows linear correlation with wind speed. Changing the orientation of the sampler with respect to the angle of the wind either decreased or increased the sampling rate depending on the angle used. The sampling rate increased when the base was tilted up pointing into the wind. However, when the base was tilted down so the top of the sampler was facing the wind the sampling rate decreased initially and then increased.The static water surrogate surface (SWSS) was hypothesized to collect deposited elemental mercury (Hg0), reactive gaseous/oxidized mercury (RGM), and mercury associated with particulate matter (Hgp) while the QFF, KCl-coated QFF, and gold-coated QFF on the KSS were hypothesized to collect Hgp, RGM + Hgp, and Hg0 + RGM + Hgp, respectively. The Hg flux measured by the deionized water was significantly smaller than that captured by the acidified water probably because Hg0 was oxidized to Hg2+ which stabilized the deposited Hg and decreased mass transfer resistance to low pH solutions. Acidified BrCl captured significantly more Hg than other solutions because of high Hg0 oxidized rate. However, of all collection media, gold-coated filters showed the highest Hg collection efficiency. Fluxes measured with the QFF and KCl-coated QFF were small and generally not different than the blanks probably because of the relatively low atmospheric concentrations at the site. For all other media (except deionized water), co-located samples agreed well and the combination of daytime plus nighttime results were comparable to 24-h samples, implying that Hg0, RGM and Hgp were not released after they deposited nor did the surfaces reach equilibrium with the atmosphere. (Abstract shortened by UMI.)...