S3 The North American mercury speciation networks: Analysis and modeling results

Thursday, 28 July, 2011

RS3-P1 — 11:00-12:00 and 17:30-18:30
Authors: DALZIEL, John1, TORDON, Robert 1, BEAUCHAMP, Stephen1
(1) Environment Canada, john.dalziel@ec.gc.ca

Environment Canada is measuring the levels of three gaseous Hg species – gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), particulate bound mercury (PBM) in the fine fraction (<2.5 µm) from two sites in Atlantic Canada. One site is located in a coastal city, Halifax (Nova Scotia), while the second site is located about 100 km from the coast in central southwestern Nova Scotia within Kejimkujik National Park. This poster illustrates and discusses the GEM, GOM and PBM data collected at both sites over a continuous twelve month period from December 1, 2009 to December 1, 2010. The year long data set from Halifax for GEM had a median of 1.60 ng m-3, and a range of 0.63 to 370 ng m-3 with GOM having a median of 1.37 pg m-3 with a range from detection limit (dl) to 62 pg m-3 and PBM a median of 1.92 pg m-3 with a range from dl to 41 pg m-3. The very elevated levels (<300 ng m-3) of Halifax GEM observed on two occasions also had corresponding elevated levels (>20 pg m-3) of PBM. The mercury species data from the site at Kejimkujik over the same time period have a GEM median of 1.39 ng m-3 and a range of 0.32 to 2.83 ng m-3 with a GOM median of 0.16 pg m-3 with a range from the dl to 12.2 pg m-3 and PBM a median of 1.85 pg m-3 and a range from dl to 33.9 pg m-3. For both sites, seasonal GEM trends were evident showing the lowest median GEM levels occurring during summer and fall. The GEM data from Kejimkujik also showed scavenging occurring in this air shed during the late night and early morning hours of summer and fall with the minimum levels in GEM occurring around 07:00. GOM from both Kejimkujik and Halifax was higher in the spring with a diurnal trend of highest GOM occurring during in the late afternoon (16:00 to 18:00) during spring and to a lesser extent summer months. Seasonal or diurnal trends for PBM form both sites were not as clearly evident in this study. This poster will illustrate and discuss trends in the atmospheric mercury speciated data from both sites from this study.

RS3-P2 — 11:00-12:00 and 17:30-18:30
Authors: PILOTE, Martin1, POISSANT, Laurier1, BEAUVAIS, Conrad2
(1) Environment Canada, Martin.pilote@ec.gc.ca; (2) Environnment Canada.

Mercury is known to impact the aquatic and terrestrial environment. The dominant form in the atmosphere is gaseous elemental mercury (GEM). The residual forms are the reactive gaseous mercury (RGM) including HgCl2, particulate mercury (PM) and various organic mercury forms. Atmospheric mercury speciation and deposition are critical to understanding the fate of mercury in the environment. Atmospheric mercury speciation and deposition were investigated in the upper St. Lawrence Valley in at rural site, St. Anicet, and in a small nearby industrial town, Valleyfield. One year measurements (2009) of GEM, RGM and PM concentrations in St. Anicet were respectively 0.48 - 12.58 ng/m3 (average 1.45 ng/m3), 0 - 548.92 pg/m3 (average 3.86 pg/m3), 0 - 2871.71 pg/m3 (average 15.49 pg/m3). The annual average of PM and RGM measured in St. Anicet were < 1.5 % of the total mercury. In addition, the backward trajectory calculated for some Hg values reveal various sources, from New England and the Maritime Provinces to northern Québec and the Hudson Bay. Spatial trends of Hg were investigated during an intensive field campaign at both sites between June 16 and August 15, 2009. GEM, RGM and PM measurements at the industrial site were respectively higher than at the rural site, 2.27 and 1.49 ng/m3, 14.18 and 2.72 pg/m3, as well as 19.51 and 12.69 pg/m3. The windrose drawn at both sites pointed out some similar industrial mercury sources. Further, mercury speciation fluxes were estimated in St. Anicet, August 24 to September 29, 2009. For this purpose, a well demonstrated technique using synchronized automated high-time-resolution mercury speciation fluxes and dry deposition measurements of GEM, RGM and PM was used. The average GEM flux was -88.92 ng/m2/h (positive sign means volatilization, negative sign indicates deposition) with a maximum deposition value of 1723.30 ng/m2/h and a maximum volatilization value of 1002.54 ng/m2/h. The average RGM flux was -0.87 ng/m2/h with a maximum deposition value of 36.25 ng/m2/h and a maximum evasion value of 26.73 ng/m2/h. The average PM flux was -3.75 ng/m2/h with a maximum deposition value of 315.75 ng/m2/h and a maximum resuspension value of 115.94 ng/m2/h. The median dry deposition velocities calculated from flux and concentration measurements were 0.03 cm/s (GEM) < 0.33 cm/s (RGM) < 1.38 cm/s (PM). Emission sources, regional atmospheric chemistry, and near-ground micrometeorological conditions may all influence the distribution of the atmospheric mercury speciation and deposition.

RS3-P3 — 11:00-12:00 and 17:30-18:30
Authors: FELTON, H Dirk1, CIVEROLO, Kevin L1
(1) New York State DEC, hdfelton@gw.dec.state.ny.us

The New York State Department of Environmental Conservation (NYSDEC) has operated two Tekran speciated Hg monitors collocated with Mercury Deposition Network (MDN) monitors for almost 3 years in urban areas. The sites in Rochester and in the Bronx were selected because little was known about the influences of the urban co-pollutant mix on the ratios of speciated Hg compounds as measured by the Tekran system, or total Hg as measured by the MDN program.

The Tekran and MDN data are presented and compared to other rural datasets in New York and other urban data sets from locations in New Jersey. The Hg and related pollutant data are summarized and diurnal, seasonal and year to year patterns are presented. Additionally, trajectory analysis and pollution roses are used to provide an indication of source regions for each of the Hg species impacting these sites.

These ongoing measurements were initiated with support from an EPA Local-Scale Community Monitoring Grant to provide baseline mercury concentrations to assess the effects of recently implemented emission reductions in the region. The Department is anticipating receiving support from the Great Lakes Commission (GLC) to continue this monitoring program. New York and other states in the Great Lakes region have implemented their own strategies, focusing on coal-fired electric generation facilities and other sources. The data prior to 2010 will be used to provide information on reference baseline ambient mercury levels, and subsequent data collection will be used to track the effects of mercury emission reductions in the region.

RS3-P4 — 11:00-12:00 and 17:30-18:30
Author: LAN, Xin1
(1) University of Houston, xlan3@uh.edu

Speciated atmospheric mercury data collected over the period from 2007 to 2010 at the Environmental Protection Agency and National Atmospheric Deposition Program Atmospheric Mercury Network (AMNet) sites were analyzed for its spatial, seasonal, and annual characteristics across the U.S. Specifically, particulate bound mercury (PBM), gaseous organic mercury (GOM) and gaseous elemental mercury (GEM) data from 25 AMNet sites will be used to determine the mercury chemical distributions and temporal variations. Rigorous data and statistical analyses will be utilized to better understand sources, sinks, and transport of atmospheric mercury in the U.S. Meteorological factors such as air temperature, pressure, and winds will be evaluated for their influence on mercury distributions. Backward air mass trajectories will be used to shed light on the impact of long-range transport. In particular, AMNet site data from CA and UT will be examined to identify influence from trans-Pacific transport from Asia to the U.S. We identified similarities and differences in seasonal and annual cycles in geographic regions of the U.S. and offer potential explanations. Data analysis shows that PBM median levels from different monitoring sites ranged from 1.27 pg m-3 to 13.66 pg m-3 (0.14 ppqv to 1.53 ppqv), whose highest value was found in Rochester, New York. The median of GOM ranged from 0.47 pg m-3 to 12.35 pg m-3 (0.05 ppqv to 1.39 ppqv), with the highest value found in Utah. The GEM median levels ranged from 1.33 ng m-3 to 2.39 ng m-3 (149 ppqv to 267 ppqv), with the highest values in New Jersey. Special focus will be placed on the Northeast, where more than 70% of the AMNet monitoring sites are located. Higher mixing ratios there may be related to major coal burning areas to the west and southwest. Our groups airborne data sets suggest that transport of mercury emitted in urban areas may be very important to downwind locations. This will be examined at U.S. sites with intensive anthropogenic activities upwind. In addition, the influence from the changing distribution of pressure systems and passage of fronts will be explored. Overall, our analysis of the AMNet data should provide a more detailed understanding of atmospheric mercury that can be used to better inform regional and global models.

RS3-P5 — 11:00-12:00 and 17:30-18:30
Author: REN, Xinrong1
(1) NOAA Air Resources Laboratory, Xinrong.Ren@Noaa.gov

Hourly speciated mercury measurements were made at an AMNet site in Grand Bay, Mississippi during the summer 2010 intensive from July 29 to August 14, 2010. In addition, other chemical and meteorological parameters were measured concurrently. These data were analyzed using HYSPLIT back trajectory and principal components analysis in order to develop source-receptor relationships for mercury species in this coastal environment. Trajectory frequency analysis shows that high mercury events were generally associated with high frequencies of the trajectories passing through the areas with high mercury emissions, while low mercury levels were largely associated the trajectories passing through relatively clean areas. Principal components analysis reveals two main factors: combustion and photochemical process that were clustered with high reactive gaseous mercury (RGM) and fine particle mercury (FPM). This study indicates that the receptor site which is located in a coastal environment of Gulf of Mexico experienced impacts from mercury sources that are both local and regional in nature.

RS3-P6 — 11:00-12:00 and 17:30-18:30
Authors: STEFFEN, Alexandra1, SCHERZ, Christina1, OLSON, Mark2, GAY, David2, BLANCHARD, Pierrette1
(1) Environment Canada, Alexandra.Steffen@ec.gc.ca; (2) National Atmospheric Deposition Program, University of Illinois;

Collecting quality atmospheric mercury species data is a challenging task and while significant advances in the measurement of them have been reported in the past 10 years, there is limited information on the quality control (QC) and assurance processes employed. Over the past several years, considerable work has been undertaken within North America (NA) to develop quality control and assurance programs. The Canadian-led Atmospheric Mercury Networks and the American-led National Atmospheric Deposition Network (NADP) both developed their own programs to QC atmospheric mercury speciation data. While these programs were developed in consultation, there exist several differences between each networks approach. The need to have comparable measurements within North America has become evident with initiatives to develop international controls of mercury emissions on the forefront of the United Nations Environment Programme.

A robust investigation of the two QC programs was made to assess the criteria on which the data is quality controlled and to assess the comparability of the final data that emerges from these networks within NA. The programs employed data collected with the TekranTM 2537, 1130 and 1135 instruments which yield measurements of gaseous elemental mercury, reactive gas phase mercury and particulate mercury. Results from this analysis show that the criteria that were established within each network used to flag data were comparable but had some different approaches. Data from four different sites in NA were put through each QC program and the number of flags that emerged was investigated. These results showed that most of the flagging criteria are equivalent. Finally, data from two different locations was put through each program and the final concentrations of the speciated data were compared. Final data produced from a mid-latitude site compared very well (less than 15% difference between the mean concentrations of each species). However, the final data from a site with strong variability and high concentrations revealed that the data sets do not compare as well.

The QC methods applied by these two networks will be discussed and compared and results from the intercomparison of the final data sets will be presented.

RS3-P7 — 11:00-12:00 and 17:30-18:30
Authors: MAESTAS, Melissa M.1, PERRY, Kevin D.1, LISONBEE, Joel R.1
(1)University of Utah, m.maestas@utah.edu

Speciated measurements of atmospheric mercury have been made at the UT96 site since July 1, 2009. The UT96 site is located on the eastern shore of the Great Salt Lake and is subjected to daily lake/land breeze wind reversals. These wind reversals alternately bring air to the receptor site from over the lake (daytime) and from the adjacent urbanized areas (nighttime). The site is also periodically impacted by northerly winds after the passage of cold fronts. These northerly breezes do not pass over the lake or urbanized regions and typically result in pristine (i.e., background) concentrations of atmospheric mercury. The net result of these wind patterns is a single site that can be used to study background mercury concentrations, interactions with a marine-like boundary layer, and local/regional pollution sources.

The annual median concentrations of gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particulate-bound mercury (PBM2.5) for the first year of operations at the UT96 site were 1.56 ng m-3, 4.2 pg m-3, and 9.9 pg m-3, respectively. The monthly-median GEM concentrations peaked during February (1.79 ng m-3) and gradually decreased to a minimum in August (1.48 ng m-3). The monthly-median GOM concentrations had a distinct summer maximum (13.5 ng m-3) and winter minimum (1.0 pg m-3). The monthly-median PBM2.5 concentrations peaked during the winter temperature inversion periods (15.6 pg m-3) and were at the minimum during May (3.7 pg m-3).

Statistically significant diurnal patterns were observed for GEM, GOM, and PBM2.5 throughout the most of the year. GEM concentrations typically peak at night and are minimized during the late afternoon. This diurnal cycle is most pronounced during the summer, but is also evident in the spring and the fall. GOM has the strongest diurnal pattern with concentrations in the late afternoon that are often an order of magnitude higher than the nighttime values during the summer months. This diurnal pattern is observed during all seasons except winter. PBM2.5 concentrations exhibit a diurnal pattern very similar to that of GEM. The only difference is that the diurnal pattern is only observed during the summer and the fall. The UT96 site is occasionally impacted by significant local/regional sources of GEM and PBM2.5. The maximum 2-hr averaged concentrations of GEM and PBM2.5 measured so far were 43.8 ng m-3 and 1412 pg m-3, respectively.

RS3-P8 — 11:00-12:00 and 17:30-18:30
Authors: MOORE, Chris1, SHERWELL, John2, BROOKS, Steve3
(1) Appalachian Laboratory, UMCES, cmoore@umces.edu; (2) Maryland DNR; (3) NOAA.

The purpose of this study was to estimate the annual dry deposition rate of gaseous oxidized mercury (GOM) at our Atmospheric Mercury Network (AMNet, MD08) site in western Maryland. Our annual estimates were made using passive ion-exchange membranes and a multilayer model (Lyman et al. 2007). Our replicated passive filters were deployed for 17 weekly sampling periods from September 2009 to October 2010. During each of our weekly sampling periods, we measured ambient air concentrations of GOM and the meteorological parameters necessary for the multilayer model. From our passive filters, the average weekly dry deposition rate was 360 pg m-2hr-1 within a range of 80 to 1512 pg m-2hr-1. Our lowest deposition rates were in the fall and the highest rates were in the spring. Our GOM dry deposition rates were strongly correlated with the weekly averaged ambient air GOM concentrations, which ranged from 5 to 38 pg m-3. Dry deposition rates of GOM could be predicted from the ambient air concentrations of GOM using this equation: GOM dry deposition rate = 4 x GOM concentration - 59, r2 = 0.7979. Dry deposition velocities, computed from GOM concentrations and passive membrane GOM deposition rates, ranged from ~0.2 in the fall up to 1.4 cm s-1 in winter and spring. For comparison, the dry deposition velocities predicted by the multilayer model ranged from 0.2 up to 5 cm s-2, with no obvious seasonal patterns. In addition, deposition velocities computed using the two different approaches were not correlated. Despite this lack of correlation between dry deposition velocities, our modeled dry deposition rates of GOM were highly correlated with the passive GOM dry deposition rates (modeled dry deposition rate = 0.91 x passive dry deposition rate + 64; r 2 = 0.8394). Using the average dry deposition rate from our passive ion-exchange membranes, we estimated an annual GOM dry deposition rate of 2.5 ug m-2 yr-1. Using the multilayer model, we predicted an annual GOM dry deposition rate of 3.2 ug m-2 yr-1. This value was computed from the sum of all modeled two-hour dry deposition rates. For comparison with wet deposition, annual rates of wet deposition at MD 08 ranged from 5.5 to 10.2 ug m-2yr-1, with an average of 7.6 ug m-2 hr-1. Our estimated annual rates of GOM dry deposition were 1/3 to 1/2 of the average annual wet deposition of mercury at AMNet site MD08.

RS3-P9 — 11:00-12:00 and 17:30-18:30
Authors: LUKE, Winston T.1, KELLEY, Paul2, COHEN, Mark 2, REN, Xinrong2, WALKER, Jake3, COLTON, Aidan4, SUGANUMA, Poai5
(1) NOAA/Air Resources Laboratory (ARL), Winston.Luke@noaa.gov; (2) NOAA/ARL; (3) Grand Bay National Estuarine Research Reserve; (4) NOAA/Earth Systems Research Laboratory (ESRL); (5) NOAA/ESRL

The Headquarters Division of NOAA’s Air Resources Laboratory administers three sites for the long-term monitoring of mercury in the atmosphere: Grand Bay NERR in Moss Point MS; Beltsville, MD; and the latest site at the Mauna Loa Observatory in Hawaii. Overall trends and results from these sites will be compared and contrasted, and data will be interpreted using a combination of regression and principal component analyses, back trajectory analysis, and modeling output. Highlights of two recent research intensives, focusing on the roles of halogen chemistry and transport processes in mercury chemistry, distribution, and conducted at the Grand Bay NERR in Summer 2010 and Spring 2011 will also be presented.

RS3-P10 — 11:00-12:00 and 17:30-18:30
Author: ECKLEY, Chris1
(1) Environment Canada, chris.eckley@ec.gc.ca

Levels of mercury in the Canadian atmosphere have been measured throughout the county as part of the Clean Air Regulatory Agenda (CARA) as well as through existing monitoring networks. The objective of this project is to compile these datasets to provide a holistic national perspective on spatial trends in atmospheric Hg concentrations as well as to highlight temporal trends at the locations where long-term monitoring has occurred. Throughout Canada, total gaseous mercury (TGM) measurements have been conducted at 17 locations and speciated mercury concentrations (gaseous elemental mercury-GEM, particulate bound mercury-PBM, and reactive gaseous mercury-RGM) at 6 locations. Measurement sites include urban, rural, and remote regions as well as areas impacted by specific localized industrial emissions. Atmospheric Hg measurements began in 1996 at a few locations and the monitoring network has expanded over the years to include additional sites. Recent publications have presented data on several aspects of atmospheric Hg trends within Canada (e.g. Temme et al., 2007) and the work presented here expands on these efforts to include contemporary data (through December, 2010) and/or includes additional spatial locations not included in previous assessments. To accurately identify spatial trends, all TGM and speciation datasets were quality controlled using the same system (Research Data Management and Quality Control System-RDMQ). The data measured at the monitoring locations across the country were compared with levels predicted by Environment Canada’s Global/Regional Atmospheric Heavy Metals (GRAHM) model. The results from this project summarize previous and current atmospheric mercury monitoring activities throughout Canada such that spatial and temporal trends can be identified at the national level.

RS3-P11 — 11:00-12:00 and 17:30-18:30
Author: OLSON, Mark1
(1) NADP, mlolson@illinois.edu

The Atmospheric Mercury Network (AMNet) became an official network within the National Atmospheric Deposition Program (NADP) in the fall of 2009. The goal of this network is the measurement of the atmospheric mercury species on a continuous basis. Currently the AMNet consists of 25 sites across North America and stores over 60 years of mercury species measurements which is available to public on the NADP web site. Prior to public release, all data needs to be validated by three different interconnected reviews: the site operator, the site liaison, and the automated AMNet Quality Assurance program. Site operators upload data directly through a FTP protocol or manual methods. Site operators fill out a NADP AMNet field form while on site which is submitted electronically to NADP monthly. This field information is brought into the AMNet database and linked to the ambient Hg records [QA level 1]. Once received, the data are processed by an automated quality assurance program to flag data as valid or invalid with corresponding QA codes [QA level 2]. After automatic review, the Site Liaison reviews all data and updates automated QA program flags [QA level 3]. Next, the site operator is requested to review data, having final say regarding the QA flags for their data. Once changes are made and agreed to by the Site Liaison, the data is finalized and ready for public display [QA level 4].

This presentation, as part of special session three, will provide an overview of the quality assurance, and a live demonstration of the automated QA program and associated flag criteria.

RS3-P12 — 11:00-12:00 and 17:30-18:30
Authors: PRESTBO, Eric1, GAY, David2, EDGERTON, Eric3, LUKE, Winston4, KELLEY, Paul4, FELTON, Dirk5, HOLSEN, Thomas6, HUANG, Jiaoyan6, CHOI, Hyun-Deok7
(1) Tekran Research & Development, eprestbo@tekran.com; (2) University of Illinois; (3) Atmospheric Research & Analysis; (4) NOAA Air Resources Laboratory; (5) New York Department of Enviromental Conservation; (6) Clarkson University; (7) National Institute of Aerospace.

Atmospheric mercury speciation measurements are being made and reported on-line as part of the new Atmospheric Mercury Network (AMNet). It is easy to overlook that mercury is the only atmospheric constituent routinely and continuously measured at the part per quadrillion level (ppqv, mixing ratio). Typical values range from a few hundred ppqv for gaseous elemental mercury (GEM) to 0.5 to 10 ppqv for gaseous oxidized mercury (GOM) and particulate-bound mercury (PBM). For contrast, background ozone concentrations are roughly 30 million times higher than average GOM concentrations. Because of the exceedingly low atmospheric mercury species concentrations, it is technically very difficult to generate and deliver stable and traceable standards to the inlet of automated measurement systems for quality assurance and calibration purposes. Thus, quality assurance has normally consisted of 1) routine automated internal calibration of the detector with a traceable elemental mercury permeation source, 2) manual injections of elemental mercury at locations upstream of the detector and 3) direct intercomparisons of measurements with two or more instruments over a short time period. Both manual injections and direct intercomparisons are done infrequently and few are reported in the literature. Fortunately, within AMNet, there have been 4 sites where two instruments have been co-located for an extended period of time and operating under harmonized and well-defined standard operating procedures. At several sites, the co-located instruments are normally run asynchronously to get hourly GEM, GOM and PBM with no data gaps. However, there are periods over the course of time where the instruments are operated synchronously. A statistical analysis of co-located, synchronous atmospheric mercury speciation data will be presented in order to quantify the uncertainty of the measurements. Additionally, a summary of historical atmospheric mercury speciation quality assurance data will be shown and compared to the results of this study.

RS3-P13 — 11:00-12:00 and 17:30-18:30
Authors: SHARAC, Timothy1, GAY, David2, SCHMELTZ, David 1, OLSON, Mark2, PRESTBO, Eric3, BERGERHOUSE, Tom2, COHEN, Mark4, LUKE, Winston4, KELLEY, Paul4
(1) US EPA, sharac.timothy@epa.gov; (2) NADP; (3) Tekran Research and Development; (4) NOAA;

The NADP Atmospheric Mercury Network (AMNet) is a collaborative effort involving many federal, state, and tribal agencies, academic researchers, and industry partners across North America. AMNet was officially adopted by the NADP Executive Committee in October 2009 The network monitors, summarizes, and reports atmospheric mercury species which contribute to dry and total mercury deposition. At each station, the following data are collected: concentrations of atmospheric mercury species from automated, continuous measuring systems; and concentrations of total mercury in precipitation. Data are collected with standardized methods, quality assured and archived in an NADP online data base.

There are several specific objectives for this network:

  • determine the status and trends in concentrations of atmospheric mercury species (gaseous oxidized, particulate-bound, and elemental) in select locations;
  • offer high-quality measurement data to estimate dry and total deposition of atmospheric mercury to aquatic ecosystems and other locations influenced by emissions and transport on the local, regional, and global scale; and
  • offer high-quality data necessary for atmospheric mercury model evaluation and development.

The NADP organization supports a network that can produce data that are accessible, quality assured, and comparable. AMNet will be one component of an assessment that will be closely associated with mercury research (e.g., dry deposition studies) and modeling efforts. Work within AMNet focuses on continued development of a quality assurance program including enhancing the data reduction program, testing and improving the standard operating procedure and posting NADP atmospheric mercury speciation monitoring data onto the website. We will also continue to reach out to a broad cross-section of agencies and institutions to coordinate mercury monitoring activities, building on current efforts and encouraging new collaborative partnerships. The poster will summarize the status of the network including the structure for participation, standardization of methods, quality assurance, site locations, data handling, and example data plots.

Thursday, 28 July, 2011