1a: Atmospheric mercury cycling and transformations: Insights from measurements and models

We examine the sources and pathways of potential mercury (Hg) exposure in the context of a case study involving the Aroostook Band of Micmacs based in Presque Isle, ME, using integrated modeling and data analysis. We use GEOS-Chem to quantify the sources of mercury deposition for Maine identifying the contribution of domestic, international, and historical sources , and the HYSPLIT trajectory model to provide further insight into source attribution. Simulated atmospheric concentrations are compared to measurement data for Hg at Presque Isle, ME. Average concentrations, seasonal variation, and diurnal variability are assessed and correlations (r2) calculated between observed and simulated time series. Simulated wet deposition from GEOS-Chem are also compared on a seasonal and annual basis to weekly samples from the Mercury Deposition Network at Caribou, ME. We examine source attribution under policy and no policy cases, identifying how observed atmospheric and wet deposition data would reflect these changes given variations in climate and other drivers. We show preliminary analysis of how these GEOS-Chem outputs can influence fish concentrations and ultimate exposure under desired levels of fish consumption.

Mercury (Hg) is known to be a persistent and toxic heavy metal that can bio-accumulate in the aquatic environment and lead to serious human health effects. Hg is released into the atmosphere from both natural and anthropogenic sources, where in the atmosphere it can be present in a gaseous phase or in particulate matter. In the gaseous phase it can be incorporated with atmospheric precipitation (e.g. rainwater), which is the portion that eventually ends up in the aquatic ecosystem, leading to serious environmental problems. It is known that air trajectory calculations can be helpful in a variety of atmospheric analyses. It enables researchers to understand the transport of pollutants via trajectory routes, assisting in gaining a deeper understanding of pollution events, but more specifically of Hg in rainwater (wet deposition route). Ongoing research at Cape Point, South Africa has shown that most rain events are associated with cold fronts, for which approximately two-thirds reached the Cape Point observatory directly across the Atlantic Ocean from the south, while a third can be attributed to air mass movement from the Cape Town metropolitan region. Results to date collected for the rainy season between May to October annually (2007-2013), have shown that Hg concentrations range between 0.03 to 52.5 ng/L (overall average: 9.91 ng/L). A close relationship was also found between the GEM concentrations in air and TotHg concentrations in rainwater during the raining season. The work in this paper evaluated the existence of any relationships between Hg concentrations in rainwater, to the influence of meteorological variables such as wind direction and air mass backward trajectories. These variables will be valuable in understanding the ongoing monitoring of Hg in wet deposition at the sampling station.

A new regional station (CHC/GAW for short) of the Global Atmosphere Watch, which is part of the World Meteorological Organization (WMO), was set to work in December 2011 at Mount Chacaltaya (5240-5400 masl, 16°21.0´S 68°07.9´). The purpose of this station is to monitor the atmospheric composition of the region, especially the greenhouse effect gases, reactive gases and particle material which is carried to the medium troposphere and to (or from the ) free troposphere. To this purpose, a consortium of European and American institutions set up various instruments in the Cosmic Ray Laboratory.

From July 2014 to February 2016, we investigated atmospheric Total Gaseous Mercury (TGM) using a Tekran 2537A. The instrument worked with a 0.8 LPM flowrate and a sampling time resolution of 15 minutes. In addition to internal calibration using the internal permeation sources, the instrument was calibrated on site with manual injection of mercury vapors. The inlet, protected by a PTFE 0.45µm filter was situated 6 meters above the ground.

The TGM average over the period is around (0.65 ± 0.25) ng/m3 which is ~30% lower than other sites of the Southern Hemisphere (Amsterdam Island, Cape Point) where average TGM is around 1 ng/m3. Records of TGM in the tropical zone of the Southern Hemisphere are scarce, and this is the sole record at such a high altitude. These low values could indicate the existence of a TGM sink in this region, although further work is needed to determine its origin.

Atmospheric signal is partly influenced by polluted air masses from La Paz /El Alto urban areas during daytime due to convective transport of air masses to Chacaltaya station. At nights, aerosol and black carbon measurements clearly indicate that we have free-tropospheric conditions. We observe a clear seasonal feature, with higher TGM values during the rainy season (Oct to Jan) and lower values during the dry season (Jul-Sept). This data set will be useful to determine the influence of regional scale mining activities and the contribution of biomass burning occuring in the Amazon basin area.

In the last few years continuous mercury monitoring has commenced at several sites in the Southern Hemisphere, providing new and more refined information. Under the frame work of the Global Mercury Observation System (GMOS) project, a monitoring station has been set up on Amsterdam Island (37°48’S, 77°34’E) in the remote southern Indian Ocean in January 2012.

For the first time in the Southern Hemisphere, a 3-year record of gaseous elemental mercury (GEM), reactive gaseous mercury (RGM) and particle-bound mercury (PBM) is presented. In 2016, the Tekran 1130-1135 unit was uninstalled, and we now record total TGM using a Tekran 2537B analyzer (1 LPM, 15 minutes). Wet deposition is also measured using an Eigenbrodt wet-only collector.

GEM concentrations are remarkably steady (1.03±0.08 ng m−3) while RGM and PBM concentrations were very low and exhibited a strong variability (mean: 0.34 pgm−3, range: This data set provides a new insight into baseline concentrations of mercury species in the Southern Hemisphere mid-latitudes and new measurement constraints on the mercury cycle, opening the way for new avenues in future modeling studies.

The GEOS-Chem biogeochemical Hg model (like every other major Hg model) has historically been developed and evaluated using observations from the northern hemisphere, which are significantly more abundant than observations in the southern hemisphere. A recent evaluation of GEOS-Chem against a global database of Hg observations found significant biases in simulation of Hg in the southern hemisphere; however, only three southern hemisphere sites were included and none were in Australia. To date, model performance in Australia has not been evaluated due to a lack of available Hg observations. Six years of continuous atmospheric gaseous elemental mercury (GEM) data have now been collected at the Cape Grim Baseline air pollution station located at the northwest tip of Tasmania. Additional time series of ambient GEM, ranging in duration from 1 to 3 years, have been measured at several locations in Australia: a tropical baseline site at Gunn Point NT, a site located near coal-fired power plants and open pit coal mining in Glenville, NSW, and an urban site in Sydney, NSW. These diverse, newly available atmospheric mercury datasets provide an opportunity to test our understanding of the Hg cycle in Australia as embedded in global models. Here, we will present evaluation of the GEOS-Chem model using the new Hg observations along with preliminary model development designed to improve model skill and utility in Australia. Our evaluations to date show that the model is able to simulate observed latitudinal gradients and seasonality, and that all Australian sites are highly sensitive to treatment of ocean Hg exchange. Uncertainties remain in simulating the southern hemisphere GEM background and the air-surface exchange over terrestrial Australian landscapes, and these are the focus of ongoing and future work.

The National Atmospheric Deposition Program (NADP) and its Total Deposition Science Committee is planning to provide estimates of speciated mercury dry deposition using gaseous concentrations measured in the Atmospheric Mercury Network (AMNet). This poster will detail the approach that NADP will use, and provide basic output from this approach. The method will provide fluxes for the three operationally defined mercury forms, i.e., gaseous oxidized mercury (GOM), particulate bound mercury (PBM), and gaseous elemental mercury (GEM). Calculations will be made hourly to multi-hourly depending on availability. Weekly aggregated values (Tuesday to Tuesday) to align with wet deposition observation.

The flux of GOM is estimated as the product of its air concentration and dry deposition velocity calculated using the dry deposition scheme of Zhang et al. (2003; 2012).The flux of PBM is estimated as the product of its air concentration of both fine particulate (<2.5 µm) and an estimated course particulate loading (PBM 2.5-10 µm) and dry deposition velocity of fine and coarse particles calculated according to Zhang and He (2014). The flux of GEM is estimated using a bi-directional air-surface exchange scheme described in Wright and Zhang (2015).

Hourly Meteorological data used are from the archived Canadian weather forecast model at a horizontal grid resolution of 15 km by 15 km and surface and the first model-layer data (typically 40-50 meters agl). Land cover in the vicinity (3 km circle) of monitoring sites is taken from the MODerate resolution Imaging Spectroradiometer (MODIS) land cover type product (MCD12Q1). Fluxes for all existing land covers will be calculated and also aggregated into site-specific (land-cover area-weighted) values. Uncertainty in the estimated fluxes will be provided in the poster. At some point in the future, these estimates will be released for public use through its AMNet webpage.

The concentration of Gaseous Elemental Mercury (GEM) recorded at the Mauna Loa Observatoryin Hawaii between during the period 2002 - 2009 has been analyzed using theEmpirical Mode Decomposition technique.This technique has been used in numerous contexts in order to identify periodical variations in time series data.The periodicities observed in the Sea Surface Temperature collected by five buoys,three in the equatorial Pacific, and two to the south of Hawaii over the sameperiod are also observed in GEM concentrations measured at the MLO.The lag times in the observed periodicities are related to the position of thebuoys with respect to the measurement site. This demonstrates a direct link between climatological phenomena, in this case Sea Surface Temperature, and measured GEMand reflects the influence of ocean Sea Surface Temperature on GEM evasion. This is the first long-term experimental evidence of such a direct effect onGEM evasion from the oceanic surface driven by temperature.

Climate, meteorology and atmospheric chemistry are scientific disciplines that study the same system: the atmosphere. Long-lasting research infrastructures covering all three areas are therefore of highest importance.

One of them, the Atmospheric Station (AS) Křen u Pacova (part of the National Atmospheric Observatory Koetice), central Czech Republic, is focused on monitoring of the occurrence and long-range transport of greenhouse gases, atmospheric aerosols, selected gaseous atmospheric pollutants and basic meteorological characteristics. The AS and its 250 m tall tower was built according to the recommendations of the Integrated Carbon Observation System (ICOS) and cooperates with numerous national and international projects and monitoring programmes. First measurements conducted at ground started in 2012, vertical profile measurements were added in 2013.

The Atmospheric Station (AS) Křen u Pacova consists mainly of a 250 m tall guyed mast of a lattice, 2.6 m wide triangular structure. It was designed and equipped exclusively for scientific purposes according to recommendations by ICOS, ACTRIS (Aerosol, Clouds, and Trace Gases Research Infrastructure Network) and GMOS (Global Mercury Observation System). Measurement data are or will be provided also to the InGOS (Integrated non-CO2 Greenhouse gas Observing System), EMEP (European Monitoring and Evaluation Programme), GAW (Global Atmosphere Watch) and ISKO (Czech Air Quality Information System) databases. The AS was built in 100 m distance from the Koetice Observatory, an infrastructure specialized in air quality and hydrological monitoring since 1988. The character of the site as rural background located in densely populated central Europe, far (> 80 km) from major pollution sources (cities, industry), has been confirmed in numerous studies using air quality data from the Koetice Observatory. Atmospheric long-range transport is expected especially from the west and northwest, comprising also marine air masses as calculated for the Koetice Observatory.

The trends and gradient in atmospheric mercury concentration are part of this measurement programme. Station is equipped by the continuous gaseous elemental mercury (GEM) measurements conducted with two Tekran 2537B instruments installed in a ground-based container and in small technological container in height of 230m on the tower. This sampling design is worldwide unique, and is thoroughly examined and tested.

The detection limit and sensitivity of a Tekran 2537B instrument is < 0.1 ng m-3 (www.tekran.com).

The data sets produced during the period from the opening of this tower until now are now evaluated and validated. Tower represents very important infrastructure for the long-term atmospheric research and monitoring.

With the severe and persistent air pollution (e.g. extremely high concentrations of PM2.5), an emerging challenge is to understand pattern and sources of atmospheric Hg speciation during haze and non-haze days. Measurement of atmospheric mercury speciation was conducted in a coastal city of the Yangtze River Delta, China from July 2013 to January 2014, in conjunction with air pollutants and meteorological parameters. The mean concentrations of gaseous elemental mercury (GEM), particulate bound mercury (HgP) and reactive gaseous mercury (RGM) were 3.26±1.63 ng m-3, 659±931 pg m-3, and 197±246 pg m-3, respectively. High percentages of HgP during haze days were found, due to the increase in direct emissions and gas-particle partitioning of RGM. The average gas-particle partitioning coefficients (Kp) during moderate or severe haze days (PM2.5>150 ug m-3) were obviously decreased. GEM and HgP were positively correlated with PM2.5, SO2, NO2 and CO, suggesting a significant contribution of anthropogenic sources. Elevated HgP concentrations in cold seasons and in the morning were observed while RGM exhibited different seasonal and diurnal pattern. The ratio of HgP/SO2 and pearson correlation analysis suggested that coal combustion was the main cause of increasing atmospheric Hg concentrations. The monitoring site was affected by local, regional and interregional sources. The back trajectory analysis suggested that air mass from northwest China and Huabei Plain contributed to elevated atmospheric Hg in winter and autumn, while southeast China with clean air masses were the major contributor in summer. These results emphasized that both the reduction of anthropogenic emissions from local sources and regional cooperation policy among different city cluster were required to decrease Hg pollution in the atmosphere. Meanwhile, PM2.5 level in developing countries should be controlled to reduce the risks of atmospheric Hg to human health and land ecosystems.

The tropics represent an important region for mercury cycling as it is home to around 40% of the worlds human population, including over 50% of people under the age of 15 - a group at greater risk of adverse effects due to mercury exposure during early development. Stationary observations of gaseous elemental mercury (GEM) taken within the tropics are rare but report significant changes in concentration as source regions shift across hemispheres with the continual drift of the chemical equator. Initiated under the Global Mercury Observation System (GMOS) in June 2014, measurements of GEM are being undertaken at the Australian Tropical Atmospheric Research Station (ATARS) east of Darwin, Australia. The shifting latitude of the inter-tropical convergence zone (ITCZ) in this region leads to significant differences in air mass sources - identified using back trajectory analyses and concomitant radon measurements - over ATARS between the wet monsoon and dry seasons. Mean GEM concentrations over the entire year were 0.95 ng m-3, with higher values over the dry season (largely terrestrial fetch) than the wet season (greater oceanic fetch). GEM concentrations also showed a significant diurnal pattern, with a notable decrease overnight. Using radon as an indicator of atmospheric stability shows that this nocturnal depletion is enhanced under calm, stable boundary layers. Due to the low latitude of Darwin (12 ˚S), impacts from air masses originating in the northern hemisphere at this site are rare, however these events are characterised by increases in GEM concentrations. Back trajectory analyses show that these air masses pass over the populated Indonesian archipelago, suggesting impacts from GEM sources in this region rather than from the northern hemisphere background GEM pool.

Since January 2016, total gaseous mercury (TGM) has been monitored at Chilbolton, Hampshire, UK. Chilbolton became one of the UKs two EMEP supersites (Level II) in 2016 as a replacement for the Harwell Supersite, Oxfordshire, England. Chilbolton is operated by Ricardo-AEA on behalf of the UK Department for the Environment, Food and Rural Affairs. Mercury in air monitoring is undertaken at the site as part of the UK Eutrophying and Acidifying Pollutants (UKEAP) monitoring network run by the UK’s Centre for Ecology & Hydrology. The monitoring site is in a rural setting surrounded by agricultural fields, a short distance from a small village and satellite observatory located at 51.144°N, 1.438°W.

Total gaseous mercury (comprising elemental and gaseous oxidised mercury) was monitored using a Tekran 2537A mercury vapour analyser, run at a resolution of 5 minutes, using dual channels allowing for continuous monitoring. The average TGM concentration at Chilbolton for January – October 2016 was 1.38 ng m-3, whilst the 2015 average for Harwell was 1.44 ng m-3. Both sites showing lower averages than the northern hemispherical background observed in other studies.

We present an initial overview of TGM measurements at the Chilbolton supersite since January 2016 and compares the data to that collected at the previous Supersite at Harwell between 2012 and 2015.

We have used wind sector analysis, cluster analysis and air-mass back trajectories in the OpenAir package in the R statistical software, we show how the TGM concentrations are influenced by local and regional sources (< 50 km) as well as long-range sources at both sites.

Previous work has shown that Harwell was significantly impacted by local sources, through its science campus location and potential mercury remissions due to activity on site, and from its relatively short distance from a coal-fired power station. We address the question of what new or different sources or influences (local or otherwise) are influencing TGM at Chilbolton? What effect has changing monitoring location had on the data? Are the sites comparable? How does that change the context and understanding of atmospheric mercury in the UK?

Mercury (Hg) is a toxic pollutant of concern throughout the northern hemisphere. Atmospheric Hg is often emitted as inorganic forms; however once inorganic Hg is deposited into aquatic ecosystems it can be transformed into MeHg, the most toxic form. Atmospheric deposition has been suggested as an important input for aquatic and terrestrial environments in many previous studies; therefore, quantification of atmospheric Hg deposition is critically needed in order to reduce MeHg levels in aquatic environment. Among atmospheric inorganic Hg species, gaseous divalent form (often called as gaseous oxidized mercury (GOM)) and particulate bound mercury (PBM) are considered to be important with respect to deposition due to high dry and wet deposition velocities although their concentrations are generally much lower than Hg0.

In this study, atmospheric concentrations of GOM and PBM were measured using KCl coated denuder and quartz filter, respectively. Size-distribution of PBM was also investigated using MOUDI sampler. Concurrently, a knife-edge surrogate surfaces using cation-exchange membrane and quartz filter were used to directly measure GOM and PBM dry deposition fluxes, respectively. The measured dry deposition fluxes were compared with the fluxes estimated using three-layer resistance model. Average concentration of GOM was higher in spring (7.47 ± 1.99 pg m-3) than in other seasons (2.76 ± 1.30 pg m-3). Average dry deposition flux of GOM was measured to be 0.74 ± 0.20 ng m-2 h-1 from surrogate surface, which was similar to the previous measurements in Yorkville (0.22 ng m-2 h-1) and Reno (0.79 ng m-2 h-1). There was a strong correlation between GOM concentration and GOM measured flux (R = 0.82), suggesting the rationality of dry deposition device. Measured flux was highly correlated with the estimated flux (R = 0.68); however, the measured flux was approximately 12 times higher than the estimated flux. Detailed results will be presented at the conference.

TraPSA (Trajectory-based Potential Source Apportionment) software is a graphical air pollution source analysis tool based on air pollutant measurements and a state-of-art air mass back trajectories modelHYSPLIT-4. TraPSA provides researchers and students an integrated, user-friendly platform for air pollutant database development and management, pollutant pattern and trend analysis, and potential source identification, by applying, comparing and exploring current popular trajectory ensemble receptor models. A database of pollutant monitoring site data can be established in TraPSA. The smart back-trajectory method in TraPSA helps users easily set up, calculate, and import trajectory data. TraPSA includes current popular algorithms for trajectory ensemble receptor models including Conditional Probability Function (CPF), Concentration Field Analysis (CFA), Concentration Weighted Trajectory (CWT), Residence Time Weighted Concentration (RTWC), Potential Source Contribution Function (PSCF), and Simplified Quantitative Transport Bias Analysis (SQTBA). TraPSA provides users sufficient GIS editing functions for mapping air pollutant source apportionment. In addition, GIS data files (ESRI shape file and Geo TIFF file) can be imported or exported by TraPSA allowing further research or editing by GIS software.

Some models of the global oxidation of gaseous elemental mercury (GEM) by atomic bromine suggest that BrHgONO is the major Hg(II) species formed. The molecular structure of this compound is analogous to that of HONO, which is known to photolyze rapidly to HO + NO. Quantum chemical calculations suggest that BrHgONO will photolyze in a similar manner to produce BrHgO + NO. The BrHgO radical has never been directly detected in the laboratory, although calculations a decade ago indicated that it is thermally stable.

In the absence of a way to produce and monitor the abundance of BrHgO, we use computational chemistry to evaluate its mechanism of reaction in the atmosphere. We estimate the rate constant for the reaction BrHgO + CH4 --> BrHgOH + CH3 to be 3 10-14 cm3 molecules-1 s-1 at 298 K. This corresponds to a BrHgO lifetime of 0.7 seconds with respect to forming BrHgOH. We are also studying analogous reactions in which BrHgO abstracts hydrogen atoms from ethane and formaldehyde. The theoretical approach used to date (PBE0 functional with a valence triple-zeta basis set) overestimates the rate constant for the analogous HO + CH4 --> HOH + CH3 reaction; as a result, the lifetime reported here is probably smaller than the true value. We are refining our results.

We find that BrHgO can also react with NO2 or NO to form thermally stable BrHgONO2 or regenerate BrHgONO, respectively. These reactions may be significant fates of BrHgO in areas heavily impacted by emissions from motor vehicles and energy production.

This work contributes to identifying the molecular identity of gaseous oxidized mercury (GOM) species formed in the Br-initiated oxidation of GEM. This work will help improve models of atmospheric mercury oxidation, aid laboratory scientists in designing kinetic and mechanistic experiments, and contribute to identifying GOM species in field work.

Temporal variation of gaseous elemental mercury (GEM) was investigated over metropolitan Taipei in Taiwan during 2007 by using an in-situ Automated Gaseous Mercury Analyzer (AGMA). The mean GEM concentration was 4.3 ng m-3 with a range from 0.9 to 15.5 ng m-3. Distinct annual patterns were observed for the GEM with a winter maximum of 5.1±2.0 ng m-3 (n=3) and low in summer (3.7±1.1, n=3). The data showed the seasonal monsoons play a crucial role in the atmospheric long-range transport of Hg with its distribution and cycling in Taiwan. During the northeast monsoon in winter, air masses came from mainland China, bringing continental- and industrial-derived GEM to Taipei. The measured GEM/CO ratio of 0.0047 ng m-3 ppbv-1 is further similar to the results, which are observed at Mount Bachelor observatory (MBO) and Hedo Station, Okinawa (HSO) from Asian long-range transport. In contrast, the southwest monsoon prevailed in summer transports marine air from the South China Sea and west Pacific Ocean with lower GEM levels. Furthermore, a distinct diurnal variation of GEM concentration was observed, which level significantly exhibited greater in daytime than in nighttime during the warm season. Diel GEM variation was positively related to the ozone, PM10, PM2.5 and solar irradiance. The daily pattern with a maximum GEM concentration observed in the early afternoon and a minimum in the mid-night was likely due to local human activities, rising solar irradiance and ambient surface air temperature in Taipei area.

Mercury is a naturally occurring metal that is toxic to many organisms. The atmospheric emission of Hg from natural and anthropogenic sources combined with the long atmospheric lifetime of gas phase elemental Hg and subsequent atmospheric deposition cause the element to occur globally in environmental systems. Atmospheric wet deposition of Hg via rain and snow is an important process in the biogeochemical cycling of Hg.

In an effort to characterize and understand processes contributing to the wet deposition of Hg, event based wet deposition samples were collected at the Thompson Farm AIRMAP site in Durham, New Hampshire from June 2006 to September 2009. Samples were analyzed for total aqueous Hg, total Pb, dissolved organic carbon (DOC), total dissolved nitrogen (TDN), nitrate, ammonium, sulfate, sodium, potassium, and chloride. Statistically significant (p<0.05) positive correlations exist between concentrations of Hg and total Pb, DOC, TDN, nitrate, ammonium, sulfate, and potassium.

This multi-year dataset allows for seasonal comparisons between the analytes. Volume weighted mean (VWM) concentrations were highest during the summer for Hg, Pb, DOC, TDN, ammonium, and sulfate. The VWM concentrations of sodium, chloride, and potassium were highest during the fall. The lowest seasonal VWM concentrations of sodium and chloride were during the summer. The VWM concentrations for the majority of the other analytes were lowest during the winter. Seasonal patterns in total deposition, the product of concentration and total rainfall, largely follow the patterns in seasonal VWM concentrations.

The interpretation of these results may be used to distinguish atmospheric sources of Hg and inform efforts to model the atmospheric deposition of Hg.

GEM was measured in downtown Toronto, Canada from Oct. 2015 to Oct. 2016 at two rooftop sites (i.e. KHN and JOR) which are 120m apart and have heights 29m and 60m respectively. Monitoring was done using mercury vapour analyzers (model 2537A, Tekran Inc., Toronto, Canada) accompanied with portable weather stations (Onset HOBOData Loggers) equipped with sensors to measure meteorological parameters.

The average atmospheric concentration of GEM was found to be 1.76 ± 0.87 ng/m3 at KHN and 1.23 ± 0.44 ng/m3 at the JOR site. The average GEM values measured at the two sites are statistically different at the 95% confidence level. This suggests local sources are contributing to the higher values measured at KHN. In addition, the difference in the height of the sampling sites and in city topography may also have contributed to the different values of Hg observed.

There is evidence of Hg pollution sources to the experiment sites from the NW and SE directions. Comparison of the data collected from KHN in 2004 and 2016 show similar Hg distribution patterns from the same direction but the average concentration observed at the sampling site dropped from 4.5 ng/m3 in 2004 to 1.76 ng/m3 in 2016. GEM measurements from downtown Toronto in 2004 and 2016 were also compared to background values from the CAMNet for the same or similar years. A decrease in average GEM was observed from 2004 to 2016 at both the urban sampling sites as well as at the rural CAMNet sites. The observed decrease of mercury may be a result of the restrictions placed on the use and disposal of mercury and mercury-containing products.

The results from the study suggest that the concentration and distribution of GEM in an urban environment is influenced by local and regional point sources, city topography and environmental policy.

Event-based wet deposition of mercury was collected in a study from 2010 to 2012 at the Patricia McInnes (AMS 6) monitoring site 30 km from the nearest upgrading facilities in Fort McMurray, AB, Canada. For the entire study period (21 months), volume weighted mean, VWM, concentration was 11.2 ng L-1 while total Hg wet deposition was 2.3 g m-2. Hg enrichment factors ranged from 105419 in rainfall, 45-599 in mixed precipitation and 73-266 in snowfall samples. Concurrent enrichment of trace elements including S, As, and Zn was also observed in samples. Our results suggest near field deposition of local anthropogenic emissions from the industrial and energy sectors impacted the AMS 6 site. Maximum Hg enrichment was found when winds transported smoke and particulate matter from forest fires 100 km away, to the sampling site. This finding corroborates previous findings that biomass burning is a source of particulate Hg that is deposited on a local scale. The magnitude of Hg wet deposition at the AMS 6 site was at the lower end of the measurement range made in the United States and Canada and limited by the low precipitation depths that occurred at this semi-arid location. This suggests that Hg dry deposition may be significant in the AOSR and should be addressed in future studies.

Gaseous elemental mercury (GEM) has been measured at the WMO Global Atmosphere Watch (GAW) station, Cape Point, South Africa, since September 1995. Two techniques have been used: a low resolution manual technique until the end of 2004 and a high resolution auto-mated technique since March 2007. Besides meteorological parameters and solar radiation data, CO, CH4, CO2, O3, N2O and 222Rn concentration values are available for the interpretation of the GEM data.

A downward GEM trend was observed between 1995 and 2004 and an upward one since 2007. A statistical analysis of GEM, 222Rn, CO, CH4, and N2O trends as well as their inter-comparison will be presented. Furthermore, the implications of this comparison and possible underlying reasons of the observed trends will be discussed.

The Listvyanka station is located at a shore of Lake Baikal, Siberia, far away from the existing mercury monitoring sites in Asia. Long-term air mercury monitoring within Global Mercury Observation System (GMOS) project started in October 2011. The station is part of the EANET network whereby numerous parameters of the air pollution, wet and dry deposition, as well as condition of the terrestrial and aquatic environment are measured. Lumex RA-915AM mercury monitor is used for the continuous air mercury monitoring in compliance with the unified GMOS network standard operational procedures.

The 5-years monitoring shows obvious seasonal variation of the background gaseous elemental mercury (GEM) concentration in air, which increases in the winter season with monthly average of 1.59 (1.43-1.79) ng/m3 and decreases in the warm season with monthly average minima of 1.25 (1.11 -1.54) ng/m3 in June-September. The same character of the seasonal variation is observed for particulate bound mercury (PBM) having average concentration of 6.2 (2.5-20) pg/m3.

Local short-term mercury concentration rises are associated mainly with the wind carrying air from industrial areas of Irkutsk and Angarsk cities where several big coal-fired power plants are located. These power plants are the main sources of the elevated acid gases and mercury concentrations in air measured at the Listvyanka site. A positive correlation between mercury, SO2 and NOX concentrations is observed both in the short-term variations and in the monthly average concentrations, whereas correlation between the mercury and ozone concentrations is negative due to the O3 depletion in the power plants plume. The short-term variations clearly show the possibility of the long distance mercury transfer with the so-called low-level atmospheric jets. In contrast to industrial emission, during huge forest fires of summer 2015, a positive correlation between mercury and ozone was observed. At the same time, no PBM increase was registered during the forest fires. Data processing reveals a moderate, statistically significant, diurnal cycle of the mercury concentration both in the warm and cold seasons with a lower level at night and higher level at daytime.

This research was carried out under FP7 project Global Mercury Observation System, grant agreement No 265113.

Our recently introduced passive air sampler (PAS) for gaseous mercury (Hg) uses a radial diffusive barrier to control uptake kinetics and sulfur-impregnated activated carbon as a sorbent. Deploying multiple PASs simultaneously, retrieving them after variable lengths of time and analyzing them for the amount of Hg taken up yields an uptake curve. If the Hg concentration is simultaneously recorded with another sampling technique, e.g. a Tekran 2537 instrument, one can calculate a sampling rate SR (defined as the volume of air stripped of Hg per unit of time) from the slope of the uptake curve, i.e., by dividing the amount of Hg collected by the PAS by the actively measured concentration during the sampling period, and the deployment time. If a SR is known a priori, it is possible to compare the mean concentration measured by PAS with the actively measured concentration, i.e. establish the accuracy of the PAS. We previously presented a year-long uptake curve measured in Toronto, Canada, yielding a SR of 0.121 ± 0.005 m3 day−1. Here, we present year-long uptake curves measured at 22 sites across the globe with ongoing active Hg measurements. The sites, located in Canada, USA, Germany, China, Taiwan, and Australia, cover a range of climatic conditions (tropical to polar), geographic settings (city, mountain, coast) and Hg concentrations. The data are used in two ways. By calculating an air concentration using the SR determined in Toronto, we obtain an estimate of the accuracy of the PAS if it is assumed that the SR is the same everywhere. Alternatively, we can calculate site-specific SRs. We show that even with a generic SR, PAS-derived Hg concentrations are generally within 10 % of those obtained by a Tekran. Better accuracy can be achieved by using site-specific SRs. For sites for which no site specific SR exists a priori, it is possible to estimate a SR using local temperature and wind speed conditions.

Gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particulate bound mercury (PBM)) were continuously measured in Rochester, NY (NY43) and Bronx, NY (NY06) from Jan 2012 to Dec 2014. Continuous measurements of ozone (O3), sulfur dioxide(SO2), carbon monoxide(CO), nitrogen oxides(NOx), particulate matter (PM2.5), and meteorological data were also made at these sites. A principle components analysis (PCA) of 15 variables for the period identified several factors including wet deposition of GOM and PBM, and oxidization of GEM. Wood and coal combustion were found in two different factors through PCA analysis at the Rochester site and conditional bivariate probability function (CBPF) was used to determine the source of these two factors. A heating oil combustion factor was found for Bronx site indicating an increasing consumption of No.6 oil for central heating systems in that area. Mobile source was significant in 2012 but not in 2014 for both sites indicating the influence of the implementation of increasing tax of diesel starting in July, 2013.

A field campaign to measure total gaseous mercury (TGM) concentrations in ambient air was conducted at the Whiteface Mountain peak, NY from 1 June 2016 to 14 Oct 2016 using an automatic atmospheric mercury analyzer (Tekran 2537X) to investigate the concentrations at high altitude and relationships with other pollutants In addition concentrations were compared to a nearby low altitude site. The average TGM concentration from the high altitude site was 0.940.34 ng/m3, which is lower than the GEM concentration measured at nearby Huntington Forest ground site (NY20). A correlation analysis of TGM with other atmospheric pollutant concentrations was used to explore the characteristic of the measured TGM. In addition factors such as temperature, wind speed, wind direction and precipitation events were evaluated to determine if they are correlated with TGM concentrations. Conditional probability function (CPF) and potential source contribution function (PSCF) models were used to determine the source of TGM measured at the receptor site on the peak of Whiteface Mountain.

Tracing pollution sources in complex terrain, such as the western United States, is a difficult task. For this work, we investigated the hypothesis that lead isotope analyses would aid in identifying sources of mercury and ozone to the western United States. Reactive mercury and lead analysis of 24 h ambient air particulate matter was used to determine sources of pollution to three different sites in Nevada, USA. Measurements were made at three sites: a lower elevation, highway-impacted site (elev. 1370 m) from December 2013 to November 2015; a high elevation site (elev. 2515 m) adjacent to the lower elevation site from December 2013 to October 2014; and in Great Basin National Park (2061 m, eastern edge of Nevada near the Utah border) from March to October 2014. Ambient reactive mercury (gaseous oxidized mercury + particulate bound mercury) was collected using one inlet with cation exchange membranes while ambient lead samples were collected through a second inlet using Teflon membranes. A Tekran total mercury system (Model 2600) was used for analyses of CEM filters for total Hg. Lead isotope samples collected on the Teflon filters were analyzed with a multi-collector inductively coupled plasma mass spectrometer (IsoProbe). Lead isotope ratios have been used to identify Asian lead sources based on the 206/207 and the 208/207 lead isotope ratios. Analysis of preliminary results suggests that both higher elevation sites see a greater influence of Asian lead than the lower elevation site during the study periods. High Asian lead influence occurred mainly from March to June when long-range transport of pollutants occurs in this area. The two sites at higher elevation typically experience higher concentrations of reactive mercury during lower Asian lead influence, suggesting regional sources of reactive mercury. Reactive mercury at the low elevation, urban site varied less with the influence of Asian lead than the two higher elevation sites.

The objectives of this study were to: (1) characterize the hourly and seasonal variations of atmospheric total gaseous mercury (TGM) and co-pollutants concentrations in urban (Seoul) and background areas (Kanghwa island) of South Korea, (2) identify the relationships between TGM and co-pollutants concentrations, (3) characterize high TGM concentration events by distinguishing between long-range transport (LRT) and local impacts, (4) estimate TGM emission flux using ΔTGM/ΔCO, and (5) identify likely source locations of TGM for LRT events using potential source contribution function (PSCF) and likely source directions and locations of TGM for local sources using conditional probability function (CPF), conditional bivariate probability function (CBPF) and PSCF.

TGM concentrations were measured every 5 min from January 2008 to December 2009 on the roof of the Graduate School of Public Health building using a Tekran 2537A in the urban area and in Kanghwa island using a Tekran 2537B.

Hourly meteorological data and concentrations of SO2, NO2, O3, CO, PM10 and PM2.5 were obtained from the Korea Meteorological Administration (KMA) and the National Air Quality Monitoring Network (NAQMN), respectively. The TGM (3.7 ± 2.3 ng m-3), CO (661.7 ± 389.7 ppbv), NO2 (35.2 ± 17.5 ppbv), SO2 (7.4 ± 3.6 ppbv) and PM10 (52.5 ± 37.3 ppbv) concentrations in urban area were statistically significantly higher than the TGM (2.0 ± 0.9 ng m-3), CO (661.7 ± 389.7 ppbv), NO2 (35.2 ± 17.5 ppbv), SO2 (7.4 ± 3.6 ppbv) and PM10 (52.5 ± 37.3 ppbv) concentrations in background area (p<0.01). However, the O3 concentrations in the background area were significantly higher than those in urban area (p<0.01). The TGM concentrations in the background area were significantly positively correlated with CO, NO2, SO2 and PM10 (p < 0.01) but negative correlated with O3 (p < 0.01). Similarly, the TGM concentrations in urban area were significantly positively correlated with CO, NO2, SO2 (p < 0.01) and PM10 (p < 0.05) but negative correlated with O3 (p < 0.01).

A total of 150 high TGM concentration events in urban area were identified during the sampling period: 107 (71%) LRT events and 43 (29%) local events. A total of 91 high TGM concentration events in background area were identified during the sampling period: 61 (67%) LRT events and 30 (33%) local events.

Backward trajectory analysis starting from two sampling sites showed that air parcels arrived mostly from China for long-range transport events.

At a “clean air” trade winds site in tropical northeastern Puerto Rico, atmospheric total mercury (THg) deposition in 2006-2007 averaged 27.9 µg m-2 yr-1, higher than any site in the USA Mercury Deposition Network. These high rates of THg deposition are driven by high rainfall amount, and evidence also supports efficient capture of THg of upper tropospheric Hg by high rain-forming clouds. The elevated THg in deposition was reflected in high THg concentration and flux in streamwater, but assimilation into the local food web was quite low. There are few mammalian or freshwater fish predators on the island, but avian blood THg concentrations (n=31, from 8 species in various foraging guilds) ranged widely from 0.2 to 32 ng g-1, with a median of 4.3 ng g-1. Avian blood THg levels were an order of magnitude lower than comparable values in the northeastern U.S. These low levels were surprising given the high Hg inputs and watershed features that would seem to favor methylmercury (MeHg) production (Hg(II)-methylation) – high soil moisture, ample organic matter and sulfur, and year-round warm temperatures. However, organic soil (0-10 cm) along a hillslope to riparian transect averaged only 0.45 ng/g MeHg, with an average MeHg/THg ratio percent of only 0.34%. Stable isotope amendment incubations (n=6) to assess 200Hg(II)-methylation and Me201Hg demethylation potentials along the upland to wetland transect indicated that rate constants for demethylation were 6-60 fold greater than those for Hg(II)-methylation, and calculated potential demethylation rates were 3-9 fold greater than Hg(II)-methylation rates. The net change in the ambient Me200Hg pool in the 6-day anoxic Me201Hg incubations revealed slight positive net methylation (mean = 15.8 ± 4.6 pg g-1 d-1 dry wt.; n = 6). This rate is considered an upper limit, as the soil samples were initially oxic (field Eh range +342 to +575 mV), while the 6-day incubations were performed under anoxic conditions. Thus, a likely resolution of the paradox is that MeHg degradation keeps pace with MeHg production on the landscape. The interplay of these microbial processes shields the food web from adverse effects of high atmospheric Hg loading on the island.

To better understand the influence of anthropogenic source and monsoon transport of atmospheric mercury(Hg) in Yangtze River Delta, measurements of total gaseous mercury (TGM), and gaseous oxidized mercury (GOM), particulate bound mercury(PBM) were carried out at Chongming Island (CM) in the eastern China from March 2014 to December 2016. The mean concentration (±SD) for TGM, GOM and PBM were 2.24 ± 1.17ng m−3, 14.12 ± 13.49 and 16.88 ± 31.69 pg m−3, respectively. In the sampling period, TGM showed an annual change trends from 2014 to 2016 with relatively lower concentrations (1.54±0.49 ng m-3) during 2016 and higher concentrations (2.81±1.46 ng m-3) during 2014. The Potential Source Contribution Function analysis suggests that the in-land source have a significant influence for GEM and PBM in CM while oceanic source contribute more for GOM. Back-trajectory-based analysis consistently indicated that TGM showed a monsoonal distribution pattern with relatively higher concentrations during the east Asia summer monsoon (EASM, from May to September) than winter monsoon. This study suggests that the anthropogenic source and monsoonal transportation have a collaborative significant influence for GEM in YRD. The EASM have a strong impact on long-range transport of Hg between YRD and east China sea. Besides the climate change in the study area, this study suggests that the decrease in anthropogenic emission also contributes to the down trend of mercury concentration at CM. The trajectory cluster analysis indicates that anthropogenic sources have more influences on the concentration variations of GEM compared to the summer monsoon. Various pollutants were also observed at CM and principal component analysis suggested that combustion emissions were the dominant anthropogenic mercury sources for the study area.

Atmospheric concentrations of gaseous elemental mercury (GEM) were measured from a shipboard system during transects around continental North America and in the Canadian Arctic in 2009 and 2015.

In the Atlantic, strong diurnal variations were observed in GEM concentrations between 40°N to 23°N, which is consistent with previous work indicating high variability in the Atlantic basin. Concentrations in the Pacific between 9.5°N and 36°N were less variable, although minimum values were higher in the Pacific relative to the Atlantic. Concentrations of GEM were low in the equatorial Pacific and Atlantic, with the exception of the Panama Canal, where transient spikes in GEM concentrations were attributed to exhaust from Canal ship traffic.

In Arctic waters, GEM concentrations were elevated in Baffin Bay relative to the Beaufort Sea where concentrations were low throughout. Summertime atmospheric mercury depletion events (AMDEs) were potentially observed in the waters surrounding Banks Island, with consistently low observed GEM concentrations (< 1.0 nm m-3) observed over several periods.

The combined data set prevent us from distinguish between spatial and temporal differences to any great extent, although overlapping regions of the Beaufort Sea were measured in both 2009 and 2015. Southwest of Banks Island GEM concentrations were higher in late-June 2009 relative to late-August 2015 in contrast to previous studies, which generally have observed maximum GEM in summer.

Overall, our data provide insight into understudied regions of GEM distributions, especially Pacific and Canadian Arctic waters. In general, observed GEM concentrations are lower than both previously published shipboard measurements and averages collected at land-based monitoring sites. The temporal and spatial variability may inform models of GEM distributions in the marine boundary layer.

Within the EU-funded project, Global Mercury Observation System (GMOS) the airborne mercury species/fractions: Gaseous Elemental Mercury (GEM), Particulate Bound Mercury (PBM) and Gaseous Oxidized Mercury (GOM) were monitored using the Tekran speciation system at the R background measurement site on the west coast of Sweden. An evaluation of mercury concentrations measured during May 2012 to May 2015 is presented. The mercury concentrations measured at the R site were found to be low in comparison to other, comparable European measurement sites. The R site receives background air about 60% of the time. However, elevated mercury concentrations arriving with air masses from source areas from the south-east are noticeable. GEM and PBM concentrations show a clear annual variation with the highest values occurring during winter, whereas the highest concentrations of GOM were obtained in spring and summer. GOM concentrations observed at the R site often show a diurnal pattern with peak concentrations at midday. This phenomenon has also been observed at other sites and has often been interpreted as oxidation of GEM driven by local atmospheric photochemistry. An analysis of the origin of air masses arriving to the R site made it possible to distinguish between air masses associated to regional mercury sources from air originating from clean background air. This analysis showed that the highest GOM concentrations were observed in conjunction to import of air masses from the north which, are not associated with major anthropogenic mercury sources. The highest GOM concentrations were obtained from air masses originating from north of Scandinavia.

Here it is proposed that a significant part of the GOM measured during summertime at the R site is due to elevated concentrations of GOM accumulated in the free troposphere from oxidation of GEM. Evidence of this sort has also have been suggested from earlier observations, e.g. Wngberg et al., 2007; Weiss-Penzias et al., 2009. Hence, like with ozone, which also is secondary air pollutant the diurnal variation in concentration can be understood in terms of local meteorology, i.e. by nocturnal inversion at night, a phenomenon that occurs during clear sky conditions. During night GOM and ozone is depleted due to deposition on vegetation and on wet aerosols. The inversion prevents GOM and ozone from above to mix with the air below until the next morning when the inversion is broken by the sun and air from above are transferred to the ground through vertical mixing.

Atmospheric mercury (Hg) deposition of three atmospheric mercury species gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particulate bound mercury (PBM) is the largest atmospheric input to most terrestrial and aquatic ecosystems. Regional air quality models are needed to quantify Hg budgets in the atmosphere but limited by large uncertainties. The Community Multiscale Air Quality model with mercury (CMAQ-Hg) has been extensively used in research and policy-related studies. However, the default CMAQ-Hg (version 5.0.2) does not include GEM oxidation by active Br species, which have been suggested as important GEM oxidants. In this study, an algorithm depicting state-of-the-art Hg and halogen chemistry mechanisms was implemented in CMAQ-Hg, and Br species were constrained with an observed vertical BrO profile. Using this new mechanisms with initial and boundary concentrations (ICs and BCs) from global model output, we conducted simulations for the months of March to November 2010 over a domain covering the northeastern United States at a horizontal resolution of 12 km. Simulated GEM mixing ratios appeared to be dominated by the BCs, and hence reflected the significant seasonal variation that was captured in the global model output as opposed to the lack of seasonal cycles using the CMAQs default constant BCs. Our improved model simulations agreed well with GEM observations with 6.5% fractional bias (FB) in the fall, but underestimated GEM in the spring (FB = 13%) and summer (FB = 20%). GOM and PBM were better simulated using the improved model with FB = 2% and 19%, respectively, compared with FB = 72% and 69% using the default model. The new chemical mechanism alone resulted in a 12% decrease in GOM and a 34% decrease in PBM mixing ratios compared to the default one, and both reached maximum decreases in the summer. With simulated GOM and PBM close to or slightly higher than observations, Hg wet deposition was underestimated (FB = -60%) by the improved model at all observational sites in the domain. A sensitivity test of including GEM oxidation by OH in the new chemical mechanism resulted in the best simulations of monthly total Hg wet deposition with 0.3% bias.

There have been growing concerns on mercury pollution on local, regional and global scales. Better understanding of the source receptor relationships for mercury deposition in the context of global change is greatly needed. We use the global GEOS-Chem coupled atmosphere-land-ocean mercury model, driven by GISS ModelE2 meteorology to examine the source attribution for mercury deposition over various regions (such as North America, East Asia and the Great Lakes region) for the present-day as well as the impacts from future changes in anthropogenic emissions, biomass burning emissions, climate, land use and land cover. Through a suite of sensitivity simulations, we quantify the contributions from various sources (e.g., anthropogenic vs natural sources) and various regions (e.g., local emissions vs long-range transport) to the total mercury deposition over specific receptor regions. The spatial-temporal patterns of the perturbations to these source-receptor relationships associated with various factors in the context of global change are examined in detail.

Although East Asia is the major atmospheric mercury (Hg) emission source region, studies about atmospheric Hg deposition in this region are still limited. Here we reported the wet and dry deposition of atmospheric Hg to the Lulin Atmospheric Background Station (LABS), a tropical mountain site in central Taiwan (23.47ºN, 120.87ºE, 2862 m a.s.l.), from 2009 to 2016. Weekly rainwater samples were collected for total Hg analysis. Wet deposition flux was calculated by multiplying rainwater Hg concentration and rainfall depth. Concentrations of speciated atmospheric Hg, including gaseous elemental Hg (GEM), gaseous oxidized Hg (GOM) and particulate Hg (PHg), were measured by the Tekran 2537A/1130/1135 speciation unit. Dry deposition of speciated Hg was estimated by multiplying concentration and deposition velocity. In 2009-2012, the annual rainfall ranged from 3172 to 4991 mm and the volume-weighted mean concentration of Hg in rainwater ranged from 8.71 to 13.53 ng L-1. Annual wet deposition fluxes ranged between 33.89 and 42.84 µg m-2. Hg wet deposition fluxes were higher in summer because of higher rainfall. Weekly wet deposition fluxes and rainfall were highly correlated (R2 = 0.81, p < 0.01). Annual dry deposition fluxes ranged from 67.41 to 75.91 µg m-2, nearly 2 times the wet deposition fluxes. Nighttime GOM dry deposition flux (6.92 µg m-2 yr-1) was higher than that of daytime (4.51 µg m-2 yr-1) due to higher GOM concentration and wind speed at night. Because of the high percentage of forest canopies at the monitoring site, average annual GEM dry deposition (59.71 µg m-2) was significantly higher than GOM (11.43 µg m-2) and PBM (0.13 µg m-2). It should be noted that the GEM dry deposition fluxes could be overestimated because GEM re-emission from the surface was not considered. We are still working on the estimation of dry Hg deposition using bi-directional surface resistance model. Moreover, we will expand the estimation of atmospheric Hg deposition to cover the whole 2009-2016 time period.