G4B Atmospheric Mercury: Measurement and Monitoring

The measurement of atmospheric mercury at Kosetice Observatory started according to the EMEP methodological guidelines in 2006. The Observatory was established as a station specializing in air quality monitoring and research at the background scale in 1988. The observatory represents the Czech Republic in activities under CLRTAP and WMO as well as in projects (EUSAAR, ACTRIS). Mercury concentrations in the air and aerosols have been detected in weekly frequency. The sampling of gaseous mercury in the air has been carried out using protective quartz filter followed by two gold traps in series; mercury in atmospheric aerosol is collected on quartz filter. Analytical process is implemented in accredited ALS laboratory in Prague. The analysis of gaseous mercury samples is made using double amalgamation CVAFS. The gold trap is mounted in series with a second analytical trap in a gas stream (Hg-free argon) leading to the CVAFS detector. The results show that the concentration levels of gaseous mercury ranges in between 0,03 and 3,71 ng.m^-3. The mean annual concentration extend to 1,21 ng.m^-3. This value is in very good agreement with general scientific consensus about the current global background concentration in the Northern Hemisphere, which varies from 1,5 to 1,7 ng.m^-3. The highest concentrations are registered in the days with very good conditions for long-range transport, when the air masses originate between the southwest and northwest sectors. The mean annual concentration of mercury in aerosol was 12,4 pg.m^-3 . Monitoring should be expanded in next year within the Czech Globe project. Two automatic analyzers for gaseous mercury measurement should be bought. The first will be used for continuation of long-term measurement at Košetice Observatory; the other will be located at the mountain station Churáňov (1118 m above sea level). The results will improve our knowledge of mercury long-range transport.

Knife-edge surrogate surfaces containing various mercury collection media were used in Cleveland (CLE), Rochester (ROC), and Huntington Forest (HF) to directly measure Hg dry deposition. Quartz fiber filters (QFF) and ion-exchange membranes (IEM) were used to capture particle-bound mercury (PBM) and gaseous oxidized mercury (GOM) dry deposition, respectively. Hg dry deposition to aqueous systems was also investigated using deionized water (DI) and acidified bromine chloride (BrCl). Gold-coated QFFs were treated as a perfect sink of GEM due to the fast and stable amalgamation reaction between GEM and Hg. The Hg mass deposited to the QFF was not significantly different from its blank value indicating QFF is not a suitable media for PBM dry deposition measurements under the conditions tested. In CLE, Hg dry deposition fluxes were measured at various intervals (from 8 hours to 10 days). Hg fluxes to up and downward facing gold-coated QFFs were not statistically different (deposition velocity (V_d) of 1.5 ± 1.0 cm s^-1). The GOM and fine PBM flux to downward facing IEM filters was 1.7 ± 0.6ng m^-2 h^-1, and to upward facing IEM was 2.0 ± 0.5ng m^-2 h^-1. This difference may have been due to large particle deposition to the upward facing filter. The results suggest that GEM controls the deposition to the gold-coated QFFs, and GOM and PBM may be equally important for deposition to the IEMs. The GEM V_d to BrCl was 0.06 ± 0.04 cm s^-1. The relative percent difference (RPD) between measured and modeled (using multiple-resistance model) Hg dry deposition fluxes range from 9 to 45 %.

The objectives of this study were to determine the seasonal variations in atmospheric wet deposition flux of total mercury (TM), and to identify possible source locations using receptor model. Atmospheric wet deposition for TM was measured with a modified MIC-B sampler on th roof of Graduate School of Public Health building in Seoul, Korea from January 2006 to December 2010. The samples were analyzed using Tekran 2600 system equipped with Cold Vapor Atomic Fluorescence Spectrometry (CVAFS).

The volume weighted mean (VWM) TM concentrations in 2006, 2007, 2008, 2009 and 2010 were 10.20 ± 16.90 ng L^-1, 16.65 ± 16.68 ng L^-1, 14.39 ± 11.86 ng L^-1, 12.23 ± 15.28 ng L^-1, 11.50 ± 10.25 ng L^-1, respectively. The TM wet deposition flux in 2006, 2007, 2008, 2009 and 2010 were 16.85 µg m^-2, 20.24 µg m^-2, 18.57 µg m^-2, 13.66 µg m^-2 and 14.44 µg m^-2, respectively. The large wet deposition fluxex observed in summers were possibly due to the intense rainfall during those periods. Overall there was a significant positive correlation between wet deposition flux and rainfall depth (r² = 0.26) (p < 0.01) and a significant negative correlation between rainfall depth and TM concentration in precipitation (r² = 0.10) (p < 0.01) possibly due to dilution. In addition, a weak positive correlation between TM concentration and wet deposition flux was shown (r² = 0.10) (p < 0.05).

Sources were identified using positive matrix factorization (PMF). A total of 206 samples were obtained during the sampling period, and 28 chemical species were measured. Lagrangian Particle Dispersition Modeling (LPDM) results showed that possible source locations contributing to the wet deposition in Seoul were the major industrial areas in China.

Acknowledgements:

This work was supported by the Ministry of Environment (Eco-technopia 21 project 2009-12001-0050-1) and Ministry of Education, Science and Technology, Republic of Korea (R01-2008-000-11165-0).

Mercury pollution is caused from small-scale gold mining along the Cikaniki River. Atmosphere is one of the most important mediums for mercury dispersion. In this study, to reveal the dispersion of mercury through air from the mining, the atmospheric mercury levels are estimated by using of native epiphytic fern Asplenium nidus L. as a biomonitor.

Samples were collected on October–November, 2008 and 2009 in the Cikaniki Basin. Site A is the headwaters. Sites B and C are located in about 12 km lower from Site A. Site D is within 1 km and about 100 m high from Site C. Sites E, F and G are located in about 18, 19 and 21 km lower from Site A, and Sites H and I are more than 2 km off from the Cikaniki River. The mining has been operated in Sites B, C, E, F and G, especially Sites B and C have been the hot spots.

All A. nidus examined were found on tree trunks, attached to 1–3 m above the ground. More than 3 fronds of each A. nidus were collected. The fronds were washed and dried at 70^oC for 1–3 days. Total mercury concentration in the frond was determined by cold vapor atomic absorption spectrometry (CVAAS) after acid–digestion. Atmospheric mercury was collected with porous gold collectors and the concentration was determined by double–amalgam CVAAS.

The highest mercury concentration of A. nidus was observed in the mining hot spot (Site C) and the lowest in the no mining sites (Sites A and I). Among the mining sites, the mercury concentrations of A. nidus were higher in Sites B and C (3.8x10³–5.4x10³ ng/g) than the other mining sites (4.0x10²–9.4x10² ng/g). Meanwhile, among the no mining sites, the concentrations in Site D (5.4x10² and 8.8x10² ng/g) were significantly higher than in Site I (7.0x10–3.1x10² ng/g). It was suggested that mercury released from the mining activity in Sites B and C was transported through air and caused the increase in mercury level in Site D.

The mercury distribution of A. nidus in this area indicated a similar tendency with that of air. A significant correlation was observed between mercury concentration in air and A. nidus (R=0.895, P<0.001, n=14). The mercury level of air can be estimated with the mercury concentration of A. nidus using a regression equation.

The three objectives of this study were to measure ambient total gaseous mercury (TGM) and co-pollutant concentrations in urban area (Seoul) and background area (Kanghwa-island); characterize TGM high concentration events by distinguishing between long-range transport events and local events; and estimate TGM emission flux using measured TGM/CO ratio. Ambient TGM was measured concurrently in urban area (the roof of Graduate School of Public Health building in Seoul, Korea) and background area (Kanghwa-island in Yellow Sea) from January 2008 to December 2009. TGM was measured with a mercury vapor analyzer (Tekran 2537A). Hourly concentration of SO₂, NO₂, CO, O₃, PM₁₀ were obtained from the National Institute of Environment Research (NIER). The average concentrations of TGM and CO over the sampling period in urban and background areas were 3.68 ± 2.31 ng m^-3, 662.73 ± 389.70 ppbv and 2.00 ± 0.87 ng m^-3, 446.00 ± 247.85 ppbv, respectively. TGM concentrations had positive correlations with SO₂, NO₂, CO, and PM₁₀ (p < 0.01) but had a negative correlation with O₃. TGM and CO concentrations were highest during the winter and lowest during the summer. In urban area, totally 150 high TGM concentration events were identified: 107 events were classified as long-range transport events and 43 events were classified as local events. In background area, totally 91 high TGM concentration events were identified: 61 events were classified as long-range transport events and 30 events were classified as local events. Five-day backward trajectory analysis starting at two sampling site for long-range transport events showed that air parcels arrived mostly from China among potential source regions.

Acknowledgements:
This work was supported by Ministry of Environment (Eco-technopia 21 project 2009-12001-0050-1) and Ministry of Education, Science and Technology, Republic of Korea (R01-2008-000-11165-0).

Because of its very high-altitude and limited industrial productions, environment of Shangri-la area is pristine. However, once released to the atmosphere, mercury(mainly in the form of Hg0)can be transported over a long distance and deposited in remote regions. From November 2009 to November 2010, measurements of total gaseous mercury(TGM), particulate mercury( PHg, Hg bounded to particulate matter with an aerodynamic diameter <2.5 µm ), and reactive gaseous mercury(RGM) were carried out at the Zhuzhang Atmosphere Watch Regional Station(3558 m above sea level)of Shangri-la area, in Tibetan Plateau, Southwestern China. The overall mean concentrations for TGM, PHg and RGM were 2.59±1.33 ng m3, 43.5±41.6 pg m-3 and 8.2±9.4 pg m-3 respectively, which are higher than the values obtained in remote areas of Northern America. TGM, PHg, and RGM in Zhuzhang showed a noticeable seasonal distribution pattern with elevated concentrations in spring and low concentrations in winter. The study suggests that long-range transport of Hg from anthropogenic sources affected the concentrations of speciated atmospheric mercury by both from the south via the Indian monsoon during summer and from the west via winter westerlies. In addition, relative humidity was a dominant factor to TPM and RGM levels.

Pengjiayu is a small remote island in the west Pacific Ocean, with an area of 1.14 km² and about 56 km distance to the north of Taiwan. Mercury wet deposition monitoring started at Penjiayu Weather Station (25°37’46”N, 122°4’16.5”E, 101.7 m a.s.l.) since late 2008 to study total Hg concentration in precipitation and the associated wet depositional fluxes. The data will later be used to evaluate the contribution of the East Asian mercury emissions to the measured Hg wet deposition. Weekly rainwater samples are collected using an automated wet-only precipitation collection system. A total of 35 rainwater samples were collected in 2009. Sample total Hg concentrations ranged between 2.25 and 22.33 ng L^-1, with a volume-weighted mean (VWM) concentration of 8.85 ng L^-1. This VWM concentration was comparable to the 2009 values reported by the Mercury Deposition Network (MDN) for the Southeastern and upper Midwestern states in USA. Annual Hg wet depositional flux was 12.62 µg m^-2, also comparable to the 2009 values of the Southeastern states in USA. Seasonal VWM concentrations were 7.23, 11.58, 7.82, and 9.83 ng L^-1 for spring, summer, fall, and winter, respectively. High rainwater Hg concentrations in summer were also observed at MDN sites. Since there is no anthropogenic Hg emission source on Pengjiayu, the observed high summertime rainwater Hg concentration hints the importance of Hg⁰ oxidation and/or scavenging of upper-altitude Hg(II) by deep convection. Direct anthropogenic Hg(II) emissions from the East Asian continent may not contribute significantly to the measured rainwater Hg concentrations; however, anthropogenic Hg⁰ emissions may be transported to the upper troposphere or marine boundary layer where it can be oxidized to produce Hg(II), which will then be effectively scavenged by cloud water and rainwater.

Between 2005 and 2008, total gaseous mercury was collected at ten sites which comprise part of the UK rural heavy metals monitoring network run by the UK’s Centre for Ecology & Hydrology on behalf of the UK Department for the Environment, Food and Rural Affairs.

Total gaseous mercury (TGM, comprising elemental, reactive gaseous and particulate mercury) was captured using the gold amalgam technique with a custom built sampler. The samples were then thermally desorbed for analysis using a Tekran 2537A mercury vapour analyser. The data showed no upward or downward trend for the period, with 4 year average concentrations for each site between 1.3 and 1.9 ng m^-3, which agree with observations of the northern hemispherical background concentrations at other monitoring sites of between 1.5 and 1.7 ng m^-3.

Data from nine of the network sites are used to show seasonality within the data and, using kriging, concentrations of TGM are interpolated across the UK, revealing a south-east to north-west declining concentration gradient. The concentration of TGM recorded in the south-east of the UK more closely matches that of background TGM observed in continental Europe, which could indicate that the TGM concentrations from the north of the UK are a better reflection of the true north Atlantic atmospheric mercury background level.

Using wind sector analysis and air-mass back trajectories, we show how speciated mercury measurements collected at one of the network sites using a Tekran 2537A instrument are influenced by regional sources (< 50 km) as well as air masses moving over the UK from continental Europe on easterly winds.

Northern South China Sea is adjacent to major atmospheric mercury (Hg) emission source regions, including the East Asian continent (anthropogenic Hg emissions) and Indochina Peninsula (biomass burning Hg emissions). Regional monsoon activity could thus transport atmospheric Hg from these regions to northern South China Sea. Additionally, high-altitude westerlies could transport the Indochina biomass burning plumes eastward to Taiwan. However, studies concerning regional atmospheric Hg distribution and cycling are very limited. Accordingly, atmospheric Hg was measured in March and April during the 2010 Dongsha Experiment to study its spatial and temporal distribution. Atmospheric Hg was manually sampled at Hengchun (Taiwan), Dongsha Island (Taiwan), Da Nang (Vietnam), Ching Mai (Thailand) and aboard a research vessel, whereas automated speciated Hg measurement was performed at Mt. Lulin (Taiwan). Atmospheric Hg concentrations ranged between 1.69 and 6.83 ng m^-3, higher than the Northern Hemisphere background value of 1.5-1.7 ng m^-3, indicating sources other than background air. Significantly elevated atmospheric Hg concentrations were observed at Da Nang, likely indicating local Hg emission sources. Compared with the summer values of 2008-2009, elevated Hg levels were observed at Dongsha Island in the spring of 2010. Summer air masses were mainly from the southern South China Sea, representing relatively clean marine air. On the other hand, air masses were mainly from the north, passing eastern China or Taiwan prior to reaching Dongsha Island. Results of this research thus demonstrated the transport of atmospheric Hg from nearby emission source regions to northern South China Sea due to regional monsoon activity. Moreover, Hg concentrations were always higher at Chiang Mai than at Mt. Lulin, which seems to support the argument that Mt. Lulin is under the influence of Southeast Asian biomass burning Hg emissions in spring due to the transport of high altitude westerlies.

Particulate mercury, which is bound with particle in atmosphere, has a negative impact on human health and the environment, also plays an important role in the biogeochemical process of mercury. The aim of this research is to learn the pollution characteristics of particulate mercury in atmosphere in a cluster of cities located on western coastal line of Taiwan strait, China, along which is the area undergoing the rapid urbanization and industrialization. In this paper, the PM_2.5, PM₁₀ and TSP were collected in 14 urban sampling sites and 1 background site in Winter of 2010, RA-915⁺ mercury analyzer was employed to determinate mercury concentrations in different size particle matters based on zeeman atomic absorption spectrometry. Preliminary results showed that the contents of particulate mercury were generally lower than that in the other cities of China, but higher than those in Europe and North-America areas, the experimental data also showed that the particulate mercury were mainly distributed in fine particles (PM_2.5), which covered more than 60.14%, and it could be concluded that the rate of particulate mercury enrichment in fine particle was much higher than that of coarse particle. The atmospheric particulate mercury concentrations in different cities followed by a order of background < islang city < tourism city < commerce city < Peri-urban < industrial city, and the concentration level in northern cities were higher than that of southern ones. It could be concluded that the distribution characteristics of particulate mercury was dominated by the function and industry layout of cities, and also affected by the meteorological conditions, such as temperature inversion and low air pressure in winter, which were favorable for the accumulation of particulate matters and resulted in higher pollution.

The chemical cycling and spatiotemporal distribution of mercury in the troposphere is poorly understood. We measured gaseous elemental mercury (GEM) and particulate mercury(p-Hg) along with SO2, ozone, aerosols and meteorological variables at the summit of Mt. Fuji (3776m a.s.l.) from 23 August to 30 August. The mean mercury concentrations were 23ng/m3 (GEM) and 4.7ng/m3 (p-Hg). We observed this event of strongly enhanced atmospheric GEM levels with maximum concentration up to 25 ng/m3. High GEM and p-Hg levels were related to pollution events, particularly SO2 transported from Asian Continent. As result of back trajectory analysis will show this phenomena.

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. Therefore the quantification of Hg atmospheric deposition is critically needed in order to evaluate the Hg transformation and transport mechanisms, and further to reduce MeHg levels in environment. Among atmospheric inorganic Hg species gaseous divalent form (Hg(II)) and particulate form (Hg(p)) are considered to be very important with respect to deposition processes including dry and wet deposition. However a reliable sampler for measuring Hg dry deposition flux has not been developed yet.

Although particle size is one of the main factors influencing dry and wet deposition for aerosol, size distribution of Hg(p) has been rarely studied. In this research size-segregated Hg(p) were extensively measured using two different multistage impactors including MOUDI and CASCADE in urban (Seoul) and rural (Chuncheon) areas of Korea. Between November 2009 and August 2010 the average concentrations of Hg(p) in 0.18 ~ 18 µm were 6.77±6.52 pg/m³ and 4.57±2.72 pg/m³ at urban and rural sites, respectively. Distribution pattern of Hg(p) was mono-modal, peaking in the range of 0.32 ~ 0.56 µm at both sites. The contribution of fine mode (<1.8 µm) exceeded 80% of total Hg(p) mass at both sites. However the contribution of coarse Hg(p) must be significant to Hg deposition flux even though the coarse portion (3.2~18 µm) occupied only 12% of total Hg(p) concentration. Based on the concentrations of size fractionated Hg(p), dry deposition fluxes of mercury were estimated using Sehmel-Hogson model, and the average fluxes were 0.281 in Seoul in winter, 0.215 in Chuncheon in winter, 0.124 in Seoul in summer, and 0.113 ng/m²/day in Chuncheon in summer. The poster will present the estimated dry deposition flux of Hg(p) based on measured size-distribution of Hg(p) as well as size-segregated concentrations of Hg(p) in urban and rural area.

The purpose of this study was to characterize atmospheric mercury deposition at forest areas by measuring dry deposition, wet deposition, throughfall, litterfall, and soil evaporation of mercury. Samples were collected at forest areas located in Yangsuri, Korea from September 2008 through February 2010. Average RGM dry deposition flux was highest in spring (1.16 ± 0.34 ng m^-2 h^-1), while Hg_p dry deposition flux was highest in summer (1.11 ± 1.05 ng m^-2 h^-1). Volume weighted mean (VWM) total mercury (TM) concentration of throughfall was about two times higher than the one of wet deposition. Wet and throughfall fluxes were higher in summer possibly due to intense rainfall. Average wet and throughfall fluxes during the sampling period were 4.33 ug m^-2 yr^-1 and 6.41 ug m^-2 yr^-1, respectively. There was a negative correlation between VWM TM concentration and rainfall depth and positive correlation between TM wet deposition flux and rainfall depth. VWM TM concentrations (24.01 ± 13.49 ng L^-1) in snow were higher than those (5.87 ± 2.22 ng L^-1) in rain.

THg concentration in soil was highest in A horizon of fir habitat (77.82 ±3.04 ug/kg) and in B horizon of pine habitat (56.37 ± 2.03 ug/kg). In this study annual litterfall flux was 9.95 ug m^-2 yr^-1. Litterfall flux was higher from October to December possibly due to the increase of leaves production. Hg emission flux in soil was highest in June (1.48 ± 2.77 ng m^-2 h^-1), while it was lowest in October (-0.15 ± 1.69 ng m^-2 h^-1). Annual Hg emission flux in soil was 4.81 ug m^-2 yr^-1. In this study, the ratio of wet deposition, throughfall, and litterfall was 1 : 1.5 : 2.3.

Acknowledgements:
This work was supported by grant No. R01-2008-000-11165-0 from the Basic Research Program of the Korea Science & Engineering Foundation.

Mercury contamination is generally considered to be of highest concern in aquatic habitats, but recent evidence has indicated that terrestrial food webs could also be experiencing negative consequences of mercury contamination. Birds in forests from high elevation areas have been shown to have elevated blood mercury concentrations even though there is no known mechanism of mercury methylation within the terrestrial environment. In order to determine how terrestrial food webs are affected by mercury, it is important to know where mercury originates and how it is subsequently cycled through a forest. In this study, we established two transects along the eastern and western sides of Whiteface Mountain, New York, USA. The 24 sample sites ranged from 450-1450 m above sea level and covered three distinct forest types: deciduous/hardwood forest, spruce/fir conifer forest, and stunted alpine/fir forest. Sites were sampled periodically during the 2009 and 2010 growing seasons. Wet deposition of mercury was measured using throughfall collectors, and dry deposition was estimated by collecting live canopy foliage samples and analyzing them for mercury. In addition, a cloud sampler was deployed at the summit to estimate mercury inputs through cloud vapor. Mercury inputs to the forest floor from litter were estimated by collecting litter in litter traps, and soil samples were collected from organic soil layers to measure mercury accumulation within the forest floor. Preliminary analysis indicates that throughfall inputs and organic soil mercury accumulation are both higher in coniferous forest areas than in deciduous forest areas, and cloud deposition could potentially be a significant source of mercury to high-elevation ecosystems.

An intensive field campaign for the measurement of elemental gaseous mercury (Hg(0)) and Particulate mercury Hg(p) concentrations in ambient air was conducted in summit of Mt. Fuji from 11 August to 17 August in 2008 using an developed measurement technology, which was the first time Hg(0) and Hg(p) were monitored at a remote area in Mt. Fuji. The overall average Hg(0) covering the sampling periods was 2.61±1.24ng/m³, which is only a little elevated comparing to global background of approximately 1.5-2.0ng/m³.

Elemental gaseous mercury concentrations range from 1.45ng/m³ to 5.42ng/m³ in ambient air. Although there is not significant difference in concentration between daytime and night time, distinct daily variability of Hg(0) observed during survey periods. The phenomenon is caused by the direction of airmass. The back trajectory analysis were shown in Fig. 2. From this result, when airmass come from East Asian continent, elemental gaseous mercury concentrations were larger when that come from the Pacific Ocean.

Acknowledgment: This research was partially supported by Mitui&Co.,Ltd. Environment Fund, the Environment Research and Technology Development Fund [B-1008] of the Ministry of the Environment, Japan, the Watanabe Memorial Foundation for the Advancement of Technology, and the financial support of Japan Post Service Co.,Ltd. In 2009. This work was performed during the period in which the NPO “Valid Utilization of Mt. Fuji Weather Station”maintained the facilities.

To understand wet deposition processes and the loading of Minamata Bay with atmospheric mercury, the total mercury concentration (dissolved mercury + particulate mercury) in the wet depositions which were collected by a weekly sampling at the site of Minamata bay area were observed during the period from September, 2008 to August, 2010. The concentrations of reactive mercury and methyl mercury in the dissolved phase were also measured. In the latter half of the period, we tried to collect the samples in each rain event as much as possible. In addition, other chemical components such as nine major ions (Cl^-, NO₃^-, SO₄^2-, Ca²⁺, H⁺, K⁺, Mg²⁺,Na⁺, NH₄⁺), acetate ions and dissolved organic carbons (DOCs) were measured. The volume weighted average concentration of total mercury and dissolved methyl mercury in all sampling periods were 5.9 ng L^-1 and 0.065 ng L^-1, respectively. Almost 90% of total mercury was in the dissolved phase. Then, dissolved reactive mercury was the dominant mercury species in the wet depositions. The concentrations of total mercury and dissolved reactive mercury were almost constant and had no seasonal variations. Therefore, their wet deposition fluxes were higher in the rainy season and had a strong positive correlation with precipitation volume. These findings were probably because Hg wet deposition is dominated by the precipitation scavenging of atmospheric gaseous Hg (reactive gaseous mercury, RGM). In our preliminary measurement of RGM in the atmosphere at the same site, RGM concentrations were significantly and positively correlated with oxidants (primarily ozone). On the other hand, the concentrations and wet deposition fluxes of methyl mercury were higher in winter and spring than in summer. In event samples (N=33), the wet deposition fluxes of methyl mercury were significantly correlated with those of nss-K⁺, acetate ion, DOC, NO₃^-, nss-SO₄^2- (r= 0.45 – 0.61, P<0.01) and NH₄⁺ (r=0.40, P<0.05). Nss- means the non-sea salt fraction of each chemical component calculated on the basis of Na⁺ concentrations in rainwater and their concentration ratios in seawater. Total mercury and dissolved reactive mercury had no correlations with these components. It is known that natural biogenic sources or photochemical reactions are possible main sources of these components except for nss-SO₄^2-. Thus, wet deposition fluxes and concentrations of methyl mercury can also be changed by the influences of biogenic sources or chemical and biological reactions related to the light in the air.

Atmospheric mercury deposition measurements are few in tropical latitudes. Here we report two years (April 2005 to March 2007) of wet Hg deposition at a tropical wet forest in the Luquillo Mountains, northeastern Puerto Rico, USA. Despite receiving unpolluted air off the Atlantic Ocean from northeasterly trade winds, the site averaged 27.9 µg m^-2 yr^-1 wet Hg deposition, or about 30% more than Florida and the Gulf Coast, the highest deposition areas within the USA. These high Hg deposition rates are driven in part by high rainfall, which averaged 2855 mm yr^-1. The volume-weighted mean Hg concentration was 9.8 ng L^-1, and was highest during the summer and lowest during the winter dry season. Rainout of Hg (decreasing concentration with increasing rainfall depth) was minimal. Low reactive gaseous mercury (RGM) concentrations (<10 pg m^-3) at ground level did not support high Hg in rain; rather, we contend that high convective cloud tops scavenge RGM from above the mixing layer. In the tropics, elevated photooxidation rates of Hg⁰ from the global pool likely maintain upper tropospheric RGM. The high wet Hg deposition at this “clean air” site suggests that high Hg deposition may occur in the tropics worldwide.

Currently, there is no coordinated global observational network for mercury that could be used by the modelling community or to establish recommendations to protect human health and ecosystems on a global scale. Current national/regional monitoring networks are inadequate for a global scale assessment. Recognizing that TGM and Hg in wet deposition are spatially heterogeneous, several studies have aimed to set up monitoring networks in order to compare trends between sites in the same region, between regions, and to determine the influence of local and regional emission sources. It is important to stress that the measurement of Hg by itself is not sufficient to improve our understanding of Hg sources and impacts. Measurements of other key atmospheric constituents at the global monitoring sites are necessary to develop a better understanding of the global redistribution of Hg, and to further refine the parameterisation of key processes in global and regional atmospheric mercury models. GMOS is a five year project (www.gmos.eu) funded by the European Commission; it is aimed to establish a worldwide observation system and will include ground-based monitoring stations, shipboard measurements over the Pacific and Atlantic Oceans and European Seas, as well as aircraft-based measurements in the UTLS. GMOS will support UNEP F&T, GEOSS, and UNECE-TF HTAP and will benefit and will cooperate with on-going monitoring programs in Europe, North America and Asia. The specific objectives of GMOS are:

1. To establish a Global Observation System for Mercury able to provide ambient concentrations and deposition fluxes of mercury species around the world, by combining observations from permanent ground-based stations, and from oceanographic and tropospheric measurement campaigns.
2. To validate regional and global scale atmospheric mercury modelling systems able to predict the temporal variations and spatial distributions of ambient concentrations of atmospheric mercury, and Hg fluxes to and from terrestrial and aquatic receptors.
3. To evaluate and identify source-receptor relationships at country scale and their temporal trends for current and projected scenarios of mercury emissions from anthropogenic and natural sources.
4. To develop interoperable tools to allow the sharing of observational and models output data produced by GMOS, for the purposes of research and policy development and implementation as well as at enabling societal benefits of Earth observations, including advances in scientific understanding in the nine Societal Benefit Areas (SBA) established in GEOSS.

The objectives of this study were to (i) collect atmospheric mercury wet and dry deposition in urban, rural and background areas in the central region of South Korea from August 2008 to December 2009, (ii) to analyze spatial variability in rainfall Hg concentrations between three sites (urban: Seoul, rural: Yangsuri, background: Kanghwa Island), (ііі) to calculate near-field attribution percentage for enhanced local and regional deposition gradients near anthropogenic sources in urban site compared to rural and background sites, and (іv) to identify likely sources areas which impact Seoul using back-trajectory and Lagrangian Particle Dispersion Model (LPDM). Analysis of wet deposition samples (Total Mercury, TM) was performed in a class-100 clean room using Tekran 2600 equipped with Cold Vapor Atomic Fluorescence Spectrometry (CVAFS). The volume weighted mean TM concentration and wet deposition flux in urban, rural and background sites were 12.21 ± 15.51 ng L^-1 and 18.45 µg m^-2 , 5.49 ± 4.92 ng L^-1 and 2.96 µg m^-2, 9.67± 8.36 ng L^-1 and 11.32 µg m^-2, respectively. Annual dry deposition fluxes of RGM, and Hg_p were 58.47 ± 2.31 ug m^-2 year^-1 and 43.24 ± 2.76 ug m^-2 year^-1 for the urban site, 2.8 ± 0.13 ug m^-2 year^-1 and 2.72 ± 0.09 ug m^-2 year^-1 for the rural site, 5.20 ± 0.48 ug m^-2 year^-1 and 2.55 ± 0.08 ug m^-2 year^-1 for the background site, respectively. Annual mean dry deposition velocities of RGM and Hg_p (cm sec^-1) were calculated using measured dry deposition flux divided by its atmospheric concentration and those at urban site during 2009 were 2.18 cm sec^-1 (r² = 0.5, p < 0.01) and 0.37 cm sec^-1, respectively showing similar results to previous studies (2.0 cm sec^-1 for RGM and 0.07 cm sec^-1 for Hg_p).

Acknowledgements:
This work was supported by the Ministry of Environment (Eco-technopia 21 project 2009-12001-0050-1) and Ministry of Education, Science and Technology, Republic of Korea (R01-2008-000-11165-0).