Toward the next generation of air quality monitoring: Mercury

Mercury is a global pollutant that is ubiquitous in the environment. Enrichment of mercury in the biosphere as the result of human activities and subsequent production of methylmercury (MeHg) has resulted in elevated concentrations in fish, wildlife and marine mammals globally. Elemental mercury (Hg0) is the most common form of mercury in the atmosphere, and the form that is most readily transported long distances from its emission source. Most mercury deposition from the atmosphere is in the highly soluble, oxidised inorganic form HgII. Thus, understanding atmospheric transport and oxidant distribution is essential for understanding mercury inputs to ecosystems. Methylmercury (MeHg) is the most toxic form of mercury that accumulates in aquatic food web and can cause a variety of negative health effects such as long-term IQ deficits and cardiovascular impairment in exposed individuals. Humans are predominately exposed to MeHg by consuming fish. Hg0 emitted from anthropogenic sources has a long (6 monthse1 year) atmospheric residence time allowing it to be transported long distances in the atmosphere. It is eventually oxidised to the highly soluble HgII (likely by atomic Br and/or OH/O3) and rapidly deposited with precipitation. Some of the mercury deposited to terrestrial and marine ecosystems is converted to MeHg, which is the only form that bioaccumulates in aquatic food webs. Recent studies suggest that there is a first-order relationship between the supply of inorganic mercury to ecosystems and production of MeHg, thus implying that declines in deposition will translate directly into reduced concentrations in biota and human exposures. However, one of the major uncertainties in this cycle is the time scale required for these changes to take place and this is known to vary from years to centuries across different environmental compartments depending on their physical and biogeochemical attributes. Thus, a key challenge in the case of mercury pollution is understanding the link between the magnitude of mercury emissions and the concentrations found in the fish that we consume. For air quality monitoring, priorities include expanding the existing data collection network and widening the scope of atmospheric mercury measurements (elemental, oxidised, and particulate species as well as mercury in precipitation). Presently, the only accurate indicators of mercury impacts on human and biological health are methylmercury concentrations in biota. However, recent advances in analytical techniques (stable mercury isotopes) and integrated modelling tools are allowing greater understanding All rights reserved. N. Pirrone et al. / Atmospheric Environment 80 (2013) 599e611 600 of the relationship between atmospheric deposition, concentrations in water, methylation and uptake by biota. This article recommends an expansion of the current atmospheric monitoring network and the establishment of new coordinated measurements of total mercury and methylmercury concentrations in seawater and concurrent concentrations and trends in marine fish. 2013 Elsevier Ltd. All rights reserved. 1. Background and objective The toxicity of mercury and its compounds for humans such as ataxia, constriction of vision, impaired hearing and death was first described in 1865 (Grandjean et al., 2010). Mercury mining and use in products continues to the present day. Consumer products still containing mercury include button cell batteries, fluorescent bulbs, and some cosmetics (McKelvey et al., 2011; Streets et al., 2011). However, most mercury emitted to the atmosphere since the beginning of the Industrial Revolution has been as a result of coal combustion, and to a lesser extent metal smelting and more recently cement production andwaste disposal through incinerator plants (Nriagu and Pacyna, 1988; Pirrone et al., 1996, 2010). Most non-occupational human exposure to mercury is from fish and marine seafood containing MeHg (e.g. Sunderland, 2007; Mahaffey et al., 2004, 2009). MeHg is a potent neurotoxin that causes a variety of reproductive and developmental disorders at high concentrations (Clarkson and Magos, 2006). New research has also suggested a link between MeHg exposure and cardiovascular health in adults, although conflicting results have been reported across epidemiological studies (Mozaffarian et al., 2011; Roman et al., 2011; Valera et al., 2011; Wennberg et al., 2012). Unlike HgII, MeHg is absorbed efficiently from food (>90%) and readily crosses the bloodebrain and placental barriers. Gaseous Hg0 is also efficiently absorbed when inhaled, although concentrations in the atmosphere, even next to point sources such as coal-fired utilities, are far below levels of toxicological concern (Clarkson and Magos, 2006). Occupational exposures to Hg0 (which can also cross the bloodebrain barrier) associated with artisanal gold mining and due to vaporisation of quicksilver used in products and other uses can cause severe health effects such as kidney failure and central nervous system impacts (Mergler et al., 2007). Atmospheric mercury exists in three forms: Gaseous elemental mercury (GEM), gaseous oxidised mercury (HgII) compounds (GOM), and mercury associated with particulate matter (HgP). Oxidised mercury compounds are emitted from anthropogenic sources and readily transferred to aquatic and terrestrial receptors by dry deposition processes and wet scavenging by precipitation. Oxidised mercury compounds are much less volatile, and most are more water soluble than Hg0. The precise chemical nature of these compounds is still not known and thus the term GOM is used to describe all forms of mercury sampled from the atmosphere using a KCl-coated denuder and analysed by CVAAFS (Landis et al., 2002). There is some debate however over the efficiency of the commonly used sampling techniques, and indeed what they sample, although alternative techniques are being developed (Lyman et al., 2010b; Ambrose et al., 2013; Gustin et al., 2013). Mercury associated with particulate matter can be emitted from anthropogenic sources, active volcanic eruptions and evaporation of cloud/aerosol droplets that contained mercury compounds. These particles are generally part of the fine aerosol fraction and their transport and deposition characteristics are defined by particle properties. Mercury in this form is thought to be mostly insoluble. All three forms are released by anthropogenic sources, primarily combustion processes, as well as by a variety of natural sources and processes. Natural sources include crustal degassing, volcanoes, a component of the reemitted mercury from soils and aquatic surfaces, weathering processes of the Earth’s crust and some forest fires (Pirrone et al., 2010). On a global scale, the dominant component of the mercury released from terrestrial and oceanic systems is previously deposited anthropogenic mercury rather then geogenic sources (Streets et al., 2011). Contributions from natural sources and processes vary geographically and over time depending on a number of factors including meteorological conditions, the presence of volcanic or geothermal activities, the presence of Hg bearing minerals such as cinnabar, the magnitude of exchange processes between waters and the atmosphere, the reemission of previously deposited Hg from top soils and plants, and also the occurrence of forest fires (Mason, 2009; Friedli et al., 2009; Pirrone et al., 2010). In addition to anthropogenic mercury emissions (z2000 Mg yr 1), reemitted mercury from soils and aquatic ecosystems presently contribute approximately 2/3 of global emissions to the atmosphere (Corbitt et al., 2011). Pirrone et al. (2010) estimated that mercury evasion from terrestrial surfaces is about 2430 Mg yr 1 and that from surface waters (oceans and lakes) is about 2780 Mg yr 1. The latest simulations based on data from GEOSChem global biogeochemical mercury model show similar estimates with emissions from land at 2100 Mg yr 1 and 3100 Mg yr 1 from the ocean (Streets et al., 2011). On an areal basis, re-emissions from the land (surface 1.46 108 km2) are higher than those from the ocean (surface 3.49 108 km2), and the majority of anthropogenic mercury appears to accumulate in the subsurface and deep ocean. Biomass burning estimates range between 200 and 675 Mg yr 1 (Friedli et al., 2009; Holmes et al., 2010), whereas desert and non-vegetated zones emit 546 Mg yr 1, followed by tundra and grassland with 448 Mg yr 1, forest with 342 Mg yr 1 and contaminated sites with 200 Mg yr 1 (Pirrone et al., 2010). Mercury in ocean waters is present as elemental mercury (Hg0), monomethyl mercury (MeHg), dimethyl mercury (Me2Hg), aqueous divalent mercury (HgII), colloidal mercury, and particulate mercury (Mason et al., 2012). Globally, total Hg concentrations in the ocean mixed layer are generally <1 pM (1 pM 1⁄4 10 12 mol l 1) (Soerensen et al., 2010) and slightly greater than 1 pM in the upper 1000m of thewater column (Mason et al., 2012). In the early 1980s, values as high as 9.6 pM were measured and may provide evidence for a historical decline in some ocean basins (Gill and Fitzgerald, 1987; Soerensen et al., 2012). Mercury airesea exchange is primarily driven by (i) the concentration gradient of Hg0 between top-water microlayer and the air above, (ii) solar irradiation, which is responsible for photo-mediated redox processes, (iii) wind speed and temperature at the airewater interface, (iv) and the supply of reducible HgII available for reduction to Hg0 and the intensity of other processes that affect this supply such as scavenging by particulate organic matter (Pirrone et al., 2003; Soerensen et al., 2010). Evasion of Hg0 competes with methylation for the available substrate of mercury in seawater and is thus a critical loss process affecting biological MeHg concentrations (Mason et al., 2012). N. Pirrone et al. / Atmospheric Environment 80 (2013) 599e611 601 Measurements of atmospheric and aquatic concentrations of mercury have now been performed on a sampling campaign basis in a number of parts of the world, and in some areas repeatedly (Sprovieri et al., 2009; Soerensen et al., 2012). Exceptions include the Southern Hemisphere and most oceans, where data are still patchy. Long-standing monitoring networks only exist in Europe and the North American continent, although a recent initiative from the EU is setting up a global monitoring network for mercury as part of the Global Mercury Observation System (GMOS) project (www.gmos.eu). Guidelines for exposure to mercury include: 1 mg L 1 for Hg in water, 1 mg m 3 for air (annual average) and 0.2 mg m 3 for longterm inhalation of exposure to elemental mercury vapour (FAO/ WHO, 2007). The U.S. EPA reference dose for daily intake of MeHg that is without an appreciable increase in risk of adverse effects over a lifetime is 0.1 mg kg 1 body weight per day. Acceptable exposures to MeHg are continually being refined with the discovery of effects of MeHg on the central nervous system at lower levels (Grandjean et al., 2010). Other governmental agencies have developed MeHg intake levels to protect public health that range from 0.10 to 0.47 mg kg 1 body weight per day, UNEP (see Table 4.1 of 2002). Last year the EFSA Panel on Contaminants in the Food Chain (CONTAM) established a Tolerable Weekly Intake (TWI) of 1.3 mg kg 1 body weight (EFSA, 2012) less than the TWI of 1.6 mg kg 1 body weight recommended by the Joint FAO/WHO Expert Committee on Food Additives (FAO/WHO, 2007). Differences in safe recommended intakes relate to methods used to estimate exposure considered to bewithout adverse effects (extrapolation of measured dose response relationships for IQ and MeHg exposure) rather than exposures producing effects (Mahaffey et al., 2011). In Europe, environmental target levels that are safe for both humans and the environment have been established by a legislative framework (i.e. Environmental Quality Standards Directive, 2008/ 105/EC). In particular, either Environmental Quality Standards (EQSs), which represent thresholds that should not be exceeded, or Environmental Assessment Criteria (EACs), which are long-term objectives close to Background Concentrations, have been adopted for different marine sediment, water and biota. Regulatory levels have been also established in community legislation for protection of public health. Following guideline levels formerly established within the Codex Alimentarius (FAO/WHO, 1995), maximum levels of mercury in certain foods have been established by Commission Regulation (EC) No 1881/2006. Maximum levels for mercury in fish range from 0.2 ppm to 1.0 ppm depending on the agency. These levels are generally derived from the above safe intake levels for humans and depend on the assumedmagnitude of fish consumption in a given population that will maintain intakes below the safety threshold. State governments in the U.S. have issued advisories for populations that are assumed to eat large quantities of fish (e.g. Minnesota), while international standards set by different federal governments tend to be between 0.3 ppm (U.S.) and 1.0 ppm. Contaminants in fish and other seafood influence both the health of the consumer and the sustainable use of marine resources (Oken et al., 2012). This article makes recommendations for improved atmospheric Hg monitoring and modelling strategies to better link emission sources to human health impacts. Since most human exposure to Hg results from consuming MeHg in seafood, we discuss possibilities for establishing a coordinated atmosphericeocean Hg measurement strategy. Such a strategy would allow us to better assess the temporal response of marine ecosystems to reductions of mercury emissions to the atmosphere from anthropogenic sources, and future changes in the global mercury cycle that may occur as the result of variability in reemissions from soils and the ocean and driven by climate. 2. Overview of existing monitoring systems and indicators 2.1. Ground-based monitoring networks Systematic long-term direct measurements of mercury in the atmosphere provide valuable information about the impact of emission controls on the global budget of atmospheric mercury and offer insight into source-receptor transboundary transport of mercury. Additional mercury species measurements such as oxidized and particle-bound mercury compounds can help to improve the understanding of local atmospheric chemistry and short-term oxidation processes regarding the removal of mercury from the atmosphere. Harmonized Standard Operating Procedures (SOPs) and QA/QC protocols for monitoring ambient concentrations of all mercury species are needed in order to assure a full comparability of site specific observational datasets with that obtained inside and outside existing monitoring networks. SOPs and QA/QC protocols should be in accordance with measurement practice adopted in well established monitoring networks and based on the most recent literature. State-of-the-art SOPs and QA/QC protocols have been developed in the frameworks of several programs including EMEP, CEN-TC 264, NADP, AMNet, CAMNet, GMOS. A large number of activities have been carried out to characterise the levels of mercury (Hg) species in ambient air and precipitation, in order to understand how they vary over time and how they depend on meteorological conditions. Monitoring of ambient mercury is focused on the three primary forms of mercury in the atmosphere: GEM, GOM, HgP. The measurement of atmospheric GEM is now routine, and can be easily implemented due to its relatively high concentration and chemical inertness. Uncertainties, detection limits and ruggedness are well established (Brown et al., 2010; Steffen et al., 2012; Gay et al., 2013). In contrast, the measurement of the atmospheric mercury species GOM and HgP are more challenging and uncertain due to their extremely low concentrations, more complex chemical reactivity and analytical challenges related to coated and non-coated glassware, heated filters, vials and tubings. However, GOM and PBMmeasurements are critical to help define and model the fate and transport of atmospheric mercury. The atmosphere provides the main environmental pathway for redistribution of Hg around the globe, and based on the existing data, there is a scientific consensus about the current global background concentration of airborne Hg which is considered to be in the range of 1.5e1.7 ng m 3 in the Northern Hemisphere and 1.1e 1.3 ng m 3 in the Southern Hemisphere (Lindberg et al., 2007; Sprovieri et al., 2010b; Slemr et al., 2011). Due to its global nature, it is critically important to quantify the transfer of Hg from the air to the Earth’s surface via wet and dry deposition and analyse the global long-term trends of mercury in the atmosphere and in precipitation. To date a lack of legislation at national as well as international levels has resulted in limited long-term measurements, the exceptions being some sites in Europe and the Arctic (Ebinghaus et al., 2011; Cole and Steffen, 2010; Cole et al., 2013). Measurement data is particularly limited in the Southern Hemisphere. There is the need to coordinate activities at the global level to ensure that future research provides the maximum benefits in terms of assessing global and regional trends in Hg concentration. The Group on Earth Observations (GEO, http://www. earthobservations.org/about_geo.shtml) has established the Task HE-02-C1 “Global Mercury Observation System” for the work plan 2012e2015, which is a continuation of the activity initiated in the GEO Work Plan (2009e2012). This task supports the achievement of the goals of GEOSS http://www.earthobservations.org/geoss. shtml and other on-going international programs such as the UNEP Mercury Program and international conventions dealing with large-scale transboundary transport of mercury such as the N. Pirrone et al. / Atmospheric Environment 80 (2013) 599e611 602 United Nations Economic Council for Europe Convention on Longrange Transboundary Air Pollution (UNECE-CLRTAP, http://www. unece.org/env/lrtap/welcome.html). Table 1 provides a summary of the existing networks for measuring mercury in ambient air and precipitation. A number of regional monitoring networks have been in operation for many years in North America and Northern Europe. Measurements have also been collected at a few sites in Asia and in the Arctic region. However, in many other parts of the world, especially the Southern Hemisphere, such extensive measurement networks do not exist. Given the mostly ad hoc nature and spatially heterogeneous distribution of the efforts to monitor and measure atmospheric Hg species, it has become clear that a coordinated global monitoring network is needed to provide information for a global assessment, and for global and regional model evaluation and extrapolation (Pirrone et al., 2008). The European Monitoring and Evaluation Program (EMEP) under the UNECE-LRTAP convention was the first international measurement network for mercury, with Swedish measurements dating from 1980 and other European sites, located mostly in the northern, western and central parts of Europe starting in 1990 (Tørseth et al., 2012). Several of the EMEP sites are also part of other regional networks such as that related to the marine Conventions OSPARCOM and HELCOM, and the Arctic Programme AMAP. Long-term monitoring of atmospheric Hg with high time resolution started at Alert, Canada in 1995, which was the first