Science of the Total Environment 509–510 (2015) 3–15

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Review

Atmospheric mercury in the Canadian Arctic. Part I: A review of recent field measurements Alexandra Steffen a,⁎, Igor Lehnherr b, Amanda Cole a, Parisa Ariya c,d, Ashu Dastoor e, Dorothy Durnford e, Jane Kirk f, Martin Pilote g a

Environment Canada, Air Quality Processes Research, Toronto M3H 5T4, Ontario, Canada University of Waterloo, Department of Earth and Environmental Sciences, Waterloo N2L 3G1, Ontario, Canada McGill University, Department of Chemistry, 801 Sherbrooke St. W., Montreal H3A 2K6, Quebec, Canada d McGill University, Department of Atmospheric and Oceanic Sciences, 801 Sherbrooke St. W., Montreal H3A 2K6, Quebec, Canada e Environment Canada, National Prediction Development Division, Dorval H9P 1J3, Quebec, Canada f Environment Canada, Aquatic Contaminants Research Division, Burlington L7R 4A6, Ontario, Canada g Environment Canada, Aquatic Contaminants Research Division, Montreal H2Y 2E7, Quebec, Canada b c

H I G H L I G H T S • This paper reviews progress made in the study of the transport, transformation, deposition and reemission of atmospheric Hg in the Canadian Arctic, focusing on field measurements. • Redox processes control the speciation of atmospheric Hg and bromine radicals are the primary oxidant of atmospheric Hg depletion in the spring • It is expected that a smaller fraction of deposited Hg will be reemitted from coastal snowpacks. • Atmospheric gaseous Hg concentrations have decreased in some parts of the Arctic but at a rate that was less than that at lower latitudes • Remaining knowledge gaps are: the identification of oxidized Hg species in the air, total deposition amounts, physical-chemical processes over sea ice and impacts of Arctic climate change

a r t i c l e

i n f o

Article history: Received 17 June 2014 Received in revised form 27 October 2014 Accepted 31 October 2014 Keywords: Mercury Arctic Atmosphere Redox Snow Temporal trends

a b s t r a c t Long-range atmospheric transport and deposition are important sources of mercury (Hg) to Arctic aquatic and terrestrial ecosystems. We review here recent progress made in the study of the transport, transformation, deposition and reemission of atmospheric Hg in the Canadian Arctic, focusing on field measurements (see Dastoor et al., this issue for a review of modeling studies on the same topics). Redox processes control the speciation of atmospheric Hg, and thus impart an important influence on Hg deposition, particularly during atmospheric mercury depletion events (AMDEs). Bromine radicals were identified as the primary oxidant of atmospheric Hg during AMDEs. Since the start of monitoring at Alert (NU) in 1995, the timing of peak AMDE occurrence has shifted to earlier times in the spring (from May to April) in recent years, and while AMDE frequency and GEM concentrations are correlated with local meteorological conditions, the reasons for this timing-shift are not understood. Mercury is subject to various post-depositional processes in snowpacks and a large portion of deposited oxidized Hg can be reemitted following photoreduction; how much Hg is deposited and reemitted depends on geographical location, meteorological, vegetative and sea-ice conditions, as well as snow chemistry. Halide anions in the snow can stabilize Hg, therefore it is expected that a smaller fraction of deposited Hg will be reemitted from coastal snowpacks. Atmospheric gaseous Hg concentrations have decreased in some parts of the Arctic (e.g., Alert) from 2000 to 2009 but at a rate that was less than that at lower latitudes. Despite numerous recent advances, a number of knowledge gaps remain, including uncertainties in the identification of oxidized Hg species in the air (and how this relates to dry vs. wet deposition), physical–chemical processes in air, snow and water—especially over sea ice—and the relationship between these processes and climate change. Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

⁎ Corresponding author. Tel.: +1 416 739 4116; fax: +1 416 739 4179. E-mail address: [email protected] (A. Steffen).

http://dx.doi.org/10.1016/j.scitotenv.2014.10.109 0048-9697/Crown Copyright © 2014 Published by Elsevier B.V. All rights reserved.

4

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atmospheric processes of Hg . . . . . . . . . . . . . . . . . . . . . . 2.1. Overview of atmospheric Hg depletion events . . . . . . . . . . 2.2. Chemical reactions of Hg in the Arctic atmosphere . . . . . . . . 2.3. Temporal trends of atmospheric Hg depletion events at Alert . . . . 2.4. Distribution of atmospheric Hg species in the Arctic . . . . . . . . 3. Deposition of atmospheric Hg . . . . . . . . . . . . . . . . . . . . . 4. Summary of Canadian atmospheric measurement data and spatial patterns . 4.1. Gaseous elemental Hg . . . . . . . . . . . . . . . . . . . . . 4.2. Reactive gaseous Hg and airborne particulate Hg . . . . . . . . . 5. Temporal trends of gaseous elemental Hg . . . . . . . . . . . . . . . . 6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction The atmosphere plays a fundamental role in the mercury (Hg) cycle in the Canadian Arctic, facilitating both the transport and deposition of this toxic contaminant. Due to its long (6–24 months; Pacyna et al., 2006) atmospheric residence time (lifetime in the air), Hg can undergo long-range transport from distant point sources to remote regions such as the Arctic. Various wet- and dry-depositional processes lead to the input of atmospheric Hg to Arctic terrestrial, freshwater, and marine ecosystems, where it may eventually bioaccumulate and biomagnify in local foodwebs following its conversion to methylmercury (MeHg) (Lehnherr, 2014). Therefore, the transport, deposition and fate of atmospheric Hg play an important role in determining Hg exposure in humans and wildlife. Efforts to characterize the atmospheric transport and delivery of Hg to the Arctic have been challenging because Hg emissions to the atmosphere occur from both natural sources and anthropogenic activities, and complex exchanges of Hg occur between the atmosphere and its interfaces with soil, water, and the cryosphere (Durnford and Dastoor, 2011). Furthermore, the cycle of atmospheric Hg has a number of features that are unique to the Arctic and Polar Regions, compared to lower-latitude regions, due to processes such as atmospheric mercury depletion events (Steffen et al., 2008a) (AMDEs, Section 2.1). Measurements of atmospheric Hg in the Canadian Arctic were initiated in the early 1990s, leading to the influential discovery of AMDEs in 1995 (Schroeder et al., 1998). During springtime AMDEs, gaseous elemental Hg is rapidly depleted from the lower atmosphere through oxidative processes and deposited on the ground or bound to aerosol surfaces (Steffen et al., 2014). This process has sparked interest in atmospheric Hg and the role of AMDEs in the Arctic Hg cycle. Phase II of the Northern Contaminants Program (NCP) (1998–2003) resulted in the establishment of automated, long-term measurements of gaseous elemental Hg (GEM) at Alert, revealing seasonal patterns resulting from AMDEs such as the rapid removal of GEM from the air within a period of hours after polar sunrise. AMDEs were also observed during springtime at subarctic latitudes, in Kuujjuarapik (Québec) and Churchill (Manitoba) for example, where incoming solar radiation remains present during the entire winter, suggesting that other processes may also occur (Poissant and Pilote, 2003; Kirk et al., 2006). Sites where AMDEs predominately occur and the chemistry responsible for these events were better characterized. The first measurements of ionic or non-elemental atmospheric Hg species, namely reactive gas phase Hg (RGM) and particulate Hg (PHg), were conducted at Alert (Nunavut) (Cole et al., 2013; Steffen et al., 2014) and Kuujjuarapik. Collectively this research provided a new understanding of Hg processes in the Arctic, and was highlighted as a key finding in the previous Canadian Arctic Contaminants Assessment Report (CACAR II) (Bidleman et al.,

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

4 4 5 5 6 7 8 11 11 12 12 13 13 13

2003). However, the report also concluded that the significance of AMDEs to the Arctic environment was yet to be determined. During Phase III of the NCP further progress has been made on elucidating the chemical reactions resulting in oxidation of atmospheric Hg in the Arctic and the critical role of halogens as oxidants. Field measurements on Hg speciation, including both GEM and its oxidized counterparts RGM and PHg, were collected in the Canadian Arctic (Kirk et al., 2006a, 2012; Steffen et al., 2014). Automated measurements of air Hg concentrations at Alert (High Arctic) and Kuujjuarapik (sub-Arctic) over the last decade provide new insights into temporal trends at different time scales. The geographic coverage of air measurements has been expanded with continuous data now available for one new site in the Yukon. Extensive modeling of atmospheric processes has been conducted (discussed in Dastoor et al., this issue). This has resulted in a new understanding of long-range atmospheric transport of Hg from various source regions around the world as well as refined estimates of atmospheric Hg deposition in the Arctic (Dastoor et al., 2008; Dastoor and Durnford, 2014; Durnford et al., 2010). Atmospheric deposition is an important source of Hg to Arctic aquatic and terrestrial ecosystems and the objective of this review is to highlight recent findings, identify knowledge gaps, and recommend directions for future research on the atmospheric deposition of Hg in the Canadian Arctic. This is Part 1 of a two part review paper describing the work that has been undertaken on atmospheric Hg in the Canadian Arctic. Part 1 describes the measurements undertaken to describe Hg from its presence in the Arctic air; to the reactions it undergoes in the atmosphere; to how it is distributed; and to how this distribution affects deposition and emission of Hg to the Arctic surfaces, including a model describing the uptake or emission of Hg in the snow. Part 2 (Dastoor et al., this issue) describes the modeling of Hg from its transport, transformation, deposition and emission which has made use of the results described in this Part 1 of the atmospheric Hg paper. 2. Atmospheric processes of Hg Three forms of Hg are considered in atmospheric Hg processes, including GEM or Hg(0), and two operationally defined forms: 1) RGM, which consists of some oxidized gaseous inorganic Hg(II) and Hg(I) species; and 2) PHg, which consists of oxidized Hg associated with particles. Each of these three categories has a different residence time in the atmosphere. GEM can remain in the atmosphere for long periods of time and, thus can be distributed globally as a result of longrange transport from point sources. In contrast, the more reactive oxidized forms of Hg are generally scavenged or deposited more locally near emission sources either directly or following atmospheric chemical conversions. Hence, it is highly likely that GEM is the dominant form of Hg that is transported to the Arctic by the atmosphere (Durnford et al.,

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

2010). The conversion of GEM to PHg and RGM in the Arctic atmosphere is a key step leading to the deposition of Hg to Arctic ecosystems. Atmospherically-deposited Hg can be subsequently methylated to MeHg in aquatic environments, and this organic toxic form of Hg can then bioaccumulate in biota and humans (Lehnherr, 2014) (see Chételat et al., this issue; Braune et al., this issue). It should be noted that RGM and PHg as discussed in this paper are operationally defined as that which stick to a KCl coated denuder and measured on particles less than 2.5 μm, using a Tekran model 2537/1130/ 1135 instrument, respectively. Measuring RGM and PHg at the pg/m3 level is very challenging. Recently, some research has shown that there may be some bias in the RGM and PHg measurements using this instrumentation (Gustin et al., 2013; Kos et al., 2013; Lyman et al., 2010); however, for the purposes of this review these measurements are considered operationally defined and accepted as is. It should also be noted that the data presented in the following have all been treated with the same quality assurance protocols that are described by Steffen et al. (2012).

2.1. Overview of atmospheric Hg depletion events AMDEs are characterized as a series of photochemically-initiated reactions in the atmosphere which result in the conversion of GEM to RGM, which can also associate with aerosols to form PHg. Both RGM and PHg have short atmospheric lifetimes and are rapidly deposited near sources to terrestrial and aquatic surfaces, which are still snow and ice-covered during polar spring when AMDEs take place. Extensive research has been conducted on AMDEs and the findings have been reviewed (Steffen et al., 2008), showing that the initiation of AMDEs requires Hg in the atmosphere, cold temperatures, a stable inversion layer, sunlight, and reactive halogens (Dommergue et al., 2009; Nguyen et al., 2009; Poissant et al., 2008). Sea salt in ice, snow or frost flowers may provide a large source of halogen species that, under certain conditions, can react to form radicals such as bromine (Br), bromine monoxide (BrO), chlorine (Cl), and chlorine monoxide (ClO) to initiate these photochemical reactions (Simpson et al., 2007). Radicals can be produced directly upon photolysis of their precursor or, through secondary reactions; the original photochemically initiated radicals can interact with a suite of molecules, generating new radicals in atmosphere or atmospheric interfaces (Subir et al., 2012). The fate of Hg deposited during AMDEs is less clear, since both or either RGM and PHg are deposited to surfaces depending on the reaction processes. Studies have concluded that some of the deposited RGM is reduced to GEM and re-emitted into the atmosphere while some remains or is reoxidized at the surface of, or within, the snowpack (see Durnford and Dastoor, 2011). Studies of the photochemistry of Hg within the snowpack have shown that reactions occur within the first few centimeters of the snow surface, and flux measurements indicate that there is net deposition of Hg to the snow in early spring followed by a net emission of Hg from the snowpack in the summer (Constant et al., 2007; Dommergue et al., 2003, 2007; Ferrari et al., 2005; Lalonde et al., 2002; Poulain et al., 2004). Global and regional models now incorporate AMDEs into their chemical mechanisms although

5

there is still an incomplete understanding of the chemical processes that drive deposition and re-emission of Hg in this environment.

2.2. Chemical reactions of Hg in the Arctic atmosphere Until recently, only indirect evidence was available for the oxidation of long-lived Hg(0) by halogen atoms — particularly bromine — to easily deposited RGM species including Hg(I) and Hg(II). This evidence was based on a strong correlation between AMDEs and ozone depletion events, which were believed to be caused by catalytic loss of ozone through reaction with halogen atoms and showed a negative correlation with concentrations of bromine oxide—one of the components of the catalytic ozone loss cycle—in the boundary layer and total air column (Simpson et al., 2007). Direct evidence for the oxidation of Hg in the gas phase has more recently been gathered through laboratory measurements (Ariya et al., 2002; Donohoue et al., 2006; Pal and Ariya, 2004; Raofie and Ariya, 2003) and theoretical calculations (Goodsite et al., 2004; Khalizov et al., 2003; Maron et al., 2008; Shepler et al., 2005) of reaction rates and reaction products. These findings are reviewed in more detail elsewhere (Ariya et al., 2008, 2009) but a summary of selected oxidation reaction rate coefficients is presented in Table 1 along with approximate oxidant concentrations ((Sander and Bottenheim, 2012) and references therein). This table illustrates how the importance of each reaction depends on both the oxidant concentration and the rate coefficient for a given reaction, though these estimates do not incorporate temperature- and pressure-dependence of the reactions. For example, the reaction of Hg(0) with Br is slower than the reaction with Cl by a factor of 2–25 but since the concentration of Br during bromine explosions can be approximately 100–2,000 times higher than that of Cl (Cavender et al., 2008), the reaction of Hg(0) with Br is considered more important. In fact, models of atmospheric Hg chemistry suggest that oxidation by bromine radicals alone can account for the rapid oxidation of GEM that occurs during AMDEs when GEM concentrations can sometimes drop by a factor of ten in a matter of hours (Ariya et al., 2004; Goodsite et al., 2004; Holmes et al., 2006, 2010; Skov et al., 2004) and may also be a significant oxidant of GEM in other seasons (Holmes et al., 2010). However, there is evidence to suggest that Cl concentrations may be significantly higher during some AMDEs and could significantly contribute to observed GEM oxidation (Stephens et al., 2012). It is also clear from Table 1 that there is a great deal of uncertainty associated with many of the rate coefficients. Some discrepancy between different experiments has been attributed to reactions on the walls of reaction chambers, suggesting not only that the gas phase reaction rates may be inaccurate but also that additional reactions on the surface of particles, snow, ice, and ocean water may contribute to GEM oxidation in the polar atmosphere (Sabir et al., 2011). Additionally, there are limited data for the temperature dependence of these reactions (Donohoue et al., 2005, 2006; Goodsite et al., 2004). Regardless of their source, these large uncertainties in reaction rates must be narrowed in order to accurately model Hg oxidation during AMDEs and throughout the year.

Table 1 A summary of selected experimental and theoretical reaction rate coefficients for the oxidation of GEM (Ariya et al., 2008, 2009)., along with estimated GEM lifetime based on estimated oxidant concentrations (Sander and Bottenheim, 2012) The calculated lifetime corresponds to the time required to oxidize 63% of initial Hg(0) with each oxidant. Oxidant

Estimated concentration in polar spring (molecules cm−3)

Rate coefficient for reaction with GEM (cm3 molecules−1 s−1)

GEM chemical lifetime with this oxidant

Br Cl BrO OH O3

107–108 104 b5 · 108 105–106 b1012

4 · 10−13–3 · 10−12 5 · 10−13–1 · 10−11 10−15–10−13 9 · 10−14–3 · 10−13 10−20–10−18

b1 h to 3 days 115 days to 5.9 years N5 h toN23 days 36 days to 3.6 years 11 days to 4 years

6

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

a

b

1996–2000, 2002 February

2003–2009

February

March

June

June

March

April May

May

April

Fig. 1. Average distribution of AMDEs in each of the spring months at Alert (Nunavut), for the years (a) 1996 to 2000 and 2002, and (b) 2003 to 2009. Correcting data for different sampling rates resulted in changes of b1% (Cole and Steffen, 2011).

0.7

2.0

0.6 1.5

0.5 0.4

1.0

0.3 0.2

0.5

0.1 0.0 0 to 5

-5 to 0

-10 to -5

-15 to -10

-20 to -15

-25 to -20

-30 to -25

-35 to -30

-40 to -35

-45 to -40

0.0

mean GEM (ng m-3) (lines)

integrated AMDE frequency (bars)

The reduction of Hg(I) and Hg(II) back to Hg(0) is another important component of the atmospheric Hg cycle in the Arctic. Reduction is generally assumed to take place in the aqueous phase, since Hg(I) and Hg(II) species are soluble, and reduction results in volatile Hg(0) that is easily released to the atmosphere. This process also occurs in lakes and oceans (Chételat et al., this issue; Braune et al., this issue). However, Hg reduction also occurs in snowpack and potentially in cloud and rain droplets, and in particles with significant water content. Emission of GEM from snow has been observed following AMDEs (Section 3) and appears to depend on exposure to ultraviolet (UV-B) radiation (Lalonde et al., 2003; Poulain et al., 2004). Little is known about the

2.3. Temporal trends of atmospheric Hg depletion events at Alert

2.0

0.6 1.5

0.5 0.4

1.0

0.3 0.2

0.5

0.1 0.0 315 to 360

270 to 315

225 to 270

180 to 225

135 to 180

90 to 135

45 to 90

0 to 45

0.0

mean GEM (ng m-3) (lines)

integrated AMDE frequency (bars)

temperature bin (oC) 0.7

wind direction bin (o) Feb

Mar

Apr

May

reactants involved in reduction reactions or their rates. For example, rate constants for the reduction of Hg(II) ions by hydroperoxyl (HO2) radicals and for the decomposition of the mercuric sulphite (HgSO3) complex to Hg(0)—both in water—have been measured but are still highly uncertain (Sabir et al., 2011, and references therein). Similarly, the photoreduction of Hg(II) species in the presence of organic matter has been suggested (Si and Ariya, 2008, 2011), however the implications and relevance of this process have yet to be evaluated in situ. Photoreduction of Hg on surfaces, such as aerosols rather than within the water itself, has not been studied (Sabir et al., 2012). It is a difficult research question to investigate because oxidation of Hg(0) back to Hg(I) or Hg(II) can occur at the same time and may be enhanced by the presence of water molecules or clusters (Shepler et al., 2007) and Cl or Br ions in the water (Sheu and Mason, 2004). Given the uncertainty in both reduction and oxidation chemistry on environmental surfaces, heterogeneous (i.e. air-surface) reactions of Hg are a major knowledge gap that currently limits our understanding of factors that drive Hg fluxes in the Arctic environment. Finally, although mercuric chloride (HgCl2) and mercuric bromide (HgBr2) were identified as products of the Hg + Cl and Hg + Br reactions in laboratory experiments (Ariya et al., 2002), the actual products of Hg reactions in the ambient environment have not yet been characterized.

Jun

Fig. 2. Frequency of AMDEs (bars) and mean GEM concentrations (lines) at Alert for spring months binned by local temperature and wind direction. Source: Cole and Steffen (2010).

Patterns in Hg speciation data collected over several years provide insight into the atmospheric processes that occur during AMDEs. Datasets reporting the AMDE phenomenon have been gathered for varying periods of time at several Arctic stations, including Alert (Nunavut), Barrow (Alaska), Amderma (Russia), Station Nord (Greenland), and Ny-Ålesund (Svalbard) (for a review see Steffen et al., 2008) and in the sub-Arctic at Kuujjuarapik (Quebec) (Gauchard et al., 2005; Poissant and Pilote, 2003; Steffen et al., 2005). Analysis of the Alert dataset from 1995 to 2007 revealed no significant time trend in the overall combined frequency and strength of depletion events over the AMDE season (Cole and Steffen, 2010). However, it did reveal significant changes with respect to when these AMDEs occur. Over time, the month of maximum AMDE activity shifted from May to April, with increased AMDE activity in March as well. This observation is illustrated in Fig. 1, which shows the average distribution of AMDE occurrence in each of the spring months for 1995 to 2002 and for 2003 to 2009. The reason for the shift in timing of depletion events is not yet understood, though it has been found that the frequency of depletion events and overall GEM concentrations exhibit complex relationships with local meteorology (Fig. 2). The top panel of Fig. 2 shows that the integrated frequency of depletion events—a quantity that combines the length, magnitude, and frequency of depletion events— is lower at higher temperatures within each month. This pattern is

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

7

800

4 GEM PHg RGM

3

600

GEM (ng m-3)

1 0 400

-1 -2 -3

200

-4

PHg and RGM (pg m-3)

2

-5

0

-6

2003

2004

2005

Churchill 2004

4

GEM (ng m-3)

2006

2007

2000

3 1500

2 1

1000

0 -1

500

-2

0

-3

Apr

May

Jun

2008

2009

2010

2011

2012

PHg and RGM (pg m-3)

2002

Jul

Fig. 3. Daily average GEM, PHg, and RGM concentrations at Alert (Nunavut) from 2002 to 2011. Left and right scales are equivalent (3000 pg = 3 ng) (Steffen, 2011). (Inset) Daily average GEM, PHg, and RGM concentrations at Churchill (Manitoba), March to August 2004. Left and right scales are equivalent (3000 pg = 3 ng). Data from Kirk et al. (2006b).

consistent with previous findings that the halogen chemistry initiating AMDEs and ozone depletion events may be temperature-dependent (Adams et al., 2002; Koop et al., 2000; Sander et al., 2006; Tarasick and Bottenheim, 2002). The frequency of AMDEs was not related to local wind speed but was found to have a relationship with wind direction (Fig. 2, bottom panel). The fact that air masses originating from over the Arctic Ocean in the spring are depleted of GEM more frequently than those originating elsewhere suggests that AMDEs occur over the ocean. This conclusion was also reached by researchers analyzing GEM data at Ny-Ålesund using particle dispersion modeling (Hirdman et al., 2009). Finally, it was found that March GEM concentrations at Alert correlated with the Polar/Eurasian Teleconnection and North Atlantic Oscillation indices (Climate Prediction Center, 2009; Sander and Bottenheim, 2012). These two indices are not independent, and suggest that March AMDEs are more frequent and/or intense in years when the circumpolar vortex is strong (shown by a positive Polar/Eurasia Teleconnectiion Pattern (PET) and negative North Atlantic Oscillation Index (NAO) phase). However, there was no temporal trend in those indices, nor did the correlation extend into April and May (the primary AMDE season).

2.4. Distribution of atmospheric Hg species in the Arctic The distribution of atmospheric Hg species differs in the Arctic from other regions, as shown by a time series of GEM, RGM, and PHg from 2002 to 2011 at Alert (Fig. 3) and a time series of GEM, RGM, and PHg measurements from several months in 2004 at Churchill in sub-Arctic Manitoba (Fig. 3 inset). While there is a significant difference in the length of these time series, the results from Churchill show that the distribution in the springtime reflects the same distribution seen in each spring in Alert. The seasonal cycles of GEM, RGM, and PHg can also be seen in Fig. 3 which clearly demonstrate the periodicity of AMDEs in springtime and the significant changes in Hg speciation that occur during these events. Background atmospheric Hg speciation in low- and mid-latitude rural areas is dominated by GEM which represents about 95–98% of total Hg (THg), with approximately 0.2–1% RGM and 0.2–2% PHg making up the balance (Peterson et al., 2009; Rutter et al., 2009; Schroeder and Munthe, 1998). The distribution of Hg species in urban areas is reported to be 79–99% GEM, 0.4–13% RGM, and 0.4–8% PHg (Rutter et al., 2009). Cobbett et al. (2007b) reported that during the 2005

Table 2 Speciation of atmospheric Hg between GEM, RGM and PHg in the Arctic during both the AMDE and non-AMDE seasons; data from low- and mid-latitude rural and urban areas shown for comparison. Site

GEM

RGM

PHg

Source

Alert, NU (Polar night) Alert (AMDE season) Churchill, MB (non-AMDE season) Churchill, MB (AMDE-season) Low- and mid-latitude rural areas

95% 88.6% 99.2% 71.3% 95–98%

0.6% 5.7% 0.5% 10.3% 0.2–1%

4.4% 5.7% 0.3% 18.3% 0.2–2%

Urban areas

79–99%

Cobbett et al. (2007) Steffen et al. (2014) Kirk et al. (2006) Kirk et al. (2006) Schroeder and Munthe (1998); Petersen et al. (2009); Rutter et al. (2009) Rutter et al. (2009)

0.4–13%

0.4–8%

8

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

polar night (mid-winter) at Alert, the distribution of Hg species was similar to that of southern rural latitudes (i.e. 95% GEM, 0.63% RGM, and 4.4% PHg; Table 2). In contrast, during AMDEs, the Arctic springtime atmosphere changed to 88.6%, 5.7%, and 5.7% of GEM, RGM, and PHg, respectively. At Churchill in 2004, the change in distribution was even more dramatic, with GEM, RGM, and PHg distributed on average over three non-AMDE months as 99.2%, 0.5%, and 0.3%, respectively. However, during the AMDE period of March to May, it changed to 71.3%, 10.3%, and 18.3% on average (based on 1 season of data). Although the distribution of atmospheric species during AMDEs is similar to that reported in an urban environment, it should be noted that the high spring concentrations of RGM and PHg in the Arctic are due to oxidation of GEM in the atmosphere, while high concentration levels in urban areas reflect direct industrial emissions of RGM and PHg. Furthermore, since RGM and PHg have a higher deposition velocity than GEM, the springtime shift to high concentrations of short-lived Hg species warrants investigation to determine the magnitude of net springtime atmospheric deposition of Hg to the Arctic environment. This was investigated by Steffen et al. (2014) using the Alert speciation data from 2002 to 2011; however, quantifying net deposition to the entire Arctic region requires a coupled surface-atmosphere model (Dastoor et al., this issue). The distribution of GEM, RGM, and PHg during AMDEs differs among Arctic measurement sites and the time in spring. For example, at Barrow in 2001, RGM was the dominant species during AMDEs (Lindberg et al., 2002). In Ny-Ålesund in 2003, the distribution was reported to vary throughout the study period but showed that RGM was the predominant oxidation product during AMDEs (Gauchard et al., 2005; Sprovieri et al., 2002), whereas during a more recent study at Ny-Ålesund, the distribution also varied but showed a predominance of PHg (Steen et al., 2009). Some researchers have suggested that the predominance of either PHg or RGM as the product of GEM oxidation can be an indication of the origin of the air mass (Gauchard et al., 2005; Lindberg et al., 2002) but perhaps the timing in which the sampling occurred more accurately reflects the distribution of these species. At Churchill and Alert, Kirk et al. (2006b) and Cobbett et al. (2007b), respectively, showed a predominance of PHg during AMDEs at the beginning of spring and a higher predominance of RGM in AMDEs occurring toward the end of spring (Fig. 3). This pattern has been shown to repeat each year at Alert (Steffen et al., 2014). It has been suggested that a relationship exists between the local meteorology and the distribution of Hg species during AMDEs, and a significant correlation exists between

February 400

March

April

increases in PHg and lower air temperatures and humidity (Cobbett et al., 2007). 3. Deposition of atmospheric Hg The atmospheric processes involved in the deposition of Hg during AMDEs have been the topic of a number of studies. Steffen et al. (2002) reported that not all Hg is lost from the air during AMDEs. Although GEM may be depleted in the air, depending on the wind conditions, there remains some of the other Hg species in the air. These species can be available for transport to another location. This implies that not all GEM oxidized during an AMDE is deposited onto the snowpack at the site of GEM depletion and oxidation. More recent work at Alert shows that the deposition of Hg from the air during AMDEs is dependent on the conditions of the atmosphere such as temperature, relative humidity and aerosol contribution (Cobbett et al., 2007a; Steffen et al., 2014). The authors found that higher temperatures and lower aerosol concentrations were the conditions under which the highest concentration of Hg in the snow was reported. Thus, in order to accurately understand the wet and dry deposition processes of Hg from air to snow that result from AMDEs, prevailing atmospheric conditions must be fully characterized. Wet deposition is defined as the scavenging of gaseous and particulate Hg species into atmospheric precipitation while dry deposition is defined as direct deposition to the surface. Currently, deposition of Hg during AMDEs is typically inferred from concentrations of THg in the snowpack and not primarily through direct measurements of dry or wet deposition. Direct measurements of RGM and PHg fluxes in the Arctic are scarce and very limited. During a 2004 study at Kuujjuarapik (March 9 to 16), the maximum fluxes of dry deposition directly quantified during AMDEs were − 14.0 (GEM), −2.3 (PHg) and −1.2 ng/m2/h (RGM). Maximum GEM, PHg and RGM deposition velocity were calculated to be 0.3, 1.2 and 1.6 cm/s, respectively. At Alert, snow/ice crystal samples have been collected on a Teflon snow table on an event basis during springtime since 2002 and give a reasonable estimate of Hg that is removed from the atmosphere and scavenged by snow (Steffen et al., 2014). Data from this monitoring show that depletion events are often accompanied by an increase in THg concentration in the fallen snow. Additionally, the time when higher concentrations of THg are found in the snow repeats from year to year. It has been observed in the snow long-term dataset that the most significant increase of THg in the snow generally occurs when

May

June

July 600 500 300

Mercury in Snow

300

250 200 200 150 100

100

Hg in Snow (pg g-1)

PHg and RGM (pg m-3)

PHg RGM

50 0

0 40

60

80

100

120

140

160

180

Julian days (Jan 31 to June 30) Fig. 4. Atmospheric Hg speciation data (PHg in green, RGM in pink) at Alert (Nunavut) from February to June, 2002 to 2011 and THg concentrations in snow (blue bars). Modified from Steffen et al. (2014).

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

there is a switch from PHg predominance to RGM predominance in the atmosphere (Fig. 4). These results show that the highest deposition of Hg through snowfall does not occur throughout the springtime period but rather when the atmosphere is characterized by an increase in relative humidity, air temperature greater than −6 °C (on average), and a significant decrease in particle concentration (Steffen et al., 2014). In addition, sea ice dynamics can further influence the overlying atmospheric chemistry, including AMDEs (Moore et al., 2014). These authors showed that the creation of sea ice leads aids in the replenishment of GEM from aloft and the refreezing of the sea ice creates the atmospheric chemistry conditions to initiate depletion events. These findings about specific atmospheric conditions leading to deposition may help explain a report from Ny-Ålesund which found that there were times when no increase in the concentration of THg in the snowpack was observed following AMDEs (Ferrari et al., 2005). Other direct wet deposition flux measurements were reported by Sanei et al. (2010), who employed a modified Mercury Deposition Network (MDN) precipitation collector in Churchill for over 27 months between 2006 and 2008 (Sanei et al., 2010). Contrary to other observations at Churchill (Kirk et al., 2006) and elsewhere in the Arctic, Sanei et al. (2010) reported that there is little evidence of higher deposition of Hg in the springtime during AMDEs. However, atmospheric models predict that there should be deposition of Hg during the springtime period at this location. This suggests that Hg deposition during AMDEs occurs as dry deposition and therefore might not be recorded during wet deposition monitoring in the Arctic. However, further investigation is warranted to reconcile wet deposition measurements with measurements of increased Hg concentrations in snow during and immediately following AMDEs. The deposition and fate of Hg during and after depletion events have been examined by measuring the concentration of THg in the snowpack (Poissant et al., 2008; Steffen et al., 2008b). Several researchers have reported that when the concentration of THg in the snowpack is elevated following an AMDE, there can be a 20–50% loss of that Hg within 24 h (Durnford and Dastoor, 2011) and 50–90% loss within a few days (Constant et al., 2007; Kirk et al., 2006b). This loss is likely a result of photochemical reduction of Hg(II) to GEM and emission to the atmosphere, and is temperature dependent (Constant et al., 2007; Lalonde et al., 2003; Outridge et al., 2008; Poulain et al., 2004; Sherman et al., 2010). An alternate approach that has also been employed is to measure the atmospheric fluxes of GEM above the Arctic snowpack to quantify the deposition and emission of GEM around AMDEs. In 2000, vertical gradients of GEM concentrations were measured at Alert above the snow surface during springtime. These measurements showed higher concentrations of GEM immediately above the snowpack, thereby suggesting an emission from the surface (Steffen et al., 2002). This evasion was observed for several days and throughout an extended AMDE during which GEM appeared to be emitted from the snow surface and then underwent photooxidation, which depleted it from the air again. In 2005 at Alert, Cobbett et al. (2007b) reported a net depositional flux of GEM to the snow during polar night between January and March when no AMDEs occur; however, during the spring there was no net GEM deposition to the snow surface and no evidence of net emission of GEM following AMDEs (Cobbett et al., 2007b). Significant cycling of Hg was observed between the surface and the atmosphere but an overall net flux of zero was reported throughout the entire season in 2005 (Cobbett et al., 2007). In March 2004 at Kuujjuarapik, net deposition of GEM was observed during AMDEs with no reemission thereafter; however, reemission of PHg and RGM was observed following one of the two AMDEs observed during this period of measurement. A flux chamber study in Barrow reported a total emission of GEM from the surface of 4–7% of the THg content in the snow within a 24-hour period and the authors suggested that this was an upper limit to the photoreduction of easily reduced Hg after AMDEs (Johnson et al., 2008). Similarly to the findings of Cobbett et al. (2007), Norwegian researchers in Ny-Ålesund reported net deposition of GEM during polar night; however they measured a net emission of GEM in the spring (Steen et al.,

9

2009). It should be noted that these measurements of GEM flux do not account for fluxes of RGM or PHg, which are the fractions thought to deposit more readily to the surface during AMDEs. Deposition of RGM and PHg species, which are less susceptible to reemission from the surface, would increase the overall net deposition of Hg to the surface during the AMDE season. A number of factors such as meteorological conditions, sea-ice conditions and snow chemistry parameters also play a key role in the deposition and fate of Hg in Arctic environments. For example, consolidated ice cover promotes depletion of GEM and AMDEs; however, the upwind presence of open sea-ice leads can result in convective forcing and mixing of non-depleted overlying air masses into the boundary layer (Moore et al., 2014). This mechanism provides an increased supply of GEM for oxidation during AMDEs and could result in increased deposition of RGM/PHg (Moore et al., 2014). This is consistent with model results, which show that deposition of atmospheric Hg is higher in years with higher wind speeds as a result of increased turbulent mixing of Hg-enriched air masses and thus higher deposition during AMDEs (Fisher et al., 2013). Conversely, the model showed that high air temperature, low sea-ice fraction, low cloudiness, and a shallow surface ocean mixed layer lead to lower deposition of atmospheric Hg and higher reemission through photoreduction (Fisher et al., 2013). Similarly, a negative correlation has been observed between air temperature and the occurrence of AMDEs (Berg et al., 2013). The meltwater flux of Hg from the cryosphere (a significant portion of which is atmospheric in origin) to the surface ocean is enhanced by high solar radiation, leading to increased melting, and a large air temperature difference between spring, when cold conditions favor deposition of atmospheric Hg, and summer, when warmer conditions leads to increased meltwater inputs from sea ice and rivers (Fisher et al., 2013). Climate change, which is projected to result in increased cloudiness and warming in spring, may therefore lead to decreased Hg inputs to the Arctic Ocean (Fisher et al., 2013), although how it will impact inputs from rivers is still uncertain (Fisher et al., 2012; Dastoor and Durnford, 2014). Durnford et al. (2012a) performed a statistical study relating observations of the 24-h fractional loss of mercury from surface snow, and mercury in surface snow, seasonal snowpacks, the snowpack meltwater's ionic pulse, and long-term snowpack-related records to simulated values of Hg deposition and meteorological variables. These authors found that oxidized Hg deposited through wet processes was retained preferentially by the snowpack-related media over oxidized Hg deposited through dry processes. This may be caused either by the burying of deposited mercury by fresh snowfalls (Witherow and Lyons, 2008; Dommergue et al., 2010) and/or by the more central location within a snowpack's snow grain of mercury deposited through wet versus dry processes (Seigneur et al., 1998; Douglas et al., 2008). This preferential retention within snowpacks of mercury deposited through wet processes has important implications for climate change. If, in a changing climate, precipitation patterns change, the spatial distribution of wet deposition will also change. Consequently, the spatial distributions of mercury concentrations in snowpacks and their meltwater, and also, therefore, of the transfer of mercury to the underlying surface from the meltwater, will change. The presence of halogen anions such as Br and Cl in snow can stabilize deposited Hg(II) by slowing the reduction of this Hg(II) back to Hg(0) and/or enhancing the oxidation of Hg(0) by halides (Poulain et al., 2004; St. Louis et al., 2007; Lalonde et al., 2003), resulting in lowered reemission of GEM over sea ice compared to tundra (Steffen et al. 2013). Therefore, it is likely that both deposition and retention of RGM and PHg in coastal snowpacks are enhanced compared to terrestrial snowpacks, and these are reflected in higher snow THg concentrations measured at coastal vs. inland sites (St. Louis et al., 2007; Poulain et al., 2007a). Furthermore, THg concentrations measured in snow under tree canopies are higher than those in adjacent open areas (Poulain et al., 2007b). Canopies can reduce reduction and volatilization significantly by attenuating incident solar radiation, thereby slowing the

10

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

Wet and dry deposition, canopy throughfall

GEM PHg RGM snowmelt increased reemission

top ~2 cm

stabilizers: Br -, Cl-

Reduction GEM UV-B; 305-320 nm

oxidation

top ~50 to 100 cm

diffusion ventilation continuous ubiquitous slow

episodic localised rapid

dominates in daylight RGM

snowmelt

ionic pulse

Fig. 5. A schematic of the physical and chemical processes that govern the behavior of cryospheric Hg modified from Durnford and Dastoor, 2011. Pink represents processes related to snow melt, blue represents processed related to vertical transport; reducers (red) are H2O2, HO2•, humic acids, sulphite-based compounds, and oxalic acid; oxidants (green) are H2O2, Br•, Br2, O3, OH•, alkenes, alkyl and nitrates.

rate of photoreduction of RGM in snow, increasing the fraction of Hg deposited as PHg and the concentration of oxidants in the snowpack, and finally, dampening wind ventilation of snowpack that promotes GEM emission. Furthermore, snow and ice crystal type and formation can play an important role in determining Hg concentrations in the cryosphere (Douglas et al., 2008). To understand and quantify key physical and chemical processes that influence Hg deposition and fate in Arctic snowpacks, researchers created a snow exchange model that includes atmospheric deposition (Fig. 5) (Durnford and Dastoor, 2011). To develop a model parameterization for the fate of deposited Hg in the Arctic, Durnford and Dastoor (2011) analyzed the basic mechanisms of the air-snow exchange of Hg using measurements in snow from around the world. Based on field and laboratory data and model results, they presented a conceptual mechanism of the physical and chemical processes governing the fate of deposited Hg to the snow as illustrated in Fig. 5. It is hypothesized that all GEM deposited onto snow-covered surfaces is likely revolatilized immediately—as indicated in the top left-hand corner of Fig. 5. However, the snowpack likely retains most, if not all, of the deposited PHg. The fate of RGM deposited onto the snowpack is far more complicated because it may undergo a series of chemical reactions within the snowpack, as described above and in Section 2.2. For instance, RGM can be reduced to GEM, which may then be oxidized back to RGM again. Synthesis of existing data suggests that oxidation of GEM tends to dominate over reduction of RGM in daylight. Furthermore, hydrogen peroxide acts as both a reductant (in pH neutral snow) and an oxidant (in acidic snow), and the presence of halides in the snowpack is found to stabilize the oxidized Hg in the snowpack. Emitted GEM is sourced more from surface snow than deeper snowpack layers, and GEM emission to the atmosphere increases significantly at the onset of snowmelt (Durnford and Dastoor, 2011). Finally, at the onset of snowmelt, Hg(0) emission to the atmosphere increases (Dommergue et al., 2003; Dommergue et al., 2010; Sommar et al., 2007; Brooks et al., 2008) and a considerable fraction of the snowpack's Hg(II) burden exits the snowpack in the melt water ionic pulse (Bales et al., 1990; Kuhn, 2001). Fig. 5 indicates that PHg deposited onto the snowpack exits during snowmelt (narrower pink arrow labeled “snowmelt”) whereas RGM within the snowpack exits during the ionic pulse

(wider pink arrow labeled “ionic pulse”). It is believed that Hg concentrations are higher during the ionic pulse period than during the remainder of the snowmelt period. According to a statistical study, the concentration of THg in seasonal snowpacks is strongly affected by the burial of deposited Hg with fresh snow—THg concentrations increase with increasing snow precipitation and its frequency (Durnford et al., 2012a). In contrast, deeper snowpacks apparently dilute the concentration of THg in the ionic pulse of meltwater from the snowpack. The concentration of THg in long-term cryospheric records varies by latitude possibly due to mid-latitude anthropogenic sources of Hg. Long-term cryospheric mercury concentrations also appear to decrease with an increasing occurrence of sunny, dry conditions; such conditions promote photoreduction and subsequent revolatilization, and reduce the likelihood of wet deposition and the burial of mercury by fresh snowfalls (Durnford et al., 2012a). Carignan and Sonke (2010) examined Hg deposition along the coast of Hudson Bay. They found that, not only is the snow surface subject to deposition of Hg from AMDEs, but the lichens on trees around Hudson Bay also accumulate this atmospherically deposited Hg. Although this vegetation is known to be a good indicator of atmospheric deposition of Hg (Bargagli and Barghigiani, 1991; Horvat et al., 2000; Riget et al., 2000), Hg retention by lichens has only been investigated at a few locations in the Canadian Arctic (Gamberg et al., this issue). In the study by Carignan and Sonke, the THg concentration in lichens decreased with distance from the coast of Hudson Bay and showed the same pattern as snow THg concentrations in the area (Constant et al., 2007). These findings, along with those of other researchers across the Arctic (Douglas and Sturm, 2004; St. Louis et al., 2007; Poulain et al., 2007a) show that the concentration of THg in the snow decreases with distance inland from the ocean and indicates that Hg deposition and retention are greater closer to ocean water (see also Section 4.1). While numerous studies have concentrated on the exchange of Hg between air and snow on land in coastal areas, there have been few investigations into the cycling of Hg over sea ice. Elevated levels of THg in frost flowers, hoar frost, sea snow, and seawater over the Arctic Ocean have been reported (Douglas and Sturm, 2004; Douglas et al., 2005). Furthermore, frost flowers may enhance, at least temporarily, the retention of atmospheric Hg deposited during AMDEs by providing

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

a surface on which GEM that is reemitted from the snowpack can be oxidized once more and adsorbed on to the crystal surface (Sherman et al., 2012). Future climate and sea-ice conditions are predicted to favor the growth of frost flowers and may therefore result in increased flux of atmospheric Hg to the Arctic Ocean (Sherman et al., 2012). A continued decrease in perennial ice may also favor the occurrence of bromine explosions, which play an important role in the deposition of atmospheric Hg to the Arctic (Nghiem et al., 2012). A recent publication of measurements over the Beaufort Sea showed a comparison of two snow-covered ice cores and suggested that elevated THg levels in surface snow, presumably due to atmospheric deposition, do not extend down to the sea ice until melting of the snow has begun (Chaulk et al., 2011). Modeling results by Hirdman et al. (2009) provide additional support for the hypothesis that AMDEs result in the net deposition of Hg from the atmosphere to the Arctic Ocean. The authors performed a statistical analysis on the results from a Lagrangian particle dispersion model (FLEXPART) and GEM concentrations measured at Ny-Ålesund to identify source regions of high- and low-Hg air masses. It was found that, in the spring, air masses depleted in Hg had undergone low-level transport over the ocean. However, in the summer, air masses with strong surface contact over the Arctic Ocean showed a high Hg concentration. This suggests that the Arctic Ocean is a net sink for Hg in the springtime but a net source in the summer, possibly due to emissions from the snow surface and evasion from the ocean (Hirdman et al., 2009). Durnford et al. (2012b), using an atmospheric model, found net deposition in the Arctic in both spring and summer, but with that of spring (63 Mg) 1.5-fold greater than that of summer (43 Mg). Using an ocean-atmosphere model, Fisher et al. (2012; 2013) demonstrate that the summer-time peak in GEM concentration observed in the Arctic atmosphere is a result of evasion from the Arctic Ocean, likely as a result of elevated riverine Hg inputs at this time of year. The atmospheric mercury model used by Dastoor and Durnford (2014) simulated an initial warm-season peak in atmospheric GEM concentrations in the Arctic driven by revolatilization from snowpacks and a second, later peak driven by ocean evasion. Hirdman et al. (2009) also found that in the winter, low Hg concentrations were associated with air masses from the free troposphere. This latter model analysis has not been replicated with data from the Canadian Arctic but the findings are in agreement with data reported from Alert for both Hg and ozone (Bottenheim and Chan, 2006; Cole and Steffen, 2010). The overall impact of atmospheric Hg deposition on the Arctic remains unclear, not only in Canada but across the circumpolar Arctic. A mass balance budget was produced for THg in the Arctic Ocean, and it was concluded that the contribution of the atmosphere to the Hg pool in the Arctic Ocean is less than what is derived for other oceans but is still significant overall (Outridge et al., 2008). Using data collected at

11

Barrow (Alaska), a mass balance budget was developed for the Arctic that showed a net surface gain in Hg in the springtime (Brooks et al., 2006). Early modeling estimates of net deposition of Hg to the Arctic, including deposition from AMDEs, range from 50 to 300 tonnes (t) of Hg per year (Ariya et al., 2004; Banic et al., 2003; Lu et al., 2001; Skov et al., 2004). However, more recent models which incorporate snowpack photoreduction and GEM reemission tend to yield lower depositional values ranging from 60 t yr− 1 (Holmes et al., 2010), to 143 t yr− 1 (Dastoor et al., this issue). Further, Durnford et al. (2012b) estimated that, in the Arctic, gross deposition during spring, when AMDEs are active, represents 59% of the annual gross deposition. Results from these investigations and comparison with field measurements of depositional fluxes are discussed in more detail in Dastoor et al. (this issue). It is clear from the findings described here that the cycling of Hg at the snow-air interface and within the snowpack and melt water is very dynamic, but the processes governing this cycling are not well understood. Some of these issues for marine snow are addressed in more depth in Section 4 of Braune et al. (this issue). 4. Summary of Canadian atmospheric measurement data and spatial patterns Atmospheric GEM concentrations have been monitored in Canada at approximately ten different sites since 1995. Of the ten sites, GEM has been continuously measured at only three sites between 1995 and the present. In the Canadian Arctic and sub-Arctic, the spatial coverage is even sparser. Alert, Kuujjuarapik, and Little Fox Lake (Yukon) are the only sub-arctic and Arctic sites where continuous atmospheric GEM measurements are made at present—measurements at Little Fox Lake began in 2007. Supplementary data from short campaigns (b1 year) are available for Resolute (Nunavut) (Lahoutifard et al., 2005), Churchill (Kirk et al., 2006b), and the Amundsen Gulf (Northwest Territories). Continuous monitoring data for speciated Hg (i.e. GEM, RGM, and PHg concurrently) are even more limited. Alert is the only site where speciated measurements have been collected over multiple years. Short periods of speciated Hg monitoring have been done at Kuujjuarapik (Steffen et al., 2003), Churchill (Kirk et al., 2006b), on Hudson Bay (Poissant et al., 2006) and around the Amundsen Gulf during the IPY (2007–2008). 4.1. Gaseous elemental Hg Geographic variation was observed in seasonal mean concentrations of GEM from Alert and the Amundsen Gulf in the High Arctic; Little Fox Lake, Churchill, and Kuujjuarapik in the sub-Arctic; and, for comparison, at the Alaskan Arctic site of Barrow and the high altitude temperate site

Table 3 Multi-annual medians and seasonal averages of GEM (ng m−3), RGM (pg m−3) and PHg (pg m−3) concentrations across Canada and Alaska. Data are from Cole et al. (2013), Kirk et al. (2006) and unpublished data. Site (measurement period)

Days of data/ days in period

Multi-annual median GEM

Spring GEM mean (SD)

Summer GEM mean (SD)

Fall GEM mean (SD)

Winter GEM mean (SD)

Spring RGM mean (SD)

Spring PHg mean (SD)

Alert (1995 to 2009) 82.5° N, 62.5° W Barrow (2007 to 2009) 71.3° N, 156.6° W Churchill (Mar–Aug 2004) 58.6° N, 94.2° W Kuujjuarapik (Aug 1999 to 2009) 55.3° N, 77.7° W Little Fox Lake (Jun 2007 to 2009) 61.3° N, 135.6° W Whistler (Aug 2008 to 2009) 50.1° N, 122.9°W Amundsen Gulf (Mar–May 2008) 71.1° N, 124.8° W

4629/5479

1.56

1.24 (0.53)

1.80 (0.35)

1.49 (0.11)

1.59 (0.17)

50 (72)

70 (65)

780/1017

1.13

1.06 (0.69)

1.40 (0.27)

1.06 (0.12)

1.03 (0.26)

138/150

N/A

1.26 (0.61)

1.81 (0.19)

N/A

N/A

a

Kuujjuarapik January–March 2008.

GEM trend (ng m−3 y−1) -0.009 N/A

213 (199)

342 (495)

a

a

18 (41)

N/A

3479/3778

1.63

1. 72 (0.64)

1.86 (0.58)

1.58 (0.46)

1.56 (0.43)

710/942

1.31

1.43 (0.11)

1.27 (0.12)

1.17 (0.14)

1.35 (0.14)

N/A

332/491

1.14

1.14 (0.09)

1.17 (0.17)

1.20 (0.25)

1.21 (0.21)

N/A

54/92

N/A

0.94 (0.58)

N/A

N/A

N/A

N/A

141 (75)

-0.038

12

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

of Whistler (British Columbia) (Table 3). This table reflects the available data of atmospheric Hg monitoring in the Canadian Arctic and, as well, shows the lack of existing spatial coverage. The data presented was collected and analyzed using the same protocols as described in Steffen et al (2012). The heights of the inlets vary between sampling location and are described elsewhere (Cole et al., 2013). Mercury is emitted from surfaces such as snow packs and sea ice and thus inlet heights may impact the comparability of data from several locations. Mean values for each season (winter: December to February, spring: March to May, summer: June to August, and autumn: September to November) and for the entire dataset were calculated over different time periods. These measurement data reveal that annual concentrations of GEM are lower at the western sites of Barrow, Little Fox Lake, and Whistler —overall median concentrations of 1.13, 1.31 and 1.14 ng m−3, respectively—compared to eastern sites at Alert and Kuujjuarapik—1.56 and 1.63 ng m−3, respectively (Table 3). Geographic differences were also observed in the seasonal patterns of GEM among these sites. At Alert, Kuujjuarapik, and Barrow, there were several days in the spring where the GEM concentration dropped rapidly to below 1 ng m− 3, typically used as the definition of an AMDE. These AMDEs occurred at Alert and Barrow primarily from March to May, at Kuujjuarapik from February to April, and were not observed at Little Fox Lake or Whistler. Data from one spring and summer at Churchill also showed AMDEs (Fig. 3 inset). Those sites with AMDE activity are Arctic and sub-Arctic sites that are in relatively close proximity to sea ice. The sub-Arctic site at Little Fox Lake does not experience AMDEs, which supports the hypothesis that these events occur over or near a maritime source of sea salt (Douglas and Sturm, 2004; Douglas et al., 2005). Continuous GEM measurements at Mingan (50°16′N; 64°14′W), north of the Gulf of St. Lawrence, has also revealed the occurrence of AMDEs pointing to an influence of sea salt from this marine environment (Poissant, 2000). Sites with AMDEs also experienced rapid increases in GEM concentrations in the late spring and summer that reached average daily concentrations of N5 ng m − 3 at Kuujjuarapik (Cole et al., 2013). As a result, GEM concentrations peaked in summer at these four sites as shown in Table 3. It should be noted that the data from Churchill are for spring and summer only and cannot be compared with fall averages. In addition, while AMDEs at Alert depressed the springtime average GEM concentration to the lowest seasonal value, the high GEM peaks in late spring at Barrow and Kuujjuarapik—which are not prevalent until June at the higherlatitude site of Alert—brought the springtime average concentrations up to values that are similar to or higher than fall and winter averages (Table 3). Finally, AMDEs and accompanying rapid increases in GEM resulted in high variability of both spring and summer GEM concentrations at sites where AMDEs occur—shown by the high standard deviation of seasonal measurements in Table 3. In contrast, at Little Fox Lake and Whistler, where AMDEs were not observed in the time series, the mean GEM concentrations were consistent between seasons and much less variable within each season. 4.2. Reactive gaseous Hg and airborne particulate Hg Long-term measurements of atmospheric Hg speciation are currently limited to Alert. Measurements of RGM and PHg from 2002 to 2011 are shown with GEM measurements in Figs. 3 and 4. Six months of GEM, RGM, and PHg data were also collected at Churchill (Fig. 3, inset) (Kirk et al., 2006b). Both sites reveal that RGM and/or PHg values were high in the springtime during periods when GEM is unusually low (Table 3), consistent with the conversion of GEM to RGM and/or PHg during AMDEs. However, it should be noted that currently these measurements are only operationally defined because there are no existing calibration standards and the exact chemical compounds that comprise RGM and PHg are unknown.

Therefore, data from these two sites have not been quantitatively compared. Identification of these chemical species remains an important knowledge gap. 5. Temporal trends of gaseous elemental Hg As global anthropogenic emissions of Hg continue to change, and as the Arctic undergoes dramatic environmental changes, such as shrinking permanent sea ice, the continuous monitoring of atmospheric Hg provides important information about long-term changes in the transport, chemistry, and deposition of this pollutant in the Arctic environment. Worldwide atmospheric measurements of GEM up to the early 2000s suggest that concentrations of atmospheric Hg increased from the 1970s to a peak in the 1980s and then decreased to a plateau around 1996 to 2001 (Slemr et al., 2003). Similarly, a recent reconstruction of GEM levels in firn air (trapped gases in porous ice similar to compacted snow that is less dense than ice) from Greenland indicated that GEM increased from the 1940s to the 1970s and reached a plateau around the mid-1990s (Fäin et al., 2009). Another study of long-term trends at Resolute Bay examined measurements from 1974 to 2000 of filterable Hg. That is, manual samples of Hg collected by passing air through particle filters and is likely PHg and RGM combined (Li et al., 2009). These authors reported a decrease of approximately 3% per year in total filterable Hg in summer and fall, which is similar to the world-wide decrease in Hg emissions from anthropogenic activities between 1983 and 1995. Considerable variability was found in the data during the winter and early spring months suggesting some influence of AMDEs in the samples. Continuous measurements of GEM have been made since 1995 at Alert, and for 11 years at Kuujjuarapik (1999–2009). Temporal trends in these datasets have been assessed using statistical techniques to account for large seasonal variability. The techniques included seasonal decomposition (Temme et al., 2004) and the seasonal Kendall test for trend (Cole and Steffen, 2010). While analysis of the Alert data to the end of 2004 found no significant trend in GEM air concentrations (Temme et al., 2007), subsequent analysis using additional years of data revealed that GEM decreased at a rate of -0.6% per year from 1995 to 2007 (Cole and Steffen, 2010). The seasonal Kendall test for trend and related Sen's estimator of slope were also used to calculate ten-year trends from 2000 to 2009 at Alert, Kuujjuarapik, Ny-Alesund, and three Canadian mid-latitude sites (Egbert, Ontario; Kejimkujik, Nova Scotia; St. Anicet, Quebec) (Cole et al., 2013). The time period was chosen to compare trends between multiple sites over the same time interval. Over this period, GEM concentrations at Kuujjuarapik and at the three mid-latitude sites declined at a rate of approximately −2% per year, while Alert concentrations declined by −0.9% per year (Cole et al., 2013) and no decrease was seen at Ny-Alesund (Berg et al., 2013). Trends at the four non-Arctic sites agreed well with the reported decrease in background GEM concentration at Mace Head (Ireland) of − 1.8 ± 0.1% per year over the period 1996 to 2009 (Ebinghaus et al., 2011). The long-term trends at the Arctic sites were also more variable from month to month, partially due to AMDE chemistry and likely also due to effects from changes in sea ice cover. Trends in RGM and PHg at Alert were also reported for a sevenyear period (2002–2009); both have increased in the spring when their concentrations are significant, but high interannual variability limits the reliability of these trends (Cole et al., 2013). At this time, it is not clear why atmospheric GEM levels have decreased more rapidly at sub-Arctic and temperate locations than at polar sites. It is important to determine the mechanism responsible for the more positive GEM trends in the Arctic, because the effect on the deposition of mercury to the Arctic ecosystem could be very different. For example, decreased oxidation chemistry would lead to increased spring GEM concentrations but decreased deposition to the surface, while increased transport of mercury from lower latitudes would increase both Arctic GEM concentrations and total mercury deposition. It is also

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

possible that Alert is more strongly influenced by source regions where Hg emissions are increasing (such as Asia) compared with Kuujjuarapik, Saint-Anicet, Mace Head, or Cape Point (see also Dastoor et al., this issue), or that large climate-related variability is impacting the trends (Cole and Steffen, 2010). Decreasing GEM trends mid-latitude sites such as at St-Anicet also likely reflect a small contribution from decreased Hg emissions as a result of improved emission control in industries such as coal-fired power plants in the north-eastern USA and Canada, and/or decreased emissions from the North Atlantic (Soerensen et al., 2012). Detailed modeling studies and better knowledge of natural emission trends would help resolve the reason(s) for latitude-dependent trends and also extrapolate the impact on mercury deposition in the Arctic. Also, continued observations in the Antarctic (Pfaffhuber et al., 2012; Brooks et al., 2008), where AMDEs occur but emission sources are more distant, and at other remote Northern Hemisphere locations like Little Fox Lake that do not experience AMDEs, may help separate the effects of chemistry, emissions, and transport. 6. Conclusions Significant advances have been made toward understanding the unique processes governing atmospheric Hg transport and fate in the Arctic and sources of Hg to this region. Several of these advances relate to the chemical processes of AMDEs—a phenomenon characterized by rapid depletion of GEM from the lower atmosphere through oxidative processes and subsequent deposition of this Hg on the ground or to the surfaces of particles. It was verified that Br atoms, thought to originate from marine waters, are the most significant oxidants of Hg in the Arctic atmosphere. The speciation of atmospheric Hg changes during springtime AMDEs in the Arctic as GEM is oxidized, leading to large increases in RGM and PHg concentrations. Long-term measurements of Hg species at the High Arctic monitoring station at Alert revealed that the timing of AMDEs has shifted to earlier in the year although the frequency of AMDEs has not changed. The timing of AMDEs was correlated with local temperature, time of year, and wind direction, but these factors alone did not completely account for observed trends and variability of AMDEs. Further study of the factors affecting AMDE chemistry is needed in order to predict their future contributions to Hg deposition in the Arctic. Field studies on Hg fluxes from the atmosphere to snow indicate that, while a fraction of the Hg deposited during AMDEs is retained in the snowpack, a significant portion is quickly reduced and emitted to the atmosphere. Field and modeling studies suggest that halogen compounds from the marine environment and frost flowers facilitate oxidative and stabilizing reactions with Hg that increase its retention in the snow. Snowpacks rich in halogen species are found near coastal regions, and reactions between Hg and halogens in the snow may in part explain the observation that the extent of Hg re-emission increases with distance from open ocean water. The concentration of THg in seasonal snowpacks is also affected by the burial of Hg with fresh snow. Therefore, snow accumulation of Hg may be higher in regions with higher and frequent snow precipitation. The outcome of these findings is that Hg retention and cycling in the Arctic seem to differ depending on the time of year, proximity to the ocean, and meteorology. Model derived estimates suggest that net deposition of Hg above the Arctic Circle (north of 66.5°) occurs at a rate of 143 t yr−1 (Dastoor et al., this issue). However, both model simulations and estimates derived from lake sediments (Muir et al., 2009) show that the deposition of atmospheric Hg generally declines with increasing latitude, and deposition of atmospheric Hg in Arctic and subarctic lakes is significantly lower than in mid-latitude North American lakes. The geographic coverage of air measurements in the Canadian Arctic is still sparse but now includes additional data for Little Fox Lake (Yukon), Churchill (Manitoba) and the Amundsen Gulf (Northwest Territories) to complement the long-term monitoring sites at Alert in the High Arctic and Kuujjuarapik in sub-Arctic Quebec. Annual

13

concentrations of GEM were lower at western Arctic sites of Little Fox Lake, and Barrow (Alaska) and similar to a western temperate site at Whistler (British Columbia). Rapid drops in atmospheric GEM concentration, associated with AMDEs, were observed for the coastal sites at Alert, Barrow, Churchill, and Kuujjuarapik but not for the inland site at Little Fox Lake, which supports the hypothesis that these events occur near marine sources of halogens. Air concentrations, measured as GEM, declined from 2000 to 2009 at Alert and Kuujjuarapik. The annual rate of decline at Kuujjuarapik of −2.0 ± 0.9% (95% confidence limits) was comparable to non-Arctic monitoring sites in North America and elsewhere, while the trend for Alert was lower (−0.9 ± 0.6%). A slower rate of decline at Alert may be related to the environmental conditions occurring in the High Arctic or exposure to different pathways of long-range atmospheric transport of Hg. According to modeling analyses in mid-latitude locations of North America, the decline in regional Hg emissions between 1990 and 2005 was well reflected by a decline in air concentrations and deposition. However, in the High Arctic, the model results (Dastoor et al., this issue) suggested that the Hg trends were mostly related to changes in meteorology and global changes in anthropogenic emissions. Results from modeled and measured GEM concentrations illustrate that Hg trends observed at temperate locations cannot be extrapolated to the Arctic. Although tremendous progress has been made toward understanding long-range atmospheric transport of Hg and its deposition in the Arctic, many knowledge gaps still impede our capacity to quantify the significance of this cycling for bioaccumulation of methylmercury in biota and humans. These gaps include uncertainties in the identification of oxidized Hg species in the air, physical-chemical processes in air, snow and water—especially over the sea ice—and the relationship between these processes and climate change. Greater spatial coverage of long term atmospheric speciated mercury needs to be initiated so that long term changes at locations other than Alert can be assessed. A fuller understanding of the impacts of air exchange and sea ice dynamics on the deposition of mercury needs to be addressed. The Arctic is covered by significant permafrost under the snow pack and how much mercury resides in the permafrost is unknown as well as how much will be emitted to the atmosphere as melting increases. Moving forward, these unknowns should be investigated. The Arctic spans a significant portion of the Canadian landscape. This region is undergoing drastic changes that affect the complex biogeochemical cycling of pollutants such as Hg, as described in this special issue. The changes are dynamic and can have a cascading effect on the transport, transformation, deposition and uptake of Hg in the Arctic. It is vital that future changes in anthropogenic activities, such as land use changes, resource development and transportation in the Arctic region be considered as both new sources of Hg to the region and for their impacts on the chemistry dictating the cycling of Hg in the area. Acknowledgments We would like to acknowledge the support of the Northern Contaminants Program. We also thank Heather Morrison who contributed to and edited an earlier version of this manuscript. Finally, we thank Melissa Antoniadis and Julie Narayan for their assistance with graphics design and editing. References Adams JW, Holmes NS, Crowley JN. Uptake and reaction of HOBr on frozen and dry NaCl/NaBr surfaces between 253 and 233 K. Atmos Chem Phys 2002;2:79–91. Ariya PA, Khalizov A, Gidas A. Reactions of gaseous mercury with atomic and molecular halogens: kinetics, product studies, and atmospheric implications. J Phys Chem A 2002;106:7310–20. Ariya PA, Dastoor A, Amyot M, Schroeder W, Barrie L, Anlauf K, et al. The Arctic: a sink for mercury. Tellus B Chem Phys Meteorol 2004;56:397–403. Ariya PA, Skov H, Grage MML, Goodsite ME. Chapter 4: gaseous elemental mercury in the ambient atmosphere: review of the application of theoretical calculations and experimental studies for determination of reaction coefficients and mechanisms

14

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15

with halogens and other reactants. In: Goodsite ME, Johnson MS, editors. Advances in quantum chemistry, vol 55: applications of theoretical methods to atmospheric science. San Diego: Elsevier Academic Press Inc.; 2008. p. 43–55. Ariya PA, Peterseon K, Snider G, Amyot M. Mercury chemical transformations in the gas, aqueous and heterogeneous phases: state-of-the-art science and uncertainties. In: Pirrone N, Mason RP, editors. Mercury Fate and TRansport in the Global Atmosphere. Emissions. New York: Measurements and Models. Springer; 2009. p. 459–501. Bales RC, Sommerfeld RA, Kebler DG. Ionic tracer movement through a Wyoming snowpack. Atmos Environ 1990;24:2749–58. Banic C, Beauchamp ST, Tordon RJ, Schroeder WH, Steffen A, Anlauf KA. Vertical distribution of gaseous elemental mercury in Canada. J Geophys Res 2003;108: 4264. http://dx.doi.org/10.1029/2002JD002116. Bargagli R, Barghigiani C. Lichen biomonitoring of mercury emission and deposition in mining, geothermal and volvanic areas of Italy. Environ Monit Assess 1991;16:265–75. Berg T, Pfaffhuber KA, Cole AS, Engelsen O, Steffen A. Ten-year trends in atmospheric mercury concentrations, meteorological effects and climate variables at Zeppelin, NyAlesund. Atmos Chem Phys 2013;13:6575–86. http://dx.doi.org/10.5194/acp-136575-2013. Bidleman T, Macdonald R, Stow J. Sources, occurrence, trends and pathways in the physical environment. Canadian Arctic contaminants assessment report II. Ottawa: Indian and Northern Affairs Canada; 2003 [pp. xxi +332]. Bottenheim J, Chan E. A trajectory study into the origin of spring time Arctic boundary layer ozone depletion. J Geophys Res 2006;111:D19301. http://dx.doi.org/10.1029/ 2006JD007055. Brooks SB, Saiz-Lopez A, Skov H, Lindberg SE, Plane JMC, Goodsite MEG. The mass balance of mercury in the springtime arctic environment. Geophys Res Lett 2006;33:L13812. http://dx.doi.org/10.1029/2005GL025525. Brooks S, Arimoto R, Lindberg S, Southworth G. Antarctic polar plateau snow surface conversion of deposited oxidized mercury to gaseous elemental mercury with fractional long-term burial. Atmos Environ 2008;42:2877–84. http://dx.doi.org/10.1016/j. atmosenv.2007.05.029. Carignan J, Sonke J. The effect of atmospheric mercury depletion events on the net deposition flux around Hudson Bay, Canada. Atmos Environ 2010;44:4372–9. Cavender AE, Bisenthal TA, Bottenheim JW, Shepson PB. Volatile organic compound ratios as probes of halogen atom chemistry in the Arctic. Atmos Chem Phys 2008;8: 1737–50. Chaulk A, Stern GA, Arnstrong D, Barber DG, Wang F. Mercury distribution and transport across the ocean–sea–ice–atmosphere interface on the Arctic Ocean. Environ Sci Technol 2011;45:1866–72. Climate Prediction Center. Northern Hemisphere Teleconnection Patterns. Accessed 2009. National Weather Service. Cobbett FD, Steffen A, Lawson G, Van Heyst BJ. GEM fluxes and atmospheric mercury concentrations (GEM, RGM and HgP) in the Canadian Arctic at Alert, Nunavut, Canada (February–June 2005). Atmos Environ 2007a;41:6527–43. Cobbett FD, Steffen A, Lawson G, Van Heyst BJ. GEM fluxes and atmospheric mercury concentrations (GEM, RGM and HgP) in the Canadian Arctic at Alert, Nunavut, Canada (February–June 2005). Atmos Environ 2007b;41:6527–43. Cole AS, Steffen A. Trends in long-term gaseous mercury observations in the Arctic and effects of temperature and other atmospheric conditions. Atmos Chem Phys 2010;10: 4661–72. Cole AS, Steffen A. Atmospheric concentrations of Hg in the Canadian Arctic, unpublished data. Air Quality Research Division, Environment Canada, 2011 Cole A, Steffen A, Aspmo Pfaffhuber K, Berg T, Pilote M, Tordon R, et al. Ten-year trends of atmospheric mercury in the high Arctic compared to Canadian sub-Arctic and midlatitude sites. Atmos Chem Phys 2013;13:1535–45. Constant P, Poissant L, Villemur R, Lean D. Fate of mercury and methylmercury within the snow cover at Whapmagoostui–Kuujjuarapik (Québec, Canada). J Geophys Res D Atmos 2007;112:D08309. http://dx.doi.org/10.1029/2006JD007961. Dastoor A, Durnford D. Arctic Ocean: is it a sink or a source of atmospheric mercury. Environ Sci Tech 2014;48:1707–17. Dastoor A, Davignon D, Theys N, van Roozendael M, Steffen A, Ariya P. Modeling dynamic exchange of gaseous elemental mercury at polar sunrise. Environ Sci Technol 2008; 42:5183–8. Douglas T, Sturm M, Simpson W, Blum J, Alvarez-Aviles L, Keeler G, et al. The influence of snow and ice crystal formation and accumulation on mercury deposition to the Arctic. Environ Sci Technol 2008;42:1542–51. Dommergue A, Ferrari CP, Gauchard P-A, Boutron CF, Poissant L, Pilote M, et al. The fate of mercury species in a sub-arctic snowpack during snowmelt. Geophys Res Lett 2003; 30:1621. http://dx.doi.org/10.1029/2003GL017308. Dommergue A, Bahlmann E, Ferrara R, Boutron CF. Laboratory simulation of Hg0 emissions from a snowpack. Anal Bioanal Chem 2007;388:319–27. Dommergue A, Ferrari CP, Amyot M, Brooks S, Sprovieri F, Steffen A. Spatial coverage and temporal trends of atmospheric mercury measurements in polar regions. In: Pirrone N, Mason RP, editors. Mercury fate and transport in the global atmosphere. Emissions, measurements and models. New York: Soringer; 2009. p. 293−321293 −321. Dommergue A, Larose C, Fain X, Clarisse O, Foucher D, Hintelmann H, et al. Deposition of mercury species in the Ny-Alesund area (79oN) and their transfer during snowmelt. Environ Sci Technol 2010;44:901–7. http://dx.doi.org/10.1021/es902579m. Donohoue DL, Bauer D, Hynes AJ. Temperature and pressure dependent rate coefficients for the reaction of Hg with Cl and the reaction of Cl with Cl: a pulsed laser photolysispulsed laser induced fluorescence study. J Phys Chem A 2005;109:7732–41. Donohoue DL, Bauer D, Cossairt B, Hynes AJ. Temperature and pressure dependent rate coefficients for the reaction of Hg with Br and the reaction of Br with Br: a pulsed laser photolysis-pulsed laser induced fluorescence study. J Phys Chem A 2006;110: 6623–32.

Douglas TA, Sturm M. Arctic haze, mercury and the chemical composition of snow across northwestern Alaska. Atmos Environ 2004;38:805–20. Douglas TA, Sturm M, Simpson W, Brooks S, Lindberg S, Perovich DK. Elevated mercury measured in snow and frost flowers near Arctic sea ice leads. Geophys Res Lett 2005;32:L04502. http://dx.doi.org/10.1029/2004GL022132. Durnford D, Dastoor A. The behavior of mercury in the cryosphere: a review of what we know from observations. J Geophys Res D Atmos 2011;116:D06305. http://dx.doi. org/10.1029/2010JD014809. Durnford D, Dastoor A, Figueras-Nieto D, Ryjkov A. Long range transport of mercury to the Arctic and across Canada. Atmos Chem Phys 2010;10:6063–86. Durnford DA, Dastoor AP, Steen AO, Berg T, Ryzhkov A, Figueras-Nieto D, et al. How relevant is the deposition of mercury onto snowpacks? Part 1: A statistical study on the impact of environmental factors. Atmos Chem Phys 2012a;12:9221–49. http://dx.doi. org/10.5194/acp-12-9221- 2012. Durnford D, Dastoor A, Ryzhkov A, Poissant L, Pilote M, Figueras-Nieto D. How relevant is the deposition of mercury onto snowpacks? — part 2: a modeling study. Atmos Chem Phys 2012b;12:2647–706. Ebinghaus R, Jennings SG, Kock HH, Derwent RG, Manning AJ, Spain TG. Decreasing trends in total gaseous mercury in baseline air at Mace Head, Ireland, from 1996 to 2009. Atmos Environ 2011;45:3475–80. Fäin X, Ferrari CP, Dommergue A, Albert MR, Battle M, Severinhaus J, et al. Polar firn air reveals large-scale impact of anthropogenic mercury emissions during the 1970s. Proc Natl Acad Sci U S A 2009;106:16114–9. Ferrari CP, Gauchard PA, Dommergue A, Magand O, Nagorski S, Boutron CF, et al. Snow to air exchange of mercury in an Arctic seasonal snow pack in Ny-Ålesund, Svalbard. Atmos Environ 2005;39:7633–45. Fisher JA, Jacob DJ, Soerensen AL, Amos HM, Steffen A, Sunderland EM. Riverine source of Arctic Ocean mercury inferred from atmospheric observations. Nat Geosci 2012;5: 499–504. http://dx.doi.org/10.1038/ngeo1478. Fisher JA, Jacob DJ, Soerensen AL, Amos HM, Corbitt ES, Streets D, et al. Factors driving mercury variability in the Arctic atmosphere and ocean over the past 30 years. Global Biogeochem Cycles 2013;27:1226–35. http://dx.doi.org/10.1002/2013GB004689. Gauchard PA, Aspmo K, Temme C, Steffen A, Ferrari CP, Berg T, et al. Study of the origin of atmospheric mercury depletion events recorded in Ny-Ålesund, Svalbard, spring 2003. Atmos Environ 2005;39:7620–32. Goodsite ME, Plane JMC, Skov H. A theoretical study of the oxidation of Hg-0 to HgBr2 in the troposphere. Environ Sci Technol 2004;38:1772–6. Gustin MS, Huang J, Miller MB, Petersen C, Jaffe DA, Ambrose J, et al. Do we understand what the mercury speciation instruments are actually measuring? Results of RAMIX. Environ Sci Tech 2013;47:7295–306. Hirdman D, Aspmo K, Burkhart JF, Eckhardt S, Sodemann H, Stohl A. Transport of mercury in the Arctic atmosphere: evidence for a springtime net sink and summer-time source. Geophys Res Lett 2009;36:L12814. http://dx.doi.org/10.1029/2009GL038345. Holmes CD, Jacob DJ, Yang X. Global lifetime of elemental mercury against oxidation by atomic bromine in the free troposphere. Geophys Res Lett 2006;33:L20808. http://dx.doi.org/10.1029/2006GL027176. Holmes CD, Jacob DJ, Corbitt ES, Mao J, Yang X, Talbot R, et al. Global atmospheric model for mercury including oxidation by bromine atoms. Atmos Chem Phys Discuss 2010; 10:19845–900. Horvat M, Jeran Z, Spiric Z, Jacimovic R, Miklavcic V. Mercury and other elements in lichens near the INA Naftaplin gas treatment plant, Molve, Croatia. J Environ Monit 2000;2:139–44. Johnson KP, Blum JD, Keeler GJ, Douglas TA. Investigation of the deposition and emission of mercury in arctic snow during an atmospheric mercury depletion event. J Geophys Res D Atmos 2008;113:D17304. http://dx.doi.org/10.1029/2008JD009893. Khalizov AF, Viswanathan B, Larregaray P, Ariya PA. A theoretical study on the reactions of Hg with halogens: atmospheric implications. J Phys Chem A 2003; 107:6360–5. Kirk JL, St. Louis VL, Sharp MJ. Rapid reduction and reemission of mercury deposited into snow packs during atmospheric mercury depletion events at Churchill, Manitoba, Canada. Environ Sci Technol 2006a;40:7590–6. Kirk JL, St. Louis VL, Sharp MJ. Rapid reduction and reemission of mercury deposited into snowpacks during atmospheric mercury depletion events at Churchill, Manitoba, Canada. Environ Sci Technol 2006b;40:7590–6. Kirk JL, Lehnherr I, Andersson M, Braune B, Chan L, Dastoor A, et al. Mercury in Arctic marine ecosystems: sources, paths and exposure. Environ Res 2012;119:64–87. Koop T, Kapilashrami A, Molina LT, Molina MJ. Phase transitions of sea-salt/water mixtures at low temperatures: implications for ozone chemistry in the polar marine boundary layer. J Geophys Res D Atmos 2000;105:26393–402. http://dx.doi.org/10. 1029/2000JD900413. Kos G, Ryzhkov A, Dastoor A, Narayan J, Steffen A, Ariya PA, et al. Evaluation of discrepancy between measured and modeled oxidized mercury species. Atmos Chem Phys 2013;13:4839–63. Kuhn M. The nutrient cycle through snow and ice, a review. Aquat Sci 2001;63:150–67. Lahoutifard N, Sparling M, Lean D. Total and methyl mercury patterns in Arctic snow during springtime at Resolute, Nunavut, Canada. Atmos Environ 2005;39: 7597–606. Lalonde JD, Poulain AJ, Amyot M. The role of mercury redox reactions in snow on snowto-air mercury transfer. Environ Sci Technol 2002;36:174–8. Lalonde JD, Amyot M, Doyon M-R, Auclair J-C. Photo-induced Hg(II) reduction in snow from the remote and temperate Experimental Lakes Area (Ontario, Canada). J Geophys Res D Atmos 2003;108:4200. http://dx.doi.org/10.1029/2001JD001534. Lehnherr I. Methylmercury biogeochemistry: a review with special reference to Arctic aquatic ecosystems. Environ Rev 2014;22:229–43. Li C, Cornett J, Willie S, Lam J. Mercury in Arctic air: the long-term trend. Sci Total Environ 2009;407:2756–9.

A. Steffen et al. / Science of the Total Environment 509–510 (2015) 3–15 Lindberg SE, Brooks S, Lin C-J, Scott KJ, Landis MS, Stevens RK, et al. Dynamic oxidation of gaseous mercury in the Arctic troposphere at polar sunrise. Environ Sci Technol 2002; 36:1245–56. Lu JY, Schroeder WH, Barrie LA, Steffen A, Welch HE, Martin K, et al. Magnification of atmospheric mercury deposition to polar regions in springtime: the link to tropospheric ozone depletion chemistry. Geophys Res Lett 2001;28:3219–22. http://dx.doi.org/ 10.1029/2000GL012603. Lyman SN, Jaffe DA, Gustin MS. Release of mercury halides from KCl denuders in the presence of ozone. Atmos Chem Phys 2010;10:8197–204. Maron L, Dommergue A, Ferrari C, Delacour-Larose M, Faïn X. How elementary mercury reacts in the presence of halogen radicals and/or halogen anions: a DFT investigation. Chem A Eur J 2008;14:8322–9. Moore CW, Obrist D, Steffen A, Staebler RM, Richter A, et al. Sea Ice Lead-induced Convective Forcing of Mercury and Ozone in the Arctic Boundary Layer Nature; 2014. http:// dx.doi.org/10.1038/nature12924. Muir DCG, Wang X, Yang F, Nguyen N, Jackson TA, Evans MS, et al. Spatial trends and historical deposition of mercury in eastern and northern Canada inferred from lake sediment cores. Environ Sci Technol 2009;43:4802–9. Nguyen HT, Kim KH, Shon ZH, Hong S. A review of atmospheric mercury in the polar environment. Crit Rev Environ Sci Technol 2009;39:552–84. Nghiem S, Rigor I, Richter A, Burrows JP, Shepson PB, Bottenheim J, et al. Field and satellite observations of the formation and distribution of Arctic atmospheric bromine above a rejuvenated sea ice cover. J Geophys Res D: Atmos 2012;117:D00S05. http://dx.doi. org/10.1029/2011JD016268. Outridge PM, Macdonald RW, Wang F, Stern GA, Dastoor AP. A mass balance inventory of mercury in the Arctic Ocean. Environ Chem 2008;5:89–111. Pacyna EG, Pacyna JM, Steenhuisen F, Wilson S. Global anthropogenic mercury emission inventory for 2000. Atmos Environ 2006;40:4048–63. Pal B, Ariya PA. Studies of ozone initiated reactions of gaseous mercury: kinetics, product studies, and atmospheric implications. J Phys Chem Chem Phys 2004;6:572–9. Peterson C, Gustin M, Lyman S. Atmospheric mercury concentrations and speciation measured from 2004 to 2007 in Reno, Nevada, USA. Atmos Environ 2009;43:4646–54. Pfaffhuber KA, Berg T, Hirdman D, Stohl A. Atmospheric mercury observations from Antarctica: seasonal variation and source and sink region calculations. Atmos Chem Phys 2012;12:3241–51. Poissant L. Total gasesous mercury in Québec (Canada) in 1998. Sci Total Environ 2000;259: 191–201. Poissant L, Pilote M. Time series analysis of atmospheric mercury in Kuujjuarapik/ Whapmagoostui (Quebec). J Phys IV 2003;107:1079–82. Poissant L, Pilote M, Constant P, Aulagnier F, Zhang H. Mercury in air, water and sediments across the Canadian Arctic: preliminary results from ArcticNet's Amundsen icebreaker cruise in 2005. 8th International Conference on Mercury as a Global Pollutant; 2006. [Madison, USA]. Poissant L, Zhang HH, Canário J, Constant P. Critical review of mercury fates and contamination in the Arctic tundra ecosystem. Sci Total Environ 2008;400:173–211. Poulain AJ, Lalonde JD, Amyot M, Shead JA, Raofie F, Ariya PA. Redox transformations of mercury in an Arctic snowpack at springtime. Atmos Environ 2004;38:6763–74. Poulain AJ, Roy V, Amyot M. Influence of temperate mixed and deciduous tree covers on Hg concentrations and photoredox transformations in snow. Geochim Cosmochim Acta 2007a;71:2448–62. Poulain AJ, Garcia E, Amyot JD, Campbell PGC, Raofie F, Ariya PA. Biological and chemical redox transformations of mercury in fresh and salt waters of the high Arctic during spring and summer. Environ Sci Technol 2007b;41:1883–8. Raofie F, Ariya PA. Reactions of BrO with mercury: kinetic studies. J Phys IV 2003;107: 1119–21. Riget F, Asmund G, Aastrup P. The use of lichen (Cetraria nivalis) and moss (Rhacomitrium lanuginosum) as monitors for atmospheric deposition in Greenland. Sci Total Environ 2000;245:137–48. Rutter AP, Snyder DC, Stone EA, Schauer JJ, Gonzalez-Abraham R, Molina LT, et al. In situ measurements of speciated atmospheric mercury and the identification of source regions in the Mexico City Metropolitan Area. Atmos Chem Phys 2009;9:207–20. Sabir M, Ariya PA, Dastoor AP. A review of uncertainties in atmospheric modeling of mercury chemistry I. Uncertainties in existing kinetic parameters — fundamental limitations and the importance of heterogeneous chemistry. Atmos Environ 2011;45:5664–76. Sabir M, Ariya PA, Dastoor AP. A review of the sources of uncertainties in atmospheric mercury modeling II. Mercury surface and heterogeneous chemistry — a missing link. Atmos Environ 2012;46:1–10. Sander R, Bottenheim J. A compilation of tropospheric measurements of gas-phase and aerosol chemistry in polar regions. Earth Syst Sci Data 2012;4:215–82. Sander R, Burrows J, Kaleschke L. Carbonate precipitation in brine — a potential trigger for tropospheric ozone depletion events. Atmos Chem Phys 2006;6:4653–8. Sanei H, Outridge PM, Goodarzi F, Wang F, Armstrong D, Warren K, et al. Wet deposition mercury fluxes in the Canadian sub-Arctic and southern Alberta, measured using an automated precipitation collector adapted to cold regions. Atmos Environ 2010;44: 1672–81. Schroeder WH, Munthe J. Atmospheric mercury — an overview. Atmos Environ 1998;32: 809–22. Schroeder WH, Anlauf KG, Barrie LA, Lu JY, Steffen A, Schneeberger DR, et al. Arctic springtime depletion of mercury. Nature 1998;394:331–2. Seigneur C, Abeck H, Chia G, Reinhard M, Bloom NS, Prestbo E, et al. Mercury adsorption to elemental carbon (soot) particles and atmospheric particulate matter. Atmos Environ 1998;32:2649–57. Shepler BC, Balabanov NB, Peterson K. Ab initio thermochemistry involving heavy atoms: an investigation of the reactions Hg + IX (X = I, Br, Cl, O). J Phys Chem A 2005;109: 10363–72.

15

Shepler BC, Balabanov NB, Peterson KA. Hg + Br → HgBr recombination and collisioninduced dissociation dynamics. J Chem Phys 2007;127:164304. http://dx.doi.org/10. 1063/1.2777142. Sherman LS, Blum JD, Johnson KP, Keeler GJ, Barres JA, Douglas TA. Mass-independent fractionation of mercury isotopes in Arctic snow driven by sunlight. Nat Geosci 2010;3:173–7. Sherman LS, Blum JD, Douglas TA, Steffen A. Frost flowers growing in the Arctic ocean-atmosphere-sea ice-snow interface: 2. Mercury exhange between the atmopshere, snow and frost flowers. J Geophys Res D: Atmos 2012;117:D00R10. Sheu G-R, Mason RP. An examination of the oxidation of elemental mercury in the presence of halide surfaces. J Atmos Chem 2004;48:107–30. Si L, Ariya P. Reduction of oxidized mercury species by dicarboxylic acids (C 2–C4): Kinetic and product studies. Environ Sci Technol 2008;42:5150–5. http://dx.doi.org/10. 1021/es800552z. Si L, Ariya P. Aqueous photoreduction of oxidized mercury species in presence of selected alkanethiols. Chemosphere 2011;84:1079–84. http://dx.doi.org/10.1016/j. chemosphere.2011.04.061. Simpson W, Von Glasow R, Riedel K, Anderson P, Ariya PA, Bottenheim J, et al. Halogens and their role in polar boundary-layer ozone depletion. Atmos Chem Phys 2007;7:4375–418. Skov H, Christensen JH, Heidam NZ, Jensen B, Wahlin P, Geernaert G. Fate of elemental mercury in the Arctic during atmospheric depletion episodes and the load of atmospheric mercury to the Arctic. Environ Sci Technol 2004;38:2373–82. Slemr F, Brunke E, Ebinghaus R, Temme C, Munthe J, Wängberg I, et al. Worldwide trend of atmospheric mercury since 1977. Geophys Res Lett 2003;30:1516. http://dx.doi. org/10.1029/2003GL016954. Sprovieri F, Pirrone N, Hedgecock IM, Landis MS, Stevens RK. Intensive atmospheric mercury measurements at Terra Nova Bay in Antarctica during November and December 2000. J Geophys Res D Atmos 2002;107:4722. http://dx.doi.org/10.1029/2002JD002057. St. Louis V, Hintlemann H, Graydon JA, Kirk JL, Barkar J, Dimock B, et al. Methylated species in the Canadian high arctic marine surface waters and snowpacks. Environ Sci Technol 2007;41:6433–41. Steen AO, Berg T, Dastoor AP, Durnford DA, Hole LR, Pfaffhuber KA. Dynamic exchange of gaseous elemental mercury during polar night and day. Atmos Environ 2009;43:5604–10. Steffen A. Atmospheric concentrations of Hg in the Canadian Arctic, unpublished data. Air Quality Research Division, Environment Canada, 2011 Steffen A, Schroeder WH, Bottenheim J, Narayan J, Fuentes JD. Atmospheric mercury concentrations: measurements and profiles near snow and ice surfaces in the Canadian Arctic during Alert 2000. Atmos Environ 2002;36:2653–61. Steffen A, Schroeder WH, Poissant L, MacDonald R. Mercury in the Arctic atmosphere. In: Bidleman T, MacDonald R, Stow J, editors. Canadian Arctic contaminants assessment report II. Ottawa: Indian and Northern Affairs Canada; 2003. p. 124–42. Steffen A, Schroeder WH, Macdonald R, Poissant L, Konoplev A. Mercury in the arctic atmosphere: an analysis of eight years of measurements of GEM at Alert (Canada) and a comparison with observations at Amderma (Russia) and Kuujjuarapik (Canada). Sci Total Environ 2005;342:185–98. Steffen A, Douglas T, Amyot M, Ariya P, Aspmo K, Berg T, et al. A synthesis of atmospheric mercury depletion event chemistry in the atmosphere and snow. Atmos Chem Phys 2008a;8:1445–82. Steffen A, Douglas T, Amyot M, Ariya P, Aspmo K, Berg T, et al. A synthesis of atmospheric mercury depletion event chemistry in the atmosphere and snow. Atmos Chem Phys 2008b;8:1445–82. Steffen A, Scherz T, Olson ML, Gay DA, Blanchard P. A comparison of data quality control protocols for atmospheric mercury speciation measurements. J Environ Monit 2012; 14:752–65. Steffen A, Bottenheim J, Cole A, Douglas TA, Ebinghaus R, Friess U, et al. Atmospheric mercury over sea ice during the OASIS-2009 campaign. Atmos Chem Phys 2013;13: 7007–21. http://dx.doi.org/10.5194/acp-13-7007-2013. Steffen A, Bottenheim J, Cole A, Ebinghaus R, Lawson G, Leaitch WR. Atmospheric mercury speciation and mercury in snow over time at Alert, Canada. Atmos Chem Phys 2014; 14:2219–31. Stephens CR, Shepson PB, Steffen A, Bottenheim JW, Liao J, Huey LG, et al. The relative importance of chlorine and bromine radicals in the oxidation of atmospheric mercury at Barrow, Alaska. J Geophys Res D Atmos 2012;117:D00R11. http://dx.doi.org/10.1029/ 2011JD016649. Soerensen AL, Jacob DJ, Streets DG, Witt MLI, Ebinghaus R, Mason RP, et al. Multi-decadal decline of mercury in the North Atlantic atmosphere explained by changing subsurface seawater concentrations. Geophys Res Lett 2012;39. L21810. Sommar J, Wangberg L, Berg T, Gardfeldt K, Munthe J, Richter A, et al. Circumpolar transport and air-surface exchange of atmospheric mercury at Ny-A°lesund (79° N), Svalbard, spring 2002. Atmos Chem Phys. 2007;7:151–66. Subir M, Ariya PA, Dastoor AP. A review of the sources of uncertainties in atmospheric mercury modeling II. Mercury surface and heterogeneous chemistry — a missing link. Atmos Environ 2012;46:1–10. Tarasick DW, Bottenheim JW. Surface ozone depletion episodes in the Arctic and Antarctic from historical ozonesonde records. Atmos Chem Phys 2002;2:197–205. Temme C, Ebinghaus R, Einax JW, Steffen A, Schroeder WH. Application of time series analysis on long-time data sets of atmospheric mercury concentrations at two different sites. Anal Bioanal Chem 2004;380:493–501. Temme C, Blanchard P, Steffen A, Beauchamp ST, Poissant L, Tordon RJ, et al. Trend, seasonal and multivariate analysis study of total gaseous mercury data from the Canadian Atmospheric Mercury Measurement Network (CAMNet). Atmos Environ 2007;41:5423–41. Witherow RA, Lyons WB. Mercury deposition in a polar desert ecosystem. Environ Sci Technol 2008;42:4710–6.

Atmospheric mercury in the Canadian Arctic. Part I: a review of recent field measurements.

Long-range atmospheric transport and deposition are important sources of mercury (Hg) to Arctic aquatic and terrestrial ecosystems. We review here rec...
2MB Sizes 0 Downloads 3 Views