Spatial-temporal distributions of gaseous element mercury and particulate mercury in the Asian marine boundary layer

Atmospheric Environment 126 (2016) 107e116

Contents lists available at ScienceDirect

Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv

Spatial-temporal distributions of gaseous element mercury and particulate mercury in the Asian marine boundary layer Chunjie Wang a, b, Zhangwei Wang a, *, Zhijia Ci a, Xiaoshan Zhang a, Xiong Tang a, b a b

Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, No. 18 Shuangqing Road, Beijing 100085, China Graduate School of Chinese Academy of Sciences, Beijing 100049, China

h i g h l i g h t s
Spatial-temporal distributions of GEM and HgP in the Asian MBL were discussed.
Size distributions of HgP were bi-modal in spring and uni-modal in fall.
The coarse mode was dominant in spring while the fine mode was dominant in fall.
The coarse particles contributed more than 90% to the total dry deposition of HgP.

a r t i c l e i n f o

a b s t r a c t :

Article history: Received 2 September 2015 Received in revised form 12 November 2015 Accepted 16 November 2015 Available online 2 December 2015

We determined the concentrations of gaseous element mercury (GEM) and particulate mercury (HgP) in the Asian marine boundary layer (MBL) during the spring and fall of 2013 and 2014 to investigate the spatial-temporal distributions of GEM and HgP. A cascade impactor was used to collect HgP in nine size fractions ranging from 10 mm to <0.4 mm. The concentrations of HgP in PM10 (hereafter referred to as HgP 10) tended to decrease from the land to the open sea both in spring and fall. The mean (±SD) concentrations of HgP 10 during spring and fall were 15.3 ± 9.1 and 15.8 ± 4.4 pg m3 respectively, while the mean GEM concentration during the entire study period was 2.02 ± 1.08 ng m3 (N ¼ 12,341), which was much higher than those of other remote oceans. Moreover, the size distributions of HgP was bi-modal during spring, and HgP was found mainly (57%) in coarse fractions (2.1e10 mm), while HgP was dominated by fine particles (<2.1 mm) during fall. The concentrations of GEM and HgP 10 in the Bohai Sea (BS) were generally higher than those in the Yellow Sea and East China Sea. Furthermore, the HgP 10 concentrations were slightly higher during fall than during spring except the data measured in the BS for its specific location. The average dry deposition fluxes of HgP were calculated to be 2.77 ng m2 d1 during spring and 1.92 ng m2 d1 during fall, respectively, which were comparable to those measured at rural sites in North America, but considerably lower than those measured in urban cities in China. Additionally, compared to fine particles, coarse particles contributed more than 90% to the total dry deposition of HgP due to higher deposition velocities. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Gaseous element mercury Size-fractionated particulate mercury Asian marine boundary layer Dry deposition

1. Introduction Mercury (Hg) is considered to be a global persistent pollutant due to its ability to undergo long range transport in the atmosphere (Schroeder and Munthe, 1998). Moreover, the bioaccumulation and biomagnification of methylmercury in the food chain is a primary ecotoxicological concern related to Hg in the global environment

* Corresponding author. E-mail address: [email protected] (Z. Wang). http://dx.doi.org/10.1016/j.atmosenv.2015.11.036 1352-2310/© 2015 Elsevier Ltd. All rights reserved.

(Ci et al., 2011c). Hg is emitted into the atmosphere from various natural and anthropogenic sources (Pacyna et al., 2003, 2006, 2010; Schroeder and Munthe, 1998; Zhang et al., 2015). The fate and transport of Hg released into the atmosphere are determined by its physical and chemical properties and transformation processes. Atmospheric Hg exists mainly in three operationally defined forms, including gaseous elemental Hg (GEM or Hg0), reactive gaseous Hg (RGM ¼ HgCl2 þ HgBr2 þ HgBrOH þ …), and particulate Hg (HgP) (Choi et al., 2008; Gabriel et al., 2005; Schroeder and Munthe, 1998). The ratios of GEM, RGM, and HgP to total atmospheric Hg (THga ¼ GEM þ RGM þ HgP) vary geographically depending on

108

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

different land surface types, chemical environments, and human influences. However, very little is known about HgP in comparison to GEM and RGM, and even less is known about the sizefractionated HgP in the atmosphere. Generally, HgP constitutes a small percentage of THga at rural sites (Feddersen et al., 2012; Kim et al., 2012; Mao and Talbot, 2012; Xu et al., 2013), but it contributes a significant portion of the deposition of Hg to the land and ocean. Previous studies in China, Japan, Europe, and North America showed that the dry deposition of HgP was similar in magnitude to the wet deposition of Hg (Gencarelli et al., 2014; Landis and Keeler, 2002; Sakata and Marumoto, 2005; Zhu et al., 2014). HgP may actually play a disproportionately large role in the deposition of atmospheric Hg. In recent decades, research on the sources, transport, and deposition of HgP has gained increasing attention as it is critical in fully understanding the behavior and cycling of Hg in the environment. Although the size distribution of HgP has been reported in many studies, most of them concentrated on urban or rural sites, for example, the HgP measurements at urban and suburban sites in Beijing showed that the highest Hg concentration was found in the size fraction less than 1.1 mm (Wang et al., 2006). Tsai et al. (2003) reported that the HgP in PM10 (hereafter referred to as HgP 10) measured in Tainan showed regular daily variation with the higher values at daytime and lower values at nighttime, and there was more HgP in fine particles (<2.5 mm) than in coarse particles (2.5e10 mm) (generally > 70%). Xiu et al. (2005) studied the sizefractionated HgP in Shanghai and found that HgP mainly concentrated on fine particles and approximately 50e60% of Hg in PM8 concentrated on PM1.6. Nine and ten size fractions of HgP were collected by Feddersen et al. (2012), Kim et al. (2012), Xu et al. (2013), and Zhu et al. (2014) to evaluate the dominant fractions and variability of HgP in North America, South Korea, and China, respectively, where the size distribution of HgP changes due to different meteorological conditions, physical and chemical processes (e.g., adsorption and nucleation), and sources of ambient particles. Atmospheric emissions of Hg from East Asia are much higher than those from other continents in global emission inventories (Pacyna et al., 2006, 2010). China is the largest contributor to global atmospheric Hg, and anthropogenic Hg emissions are likely to further increase with the expansion of nonferrous production and coal combustion (Streets et al., 2005; Wu et al., 2006; Zhang et al., 2015). Several atmospheric GEM measurements had been conducted in the downwind regions and sites of East Asia, including the Yellow Sea (YS), East China Sea (ECS), and some coastal sites of the YS and ECS (Ci et al., 2011a, b; 2015; Friedli et al., 2004; Jaffe et al., 2005; Nguyen et al., 2007, 2010; Xia et al., 2010). However, there is little documentation on the GEM levels in the Bohai Sea (BS) and ECS. Besides, little equivalent information on the spatial and size distributions of HgP in the Asian marine boundary layer (MBL) is available. The descriptions of source, spatial and size distributions, and removal processes of HgP in the MBL are the prerequisite to assess its impact on marine environment. Additionally, atmospheric deposition is the main source of Hg to the open ocean and plays a large role in determining the pool of HgII available for reduction (Mason and Sheu, 2002). Therefore, this paper aims to identify the spatial distributions of GEM and HgP in the Asian MBL, compare the seasonal variability of GEM and HgP, and estimate the dry deposition flux of HgP based on the concentrations of each sizefractioned HgP. To our knowledge, this is the first comprehensive study of the size distributions of atmospheric HgP in the Asian MBL. Through the data presented here, we hope to provide fundamental data for the further research on GEM and HgP cycling in the marine environment.

2. Materials and methods 2.1. Study area Fig. 1 shows the sampling regions, including the majority areas of the BS, YS, and ECS. The BS is an inner sea, while the YS and ECS are semi-enclosed marginal seas. In addition, the sampling regions are located in the downwind of the East Asia, which contributed about half to the global Hg emission from anthropogenic sources (Pacyna et al., 2006, 2010), thus the study regions may be significantly influenced by the anthropogenic emissions of Hg (e.g., fossil fuel combustion, nonferrous metals smelting, and cement production) from China, Korean Peninsula, and Japan (Pacyna et al., 2006, 2010; Wu et al., 2006; Zhang et al., 2015). 2.2. Sampling and analytical methods 2.2.1. Atmospheric HgP measurements Two oceanographic cruises onboard the R/V Kexue III were conducted in the ECS during the spring (21 June to 19 July) and fall (27 October to 18 November) of 2013. An Andersen cascade impactor was deployed on the top deck of the R/V at a height of 15.5 m above the sea level (a.s.l.). Another two campaigns onboard the R/V Dongfanghong II were carried out in the YS and BS during the spring (27 April to 20 May) and fall (5e24 November) of 2014. The impactor was mounted on the top of the R/V at a height of 17.0 m a.s.l.. To reduce the contamination from exhaust plume of the ships as little as possible, we stopped sampling when R/V arrived at sampling stations. The Andersen impactor has been widely used to collect HgP in the atmosphere (Feddersen et al., 2012; Kim et al., 2012; Zhu et al.,

Fig. 1. The location of the sampling regions (Bohai Sea, Yellow Sea, and East China Sea) during the spring and fall cruises of 2013 and 2014.

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

2014), which has nine-stage, and the particle cut-off diameters are 10.0, 9.0, 5.8, 4.7, 3.3, 2.1, 1.1, 0.7, and 0.4 mm, respectively, particles less than 0.4 mm were collected on the back-up filter. The sampler was operated at a flow rate of 28.3 l min1 to maintain the maximum efficiency. Generally, sample collection began at daytime and continued for 2e4 days (Zhu et al., 2014), the details of each set are listed in Table 1. A total of four sets of field blank samples (two sets in the ECS and two sets in the YS and BS) were collected. For the sake of simplicity, the sampling time that from 22 June to 20 July in 2013 was referred to as spring. Quartz filter had been widely used to collect HgP in the atmosphere (Feddersen et al., 2012; Kim et al., 2012; Landis et al., 2002; Liu et al., 2011; Xu et al., 2013). Therefore, the 81 mm diameter quartz filters (Whatman) were used as impaction surfaces in this study, which were pre-cleaned at 900  C for 3 h to remove any Hg present on them, and then were placed in clean petri-dishes, tightly sealed until used in sampling. Field blanks for HgP were collected by placing nine unused pre-cleaned quartz filters in the impactor for five minutes without drawing air through the system. After that, the filters were packed using filter boxes and sealed in plastic bags and then immediately preserved in a refrigerator at 20  C until the end of the cruises. Particle-free gloves were always worn when handing the samples in the laboratory as well as in the field. Analysis procedures for HgP described by Kim et al. (2009, 2012) and Lu et al. (1998) are followed. Before analysis, each filter was conditioned in a desiccator for more than 24 h. The quartz filters were thermally desorbed in a tube furnace using ultra zero air to 900  C, the heated air was transported into a gold trap, and then the Hg content in gold trap was quantified using the dualamalgamation CVAFS method (Fitzgerald and Gill, 1979). It should be noted that a soda-lime (Ultra-trace Hg grade) trap (Teflon, 12 cm in length and 1 cm in inner diameter) was used to protect the gold traps from being passivated by non-Hg compound(s). The average blank value obtained in this study was <5 pg Hg per filter (equivalent of <0.04 pg m3 for a 3-day sample). The detection limit was 1.49 pg m3 based on 3 times the standard deviation of field blanks (n ¼ 6). The field blank was subtracted from the samples, which contained 6e560 times more Hg. 2.2.2. Atmospheric GEM measurements GEM in the marine atmosphere was continuously measured using an automatic dual channel, single amalgamation cold vapor atomic fluorescence analyzer (Model 2537B, Tekran, Inc., Toronto, Canada) with a Teflon filter (47 mm in diameter and 0.2 mm in pore size) and a soda-lime trap (two pieces of pre-clean fluffy quartz wool were placed in the both ends of the soda lime trap) just upstream to remove the atmospheric particulates and water, and the soda-lime was changed every 3 days during the cruises. In this study, the sampling intervals were 5 min and the sampling flow

109

rate was 1.5 l min1. To maintain high standard measurements, the instrument was calibrated every 25 h using the internal calibration source and these calibrations were checked against manual injections of known quantities of Hg0 before and after the cruises, and the accuracy was estimated at 95%. The detection limit of GEM in this study was less than 0.1 ng m3. To diminish the contamination from domestic waste and exhaust plume of the ship as possible, we installed the Tekran system inside the ship laboratory and mounted the sampling inlet at the front deck 2 m above the top deck using a polytetrafluoroethylene (PTFE) tube (¼ inch in outer diameter and 23 m in length). 2.2.3. Back-trajectory model To identify the long range transport of GEM and investigate the influence of air masses movements on the GEM levels, we calculated 72-h back-trajectories of air masses using the HYSPLIT model (Draxler and Rolph, 2012) and TrajStat software (Wang et al., 2009) based Geographic Information System (GIS). Global Data Assimilation System (GDAS) meteorological dataset (spatial resolution: 1 latitudeelongitude) was used as the HYSPLIT model input (ftp:// arlftp.arlhq.noaa.gov/pub/archives/gdas1/). The start height of back trajectory was 500 m a.s.l., generally representing the typical height of the MBL. The start locations of back trajectories were some representative stations, and the start time of each back trajectory was matched to the GEM sampling time (Greenwich Mean Time, GMT). 3. Results and discussion 3.1. Potential artifacts for GEM and HgP Though HgP would be lost on the Teflon filter, and most of the RGM in air would be removed when passing the long sampling PTFE tube during the measurement of GEM, there are still some artifacts in determining the GEM. For example, the PTFE sampling tube was not heated in this study and RGM may be lost to the walls of sampling tube (Gustin et al., 2013), but it is not known whether the soda lime trap captures and retains RGM and there are some RGM may be transported to the gold cartridge in the Tekran, which may result in higher GEM concentrations. However, the GEM was reported at a level at least 2 orders of magnitude higher than RGM in the MBL (Chand et al., 2008; Soerensen et al., 2010). Therefore, the atmospheric Hg measured in this study can be referred to as GEM. Another problem is related to the interferences from the halide radicals to the gold cartridge efficiency. This may cause the gold cartridge ruined and its efficiency become erratic. Thus it is likely that GEM was underestimated in this study despite the soda line trap was employed. Gustin et al. (2015) suggested using a pyrolyzer at the sampling inlet if TGM measurement is desired.

Table 1 The detailed description of each set of samples collected during 2013 and 2014. Season

Date

Sampling site or region

Season

Date

Sampling site or region

Spring 2013

2013/6/226/26 2013/6/266/30 2013/6/307/4 2013/7/4e7/8 2013/7/87/11 2013/7/177/20 2014/4/274/30 2014/4/305/3 2014/5/3e5/7 2014/5/85/11 2014/5/115/15 2014/5/155/18

YS Zhoushan Islands ECS ECS ECS ECS YS YS YS YS BS BS

Fall 2013

2013/10/1410/18 2013/10/1810/22 2013/10/2710/30 2013/11/2e11/5 2013/11/5e11/8 2013/11/1411/17 2014/11/7e11/10 2014/11/10e11/14 2014/11/1511/16 2014/11/1611/19 2014/11/1911/21 2014/11/2111/23

ECS ECS ECS ECS ECS ECS YS YS Qingdao YS BS BS

Spring 2014

Fall 2014

110

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

Compared to GEM and RGM, HgP measurements have received less systematic study. Because of the very low concentration HgP (pg m3) in the air, measurements are challenging and trace metal clean handling protocols are required. Tekran has developed a system (2537, 1130 and 1135 units) that includes an annular denuder to first collect RGM, followed by a quartz fiber filter to collect HgP. Though the Tekran system is currently the most widely used configuration for measuring HgP, the magnitude of the Tekran measured HgP bias is presently unclear (Gustin et al., 2015). For HgP a variety of filter systems have been applied (Engle et al., 2008; Keeler et al., 1995; Lynam and Keeler, 2005), and the filter-based methods have been tested in recent years (Kim et al., 2012; Malcolm and Keeler, 2007; Rutter et al., 2008; Talbot et al., 2011). Since HgP is associated with airborne particles, such as dust, soot, and sea salt aerosols, and HgP is likely produced by adsorption of RGM and GEM onto atmospheric particles (Sakata and Marumoto, 2002; Xiu et al., 2005, 2009), there may be some negative and positive artifacts for the filter method of determining HgP in this study such as loss of Hg from collected particulates when sampling time of more than a few hours are employed (Lynam and Keeler, 2005; Malcolm and Keeler, 2007; Rutter et al., 2008) or adsorption of RGM onto collected particulates (Landis et al., 2002). Interestingly, according to our previous studies conducted in the Changdao County, Shandong Province, China (ocean environment), there was no significant difference on the HgP 10 concentration between one day and three days sampling time in the clean weather conditions (the relative percent difference was less than 16%, n ¼ 6), and previous work has suggested that the filter HgP values (24-h) were on the average 21% higher than the Tekran HgP (3-h) (Talbot et al., 2011), therefore the sampling error derived by different sampling duration was estimated to be less than 25% in this study. Recently, Gustin et al. (2015) discussed the HgP measurements and potential artifacts, and there are still some uncertainties in determining the HgP. Therefore, more research is needed to accurately measure the HgP concentration, and potential interferences due to environmental conditions should be systematically investigated. 3.2. Spatial-temporal distribution of GEM Fig. 2 shows the spatial-temporal distribution of GEM in the Asian MBL. Concentrations of GEM in the BS were significantly higher than those in the YS and ECS both in spring (t-test, p < 0.01) and fall (t-test, p < 0.01) due to the influence of anthropogenic sources of Hg, which suggests that the BS was much more easily polluted by anthropogenic Hg than the YS and ECS, and the large emissions of Hg from East Asia have seriously impacted the GEM distribution in the BS. In addition to the anthropogenic emissions, the air-sea exchange of gaseous Hg from the surface sea water to the marine atmosphere is another important source of GEM in the MBL (Ci et al., 2015; Soerensen et al., 2013, 2014). Source tracking through backward air trajectory analysis demonstrated that air masses mainly came from China during the northwest monsoon (fall), bringing continental-and industrial-derived GEM to the study area. In contrast, during spring air masses mainly came from the South China Sea (SCS), West Pacific Ocean, Japan Sea, and Korea Peninsula. The mean (±SD) GEM concentration in the Asian MBL during the entire study period was 2.02 ± 1.08 ng m3 (N ¼ 12,341), which was generally higher than those of remote oceans, such as the Indian Ocean (1.2 ng/m3, Witt et al., 2010), the West Atlantic Ocean (<1.5 ng/m3, Soerensen et al., 2013), and the equatorial Pacific Ocean (<1.4 ng/m3, Soerensen et al., 2014). The statistical summary of GEM concentrations is presented in Table 2. The average levels of GEM in the BS, YS, and ECS were 2.51 ± 0.77 ng m3, 1.89 ± 0.64 ng m3, and 1.61 ± 0.32 ng m3

respectively during spring, and they were 3.64 ± 2.54 ng m3, 1.59 ± 0.44 ng m3, and 2.20 ± 0.58 ng m3 respectively during fall (Table 2). The long-lived GEM shows lower variability especially in the YS and ECS. The results showed that the GEM concentrations in the BS both in spring and fall were much higher than the background level (1.5e1.8 ng m3) in the Northern Hemisphere (Schroeder and Munthe, 1998; Sprovieri et al., 2010), and also higher than those of other sea environments (Chand et al., 2008; Ci et al., 2011a, b; 2015; Fu et al., 2010). It is notable that the GEM concentrations in the BS during fall were comparable to those of polluted marine environments (Nguyen et al., 2007; Tseng et al., 2012) and urban or rural sites in the East Asia, such as Xiamen (3.50 ± 1.21 ng m3, Xu et al., 2015) and Mt. Changbai (3.58 ± 1.78 ng m3, Wan et al., 2009a) in China, Jeju Island (3.85 ± 1.68 ng/m3, Nguyen et al., 2010) and Seoul (3.22 ± 2.10 ng m3, Kim et al., 2009) in South Korea, while the GEM concentrations in the ECS (spring) and YS (fall) were comparable to the background level in the Northern Hemisphere. In addition, the previous study conducted in the YS showed that the GEM level in spring (1.86 ± 0.40 ng m3, Ci et al., 2015) was comparable to our observation in spring (1.89 ± 0.64 ng m3). Furthermore, it could be found that the GEM concentrations in the coastal area were generally higher than those in the open sea, which consists with the results of other studies that atmospheric Hg emission from anthropogenic sources in East Asia enhances the atmospheric Hg levels in the downwind regions (Friedli et al., 2004; Fu et al., 2010; Jaffe et al., 2005; Obrist et al., 2008). 3.3. Spatial-temporal distribution of HgP The HgP 10 was obtained by summing the nine fractions of the same set samples. The mean concentration of HgP 10 during spring was 15.3 ± 9.1 pg m3 with a range of 3.6e34.2 pg m3 except the set sampled at the coastal site (Shenjiamen) in the Zhoushan Islands (ZI), Zhejiang Province, China, while the mean HgP 10 concentration during fall was 15.8 ± 4.4 pg m3 with a range of 9.6e23.7 pg m3 except the set sampled at Jiaozhou Bay, Shandong Province, China (Table 3). The ratios of the HgP 10 to the total Hg of GEM and HgP 10 (spring: 1.94 ng m3; fall: 2.11 ng m3) in the Asian MBL were all < 1% both in spring and fall, which were much lower than those in urban cities, such as Tokyo and Nanjing (Sakata and Marumoto, 2002; Zhu et al., 2014), but slightly higher than those at the marine site (Appledore Island) (Feddersen et al., 2012; Mao and Talbot, 2012). The average concentration of HgP 10 during the whole study period was 18.3 pg m3, which was much lower than the total particulate Hg (TPM) measured in metropolitan cities (see Table 3), such as Changchun, Beijing, and Shanghai (Fang et al., 2001; Wang et al., 2006; Xiu et al., 2005, 2009), slightly lower than the TPM measured at the Mt. Gongga (Fu et al., 2008), but higher than those at rural and marine sites, such as the Chuncheon (Kim et al., 2012) in Korea and the Appledore Island (Feddersen et al., 2012; Mao and Talbot, 2012) in the US. Previous studies conducted in urban Nanjing (Zhu et al., 2014) and Tokyo (Sakata and Marumoto, 2002) showed that the HgP 10 concentrations were considerably higher than those of our measurements. Additionally, the HgP 2.5 measured at the Mt. Changbai in China (Wan et al., 2009b) and the Alert in Canada (Cobbett et al., 2007) were higher than the HgP 10 measured in this study, indicating that the HgP 10 concentrations at the Mt. Changbai and Alert may much higher compared to the HgP 10 in this study area. The spatial-temporal distribution of HgP is illustrated in Fig. 3. It could be found that the HgP 10 concentrations tended to decrease from the land to the open sea both in spring and fall except the measurements in the BS due to its specific location. During spring,

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

111

Fig. 2. Concentrations of GEM and 72-h back-trajectory analysis of air masses in the Asian marine boundary layer during spring and fall.

Table 2 Statistical summary of GEM concentrations (ng m3) in the Bohai Sea, Yellow Sea, and East China Sea during spring and fall. Location

Bohai Sea Yellow Sea East China Sea

Spring

Fall

Min

Max

Media

Mean

SD

N

Min

Max

Media

Mean

SD

N

1.20 0.99 0.85

5.68 4.95 4.48

2.43 1.75 1.56

2.51 1.89 1.61

0.77 0.64 0.32

1255 3988 1505

1.17 1.03 1.16

17.41 3.95 6.79

2.96 1.47 2.08

3.64 1.59 2.20

2.54 0.44 0.58

1080 3332 1181

Table 3 The concentrations and dry deposition fluxes of HgP in this study area and other sites and areas. Location

Sampling time

Characteristic

Cut off size

HgP conc. (pg m3)

Dry deposition (ng m2 d1)

Reference

BS,YS, and ECS

2013e2014 (spring) 2013e2014 (fall) 2003e2004 2003e2004 2011e2012 1999e2000 2003e2006 2005e2006 2010 2009e2010 2009e2010 1994 2009e2010 (summer) 2009e2010 1994e1995

Ocean Ocean Urban Rural Urban Urban Urban Rural Industry Urban Rural Urban Ocean Rural Lake

9-stage Andersen 9-stage Andersen 7, 3.3, 2.0, 1.1 mm 7, 3.3, 2.0, 1.1 mm 9-stage Andersen TPM 18, 8, 3.7, 1.6 mm TPM 18, 10, 2.5, 1.0 mm 8-stage MOUDI 8-stage MOUDI TPM 10-stage Andersen 10-stage Andersen TPM

3.6e34.2 9.6e23.7 180e3510 130e2400 320e2040 145 70e1450 5.2e135.7 2551 1.1e18.5 1.0e8.9 22e225 2.1e7.6 1.6e9.6 n.a.

2.77 1.92 1115.1 739.7 172.3 118.1 n.a. n.a. 288 n.a. n.a. n.a. 2.8 3.0 26.6

This work This work Wang et al., 2006 Wang et al., 2006 Zhu et al., 2014 Fang et al., 2001 Xiu et al., 2005, 2009 Fu et al., 2008 Chen et al., 2012 Kim et al., 2012 Kim et al., 2012 Keeler et al., 1995 Feddersen et al., 2012 Feddersen et al., 2012 Landis and Keeler, 2002

Beijing, China Nanjing, China Changchun, China Shanghai, China Mt. Gongga, China Quan-xing, Taiwan Seoul, Korea Chuncheon, Korea Detroit, USA Appledore Island, USA Thompson Farm, USA Lake Michigan, USA

TPM denotes total particulate mercury. n.a. denotes not applicable.

112

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

Fig. 3. Spatial distribution of HgP 10 and HgP 2.1/HgP 10 ratios in the Asian marine boundary layer and cruise tracks during spring and fall (HgP 2.1 and HgP 10 denote the HgP in PM2.1 and HgP in PM10 respectively).

one striking feature is that the HgP 10 concentration at the ZI (76.6 pg m3) was significantly higher than those of other open sea areas (left panel of Fig. 3), and the ratio of HgP 2.1 to HgP 10 was up to 60% at the ZI, indicating that a large amount of anthropogenic Hg was released at the ZI because it is a capacious harbor (there are lots of cargo ships and fishing boats). The HgP 10 concentration (34.2 pg m3) in the offshore area of the Yangtze River Delta was lower than that at the ZI, but higher than those of other marine environments in this study. The average HgP 10 concentrations in the open sea areas of the BS, YS, and ECS during spring were 24.5 pg m3, 8.2 pg m3, and 13.2 pg m3 respectively, and the HgP 10 concentrations in the BS were 2e7 times higher than those in the YS. During fall, the average HgP 10 levels in the BS, YS, and ECS were 18.2 pg m3, 14.3 pg m3, and 16.6 pg m3 respectively. The large export of anthropogenic Hg from the surrounding areas (e.g., Shandong, Hebei, Tianjin, Beijing, and Liaoning Provinces) was probably the main reason for the higher HgP 10 concentrations in the BS. The ratios of HgP 2.1 to HgP 10 would be further discussed in the following section. Interestingly, the results showed that HgP 10 concentrations were correlated positively with GEM concentrations both in spring and fall (see Figs. 2 and 3), for example, the GEM and HgP 10 simultaneously exhibited higher levels in the BS and lower levels in the YS. We speculate that there are two main reasons for the higher HgP 10 concentrations in the BS compared to those in the YS and ECS, including: (a) the directly discharge of HgP from the surrounding land (mainly East China) and (b) the oxidation of GEM by hydroxyl radical (OH) and O3 (Calvert and Lindberg, 2005; Selin et al., 2007) and reactive halogen species (Br, Cl, and BrO etc.) (Auzmendi-Murua et al., 2014; Holmes et al., 2009, 2010; Seigneur and Lohman, 2008) in the marine atmosphere (the higher GEM and

RGM concentrations may induce higher HgP levels in the BS). It should be noted that there was large difference in the discharge of HgP among the four different seasons in China (Fu et al., 2012). Previous studies have showed that there were a seasonal difference in halogen chemistry and a distinct diurnal variation of halogen species with higher value during daytime in the MBL (Sander et al., 2003; von Glasow et al., 2002). Additionally, it was obviously found that the HgP 10 concentrations in the YS and ECS were generally higher in fall than in spring, while the average HgP 10 concentration during spring was comparable to that during fall in the BS (see Fig. 3). The seasonal variability of HgP 10 can be explained by the different wind patterns during spring and fall. The prevailing wind directions were northwest during fall (see Fig. 2), indicating that large particulate matters could be transported to the YS and ECS from China, which is the biggest contributor to the global Hg emissions (Pacyna et al., 2006). In contrast, the prevailing wind directions were south in the ECS and northeast in the YS, indicating that the air masses largely originated from the oceans and remote areas. The higher concentrations of TPM or HgP 10 at rural and urban sites in China (Fang et al., 2001; Fu et al., 2008; Wang et al., 2006; Xiu et al., 2005, 2009; Xu et al., 2013, 2015; Zhu et al., 2014) and lower concentrations in the Korea Peninsula and the Okinawa Island (Chand et al., 2008; Kim et al., 2012) may support the spatial distribution of HgP 10 in this study area. To some extent, this was due to the transport of anthropogenic sources of HgP from China to marine environment. 3.4. Size distribution of HgP The size distributed concentrations of HgP in the Asian MBL during spring and fall are shown in Fig. 4. The concentrations of all

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

113

Fig. 4. Size distributed concentrations of HgP in the Asian marine boundary layer during spring and fall. The data shown are the mean ± standard error.

size fractioned HgP during spring and fall were summarized in Supplemental Information (Table S1 and Table S2). The size distributions of HgP were observed to be bi-modal during spring and unimodal during fall. There were two peaks during spring: a higher peak and a lower peak were found in the size range of 0.7e1.1 mm and 3.3e4.7 mm respectively, and HgP 2.1 contributed approximately 43% (19%e62%, see Fig. 3) to the HgP 10. In contrast, the fine mode (<2.1 mm) was the dominant size during fall, peaking in the 0.7e1.1 mm size range, approximately 53% (44%e61%, see Fig. 3) of the HgP 10 was found in fine particles. The main reason for the different HgP size distributions between spring and fall is the discrepant particles sources. Fig. 5 shows the frequency distribution of wind speed segregated into five categories (<3, 3e6, 6e9, 9e12, and >12 m s1) per 15 bins of wind direction. The prevailing wind direction was south (135 e225 ) during spring (Fig. 5a). In contrast, it was north during fall (Fig. 5b). Furthermore, the back trajectories analysis demonstrated that the prevailing wind direction during spring was south in the ECS (Fig. 2), indicating the air masses largely originated from the West Pacific Ocean and SCS. The ratios of HgP 2.1 to HgP 10 in the ECS and northern YS were significantly low (left panel of Fig. 3), this is due to the fact that these air masses contain high concentrations of seasalts which generally exist in the coarse mode (1e10 mm) (Athanasopoulou et al., 2008; Mamane et al., 2008). Moreover, coarse particles of sodium nitrate and chlorides of potassium and

sodium have high partitioning coefficients, shifting the RGM partitioning toward the particle phase (Rutter and Schauer, 2007), which may increase the concentrations of HgP in coarse particles. While HgP mainly concentrated on fine particles in the BS and coastal marine environments (e.g., Yangtze River Delta) during spring, and approximately 60% of Hg in PM10 concentrated on PM2.1. Therefore, we found the bi-modal distribution of HgP during spring. In contrast, winds were predominantly from the northwest during fall, almost all the air masses in the study regions originated from land, including the China and Japan (right panel of Fig. 2). Furthermore, numerous researches on the atmospheric HgP have been conducted in many cities in East China, such as Changchun and Beijing (Fang et al., 2001; Wang et al., 2006) in the north of China, and Shanghai (Xiu et al., 2005, 2009), Nanjing (Zhu et al., 2014), Tainan (Tsai et al., 2003), and Xiamen (Xu et al., 2013) in the southeast of China, where HgP mainly concentrated on fine particles. Therefore, there was only one distinct peak during fall and the ratios of HgP 2.1 to HgP 10 were higher in fall (generally > 50%) compared to in spring (see Figs. 3 and 4). 3.5. Dry deposition of HgP In this study, the total dry deposition flux of HgP was obtained by summing the fluxes of all size-fractionated of HgP. The dry deposition flux of HgP is calculated using Eq. (1).

Fig. 5. The average frequency of wind direction during the spring (a) and fall (b) cruises.

114

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

F¼

P

CHgP  Vd

(1)

where F is the total dry deposition flux of HgP (ng m2 d1), CHgP is the concentration of HgP in each size fraction (pg m3), and Vd is the corresponding dry deposition velocity (cm s1). Dry deposition velocity not only depends on particle properties, but also on meteorological conditions and surface characteristics (Landis and Keeler, 2002; Zhang et al., 2001). In this study, the dry deposition velocities 0.03, 0.01, 0.06, 0.15, and 0.55 cm s1 (Giorgi, 1988; NhoKim et al., 2004; Pryor and Sorensen, 2000) over ocean surface were chosen for the following size-fraction particles: <0.4, 0.4e1.1, 1.1e2.1, 2.1e5.8, and 5.8e10 mm respectively. The average dry deposition fluxes of HgP were estimated to be 2.77 ng m2 d1 during spring and 1.92 ng m2 d1 during fall based on the concentrations of size-fractionated HgP in the Asian MBL (see Table S3 and Table S4). During spring, the highest and lowest dry deposition fluxes of HgP were observed in the ZI and YS respectively, where we observed the highest and lowest concentrations of HgP respectively. The average dry deposition fluxes of HgP in the BS, YS, and ECS were 2.23, 1.78, and 3.97 ng m2 d1, respectively. The results showed that the average dry deposition flux of HgP in the BS was lower than that in the ECS though the HgP 10 in the BS was higher than that in the ECS, this is due to the fact that the concentrations HgP in coarse particles in the ECS were generally higher than those in the BS and the coarse particles have higher deposition velocities (Nho-Kim et al., 2004). Interestingly, there was no significant difference on the average dry deposition fluxes among the BS (1.97 ng m2 d1), YS (1.76 ng m2 d1), and ECS (2.00 ng m2 d1) during fall because the HgP 10 levels in the BS, YS, and ECS were similar and ratios of HgP 2.1 to HgP 10 in the BS, YS, and ECS were also similar. Table 3 shows dry deposition fluxes of HgP in this study and other studies. The dry deposition fluxes of HgP in this study area were much lower than those of urban cities such as Quan-xing, Changchun, Beijing, and Nanjing (Chen et al., 2012; Fang et al., 2001; Wang et al., 2006; Zhu et al., 2014), and were also lower than those of remote areas, such as the Lake Michigan (Landis and Keeler, 2002) and the Central and Northern Europe (17.5 mg m2 yr1; Petersen et al., 1995), but were comparable to that measured at a marine (the Appledore Island) and coastal site (Thompson Farm) (Feddersen et al., 2012). Gencarelli et al. (2014) evaluated the dry deposition flux of HgP in the Mediterranean Sea using a model, and their value was slightly higher than those obtained in this study. However, the dry deposition flux of HgP is very difficult to measure directly, and more research work is needed in the future. The importance of the coarse particles can be found in the contribution of the dry deposition of HgP in coarse particles to the total dry deposition flux of HgP 10. The results showed that HgP in coarse particles contributed more than 90% to the total dry deposition of HgP due to the higher deposition velocities of coarse particles both in spring and fall. In other words, the dry deposition fluxes of HgP in coarse particles were approximately 10 times higher than those of HgP in fine particles. Therefore, the coarse particles play an important role in the dry deposition of HgP in the Asian MBL. However, it should be noted that there are still a lot of uncertainties in estimating the dry deposition of HgP because it is closely related to the physical and chemical properties of the particulates (particle size, shape, solubility, and density), meteorological conditions (wind speed, atmospheric stability, and relative humidity), and the depositing surface properties (Landis and Keeler, 2002; Zhang et al., 2001, 2009, 2012). 4. Conclusions In order to identify the spatial distributions of GEM and HgP and

compare the seasonal variability of GEM and HgP in the Asian MBL, the concentrations of GEM and HgP 10 in the Asian MBL were determined. The mean levels of GEM and HgP 10 in the Asian MBL were higher than those at rural sites in North America and South Korea, but considerably lower than those at urban or rural sites in East China. Concentrations of GEM in the BS were significantly higher than those in the YS and ECS both in spring and fall, indicating that the BS has been seriously polluted by the anthropogenic Hg. The concentrations of HgP 10 in the offshore area were higher than those in the open sea both in spring and fall. The size distributions of HgP were observed to be bi-modal during spring and unimodal during fall, and the fine mode was the dominate size during fall. The average deposition flux of HgP during spring was slightly higher than that during fall, but they were much lower than those measured at urban sites in China. Coarse particles play a critical role in the dry deposition of HgP in the Asian marine atmosphere. Since we only determined the HgP 10 concentrations in spring and fall, and there was no data on the HgP 10 levels in summer and winter, thus more research work (especially during summer and winter) is needed in the future. Acknowledgments This research was funded by the National Key Basic Research Program of China (No. 2013CB430002), National Natural Science Foundation of China (No. 41176066) and “Strategic Priority Research Program” of the Chinese Academy of Sciences, Grant No. XDB14020205. We gratefully acknowledge the open cruises in 2013 organized by the Institute of Oceanology, Chinese Academy of Sciences (IOCAS) and the open cruises in 2014 organized by the Ocean University of China (OUC). We thank the personnel from the IOCAS and OUC for their efforts. The authors would like to thank the captains and crews of the R/V of Kexue III (IOCAS) and Dongfanghong II (OUC) for their assistance on sample and data collections. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.atmosenv.2015.11.036. References Athanasopoulou, E., Tombrou, M., Pandis, S.N., Russell, A.G., 2008. The role of seasalt emissions and heterogeneous chemistry in the air quality of polluted coastal areas. Atmos. Chem. Phys. 8, 5755e5769.
Bozzelli, J.W., 2014. Mercury oxidation via chlorine, Auzmendi-Murua, I., Castillo, A., bromine, and iodine under atmospheric conditions: thermochemistry and kinetics. J. Phys. Chem. A 118 (16), 2959e2975. Calvert, J.G., Lindberg, S.E., 2005. Mechanisms of mercury removal by O3 and OH in the atmosphere. Atmos. Environ. 39, 3355e3367. Chand, D., Jaffe, D., Prestbo, E., Swartzendruber, P.C., Hafner, W., Weiss-Penzias, P., Kato, S., Takami, A., Hatakeyama, S., Kajii, Y., 2008. Reactive and particulate mercury in the Asian marine boundary layer. Atmos. Environ. 42, 7988e7996. Chen, S., Lo, C., Fang, G., Huang, C., 2012. Particulate-bound mercury (Hg[p]) size distributions in central Taiwan. Environ. Forensics 13 (2), 98e104. Choi, H.D., Holsen, T.M., Hopke, P.K., 2008. Atmospheric mercury (Hg) in the adirondacks: concentrations and sources. Environ. Sci. Technol. 42, 5644e5653. Ci, Z., Zhang, X., Wang, Z., 2011a. Elemental mercury in coastal seawater of Yellow Sea, China: temporal variation and airesea exchange. Atmos. Environ. 45, 183e190. Ci, Z., Zhang, X., Wang, Z., Niu, Z., 2011b. Atmospheric gaseous elemental mercury (GEM) over a coastal/rural site downwind of East China: temporal variation and long-range transport. Atmos. Environ. 45, 2480e2487. Ci, Z., Zhang, X., Wang, Z., Niu, Z., 2011c. Phase speciation of mercury (Hg) in coastal water of the Yellow Sea, China. Mar. Chem. 126, 250e255. Ci, Z., Wang, C., Zhang, X., Wang, Z., 2015. Elemental mercury (Hg(0)) in air and surface waters of the Yellow Sea during late spring and late fall 2012: concentration, spatial-temporal distribution and air/sea flux. Chemosphere 119, 199e208. Cobbett, F.D., Steffen, A., Lawson, G., Van Heyst, B.J., 2007. GEM fluxes and atmospheric mercury concentrations (GEM, RGM and Hgp) in the Canadian Arctic at Alert, Nunavut, Canada (FebruaryeJune 2005). Atmos. Environ. 41, 6527e6543.

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116 Draxler, R.R., Rolph, G.D., 2012. HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory) Model access via NOAA ARL READY. NOAA air Resources laboratory, Silver Spring, MD. Website. http://www.arl.noaa.gov/ready/hysplit4. html. Engle, M.A., Tate, M.T., Krabbenhoft, D.P., Kolker, A., Olson, M.L., Edgerton, E.S., DeWild, J.F., McPherson, A.K., 2008. Characterization and cycling of atmospheric mercury along the central US Gulf Coast. Appl. Geochem. 23 (3), 419e437. Fang, F., Wang, Q., Li, J., 2001. Atmospheric particulate mercury concentration and its dry deposition flux in Changchun City, China. Sci. Total Environ. 281, 229e236. Feddersen, D.M., Talbot, R., Mao, H., Sive, B.C., 2012. Size distribution of particulate mercury in marine and coastal atmospheres. Atmos. Chem. Phys. 12, 10899e10909. Fitzgerald, W.F., Gill, G.A., 1979. Subnanogram determination of mercury by twostage gold amalgamation and gas phase detection applied to atmospheric analysis. Anal. Chem. 51 (11), 1714e1720. Friedli, H.R., Radke, L.F., Prescott, R., Li, P., Woo, J.H., Carmichael, G.R., 2004. Mercury in the atmosphere around Japan, Korea and China as observed during the 2001 ACE-Asia field campaign: measurements, distributions, sources, and implications. J. Geophys. Res. 109 (D19S25). http://dx.doi.org/10.1029/2003JD004244. Fu, X., Feng, X., Zhu, W., Zheng, W., Wang, S., Lu, J.Y., 2008. Total particulate and reactive gaseous mercury in ambient air on the eastern slope of the Mt. Gongga area, China. Appl. Geochem. 23 (3), 408e418. Fu, X., Feng, X., Zhang, G., Xu, W., Li, X., Yao, H., Liang, P., Li, J., Sommar, J., Yin, R., Liu, N., 2010. Mercury in the marine boundary layer and seawater of the South China Sea: concentrations, sea/air flux, and implication for land outflow. J. Geophys. Res. 115, D06303. http://dx.doi.org/10.1029/2009JD012958. Fu, X., Feng, X., Sommar, J., Wang, S., 2012. A review of studies on atmospheric mercury in China. Sci. Total Environ. 421e422, 73e81. Gabriel, M.C., Williamson, D.G., Brooks, S., Lindberg, S., 2005. Atmospheric speciation of mercury in two contrasting Southeastern US airsheds. Atmos. Environ. 39, 4947e4958. Gencarelli, C.N., De Simone, F., Hedgecock, I.M., Sprovieri, F., Pirrone, N., 2014. Development and application of a regional-scale atmospheric mercury model based on WRF/Chem: a mediterranean area investigation. Environ. Sci. Pollut. Res. 21, 4095e4109. Giorgi, F., 1988. Dry deposition velocities of atmospheric aerosols as inferred by applying a particle dry deposition parameterisation to a general circulation model. Tellus 40B (1), 23e41. Gustin, M.S., Huang, J., Miller, M.B., Peterson, C., Jaffe, D.A., Ambrose, J., Finley, B.D., Lyman, S.N., Call, K., Talbot, R., Feddersen, D., Mao, H., Lindberg, S.E., 2013. Do we understand what the mercury speciation instruments are actually measuring? Results of RAMIX. Environ. Sci. Technol. 47, 7295e7306. Gustin, M.S., Amos, H.M., Huang, J., Miller, M.B., Heidecorn, K., 2015. Measuring and modeling mercury in the atmosphere: a critical review. Atmos. Chem. Phys. 15, 5697e5713. Holmes, C.D., Jacob, D.J., Mason, R.P., Jaffe, D.A., 2009. Sources and deposition of reactive gaseous mercury in the marine atmosphere. Atmos. Environ. 43, 2278e2285. Holmes, C.D., Jacob, D.J., Corbitt, E.S., Mao, J., Yang, X., Talbot, R., Slemr, F., 2010. Global atmospheric model for mercury including oxidation by bromine atoms. Atmos. Chem. Phys. 10, 12037e12057. Jaffe, D., Prestbo, E., Swartzendruber, P., Weiss-Penzias, P., Kato, S., Takami, A., Hatakeyama, S., Kajii, Y., 2005. Export of atmospheric mercury from Asia. Atmos. Environ. 39, 3029e3038. Keeler, G., Glinsorn, G., Pirrone, N., 1995. Particulate mercury in the atmosphere: its significance, transport, transformation and sources. Water Air Soil Poll. 80, 159e168. Kim, S., Han, Y., Holsen, T.M., Yi, S., 2009. Characteristics of atmospheric speciated mercury concentrations (TGM, Hg(II) and Hg(p)) in Seoul, Korea. Atmos. Environ. 43, 3267e3274. Kim, P., Han, Y., Holsen, T.M., Yi, S., 2012. Atmospheric particulate mercury: concentrations and size distributions. Atmos. Environ. 61, 94e102. Landis, M.S., Keeler, G.J., 2002. Atmospheric mercury deposition to Lake Michigan during the Lake Michigan mass balance study. Environ. Sci. Technol. 36, 4518e4524. Landis, M.S., Stevens, R.K., Schaedlich, F., Prestbo, E.M., 2002. Development and characterization of an annular denuder methodology for the measurement of divalent inorganic reactive gaseous mercury in ambient air. Environ. Sci. Technol. 36, 3000e3009. Liu, N., Qiu, G., Landis, M., Feng, X., Fu, X., Shang, L., 2011. Atmospheric mercury species measured in Guiyang, Guizhou province, southwest China. Atmos. Res. 100, 93e102. Lu, J.Y., Schroeder, W.H., Berg, T., Munthe, J., Schneeberger, D., Schaedlich, F., 1998. A device for sampling and determination of total particulate mercury in ambient air. Anal. Chem. 70 (11), 2403e2408. Lynam, M.M., Keeler, G.J., 2005. Artifacts associated with the measurement of particulate mercury in an urban environment: the influence of elevated ozone concentrations. Atmos. Environ. 39, 3081e3088. Malcolm, E.G., Keeler, G.J., 2007. Evidence for a sampling artifact for particulatephase mercury in the marine atmosphere. Atmos. Environ. 41, 3352e3359. Mamane, Y., Perrino, C., Yossef, O., Catrambone, M., 2008. Source characterization of fine and coarse particles at the East mediterranean coast. Atmos. Environ. 42, 6114e6130. Mao, H., Talbot, R., 2012. Speciated mercury at marine, coastal, and inland sites in

115

New EnglandePart 1: temporal variability. Atmos. Chem. Phys. 12, 5099e5112. Mason, R.P., Sheu, G.R., 2002. Role of the ocean in the global mercury cycle. Global Biogeochem. Cy. 16 (4), 1093. http://dx.doi.org/10.1029/2001GB001440. Nguyen, H.T., Kim, K.H., Kim, M.Y., Hong, S., Youn, Y.H., Shon, Z.H., Lee, J.S., 2007. Monitoring of atmospheric mercury at a global atmospheric Watch (GAW) site on an-myun island, Korea. Water Air Soil Poll. 185, 149e164. Nguyen, H.T., Kim, M.Y., Kim, K.H., 2010. The influence of long-range transport on atmospheric mercury on Jeju Island, Korea. Sci. Total Environ. 408, 1295e1307. Nho-Kim, E.Y., Michou, M., Peuch, V.H., 2004. Parameterization of size-dependent particle dry deposition velocities for global modeling. Atmos. Environ. 38, 1933e1942. Obrist, D., Hallar, A.G., McCubbin, I., Stephens, B.B., Rahn, T., 2008. Atmospheric mercury concentrations at Storm Peak laboratory in the RockyMountains: evidence for long-range transport from Asia, boundary layer contributions, and plant mercury uptake. Atmos. Environ. 42, 7579e7589. Pacyna, J.M., Pacyna, E.G., Steenhuisen, F., Wilson, S., 2003. Mapping 1995 global anthropogenic emissions of mercury. Atmos. Environ. 37, S109eS117. Pacyna, E.G., Pacyna, J.M., Steenhuisen, F., Wilson, S., 2006. Global anthropogenic mercury emission inventory for 2000. Atmos. Environ. 40, 4048e4063. Pacyna, E.G., Pacyna, J.M., Sundseth, K., Munthe, J., Kindbom, K., Wilson, S., Steenhuisen, F., Maxson, P., 2010. Global emission of mercury to the atmosphere from anthropogenic sources in 2005 and projections to 2020. Atmos. Environ. 44, 2487e2499. Petersen, G., Iverfeldt, Å., Munthe, J., 1995. Atmospheric mercury species over Central and Northern Europe model calculations and comparison with observations from Nordic air and Precipitation Network for 1987 and 1988. Atmos. Environ. 29, 47e67. Pryor, S.C., Sorensen, L.L., 2000. Nitric acidesea salt reactions: implications for nitrogen deposition to water surfaces. J. Appl. Meteorol. 39, 725e731. Rutter, A.P., Schauer, J.J., 2007. The impact of aerosol composition on the particle to gas partitioning of reactive mercury. Environ. Sci. Technol. 41, 3934e3939. Rutter, A.P., Hanford, K.L., Zwers, J.T., Perillo-Nicholas, A.L., Schauer, J.J., Olson, M.L., 2008. Evaluation of an offline method for the analysis of atmospheric reactive gaseous mercury and particulate mercury. J. Air Waste Manag. Assoc. 58, 377e383. Sakata, M., Marumoto, K., 2002. Formation of atmospheric particulate mercury in the Tokyo metropolitan area. Atmos. Environ. 36, 239e246. Sakata, M., Marumoto, K., 2005. Wet and dry deposition fluxes of mercury in Japan. Atmos. Environ. 39, 3139e3146. Sander, R., Keene, W.C., Pszenny, A.A.P., Arimoto, R., Ayers, G.P., Baboukas, E., €nninger, G., Huebert, B.J., Maenhaut, W., Cainey, J.M., Crutzen, P.J., Duce, R.A., Ho Mihalopoulos, N., Turekian, V.C., Van Dingenen, R., 2003. Inorganic bromine in the marine boundary layer: a critical review. Atmos. Chem. Phys. 3, 1301e1336. Schroeder, W.H., Munthe, J., 1998. Atmospheric Mercurydan overview. Atmos. Environ. 32, 809e822. Seigneur, C., Lohman, K., 2008. Effect of bromine chemistry on the atmospheric mercury cycle. J. Geophys. Res. 113, D23309. http://dx.doi.org/10.1029/ 2008JD010262.
, L., Jaffe, D., 2007. Selin, N.E., Jacob, D.J., Park, R.J., Yantosca, R.M., Strode, S., Jaegle Chemical cycling and deposition of atmospheric mercury: global constraints from observations. J. Geophys. Res. 112, D02308. http://dx.doi.org/10.1029/ 2006JD007450. Soerensen, A.L., Skov, H., Jacob, D.J., Soerensen, B.T., Johnson, M.S., 2010. Global concentrations of gaseous elemental mercury and reactive gaseous mercury in the marine boundary layer. Environ. Sci. Technol. 44, 7425e7430. Soerensen, A.L., Mason, R.P., Balcon, P.H., Sunderland, E.M., 2013. Drivers of surface ocean mercury concentrations and airsea exchange in the West Atlantic Ocean. Environ. Sci. Technol. 47, 7757e7765. Soerensen, A.L., Mason, R.P., Balcom, P.H., Jacob, D.J., Zhang, Y., Kuss, J., Sunderland, E.M., 2014. Elemental mercury concentrations and fluxes in the tropical atmosphere and ocean. Environ. Sci. Technol. 48, 11312e11319. Sprovieri, F., Pirrone, N., Ebinghaus, R., Kock, H., Dommergue, A., 2010. A review of worldwide atmospheric mercury measurements. Atmos. Chem. Phys. 10, 8245e8265. Streets, D.G., Hao, J., Wu, Y., Jiang, J., Chan, M., Tian, H., Feng, X., 2005. Anthropogenic mercury emissions in China. Atmos. Environ. 39, 7789e7806. Talbot, R., Mao, H., Feddersen, D., Smith, M., Kim, S.Y., Sive, B., Haase, K., Ambrose, J., Zhou, Y., Russo, R., 2011. Comparison of particulate mercury measured with manual and automated methods. Atmosphere 2 (1), 1e20. Tsai, Y.I., Kuo, S., Lin, Y., 2003. Temporal characteristics of inhalable mercury and arsenic aerosols in the urban atmosphere in southern Taiwan. Atmos. Environ. 37, 3401e3411. Tseng, C.M., Liu, C.S., Lamborg, C.H., 2012. Seasonal changes in gaseous elemental mercury in relation to monsoon cycling over the Northern South China Sea. Atmos. Chem. Phys. 12, 7341e7350. von Glasow, R., Sander, R., Bott, A., Crutzen, P.J., 2002. Modeling halogen chemistry in the marine boundary layer 1. Cloud-free MBL. J. Geophys. Res. 107 (D17), 4341. http://dx.doi.org/10.1029/2001JD000942. Wan, Q., Feng, X.B., Lu, J.L., Zheng, W., Song, X.J., Han, S.J., Xu, H., 2009a. Atmospheric mercury in Changbai mountain area, northeastern China I. The seasonal distribution pattern of total gaseous mercury and its potential sources. Environ. Res. 109, 201e206. Wan, Q., Feng, X.B., Lu, J., Zheng, W., Song, X.J., Li, P., Han, S.J., Xu, H., 2009b. Atmospheric mercury in Changbai mountain area, northeastern China II. The distribution of reactive gaseous mercury and particulate mercury and mercury

116

C. Wang et al. / Atmospheric Environment 126 (2016) 107e116

deposition fluxes. Environ. Res. 109, 721e727. Wang, Z., Zhang, X., Chen, Z., Zhang, Y., 2006. Mercury concentrations in sizefractionated airborne particles at urban and suburban sites in Beijing, China. Atmos. Environ. 40 (12), 2194e2201. Wang, Y.Q., Zhang, X.Y., Draxler, R.R., 2009. TrajStat: GIS-based software that uses various trajectory statistical analysis methods to identify potential sources from long-term air pollution measurement data. Environ. Modell. Softw. 24 (8), 938e939. Witt, M.L.I., Mather, T.A., Baker, A.R., De Hoog, J.C.M., Pyle, D.M., 2010. Atmospheric trace metals over the south-west Indian Ocean: total gaseous mercury, aerosol trace metal concentrations and lead isotope ratios. Mar. Chem. 121, 2e16. Wu, Y., Wang, S., Streets, D.G., Hao, J., Chan, M., Jiang, J., 2006. Trends in anthropogenic mercury emissions in China from 1995 to 2003. Environ. Sci. Technol. 40, 5312e5318. Xia, C., Xie, Z., Sun, L., 2010. Atmospheric mercury in the marine boundary layer along a cruise path from Shanghai, China to Prydz Bay, Antarctica. Atmos. Environ. 44, 1815e1821. Xiu, G., Jin, Q., Zhang, D., Shi, S., Huang, X., Zhang, W., Bao, L., Gao, P., Chen, B., 2005. Characterization of size-fractionated particulate mercury in Shanghai ambient air. Atmos. Environ. 39 (3), 419e427. Xiu, G., Cai, J., Zhang, W., Zhang, D., Büeler, A., Lee, S., Shen, Y., Xu, L., Hunag, X., Zhang, P., 2009. Speciated mercury in size-fractionated particles in Shanghai ambient air. Atmos. Environ. 43, 3145e3154. Xu, L., Chen, J., Niu, Z., Yin, L., Chen, Y., 2013. Characterization of mercury in

atmospheric particulate matter in the southeast coastal cities of China. Atmos. Pollut. Res. 4 (4), 454e461. Xu, L., Chen, J., Yang, L., Niu, Z., Tong, L., Yin, L., Chen, Y., 2015. Characteristics and sources of atmospheric mercury speciation in a coastal city, Xiamen, China. Chemosphere 119, 530e539. Zhang, L., Gong, S., Padro, J., Barrie, L., 2001. A size-segregated particle dry deposition scheme for an atmospheric aerosol module. Atmos. Environ. 35, 549e560. Zhang, L., Paige Wright, L., Blanchard, P., 2009. A review of current knowledge concerning dry deposition of atmospheric mercury. Atmos. Environ. 43, 5853e5864. Zhang, L., Blanchard, P., Gay, D.A., Prestbo, E.M., Risch, M.R., Johnson, D., Narayan, J., Zsolway, R., Holsen, T.M., Miller, E.K., Castro, M.S., Graydon, J.A., Louis, V.L.St., Dalziel, J., 2012. Estimation of speciated and total mercury dry deposition at monitoring locations in eastern and central North America. Atmos. Chem. Phys. 12, 4327e4340. Zhang, L., Wang, S., Wang, L., Wu, Y., Duan, L., Wu, Q., Wang, F., Yang, M., Yang, H., Hao, J., Liu, X., 2015. Updated emission inventories for speciated atmospheric mercury from anthropogenic sources in China. Environ. Sci. Technol. 49, 3185e3194. Zhu, J., Wang, T., Talbot, R., Mao, H., Yang, X., Fu, C., Sun, J., Zhuang, B., Li, S., Han, Y., Xie, M., 2014. Characteristics of atmospheric mercury deposition and sizefractionated particulate mercury in urban Nanjing, China. Atmos. Chem. Phys. 14, 2233e2244.