Key Laboratory of Global Change and Marine-Atmospheric Chemistry, Third Institute of Oceanography, State Oceanic Administration, Xiamen, China
Corresponding author: L. Chen, Key Laboratory of Global Change and Marine-Atmospheric Chemistry, Third Institute of Oceanography, State Oceanic Administration, Xiamen 361005, China. (email@example.com)
 To characterize the spatial distributions of methane-sulfonic acid (MSA) as represented by measured aerosol methane sulfonate (MS) and its relationships with non-sea-salt (nss) sulfate (SO42−) in the marine atmospheric boundary layer over high-latitude regions, bulk aerosol samples were collected during eight cruises during the Chinese National Antarctic and Arctic Research Expeditions from 1998 to 2008. The concentrations of MSA (an indicator of marine biogenic sulfur production), sulfate, sodium and chloride in samples were analyzed using ion chromatography. Increases in the aerosol MSA concentrations and MSA/nss-SO42− ratios were observed as functions of latitudes in the Pacific Ocean, more abruptly near high southern latitudes as compared to those in high northern latitudes. The MSA concentrations increased from 0.011 μg m−3 near the equator to 0.26 μg m−3 at 63°S, 23°W and from 0.0013 μg m−3 at northern midlatitudes to 0.19 μg m−3 at 58°N, 175°E. However, MSA decreased in the latitudes north of 58°N in the Pacific, where air temperature was lower. MSA/nss-SO42− ratios increased from 0.024 near the equator to 0.93 at 62°S, 4°E and from 0.0031 around northern midlatitudes to 0.39 at 68°N, 169°W. The MSA concentrations were more correlated with MSA/nss-SO42− (R2 = 0.43, n = 60) in Southern Hemisphere than Northern Hemisphere (R2 = 0.091, n = 40). No significant correlation was found between MSA/nss-SO42− and air temperature at high latitudes, indicating latitudinal temperature variations were not a main factor responsible for the MSA/nss-SO42− variation in those regions. Substantial increases in the concentrations of MSA in coastal Antarctica may indicate additional sources of biogenic S besides the emissions of dimethylsulfide from the sea.
 The concentrations of aerosol MSA in high-latitude regions, may reflect the level of the marine biogenic production. The fluctuations in the MSA concentrations in the ice cores have been used as records of marine primary production to investigate climate history [Saigne and Legrand, 1987; Legrand et al., 1991; Saltzman et al., 2006], as high MSA concentrations during the last glacial period suggest an increase in oceanic DMS emissions and this in turn suggests that the S cycle has been sensitive to climate change. However, the MSA concentrations in the marine atmosphere could be affected by multiple processes relating to primary productivity, such as spatial variations of phytoplankton species, air-sea exchange rates of DMS, and different oxidation pathways of DMS. In addition, variations in environmental conditions such as temperatures, precipitation patterns, sea-ice conditions, winds and ocean currents could also affect the concentrations of MSA. All these parameters are intimately linked through the complex interactions of the climate system [O'Dwyer et al., 2000].
 Because of the importance of the S cycles and Antarctica in the global climate system, many efforts were made focusing on understanding sulfur chemistry involving DMS and its effects on the climate in the Antarctic regions; major experiments include: The Sulfur Chemistry in the Antarctic Troposphere Experiment [Berresheim and Eisele, 1998; Minikin et al., 1998], Investigation of Sulfur Chemistry in the Antarctic Troposphere (ISCAT) [Arimoto et al., 2001; Davis et al., 2004], and Antarctic Tropospheric Chemistry Investigation (ANTCI) [Arimoto et al., 2008; Eisele et al., 2008]. Marine biogenic sources dominate the sulfur budget while the contributions from the long-range transport and stratospheric reservoir remains weak in this region [Minikin et al., 1998]. Although MSA and nss-SO42− were correlated based on these studies, the data sets are not sufficiently robust to draw a quantitative conclusion concerning the fraction of marine biogenic nss-SO42− in the region [Arimoto et al., 2001]. The ANTCI mean nss-sulfate (124 ng m−3) and MSA (9.1 ng m−3) concentrations were comparable to those during ISCAT, but the high concentrations of MSA and sodium and high MSA/sulfate ratios in late November/early December indicated pervasive maritime influences during that time [Arimoto et al., 2008]. The modes of nss-sulfate and MSA obtained during the cruise over the Atlantic Ocean to the Antarctic site Aboa in November–December 1999 indicated that aerosol particles were internally mixed [Virkkula et al., 2006]. The ambient conditions, such as temperature, may also affect the observed MSA/nss-SO42− ratios, as reported by Bates et al.  that an inverse correlation between air temperature and MSA/ nss-SO42− ratios was observed during their cruise from 20°N to 60°S in the Pacific Ocean, with the ratios being increased while air temperature decreased toward the high southern latitudes, and that was in agreement with the results of laboratory studies on the MSA-temperature dependent theory. Other studies conducted through research cruises and aircrafts reveal the latitudinal variations of DMS and MSA, showing changes in the atmospheric concentrations of DMS and MSA near the equator and in the higher southern latitudes [Savoie et al., 1992; Huebert et al., 1996; Davison et al., 1996; Dibb et al., 1999; Kettle et al., 1999; Leck et al., 2002].
 These studies provided important results on the processes controlling the formation of S-containing species in the remote marine atmosphere around coastal Antarctica and on the Antarctic Plateau. However, data on the latitudinal distributions of the key S species, in particular the comparison between the sub-Antarctic and the Arctic regions, remain insufficient. Here we present the results of atmospheric MSA, nss-sulfate, and selected inorganic species in aerosol particles collected from eight cruises from Shanghai, China to both Antarctica and the Arctic. We characterized the concentrations of these species in the marine atmospheric boundary layers from 78°N to 70°S. Our goal was to investigate the spatial distributions of these important S-containing species and to explore the possible mechanisms of their formation in the high latitude regions. Results from this work will add new data to the global database of MSA and nss-sulfate. We hope that our results may also encourage more explorations for additional sources of MSA beside the DMS source toward a better understanding of the atmosphere-marine ecosystem interactions.
2. Materials and Methods
 A total of 164 aerosol samples were collected on R/V Xue Long during five Antarctic (ANT) and three Arctic (ARC) cruises during the Chinese National Antarctic and Arctic Research Expeditions (CHINARE). Although the Antarctic cruises normally started in November and ended in early April of the next year, most of aerosol samples on these cruises were collected between December and February (austral summer). The Arctic cruises were carried out during the period of July–September (Table 1 and Figure 1). Thus the results from these cruises allow determine the summer-time characteristics of aerosol distributions over both Antarctica and the Arctic. Bulk aerosol samples were collected by a high-volume bulk sampler Model M241 with wind direction and speed controller (University of Miami, USA) that was installed on the upper deck of R/V Xue Long. Samples were collected on 20 × 25 cm Whatman41 filters for ∼48 h, with a sampling flow rate of ∼1 m3 min−1. A wind direction and speed controller was used during sampling to avoid possible contamination from the ship stack emissions. One sixteenth of the filters was rinsed with 5 ml of deionized water and leached overnight. The cations were analyzed with a CS12A analytical column, and the anions were analyzed with an AG11-HC guard column and an AS11-HC analytical column, using a Dionex DX-500 ion chromatograph (IC) system. Sample injections were done through the use of a 100 μl sample loop. The IC method detection limits are <0.05 mg l−1 for major anions including sulfate and MSA and for major cations including Na+. The blank values of the filters were subtracted from the measurements of air filter samples. The concentrations of nss-SO42− were calculated through the use of the equation: [nss-SO4−2] = [SO4−2]total − [Na+] × 0.25, where 0.25 is the mass ratio of SO42−/Na+ in seawater [Millero and Sohn, 1992].
Time Periods refer to the starting date from and ending date at Shanghai of each cruise. Dates are given as year.month.
3. Results and Discussions
3.1. Latitudinal Distributions of Aerosol MSA and nss-SO42−
 The latitudinal distributions of aerosol MSA obtained during the CHINARE-Arctic Cruises I in 1999, II in 2003, and III in 2008 were similar (Figure 2a). The concentrations of aerosol MSA stayed at low levels (0.0013 to 0.050 μg m−3) from ∼30°N to ∼50°N and reached its highest value of approximately 0.11 μg m−3 at 50°N in Cruise I and 57°N in Cruises II, and 0.19 at 58°N in Cruise III (Figure 2). These results were consistent with previous observations, such as those by Savoie and Prospero  who observed the highest MSA concentration of 0.097 μg m−3 at Shemya (52°N) among several islands in the North Pacific. As the exclusive source of MSA, DMS also showed high concentrations at the sea surface in this region reported by Savoie and Prospero  and re-confirmed later by Arimoto et al.  and Kettle et al. .
 The MSA concentrations, however, had different latitudinal patterns in the southern hemisphere. The MSA concentrations were 0.011 μg m−3 near the equator and increased to 0.16 μg m−3 at ∼55°S, followed by a sharp increase. The MSA concentrations reached the highest levels near the Antarctic continent (60–70°S), which is consistent with recent results from the cruise conducted during Nov 2010–Feb 2011 (Y. Gao, unpublished data, 2012). The maximum MSA value of 0.26 μg m−3 was observed on December 29 at ∼63°S, 23°W during ANT 18th cruise. The next highest value of MSA, 0.24 μg m−3, was observed on January 22 at 68°S, 18°W during the ANT 21st cruise. Both occurred in mid austral summer. The air mass back trajectory analyses indicated that the air masses arriving at the above locations on those dates were from the western Antarctic coastal regions, in particular the Antarctic Peninsula. Western Antarctica is undergoing dramatic climate changes, with temperature being rising during the past half century, resulting in the shortened sea ice seasons and accelerated glacier losses [Rignot et al., 2008; Jenkins et al., 2010]. Biological processes are often enhanced, associated with the melting of the ice, as seen with the elevated levels of chlorophyll in the surface waters observed via satellite imagery in polynya areas [Marrari et al., 2008]. Therefore the observed high MSA concentrations in the mid-austral summer in western Antarctica from this study may reflect the enhanced marine biological activities driven by the environmental changes in that region. However, MSA concentrations as low as 5 ng m−3 were found over Antarctica's marginal seas in the CHINARE-Antarctic cruises. These results are consistent with the cruise from UK to Antarctica in October 1992 to January 1993 where the maximum MSA concentration varied from 2 ng m−3 to 0.36 μg m−3 near 75°S [Davison et al., 1996]. However, during the cruise from USA to Antarctica in 1989, the concentrations of MSA varied from 3 ng m−3 to 0.07 μg m−3, with the highest concentrations occurring at 15°N and from 30° to 40°S, while at latitudes south of 40°S, MSA slightly decreased and reached 0.03 μg m−3 near 60°S [Bates et al., 1992]. The maximum concentrations of MSA were also observed at Dumont D'Urville (66.39°S, 140°W), Halley (75.35°S, 26.34°W) and Neumayer (79.40°S, 8.16°W), approximately 0.075, 0.15 and 0.35 μg m−3, respectively [Minikin et al., 1998]. The differences of MSA between these stations might be due to the variations of different factors such as biological productivity, climatology of the regions, sea ice coverage, etc [Minikin et al., 1998]. The MSA maximum from the CHINARE cruises was within the maxima observed at these stations.
 The concentrations of nss-SO42− stayed at relatively low levels during the cruises in this study, with >80% of the values below 1 μg m−3 and 16% of the rest of the values ranging from 1 to 5 μg m−3 in both hemispheres (Figure 2b). In the northern hemisphere, however, the highest values of nss-SO42− (22, 21 and 12 μg m−3) which are not included in Figure 2b, were observed near the starting pier (Shanghai, China). Results of air mass back-trajectory analyses clearly showed that the air masses affecting these high nss-SO42− values passed over urban regions in northeastern China, suggesting the influences of pollution emissions in land on the nss-SO42− levels in the coastal marine atmosphere. Earlier studies also revealed that the maximum atmospheric nss-SO42− concentrations exceeded 40 μg m−3 at coastal Qingdao in East China Sea and 20 μg m−3 at coastal Xiamen in South China [Gao et al., 1996]. These values agreed with the results from recent studied in Shanghai city and the North Yellow Sea [Wang et al., 2006; Yang et al., 2009], reflecting the contributions of pollution emissions in the continent on the nss-SO42− concentrations in the marine atmosphere in this region, although the DMS emissions in China's marginal seas could also be a source for nss-sulfate in the region [Hu et al., 2003]. The nss-SO42− concentrations decreased as the Xue Long Vessel moved away from the China mainland. The other high nss-SO42− concentrations during the Arctic cruises were 5.6 μg m−3 observed during Arctic I and 5.0 μg m−3 observed during Arctic II around 64–65°N at the entrance of the Bering Sea toward the Chukchi Sea. These high values of nss-sulfate concentrations at these areas were affected by air masses passing over North America as identified by air mass back trajectory analyses, and these results also suggest the influences of anthropogenic emissions in the continents on the atmospheric concentrations of nss-sulfate in Northern Hemisphere. These results from this study are consistent with both earlier and recent studies. Langner et al.  used a global atmospheric transport–chemistry model to estimate the changes in the distributions of tropospheric sulphate aerosol occurred since pre-industrial times, and found human activities have increased global emissions of sulphur gases during the past century, leading to increased sulphate aerosol concentrations, mainly in the Northern Hemisphere. They noticed that the increase in sulphate aerosol concentration was small over the Southern Hemisphere oceans. Quinn et al.  showed the long range transport of pollution associated with Arctic Haze based on their 30-years aerosol measurements conducted at Barrow, Alaska, and indicated that the anthropogenic emissions in East Asia and North America have already had long lasting impacts on the atmospheric composition in Northern Hemisphere. Because a fraction of nss-sulfate in the Northern Hemisphere comes from anthropogenic emission that is different from the Southern Hemisphere, this may contribute to the different MSA/nss-sulfate ratios in the two hemispheres. Our back-trajectory analyses results also indicate that in Northern Hemisphere in general, the air masses reaching the ship mainly came from East Asia for low and midlatitudes and from Eurasia or the Arctic Ocean for high latitudes, which is different from those in southern hemisphere that were mainly from the oceans or the Antarctic continent. These differences in air mass source regions could also contribute to the observed latitudinal variations of nss-sulfate.
3.2. Distributions of MSA/nss-Sulfate Ratios
 The MSA/nss-SO42− ratios from the two hemispheres obtained in this study showed different patterns (Figure 2c). In the southern hemisphere, the patterns of MSA/nssSO42− ratios were similar to those of MSA (Figure 2a). The MSA/nss-SO4−2 increased slowly from the tropic region to the midlatitude, and approximately at 50°S, the ratios increased dramatically. The MSA/nss-SO42− ratios were approximately 0.0024–0.06 in the tropical regions and 0.06–0.12 in the unpolluted midlatitudes and increased sharply near coastal Antarctica, ranging from 0.15 to 0.93. The highest ratio, 0.93, was observed at 62°S, 4°E around 31 December 2001 during the ANT 18th cruise, while the concentration of nss-SO42− was 0.19 μg m−3 that was at the lower end of the nss-SO42− concentration range. Based on the air mass back-trajectory analyses, the aerosol sample with this low nss-SO42− concentration was affected by air masses from the inland of Eastern Antarctica, suggesting the minimum production of sulfate aerosol from marine biogenic sources in the region. The latitudinal variations of this ratio were similar to those by other studies at different oceanic regions, with higher values at high latitudes [Berresheim, 1987; Bates et al., 1992; Savoie et al., 1992; Davison et al., 1996; de Mora et al., 1997; Legrand and Pasteur, 1998; Gondwe et al., 2004]. Results from Mawson, Antarctica (67°36′S, 62°30′E) from February 1987 to October 1989 showed that MSA/nss-SO42− was about 0.31 (about five times higher than that observed over the tropical and subtropical oceans), with DMS being considered as the main source for MSA and nss-SO42− [Savoie et al., 1992]. In the cruise from the eastern Pacific Ocean to the Antarctic in 1989, the molar ratio of MSA/nss-SO42− ranged from 0.2 near the equator to 0.32 at 55°S. The pattern repeatedly observed in the CHINARE-Antarctica cruises confirmed a unique feature in spatial variations of atmospheric MSA/nss-SO42− ratios.
 In the northern hemisphere, although the increases in MSA/nss-sulfate ratios were found in high latitudes, the increase in this ratio from 0.0031 at midlatitudes to the maximum of 0.39 (68°N, 169°W) were not as dramatic as those found in the high latitudes of Southern Hemisphere. This pattern of relatively low MSA/nss-sulfate ratios observed during the CHINARE Arctic cruises could be attributed to the influence of air masses containing anthropogenic sulfate from nearby continents. The lowest MSA/nss-sulfate ratio of 0.0051 among all air samples collected during the Arctic cruises was associated with the air sample collected near 60°N, 177°E toward the end of August 1999, corresponding to the highest nss-sulfate concentration of 5.6 μg m−3. Air mass back trajectory analyses indicate that this sample was impacted by air masses passing over Russia on the east of Bering Sea and the Bering Strait, suggesting the anthropogenic contributions from the continents to the observed high nss-sulfate concentration at this location. Another low MSA/nss sulfate ratio of 0.013 from the Arctic cruise air samples was also associated with high nss-sulfate concentration of 5.0 μg m−3 from the sample collected near 59°N, 175°E in September 2004 with similar continental impacts as indicated by air mass back trajectory analyses. This finding from the CHINARE Arctic cruises is consistent with the results from several recent studies in the Arctic [Douglas and Sturm, 2004; Quinn et al., 2009; Fisher et al., 2011], showing that the nss-sulfate over the Arctic was often influenced by contributions from continental sources. On the other hand, the high values of MSA/nss-SO42− observed during the CHINARE Arctic cruises were different from those found in other oceanic regions in Northern Hemisphere. Over the North Atlantic, the ratios of MSA/nss-SO42− were 0.053 at Barbados (13°N, 59°W), 0.051 at Bermuda (32°N, 65°W), and 0.33 at Mace Head, Ireland (53°N, 10°W) [Savoie et al., 2002]. At Alert, Canada (83°N, 62°W), the MSA to nss-SO42− ratios ranged from 0.006 to 0.02, with the highest value occurring in summer, similar to the results observed over the midlatitude regions [Li and Barrie, 1993; Li et al., 1993]. On the other hand, the patterns of MSA/nss-SO42− ratios and of MSA concentrations observed during CHINARE-Arctic cruises were similar (Figures 2a and 2c), with the highest MSA/nss-SO42− occurring from 40°N–68°N where sea surface DMS and air MSA concentrations were highest [Kettle et al., 1999].
 During the three CHINARE-Arctic cruises there were no strong correlations (R2 = 0.091, n = 40) between the MSA concentrations and MSA/nss-SO42− ratios. However, a stronger correlation (r2 = 0.43, n = 60) between the two was observed in the southern hemisphere during five CHINARE-Antarctica cruises (Figure 3). Although the variations of aerosol nss-SO42− in the northern hemisphere are influenced by more complex sources as compared to the southern hemisphere, the concentrations of nss-SO42− at the high latitudes of both hemispheres from this study varied, with higher values being associated with air masses of continental origins. After removing the nss-SO42− values associated with air masses affected by anthropogenic sources, the variation in the MSA concentrations seems to be the major cause for large differences in the ratios of MSA/nss-SO42− between the two hemispheres. Possibly the higher MSA concentrations near Antarctica may lead to the higher MSA/nss-SO42− ratios observed in that region, although this is not a simple linear relationship. The three lowest MSA/nss-SO42− ratios, 0.059 from the ANT 16th cruise, 0.040 from the ANT 18th cruise, 0.054 from the ANT 19th cruise, were observed in the Antarctic coastal waters, and all were associated with relatively low MSA concentrations (0.033, 0.023, 0.024 μg m−3, respectively). This result suggests the important role of MSA in shaping the spatial variation in MSA/nss-SO42− in the southern hemisphere. This result is consistent with the results by Arimoto et al.  who found that it is MSA more than sulfate that drives the spatial differences in this ratio.
3.3. Factors Influencing the Distributions of MSA and MSA/nssSO42− Ratios
 The significantly high ratios of MSA/nss-SO42− found near coastal Antarctica have been considered mainly to be due to the oxidation of marine biogenic DMS at low temperatures [Hynes et al., 1986; Turnipseed et al., 1996; Barnes et al., 2006; Ramírez-Anguita et al., 2009]. The high strength of the DMS source and the temperature-dependent kinetics of MSA formation have been used to explain not only the DMS - MSA relationships but also the inverse relationship between MSA concentration and the ambient temperature, as lower temperature may lead to higher yield of MSA [Bates et al., 1992; Arsene et al., 1999]. Hence, at certain temperatures, the ratios of MSA/nss-SO42− may probably remain at certain values. Since the temperature patterns of the northern and southern hemispheres are similar as shown in Figure 4, similar patterns of MSA/nss-SO42− should be expected in both hemispheres if DMS is the main source of both MSA and nss-SO42−. However, neither the MSA concentrations nor the MSA/nss-SO42− ratios showed similarities with air temperature in the two hemispheres. In the Arctic regions, the highest MSA values shown in Figure 2c were not associated with the lowest temperature shown in Figure 4, and the correlation between MSA and air temperature was weak, with R2 being equal to 0.033. The high values of MSA concentrations occurred in the areas with high DMS emission at the sea surface, but these values were not related to the low air temperature while aerosol sampling took place. During the ANT cruises, the MSA/nss-SO42− ratios varied from 0.024 to 0.93 in the high southern latitudes (Figure 2c), but strong relationships between MSA concentrations or MSA/nss-SO42− and air temperature during aerosol sampling were not found (Figure 5). For example, during the 18th CHINARE-Antarctica cruise, the sampling temperature near the maritime Antarctic normally ranged from −5°C to 0°C to (Figure 4), with MSA/nss-SO42− varying from 0.13 to 0.94 (Figure 2c). The maximum value of MSA was obtained at −2.00°C while the maximum MSA/nss-SO42− ratio of 0.94 was obtained at −1.36°C, and both were not associated with the lowest temperature (−5°C) (Figures 4 and 5). The correlations were R2 = 0.18 between MSA and air temperature and R2 = 0.099 between MSA/nss-SO42− and air temperature for the ANT-SH data, indicating weak relationships among them. On the other hand, as the air temperature along the paths of air masses indicated by trajectory analyses fluctuated only within a range of −5°C to 0°C for most of samples collected in coastal Antarctica based on which this MSA/nss- SO42− temperature relationship was addressed, such small variation in air temperature may not be sufficient to observe a significant correlation with air temperature. Thus the air temperature variation within such a scale did not appear to be a main factor influencing the changes of MSA/nss-SO42− ratios in coastal Antarctica.
 The robustness of the theory of the MSA production mechanism at low temperature has been questioned by other studies. The DMS-MSA relationship in the high latitudes in the southern hemisphere is not linear. Prospero et al.  found that nss-SO42− peaks in the December–January period, and MSA reaches a maximum a little later and then subsequently drops rapidly; the highest ratio of MSA to nss-SO42− molar ratios appeared around the March/April equinox period. Wagenbach  also observed abrupt changes in the MSA/nss-SO42− ratio during summer but with no concomitant changes in temperature in coastal Antarctica. The MSA/nss-SO42− ratios in coastal Antarctica during the summer months were 3–10 times higher than those obtained on the Antarctic plateau where the temperature is much lower than coastal Antarctica [Wagenbach, 1996; Legrand and Pasteur, 1998; Arimoto et al., 2001], although this may be attributed to possible processes where aerosol MSA tends to be more rapidly depleted than nss-SO42− during the transport from the coast to the plateau. High summertime DMS concentrations have been observed in seawater and marine atmosphere near coastal Antarctica due to the retreat of the sea-ice and the blooming of algae (especially DMSP-rich species) [Kettle and Andreae, 2000; Scarratt et al., 2002], which may give high production of MSA. However, as described by Kleefeld , the DMS concentrations at Neumayer were enhanced by one order of magnitude from December to January, while nss-SO42− and MSA increased by only ∼30% and a factor of 2, respectively. Thus the overall nss-SO42− and MSA levels observed in summer at coastal Antarctic sites may not be necessarily controlled by local DMS emissions and ambient temperature [Kleefeld, 1998; Minikin et al., 1998]. These results with the MSA/nss-SO42− ratios consistently showing summer maxima and winter minima raised questions about the robustness of the theory of the MSA production mechanism of MSA being preferentially produced under low temperatures. On the other hand, diminished marine biological activity and low seawater DMS conditions in winter have widely been cited as the cause of this observed trend. Gondwe et al.  proposed that as the photochemical production of hydroxyl radical (OH) during the dark winter at polar latitudes is non-existent, reduced wintertime oxidation of DMS by OH to form MSA results in summer maxima and winter minima in MSA concentrations in these regions. In addition, in the cruise from USA to the Antarctica in 1989, no significant spatial correlation was found between the latitudinal distributions of DMS, SO2, MSA, and nss-SO42− [Bates et al., 1992]. A similar conclusion was drawn from the study of another cruise carried out from UK to Antarctica in 1992 that good correlations did not exist between individual DMS-MSA data pairs, although broad agreements were seen in the atmospheric concentrations of DMS and MSA [Davison et al., 1996; O'Dowd et al., 1997].
 Results from this study and previous studies suggest that the DMS source and temperature-dependent kinetics of MSA formation cannot fully explain the variations of the MSA distributions, particularly while comparing the data from the maritime Antarctic with those in the Arctic. The production of aerosol MSA seemed to be influenced by more complex factors in addition to air temperature, such as the atmospheric concentration of DMS, the DMS oxidation rate, factors controlling DMS emissions from the sea surface, etc. In fact, it is even difficult to justify a linear relationship between air temperature and MSA/nss-sulfate ratio as this ratio could be affected by anthropogenic emissions, particularly in the northern hemisphere, or between DMS and MSA as the concentrations of MSA and nss-sulfate from DMS are influenced differently by the production and oxidation rates of a variety of intermediate reaction products, including dimethyl sulfoxide (DMSO), dimethyl sulfone (DMSO2), methane sulfinic acid (MSIA), methylsulfenic acid (MSEA), and SO2; at present, it is difficult to evaluate the roles of these intermediate products on the reactions that produce MSA and nss-sulfate [Davis et al., 1998; Lucas, 2003]. In addition to the competition between OH addition to and hydrogen abstraction from DMS, the decomposition of CH3SO2 and reaction to form CH3SO3 have been proposed to be an additional factor controlling the observed temperature dependence of the MSA/nss-sulfate ratios. On the other hand, the production rates of DMS are dependent of different groups of phytoplankton and could be affected by different regional nutrient levels [Turner et al., 2004], and these may also contribute to the different distributions of the MSA and nss-sulfate concentrations and MSA/nss-sulfate ratios.
3.4. Alternative Source of MSA Near Antarctica
 For sources of S-containing species in the Antarctic coastal air, in addition to the emission of marine DMS, volcanic emissions, etc., contributions from sea animals, in particular penguins, to the sulfur cycles near coastal Antarctica are possible. Xie et al.  detected certain organosulfur compounds in the volatile components of fresh penguin feces, including dimethyl trisulfide (DMTS), dimethyl tetrasulfide (DMTTS), and dimethyl pentasulfide (DMPS), and their results indicated that organosulfur compounds from fresh penguin feces had been emitted into the air, contributing to the S-containing species over Antarctica. They further estimated that 5.5 × 10–5 nmol m−3 sulfur emission to the atmosphere could be inferred from each penguin, accounting for 5–15% of the total Nss-sulfur emissions near Zhong Shan Station (69°S, 76°E) in Antarctica [Xie et al., 2002]. Legrand et al.  observed that near Dumont D'Urville on Ile des Petrels near coastal Antarctica, a place with a penguin population of 12,000, the concentrations of nss-SO42− were 3.77 nmol m−3 for a downwind location and 3.11 nmol m−3 for an upwind location. In this study, two of the three highest MSA concentrations, 0.26 μg m−3 observed during the ANT 18th cruise and 0.24 μg m−3 observed during the ANT 21th cruise in austral summer, were affected by air masses originating from sub-Antarctica and Antarctic Peninsula (Figure 6), where large populations of Antarctic penguins species live. These results support the idea that Antarctic penguins may contribute to atmospheric S-containing compounds in the region. The breeding seasons of sea animals are broadly consistent with those of algae blooms, and therefore the contributions from penguins to the MSA concentrations in the air may need to be considered.
 Large population of seabirds, in particular penguins, is a unique feature in the Antarctic ecosystem that is different from that of the Arctic [Woehler and Croxall, 1997], and this ecological difference may partially explain the different levels of atmospheric MSA in the two high latitude regions. The volatile organosulfur compounds, such as DMDS, DMTS, DMTTS, and DMPS, in the feces of penguins in coastal Antarctica as detected by Xie et al. , which have higher S content in their molecules than that of DMS, may be converted to MSA in some ways, and might contribute to the observed high MSA levels in aerosols over coastal Antarctica. At present, however, the reaction pathways for MSA production from sea animal feces are uncertain and investigations focusing on the production pathways from these organosulfur compounds to MSA are few. Therefore, additional field work and laboratory kinetic studies relating to possible marine animal sources for MSA should be explored to better understand the sulfur chemistry in coastal Antarctica. Additionally, certain environmental factors, such as winds, sea ice coverage, and light intensity, etc., may also contribute to the variability of S-species in the high latitude marine atmosphere and should be considered through future studies.
 We thank the Chinese Arctic and Antarctic Administration (CAA) of State Oceanic Administration (SOA) and the crew of R/V Xue Long for support with field operation. We thank Junying Sun for sharing of the CHINARE-Arctic I data. This research was jointly sponsored by National Natural Science Foundation of China (NSFC) (40671062 and 41106168), State High Technique Research Development Project (2008AA121703), Ministry of Science and Technology of China (MOST) (2004DIB5J178 and 2009DFA22920) and Chinese Arctic and Antarctic Administration (CAA) cooperation program (IC201114 and IC201201), Scientific Research Foundation of Third Institute of Oceanography, SOA (2007015) and the U.S. National Science Foundation Award 0944589. We thank three anonymous reviewers whose constructive comments helped significantly improve this paper.