Seasonal and geographic variations of methanesulfonic acid in the arctic troposphere

Atmospheric Environment Vol. 27A, No. 17/18, pp. 3011-3024, 1993.

0004-6981/93 $6.00+0.00 ((:) 1993 Pergamon Press Ltd

Printed in Great Britain.

SEASONAL A N D GEOGRAPHIC VARIATIONS OF METHANESULFONIC ACID IN THE ARCTIC TROPOSPHERE S.-M. LI a n d L. A. BARRIE Atmospheric Environment Service, 4905 Dufferin Street, Downsview, Ontario, Canada R. W. TALBOT a n d R. C. HARRISS Institute for the Study of Earth, Oceans and Space, Complex Systems Research Center, University of New Hampshire, Durham, NH 03824, U.S.A. and C. I. DAVIDSON a n d J.-L. JAFFREZO Department of Civil Engineering,Carnegie Mellon University, Pittsburgh, PA 15213, U.S.A. (First received 14 February 1992 and in final form 2 April 1993)

Abstract--Measurements in the Arctic troposphere over several years show that MSA concentrations in the atmospheric boundary layer, 0.08-6.t parts per trillion (ppt, molar mixing ratio), are lower than those over mid-latitude oceans. The seasonal cycleof MSA at Alert, Canada (82.5°N,62.3°W),has two peaks of 6 ppt in March-April and July-August and minima of 0.3 ppt for the rest of the year. At Dye 3 (65°N, 44°W)on the Greenland Ice Sheet, a similar seasonal MSA cycle is observed although the concentrations are much lower with a maximum of I ppt. Around Barrow, Alaska (71.3°N, 156.8°W),MSA is between 1.0 and 25 ppt in July, higher than 1.5_+1.0 ppt in March-April. The mid-tropospheric MSA level of 0.6-1 ppt in the summer Arctic is much lower than about 6 ppt in the boundary layer. At Alert,the ratio of MSA to non-sea-salt (nss) SO2- ranges from 0.02 to 1.13and is about l0 times higher in summer than in spring. The summer ratios are higher than found over mid-latitude regions and, when combined with reported sulfur isotope compositions from the Arctic, suggest that on average a significantfraction (about 16-23%) of Arctic summer boundary layer sulfur is marine biogenic. The measurements show that the summer Arctic boundary layer has a significantlyhigher MSA/nss-SO2- ratio than aloft. Key word index: Aerosol, Arctic, biogeochemicalcycle, Greenland ice, methanesulfonicacid, organic sulfur.

INTRODUCTION Atmospheric organic sulfur compounds are important in the understanding of climate change, atmospheric chemical processes, and glacier chemistry. Much has been published on the oceanic emissions of organic sulfur compounds, most prominently dimethylsulfide (DMS), and their effects on the global atmospheric sulfur budget (Andreae and Raemdonck, 1983; Andreae et al., 1985; Bates et al., 1990; Sze and Ko, 1980). Models have shown that organic sulfur emissions have a major impact on the sulfur budget of the marine troposphere (Toon et al., 1987). Laboratory studies indicate that the oxidation of DMS in the atmosphere leads to various end products (Barnes et al., 1987; Daykin and Wine, 1990; Grosjean, 1984; Hatakeyama et al., 1982, 1985; Jensen et al., 1991), of which particulate methanesulfonic acid (MSA) and H2SO 4 are predominant. These products can serve as cloud condensation nuclei and may be crucial to a climate feedback mechanism involving cloud albedo (Charlson et al., 1987; Bates et al., 1987). In addition, DMS plays a role in atmospheric oxidant chemistry.

Recent studies (Chatfield and Crutzen, 1990; Platt et al., 1990; Yin et al., 1990) raised the possibility of interactions of DMS with radicals in the marine atmosphere and thereby production of a large number of chemical species including radicals, thus affecting/~he atmospheric chemistry of these species. For instance, Plattet al. (1990) showed that under nighttime conditions, the reaction of the NO3 radical with DMS can potentially generate a series of peroxy radicals that would otherwise be present only during daytime. MSA is therefore of interest as a tracer of DMS oxidation products (Andreae et al., 1988; Ayers et al., 1986, 1990; Berresheim, 1987; Saltzman et al., 1983, 1985, 1986) and can serve as a diagnostic tool for models of atmospheric oxidation pathways of DMS (Chatfield and Crutzen, 1990). In addition, measurements of MSA and other sulfur species in polar regions are needed to help understand the global distribution and budget of sulfur and to determine the overall transfer function of those compounds to the deep ice in glaciers (Legrand et al., 1991; Whung et al., 1989). However, natural cycles are disturbed in the Northern

3011

3012

S.-M. LI et al.

Hemisphere where it has been suggested that the impact of anthropogenic inputs may explain the decrease of Greenland ice core MSA/nss-SO~- ratios over the past 100 years from 0.17 to 0.05 (Whung et al., 1989). Furthermore, it is well-known that anthropogenic sulfate comprises much of the total SO~- in the Arctic, at least during winter/spring seasons (Barrie and Hoff, 1985). Good measurements of seasonal and spatial variations of Arctic atmospheric MSA concentrations and the MSA/nss-SO]- ratios, coupled with an analysis of potential source regions, are needed to determine how the global sulfur cycle is affected by the coupling between the mid-latitudes and the Arctic regions, and to determine whether patterns in MSA concentrations observed in the ice cores are representative only of local conditions or are indicators of a larger region. Very little is known about atmospheric MSA and other organosulfur compounds in the Arctic. Spring MSA concentrations at Barrow, Alaska, were shown to be similar to remote oceanic levels in the Pacific (Li and Winchester, 1989). However, to date, seasonal variations, geographic distributions, and vertical profile are not known. In this paper, we report measurements of aerosol MSA in the Arctic from different seasons at different locations, thereby providing for the first time its spatial and temporal occurrence in this region.

samples for ATM 90, June-August 1990(Jaffrezo and Davidson, 1992). The location of the study sites, areas, and flight paths are shown in Fig. 1. The ABLE3A samples were analysed within 24 h of collection at field laboratories operated at Barrow and Bethel, Alaska (Talbot et al., 1992), while samples from Alert, Barrow, and Greenland were shipped to three central laboratories for extraction with deionized water and for analysis for methanesulfonate ion by ion chromatography.

RESULTS

Sea-level sites

The year long experiment at Alert shows a typical pattern with a strong seasonal variation in MSA concentrations with two broad peaks (Fig. 2a). A spring peak occurs during April-May, with a gradual increase from 1 ppt in April to a peak of 6 ppt in May. A summer peak between June and October has a concentration of 2-6 ppt. For the rest of the year from October through March MSA remains low in concentration at about 0.3 ppt. Based on measurement results of long term records obtained recently, it is found that this dual-peak feature in the MSA seasonal cycle at Alert is persistent from year to year (Li and Barrie, 1993; Li et al., 1993). Ground-level atmospheric MSA observations for 3 other sea-level sites north of 50°N latitude are shown in Fig. 3. They all show similar patterns with some variation in timing and concentration level. Note that MSA seasonal patterns at Alert are similar to those at EXPERIMENTAL Shemya Island (52°N, 174°E) in the Aleutian Islands The data were obtained at several locations in the Arctic. (Saltzman et al., 1986; Savoie and Prospero, 1989) and They essentially fall into four groups: (i) those collected by in the North Atlantic as well as on the southern coast high-volume samplers at Alert, Northwest Territories on of the United Kingdom (50.3°N, 4.1°W) (Watts et al., Whatman 41 filters, (ii) those collected at Barrow Alaska on 47-ram Nuclepore filters, (iii) those collected during the 1987, 1990) (Fig. 3). They have a minimum in winter NASA GTE ABLE3A aircraft mission on 90-mm Zcfluor and a broad maximum in summer. Consistent with Teflon filters, and (iv) those collected on the Greenlarld Ice these, Prospero and Savoie (1990) recently reported an Sheet, both during the Dye 3 Gas and Aerosol Sampling MSA seasonal cycle at Mace Head, Ireland, with a Program (DGASP) and during the first two seasons of the peaking concentration of 70 ppt (300 ng m - a) in June. ATM Program; collections in Greenland were performed with high volume sampling on 90 mm Zefluor Teflon filters. However, there are marked differences in detail among At Alert, two sets of observations were taken. One was a the sites. While winter MSA concentrations are simyear-long series (May 1986 to April 1987) of weekly samples ilar at all four sites (sub-ppt levels), the summer which are routinely collected and analysed for many aerosol concentrations are significantly different with the constituents (Barrie and Hoff, 1985; Barrie and Barrie, 1990). The second was a series of 70 daily samples taken from highest at Shemya (60 ppt) and Mace Head (70 ppt) February to April 1988 (Bottenheim et al., 1990). At Barrow, and the lowest at Alert (2-6 ppt). Furthermore, there Alaska, from March to April 1989, 70 12-h aerosol samples are double MSA peaks in May and July at Alert, in were collected with an experimental setup similar to the contrast to a single peak in July at Shemya Island and previous study (Li and Winchester, 1989). During the NASA in June at Mace Head. In the North Atlantic, two GTE ABLE3A aircraft mission in the summer of 1988 (Harriss et al., 1992), aerosols were sampled during horizon- MSA peaks are observed in May and September, tal flight legs of 30-60 min duration over the adjacent respectively (Watts et al., 1990). Beaufort Sea areas of the Arctic, the Bering Sea, and during On a finer time resolution, daily MSA observations transit flights between Barrow and Thule, Greenland (Talbot obtained at Alert during February-April 1988 (Fig. et al., 1992). At Dye 3, on the Greenland Ice Sheet (65°11'N, 43°50'W, elevation 2.5 km) a total of 89 samples of 24-48 h 4A) show similar concentrations and temporal proeach were collected from August 1988 to July 1989 during the grcssion as that revealed by the weekly samples from year long DGASP experiment (Jaffrezo and Davidson, 1993; the previous year (Fig. 2A). Prior to spring equinox, 21 Jaffrezo et al., 1993a). Two other sets of samples are also March (Day 80 of Year), the MSA concentration presented from the ATM program during the summers of averaged 0.44_+ 0.20 ppt, compared to 1.7 + 0.6 ppt 1989 and 1990 at Summit on the Greenland Ice Sheet (72°20'N, 38°45'W, elev. 3.2 km). They include 12 samples of after equinox. Two processes can lead to the gradual 24-48 h each for ATM 89, June-July 1989, and 49 daily increase: the increased photochemical reactivity in-

Seasonal and geographic variations of methanesulfonic acid

3013

Fig. 1. Sampling locations in the Arctic. E denotes sampling site in England. The 1988 NASA GTE/ABLE3A field study areas and transit flights are given by the shaded areas and the dashed lines. The mean positions of the Arctic Front in winter (January) and summer (July) are shown by the solid lines. duced by sunlight at polar sunrise, and the increased transport of DMS/MSA from lower latitudes at this time. Evidence of the first process is found when the daily MSA concentrations were plotted against measurements of global irradiance taken periodically at this site (Barton et al., 1988). Figure 4A shows a significant correlation between the two. Furthermore, the very smooth increase in MSA concentrations at Alert is consistent with the concept of a reservoir of DMS that becomes increasingly converted to MSA over time as insolation increases. One might expect more variable concentrations due to episodic transport if the second process were dominant. Nevertheless, many studies have shown the importance of longrange transport of anthropogenic species from the mid-latitudes during the winter and spring (e.g. Barrie, 1986), and hence the second process cannot be ruled out. Elsewhere in the Arctic at sea level, year-round MSA measurements have not been performed, but summer and spring data are available from the short term aircraft and ground-level studies at Barrow, Alaska. MSA concentrations near Barrow were between 1.0 and 25 ppt at 0.15 km altitude in July 1988 and 1.5__+1.0 ppt at ground level in March-April 1989.

To compare with the data from the other northern sites in Fig. 3, the four Barrow monthly averages, also plotted in Fig. 3, are calculated from the measurements made in 1986 (Li and Winchester, 1989) and 1989 at ground level and in 1988 by aircraft at 150m altitude during ABLE3A. The data from Barrow indicate that there is an increasing trend in MSA concentration in spring from March to May, which could extend to mid-summer. The summer MSA levels at Barrow are higher than the corresponding concentrations at Alert in 1986, whereas the spring MSA levels are similar at both locations. However, in contrast to Alert, the semi-daily MSA observations at Barrow in spring 1989 do not show a smoothly increasing trend (Fig. 4B). There is no obvious monotonic increase in MSA at polar sunrise; rather, the data show many short episodes of elevated concentrations. Furthermore, its concentrations are more variable on an inter-annual basis than at Alert. The average of 1.5__+1.0 ppt MSA in March-April 1989 is contrasted by the average of 2.9+0.5 ppt in March-May 1986 (Li and Winchester, 1989). This inter-annual difference is statistically significant and could reflect the geographic location of Barrow close to the Arctic Front (Fig. 1), as further discussed below.

3014

S.-M. L[ et al.

6.5

MSA (ppt)

3.9

2.6

1.3

0 May

Jun

Jul

Aug

Sep

Oct

Nov

Dee

Jan

Feb

Mar

0.5}

kpr

MSA/nssS04=

(s)

0.4 .............................................................................................. 0.3

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

0o~. ....................................................................................... 0"10 May

Jun

Jul

Aug

Sep

Oct

1986

Nov

Dec

~an

Feb

Mar

Apr

1987

Fig. 2. (A) Seasonal MSA variation at Alert, NWT from April 1986-April 1987.(B) Seasonal variations in MSA/nss-SO2- ratio (R) at Alert, NWT during the same period.

Table 1 also presents MSA observations made in other parts of the world. It can be seen that concentrations found in the Arctic are generally lower than those over mid-latitude ocean, and that there are also distinct seasonal variations. These are discussed further below. Sampling above sea level: aircraft and Greenland ice sheet measurements Aircraft observations from the NASA ABLE3A study in summer 1988 yield information on MSA vertical profiles over the Arctic Ocean. Table 1 shows that MSA concentrations in the Arctic boundary layer

and at ground level are 1-3 ppt. These are lower than to those at mid-latitude marine locations (typically 5-10 ppt). Free tropospheric levels in the Arctic are smaller still, often less than 1 ppt. MSA concentrations aloft are also spatially less variable than in the boundary layer. Over the Greenland summit (alt. 3048 m), at 0.87 + 1.3 ppt, MSA is a factor of five lower than at Alert during summer, while nss SO 2- decreases only by a factor of two. This vertical pattern of MSA is probably linked to the surface emission of DMS, its subsequent oxidation to MSA at sea level, and the upward mixing of MSA partially prevented by its uptake by cloud as cloud condensation nuclei (CCN)

Seasonal and geographic variations of methanesulfonic acid

60

50

3015

MSA (ppt)

I~ ~O'~rrO'..................~

.............................................. A

40

30

20

10

l

2

3

4

5

8

7

8

9

lo

11

12

Month Fig. 3. Comparison of MSA seasonal variations based on monthly means at Shemya, North Atlantic, Alert, and Barrow. For Barrow MSA means, data from 1986 and 1989 at ground level and 1988 aircraft flights are used. See text for references.

4

MSA (ppt) z' /

(A) Mort, Nrl' 1988

,5

J

•

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

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

, . . , . o . ~. . . . . . . . ~

•. ..................

tt

tit:

'i;i:i;iririi;r, , ;irrir;, .,,.T. .Jl"itt'tttll'l!'eiltlt't['t[t{t't['; .......

0

50

54

58

62

66

70

74

78

82

86

90

94

98 102 106 110 114 118

94

98 102 108 110 1[4 118

Day of Year MSA (ppt) 5

0 50

54

58

62

66

?0

74

78

{}2 86

90

Day of Year F i g 4 (A) Polar sunrise MSA at Alert, NWT, in spring 1988 Inserted is the scatter plot of MSA vs available concurrently measured 24-h mean global irradiance at Alert in 1988 (B) MSA at Barrow, Alaska in spring 1989 Local Barrow pollution sources have no apparent influence on the MSA observations at the N O A A - G M C C site

3016

S.-M. Ll et al.

Ox

C5 C5 C5 C5

C~

~

C~ C3

~

C~ C3 C~

0

.<

t~ e~

+

I

~

+I

~

t~

o ,z3

N

e~ 0

M

.13 0 b. 0

~0

c~i 0 o

•

~

~

~'~

~

~

~

~

~

"~o ~

~o~ ~ . ~~ ~~,-"~ o .

"-~.~. ~ o t~-.~

~ ~.~

Seasonal and geographic variations of methanesulfonic acid or by scavenging. For example, over the pack ice near Barrow, at 2 - 6 k m altitude MSA averaged 0.65 + 0.77 ppt, much lower than its average concentration of 5.3+8.5 ppt (range 1.0-25 ppt) at 0.15 km altitude in the boundary layer below the thick stratus cloud deck that commonly caps the surface atmospheric mixed layer. The aircraft flights in the boundary layer occasionally intercepted regions of intermittent fog below the stratus deck. In these air masses, MSA and nss-SO 2- were usually undetectable at ~<0.5 and < 5 ppt, respectively, and were probably scavenged by cloud. In contrast, clear air conditions during the same flights produced the highest concentrations for both sulfur species. Figure 5 presents the variation of MSA during the DGASP year at Dye 3. Once again the same seasonal pattern can be seen, with a minimum during winter, an increase starting in late spring (April), and a broad maximum during summer. But several features of this data can be pointed out. First, the concentrations are generally much lower than those observed at lower altitudes in the Arctic, with a minimum value of 0.007 ppt (average for the winter 0.08 ___0.12 ppt, as in Table 1). At 1.2 + 0.8 ppt, the summer average is still lower than the yearly average observed at Barrow or Alert. Second, it shows short episodic events peaking above a background, similar to the data for Barrow in Fig. 4b. These peaks can appear in the early winter (November or December) and to a lesser extent in mid-winter (February). In both seasons, the peaks are always synchronous with peaks in SO 2- (Jaffrezo and Davidson, 1992). Third, the background level is somewhat higher during the summer, when SO 2- peaks are not accompanied by elevated MSA concentrations. MSA concentrations measured during the ATM experiments at the Greenland Summit during the summers of 1989 and 1990 are very comparable to the summer average from Dye 3 and are of the same order of magnitude as values for the free troposphere from the aircraft sampling. Nevertheless, direct comparison

3017

between the free troposphere and the surface of the Ice Sheet should be made with caution. It has been shown that the surface-based temperature inversions over the Ice Sheet (present even during summer) has a large impact on atmospheric concentrations measured at ground level (Dibbet al., 1992a). M S A / n s s - S O ~ - ratios

The MSA/nss-SO 2- ratios (R) in the Arctic atmosphere are highly variable seasonally (Fig. 2b) but much less so spatially. The highest R value in the Arctic occurs in the boundary layer in summer when advection of pollution from mid-latitudes is at a minimum as shown by low nss-SO~- and anthropogenic trace metal concentrations (Barrie, 1986; Barrie and Hoff, 1985; Heidam, 1984). During aircraft flights, we observed higher MSA than nss-SO 2- on two occasions in the boundary layer; once over the Arctic peak ice ( R = 1.14) and once over the Bering Sea (R = 1.09). At ground level at Alert and Barrow, R was never greater than 1. Overall, mid-summer Arctic R values (0.2-0.3) are much higher than those in spring (<0.01) and considerably higher than those over temperate oceans (0.011-0.086) (Andreae et al., 1988; Ayers et al., 1986; Saltzman et al., 1983, 1985, 1986; Burgermeister and Georgii, 1991) but similar to those over the southern polar oceans in austral summer and fall (0.1-- 1.8) (Berresheim, 1987; Berresheim et al., 1990; Pszenny et al., 1989; Savoie et al., 1992). Summer average values of R in the boundary layer at Barrow and Alert are very similar (0.25 and 0.17, respectively), but R is much more variable at Barrow (a = 0.40 compared to 0.10 at Alert). In contrast, values obtained in the troposphere on the Ice Sheet are much lower than those from the boundary layer, both in winter and in summer (Table 1). The values from Dye 3 indicate a large (25 fold) increase of the average ratios from winter to summer but still lower than what is seen at sea level at Alert (60 fold). The data for the summer show similar values at Dye 3 and Summit,

3.5 3.0 2.5 ~2.0

< r/~ 1.5 1.0 0.5 0.0

Aug

Sep

Oct

1988

Nov

Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

1989

Fig. 5. Seasonal cycles of atmospheric MSA at Dye 3 on the Greenland Ice Sheet from August 1988 to July 1989. Tick marks indicate the beginnings of the months. ~.(A) ZT:Z711B-L

3018

S.-M. LI et al.

and values for summer 1989 and summer 1990 at ATM are also similar; Values of R for the free troposphere are very low, averaging only 0.011 at 2-6 km altitude in the vicinity of Barrow (Table 1). Possible reasons for these values of R are discussed below. DISCUSSION

Seasonal variations of MSA.

We propose that seasonal variations in (i) the circulation of the northern troposphere, (ii) biogenic sources of DMS in the northern oceans, (iii) the photochemical oxidation of DMS to MSA, and (iv) removal of MSA by precipitation are the controlling factors which determine the observed MSA seasonal variations in the Arctic. The influence of circulation and DMS sources on seasonal variations in MSA at Alert and Barrow can be better understood in terms of air mass transport climatology. Figure 1 shows the mean positions of the summer and winter Arctic fronts (Barrie, 1986; Shaw, 1988). These conform approximately to the mean positions of the - 5 ° C surface-level isotherm (Prik, 1959; Godske et al., 1957; Barry, 1967). In winter to spring, air masses arriving at Alert often originate in the high Arctic with frequent advection from the North Atlantic and Europe (Barrie, 1986). In contrast, air masses from the North Pacific seldom influence the Alert region. In summer, air mass advection from midlatitude to Alert is weaker than in winter (Barrie, 1986). Thus, northern marine emission sources of DMS, e.g. those in the North Atlantic, could affect organic sulfur at Alert to a certain degree, especially during winter to spring. This, combined with photochemical production as discussed further below, exerts an influence on the observed MSA variation at Alert. In contrast, MSA at Barrow reflects the alternating influence of sources in the North Pacific air masses and Arctic air masses, with weaker seasonal changes in DMS photochemistry superimposed. Shaw (1988) suggested that the meandering of the Arctic Front in the winter could bring the North Pacific air mass up to Barrow in spring. Sporadic incidences of Pacific air masses were observed at Barrow in spring during a previous study (Li and Winchester, 1990). In summer, considering the relative position of Barrow to the Arctic Front (Fig. 1), advection of marine air masses from the North Pacific may also have an important influence on the synoptic meteorology at Barrow. In addition, the Arctic Ocean around Barrow is not frozen, which can provide a regional DMS source. This picture is consistent with Arctic circulation as indicated by studies of Arctic air pollution from midlatitudes (Barrie, 1986; Barrie and Hoff, 1984, 1985) and air mass transport climatology. While the northern oceans have been shown to have high DMS emissions (Burgermeister et al., 1990; Burgermeister and Georgii, 1991; Erickson et al., 1990), DMS emission from the Arctic Ocean is not quantified, despite evidence of its potential importance. For example,

during ABLE3A air masses (with the highest MSA concentrations in the Arctic boundary layer) which influenced the Barrow region during July 1988 originated exclusively from high latitude source areas with no apparent intrusion of North Pacific air (Shipham et al., 1992). Insight into the processes that control the Arctic concentrations of MSA and DMS can be drawn from a comparison of MSA observation with those of known marine biogenic products such as bromine and iodine. For particulate bromine, a singular annual peak from February-May (Berg et al., 1983; Barrie and Barrie¢ 1990) has been attributed to accumulation of organobromines (Cicerone et al., 1988) from marine source emissions (Dyrssen and Fogelqvist, 1981) and subsequent photooxidation at polar sunrise (Barrie et al., 1988). Particulate iodine has two peaks in February-May and August-October which may be due to photooxidation at polar sunrise of marine iodine compounds and to direct marine biogenic sources in northern oceans (Barrie and Barrie, 1990), respectively. If indeed the spring peak in MSA at Alert is caused by enhanced photochemical activity owing to increasing solar radiation, then the spring peaks in MSA, particulate bromine, and particulate iodine may have in common photochemically induced oxidation of a winter/spring reservoir of precursor gases such as DMS, marine bromine compounds (Cicerone et al., 1988; Berg et al., 1984), and marine iodine compounds (WMO, 1985) accumulated during the winter to spring season. Although the Arctic Ocean DMS sources such as biota cannot be ruled out, the release of Arctic DMS should be at its lowest during winter with the ocean frozen inside the boundary of the polar front. Alternatively, it is possible that such an accumulation would come from enhanced long-range transport of DMS from source areas in the North Atlantic and North Pacific during winter/spring season and subsequent photochemistry leading to MSA at polar sunrise. The relative steady increase of MSA at polar sunrise at Alert is supportive of such an explanation; we would expect more chaotic MSA concentrations such as those observed at Barrow (influence of the North Pacific) if long-range transport were dominant for MSA. On the other hand, one cannot rule out that MSA is being supplied at an increasing rate to Alert and Barrow in spring by transport from sources in the North Pacific, Bering Sea, and the North Atlantic Ocean. Marine biogenic activities in the Bering Sea in spring (WaJsh and McRoy, 1986; Walsh et al., 1989) indicate high DMS production at this time. This is supported by model-predicted DMS emissions in both North Pacific and North Atlantic ~Erickson et al., 1990) and atmospheric DMS measurements in the North Atlantic (Burgermeister et al., 1990; Burgermeister and Georgii, 1991) and Baltic Sea (Leek et al., 1990). This DMS is available for photooxidation to MSA before being transported to Barrow, as sup-

Seasonal and geographic variations of methanesulfonic acid ported by atmospheric MSA observations at Shemya (Fig. 3) and further in the North Atlantic (Prospero and Savoie, 1990). Transport of trace metals to the Arctic in spring occurs mostly from the mid-latitudes; this transport could also include DMS and MSA. It is clear that direct DMS measurements during the polar winter are needed to help determine the relative importance of transported and photochemicaily derived MSA in the Arctic during the spring. The MSA summer peak from June to October at Alert has no counterpart in particulate bromine, suggesting no similar marine biogenic sources of bromine and DMS. The summer season is the time of lowest air mass advection from mid-latitudes into the Arctic and shortest particle residence times as evidenced by the low levels of nss-SO~- and anthropogenic trace metals in the Arctic (Barrie, 1986; Barrie and Hoff, 1985; Barrie and Barrie, 1990), which implies low levels of MSA or DMS transport from the midlatitudes. Since in the summer, oxidation of DMS is relatively fast (residence time ~0.75 d (Berresheim et al., 1990)) and the residence time of the aerosol products is lowest (several days), summer MSA concentrations should be influenced more by regional DMS emissions in the Arctic Ocean than in spring. At Barrow, the background levels of CO and NOy measured during ABLE3A (Wofsy et al., 1992) suggest that the Arctic Ocean indeed is an important source of atmospheric MSA, as also indicated by the meteorology during the ABLE3A study (Shipman et al., 1992). The summer MSA concentrations at Barrow are somewhat higher than those at Alert. This is likely due to the proximity of Barrow to more productive open Arctic Ocean areas than Alert. In addition, Alert is likely to receive more free tropospheric air as a result of katabatic winds from the surrounding mountainous terrain which has lower MSA concentrations. In contrast, Barrow less frequently receives such high altitude air mass. Although the seasonal variation observed on the Greenland Ice Sheet is similar to that observed at Barrow and Alert at sea level, concentrations are much lower. During summer they are close to the concentrations measured by aircraft in the free troposphere. On the Ice Sheet, there is an increasing MSA level from spring into the summer. It has been shown by considerations of synoptic meteorology and backward airmass trajectories (Davidson et al., 1993) that the SO~- peaks in spring at Dye 3 are mostly due to anthropogenic material transported from mid-latitudes either directly or via the Arctic basin. This transport would occur through the troposphere above the boundary layer (Dibb and Jaffrezo, 1993). There are at present two hypotheses to explain the increase in background MSA concentration over the Ice Sheet in summer. First, MSA may be transported through the free troposphere from distant oceanic sources. This assumes that large-scale transport processes are delivering MSA to the atmosphere over Greenland. Second, MSA may be brought from high

3019

productivity regions of the ocean surrounding Greenland; this implies that local transport is more important. Evidence of the second hypothesis is provided by the backward airmass trajectories showing short distances of transport in summer. As well, the low coherence in the MSA and SO~- episodes in summer is a good indication that summer MSA is more influenced by regional sources with low nss-SO~than by long-range transport. This is consistent with the information from Alert and is also supported by the aerosol MSA/SO~- ratios over the Ice Sheet, which sometimes approach 0.2. Such high ratios only occurred in air masses from high latitudes. It is important to determine whether long range transport or local transport is influencing the Ice Sheet in the different seasons in order to properly interpret the MSA record in the ice. M S A / n s s - S O ~ - ratio (R) at sea level

Long-term records of MSA and nss-SO~-, such as those from Antarctic ice cores (Legrand et al., 1992), have been interpreted with R value used as a tracer of transport pathways (Legrand et al., 1992). However, R depends on the complex pathways of DMS oxidation by either OH or NO 3 radicals (Atkinson et al., 1984; Hatakeyama et al., 1982, 1985; Jensen et al., 1991), on air temperature (Hynes et al., 1986), and on the competing rates of SO2 removal to the Earth's surface and oxidation to SO~- when in boundary layer. Although R at mid-latitude oceanic sites (Andreae et al., 1988; Ayers et al., 1986; Prospero and Savoie, 1990; Saltzman et al., 1983, 1985, 1986) has a range of low values from 0.011 to 0.086 and appears to be seasonally independent, it is not clear how much pollution-derived nss-SO~- affect these ratios. In remote marine environments where pollution is absent, R can be'considerably higher, e.g. 0.2-0.4 at Mawson, Antarctica (Prospero et al., 1991; Savoie et al,, 1992), 0.55 in the South Ocean south of Australia in austral summer (Berresheim et aL, 1990), and 1.32 in the Drake Passage in austral fall (Berresheim, 1987). These results indicate an increasing latitudinal trend in the values of R from the equator toward the Southern Oceans. The new information in Table 1 suggests a similar latitudinal trend in the Arctic for times of minimal anthropogenic influence, with R in summer increasing as one moves northward from the equator. R values increase to roughly I at 150 m over the Arctic pack ice (max. 1.04), 150 m over the Bering Sea (max. 1.09) and Barrow (max. 1.14). Values of R at Alert are smaller, averaging 0.17 with a maximum of 0.42 from the weekly samples, probably reflecting anthropogenie influences in the sample averaging. Several hypotheses have been proposed to explain this pattern of increasing R as one moves into the polar regions. (i) One key mechanism for MSA formation is by reaction of DMS with the OH radical. This

3020

S.-M. L! et al.

mechanism is temperature dependent, with increasing production of MSA as the temperature decreases. If this is true, with other factors being equal, one would expect more MSA relative to biogenic nss-SO 2- as the temperature decreases with increasing latitude (Hynes et al., 1986; Berresheim, 1987). (ii) Once MSA is formed, it may disappear by several processes. Two of the most important are believed to be oxidation by OH radical to SO 2- and precipitation scavenging. It has been hypothesized that the oxidation to SO 2- takes place in the liquid phase, e.g. cloud droplets or aerosols (Saltzman et al., 1983; Saltzman et al., 1984). Saltzman et al. (1983, 1984) estimated an MSA lifetime of about 40 d in aerosols but shorter in clouds. Assuming roughly equal rates of cloud scavenging, one would expect MSA in the atmosphere to decrease in concentration faster than SO 2-. There are more clouds in the mid-latitudes than in the polar regions; this suggests more rapid MSA loss in the mid-latitudes, leading to lower R values. (iii) Precipitation scavenging rates for MSA and SO 2- may not be equal. Limited data (e.g. Saltzman et al. (1983); Pszenny et al. (1989); Pszenny (1992)) suggest that MSA may be associated with slightly larger particles than SO 2-. This may imply greater scavenging rates for MSA than for SO 2-, leading to lower R values in the midlatitudes where precipitation is highest. (iv) Dry deposition of SO 2 to the Earth's surface. Since DMS oxidizes to MSA and SO2 and the latter further oxidizes to SO 2-, deposition of SO2 to the Earth's surface (particularly in the marine boundary layer) compared to air aloft or air over relatively SO2-inert cold snow can increase the ratio of MSA/nss-SO 2-. (v) All measured values of R may have varying degrees of influence by anthropogenic nss-SO 2-, even in marine atmospheres; at mid-latitude, the influence of anthropogenic nss-SO 2- can be expected to be greater than at high altitudes in the Southern Hemisphere, where the influence may be minimal, and Northern Hemisphere in sum-

mer (in winter anthropogenic nss-SO 2- in the Arctic is dominant). At this time there is little background information available to evaluate each of these hypotheses. The data presented here cannot be used to identify the most likely ones. A very limited support of (iii) is provided by considering that R is greater on average for precipitation (fresh snow) than associated aerosol during summer at Dye 3 (Jaffrezo et al., 1993a), indicating a slightly more efficient uptake of MSA than SO~- in cold clouds. Recently, Berresheim et al. (1991) reported slightly higher R values in precipitation than in aerosols over the western North Atlantic Ocean, also in support of hypothesis (iii). The influence of man-made SO,~- on the values of R in the summer at Alert can be assessed with the information on the isotopic composition of SO 2-. In the period June to September at Alert, Nriagu et al. (1991) reported that aerosol SO 2- had an isotopic 3"S to a2S ratio (63"S) of 8-9%0. The average fraction of SO 2- contributed by sea salt from this data was about 5% based on Na ÷. In addition, they showed that anthropogenic aerosols from Eurasia (the main source of anthropogenic pollution to the Arctic) have an average 33"S of 4.5-5.5%0 depending on source region. Table 2 shows estimates of the fractions of SO 2- at Alert from anthropogenic and biogenic sources. They were made assuming a 63'*5 of 20%0 in sea salt SO 2- and of 5%0 in anthropogenic nss-SO 2-. The assumed 334S of 15-19%o for DMS-derived nss-SO 2is based on the recent measurements of c534S in MSA over marine areas (Wagenbach, University of Heidelberg, personal communication). The results indicate that depending on the ~34S value for DMS-derived nss-SO2, -, marine biogenic sources contribute substantially to aerosol sulfate but that anthropogenic sources are still dominant. As shown in Table 2, the average contribution of biogenic sources is found to be between 16 and 32%, depending on the different initial conditions. Thus the R values observed in summer at Alert are lower than the actual branching ratio R* (R*=MSA/biogenic SO~=(MSA/nss-SO 2-) x (nss-SO2-/biogenic SO2-)) from DMS oxidation by a factor of 3-6, as given in

Table 2. Estimates of the fractions of SO2- in aerosols at sea level at Alert between July and August based on sulfur isotope ratios of Nriagu et al. (1991) 6a4S (~) Bio SO215 19

Fractions of SO24- (%) in Arctic SO429 8 9 8

Sea salt* 5 5 5 5

Biogenict 32 23 23 16

Anthrot 63 72 72 79

Ratios nss-SO2-/bio SO2-

R*:~

3 4.1 4.1 5.9

0.51 0.70 0.70 1.00

*Based on Na + aerosol observations. tEstimated from sulfur isotope balance made assuming 634Sa.thro= 596oand c$3#Sseasalt'20%e~. :~LR*-- MSA/biogenie SO2- ---(MSA/nss-SO2- ) x (nss-SO2-/biogenic SO2- ) = R x (nss-SO,2-/biogenic SO,2- ), where R =0.17 from Table 1.

Seasonal and geographic variations of methanesulfonic acid Table 2. These correction factors give a range for average R* values from DMS oxidation at Alert in summer between 0.51 to 1.00, also given in Table 2, based on the original average R = 0.17 in summer at Alert. By using much longer data sets for 634S, MSA, SO 2-, and Na +, a careful analysis of the contribution of biogenic source to aerosol SO 2- is recently presented by Li and Barrie (1993) and gives an R* of 0.5-0.9 for the summer months at Alert which is in excellent agreement with the values derived above. The range of average R* values can be compared to several observations during the aircraft observations. The two high R values of 1.14 and 1.09 were observed in the boundary layer below a stratus deck. This boundary layer is well scavenged by constant drizzle and has a wet surface ocean underneath, and appears to have little anthropogenic influence as shown by the background levels of CO and SO 2- (Talbot et al., 1992). Thus the two values are close to actual ratios for the boundary layer. R* of 1.0 calculated for the last scenario in Table 2 appears to be in good agreement with these aircraft observations and with observations from around the Antarctic (Berresheim et al., 1990; Berresheim, 1987). In other words, if the MSA/SO~ratio from DMS oxidation in the summer Arctic boundary layer is about 1, then the last scenario of Table 2 indicates a biogenic contribution of about 16-23% to the summer nss-SO 2- at Alert, with sea salt contributing 5% and anthropogenic source 72-79%. R as a function o f elevation

The overall results from ABLE3A point to the stratosphere as an important source of selected species (e.g. O a, NOy, 21°pb, and SO 2- in the Arctic midtroposphere during summer time (Dibb et al., 1992b; Jacob et al., 1992; Talbot et al., 1992; Wofsy et al., 1992). On several occasions, flying out over the Arctic Ocean pack ice we observed significantly higher concentrations of aerosol nss-SO 2- at 6 km altitude compared to levels lower in the ~oposphere (Talbot et al., 1992). It appears that stratospheric nss-SO 2-, possibly of anthropogenic origin, may dominate the nss-SO 2- budget in the Arctic mid-troposphere but that at lower altitudes biogenic sulfur sources become increasingly more important. Average values of R at Dye 3 range from 0.003 in mid-winter to 0.06 in summer, as shown in Table 1. Note that SO~- data from Dye 3 and ATM used to compute these R values (given by Jaffrezo et al. (1993a)) have not been corrected for marine influence, leading therefore to lower estimates of R, particularly in summer. However, this influence is expected to be small (<10% of the total SO 2- as sea salt SO~-) based on past data on aerosol (Davidson et al., 1985) and snow (Jaffrezo et al., 1993a) at Dye 3. R appears to decrease monotonically with elevation in summer, shown by considering the following range of values in order of increasing elevation: Alert (0.17), Dye 3 (0.06), Summit (0.05), and flights in the free

7

3021

Altitudes (kin)

Alrorsft meeeurementa from ABLE3A

Greenland Summit air and present I©e

/

oorrected with isotope Alert Summer

1

R*

,

durln0 A I L E S A

~

row

0 o

0.2

J

I

0.4

0.8

I

0.8

1

1.z

R or R* Fig. 6. SketchesofverticalprofilesofMSA/nss-SO I- (R) and MSAfoiogenic SO~- (R*I for summer Arctic.

troposphere at 6 km (0.01). These R values are similar to those in present ice at Summit (0.05) (Whung et al., 1989). This vertical trend, shown in Fig. 6, is consistent with the data reported by Berresheim et al. (1990) and by Andreae et el. (1987) showing decreasing R with elevation over Tasmania and northwest North Pacific, respectively. It suggests that the importance of nss-SO42- from DMS oxidation decreases relative to that of anthropogenic nss-SO24- with increasing altitudes, partially because of injection of the latter at higher latitudes (Talbot et el., 1992) while the sources of MSA are essentially at sea level. Furthermore, on average the high free troposphere is more likely to receive inputs from long-range transport from lower latitudes, with lower R values, while the boundary layer reflects much more local conditions of the high latitudes. Other factors may include differences in the scavenging processes (e.g. that caused by particle size difference) that would preferentially deplete MSA over SO~-, and oxidation of MSA in cloud droplets en route to the Ice Cap. Partial evidence of these effects associated with the altitude can be seen in the average values of R for preindustrial ice from central Greenland that are not higher than 0.17 (Whung et al., 1989). With biogenic emissions from local sources arriving onto the Ice Sheet in summer, the ratio should be higher due to the absence of anthropogenic S O l - , as probably was the case during the pre-industrial era. The effect of the altitude is also apparent from the smaller increase of the average MSA concentrations from winter to summer at Dye 3 than at Alert, as already mentioned. The direct impact of anthropogenic emissions of SO42- is seen with the decrease of this ratio in modern ice down

3022

S.-M. LI et al.

to averages of 0.05 (Whung et al., 1989) or 0.06 and 0.10 for fresh snows at Dye 3 (Jaffrezo et al., 1993a) and ATM (Jaffrezo et al., 1993b), respectively. Nevertheless, to quantify each of these effects, a detailed knowledge of the photochemistry of these organosulfur compounds, source strengths, and atmospheric dynamic processes will be needed.

SUMMARY

AND CONCLUSION

Unlike the Antarctic, the Arctic is a region in which atmospheric sulfur cycle is dominated by anthropogenie inputs. Only during summer months do marine biogenic sources contribute a significant yet generally minor fraction of the atmospheric sulfate. The measurements of MSA concentrations reported here are in good agreement with our understanding of the source regions, the transport pathways, and reaction processes involved in the cycle of this species. The results of the present study and their implications may be summarized as follows: 1. The seasonal variation of MSA in the lower Arctic troposphere near the North Pole has two maxima of 6 ppt in April-May and July-August and one minimum from November to January. A very similar seasonal pattern is observed at Dye 3 on the Greenland Ice Sheet. It is completely in anti-phase to seasonal variations in the southern polar atmosphere. Broadly speaking this is similar to the seasonal variation observed in other remote areas of the northern hemispheric oceans. Based on limited data of sulfur isotope composition of SO 2- in aerosols from Alert, it is estimated that 16-23% of summer SO 2- is derived from biogenic sources. 2. The April-May peak seen at Alert (82.5°N) may be associated with photo-induced oxidation of organic sulfur compounds at polar sunrise or increased transport from mid-latitude oceans during this time of year. The summer peak is attributed to local or regional Arctic oceanic sources. 3. Dye 3 on the Greenland Ice Sheet has the lowest MSA concentration observed while Barrow has the highest and most variable MSA concentrations. Alert has the most smooth seasonal changes and intermediate MSA concentrations. 4. Average summer MSA concentrations are lower in the high Arctic than those over mid-latitude regions or over the Southern Ocean. The concentrations decrease dramatically and are less variable with increasing altitude. 5. The ratio MSA/nss-SO2- is impacted by longrange transport of anthropogenic SO4. However, measurements in pristine conditions in the summer give values as high as 1 in the Arctic in the boundary layer. This is the first evidence of a situation similar to that seen in the Southern Hemisphere, with increasing values of this ratio with toward higher latitudes. The MSA/nss-SO2- ratio in the summer decreases with

altitude, in agreement with the respective altitudes of sources and transport pathways of MSA and SO 2-. 6. The evolution of the MSA/nss-SO2- ratio on the Ice Sheet between spring and summer is another piece of evidence of very distinct meteorological regime (and therefore distinct input of chemical species) during these two seasons, with local sources preferentially affecting the Ice Sheet in summer. Acknowledoement--SML acknowledges support of the Na-

tional Center for Atmospheric Research, Boulder, Colorado in obtaining the MSA data from Barrow, Alaska and the Atmospheric Environment Service in data analyses. RWT and RCH acknowledge support from the NASA Global Tropospheric Chemistry program and thank the personnel of the Wallops Flight Facility for supporting operations aboard the NASA Electra research aircraft during ABLE3B. REFERENCES

Andreae M. O. and Raemdonck H. (1983)Dimethyl sulfidein the surface ocean and the marine atmosphere: a global view. Science 221, 744-747. Andreae M. O., Ferek R. J., Bermond F., Byrd K. P., Engstrom R. T., Hardin S., Houmere P. D., LeMarrec F., Raemdonck H. and Chatfield R. B. (1985)Dimethyl sulfide in the marine atmosphere. J. geophys. Res. 90, 12,891-12,901. Andreae M. O., Berresheim H., Andreae T. W., Kritz M. A., Bates T. B. and Merrill J. T. (1988)Vertical distribution of dimethylsulfide, sulfur dioxide, aerosol ions, and radon over the northeast Pacific Ocean. J. atmos. Chem. 6, 149-173. Atkinson R., Pitts J. N. Jr. and Aschmann S. M. (1984) Tropospheric reactions of dimethyl sulfide with N O 3 and OH radicals. J. Phys. Chem. 88, 1584-1587. Ayers G. P, Ivey J. P. and Goodman H. S. (1986)Sulfate and methanesulfonate in the maritime aerosol at Cape Grim, Tasmania. J. atmos. Chem. 4, 173-185. Ayers G. P., Ivey J. P. and GiUett R. W. (1990) Coherence between seasonal cycles of dimethylsulphide,methanesulphonate and sulphate in marine air. Nature 349, 404-406. Barnes I., Becker K. H., Carlier P. and Mouvier G. (1987) FTIR study of the DMS/NO2fI2/N2 photolysis system: The reaction of IO radicals with DMS. Int. J. chem. Kinet. 19, 489-501. Barrie L. A. (1986)Arctic air pollution: an overviewof current knowledge. Atmospheric Environment 20, 643-663. Barrie L. A. and Hoff R. M. (1984) The oxidation rate and residencetime of sulphur dioxide in the Arctic atmosphere. Atmospheric Environment 18, 1995-2010. Barrie L. A. and HoffR. M. (1985) Five years of air chemistry observations in the Canadian Arctic.Atmospheric Environment 19, 1995-2010. Barrie L. A. and Barrie M. J. (1990)Chemical components of lower tropospheric aerosols in the high Arctic: six years of observations. J. atmos. Chem. 11, 211-246. Barrie L. A., Bottenheim J. W., Schnell R. C., Crutzen P. J. and Rasmussen R. A. (1988)Ozone destruction and photochemical reactions at polar sunrise in the lower Arctic atmosphere. Nature 334, 138-141. Barry R. B. (1967) Seasonal location of the Arctic front over North America. Geoar. Bull. 9, 79-95. Barton D. V., Grajnar T. S., McArthur L. J. B. and Wardle D. I. (1988) Summary of Alert solar radiation data. Internal Report ARPD 131X54, Atmospheric Environment Service, Environment Canada. Bates T. S., Charlson R. J. and Gammon R. H. (1987) Evidence for the climatic role of marine biogenic sulphur. Nature 329, 319-321.

Seasonal and geographic variations of methanesulfonic acid Bates T. S., Johnson J. E., Quinn P. K., Goldan P. D., Kuster W. C., Covert D. C. and Hahn C. H. (1990) The biogeochemical sulfur cycle in the marine boundary layer over the northeast Pacific Ocean. d. atmos. Chem. 10, 59-81. Berg W. W., Sperry P. D., Rahn K. A. and Gladney E. S. (1983) Atmospheric bromine in the Arctic. J. geophys. Res. 88, 6719-6736. Berg W. W., Heidt U E., Pollock W., Sperry P. D., Cicerone R. J. and Gladney E. S. (1984) Brominated organic species in the Arctic atmosphere. Geophys. Res. Lett. 11,429-432. Berresheim H. (1987) Biogenic sulfur emissions from the subAntarctic and Antarctic oceans. J. geophys. Res. 92, 13,245-13,262. Berresheim H., Andreae M. O., Ayers G. P., Gillett R. W., Merrill J. T., Davis V. J. and Chameides W. L. (1990) Airborne measurements of dimethylsulfide, sulfur dioxide, and aerosol ions over the southern ocean south of Australia. J. atmos. Chem. 10, 341-370. Berresheim H., Andreae M. O., Iverson R. L. and Li S.-M. (1991) Seasonal variations of dimethylsulfide emissions and atmospheric sulfur and nitrogen species over the western North Atlantic Ocean. Tellus 43B, 353-372. Bottenheim J. W., Barrie L. A., Atlas E., Heidt L. E., Niki H., Rasmussen R. A. and Shepson P. B. (1990) Depletion of lower tropospheric ozone during Arctic spring: the polar sunrise experiment 1988. J. geophys. Res. 95, 18,555-18,568. Burgermeister S. and Georgii H.-W. (1991) Distribution of methanesulfonate, nss sulfate and dimethylsulfide over the Atlantic and the North Sea. Atmospheric Environment 25A, 587-595. Burgermeister S., Zimmermann R. L., Georgii H.-W., Bingemer H. G., Kirst G. O., Janssen M and Ernst W. (1990) On the biogenic origin of dimethylsulfide: relation between chlorophyll, ATP, organismic DMSP, phytoplankton species, and DMS distribution in Atlantic surface water and atmosphere. J. geophys. Res, 95, 20,607-20,615. Charlson R. J., Lovelock J. E., Andreae M. O. and Warren S. G. (1987) Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate: a geophysiological feedback. Nature 326, 655-661. Chatfield R. B. and Crutzen P. J. (1990) Are there interactions of iodine and sulfur species in marine air photochemistry? J. geophys. Res. 95, 22,319-22,341. Cicerone R. J., Heidt L. E. and Pollock W. H. (1988) Measurements of atmospheric methyl bromide and bromoform. J. geophys. Res. 93, 3745-3749. Davidson C. I., Santhanam S., Fortmann R. C. and Olson M.P. (1985) Atmospheric transport and deposition of trace elements onto the Greenland ice sheet. Atmospheric Environment 19, 2065-2081. Davidson C. I., Jaffrezo J. U, Small M. J., Summers P. W., Olson M. P. and Borys R. D. (1993a) Trajectory analysis of source regions influencing the South Greenland Ice Sheet during the Dye 3 Gas and Aerosol Sampling Program. Atmospheric Environment 27A, 2739-2749. Daykin E. P. and Wine P. H. (1990) Rate of reaction of IO radicals with dimethylsulfide. J. geophys. Res. 95, 18,547-18,553. Dibb J. E. and Jaffrezo J. L. (1993) Beryllium-7 and lead-210 in the aerosol and snow during DGASP. Atmospheric Environment 27A, 2751-2760. Dibb J. E., Jaffrezo J. L. and Legrand M. (1992a) Initial findings of recent investigations of air-snow relationship in the Summit region of the Greenland Ice Sheet. J. atmos. Chem. 14, 167-180. Dibb J. E., Talbot R. W. and Gregory G. L. (1992b) Beryllium 7 and lead 210 in the western hemispheric Arctic atmosphere: observations from three recent aircraft-based sampling programs. J. geophys. Res. 97, 16,709-16,715. Dyrssen D. and Fogelqvist E. (1981) Bromoform concentrations of the Arctic ocean in the Svalbard area. Oceanologica Acta 4, 313-317.

3023

Erickson D. J., Ghan S. J. and Penner J. E. (1990) Global ocean-to-atmosphere dimethyl sulfide flux. J. geophys. Res. 95, 7543-7552. Godske C. L., Bergeron T., Bjerknes J. and Bundgaard R. C. (1957) Dynamic Meteorology and Weather Forecasting. Am. Met. Soc. and Carnegie Inst. of Washington publication. Grosjean D. (1984) Photooxidation of methyl sulfide, ethyl sulfide, and methanethiol. Envir. Sci. Technol. 18, 460-468. Harriss R. C., Wofsy S. C., Bartlett D. S., Shipham M. C., Hoell J. M. Jr., Bendura R. J., Drewry J. W., McNeal R. J., Navarro R. L., Gidge R. and Rabine V. (1992) The Arctic Boundary Layer Expedition (ABLE-3A): July-August 1988. J. geophys, Res. 97, 16,383-16,394. Hatakeyama S., Okuda M. and Akimoto H. (1982) Formation of sulfur dioxide and methane sulfonic acid in the photooxidation of dimethyl sulifde in the air. Geophys. Res. Lett. 9, 583-586. Hatakeyama S., Izumi K. and Akimoto H. (1985) Yield of SO 2 and formation of aerosol in the photo-oxidation of DMS under atmospheric conditions. Atmospheric Environment. 19, 135-141. Heidam N. Z. (1984) The components of the Arctic aerosol. Atmospheric Environment 1B, 329-343. Hynes A. J., Wine P. H. and Semmes D. H. (1986) Kinetics and mechanism of OH reactions with organic sulfides. J. Phys. Chem. 90, 4148-4156. Jacob D. J., Wofsy S. C., Bakwin P. S., Fan S.-M., Harriss R. C., Talbot R. W., Bradshaw J. D., Sandholm S. T., Singh H. B., Gregory G. L., Sachse G. W., Shipham M., Blake D. R. and Fitzjarrold D. R. (1992) Summertime photochemistry of the troposphere at high northern latitudes. J. yeophys. Res. 97, 16,421-16,431. Jaffrezo J. L. and Davidson C. I. (1992) Sulfate in the air, surface snow, and snowpit at Dye 3 Greenland. Submitted to the 5th International Conference on Precipitation Scavenging and Atmosphere-Surface Exchange Processes, Richland, WA, 15-19 July. Jaffrezo J. L. and Davidson C. I. (1993) DGASP: an overview. Atmospheric Environment 27A, 2703-2707. Jaffrezo J. L., Davidson C. I. and Dibb J. E. (1993a) Major ions in the aerosol and snow during DGASP. Atmospheric Environment (manuscript in preparation). Jaffrezo J. L., Davidson C. I., Dibb J. E. and Legrand M. (1993b) Major ions in the aerosol and snow during the first two ATM seasons. Atmospheric Environment (manuscript in preparation). Jensen N. R., Hjorth J., Lohse C., Skov H. and Restelli G. (1991) Products and mechanism of the reaction between NO 3 and dimethylsulphide in air. Atmospheric Environment 25A, 1897-1904. Krebs J. S. and Barry R. G. (1970) The Arctic front and the tundra-taiga boundary in Eurasia. Geoor. Rev. 4, 548554. Leck C., Larsson U., Bagander L. E., Johansson S. and Hajdu S. (1990) Dimethyl sulfide in the Baltic Sea: annual variability in relation to biological activity. J. geophys. Res. 95, 3353-3363. Legrand M., Feniet-Saigne C., Saltzmafi E. S., Germain C., Barkov N. I. and Petrov V. N. (1991) An ice-core record of oceanic emissions of dimethylsulfide during the last climate cycle. Nature 350, 144-146. Legr.and M., Feniet-Saigne C., Saltzman E. S. and Germain C. (1992) Spatial and temporal variations of methanesulfonic acid and non sea salt sulfate in Antarctic ice. J. atmos. Chem. 14, 245-260. Li S.-M. and Barrie L. A. (1993) Biogenic sulfur aerosols in the Arctic troposphere: I. contributions to total sulfate. 2. geophys. Res. (in press). Li S.-M. and Winchester J. W. (1989) Resolution of ionic components of the late winter Arctic aerosol. Atmospheric Environment 23, 2387-2399. Li S.-M. and Winchester J. W. (1990) Haze and other aerosol

3024

S.-M. LI et al.

components in late winter Arctic Alaska, 1986. J. oeophys. Res. 95, 1797-1810. Li S.-M., Barrie L. A. and Sirois A. (1993) Biogenic sulfur aerosol in the Arctic troposphere: II. Trends and seasonal variations. J. geophys. Res. (in press). Nriagu J. O., Coker R. D. and Barrie L. A. (1991) Origin of sulphur in Canadian Arctic haze from isotope measurements. Nature 349, 142-145. Platt U., LcBras G., Poulet G., Burrows J. P. and Moortgat G. (1990) Peroxy radicals from night-time reactions of NOa with organic compounds. Nature 348, 147-149. Prik Z. M. (1959) Mean position of surface pressure and temperature distribution in the Arctic. Tr. Arkticheskogo Nausch.-lssled. Inst. 217, 5-34. Prospero J. M. and Savoie D. L. (1990) Temporal and spatial variations of methanesulfonate concentrations and assessments of the impact of the biogenic source on non-sea salt sulfate over the North Atlantic Ocean. EOS Trans. AGU 71, 1234. Prospero J. M., Savoie D. L,, Saltzman E. S. and Larsen R. (1991) Impact of oceanic sources of biogenic sulphur on sulphate aerosol concentrations at Mawson, Antarctica Nature 350, 221-223. Pszenny A. A. P. (1992) Particle size distributions of methanesulfonate in the tropical Pacific marine boundary layer. J. atmos. Chem. 14, 273-284. Pszanny A. A. A. P., Castelle A. J., Galloway J. N. and Duce R. A. (1989) A study of the sulfur cycle in the Antarctic marine boundary layer. J. geophys. Res. 94, 9818-9830. Quinn P. K., Bates T. S., Johnson J. E., Covert D. S. and Charlson R. J. (1990) Interaction between the sulfur and reduced nitrogen cycles over the central Pacific Ocean. J. geophys. Res. 95, 16,405-16,416. Saltgraan E. S., Savoie D. L., Zika R. G. and Prospero J. M. (1983) Methane sulfonic acid in the marine atmosphere. J. geophys. Res. 88, 10,897-10,902. Saltzman E. S., Gidel L. T., Zika R. G., Milne P. J., Prospero J. M., Savoie D. L. and Cooper W. J. (1984) Aerosol chemistry of methane sulfonic acid. Extended abstract presented at the Conference on Gas-Liquid Chemistry of Natural Waters, Brookhaven National Laboratory, Upton, Long Island, New York. Saltzman E. S., Savoie D. L., Prospero J. M. and Zika R. G. (1985) Atmospheric methanesulfonic acid and non-sea salt sulfate at Fanning and American Samoa. Geophys. Res. Lett. 12, 437-440. Saltzman E. S., Savoie D. L., Prospero J. M. and Zika R. G. (1986) Methanesulfonic acid and non-sea salt sulfate in Pacific air. Regional and seasonal variations. J. atmos. Chem. 4, 227-240. Savoi¢ D. L. and Prospero J. M. (1989) Comparison of oceanic and continental sources of non-sea salt sulphate over the Pacific Ocean. Nature 339, 685-687. Savoie D. L., Prospero J. M., Larsen R. J. and Saltzman E. S.

(1992) Nitrogen and sulfur species in aerosols at Mawson, Antarctica, and their relationship to natural radionuclides. J. atmos. Chem. 14, 181-204. Schwartz S. E. (1988) Are global cloud albedo and climate controlled by marine phytoplankton? Nature 336, 441-445. Shaw G. E, (1988) Chemical air mass systems in Alaska. Atmospheric Environment 22, 2239-2248. Shipham M. C., Bachmeier A. S.,Cahoon D. R. and Browell E. V. (1992) Meteorological overview of the Arctic Boundary Layer Expedition (ABLE3A) flight series. J. geophys. Res. 97, 16,395-16,419. Sza N. D. and Ko M. K. W. (1980) Photochemistry of COS, CS2, CHaSCHa and H2S: implications for the atmospheric sulfur cycle. Atmospheric Environment 14, 1223-1239. Talbot R. W., Vijgen A. S. and Harriss R. C. (1992) Soluble species in the Arctic summer troposphere: Acidic gases, aerosols, and precipitation. J. geophys. Res. 97, 16,531-16,543. Teen O. B., Kasting J. F., Turco R. P. and Liu M. S. (1987) The sulfur cycle in the marine atmosphere. J. geophys. Res. 92, 943-963. Walsh J. J. and McRoy C. P. (1986) Ecosystem analysis in the southeastern Bering Sea. Cont. Shelf Res. 5, 259-288. Walsh J. J. et al. (1989) Carbon and nitrogen cycling within the Bering/Chukchi Seas: Source regions for organic matter affecting AOU demands of the Arctic Ocean. ProO. Oceanogr. 22, 279-361. Watts S. F., Watson A. and Brimblecomb¢ P. (1987) Measurements of the aerosol concentrations of methanesulphonic acid, dimethyl sulphoxid¢ and dimethyl sulphone in the marine atmosphere of the British Isles. Atmospheric Environment 21, 2667-2672. Watts S. F., Brimblecomb¢ P. and Watson A. J. (1990) Methanesulphonic acid, dimethyl sulphoxide, and dimethyl sulfone in aerosols. Atmospheric Environment 24A, 353-359. Whung P.-Y., Saltzman E. S. and Spencer M. J. (1989) A 200 year methanesulfonic acid record from the south Greenland ice core. EOS Trans. AGU 70, 1151. WMO, World Meteorological Organization (1985) Atmospheric Ozone 1985 Vol. 1. World Meteorological Organization, Geneva. Wofsy S. C., Sachse G. W., Gregory G. L., Blake D. R., Bradshaw J. D., Sandholm S. T., Singh H. B., Barrick J. A., Harriss R. C., Talbot R. W., Shipham M. A., Browell E. V., Jacob D. J. and Logan J. A. (1992) Atmospheric chemistry in the Arctic and sub-Arctic: Influence of natural fires, industrial emissions, and stratospheric inputs. J. oeophys. Res. 97, 16,731-16,746. Yin F. D., Grosjean D. and Seinfeld J. H, (1990) Photooxidation of dimethyl sulfide and dimethyl disulfide. I: Mechanism development. J. atmos. Chem. 11, 309-364.