THE FIRST GREENLAND ICE CORE RECORD OF METHANESULFONATE AND SULFATE OVER A FULL GLACIAL CYCLE

. Methanesulfonate (MSA) in ice cores has attracted attention as a possible tracer of past oceanic emissions of dimethylsu!fide (DMS). After sulfate MSA is the second most prevalent aerosol oxidation product of DMS, but in contrast to sulfate, DMS oxidation is the only known source of MSA. The hypothesis by Charlson et al., [1987] of a climate feedback mechanism with sulfur emissions f¾om marine phytoplankton influencing the cloud albedo adds to the interest in establishing long records of MSA and non-seasalt sulfate spanning large climatic changes. Records of MSA and non-seasalt sulfate covering time periods from a few years to thousands of year have been extracted from antarctic ice cores [Ivey et al., 1986; Saigne and Legrand, 1987; Legrand and Feniet-Saigne, 1991' Mulvaney et al., 1992] but only the record from the Vostok ice core [Legrand et al., 1991] covers a fhll glacial cycle. The concentrations of MSA and non-seasalt sulfate in Antarctica have been found to increase under glacial conditions. Here we present the first Northern Hemisphere record of MSA, and the first continuous record of non-seasalt sulfate, both extracted from the Renland ice core, East Greenland. The records are extending from the Holocene to the Eem interglacial !30,000 years B.P. The contrast to the Southern Hemisphere records is striking, with a decreasing concentration of MSA with the advance of glaciation but an increasing concentration of non-seasalt sulfate. A strong linear relationship is found in the Renland ice core between the ratio of MSA to non-seasalt sulfate and the temperature, with higher ratios associated with warmer climatic stages, while the opposite relationship to temperature is found in

The ice core was cut with a stainless steel microtome knife under clean room (class 100) conditions into 5 cm samples as follows: 55 continuous samples were cut between 86 and 89 m depth representing the time period ^.D. 1812-1820 and 374 continuous samples were cut in the deepest 20 m of the ice core representing approximately 10-145 ka B.P. [Johnsen and Dansgaard, 1992]. The samples were analyzed by chemically suppressed ion chromatography. Major anions (CI', NO3', SO42') and cations (Na +, NH4 +, K +, Mg 2+, Ca 2+) were determined on an integrated Dionex (4000i/2000i) system using AGSA/AS5ASg and Cationfastsep VII columns, respectively. MSA was determined in the same samples but on an other system using AG4/AS4 columns. The uncertainty in the analyses (! sigma) are estimated at less than +_10% for all major ions and +_30% for MSA. The seasalt component of the sulfate (based on sodium) is on average 10% of the total sulfate.

Result and discussion
The profiles of MSA and non-seasalt (nss) su!fate for the period 10 to 145 ka B.P. are shown in Figure 1. Mean values for the climatic stages (i.e. conventional marine oxygen isotope stages) are presented in Table l. The MSA concentration is a factor of 2.0-2.5 lower during glacial maximum (stage 2) than in interglacial periods (.stage 5e and stage 1, the latter represented by two limited sample series in Pre-boreal and late Holocene). Preliminary results from the G ISP2 ice core (ongoing deep drilling in the Greenland Summit area) support the glacial/interglacial MSA trend observed at Renland, suggesting that it is characteristic of the main Greenland ice sheet. The nss sulfate concentration is a factor of 1.9-2.7 higher during glacial maximum in the same comparison as above. This increase in nss sulfate concentration during the glacial period has previously been observed in both the Dye 3 and Camp Century deep ice cores [Herron and Langway, Jr., 1985].
The molar ratio (R) of MSA to nss sulfate in the Renland ice core is strongly related to the climate. This relationship to temperature is inverse to the temperature dependence of MSA production inferred [Berresheim, 1987] from laboratory oxidation studies [Hynes et al., 1986]  In an interglacial climate, the marine biogenic sulfur compounds in Greenland precipitation can come from either regional sources (i.e. Greenland coastal waters) or more distant, lower latitude sources. Trajectory analyses of air masses influencing the South Greenland ice sheet [Davidson et al., 1993a] indicate that the influence of different source areas varies with season, favouring the regional sources during summer. In contrast, the regional sources are closed off in a glacial climate due to the sea ice cover. Palaeooceanographic reconstructions [Ruddiman and Mcintyre, 1981 ] place the winter sea ice extent as far south as 45'•N during the most advanced glacial stage.
The interglacial R-values (Table 1 )    sulfate have yet to be found. Model studies [Langner, 1991] of  [Legrand et al., 1992] are found in both ice cores between R and isotopic temperature but with opposite signs. It shows that R in polar precipitation can not simply be a function of atmospheric temperatures. However, it would also be remarkable if the strong linear relationship to temperature in the Renland ice core is caused by changing source strengths of two, or more, independent sources of sulfur.

Conclusion
The Ren!and ice core data suggests a glacial Arctic atmospheric sulfur cycle influenced by low latitude oceanic emissions, possibly decreased, and non-biogenic sources in contrast to Antarctica which appears dominated by increased high latitude oceanic emissions. The isotopic temperature profiles suggest a roughly parallel climatic development [Johnsen et al., 1992a] in Greenland and Antarctica which calls in question the role of marine biogenic sulfate in climate forcing, or otherwise its exclusivity among the atmospheric sulfate aerosols. However, it has to be stressed that it is the composition and size distribution of the total aerosol in the atmosphere that affects the direct radiation balance and the cloud albedo, why records of non-seasalt sulfate alone demand cautious use when discussing climate forcing. The nss sulfate represents less than 5% of the total aerosol mass in the Renland ice core during full glacial conditions. In addition, the apparent influence from changing climatic conditions on the ratio R has to be more understood before using MSA as a quantitative tracer over long time-scales with major climatic changes, either as indicator of the relative strengths of biogenic versus nonbiogenic sulfur sources or indicative of changes in the principle source region. Further studies are necessary. to take the full advantage of the detailed records that will be available from the two deep ice cores drilled on the Greenland Summit.