Portal de Periódicos da CAPES

Modeled methanesulfonic acid (MSA) deposition in Antarctica and its relationship to sea ice

2011; American Geophysical Union; Volume: 116; Issue: D23 Linguagem: Inglês

10.1029/2011jd016383

2156-2202

Paul Hezel, Becky Alexander, Cecilia M. Bitz, Eric J. Steig, Christopher D. Holmes, Xin Yang, Jean Sciare,

Ocean Acidification Effects and Responses

Journal of Geophysical Research: AtmospheresVolume 116, Issue D23 Aerosol and CloudsFree Access Modeled methanesulfonic acid (MSA) deposition in Antarctica and its relationship to sea ice P. J. Hezel, P. J. Hezel [email protected] Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorB. Alexander, B. Alexander Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorC. M. Bitz, C. M. Bitz Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorE. J. Steig, E. J. Steig Department of Earth and Space Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorC. D. Holmes, C. D. Holmes Department of Earth System Science, University of California, Irvine, California, USASearch for more papers by this authorX. Yang, X. Yang NCAS-Climate, Centre for Atmospheric Science, Department of Chemistry, University of Cambridge, Cambridge, UKSearch for more papers by this authorJ. Sciare, J. Sciare LSCE, CNRS-CEA-UVSQ, Gif-sur-Yvette, FranceSearch for more papers by this author P. J. Hezel, P. J. Hezel [email protected] Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorB. Alexander, B. Alexander Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorC. M. Bitz, C. M. Bitz Department of Atmospheric Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorE. J. Steig, E. J. Steig Department of Earth and Space Sciences, University of Washington, Seattle, Washington, USASearch for more papers by this authorC. D. Holmes, C. D. Holmes Department of Earth System Science, University of California, Irvine, California, USASearch for more papers by this authorX. Yang, X. Yang NCAS-Climate, Centre for Atmospheric Science, Department of Chemistry, University of Cambridge, Cambridge, UKSearch for more papers by this authorJ. Sciare, J. Sciare LSCE, CNRS-CEA-UVSQ, Gif-sur-Yvette, FranceSearch for more papers by this author First published: 15 December 2011 https://doi.org/10.1029/2011JD016383Citations: 22AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract [1] Methanesulfonic acid (MSA) has previously been measured in ice cores in Antarctica as a proxy for sea ice extent and Southern Hemisphere circulation. In a series of chemical transport model (GEOS-Chem) sensitivity experiments, we identify mechanisms that control the MSA concentrations recorded in ice cores. Sea ice is linked to MSA via dimethylsulfide (DMS), which is produced biologically in the surface ocean and known to be particularly concentrated in the sea ice zone. Given existing ocean surface DMS concentration data sets, the model does not demonstrate a strong relationship between sea ice and MSA deposition in Antarctica. The variability of DMS emissions associated with sea ice extent is small (11–30%) due to the small interannual variability of sea ice extent. Wind plays a role in the variability in DMS emissions, but its contribution relative to that of sea ice is strongly dependent on the assumed DMS concentrations in the sea ice zone. Atmospheric sulfur emitted as DMS from the sea ice undergoes net transport northward. Our model runs suggest that DMS emissions from the sea ice zone may account for 26–62% of MSA deposition at the Antarctic coast and 36–95% in inland Antarctica. Though our results are sensitive to model assumptions, it is clear that an improved understanding of both DMS concentrations and emissions from the sea ice zone are required to better assess the impact of sea ice variability on MSA deposition to Antarctica. Key Points To model DMS-MSA sulfur cycle and compare Antarctic MSA deposition to ice cores Transport/lifetimes of DMS/MSA show MSA is not clearly associated with sea ice Need better characterization of DMS emissions from sea ice zone 1. Introduction [2] Methanesulfonic acid (MSA) concentrations in Antarctic ice cores are thought to be influenced by the distribution of nearby sea ice and have therefore been investigated with the goal of producing a proxy for sea ice cover in past climates [Saigne and Legrand, 1987; Legrand and Feniet-Saigne, 1991; Welch et al., 1993; Curran et al., 2003; Abram et al., 2007; Mayewski et al., 2009]. MSA is one of several oxidation products of dimethylsulfide (DMS), which originates in the atmosphere from biological production in the surface ocean. Observations demonstrate particularly high DMS concentrations in surface seawater in the seasonal sea ice zone around Antarctica throughout the austral spring and summer [e.g., Kettle et al., 1999; Tortell and Long, 2009; Gabric et al., 2005; Jones et al., 2010]. This observed relationship constitutes the basis of the proposed link between sea ice and MSA deposition on the Antarctic continent. [3] MSA concentrations in ice cores from various locations in Antarctica have shown inconsistent relationships with satellite observations of sea ice extent. Positive correlations of MSA with nearby sea ice extent have been reported by Welch et al. [1993], Curran et al. [2003], Foster et al. [2006], and Abram et al. [2010], with years of greater sea ice extent inducing larger areas of high DMS production and thereby increasing MSA concentrations in the ice core record. Using three ice cores from the Weddell Sea region, Abram et al. [2007] found negative correlations of MSA with sea ice extent in the Weddell Sea and positive correlations of MSA with sea ice extent to the west in the Bellingshausen Sea. Pasteur et al. [1995] and Sun et al. [2002] found negative correlations of MSA with sea ice extent at Dolleman Island and on the Lambert Glacier respectively. [4] The observational literature has not yet established which processes govern MSA deposition in Antarctica and the spatial extent over which such relationships hold. In addition to sea ice, MSA concentrations have also been linked to changes in atmospheric circulation [e.g., Becagli et al., 2009; Fundel et al., 2006]. The extent to which MSA is influenced by sea ice compared to atmospheric circulation may be regionally dependent. Recent studies have questioned whether sea ice plays a significant role in measured atmospheric MSA concentrations at all. Preunkert et al. [2007] found that atmospheric concentrations of high DMS coincided with high MSA at Dumont D'Urville (DDU) (140°1′E, 66°40′S) on short timescales (days) and was related to simultaneous high regional chlorophyll-a measured by satellite, though they concluded that a relationship to sea ice was not straightforward. Weller et al. [2011] compared atmospheric measurements of MSA and non sea-salt sulfate (nssSO42−) at Neumayer station (8°15′W, 70°39′S) and found no significant relationship between these and sea ice extent or any other climate indicators. [5] Nearly all of the MSA-sea ice proxy studies have compared observed MSA concentrations in ice cores from point locations in Antarctica to satellite observations of sea ice extent around the continent. In this study, we use a chemical transport model to examine some of the spatial characteristics of the link between MSA and sea ice extent. The experiments were designed to understand the impact of DMS emissions from the sea ice zone on MSA deposition patterns in Antarctica, since modeling studies of the sulfur cycle have neglected this important regional source [e.g., Cosme et al., 2002]. Though this sulfur source is important to understand, more fundamental characteristics of DMS emissions and MSA deposition in the high latitudes also emerge as a compelling story. In a series of model simulations with the chemical transport model GEOS-Chem [Bey et al., 2001] (http://geos-chem.org), we explore the sensitivity of MSA deposition in Antarctica to modification of DMS emissions from the sea ice itself. We find that the estimates of seawater DMS emissions play a prominent role in determining whether Antarctic sulfur deposition is dominated by DMS emissions from the sea ice zone or by southward transport of sulfur emissions from lower latitudes. We fail to find credible correlations between sea ice extent and MSA deposition in Antarctica within the model given reasonable estimates of the influence of sea ice on DMS surface concentrations. Interannual variability of DMS emissions in the model, and hence interannual variability of MSA deposition, is not strongly influenced by variability in sea ice extent. 2. Model Description and Methods Background to Model Simulations [6] Previous global sulfur cycle modeling studies 'cap' DMS emissions in the presence of sea ice [e.g., Chin et al., 2000; Cosme et al., 2002], which assumes that sea ice prevents gas exchange from the ocean, even from water among sea ice floes. Though these studies broadly capture features of the Antarctic regional sulfur cycle including atmospheric concentrations, seasonal cycles, and spatial gradients in deposition fluxes [Chin et al., 2000; Cosme et al., 2002, 2005; Gondwe et al., 2004; Castebrunet et al., 2006], estimates of MSA and nssSO42− deposition to the Antarctic continent miss this important regional source. Surface seawater DMS concentrations are typically prescribed from a climatology [e.g., Kettle et al., 1999; Simó and Dachs, 2002; Lana et al., 2011], with ocean-to-air gas exchange fluxes parameterized based on empirical relationships with wind speed and SST [e.g., Liss and Merlivat, 1986; Wanninkhof, 1992; Nightingale et al., 2000; Huebert et al., 2010]. [7] Current understanding of both sea ice biology and gas exchange processes suggests that there may be multiple ways in which the DMS source to the atmosphere is enhanced by the presence of sea ice [Levasseur et al., 1994]. Certain high DMS-producing species, including Phaeocystis sp., are prevalent in Southern Ocean waters, including the sea ice zone [Malin and Kirst, 1997]. Melting sea ice is thought to release nutrients that stimulate a phytoplankton bloom near the ice edge [Curran et al., 2003, and references therein]. Algal communities within sea ice brine pockets [Delille et al., 2007] may release dimethylsulfoniopropionate (DMSP), a DMS precursor, to the water column; DMSP is then converted to DMS by enzymatic cleavage of DMSP via bacterial consumption. DMS has also been measured in pore spaces among snow crystals on sea ice [Zemmelink et al., 2008], suggesting DMS may pass from ice to the atmosphere, and possibly from the water column through the ice as a result of direct gas exchange [Gosink et al., 1976; Semiletov et al., 2004]. Measurements of gas exchange of oxygen and sulfur hexafluoride through laboratory sea ice suggest that diffusion of gases through sea ice is much smaller than gas transfer to the atmosphere through open water leads, even when the fraction of open water is less than 1% of the ice area [Loose et al., 2011]. [8] We use the GEOS-Chem model version v8-01-03 for our study, at a horizontal resolution of 2° latitude by 2.5° longitude and vertical resolution of 30 hybrid pressure-sigma levels. GEOS-Chem is a global 3-D chemical transport model [Bey et al., 2001], driven by GEOS-4 meteorological fields from the Goddard Earth Observing System of the NASA Global Modeling and Assimilation Office [Bloom et al., 2005]. The input meteorological fields are 3-hour averages for surface fields, 6-hour averages for upper level fields, and 6-hour instantaneous fields for sea level and surface pressures and ice extent. Meteorological fields are originally computed at a resolution of 1° latitude × 1.25° longitude, 55 vertical hybrid sigma levels, and degraded to the GEOS-Chem model resolution. We use the offline-aerosol version of GEOS-Chem, described by Park et al. [2004], which uses monthly mean oxidant concentrations from a full-chemistry simulation. The set of simulations are described at the end of this section, and each was begun after a 1.5 year spinup of the chemistry model. Sea Ice and Sea Surface Temperature [9] Sea ice extent and sea surface temperature (SST) are specified from NOAA Optimum Interpolation (OI v.2) weekly fields [Reynolds et al., 2002], interpolated to 6-hour instantaneous fields at 1° × 1° resolution. Sea ice in GEOS-Chem is a binary field (the ocean fraction of a grid cell is either all ice or ice free) on the model grid. When the OI v.2 is degraded to 2° × 2.5° resolution, a grid cell is designated 'ice' if more than 50% of the area is covered with ice concentration of 15% or greater. Ocean-to-Air DMS Emissions and Surface Seawater DMS Concentrations [10] Atmospheric DMS emissions from seawater are parameterized using a climatology of surface seawater DMS concentrations [Simó and Dachs, 2002; Kettle et al., 1999; Lana et al., 2011] combined with a sea-to-air transfer velocity computed as a function of instantaneous 10-meter wind speed and SST [Nightingale et al., 2000; Huebert et al., 2010]. [11] The Nightingale et al. [2000] parameterization is used primarily in our simulations and specifies a quadratic dependence on wind speed, which implies more efficient gas transfer at higher wind speeds than the linear dependence specified in the widely-used Liss and Merlivat [1986] parameterization. Recent measurements, however, suggest that the DMS sea-to-air transfer velocity at medium wind speeds (4–12 m s−1) is at best described by a linear dependence on wind speeds [Huebert et al., 2010]. A sensitivity test using the Huebert et al. [2010] parameterization shows that our results are not strongly dependent on the difference between these parameterizations. [12] The three DMS climatologies used in this study are from Simó and Dachs [2002], Kettle et al. [1999], and Lana et al. [2011]. All three originate from a compilation of in situ surface seawater DMS concentration measurements in the Global Surface Seawater DMS Database (GSSDD, http://saga.pmel.noaa.gov/dms) begun following the publication of the climatology by Kettle et al. [1999]. We primarily use DMS concentrations from a monthly climatology following Simó and Dachs [2002], and present results using the other two [Kettle et al., 1999; Lana et al., 2011] in sensitivity studies. Simó and Dachs [2002] determined an empirical relationship between the seawater DMS concentrations in the GSSDD and simultaneous measurements of chlorophyll-a (Chl-a) concentrations at the surface and the mixed layer depth (MLD). A global DMS data set is then derived from global data sets of Chl-a and MLD. We use a Simó and Dachs [2002] DMS climatology derived from Chl-a estimates from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) monthly climatological data from 2001–2006 and ocean MLD monthly climatology from de Boyer-Montegut et al. [2004]. The widely-used Kettle et al. [1999] DMS data set interpolates the GSSDD measurements into a monthly climatology. The Lana et al. [2011] DMS climatology was constructed using the three-fold increase in GSSDD measurements over Kettle et al. [1999] projected onto biogeographic provinces and processed with objective techniques to obtain monthly fields. [13] Since SeaWiFS data do not provide reliable Chl-a estimates in areas with sea ice, persistent cloud cover or during the polar night, the Simó and Dachs [2002] DMS data set does not report high latitude DMS concentrations under extensive ice or cloud conditions. We replace missing DMS concentrations in the Southern Ocean using an average of concentrations immediately to the north of the missing data. This simple fix enables the spatial continuity of DMS emissions in the Southern Ocean without the additional impact of missing concentration values in the sea ice zone. High latitude DMS concentrations in the Simó and Dachs [2002] DMS data set are below the peak in situ measurements in the GSSDD and hence are likely a lower bound on the true concentration fields in the sea ice zone under extensive ice coverage. [14] Figure 1 shows the January surface seawater DMS concentrations from the Simó and Dachs [2002], Kettle et al. [1999], and Lana et al. [2011] DMS data sets and the corresponding mean DMS emission fluxes to the atmosphere from our simulations using the Nightingale et al. [2000] parameterization. The interpolation of sparse spatial and temporal in situ DMS measurements in the Kettle et al. [1999] data set results in extensive hot-spots and strong zonal gradients in DMS concentrations and emissions within the sea ice zone. We use the Simó and Dachs [2002] DMS climatology in our primary simulations because it lacks this strong zonal asymmetry in DMS concentrations. Figure 1Open in figure viewerPowerPoint January surface DMS concentrations (nM) from (a) Simó and Dachs [2002], (b) Kettle et al. [1999], and (c) Lana et al. [2011] DMS data sets. January climatological DMS emissions fluxes (μg m−2 d−1) in the (d) Simó and Dachs 'capped', (e) Kettle1999 'capped', and (f) Lana 2011 'capped' simulations, where DMS emissions fluxes are zero in the presence of sea ice. Scales are not linear. See section 2.8 and Table 1 for the description of the simulations. [15] Global DMS surface concentrations are difficult to estimate from present measurements, as seen in the wide variation of DMS concentrations and resulting emissions in Figure 1. Ecological and biogeochemical dynamics within planktonic communities also play a significant role in determining DMS production and hence DMS emissions [Elliott, 2009; Cameron-Smith et al., 2011]; these dynamics are not represented in the DMS seawater climatologies used in our model simulations. The use of a surface seawater DMS concentration climatology in the model neglects any interannual variability in surface DMS concentrations, though the interannual variability of DMS emissions is approximated to some extent via the interannual variability in SSTs and winds that govern the DMS ocean-to-air flux parameterizations. Other Sulfur Emissions [16] Natural (except DMS), anthropogenic, and biomass burning sulfur emissions are from the Global Emissions Inventory Activity (GEIA v.1) database (http://www.geiacenter.org/). Non-DMS emissions in all simulation years are prescribed at 1995 magnitudes to restrict interannual sulfur cycle variability to processes associated with DMS. Volcanic emissions are included in the GEIA as part of the natural source emissions, and include an average of continuous volcanic emissions over 25 years plus eruptive volcanic emissions. Ship SO2 emissions follow Corbett et al. [1999]. The model does not include sea-salt sulfate emissions, and therefore all sulfate in the model is non-sea-salt sulfate (nssSO42−). DMS Chemistry [17] The oxidative pathways of DMS are complex and not completely understood [von Glasow and Crutzen, 2004; Lucas and Prinn, 2005] and typically are simplified in large scale models. [e.g., Chin et al., 1996; Cosme et al., 2002]. DMS is thought to be oxidized in the atmosphere to SO2 and MSA primarily by the hydroxyl radical (OH) and the nitrate radical (NO3). Measurements of BrO have been shown to be high in the presence of first-year sea ice [Wagner et al., 2007; Alvarez-Aviles et al., 2008; Simpson et al., 2007], and halogen (BrO) chemistry is also known to play a role in DMS oxidation [Boucher et al., 2003; Read et al., 2008]. The magnitude, variability, and relative importance of BrO is not yet well understood. SO2 is further oxidized to sulfate and contributes the dominant fraction of nssSO42− aerosols in remote marine locations. Sulfate has other natural and anthropogenic origins which confound its relationship to DMS, but it is thought that MSA is a product only of DMS oxidation. In a study of seven DMS oxidation schemes, Karl et al. [2007] showed that the relatively simple DMS chemistry used by Chin et al. [1996] (and used in this study) reproduced observed features of the sulfur cycle extremely well compared to more complicated mechanisms. [18] The DMS-MSA chemistry in GEOS-Chem is simplified as described by Chin et al. [1996, 2000]. Yields for DMS oxidation via gas-phase reaction with OH are 100% SO2 (abstraction channel) and 75% SO2 and 25% MSA (addition channel). The fraction of DMS oxidized to MSA is larger in the high latitudes than globally, due to the temperature dependence of the OH reaction. DMS reaction with NO3 in the gas phase also yields 100% SO2, however reaction with NO3 is limited to periods of no solar insolation. Globally, 26% of DMS oxidation occurs via reaction with NO3. [19] Gas phase DMS oxidation by BrO produces dimethylsulfoxide (DMSO) which is further oxidized in the gas phase to MSA [Breider et al., 2010]. A simplified BrO oxidation mechanism was added in a sensitivity study as: [Pham et al., 1995; Boucher et al., 2002; Chatfield and Crutzen, 1990; Breider et al., 2010] using a pseudo-first order rate constant of k = 1.5 × 10−14 exp (1000/T)[BrO] [IUPAC, 2007]. This reaction assumes all DMS is oxidized to SO2 and MSA and neglects the intermediate product dimethylsulfoxide (DMSO) and its deposition, since DMSO is not explicitly included in the current model. This simplified chemistry will lead to an overestimate of MSA produced via DMS oxidation by BrO. A simple diurnal cycle was imposed by setting the rate constant to zero in the absence of sunlight. Aerosol Deposition [20] Wet deposition of aerosols is described by Liu et al. [2001] and includes contributions from scavenging in convective updrafts, rainout and washout from convective anvils and large scale precipitation, and return to the atmosphere following re-evaporation. Dry deposition velocities are computed with a standard resistance-in-series scheme based on work by Wesely [1989] as described by Wang et al. [1998]. Since deposition processes on Antarctica occur primarily on snow surfaces for both dry and wet deposition, sulfate deposition is calculated as the sum of both SO2 and sulfate for both wet and dry processes. Snow concentrations of MSA and other species are computed as , where Fi is the monthly mean deposition flux of chemical tracer i (kg m−2 d−1), and p is the monthly mean precipitation (mm d−1). Seasonal and yearly averages of MSA snow concentration are computed as precipitation-weighted means. Meteorology [21] Figure 2 shows GEOS-4 annual precipitation (mm yr−1 water equivalent) in Antarctica with maximum values (400–1000 mm yr−1) at the coast decreasing inland (10–100 mm yr−1). Comparison of GEOS-4 precipitation rates for Antarctica with those from Monaghan et al. [2006] reveal no significant bias. Monaghan et al. [2006] derived a 50-year time series of snowfall accumulation over Antarctica by combining model simulations and observations primarily from ice cores. Bloom et al. [2005] evaluated the GEOS-4 analysis data set compared to observations and other reanalysis products and note an underestimation by up to a factor of two in precipitation of the Southern Hemisphere extratropics (30–60°S) and a high bias in zonal austral summer winds in the Southern Ocean (40–60°S). Figure 2Open in figure viewerPowerPoint Mean annual precipitation rate (mm yr−1 water equivalent) from GEOS-4 meteorological fields. [22] We also conduct a sensitivity study with GEOS-5 meteorology data, which has higher boundary layer resolution than GEOS-4. We examine whether the reduced boundary layer resolution of GEOS-4 overestimates transport out of the boundary layer into the free troposphere. GEOS-5 also corrects some of the low bias in precipitation in the Southern Ocean, with increases of 25–100% over GEOS-4 precipitation in the 30–60°S region. Antarctic continental precipitation, however, is lower in GEOS-5 compared to GEOS-4. In most locations in Antarctica precipitation is within 25% of GEOS-4 values, though in some sections of West Antarctica precipitation is lower by 50% in the GEOS-5 data. The effects of the GEOS-5 meteorology on our simulations are discussed in detail in section 4.7, but the sensitivity study indicates that the choice of meteorological input does not change our conclusions. We use the GEOS-4 meteorology for our primary simulations because of the longer time period over which meteorological data is available for GEOS-Chem (1985–2006 for GEOS-4, compared to 2004-present for GEOS-5). Simulations [23] As in most sulfur models, GEOS-Chem implements a 'cap' associated with gas exchange through the sea ice, and therefore assumes that DMS emissions are zero within the sea ice by default. We experiment with relaxing this condition in our scenarios, whereby we use the same gas transfer parameterization over sea ice as over open ocean. We assume that the gas transfer parameterization over sea ice is appropriate for the sea ice fraction as well as open water among sea ice floes, which may not be realistic [e.g., Loose et al., 2011]. This is, however, useful as a first order determination of the influence of DMS emissions from within the sea ice itself. In three model scenarios we vary the DMS concentrations in the presence of sea ice from October through March. The 'capped' scenario uses this default cap on DMS emissions in the presence of sea ice in all months. The 'uncapped' scenario removes the cap from October through March, effectively removing the influence of sea ice on DMS fluxes from the ocean except via the influence of sea ice on SST. A third scenario is an 'enhanced' run where the surface DMS concentration is set to 6 nM wherever sea ice is present from October through March. Figure 3 illustrates the DMS concentrations imposed in the sea ice in each of these scenarios. We also examine the difference between the uncapped and capped runs to examine the impact of DMS emissions only from within the sea ice. Figure 3Open in figure viewerPowerPoint Illustration of seawater DMS concentrations applied in the sea ice zone for each of the scenarios. (top) DMS concentrations for the DMS data set are applied to the open ocean and the sea ice in the UNCAPPED scenario (solid red line), with no influence of sea ice. For the CAPPED scenario, DMS concentrations are set to 0 (dashed red line). In the ENHANCED scenario, DMS concentrations are set to a scalar value above the mean DMS concentrations of the ice extent (fine dashed red line). (bottom) The difference (UNCAPPED minus CAPPED) between two scenarios allows evaluation of effect of DMS concentrations of the DMS data set from within sea ice only (red dot-dot-dashed line). In all simulations, the sea ice expands and retreats through its seasonal cycle (between maximum and minimum ice extent). Each scenario is applied for the to the sea ice extent at each point in time for the months October–March. All simulations have a 'capped' scenario for April–September. [24] Table 1 summarizes all simulations discussed in the paper. Our primary simulations use the Simó and Dachs [2002] DMS climatology with the Nightingale et al. [2000] gas transfer parameterization under the capped, uncapped, and enhanced scenarios as described above. These are referred to in the text below as 'Simó and Dachs capped', 'Simó and Dachs uncapped', etc. The Simó and Dachs [2002] DMS data set has a mean DMS concentration from October-March over the climatological sea ice area of 1.2 nM, with a peak concentration of 2.0 nM in January. The 6 nM concentration is thus 5 times larger than the October-March mean of the uncapped run. We compute the difference between the uncapped and capped and the enhanced and capped model scenarios to isolate the effect of DMS emissions from within sea ice itself. The Simó and Dachs capped, uncapped, and enhanced simulations are run for 17 years from July 1985 through June 2004 after a 1.5 year spin up, excluding the period July 1997 through June 1999. This overlaps the period of satellite observations of sea ice as well as several ice core records of MSA concentrations. Table 1. Summary of Simulations and Sensitivity Simulations Discussed in the Text Simulation Name DMS Conc in Ice DMS Data Seta DMS Emissions Parameterizationa Simulation Length Sensitivity Test Simó and Dachs capped 0 SD2002 N2000 17 years (7/1985-6/2004)b Simó and Dachs uncapped Oct-Mar: Data set; Apr-Sep: 0 SD2002 N2000 17 years (7/1985-6/2004)b Simó and Dachs enhanced Oct–Mar: 6 nM; Apr–Sep: 0 SD2002 N2000 17 years (7/1985–6/2004)b Huebert2010 capped 0 SD2002 H2010 3 yrs (7/1985–6/1988) Emissions linear with wind speed Huebert2010 uncapped Oct–Mar: Data set; Apr–Sep: 0 SD2002 H2010 3 yrs (7/1985–6/1988) Emissions linear with wind speed BrO capped 0 SD2002 N2000 3 yrs (7/1985–6/1988) BrO oxidation mechanism BrO uncapped Oct-Mar: Data set; Apr-Sep: 0 SD2002 N2000 3 yrs (7/1985–6/1988) BrO oxidation mechanism GEOS-5 capped 0 SD2002 N2000 2 yrs (7/2005–6/2007) GEOS-5 meteorological data set GEOS-5 uncapped Oct-Mar: Data set; Apr-Sep: 0 SD2002 N2000 2 yrs (7/2005-6/2007) GEOS-5 meteorological data set Kettle1999 capped 0 K1999 N2000 6 yrs (7/1985–6/1991) Alternate DMS data set Kettle1999 uncapped Nov–Mar: Data set; Apr–Oct: 0 K1999 N2000 6 yrs (7/1985–6/1991) Alternate DMS data set Lana2011 capped 0 L2011 N2000 3 yrs (7/1985–6/1988) Alt. DMS data set emissions only Lana2011 uncapped Oct–Mar: Data set; Apr–Sept: 0 L2011 N2000 3 yrs (7/1985–6/1988) Alt. DMS data set emissions only a