At its greatest extent,Antarctic sea ice covers a marine area larger than that of the continent. Its seasonal formation and melting greatly affects the climate of marine and coastal areas, essentially through changes in radiative energy and mass exchange processes. As the albedo of sea ice covered by fresh snow can be as high as 90 %, in the Southern Ocean during winter most incoming solar radiation is reflected back (King and Turner 1997). The capping of the ocean prevents the exchange of heat, gases and moisture between the sea and atmosphere, and has important effects on the stability of the atmosphere, cloud formation and precipitation. Much lower albedos and significant fluxes of heat and moisture occur in areas with open water surrounded by sea ice and/or land ice, such as leads and polynyas.
Even when the ice cover in the Southern Ocean is at its maximum, about 20% of the marine area is ice-free (Gloerson et al. 1992). Persistent coastal polynyas are usually formed by the drainage flow of katabatic winds and by the continuous freezing of water and advection of formed ice by the wind, which removes a large amount of latent heat from surface waters, thereby producing cold, saline, dense waters. These water masses have important implications in deep-water formation and the primary productivity of shelf areas (Massom 1988). Bromwich (1989) estimated that a small (about 50x50 km) and recurring coastal polynya in Terra Nova Bay produces about 10 % of sea ice in the entire Ross Sea region. Polynyas can also form in the open ocean (Fig. 24). In the eastern Weddell Sea, for instance, a polynya of about 2.5x105 km2 developed roughly at the same location for three consecutive winters (1974-1976; Zwally and Gloersen 1977). Carsey (1980) found upwelling of water in the centre of this large ice-free area, while the edge was characterised by thermohaline convection, formation of ice and its transport away from the polynya. It was thought that these characteristics of the polynya were related to the upwelling of warm subsurface water over a topographic high in the seabed (the Maud Rise; Comiso and Gordon 1987). It was later suggested (Enomoto and Ohmura 1990) that polynya formation was probably due to their location beneath the circumpolar trough, in an area with variable position of surface wind fields. The formation of coastal and open ocean leads within sea ice seems essentially due to interactions between ocean surface and atmospheric circulation patterns, especially the strength and persistence of wind (Bromwich et al. 1998). A regional warming and/or change in the extent of sea ice will affect air mass circulation, wind regimes and the formation of polynyas, with consequent changes in ice production,formation of dense water masses,spring disintegration of sea ice, and marine productivity.
Antarctic sea ice is not confined by land margins, and it exhibits large inter-annual and regional variability throughout the year (Gloersen et al. 1992). The annual cycle of expansion and contraction in a roughly concentric zone around Antarctica is controlled by the equilibrium between air temperature, upper ocean structure, pycnocline depth, wind direction and strength, leads, and variation in the ACC and other major currents. Another factor affecting the nature and extent of sea ice is the depth and thermal properties of overlying snow, which controls most of the radiative exchange and sub-ice primary productivity. The exact role of overlying snow is largely unknown, and this parameter is difficult to model (Iacozza and Barber 1999). Besides the lack of reliable data on the thickness of snow, a further complication for modelling studies is the timing of snowfalls in relation to the formation of sea ice (Barber and Nghiem 1999).
Most scientists believe that Antarctic sea ice is very sensitive to climate change, and that it will provide an early indication of global warming. A reduction of the area covered by sea ice will increase the absorption of solar radiation, determining a further increase in temperature and change in atmosphere-sea coupling. However, at present there is no firm evidence indicating any significant long-term trend in the extent of Antarctic sea ice (King and Turner 1997), although it cannot be excluded that an early effect of warming could be a reduction of the total mass of Antarctic sea ice, rather than its extension.
Through the analysis of whaling records, de la Mare (1997) suggested that the Antarctic summer sea-ice edge moved 2.8 degree of latitude southwards between the mid-1950s and early 1970s. Based on an atmosphere-ocean sea ice model,Wu et al. (1999) concluded that the extent of sea ice in the Southern Ocean decreased by 0.4-1.8° latitude over the 20th century. Broad-scale records from passive microwave imagers flown on polar orbiting satellites have shown a decrease in the late winter maximum (between 1973 and 1977) and summer minimum (especially in 1980 and 1981; Gloersen et al. 1992). However, the situation reversed during the 1980s, with an increase in the winter maximum and summer minimum.At present, in spite of the growing satellite database (about 30 years), no statistically significant overall trends have been detected (Gloersen et al. 1992; Johannessen et al. 1994; Jacka and Budd 1998); only some local trends, such an increase in sea ice in the sector from 0 to 40° E and a larger decreasing sector from about 65 to 160° W in the Bellingshausen and Amundsen seas, have been detected.
As for the future, according to a CSIRO coupled model (Gordon and O'Far-rell 1997), a reduction of about 25-45 % of the sea-ice volume is predicted for a doubling of CO2 concentrations and a global warming of 2.1 °C. Through a quite similar coupled atmosphere-sea ice model, Wu et al. (1999) calculated that a global warming of 2.8 °C and higher albedo feedback by surface snow would reduce the extent of Antarctic sea ice by about 2° latitude. However, interactions among sea ice, atmosphere and ocean are complex and largely unknown. The scarce knowledge of a number of parameters such as the thickness and density of sea ice and that of the overlying snow makes it impossible to depict reliable scenarios. In current models it is very difficult to represent even detailed aspects of sea-ice distribution and timing.
In addition to brine released by the freezing of seawater, especially in coastal polynyas, the formation of dense water masses on the Antarctic continental shelf is also due to the cooling of seawater by submerged ice flowing off the continent, especially in the large Ross and Filchner-Ronne ice shelves. When flowing under deep ice shelves, seawater may reach temperatures below -1.95 °C (i.e. below the typical surface freezing point of-1.91 °C; Foldvik and Gammelsr0d 1988). This "supercooled" water spills off the shelf, descends the continental slope and mixes with Warm Deep Water to form Antarctic Bottom Water. Significant changes in the extension of Antarctic ice shelves will thus affect water circulation in the Southern Ocean.
As discussed in previous chapters, it is not yet clear whether the warming of the Antarctic Peninsula is in response to global changes or whether it is a natural, exaggerated local response to regional climate variations. Whatever the case, the Antarctic Peninsula shows a remarkable atmospheric warming trend, and its most notable effect is the progressive and continuous retreat of the Wordie, Müller, George VI and Wilkins ice shelves on the west coast, and of the Larsen (A and B) Ice Shelf on the east coast. These events are attracting considerable media coverage and scientific interest, but available data are still inadequate to identify significant changes in ocean circulation and impacts on the marine environment. The Antarctic Peninsula is not well resolved in the present generation of coupled models, and the models cannot help forecast possible changes in the near future. As ice shelves float, their melting has probably no effect on sea levels; moreover, as shelves are usually replaced by the formation of a sea-ice cover, there is probably no dramatic change in the overall albedo. There is, however, evidence of relationships between the warming trend and regional changes in sea-ice extension (Stammerjohn and Smith 1996) and sedimentary processes. Domack and McClennen (1996) found a particularly high sedimentation rate of terrigenous material in two cores collected in the northern region of the western Antarctic Peninsula. This process was consistent with historical records of increasing surface air temperature and was due to the enhanced input of glacial meltwater, which is usually laden with large amounts of terrigenous materials.
Analyses of bones and other organic remains in abandoned penguin rookeries in the near vicinity of Palmer Station show that they were inhabited by Adelie penguins for at least the past 600 years. No remains of chinstrap or gentoo penguins were found (Emslie et al. 1998). The latter two species currently inhabit the Palmer region, where they are believed to have established in the last 20-50 years (Parmelee 1992). Given that Adelie penguins are obligate associates of winter pack ice, whereas their congeners, the chinstrap penguins, occur almost exclusively in ice-free waters (Ainley et al. 1994), it has been hypothesised that changes in seabird distribution are related to regional warming in the Antarctic Peninsula (Fraser et al. 1992; Smith et al. 1999).
In contrast to Antarctic Peninsula ice shelves, those on the continent will probably not be significantly affected by global warming in the next century (Vaughan and Doake 1996); in any case, long-term responses of ice shelves to warming are uncertain, and possible consequences for Southern Ocean circulation are unpredictable. A model study by Warner and Budd (1998) shows that with a global warming of 3 °C, it will take at least several centuries for Antarctic ice shelves to disappear. The warming will initially increase the melting of ice, thereby introducing a layer of freshwater (Jenkins et al. 1997; O'Farrell et al. 1997); however, this layer will probably mitigate melting if it is not flushed away by currents or tides.
3.2.5 Biogeochemical Cycles of C, Fe, S and Other Elements in the Southern Ocean
About 3.5 % (by mass) of seawater consists of solutes, most derived from weathering of terrestrial ecosystems, which are carried by rivers to the sea. As in Antarctica, there are only very small ice-free areas and no rivers; the input of organic matter and soluble ions of continental origin in the Southern Ocean is therefore negligible. The main inputs are from the atmosphere (espe cially volcanic gases and particulate materials, which become condensation nuclei and fall out in precipitation), and hydrothermal vents which release He, Ca, Mn, K, H and other elements from the oceanic crust. The loss of elements mainly occurs through biological processes (such as organisms forming siliceous frustules or calcareous shells), the formation of sea spray, or through the adsorption of ions to clay particles and new minerals in sediments. Thus, in the Southern Ocean many elements such as Al, Bi, Ce, Co, Pb, Mn and Th, which have prevailing terrestrial sources and are highly reactive in seawater, are probably scavenged by particulate matter and may occur in lower concentrations than in other seas.
The Southern Ocean plays a key role in the oceanic cycle of Si. The downward flux of highly silicified frustules of diatoms (the so-called silicate pump, Dudgale et al. 1995) is a very important transfer mechanism of Si, C and other elements of the upper water column which are incorporated into diatoms (e.g. Ra, Ba and Ge; Holm-Hansen 1985). Diatomaceous ooze accumulated on the Southern Ocean floor represents a large proportion of total recent Si deposition throughout the world ocean. An increasing body of evidence (e.g. Nelson et al. 2001; Sigmon et al. 2002) indicates that silicic acid concentrations are very low in very productive areas of the Southern Ocean, and that Si may constitute a limiting factor for primary production.
Because of its high solubility and chemical reactivity, CO2 is taken up by the oceans much more effectively than most other gases released by anthropogenic sources. The total amount of carbon in the ocean is about 50 times higher than in the atmosphere; however, because CO2 solubility is temperature dependent, net fluxes show regional and seasonal patterns. As the cooling of surface waters tends to drive CO2 uptake, while warming drives outgassing, the Southern Ocean plays a very important role in the uptake of C, producing vertical gradients and its transport from polar to tropical regions in dense bottom waters. The seasonal and regional distribution of CO2 is also driven by primary production (about 100 petagram C year1; petagram, Pg; Falkowski et al. 1998). Part of this C is transformed into dissolved inorganic carbon through autotrophic respiration, while the remainder is the net primary production (on the basis of global remote sensing data, estimated to be about 45 Pg C year1; Falkowski et al. 1998). According to these authors, the global export production (i.e. the sinking of dead organisms, detritus and dissolved organic C) ranges from 10 to 20 Pg C year1, but only a small fraction of C sinks into the sediments. Heterotrophic respiration at depth converts the remaining organic matter back into dissolved inorganic C, which is transported by deep-water masses to other locations where it upwells and re-equilibrates with atmospheric CO2. The presence of dissolved inorganic C at depth contributes to lower concentrations of atmospheric CO2 (about 200 ppm; Maier-Reimer et al. 1996).
The formation of calcium carbonate shells by marine organisms depletes surface carbonate ions, and reduces alkalinity and uptake of CO2 from the atmosphere. Thus, although overall ocean productivity is largely determined by nitrate and phosphate (and silicon for specific types of phytoplankton) supplied from deep water, it is also affected by CaCO3 formation in surface waters.
In the Southern Ocean the production of CaCO3 by marine organisms is much lower than in temperate and tropical seas, and the water column is rather unstable due to dominant upwelling, ice formation and melting. The distribution pattern of chemical elements is therefore less "structured" than at lower latitudes. The seasonal cycle of primary productivity is mainly limited by insufficient light penetration during the austral spring (Smith and Gordon 1997), and by Si or micronutrient availability in summer (Boyd et al. 1999). The depletion of nitrate and phosphate concentrations in surface waters, a regular feature of many marine areas, rarely occurs in the Southern Ocean. This is partly the result of upwelling and of the high rate of nutrient recycling by microbe populations in the euphotic zone (N, for instance, is recycled six to seven times before settling to deeper waters as particulate N; Holm-Hansen 1985). Like in the equatorial Pacific, the surplus of nutrients in the Southern Ocean (Coale et al. 1996) is often associated with a relatively low phytoplankton biomass (the Antarctic paradox; Treguer and Jacques 1986). There is evidence (Martin et al. 1990; Boyd et al. 2000) that High-Nutrient, Low-Chlorophyll (HNLC) regions are characterised by low concentrations of certain biolimiting trace elements such as Fe and Mn. The artificial addition of Fe to Southern Ocean seawater stimulates phytoplankton growth and increases the uptake of atmospheric CO2 (e.g. Timmermans et al. 1998; Boyd et al. 2000). A recent study by Hiscock et al. (2003) in the Pacific sector of the Southern Ocean (from 54 to 72° S) shows that in zones where photosynthetic performance was low, Fe-enrichment response was high; on the contrary, where performance was high (low-silicic acid waters), the Fe-enrichment response was low. Based on these results and on silicic acid limitation (Nelson et al. 2001; Sigmon et al. 2002; Hiscock et al. 2003), there is reason to believe that the region south of the southern boundary of the ACC and north of the continental margin would respond positively to the addition of Fe; the zone between the APF and the southern boundary of the ACC (the Seasonal Ice Zone) would give the same response in spring but not in summer (during the latter season the zone has little silicic acid, while Fe is sufficient) and, finally, the region north of the APF would not respond to the addition of Fe.
As will be discussed in Chapter 5, on Antarctic continental shelves the concentration of most trace elements in seawater are similar to or higher than those reported elsewhere and seem adequate to sustain biological growth. In contrast, offshore surface waters receive scarce quantities of lithophilic elements from shelves, melting icebergs, rivers or aeolian transport. Owing to very low concentrations of Fe and other trace elements, phytoplankton can use <10% of available major nutrients (Martin et al. 1990; Westerlund and Ohman 1991a; Fung et al. 2000). Long-term series data and global surveys indicate that oceanic nitrogen fixation varies spatio-temporally and is sensitive to climatic conditions (Hansell and Feely 2000). It has been hypothesised that, over glacial-interglacial timescales, Fe can indirectly influence the nitrate content of oceans (Falkowski et al. 1998). Martin et al. (1990), for instance, postulated that 50-fold higher aeolian Fe supplies to the Southern Ocean during the last glacial maximum could have stimulated enough new productivity to reduce atmospheric CO2 from 280 to less than 200 ppm. However, more recent circulation and mixing models predict that the effect of increased nutrient uptake on atmospheric CO2 concentrations would be much lower, and that a more efficient Southern Ocean biological CO2 pump alone cannot explain lower CO2 levels during ice ages (Watson et al. 2000). Stephens and Keeling (2000) suggested that low atmospheric CO2 levels during glacial periods may result from reduced deep-water ventilation associated with year-round Antarctic sea-ice coverage or wintertime coverage, combined with ice-induced stratification during summer.
Although the ongoing debate on nutrient uptake in the Southern Ocean underlines the importance of studying how it is regulated and whether it has changed in the recent geologic past, it is now believed (IPCC 2001) that the role of marine organisms in driving the C cycle in oceans is probably close to the steady state, and that the oceanic uptake of CO2 is above all a physically and chemically controlled process. Indeed, in deep ocean water, the concentrations of nutrients and dissolved organic carbon are closely correlated and their ratios match the nutritional requirements of phytoplankton. Primary production may thus have little potential to drive a net air-sea transfer of C.
Hypothetically, oceans can incorporate most of the C released by anthropogenic activity, but the uptake of atmospheric CO2 mostly occurs in oceanic regions where waters which have spent many years in the ocean upwell (i.e. are re-exposed to the contemporary atmosphere which contains a greater amount of C due to human activity; Doney 1999). Owing to the finite rate of exposure of "older" and deeper waters to the atmosphere, the uptake process will take several hundred years to complete; therefore, even assuming that there is no further increase in anthropogenic emissions of C, atmospheric concentrations of CO2 will increase. Deep ocean sediments may also contribute to reducing atmospheric C contents through CO2 reaction with CaCO3, but a response time of about 5,000 years has been estimated (Archer et al. 1997). Studies on the Southern Ocean by Caldeira and Duffy (2000), for instance, have shown high fluxes of anthropogenic CO2 but very little burial of organic carbon in sediments. Notwithstanding this, the Southern Ocean is very important in determining the uptake of CO2 from the atmosphere (Fig. 25). Most "in-situ" measurements performed during the last decade have shown undersaturation (e.g. Robertson and Watson 1995; Stoll et al. 1999), although local source areas are also present (Bakker et al. 1997). Through the application of a one-dimensional model to quantify the distribution of CO2
sources and sinks in the Southern Ocean and to simulate its variability over the period 1986-1994, Louanchi and Hoppema (2000) calculated a mean uptake of 0.53 Pg C year-1, with an increase between 1986 and 1994. The inter-annual variability (0.15 Pg C year-1) was related to the Antarctic circumpolar wave, which affects the sea surface temperature, wind speed and sea-ice extent. These estimates agree with the results of other studies based on either field data or models (e.g. Poisson et al. 1994; Bakker et al. 1997), and indicate that in the early 1990s the Southern Ocean may have helped decrease the atmospheric CO2 growth rate by about 3-5 Pg C year-1. Global warming experiments using coupled ocean-atmosphere models show that under climate-change forcing, the CO2 uptake rate is likely to decrease more in the Southern Ocean than in any other ocean (Sarmiento and Le Quere 1996). The increased stratification of the upper water column will cause a decrease in vertical mixing along isopycnals, vertical transport of C, convective overturning, and upward mixing of warm waters from below (Sarmiento et al. 1998). The temperature of Southern Ocean surface waters will thus increase less than elsewhere, and may even decrease in some regions (Manabe and Stouffer 1994). The increased stratification will also cause a gradual collapse of thermohaline circulation in the entire deep ocean, which has a major role in the C cycle on century timescales.
The surface layer of the ocean influences the climate system not only through the exchange of greenhouse gases but also through the release of dimethylsulphide (DMS), which is thought to exert a cooling influence. DMS is a breakdown product of dimethylsulphoniopropionate (DMSP), a constitutive osmoprotectant of algal cells which is released by exudation and through autolysis, grazing or bacterial and viral attacks. Oceanic emission of DMS rep resents the largest biogenic flux of S to the atmosphere (Andreae 1990; Kettle and Andreae 2000). Once in the atmosphere, DMS forms sulphur dioxide, sulphates and methane sulphonic acid. Sunlight-scattering sulphur aerosols and cloud condensation nuclei can potentially affect the radiative balance and global climate system. Charlson et al. (1987) hypothesised a phytoplankton-climate link, and that the global warming trend could be mitigated to some extent by increased DMS emission stimulated by the warming. However, the existence of this homeostatic feedback and the strength of any cooling effect by increased DMS emission are still uncertain (Andreae and Crutzen 1997). Moreover, any such feedback acting through the S cycle must include other climatic effects. Global warming, for instance, decreases the uptake of CO2 by oceans, leading to its accumulation in the atmosphere and producing an opposite effect with respect to DMS (Kiene 1999).
In spring and summer, Antarctic coastal waters and marine areas at the retreating pack-ice edge are characterised by phytoplankton blooms, with concentrations higher than 100 nM, i.e. much greater than the average 3 nM in the world's ocean (Gibson et al. 1990; di Tullio and Smith 1995). There is evidence that higher values of the DMSP:chl a ratio are often associated with phytoplankton assemblages dominated by the colonial prymnesiophyte Phaeocystis antarctica. Research by di Tullio et al. (1998) in the southern Ross Sea found that ice diatoms are sometimes as important as P. antarctica in DMSP production. The Southern Ocean plays a prominent role in the total global flux of DMS into the atmosphere, because diatoms and P. antarctica represent a large proportion of Antarctic phytoplankton, and areas with seasonal and intense algal blooms extend for several million square kilometres. Projections based on the assumption that DMS fields and ice cover will not change between the year 2000 and 2100 indicate that the global flux of DMS will increase in the next century from 26.0 to 27.7 tg S year-1, and localised increases are foreseen in areas of the Southern Ocean immediately adjacent to the continent (IPCC 2001). However,the depletion of the ozone layer increases the amount of ultraviolet radiation reaching the ocean surface, and there is evidence that this radiation inhibits DMPS production by P. antarctica and increases the oxidation of DMS, thus reducing the flux of DMS to the atmosphere (Hefu and Kirst 1997).
Was this article helpful?