The present phase of climate change has drawn considerable attention to the behaviour of Antarctica and the Southern Ocean (e.g. Gille, 2002; Jacobs et al., 2002; Walther et al., 2002; Cook et al., 2005). However, confident identification of any effect, especially in the oceans, has been hindered by; (i) sampling bias, (ii) the short history of observations, which typically only encompass the last 50-60 years, (iii) a multiplicity of forcing mechanisms, some with and some without clear connections to a warming climate and (iv) marked variability at a range of temporal and spatial scales (e.g. Jacobs and Giulivi, 1998; Orsi et al., 2001; Vaughan et al., 2001). Of special note is the Southern Annular Mode (SAM), which appears to dominate inter-annual to centennial variability in the Southern Ocean (Hall and Visbeck, 2002; Lovenduski and Gruber, 2005; Lovenduski et al., 2007). SAM is essentially a zone of climate variability that encircles the South Pole and strongly influences zonal winds, sea ice formation and oceanic circulation. When in a positive phase, SAM is typically associated with enhanced westerly winds over the ACC and weakened winds further north. This favours a strengthening of the ACC, a reinforcement of the northward Ekman drift and subduction of surface waters, and a northward expansion of sea ice. To compensate, there is enhanced poleward transport and rising of deep water at the Antarctic continental margin. Under a negative SAM, windiness and storminess appear to migrate to mid-latitudes thus weakening both zonal and meridional ocean transport.
Despite uncertainties associated with the aforementioned limitations, it is nonetheless important to examine and evaluate recent changes in Antarctica and the Southern Ocean in light of their actual or potential influence on the global ocean and climate.
Climatic trends for the past 50 years have been identified by Turner et al. (2005) using records from 19 Antarctic meteorological stations. The results emphasise the marked geographic variability of the continental climate (e.g. Vaughan et al., 2001) as well as its temporal variability at interdecadal scales. The Antarctic Peninsula has warmed at a statistically significant rate of +0.56°C/decade from 1951 to 2000. The next largest warming trend outside of the Antarctic Peninsula is in the western Ross Sea (Scott Base) with a rise of +0.29°C/decade, although the rise is not statistically significant. Elsewhere, significant trends are unclear. Annual temperature trends for coastal and interior sites on the East Antarctic Ice Sheet suggest a slight cooling. However, all but one site exhibited warmer winters.
Notwithstanding its temporal and spatial variability, the upper Southern Ocean has warmed between 1955 and 2003 in concert with the world ocean (Levitus et al., 2005). Off the Antarctic Peninsula, the pronounced atmospheric warming has been accompanied by an equally marked warming of the surface ocean with summer temperatures increasing by 1.2°C over the latter half of the twentieth century (Meredith and King, 2005). At water depths of 700-1,100 m, Gille (2002) reported an average 0.17°C rise since 1950, which accounts for about two-thirds of the total increase in heat content in the ocean from 0 to 3,000m depth (IPCC, 2007). Much of the warming is concentrated within the Subantarctic Front of the ACC. This raises the possibility that the warming at depth may result from the sinking of atmospherically warmed SAMW. Gille (2002) further suggests that warming may also be related to a general southward displacement of the ACC. This would be consistent with a suite of model simulations that suggest warming of the Southern Ocean will be accompanied by a southward shift of zonal westerly winds together with a narrowing and intensification of the ACC (IPCC, 2007 and references therein).
Coherent historical trends are also evident for salinity. Subpolar regions have generally become fresher between 1955 and 1998 in contrast to subtropical and tropical regions, which display increased salinity with the exception of the central Pacific Ocean (Curry et al., 2003; Boyer et al.. 2005). Such freshening of the upper Southern Ocean may be responsible for a reduction in the salinity of AAIW (Wong et al., 1999; Curry et al., 2003). Reduced salinity implies increased freshwater input that may result from one or more of the following causes: (i) greater net precipitation; (ii) changes in the extent of sea ice - a major contributor to winter salinity through brine rejection; (iii) increased melting of ice shelves, ice sheets and glaciers (e.g. Cook et al., 2005), and (iv) changes in the oceanography, especially any reduced upwelling of saline LCDW (Wong et al., 1999; Jacobs et al., 2002; Curry et al., 2003). Nearer Antarctica, hydrographic records spanning over 40 years show local variability in salinity trends. The upper 50 m of the ocean, west of the Antarctic Peninsula, has become more saline although underlying waters have freshened slightly in line with the regional trend. While the more saline surface conditions appear to be out-of-step with the strong glacier retreat on the Peninsula (e.g. Cook et al., 2005), Meredith and King (2005) suggest more saline conditions are consistent with the reduced sea ice cover and the timing of the hydrographic measurements. With less sea ice production there is less freshening of the ocean in the summer when most of the hydrographic measurements are made. On the opposite side of the continent, at Law Dome, ice core records spanning a century and longer, identify a 20% loss of sea ice since 1950 although this trend is strongly overprinted with cyclical variations with an ~11 year frequency (Curran et al., 2003). Reduced sea ice along with increased precipitation and melt water from the West Antarctic Ice Sheet have been cited as contributing to the observed freshening of surface waters associated with the Ross Sea Gyre (Jacobs et al., 2002). But like Law Dome, the trend is obscured by cycles, this time by 5-6- and 9-year oscillations in HSSW formation (Assmann and Timmerman, 2005).
Salinity and temperature changes have the potential to affect bottom water production and the THC. Consequently, these changes have received considerable attention (IPCC, 2007 and references therein). At glacial-interglacial time scales, palaeoceanographic evidence reveals marked variations in the position and degree of convective overturning of the N Atlantic sector of the THC (Rahmstorf, 2002 and references therein). During interglacial periods, overturning is most active and reaches its most northerly extent. In contrast, glacial periods are likely to witness a southward shift and possible slow-down in overturning. Any slow-down may be compensated by a greater production of bottom water from Antarctica as suggested by grain size (Hall et al., 2001), magnetic properties (Venuti et al., 2007) and diatom proxies (Stickley et al., 2001). However, such conclusions are sometimes at odds with geochemical tracers such as 813C (e.g. Moy et al., 2006) that point to little change in the passage of NADW through the Indian and Pacific sectors at least over recent glacial-interglacial cycles. At millennial time scales, abrupt changes such as those associated with Heinrich events, may stop N Atlantic overturning altogether as the density of the surface ocean is reduced by rapid influxes of melt water (Rahmstorf, 2002). Again, cessation of N Atlantic production may be compensated by enhanced Antarctic production. However, responses to the latest phase of climate/ocean warming are unclear. In the N Atlantic, which is the best observed deep-water source, long-term trends are equivocal due to decadal variability, a paucity of long-term observations and other factors (IPCC, 2007). A similar situation applies to Antarctica where estimations of bottom-water production are inconsistent in response to: (i) natural cycles; (ii) differences in the definitions and techniques to estimate production rates, and (iii) a bias towards summer observations (Jacobs, 2004). On the basis of chlorofluor-ocarbons and 14C data, which allow water masses to be traced at decadal to century scales respectively, Orsi et al. (2001) revealed no decline in bottom water production over the twentieth century as indicated earlier by Broecker et al. (1999). Nevertheless, the changes recorded in recent historical times cannot be ignored. The freshening of the Ross Gyre over the last 50 years (Jacobs et al., 2002) and an accompanying downstream freshening of AABW in the adjacent Australia-Antarctic Basin (Aoki et al.. 2005) are consistent with increased freshwater input. On a larger scale, the historical salinity data of Curry et al. (2003) also reveals a freshening of deep and bottom water at Antarctic and N Atlantic sources.
Simulations by 19 model runs under IPCC greenhouse gas scenario, A1B (rapid economic growth, world population peaks mid-century, new and efficient energy technologies with reliance on a range of sources) point to an average 25% reduction in N Atlantic overturning by the year 2100 (IPCC, 2007).
None of the runs point to a shut-down; rather they favour reductions in overturning of up to 50%. While the Southern Ocean sector has received less attention from modellers, simulations based on a warmer or fresher ocean may enhance or stabilise N Atlantic overturning (Saenko et al., 2003; Weaver et al., 2003). To further emphasise the complexity of north-south relationships, the projected strengthening of the Southern Hemisphere westerly winds will increase the northward Ekman transport of upper ocean. To compensate, the poleward flow of deep water below 2,000-2,500 m depth, could be expected to strengthen and possibly stimulate the southward flow of NADW (e.g. Toggweiler and Samuels, 1995; Toggweiler et al., 2006). Again, such a trend is overprinted with marked inter-annual variability.
Because of the importance, size and complexity of the Southern Ocean, the incompleteness of observations, and its high variability at a range of temporal and spatial scales, it is critical to improve our knowledge of this ocean/climate system. To re-emphasise the introduction to this chapter, the Southern Ocean has a profound influence of the distribution of salt, heat and ventilating gases throughout global seas. At the same time it is also undergoing some of the most rapid environmental changes on Earth highlighted by the warming, glacial retreat and ice shelf collapse around the Antarctic Peninsula. Thus, to address the inevitable questions relating to impacts of a rapidly changing climate on the Southern Ocean we require a strong modelling effort, supported by multi-seasonal oceanographic and remotely sensed observations and high-resolution palaeoceanographic records of past warm extremes. While this may seem to be a well-worn message, it is still appropriate at a time of certain change and uncertain consequences.
Was this article helpful?