The Southern Ocean is an active component of the climate system that causes natural climate variability on time scales from years to centuries. It is a major thermal regulator and has the potential to play an active role in climate change by providing important feedbacks. If climate change is sustained, it is likely to produce long-term and perhaps irreversible changes in the Southern Ocean. These changes will alter the nature and amount of life in the ocean and on surrounding islands, shores, and ice. The impact of climate change is likely to be manifest in the large-scale physical environment and in biological dynamics. There even is the possibility that changes in the Southern Ocean could trigger abrupt and very long-term changes that affect the climate of the entire globe.
The Southern Ocean is of special significance to climate change because of
the variety of water masses produced there and the ability of these waters to
spread throughout the global ocean. In particular, shelf waters with temperatures
near the freezing point (about -1.9°C) are produced along the margin of
the Antarctic continent. They contribute to the formation of Antarctic bottomwater
and continuing ventilation of this water in the world's oceans (Whitworth
et al., 1998). There already is evidence that climate change has altered SAMW
and AAIW (Johnson and Orsi, 1997; Wong et al., 1999; Bindoff and McDougall,
2000). Wong et al. (1999) compared trans-Pacific sections 20 years apart and
found temperature and salinity changes that are consistent with surface warming
and freshening in formation regions of the water masses in the Southern Ocean
and their subsequent subduction into the ocean.
The main contributions to the production of Antarctic bottomwater come from
cooling and transformation of seawater under the floating Filchner-Ronne and
Ross ice shelves, as well as from dense water formed as a result of freezing
of sea ice. Polynyas also are important. These areas of combined open water
and thin sea ice (surrounded by sea ice or land ice) play an important role
in ocean-atmosphere heat transfer, ice production, formation of dense shelf
water, spring disintegration of sea ice, and sustenance of primary and secondary
productivity in polar regions. Their formation is linked to the strength and
persistence of cold outflow winds off the Antarctic continent and the position
of the polar jet stream and Southern Ocean atmospheric circulation patterns
(Bromwich et al., 1998).
Research by Wu et al., (1997), Budd and Wu (1998), and Hirst (1999)using
the CSIRO coupled, transient model and IPCC IS92a emissions scenariosuggests
that certain aspects of Southern Ocean circulation may be very sensitive to
climate change. With an increase in atmospheric equivalent CO2, there
is a substantial increase in the strength of the near-surface halocline and
a potential reduction in the formation of Antarctic bottomwater and convection
in the Southern Ocean. The decrease in downwelling of cold, dense water, as
shown in Figure 16-5, leads to a slowdown in the
thermohaline circulation (Broecker et al., 1985). A tripling of GHGs
is sufficient to induce a shutdown of the thermohaline circulation. Furthermore,
this shutdown continues for at least several centuries, to the end of the model
integration. The main processes causing a decline in the thermohaline circulation
are an increase in precipitation minus evaporation, an increase in glacial melt
from the Antarctic, and reduced sea-ice formation, so there is less brine injection
into the ocean.
According to this simulation, if concentrations of GHGs continue to grow to the end of the 21st century, there will be significant weakening of the ocean's thermohaline circulation and even potential for its shutdown. These changes might become irreversible in subsequent centuries, despite possible stabilization of GHG concentrations. Modeling by Cai and Baines (1996) indicates that the Antarctic Circumpolar Current is driven at least partly by the thermohaline circulation associated with the formation of Antarctic bottomwater, so this major current of the Southern Ocean also can be expected to slow down. This result is confirmed by Cai and Gordon (1998), who found a clear decreasing trend in the magnitude of the Antarctic Circumpolar Current as CO2 increases. The foregoing results indicate an ocean circulation that is significantly more sensitive to climate change than that found by Manabe and Stouffer (1994). The possibility remains that certain factors and parameterizations in the Hirst (1999) model could make it overly sensitive and that future changes may not be as dramatic as indicated. For example, in other models the ocean component is much less diffusive, allowing for prolonged maintenance of an enhanced halocline. None of the present generation of climate models can resolve polynyas, where the majority of brine rejection takes place and dense water masses form. These processes need to be better represented in coupled ocean-climate models before the timing of any substantial slowing of Antarctic bottomwater production or of any shutdown of the ocean thermohaline circulation can be predicted confidently. Changes in this circulation also may be driven by variable fluxes at lower latitudes, and results from different ocean models indicate uncertainty about the magnitude and timing of thermohaline slowdown (see TAR WGI Chapter 7 for more detailed discussion of the global thermohaline circulation).
The sensitivity of the Southern Ocean to changes in ice shelves remains uncertain (Williams et al., 1998a,b); it depends on whether there is free access of oceanic water to their underside (Nicholls, 1997). Circulation of meltwater under the ice shelves is complex. Any increase in melting as a result of global warming initially will introduce a fresher layer of water (see Jacobs et al., 1992; Jenkins et al., 1997; O'Farrell et al., 1997). This could restrict further melting, unless the current or tides are strong enough to flush the fresher layer away or entrain it into other water masses. Using a three-dimensional model adapted to the underside of the Amery Ice Shelf, Williams et al. (1998b) conclude that when the adjacent seas were warmed by 1°C, net melt increased more than three-fold. Grosfeld and Gerdes (1998) also show increased net melt with warming, but the situation stabilizes if there is less sea-ice formation. In considering the long-term response of the Antarctic ice sheet to global warming of 3°C, Warner and Budd (1998) show that it takes at least several centuries for the ice shelves to disappear in their model.
Other reports in this collection