|
B. Geerts |
6/'98 |
|
In late summer (February-March) the areal extent of sea ice around Antarctica is about 4 106 km2. Some coastlines, in particular south of Africa (0-70ºE), are entirely ice-free in some years. The ice-covered area increases in autumn at times when passing weather systems drive ice away from land, or induce low-level divergence over open ocean. This causes cracks of open water (i.e. 'leads' or polinyas). Ice formation is most rapid in these leads, because heat and vapour fluxes into the atmosphere are largest there, causing the surface water to cool. By late winter (August-September), the typical extent of sea ice around Antarctica is about 19 106 km2, i.e. more than twice the area of Australia, and more than the area of Arctic sea ice. Sea ice commonly occurs as far north as 60°S in the South Pacific and 55°S in the South Atlantic, as can be seen in the picture on the right. The ice coverage has been documented since 1973 by means of passive microwave sensors on polar orbiting satellite. In the microwave part of the electromagnetic spectrum (at 16 mm wavelength, to be exact), ice has a high emissivity (0.8 to 0.97) whereas the emissivity of liquid water is much lower (0.44). So ice fields can be easily detected, and the fraction of ice coverage can be estimated. Passive microwave data since 1973 have shown considerable interannual variability. In 1974 a large polinya (about the size of the UK) was detected near the Weddell Sea, a vast bay between 60-40ºW which normally remains largely ice-covered year-round. Questions arose with regard to climate change and stability of the Antarctic ice shield. In late 1976, when scientific interest in the Weddell Polinya had been mobilized, it disappeared and has not re-appeared since. This experience showed that the Weddell Sea ice slowly rotates clockwise, with a period of about two years (1), the same period for the Antarctic Circumpolar Circulation. |
|
|
|
Fig 1 (click on picture for more detail) |
The passive microwave data show very little long-term trend. From 1978 to 1996, the average ice cover around Antarctica showed almost no trend (a slight increase 1.3% per decade), however ice decreased by 2.9% per decade on average over the arctic seas (1). This change is inconsistent with the change in surface temperature: 15 coastal stations around Antarctica recorded an average warming of 0.028 degrees annually during 1959-88, i.e. three times the global average. At the same time, there was no discernible warming in the Arctic.
The interannual variation of sea ice coverage is much larger than any trend (Fig 2). It is very difficult to assess ice cover or ice thickness changes over a longer period, before 1978. Whalers' logbooks dating from as early as the early 20th century provide anecdotal evidence. At southern whaling stations the ice cover has decreased by about 25% between 1950 and 1975, i.e. -10% per decade (2).
Fig 2 Variations of sea-ice area since 1972, showing a recent reduction in the amplitude of the fluctuations (4).
General circulation models predict that global warming over the next few decades will occur mainly in the polar regions. A knowledge of sea ice extent is important because ice cover has a strong albedo warming feedback. Also, upon freezing the separation of nearly pure water (which becomes ice) leaves behind a relatively heavy brine, which sinks to form Bottom Water in the great oceans. In other words, sea ice relates to the deep-ocean circulation, which in turn relates to global climate.
Therefore the sea ice extent is carefully monitored. Satellite data do not indicate ice thickness, so very little large-scale information is available about this. This is currently being investigated around Antarctica (3), by direct measurement along the tracks of ships traversing the sea ice.
References
