At least five meteorological records from scientific stations on the Antarctic Peninsula show marked decadal warming trends (King, 1994; Harangozo et al., 1997; King and Harangozo, 1998; Marshall and King, 1998; Skvarca et al., 1998). Although the periods and seasons of observations have been different, the records show consistent warming trends—as much as 0.07°C yr^-1, considerably higher than the global mean. This atmospheric warming has caused several notable changes in the ice cover of the Antarctic Peninsula, including changes in snow elevation (Morris and Mulvaney, 1995; Smith et al., 1999a) and the extent of surface snow cover (Fox and Cooper, 1998). The most important change has been the retreat of ice shelves. Seven ice shelves on the Antarctic Peninsula have shown significant, progressive, and continued retreat. On the West Coast, the Wordie Ice Shelf, Müller Ice Shelf, George VI Ice Shelf, and Wilkins Ice Shelf have retreated (Ward, 1995; Vaughan and Doake, 1996; Luchitta and Rosanova, 1998). On the east coast, the ice shelves that occupied Prince Gustav Channel and Larsen Inlet and Larsen Ice Shelf A have retreated (Rott et al., 1996; Vaughan and Doake, 1996; Skvarca et al., 1998). Following Mercer (1978), Vaughan and Doake (1996) show that the pattern of retreat could be explained by a southerly movement of the 0°C January isotherm, which appears to define a limit of viability for ice shelves. To date, about 10,000 km² of ice shelf have been lost.

Retreat of the ice shelves on the Antarctic Peninsula has attracted considerable media coverage, and environmental campaigns of some nongovernmental organizations (NGOs) have expressed concern that these events presage a more important collapse of the West Antarctic ice sheet. However, few direct impacts result from the loss of these ice shelves. The ice shelves were floating, so their melting does not directly add to sea level, and they usually are replaced by sea-ice cover, so overall albedo changes very little. Because the Antarctic Peninsula is steep and rugged, there is no evidence that removal of ice shelves will cause melting of the glaciers that fed them to accelerate and add to sea-level rise (Vaughan, 1993). Terrestrial ecosystems generally will be unaffected by ice-shelf retreat. Most polar benthic organisms, especially in the Antarctic, grow extremely slowly, so colonization of exposed seabed will be slow.

Because warming on the Antarctic Peninsula exceeds that over much of the rest of the continent (Jacka and Budd, 1998), migration of the limit of viability for ice shelves probably will not affect the Ronne-Filchner or Ross ice shelves in the next 100 years (Vaughan and Doake, 1996). A substantial increase in Antarctic summer temperatures, however, would threaten other large ice shelves beyond the Antarctic Peninsula. The real implications of ice-shelf retreat are that it highlights issues of risk perception and public understanding of climate change, rather than real physical impacts. These issues include questions such as: Does the Antarctic Peninsula warming result from a global effect, or simply from natural regional climate variations? Such problems of attribution of climate change will recur, especially if the costs of adaptation are to be spread across nations. The inherent nonlinearity of natural systems sometimes causes exaggerated local responses to small climate changes. In the present generation of climate models, the Antarctic Peninsula is not well resolved, so local effects on these scales cannot yet be reproduced. Confidence in predicting such changes in the natural ecosystems of the Antarctic Peninsula comes from observed changes on the terrestrial biota of southern ocean islands (Bergstrom and Chown, 1999).

Rapidly increasing temperatures have been directly implicated in colonization by introduced species and displacement of indigenous biota. Thus, the extent of higher terrestrial vegetation on the Antarctic Peninsula currently is increasing (Fowber and Lewis Smith, 1994). In the marine environment, the warming trend on the Antarctic Peninsula has been linked statistically to an associated change in sea ice for this region (Stammerjohn and Smith, 1996). For example, Smith et al. (1999b) show that rapid climate warming and associated reduction in sea ice are concurrent with a shift in the population size and distribution of penguin species in the Antarctic Peninsula region.

IPCC Homepage