J.C. King
British Antarctic Survey, Natural Environment Research Council Cambridge CB3 0ET, UK
Email: jcki@bas.ac.uk

It is fortunate that the Antarctic Peninsula, which is arguably the most climatologically interesting part of Antarctica, possesses some of the longest and most complete climate records from the continent. Surface meteorological records are available from a number of stations from the late 1940s onward, with upper-air measurements commencing in 1956. Analysis of these records shows striking differences between climate variations in the Peninsula and those elsewhere in Antarctica. Interannual variations in surface air temperature at Antarctic Peninsula stations are considerably larger than those seen elsewhere in the continent. While records from East Antarctic stations exhibit a mixture of small warming and cooling trends, Peninsula records show significant warming in all seasons. At Faraday/Vernadsky, summer temperatures have warmed at 0.03 degC/year over 1951-2000, which is not significantly different from the southern hemisphere average. However, winter trends at this station (+0.11 degC/year) are amongst the largest in the southern hemisphere. Mid-tropospheric temperatures have warmed at 0.03 degC/year, with little variation seen between seasons. Temperature trends at these levels are not significantly different from southern hemisphere average values.

Interannual temperature variations on the Peninsula have been shown to be associated with changing atmospheric circulation patterns and with sea ice extent to the west of the Peninsula. It is therefore natural to speculate that the Peninsula temperature trends might be caused by long-term trends in circulation or sea ice. Unfortunately, reliable atmospheric analyses for the Antarctic exist only back to the early 1970s and no significant trends are seen in this rather short series. Mid-tropospheric winds show no significant changes, supporting the view that there have been no long-term circulation changes in this region. Satellite observations of sea ice, which began in 1973, show some evidence for a decline in sea ice extent to the west of the Peninsula. Feedbacks involving sea ice have the potential to amplify a small warming. The seasonal pattern of warming trends and their vertical variation provide evidence for the importance of air-sea-ice interactions in controlling the climate of this region.

The observed pattern of Peninsula warming is not inconsistent with a regional enhancement of anthropogenic "greenhouse" warming as a result of local air-sea-ice interaction processes. However, runs of General Circulation Models forced with observed increases in greenhouse gases do not reproduce the strong warming observed in the Peninsula. While this might be viewed as evidence for a contrary view, it must be recognised that the spatial resolution of current models and the way in which important physical processes are represented may not properly capture the complex interactions occurring in the Peninsula region. Unequivocal attribution of Peninsula warming awaits the development of improved models.