High quality satellite-based measurements of Antarctic sea ice extent available since the late 1970s clearly show there is considerable variability on interannual time scales in many areas as well as contrasting trends in different regions. At present, however, the causes of this variability are far from clear but the atmospheric circulation with its large interannual variability is a likely important factor in at least some regions. To date, systematic studies of ice extent-atmospheric circulation relationships have obtained mixed results. For example, Godfred-Spenning and Simmonds (1996) found mostly weak correlations between ice extent and extratropical cyclone density in all seasons of the year. Various Arctic and Antarctic studies, e.g. Kimura and Wakatsuchi (1999), Stammerjohn et al. (2003) Harangozo (2004), indicate useful insights can be gained by relating daily winds to ice extent, and suggest the north-south, component of the surface wind may be crucial to explaining the observed interannual variations in the winter maximum extent.

This paper will report results from a systematic study that is seeking to explain interannual variations in the Antarctic winter maximum ice extent over 22 years. It uses a novel approach in which daily 10 m winds at the ice edge at a given longitude are aggregated for each winter season and then correlated with the final winter ice extent (IE). Winds are obtained from general circulation model (GCM) reanalyses. Early results indicate that southerly winds (off ice) systematically increase as IE increases in many regions. An exception is the Ross Sea where strongest correlations are found with northerly winds. Further analysis reveals the meridional circulation modulates IE in two ways: via southerly winds and advance but also through northerly winds inducing retreat. In the RS the impact of strong and frequent northerly winds on IE is particularly marked giving rise to frequent retreats in some winters of low IE.