Antarctic Sea Ice Extent Variability and Its Global Connectivity*

This study statistically evaluates the relationship between Antarctic sea ice extent and global climate variability. Temporal cross correlations between detrended Antarctic sea ice edge (SIE) anomaly and various climate indices are calculated. For the sea surface temperature (SST) in the eastern equatorial Pacific and tropical Indian Ocean, as well as the tropical Pacific precipitation, a coherent propagating pattern is clearly evident in all correlations with the spatially averaged (over 12° longitude) detrended SIE anomalies (〈SIE*〉). Correlations with ENSO indices imply that up to 34% of the variance in 〈SIE*〉 is linearly related to ENSO. The 〈SIE*〉 has even higher correlations with the tropical Pacific precipitation and SST in the tropical Indian Ocean. In addition, correlation of 〈SIE*〉 with global surface temperature produces four characteristic correlation patterns: 1) an ENSO-like pattern in the Tropics with strong correlations in the Indian Ocean and North America (r > 0.6); 2) a teleconnection pattern between the eastern Pacific region of the Antarctic and western–central tropical Pacific; 3) an Antarctic dipole across the Drake Passage; and 4) meridional banding structures in the central Pacific and Atlantic expending from polar regions to the Tropics, even to the Northern Hemisphere. The SIE anomalies in the Amundsen Sea, Bellingshausen Sea, and Weddell Gyre of the Antarctic polar ocean sectors show the strongest polar links to extrapolar climate. Linear correlations between 〈SIE*〉 in those regions and global climate parameters pass a local significance test at the 95% confidence level. The field significance, designed to account for spatial coherence in the surface temperature, is evaluated using quasiperiodic colored noise that is more appropriate than white noise. The fraction of the globe displaying locally significant correlations (at the 95% confidence level) between 〈SIE*〉 and global temperature is significantly larger, at the 99.5% confidence level, than the fraction expected given quasiperiodic colored noise in place of the 〈SIE*〉. Based on EOF analysis and multiplicity theory, the four teleconnection patterns the authors found are the ones reflecting correlations most likely to be physically meaningful.