OVER WHAT TIME AND SPATIAL SCALES ARE NORTHERN AND SOUTHERN HEMISPHERE CLIMATES LINKED?

The question of whether climate change is synchronous between the Northern and Southern Hemispheres has been a primary motivation for research in Quaternary paleoclimatology. Much of the debate focuses on the lag or lead relationships between specific climate events in the Northern Hemisphere and their possible analogues in the Southern Hemisphere; for example, the “Younger Dryas” and the “Antarctic Cold Reversal”. In determining such phase relationships, one must also account for the possibility that the events are unrelated; that is, we must consider the extent to which similar-appearing climate variations in different parts of the globe may be due to independent climate variability. The magnitude of this regional climate “noise” is not well known for century to millennial timescales.

Quantitative analysis of climate variability on millennial timescales has centered on evidence from the best-dated Antarctic and Greenland climate records. Among the findings from comparing Byrd and GISP2 is that the relationship commonly referred to as a “seesaw” is significant, and clearly reflects a global mechanism linking the hemispheres. However, this appears to be robust only for timescales of millennial and longer. At present, the evidence is scant that the influence of Greenland on Antarctic climate (or vice versa) exceeds the regional, independent climate variability for any but the largest climate events of the last glacial period (i.e. interstadials 8 and 12 in Greenland). This implies that the regional noise on millennial timescales tends to overwhelm the global signal for all but the largest events.

Newly cross-dated Antarctic ice core records – from Taylor Dome, Siple Dome, Dome C, and Law Dome – provide a means to further examine the magnitude of regional climate noise by comparing records within the same hemisphere. Within Antarctica, as for the Greenland/Antarctica comparison, only the largest millennial-scale climate changes are expressed broadly, while smaller magnitude, higher frequency variations show more complex spatial patterns. At least one large, rapid climate change event (recorded at Siple Dome) does not occur in other ice cores. These results suggest that for century scale and smaller magnitude millennial-scale events, regional noise tends to overwhelm any possible global signal. This may apply, for example, to the Little Ice Age.