During the last glacial cycle, Earth's climate was punctuated by two series of abrupt changes of millennial duration. One of these, known as Dansgaard-Oeschger cycles, contains oscillations of 1000 to about 5000-yr duration that are best developed in the North Atlantic in Marine Isotope Stage 3 (Bond et al., 1993; Broecker, 1994). As revealed in records from Greenland ice and North Atlantic sediments, each D-O cycle began with a remarkably abrupt warming of several degrees, probably occurring within decades, if not several years, followed by increasing amounts of ice-raited debris (IRD) and a gradual cooling to a full stadial (Figs 24.1 & 24.2).
The second series, known as Heinrich events, consists of cycles that occurred every 7000 to 10,000 yr. As best documented in records from the glacial North Atlantic (Fig. 24.2), Heinrich
1 Sadly, Gerard Bond passed away during the production of this volume. We are extremely grateful for his contribution and proud to be able to publish one of the last writings of a great scientist who has contributed so much to the field.
events occurred suddenly at the end of cooling ramps containing D-O cycles, and were accompanied by marked increases in ice-rafted debris (IRD) and cold sea-surface and land temperatures that exceeded those of D-O cycles (Heinrich, 1988; Bond et al., 1992; Hemming, 2004). Each Heinrich event was followed by an exceptionally abrupt warming that raised sea-surface temperatures by several degrees to nearly Holocene values.
Both the D-O cycles and Heinrich events are thought to have been part of nearly global reorganizations of ocean-atmosphere climates. Within the limits of dating error, apparent correlatives of both series have been found in marine and terrestrial records from high northern latitudes, through the tropics, and into at least subtropical southern latitudes (Voelker, 2002). In deep-sea cores from the southeast Atlantic Ocean, Kanfousch et al. (2000) found apparent correlatives of Heinrich events in records of IRD discharge from Antarctica. Distinct millennial temperature variations also occur in Antarctic ice but whether these lead, lag or are in phase with Greenland's D-O cycles is widely debated owing to uncertainties in ice-core age models (e.g., Steig & Alley, 2002, Roe & Steig, 2004, Shackleton et al., 2004).
Was this article helpful?