Russell Drysdale et al.

DP0773700

The Earth’s climate has fluctuated on a regular basis over the last million years. The most significant fluctuations are driven by a combination of ‘external’ and ‘internal’ forcings. Variations in the Earth’s orbit, which control the distribution of solar radiation across the Earth’s surface, are the most important set of external forcings and are largely responsible for glacial-interglacial [GIG] cycles (Hays et al. 1976). The response of the climate system to orbital forcing is modulated at millennial time scales by internal feedbacks involving ocean circulation and ice-sheet dynamics (Clark et al. 2002). One of the fundamental challenges in palaeoclimatology is to discover how orbitally forced climate changes are transmitted to different parts of the Earth via the oceanatmosphere system and if this occurs synchronously (Lynch-Stieglitz 2004). The crucial evidence required to meet this challenge lies in the palaeoclimate record.

Millennial-scale climate variations were a persistent feature of virtually the entire last glacial period (115,000–10,000 yr, i.e. 115-10 ka) (Dansgaard et al. 1993). In the North Atlantic region, these occurred as cycles of rapid warming followed by steady cooling (‘Dansgaard-Oeschger events’). Comparisons of ice cores from both polar regions (Blunier et al. 1998, 2001; Caillon et al. 2003) show that these variations were anti-phase between the hemispheres: when cooling was occurring in the North Atlantic, the Southern Ocean was warming, and vice versa. Broecker (1998) proposed that this anti-phasing was due to a ‘bipolar seesaw’ effect on deep-ocean ventilation. At the end of the last glacial period (i.e. Termination I: 20-10 ka), the Southern Ocean warmed before the North Atlantic (Lamy et al. 2004; Turney et al. 2003; Williams et al. 2005; Vandergoes et al. 2005). Modelling suggests that this warming may have ‘kick-started’ North Atlantic thermohaline circulation, leading to a strengthening of the warm North Atlantic Current and the collapse of Northern Hemisphere ice sheets (Knorr & Lohmann 2003). Anti-phasing occurred until the start of the Holocene (Blunier et al. 1998). A similar scenario has been proposed for the previous deglaciation (Termination II: ca. 140 – 125 ka) (Knorr & Lohmann 2003).

The above research suggests the bipolar seesaw, or some other anti-phasing mechanism, operated throughout almost the entire last glacial period, including Termination I. It may also be instrumental in governing which part of the Earth warms first during glacial terminations. While the mechanisms driving ocean circulation changes are hotly debated (Wunsch 2002; Kerr 2005; Shiermier 2006), explanatory models for anti-phasing must be tested. It is not currently known whether a bipolar seesaw model can be invoked for previous glacial periods because we lack the appropriate palaeoclimate data. Crucial ice-based comparisons, which require coeval records from both poles, are unlikely to extend beyond 122 ka due to deformation of the Greenland ice sheet (NGRIP Project Members 2004). To evaluate the long-term persistence of anti-phasing, new proxy evidence from alternative sources of palaeoclimate data is needed. These sources must meet two strict criteria: sensitivity to changes in North Atlantic and Southern Ocean circulation, and an ability to be dated by radiometric techniques so that precise chronologies can be developed for accurate crosshemispheric comparisons. Speleothems (stalagmites and flowstones from limestone caves) present an excellent opportunity: they can be dated precisely and accurately (Edwards et al. 1986; Hellstrom 2003), they can grow continuously for long periods (104-105 years) and their growth and geochemical properties are responsive to climate change (McDermott 2004).

AWI contribution: Earth system modelling; modelling isotopes; statistical analyses