IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group I: The Physical Science Basis What Is Known About the Mechanism of these Abrupt Changes?

There is good evidence now from sediment data for a link between these glacial-age abrupt changes in surface climate and ocean circulation changes (Clark et al., 2002). Proxy data show that the South Atlantic cooled when the north warmed (with a possible lag), and vice versa (Voelker, 2002), a seesaw of NH and SH temperatures that indicates an ocean heat transport change (Crowley, 1992; Stocker and Johnsen 2003). During D-O warming, salinity in the Irminger Sea increased strongly (Elliot et al., 1998; van Kreveld et al., 2000) and northward flow of temperate waters increased in the Nordic Seas (Dokken and Jansen, 1999), indicative of saline Atlantic waters advancing northward. Abrupt changes in deep water properties of the Atlantic have been documented from proxy data (e.g., 13C, 231Pa/230Th), which reconstruct the ventilation of the deep water masses and changes in the overturning rate and flow speed of the deep waters (Vidal et al., 1998; Dokken and Jansen, 1999; McManus et al., 2004; Gherardi et al., 2005). Despite this evidence, many features of the abrupt changes are still not well constrained due to a lack of precise temporal control of the sequencing and phasing of events between the surface, the deep ocean and ice sheets.

Heinrich events are thought to have been caused by ice sheet instability (MacAyeal, 1993). Iceberg discharge would have provided a large freshwater forcing to the Atlantic, which can be estimated from changes in the abundance of the isotope 18O. These yield a volume of freshwater addition typically corresponding to a few (up to 15) metres of global sea level rise occurring over several centuries (250–750 years), that is, a flux of the order of 0.1 Sv (Hemming, 2004). For Heinrich event 4, Roche et al. (2004) have constrained the freshwater amount to 2 ±1 m of sea level equivalent provided by the Laurentide Ice Sheet, and the duration of the event to 250 ±150 years. Volume and timing of freshwater release is still controversial, however.

Freshwater influx is the likely cause for the cold events at the end of the last ice age (i.e., the Younger Dryas and the 8.2 ka event). Rather than sliding ice, it is the inflow of melt water from melting ice due to the climatic warming at this time that could have interfered with the MOC and heat transport in the Atlantic – a discharge into the Arctic Ocean of the order 0.1 Sv may have triggered the Younger Dryas (Tarasov and Peltier, 2005), while the 8.2 ka event was probably linked to one or more floods equal to 11 to 42 cm of sea level rise within a few years (Clarke et al., 2004; see Section 6.5.2). This is an important difference relative to the D-O events, for which no large forcing of the ocean is known; model simulations suggest that a small forcing may be sufficient if the ocean circulation is close to a threshold (Ganopolski and Rahmstorf, 2001). The exact cause and nature of these ocean circulation changes, however, are not universally agreed. Some authors have argued that some of the abrupt climate shifts discussed could have been triggered from the tropics (e.g., Clement and Cane, 1999), but a more specific and quantitative explanation for D-O events building on this idea is yet to emerge.

Atmospheric CO2 changes during the glacial antarctic warm events, linked to changes in NADW (Knutti et al., 2004), were small (less than 25 ppm; Figure 6.7). A relatively small positive feedback between atmospheric CO2 and changes in the rate of NADW formation is found in palaeoclimate and global warming simulations (Joos et al., 1999; Marchal et al., 1999). Thus, palaeodata and available model simulations agree that possible future changes in the NADW formation rate would have only modest effects on atmospheric CO2. This finding does not, however, preclude the possibility that circulation changes in other ocean regions, in particular in the Southern Ocean, could have a larger impact on atmospheric CO2 (Greenblatt and Sarmiento, 2004).