IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group I: The Physical Science Basis Are There Long-Term Modes of Climate Variability Identified During the Holocene that Could Be Involved in the Observed Current Warming?

An increasing number of Holocene proxy records are of sufficiently high resolution to describe the climate variability on centennial to millennial time scales, and to identify possible natural quasi-periodic modes of climate variability at these time scales (Haug et al., 2001; Gupta et al., 2003). Although earlier studies suggested that Holocene millennial variability could display similar frequency characteristics as the glacial variability in the North Atlantic (Bond et al., 1997), this assumption is being increasingly questioned (Risebrobakken et al., 2003; Schulz et al., 2004). In many records, there is no apparent consistent pacing at specific centennial to millennial frequencies through the Holocene period, but rather shifts between different frequencies (Moros et al., 2006). The suggested synchroneity of tropical and North Atlantic centennial to millennial variability (de Menocal et al., 2000; Mayewski et al., 2004; Y.J. Wang et al., 2005) is not common to the SH (Masson et al., 2000; Holmgren et al., 2003), suggesting that millennial scale variability cannot account for the observed 20th-century warming trend. Based on the correlation between changes in cosmogenic isotopes (10Be or 14C) – related to solar activity changes – and climate proxy records, some authors argue that solar activity may be a driver for centennial to millennial variability (Karlén and Kuylenstierna, 1996; Bond et al., 2001; Fleitmann et al., 2003; Y.J. Wang et al., 2005). The possible importance of (forced or unforced) modes of variability within the climate system, for instance related to the deep ocean circulation, have also been highlighted (Bianchi and McCave, 1999; Duplessy et al., 2001; Marchal et al., 2002; Oppo et al., 2003). The current lack of consistency between various data sets makes it difficult, based on current knowledge, to attribute the millennial time scale large-scale climate variations to external forcings (solar activity, episodes of intense volcanism), or to variability internal to the climate system.

6.5.2 Abrupt Climate Change During the Current Interglacial What Do Abrupt Changes in Oceanic and Atmospheric Circulation at Mid- and High-Latitudes Show?

An abrupt cooling of 2°C to 6°C identified as a prominent feature of Greenland ice cores at 8.2 ka (Alley et al., 1997; Alley and Agustsdottir, 2005) is documented in Europe and North America by high-resolution continental proxy records (Klitgaard-Kristensen et al., 1998; von Grafenstein et al., 1998; Barber et al., 1999; Nesje et al., 2000; Rohling and Palike, 2005). A large decrease in atmospheric CH4 concentrations (several tens of parts per billion; Spahni et al., 2003) reveals the widespread signature of the abrupt ‘8.2 ka event’ associated with large-scale atmospheric circulation change recorded from the Arctic to the tropics with associated dry episodes (Hughen et al., 1996; Stager and Mayewski, 1997; Haug et al., 2001; Fleitmann et al., 2003; Rohling and Palike, 2005). The 8.2 ka event is interpreted as resulting from a brief reorganisation of the North Atlantic MOC (Bianchi and McCave, 1999; Risebrobakken et al., 2003; McManus et al., 2004), however, without a clear signature identified in deep water formation records. Significant volumes of freshwater were released in the North Atlantic and Arctic at the beginning of the Holocene by the decay of the residual continental ice (Nesje et al., 2004). A likely cause for the 8.2 ka event is an outburst flood during which pro-glacial Lake Agassiz drained about 1014 m3 of freshwater into Hudson Bay extremely rapidly (possibly 5 Sv over 0.5 year; Clarke et al., 2004). Climate models have been used to test this hypothesis and assess the vulnerability of the ocean and atmospheric circulation to different amounts of freshwater release (see Alley and Agustsdottir, 2005 for a review; Section Ensemble simulations conducted with EMICs (Renssen et al., 2002; Bauer et al., 2004) and coupled ocean-atmosphere GCMs (Alley and Agustsdottir, 2005; LeGrande et al., 2006) with different boundary conditions and freshwater forcings show that climate models are capable of simulating the broad features of the observed 8.2 ka event (including shifts in the ITCZ).

The end of the first half of the Holocene – between about 5 and 4 ka – was punctuated by rapid events at various latitudes, such as an abrupt increase in NH sea ice cover (Jennings et al., 2001); a decrease in Greenland deuterium excess, reflecting a change in the hydrological cycle (Masson-Delmotte et al., 2005b); abrupt cooling events in European climate (Seppa and Birks, 2001; Lauritzen, 2003); widespread North American drought for centuries (Booth et al., 2005); and changes in South American climate (Marchant and Hooghiemstra, 2004). The processes behind these observed abrupt shifts are not well understood. As these particular events took place at the end of a local warm period caused by orbital forcing (see Box 6.1 and Section 6.5.1), these observations suggest that under gradual climate forcings (e.g., orbital) the climate system can change abruptly.