Overall, the Eemian is neither stable and boring, nor extraordinarily variable. Most regions for which good data are available record significant and important fluctuations, some of which were abrupt, but with reduced variability compared to during the cooling into and warming out of ice ages. Attention is especially focused on drought conditions in low latitudes rather than temperature in high latitudes.
The relevance of abrupt climate change of the ice age to the modern warm climate or future warmer climates is unclear. However, although glacial and deglacial rapid shifts in temperature were often larger than those of the Holocene (the last roughly 10,000 years), Holocene events were also important with respect to societally relevant climate change (Overpeck, 1996; Overpeck and Webb, 2000). For example, there were large rapid shifts in precipitation (droughts and floods) and in the size and frequency of hurricanes, typhoons, and El Niño/La Niña events. If they recurred, these kinds of changes would have large effects on society. It is not surprising that many past examples of societal collapse involved rapid climate change to some degree (Weiss and Bradley, 2001; deMenocal, 2001a).
This section summarizes some of the compelling evidence of rapid change during the Holocene. When we view the available evidence of abrupt climate change in the Holocene, it is apparent that their temporal and spatial characteristics are poorly understood. In addition, the causes of abrupt change are not well constrained. The lack of a mechanistic understanding regarding past abrupt climatic change is one of the unsettling aspects of the state of the art.
Among the most widely investigated rapid climate events of the early to middle Holocene are two that took place about 8,200 and 4,000-5,000 years ago. The former event (Figures 2.3 and 2.4) has been recognized in Greenland ice, the North Atlantic, North America, Europe, Africa, and elsewhere and has been tied to a temporary reduction in the North Atlantic thermohaline circulation generated by late-stage melting of the North American ice sheets that released a large, abrupt meltwater flood from ice-marginal lakes through Hudson Strait to the North Atlantic (Bjorck et al., 1996; Alley et al., 1997; Barber et al., 1999; Gasse, 2000; Gasse and van Campo, 1994; Kneller and Peteet, 1999, von Grafenstein et al., 1999; Yu and Eicher, 1998; cf. Stager and Mayewski, 1997). If the mechanism for this event has been identified correctly, the event was a final deglacial, or