muted Younger Dryas-like event. Changes locally might have been as large as 10°C in the North Atlantic, with changes of about 2°C extending well into Europe (Renssen et al., 2001). High-resolution pollen studies show substantial and rapid vegetation response to the event in central Europe, with early biological changes lagging climate by less than 20 years (Tinner and Lotter, 2001). Because so many Holocene climate records are available and the cause of the event is rather clear, it provides an opportunity for an especially well-documented test case of model sensitivity. The event is also important because it punctuated a time when temperatures were similar to or even slightly above more recent levels, demonstrating that warmth is no guarantee of climate stability.
A less well-understood hydrologic event from wet to dry conditions, occurring roughly 5,000 years ago, also took place during a warm period. This event is not as well documented and suffers from less than ideal temporal resolution of available records. It is most evident in African records (Gasse and Van Campo, l994; Gasse, 2000), the North Atlantic (Duplessy et al., 1992; Bond et al., 1997; deMenocal et al., 2000b; Jennings et al., in press), the Middle East (Cullen et al., 2000), and Eurasia (Enzel et al., 1999; Morrill et al., in review). Four mechanisms have been proposed to explain the event, all of which could have contributed. First, it might have been associated with a cooling in the North Atlantic, perhaps related to a slow-down in thermohaline circulation (Street-Perrott and Perrott, 1990; Gasse and van Campo, 1994; Kutzbach and Liu, 1997; deMenocal et al., 2000b). Second, it might be related to a subtle (and variable) ca. 1500-year oscillation in Atlantic variability (Bond et al., 1997) of poorly understood origin, but almost certainly involving ocean processes (Alley et al., 1999), and extending beyond the North Atlantic regions; recent work (Jennings et al., in press and Morrill et al., in review) indicated that the spatio-temporal dimensions of this variability could be complex. Third, an abrupt shift in the El Niño-Southern Oscillation (ENSO) system might have led to a more widespread event at about the time in question (Morrill et al., in review). Fourth, atmosphere-vegetation feedbacks triggered by subtle changes in the earth’s orbit might have triggered the event (Claussen et al., 1999) or at least amplified it (Kutzbach et al., 1996; Ganopolski et al., 1998; Braconnot et al., 1999).
Increasing attention is also being focused on the possibility that the ENSO system has changed its pattern of variability, perhaps rapidly. The best-documented shift in the frequency of ENSO variability occurred in 1976 (Trenberth, 1990), and it was probably one of several shifts in frequency to occur over the