haps the most studied rapid temperature shift of the Holocene is the change that began in the latter half of the nineteenth century and ended the so-called Little Ice Age. The shift and later state of substantial global warming were unprecedented in the context of the last 500 years and might be due to a combination of natural (such as solar and volcanic) and human-induced (such as trace-gas) forcing (Overpeck et al., 1997; Jones et al., 1998; Mann et al., 1998, 1999, 2000; Huang et al., 2000; Crowley, 2000; Briffa et al., 2001; Intergovernmental Panel on Climate Change, 2001a).
In contrast with the abrupt late nineteenth to early twentieth century warming, timing of the onset of the Little Ice Age is difficult to establish in that the change manifests itself as a period of slow Northern Hemisphere cooling beginning at or before ca 1000 (Mann et al., 1999; Crowley, 2000; Crowley and Lowery, 2000; Briffa et al., 2001) with several sustained cooler intervals thereafter (for example, the seventeenth century and early nineteenth century).
There are insufficient paleoclimate records to allow complete reconstruction of the last 1,000 years of change in the Southern Hemisphere, and uncertainty remains on the amplitude of Northern Hemisphere change in this interval (e.g., Briffa et al., 2001; Huang et al., 2000). There is still debate as to whether the “Medieval Warm Period” was more than a Northern Hemisphere warm event (Mann et al., 1999; Crowley, 2000; Crowley and Lowery, 2000; Briffa et al., 2001; Broecker, 2001). Moreover, evidence is scarce outside the North Atlantic-European sector (Jennings and Weiner, 1996; Keigwin, 1996; Broecker, 2001) for medieval temperatures that were close to mean twentieth century levels. Additional annually resolved records for the last 2,000 years are needed to answer such fundamental questions.
The existing temperature records, as described above, make it clear that natural variability alone can generate regional to hemispheric temperature anomalies that are sufficient to affect many aspects of human activity. However, the record of hydrologic change over the last 2,000 years suggests even larger effects: there is ample evidence that decadal, even century-scale, drought can occur with little or no warning.
A synthesis of US drought variability over the last 2,000 years (Woodhouse and Overpeck, 1998) used records from a diverse array of proxy sources (cf. Cronin et al., 2000; Stahle et al., 1998). From this synthesis, it was concluded that multi-year droughts similar to the 1930s Dust