bined signal of temperature and precipitation—along with source area changes—but each of these signals may be dominant in different settings. In tropical and dryland regions, rainfall amount is usually the main determinant of the δ18O content of cave carbonates (Bar-Matthews et al. 2003). Thus, oxygen isotope records obtained from caves in southern China (Wang et al. 2005b) (Figure 5-8) and Oman (Fleitmann et al. 2003) suggest that strong climatological changes have occurred in Asian monsoon intensity over the last several thousand years. Episodic submillennial variations within these Holocene records generally match the North Atlantic ice-rafted debris pattern of Bond et al. (2001). The correlation of speleothem records from cave sites associated with the Asian monsoons to a marine record of ice advance and retreat in the North Atlantic provides a suggestion of hemisphere-wide century-scale climate changes resulting in different local manifestations.
In some cool and wet regions, the cave temperature signal may be dominant in controlling the δ18O and other measurements, and this is especially valuable because cave temperature is stable throughout the year and represents the mean annual temperature of the outside environment. Modern speleothem properties have been used to calibrate cave sequences from northern Scandinavia in terms of Holocene temperature variability (Lauritzen and Lundberg 1999). In addition, some speleothems contain annual bands much like lake varves or tree growth rings, and these have allowed very high resolution measurements of cave isotopic changes during the last 2,000 years. Proctor et al. (2000, 2002) used one such record from northern Scotland to reconstruct annual-to-decadal climatic changes in the North Atlantic during the last three millennia.