temperatures and water vapor obtained with tropical radiosondes for the past couple of decades have been documented by Flohn et al. (1992), although the drift toward less negative d18O ratios appears to have begun earlier in the ice-core record than in the instrumental climate record.
Changes in interdecadal variability are summarized in Table 2 (see indices 1-4 and 14-15). There is a suggestion that decade-to-decade changes in d18O ratios have increased in amplitude at most of these sites. In particular, the calculated interdecadal d18O variability at Quelccaya has increased sharply in the last 100 years. This increase is evident even though the greater number of samples in the upper part of the record permitted better averages to be created for the recent decades than for earlier periods. A similar signal is apparent on the Dunde ice cap as well. These changes, however, may reflect a more localized temperature signal, changes in moisture sources for snowfall on the ice caps, or some other unknown cause unrelated to climate. It should also be noted that these two sites are located at very high elevations, and thus differ from most of the other
paleoclimate indices, which are derived from locations at much lower elevations.
Decadal means of winter (December-to-January) and summer (June-to-August) mean temperature for central England (Manley, 1974, and subsequent updates) from 1700 to present are shown in Figure 11. The recent decades do not appear to be exceptional in comparison with the total record. The variability curves (not shown) indicate little overall change, with a tendency for the summer and winter seasons to have opposite changes in interdecadal variability (when one increases, the other tends to decrease). Jones and Bradley (1992, their Figure 13.1) have compared the central England temperature record with 12 other long-period station records in Eurasia and North America. Although there are obvious differences among the various climatic series, the twentieth century does not stand out as being a period of higher decadal-scale temperature variability. Most records (although not all) do exhibit the general warming trend of the past two centuries that is associated with the rebound from the Little Ice Age (see Bradley and Jones, 1992).
The purpose of this study was to evaluate whether the climate in recent decades was ''changing," both with regard