Page 280 - Earth's Climate Past and Future

256 PART IV • Deglacial Climate Changes

Take the example of an oscillation that lasts for 2000
IN SUMMARY, apparently synchronous millennial years and is perfectly dated in one region but has a dating
oscillations of very large amplitude are recorded in uncertainty of +/– 1000 years in another (Figure 14–5).
Greenland ice and North Atlantic sediments. These In this uncertainty, the two signals could be varying syn-
oscillations indicate coupled changes in several of chronously or changing with totally opposed tempos or
the most important components of Earth’s climate varying with subtle leads and lags. At orbital time scales,
system: air and surface-ocean temperature, ice sheet leads and lags between climate signals provide critical
margins, and ice rafting.
clues to cause-and-effect relationships (Chapter 11). In
the same sense, scientists who examine millennial-scale
14-3 Detecting and Dating Oscillations changes need accurate information about leads and lags,
in Other Regions but dating uncertainties make it difficult and often
impossible to obtain the required accuracy.
The verification of millennial oscillations in both It is much easier to show that millennial-scale oscil-
Greenland ice and North Atlantic sediment set off a lations are present or absent in a given climate record,
vigorous search for similar oscillations in other regions. whatever their exact ages and correlations to other
Scientists who took part in this search (and are still records. In a few cases, the millennial oscillations match
doing so) faced two major problems: (1) Is the climatic the pattern of the changes in Greenland and the North
archive that is being examined capable of recording Atlantic so closely that little doubt can exist that we are
such brief oscillations? (2) How accurately can the oscil- looking at the same millennial oscillations. But even in
lations be dated? these cases, small leads or lags between the signals could
Resolution of climate signals varies from archive to exist.
archive and from region to region. The ideal archive is
one that allows resolution of annual changes and pro- 14-4 Oscillations Elsewhere in the
vides a record stretching well back into the last glacia- Northern Hemisphere
tion or preferably beyond. Unfortunately, few archives
combine these characteristics, but many archives that Other regions of the northern hemisphere also show
reach far back in time can resolve climate changes millennial-scale oscillations, including records from
lasting for tens or hundreds of years. This resolution is Western Europe (Figure 14–6). Short-term oscillations
sufficient for detecting millennial oscillations, even if appear in changes in the character of European soils.
not at full amplitude. These soils were richer in organic material during
The second (and more difficult) problem is deter- warmer episodes but almost free of organic carbon
mining whether or not the oscillations detected in other during colder oscillations. In addition, European pollen
archives correlate with those found in Greenland ice and records that showed orbital-scale changes from inter-
North Atlantic sediments. The C method used to date glacial forests to glacial tundra also show short-term fluc-
14
most continental records (after correcting to calendar tuations within glacial intervals from full tundra to mixed
years) has analytical uncertainties of several thousand grass steppe and forest vegetation. Although the correla-
18
years for glacial-age material. Because these dating tion of the European records with the Greenland δ O
14
errors are comparable in size to the length of the oscilla- fluctuations are not obvious, C dating of the younger
tions, it is more often than not impossible to determine fluctuations indicates that cold-adapted vegetation
how the observed oscillations actually correlate. occurred in Europe during times of colder air over

Climate signal 1
2000 years

? ? ? FIGURE 14-5 Uncertainties in
dating millennial oscillations Typical
dating uncertainties of 2000 years
make it difficult to determine whether
millennial oscillations in two separate
climate records are synchronous,
Climate signal 2 exactly opposite in timing, or offset
Dating uncertainty: +/– 1000 years
slightly in timing.