Ice core methodology - Ice Ages and Interglacials: Measurements, Interpretation, and Models - page 125

insignificant correlated climate records may well be linked in a causal manner.
Caution is required, however: short records of processes that are even slightly
reddish in spectral character can easily lead to unwarranted, and incorrect,
inferences if simple stochastic superposition is confused with deterministic
causes. Sometimes there is no alternative to uncertainty except to await the arrival
of more and better data'' [emphasis added].

In a later paper, Wunsch (2003) further discussed the problem of relating
climate curves, particularly in relating the timing of events at Antarctica to those
at Greenland. He found that over long time spans (

10,000 years) the evidence is
strong that Greenland changes lag Antarctic changes by perhaps 1,000 years.
However, over shorter time intervals no correlation of Antarctica with Greenland
can be made. Unfortunately, this paper is dicult to read.
A problem with much of the ice core data is that there is so much cross-
fertilization between time scales from various sources that it is often dicult to
determine the solid ground (if any) time scales rest upon.
The first (and perhaps most influential) means of relating curves was the
development of the so-called SPECMAP in which ocean sediment data were dated
by comparison with the astronomical theory (Imbrie et al., 1984). Ocean sediment
data can be dated back to about 30,000 years by radiocarbon measurements. For
a number of years it was common to assume that the final peak of the previous
interglacial period occurred 100 kybp , and this was used as a marker. This was
later raised to 125 kybp . However, there is evidence that warming started as early
as 140 kybp (see Figure 6.3 and discussions in Sections 6.1.4, 6.2, and 6.5.2).
The magnetic polarity change that is believed to have occurred some
780,000 ybp can sometimes be useful to relate curves. But ocean sediment data
date back to 3 mybp and comparison with the astronomical theory (''tuning'') has
been the most widely applied method for dating over this long time period.
However, much of the correspondence between the theory and the data seems to
lie in the eyes of the beholders.

>

3.3 PROCESSING ICE CORE DATA

3.3.1 Temperature estimates from ice cores
3.3.1.1 Correlation of d 18 O with temperature based on recent surface data
As we pointed out in Section 3.1, the temperature at which ice was deposited on
ice sheets in the past is believed to be in some way proportional to d 18 O. The
isotope ratio

0 ! 1

18 O
16 O

@ A

Sample

d 18 O

!

1 1,000

18 O
16 O

Reference