Signals in the Noise - The Hockey Stick and the Climate Wars - page 23

this method. Fortunately, other types of proxy data can help fill the gaps.
Corals, for example, are found in oceanic environments, generally in tropical regions, making
them complementary to tree rings in the regions of the globe they sample. Corals assimilate oxygen
atoms from the ocean in the form of the carbonate ion, CO 3 − 2 , as their layers of calcium carbonate
(CaCO 3 ) skeletons are formed. The ratios of the two main stable isotopes of oxygen (the common,
lighter 16 O versus the less common, heavier 18 O) in the corals' annual growth bands, as it turns out,
depend on the temperature and salinity (in turn influenced primarily by rainfall) of the near surface
ocean waters they grow in. Other chemical measurements from corals such as the ratio of the elements
strontium and calcium in the annual skeletal layers provide complementary climate information (the
ratio tends to decrease as temperature increases).
In the polar regions, where neither tree ring nor coral records are available, yet another
important climate proxy, ice cores, proves useful. The ratios of the 16 O and 18 O oxygen isotopes in the
frozen water (H 2 O) that constitutes the annual ice layers recovered from ice cores can tell us about
past atmospheric temperatures at the time the snow or ice was laid down. Other information from ice
cores, such as dust layers and trace chemical constituents, can provide further hints about both past
climate variations and what may have driven them, such as volcanic eruptions and solar output
changes. Sediment cores recovered from closed basin lakes in high-latitude environments can also
tell us about past seasonal temperature variations, as these variations influence the amount of
sediment flushed into the lake during a particular snowmelt season.
These are but a few examples of climate proxy records that can potentially be analyzed to assess
past climate phenomena. As the different types of proxy records provide information from
complementary regions, and often for complementary seasons, a diverse (multi-proxy) network of
these data proves particularly useful in characterizing past climate changes. 8 Since my Ph.D. research
was aimed at isolating and understanding long-term climate oscillations, such a network of proxy
records could yield a desperately needed longer-term dataset to analyze. It was a logical scientific
marriage. Ray and I proceeded to collaborate, even as I was still finishing up my Ph.D. research.
Although proxy records are admittedly indirect and imperfect measures of climate, they could
allow us to do something that the brief instrumental record simply could not: to go back centuries in
time. Now we could see whether those multidecadal oscillations discovered in the data were real.
Were they evident several centuries ago, or were they just figments of the modern record? To answer
this question, we applied the oscillation detecting machinery that Jeffrey Park and I had developed to
a considerably longer multi-proxy dataset that Ray had used in previous paleoclimate research. We
concluded that these oscillations were indeed real. 9 They continued to occur throughout a period
stretching over six centuries, and they followed the regional pattern we expected, the signal seeming
to be especially large in the North Atlantic.
That wasn't all we found in the data, however. We were able to detect an even longer-term
apparent oscillation in the proxy data with a periodicity of roughly 250 years. This oscillation, too,
had greatest amplitude in the region surrounding the North Atlantic, and it described a transition in
climate conditions that occurred sometime prior to A.D. 1500, coinciding approximately with what is
now generally accepted as the boundary, at the global scale, between the so-called Medieval Warm
Period and the Little Ice Age. Could our longest-term oscillation, we wondered, be telling us
something about that transition?