A Candle in the Dark - The Hockey Stick and the Climate Wars - page 57

of the planet. McKitrick and Michaels purported to demonstrate that surface warming was an artifact
of contaminated temperature records, based on a correlation they claimed to have found between
economic activity and temperature change in various regions of the globe. Such a relationship, the
pair argued, was unphysical and thus could only be explained by some sort of contamination of the
underlying temperature data.
Though Michaels claimed the paper was subject to “four years of one of the most rigorous peer
reviews ever,” Australian computer science expert Tim Lambert discovered a fatal problem just
weeks after its publication. 14 In attempting to account for the effect of latitude on temperature trends,
McKitrick confused the mathematical entities of degrees and radians (two different ways to measure
angles that differ by a factor of roughly fifty in magnitude!) and had in essence been feeding his
analysis procedure random noise. This alone was enough to render the conclusions erroneous, but
other equally significant problems in the paper were subsequently documented as well. 15
Another article, this one published in Nature in 2005 by British physical oceanographer Harry
Bryden and colleagues, 16 though its thesis did not pan out in the end, provides an example of the
erratic path of scientific progress. The paper appeared to demonstrate a dramatic slowing over the
previous several decades in the North Atlantic “thermohaline” or “conveyor belt” ocean circulation
pattern—the scenario envisioned a year earlier in the movie The Day After Tomorrow . While nothing
as dramatic as Hollywood had envisioned could conceivably befall us, credible climate model
simulations indicated that the conveyor belt could indeed weaken or even collapse over several
decades, with some regions in or neighboring the North Atlantic ocean paradoxically cooling, even as
the globe continues to warm. Bryden and colleagues had claimed that such a thing was not only
conceivable, but was possibly happening already—a 25 percent reduction in the strength of the
conveyor belt had apparently taken place in just decades, they claimed. That no inkling of cooling in
the North Atlantic had been witnessed and no state-of-the-art climate models had predicted any
weakening of the conveyor belt to have taken place yet, however, suggested that considerable caution
was needed before jumping to the conclusion that such a change in circulation was in fact underway.
More detailed measurements taken later showed that the Nature study likely suffered from a
phenomenon known as “aliasing”—mathematically equivalent to the “wagon wheel effect” (the
appearance that wagon wheels are spinning in the wrong direction in very old movies with limited
frame rates). We know that the strength of the ocean current system varies on all timescales. The
apparent thirty-year-long trend probably reflected sparse temporal sampling; there were only five
measurements available over the entire time period. The early measurement appears to have caught
the ocean current system, by chance, in a relatively strong phase, while the recent measurement, by
contrast, happened to catch the ocean current system in a relatively weak phase. This chance set of
measurements thus gave the false appearance of a negative long-term trend.
This episode reflects science working as it should. Experts analyzed a set of observations that
motivated a rather unexpected hypothesis—that climate change was already leading to a dramatic
weakening of the North Atlantic current system. This spurred scientists to make more refined and
detailed measurements. Those measurements showed the original hypothesis to be wrong, but
consider what happened in the process. We learned something new and important about the way a key
ocean current system operates: that it may vary erratically in its properties on short timescales—a
perfect example of science's own wonderfully erratic path toward greater understanding.