The hockey stick: a retrospective - Climate Change: The Facts - page 145

14 The hockey stick: a retrospective
Ross McKitrick
The fact that times in the past experienced a warmer climate than today is highly
inconvenient for the proponents of the theory of human-induced global warming. So the
finding in 1998 that temperature trends were much higher in the present day than the past
by the American climatologist Michael E. Mann was a key part of the political move towards
climate change policy.
Rather than the climate oscillating between hotter periods (for instance, the medieval
warm period) and cooler periods (for instance, the little ice age between the fourteenth and
nineteenth centuries), Mann's findings suggested that global warming was out of control.
The graph of temperatures in the northern hemisphere looked less like a wave, and more
like a 'hockey stick'.
Mann's redrawn temperature graph was in part based on a study of Siberian tree rings.
Ross McKitrick and Steve McIntyre studied the data behind the evidence and the statistical
techniques used to interpolate it and found key errors in the statistical technique used to
combine Mann et al.'s blended data would almost invariably produce a 'hockey stick'.
The controversy led to a US Senate Committee setting up an inquiry under Professor
Edward Wegman, who essentially reported that Mann's claims—that the 1990s were the
hottest decade in 1000 years—could not be supported. We asked Ross McKitrick to write
this synopsis of the issue.
—Alan Moran, Editor
The best place to start when learning about the hockey stick is Andrew Montford's superb
topic The Hockey Stick Illusion . 1 OtheressentialsourcesaretheoriginalMannetal.papers, 2
the McIntyre and McKitrick papers, 3 Steve McIntyre's and my presentation to the National
Academy of Sciences Panel, 4 McIntyre's Ohio State University presentation, 5 a few survey
papers and chapters of mine, 6 and McIntyre's climateaudit. org posts over the past decade
on proxy quality, the Yamal substitution, the Briffa truncation, data secrecy, and some other
issues. 7
It is sometimes said that we found Michael Mann's algorithm would always produce a
hockey stick, even from random numbers (Figure 2). That is not quite right: we found that
the algorithm could do so, given the right kind of random numbers (autocorrelated, rather