Hockey Stick Slapped: Climate Change's Bellesiles?

Hockey Stick Slapped: Climate Change's Bellesiles?

Murray Op-Ed at National Review Online
November 02, 2003

One of the pillars on which the alarmist case for doing something about global warming rests is the contention that the 20th century was the warmest in the last thousand years.  This proposition is most dramatically expressed in the "hockey stick" graph contained in the Intergovernmental Panel on Climate Change's third assessment report.  The graph shows temperatures mostly flat for the last 1,000 years, before a sudden, sharp rise in the 20th century that, on a graph, looks like the blade of a hockey stick.  The graph is very persuasive — it caused Sen. John McCain (R., Ariz.) to express alarm when he first saw it in May 2000, and played a part in McCain's bringing to the Senate floor the misguided and economically destructive Climate Stewardship Act, which was defeated last week. <?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

One of the problems with the graph is that its smooth progression through the first 900 years is at variance with established scholarship in the field.  We know that the 1600s-1800s saw both a Medieval Warm Period — when the Vikings colonized <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" />Greenland — and a Little Ice Age — when the River Thames froze over regularly in London.  The graph does not show much variance for either of these occurrences.  It was constructed by calibrating many different proxies for temperature, such as tree-ring widths or measurements from ice deposits at the poles.  It should also be noted that the blade of the hockey stick consists of actual temperature readings from thermometers, not the proxies, which themselves do not show nearly as great a spike.  But the graph's defenders argue that the well-documented temperature swings of history were localized, and only the 20th-century warming is a global phenomenon.

The debate over the graph exploded into new life in late October when two Canadian experts in statistical analysis — Stephen McIntyre and Ross McKitrick (known as "M&M" in discussions of the climate debate) — published what they called an "audit" of the data underlying the hockey-stick graph.  Their audit was based on information provided by an associate of Michael Mann, whose articles inspired the hockey-stick graph in the first place.  M&M investigated the data underlying the graph with the original source data.

They found numerous and worrying errors.  As they put it, the data "for the estimation of temperatures from 1400 to 1980 contains collation errors, unjustifiable truncation or extrapolation of source data, obsolete data, geographical location errors, incorrect calculation of principal components and other quality control defects.”  They used the original source data to correct these errors, after which they concluded, "The particular 'hockey stick' shape . . . is primarily an artefact of poor data handling, obsolete data, and incorrect calculation of principal components."

If true, this conclusion is devastating to the "hockey stick" argument, and we must conclude that 20th century was not unusual at all.  The warming trend could simply be natural phenomenon; support for the greenhouse theory dissipates; and we therefore have little need to enact restrictions on carbon-dioxide emissions (restrictions that, in practice, amount to restrictions on energy use and, therefore, restrictions on wealth).

Michael Mann's reaction has so far been dismissive.  He attacked the M&M paper as a "political stunt," seemingly before reading it.  His retort was transmitted through the website of a sympathetic freelance journalist, and has muddied the debate considerably.

Mann claims that M&M used the wrong data set, and that they only used 112 proxies when 159 were needed.  But M&M point out on their web site that Mann's original research paper contained only 112 proxies, and that these were the proxies Mann instructed his associate to provide to them.

Mann also asserts that many other paleoclimatologists have been able to replicate his results closely.  But this response does not address the question whether those experts uncovered the same errors in the data that M&M, coming to the issue fresh, were able to find.

The whole affair bears strong resemblance to the recent Bellesiles controversy.  Emory University historian Michael Bellesiles won a Bancroft Prize for his argument that gun ownership in early America was not widespread.  It took an amateur historian, Clayton Cramer, to point out that this claim could not be substantiated on the basis of actual gun-ownership records.  Eventually, an Emory University investigation strongly criticized Bellesiles, and the Bancroft Prize was withdrawn.

So far, it looks like the errors in Mann's data set were accidental.  Yet it will be interesting to see how far the proponents of strong action on climate change go to defend the data without addressing the fundamental question: Are the numbers as proposed by M&M right?  If they are, then the climate debate will need to change.