Posts
Hockey Stick History: Past and Present
The public and political controversy over the science of global warming has become so tangled and has so many parts to it, that it is difficult to sort the whole thing out. I can't say that I have it all sorted out myself. This post is to help organize some information for myself, as well as to provide background to some new developments.
In the Beginning
Once upon a time (1990, to be specific), the IPCC issued their first climate report, which contained a figure representing the global average temperature over the last millennium. The figure looked like this, with the dotted line representing conditions at the beginning of the twentieth century:
We should note here that there are three problems with this figure that are often overlooked. First, in the IPCC report it is called a "schematic." Second, it is based on data from mostly central England, and cannot, by itself, be considered representative of the whole globe, or even the whole Northern Hemisphere. Third, although it appears to be current up to the end of the twentieth century, the data it is based on do not include the last few decades of the twentieth century which is when much of the current warming has taken place. (See Common graphical tricks and the Medieval Warm Period.)
Then in 1998 a young climatologist named Michael Mann did a study of tree rings, corals, and other indicators in order to reconstruct the average temperature of the Northern Hemisphere and compared it to the recent record taken by instruments. The results were published in 1998 and 1999, and looked like this:
This image was the basis for a figure featured prominently in the 2001 IPCC report, which looked like this:
Thus the 'hockey stick' was born and the conditions for controversy were set. The last figure suggests that the current warming of the planet is unprecedented in the last 1,000 years and helps to bolster the case that humans are warming the planet. In contrast, in the first figure the current warming (which actually is not shown, as noted above) appears insignificant compared to the so-called medieval warming period (MWP). Skeptics of global warming have seized upon this difference. They claim that the first figure (along with other data) shows that global warming is not significant and charge that bias, if not outright fraud, led the IPCC to embrace the later figure and discount the MWP.
Retired mining statistician Stephen McIntyre and economist Ross McKitrick became interested in Mann's figures and published criticisms of the statistical analysis behind them. Their basic claim was that the shape of Mann's graphs were an artifact of incorrect application of statistical analysis, and that the analysis relied too heavily on certain data sets which bias the early points toward the low side. Their 'corrected' version, published in 2003, looked like this (note that the instrumental record is not shown):
These criticisms touched off a number of papers with claims and counter claims, and here those of us without statistical training see everything blur into a nerd-fight over statistical methods. Added into this mix is a spat over data and whether Mann shared enough of his data and code with McIntyre and McKitrick, and whether he was even obligated to. To hear McIntyre and McKitrick tell it, they were able to uncover flaws in Mann's work through heroic efforts in overcoming Mann's stonewalling. To hear Mann tell it, McIntyre and McKitrick made ill-founded criticisms and were demanding data and computer code that they were not entitled to. You can get a sense for this in the Wall Street Journal article, In Climate Debate, The 'Hockey Stick' Leads to a Face-Off.
What I have outlined thus far accounts for a significant portion of the public controversy. Meanwhile, in 2005 a group of scientists in Sweden and Russia published their own reconstruction of Northern Hemisphere temperatures over the last 2000 years, which looked like this:
Investigation
Eventually Congress got involved. It asked the National Research Council (NRC) of the National Academy of Sciences (NAS) to look into Mann's methodology and conclusions. The report, Surface Temperature Reconstructions for the Last 2,000 Years, was released in 2006. Although it did contain some criticism of Mann's statistical analysis and conclusions, it was basically supportive. In 2008 Mann published a new reconstruction that took into account the NRC criticisms. The result looked like this (note that the figure also shows other published reconstructions for comparison, such as the Swedish one above (Moberg et al.)):
So, we end up with a picture that is a little more wavy on long time scales than Mann's original reconstruction, but with current warming still elevated relative to the last two millennia.
Meanwhile, back in 2006, the House Committee on Energy and Commerce, chaired by Rep. Joe Barton (R-TX), held its own hearing (here, here, and here). The committee asked George Mason University statistician Edward Wegman to assemble a panel and to essentially referee the running dispute between Mann verses McIntyre and McKitrick. While not denying the reality of global warming (something sometimes glossed over by skeptics), the report, dubbed the Wegman report, basically upheld all of McIntyre and McKitrick's criticisms of Mann's early work. It also went on to suggest that the principle researchers within the niche of paleoclimatology were too cozy with one another, which compromised the peer-review process (something Mann denies since, among other reasons, most of his scientific relationships were formed after the publication of the papers in question).
Because they were released in the same year, and because Wegman has been a chair of the NRC's Committee on Applied and Theoretical Statistics, the NRC report and the Wegman report are sometimes conflated. They are not the same. As mentioned above, the NRC report was produced under the auspices of the NAS with a panel of at least 12 scientists and an additional 13 reviewers. It focused less on Mann's early papers than the current state and reliability of the science. In contrast, other than Wegman and one other respected statistician (David Scott), the other contributors to the Wegman report were then-current or former graduate students of Wegman, including Yasmin Said, the third co-author. The Wegman report was ostensibly limited to the question of McIntyre and McKitrick's criticisms of Mann's early papers and gave little consideration to later developments in the field. It specifically disclaimed any position on whether global warming was caused by humans. There is some overlap in the basic findings in these two reports. However, the Wegman report is more negative and is naturally more popular with climate skeptics.
Back to the Future
All of this might be water under the bridge were it not for the Climategate hoopla of a year ago. The content of the private emails, made public, re-ignited the old disputes. For the six months following the release of the emails, climate scientists took a public beating, often for passages taken out of context. The controversy has died down as official investigations have cleared the climate scientists of wrongdoing, but based on the emails the Attorney General of Virginia is seeking to investigate Mann for fraud and his brief relies in part on the Wegman report. (I didn't see any references to the NRC report.) Despite its limited scope, the Wegman report remains a touchstone of climate skepticism.
While the climate scientists were being pilloried in the press, a Canadian blogger who goes by the name of "Deep Climate" was taking a closer look at the Wegman report. He found that some of the background material on paleoclimatology was plagiarized from a textbook that was authored, ironically, by one of the co-authors of the very papers being criticized, and that other portions were taken from Wikipedia. These findings inspired computer scientist John Mashey to take a detailed look at the Wegman report and the circumstances around it. He found additional examples of plagiarism and other alleged forms of poor scholarship. His report is lengthy and overwhelmingly detailed, but his bottom-line conclusion is that the Wegman report was designed to provide a scientific facade for the PR campaign of Washington D.C. think tanks.
Deep Climate is continuing to digest the Wegman report and has recently posted analysis (here and here) suggesting that the Wegman report simply re-capitulated the analysis of McIntyre and McKitrick without any real attempt at understanding the underlying details. And in another twist of irony, Stanford emeritus physics professor David Ritson has been trying to get details on Wegman's analysis for four years, but to no avail. Wegman has thus far exempted himself from following principles of openness that his report lectured Mann about.
Ravings of partisans? Perhaps. However, the plagiarism charge has gained traction. As reported by USA Today (here and here) Wegman is under investigation by his university and outside experts agree that there are clear examples of unattributed quoting or nominal re-wording of text. (As if experts were needed; it's pretty obvious when you look at the comparison.)
According to Mashey's analysis, the plagiarism extends into the dissertations of Wegman's graduate students, one of which was a co-author of the Wegman report. It seems plausible to me that Wegman farmed out the basic composition of the report to his student and that she wrote most of the text and is responsible for the plagiarism. Nevertheless, aside from being an academic sin in and of itself, the plagiarism raises questions about the level of scholarship of the report as a whole. Whether further criticisms by Mashey and Deep Climate will stick remains to be seen.
Over a decade since its initial publication, the hockey stick controversy lives on.
