There He Goes Again, Version 4.0: Mann Claims His Hockey Stick was Affirmed by the NAS

In a letter to the editor published this month in 'Scientific' American, spinmeister Michael Mann fires off ad hominem attacks and yet again claims his "hockey stick" was affirmed by the National Academy of Sciences. The NAS report did nothing of the sort, and in fact validated all of the significant criticisms of McIntyre &  McKitrick (M&M) and the Wegman Report:

1. The NAS indicated that the hockey stick method systematically underestimated the uncertainties in the data (p. 107).

2. In subtle wording, the NAS agreed with the M&M assertion that the hockey stick had no statistical significance, and was no more informative about the distant past than a table of random numbers. The NAS found that Mann's methods had no validation (CE) skill significantly different from zero. In the past, however, it has always been claimed that the method has a significant nonzero validation skill. Methods without a validation skill are usually considered useless. Mann’s data set does not have enough information to verify its ‘skill’ at resolving the past, and has such wide uncertainty bounds as to be no better than the simple mean of the data (p. 91). M&M said that the appearance of significance was created by ignoring all but one type of test score, thereby failing to quantify all the relevant uncertainties. The NAS agreed (p. 110), but, again, did so in subtle wording.

3. M&M argued that the hockey stick relied for its shape on the inclusion of a small set of invalid proxy data (called bristlecone, or “strip-bark” records). If they are removed, the conclusion that the 20th century is unusually warm compared to the pre-1450 interval is reversed. Hence the conclusion of unique late 20th  century warmth is not robust—in other word it does not hold up under minor variations in data or methods. The NAS panel agreed, saying Mann’s results are “strongly dependent” on the strip-bark data (pp. 106-107), and they went further, warning that strip-bark data should not be used in this type of research (p. 50).

4. The NAS said " Mann et al. used a type of principal component analysis that tends to bias the shape of the reconstructions", i.e. produce hockey sticks from baseball statistics, telephone book numbers, and monte carlo random numbers.

5. The NAS said Mann downplayed the "uncertainties of the published reconstructions...Even less confidence can be placed in the original conclusions by Mann et al. (1999) that ‘the 1990s are likely the warmest decade, and 1998 the warmest year, in at least a millennium.’

Mann never mentions that a subsequent House Energy and Commerce Committee report chaired by Edward Wegman totally destroyed the credibility of the ‘hockey stick’ and devastatingly ripped apart Mann’s methodology as ‘bad mathematics’. Furthermore, when Gerald North, the chairman of the NAS panel -- which Mann claims ‘vindicated him’ – was asked at the House Committee hearings whether or not they agreed with Wegman’s harsh criticisms, he said they did:

CHAIRMAN BARTON: Dr. North, do you dispute the conclusions or the methodology of Dr. Wegman’s report?

DR. NORTH [Head of the NAS panel]: No, we don’t. We don’t disagree with their criticism. In fact, pretty much the same thing is said in our report.

DR. BLOOMFIELD [of the Royal Statistical Society]: Our committee reviewed the methodology used by Dr. Mann and his co-workers and we felt that some of the choices they made were inappropriate. We had much the same misgivings about his work that was documented at much greater length by Dr. Wegman.

WALLACE [of the American Statistical Association]: ‘the two reports [Wegman's and NAS] were complementary, and to the extent that they overlapped, the conclusions were quite consistent.’ 

Mann uses the 5 rules of propaganda in his defense, including the rule of orchestration: endlessly repeating the same messages in different variations and combinations [e.g. the NAS gave my hockey stick a clean bill of health].



STICKING TO CLIMATE SCIENCE by Michael Mann


As an undergraduate physics major in the mid-1980s at the University of California, Berkeley, I knew about Richard Muller—the physics professor who was the subject of Michael D. Lemonick’s interview, “‘I Stick to the Science’”—and his controversial theory that a “death star” was responsible for major mass extinctions. Later, as a graduate student studying climate, I became aware of Muller’s work attempting to overthrow the traditional Earth orbital theory of the ice ages—that, too, didn’t pan out. To be clear, there is nothing wrong in science with putting forth bold hypotheses that ultimately turn out to be wrong. Indeed, science thrives on novel, innovative ideas that—even if ultimately wrong—may lead researchers in productive new directions.


One might hope, however, that a scientist known for big ideas that didn’t stand the test of time might be more circumspect when it comes to his critiques of other scientists. Muller is on record accusing climate scientists at the University of East Anglia Climatic Research Unit of hiding data—a charge that was rejected in three separate [whitewashed] investigations. In his interview, Muller even maligned my own work on the “hockey stick” reconstruction of past temperatures. He falsely claimed “the hockey-stick chart was in fact incorrect” when in fact the National Academy of Sciences affirmed our findings in a major 2006 report that Nature summarized as “Academy affirms hockey-stick graph.” Scientific American itself recently ran [pre-climategate] an article it billed as “Novel analysis confirms climate ‘hockey stick’ graph” [“Still Hotter Than Ever,” by [uber-warmist] David Appell, News Scan; Scientific American, November 2009].


Rather than providing a platform for Muller to cast aspersions on other scientists, Lemonick could have sought some introspection from him. How, for example, have the lessons learned from his past failures influenced the approach he has taken in his more recent forays into the science of human-caused climate change? More than anything else, the interview was simply a lost opportunity. Not only can Scientific American do better, it will need to.


Michael E. Mann
Pennsylvania State University