Law and order in paleoclimate

Unfortunately, lawyers are again making forays in the realm of climate science, which is distracting to say the least and probably only damaging. It is clearly not helpful.

I am not acquainted with legal terms and definitions and so this post is just an account, from an interested lay person, of my thoughts about two cases that have grabbed my attention in the last days/months. In both cases I have some background knowledge, though probably not complete.

The first one is related to the subpoenas issued by the Virginia General Attorney trying to prod the University of Virginia to deliver a series of older emails of Micheal Mann, based apparently on the reasoning that Mann could have fraudulently applied for public funds to fund his research activities. This case is for me pretty clear and I do not need to add much to what has been already expressed in blogs of different orientations, Roger Pielke Jr. , Climate Audit, and realclimate, among others. Independently of whether or not one agrees with Mann's scientific results or activities, the eventual accusation of fraud appears to me to have no base. I have no reason to doubt that all his research proposals were duly reviewed by peers and approved or rejected by the corresponding funding agencies based on those reviews. One thing is to argue that this or that paper might be not correct and a completely different thing is to accuse someone of fraud on the basis that perhaps he performed an un-centered principal components analysis instead of a centered principal component analysis. This case smacks of being a modern witch-hunt, and it might prove to be a dangerous precedent should the Republicans regain control of the House in November.

In the second case, Edward Wegman, the author of the so called Wegman report on paleoclimate reconstructions published in 2006, would be accused of plagiarism for extended use in that report of passages taken from the book 'Paleoclimatology' by Raymond Bradley. I think the accusation here is exaggerated, in the same way that  the relevance of the Wegman report was also exaggerated at the time of publication. Wegman is a professional statistician and was commissioned by Congressman Barton (Rep.) to review the climate reconstructions of the past millennium published by Mann and co-authors in 1998 and 1999. The report contains a general background part, which I guess was included to help the members of the House committee to better understand the following, more technical parts. It is quite possible that Wegman made heavy use of paleoclimatology text books to prepare this chapter, among which Bradley's book is very well known. It seems that Wegman did not include the citations in place that are commonly used in scientific texts, although Bradley's book appears cited at the end of the report in the literature list. If I had been in Wegman's place I would have included a couple of sentence explaining that, since Wegman himself is not a climate scientist, he had freely used material from this and that book to convey the necessary introduction for the policy makers. Honestly, I do not think this a big deal, and certainly not a cause for litigation - but what is not possible in paleoclimate science these days ?

A much more substantive matter would be to discuss the findings described in the Wegman report itself. The years 2004-2005 were, as far palaeoclimatology is concerned, different from the present time. At that time, the Wegman report, together with the National Research Council  assessment of climate reconstructions over the past 2 thousand years, appeared as quite a novelty , as they kicked-off a public and unbridled discussion about the hockey-stick. However, from today's perspective, the Wegman report appears to me as quite insubstantial, focused on unimportant aspects of the Mann-Bradley-Hughes reconstruction method, and - most worryingly, having been written by a statistician - not mentioning the real problems affecting this and other reconstruction methods, and that have been (re)-discovered later, not by statisticians, but by humble climate scientist. In some sense the Wegman report reminds me of the recent manuscript by McShane and Wyner, more focused on politically sensitive issues that on real scientific ones. In essence, the Wegman report, as the first part of the McShane and Wyner paper, indicated that the un-centered principal component analysis applied to some sets of proxies used by MBH was incorrect, and that it could lead to an artificial 'flattening' of the reconstructed past temperatures. Although this mistake was correctly pointed out, it bears small relevance for paleoclimate reconstructions. The big problems lie elsewhere: in the tendency of virtually all methods to underestimate past variability. This was completely missed by Wegman at that time. Five years later, we - dumb climatologist- are still awaiting  the smart statisticians to enlighten us with the proper method to conduct climate reconstructions.