07 July 2010

The Muir Russell Review

The Muir Russell Review is out (here in PDF), and it has plenty in it for everyone on all sides of the debate over the East Anglia emails to crow about and to complain about. It has some strong rebukes of the scientists involved in the emails and dismisses many, but accepts some, of the criticisms raised by their strongest critics.

In this post I want to highlight a very puzzling statement in the report. The Muir Russell report characterizes the IPCC as follows (p. 41):
The IPCC produces assessments of the current state of understanding of climate change, its causes and implications. Its approach is to produce the most probable account of these issues; together with their uncertainties, and to identify where there is insufficient evidence to discriminate between different interpretations of a phenomenon. Its purpose is to produce a "best estimate" of what is currently understood, through the work of a group of scientists chosen for their expertise and experience to make reasoned assessments on the balance of evidence. It is not to produce a review of the scientific literature.
The idea that the IPCC presents a "best estimate" understanding based on the views of a selected group of scientists is completely contrary to how the IPCC characterizes its own work. To suggest that the IPCC is "not to produce a review of the scientific literature" is just plain wrong.

Here is how the IPCC WG I (the relevant working group for the MR inquiry) characterizes the scope of its own assessment process (emphasis added):
All chapters undergo a rigorous writing and open review process to ensure consideration of all relevant scientific information from established journals with robust peer review processes or from other sources which have undergone robust and independent peer review.
Note that it says "all relevant scientific information" -- it says nothing about a "best estimate." The IPCC states very clearly in its principles for report preparation (PDF) that its reports are supposed to,
present a comprehensive, objective, and balanced view of the areas they cover
And it explains that authors of IPCC reports,
should clearly identify disparate views for which there is significant scientific or technical support, together with the relevant arguments
IPCC reports are further supposed to
represent the latest scientific, technical and socio-economic findings and are as comprehensive as possible . . . [and] provide a balanced and complete assessment of current information
The notion that IPCC reports are supposed to present a selective view of climate science, representing the judgments of a select group of experts is in fact contrary to the mission of the IPCC.

The Muir Russell mischaracterization of the IPCC becomes relevant in the report when it uses the characterization as a criterion for evaluating the efforts revealed in the emails to minimize or exclude certain perspectives. For instance, the Muir Russell report explains with respect to one alleged instance of exclusion of peer reviewed literature from IPCC drafts that (p. 76):
Those within the [IPCC] writing team took one view, and a group outside it took another. It is not in our remit to comment on the rights and wrongs of this debate, but those within the team had been entrusted with the responsibility of forming a view, and that is what they did.
This speaks directly to problems of the IPCC, revealed to some degree by the emails, but of much broader concern. The IPCC is supposed to "identify disparate views" not hide them from view or take the side held by the author team. Had the Muir Russell review actually taken an accurate view of the IPCC, it is likely that its judgment about the appropriateness of the behaviors revealed by the emails would be considerably different.

It is not the job of the IPCC authors to serve as selective arbiters of the peer reviewed literature and judge which peer reviewed science they agree with and disagree with. This only invites extra-scientific considerations into the assessment process and a cherrypicking of the literature, rather than a considered assessment. The job of the IPCC should be exactly as it says it is -- to produce a comprehensive, balanced and complete review of the relevant literature. If the IPCC finds itself in a situation where its author team reflects a perspective represented by only a subset of the literature, then the IPCC has a problem.

The released East Anglia emails -- for better or worse -- revealed some problems associated with in-group control of parts of the IPCC. Muir Russell's sanctioning of in group behavior in the preparation of IPCC reports is a notable shortfall in what otherwise appears to be a nuanced and comprehensive assessment of the implications of the East Anglia emails.