Climate Audit

One of my long-standing concerns of Climate Audit and its readers has been a concern over the role of Eugene Wahl in changing the IPCC assessment of the McMc-Mann dispute in the Final Report – a role that Fred Pearce described in The Climate Files as a “subversion” of IPCC policies of openness and transparency. Jones’ delete-all-emails request was a direct request to erase all evidence of Wahl’s surreptitious contact with Briffa. Even this year, the University of East Anglia has continued to refuse FOI requests for Wahl’s actual proposed changes to IPCC AR4 text. The indolent Muir Russell inquiry doesn’t appear to have bothered obtaining and examining the critical attachments to Wahl’s emails, seeing no need to bother looking behind Briffa’s soothing reassurances on the matter.

For review, the assessment change directly affected Julia SLingo’s evidence to the SciTech Committee that the McMc criticisms of MBH did not have much impact – a point with which we disagree. This observation was based on the IPCC assessment which stated in the Final Report (quoted on page 80 of Muir Russell):

Wahl and Ammann (2006) also show that the impact on the amplitude of the final reconstruction is very small (~0.05°C; for further discussion of these issues see also Huybers, 2005; McIntyre and McKitrick, 2005c,d; von Storch and Zorita, 2005).

This was a substantial change in Wahl’s (and Mann’s) favor from the assessment in the Second Draft – the assessment sent to reviewers.

it is unclear whether it [the MM criticisms] has a marked impact upon the final reconstruction (Von Storch et al., 2004; Huybers, 2005; McIntyre and McKitrick, 2005).

No review comments on the record supported the changed assessment. Together with some puzzling content in the author responses to Review Comments, both David Holland and I surmised that there had been off-balance-sheet contact between Ammann and Wahl and Briffa that had not been archived with other review comments. (This becomes a long story.)

When the Climategate emails became available, it became clear that there had in fact been surreptitious contact between Wahl and Briffa – the emails evidenced that the correspondents were aware of the furtiveness of their exchange, the emails being marked “confidential”, were not copied to their usual cc list. Muir Russell didn’t bother asking Briffa why they were being so furtive.

One of the emails showed that Wahl had sent proposed changes to the assessment. Since reading the emails, it has been my opinion that Wahl drafted the language of the Final Draft that favored his assessment and that Briffa used Wahl’s amended language in the final assessment. While theoretically, there is joint responsibility for the final product, I very much doubted whether the Wahl-Briffa changes to the final assessment on this point were specifically reviewed by the overall committee at the time.

This raised the interesting question: could Wahl and Briffa between the two of them have inserted the changed assessment without anyone else in the world paying any attention prior to the publication of the final report?

In one of the documents at the Muir Russell website, this question is more or less asked of Chapter 6 Review Editor John Mitchell (whose untrue prevarications to refuse FOI requests have been discussed here in the past.) The question:

It would be helpful to understand the particular practices of the chapter 6 writing team as follows: …

3. to what extent would it be possible for individual team members to ensure inclusion of material in a report without oversight or challenge by other team members?

Good question in respect of the Wahl-Briffa exchange. Could it have been inserted in the Final Draft without oversight? (This is my opinion.)

Mitchell’s answer is as follows:

8. The palaeoclimate chapter depends on inputs from a wide variety of scientific fields. The early stage of assembly of evidence was done by individuals and groups of authors according to their discipline. It was then assembled into a first draft on which expert reviewers comments were received. The review editors were involved in plenary meetings prior to the second order and final drafts, where they ensured that reviewers comments had been properly dealt with.

9. All the authors would meet at least two or three times in each of the meetings I attended to discuss main issues in the chapter, the wording of the summary and cross chapter issues (I missed the beginning of the Bergen meeting). There were over a thousand comments in each review, and over sixty pages in the final chapter, so it was not practical or the best use of time for everyone to go over every part of the chapter line by line. (Nevertheless, I think after the first review, I think most of the authors went through most of the reviewers’ comments together at the beginning of the New Zealand meeting). The so-called “hockey stick” curve of Mann et al (1998) was widely discussed, as was the relevant wording. The executive summary was reviewed by the whole group to ensure that it accurately represented the evidence presented in the chapter, and it is the whole group that should be regarded as responsible, rather than any one person. [my bold]

10. In particular, it was accepted, including by Briffa, that the original Mann et al (98) paper had flaws relating to the particular use of principle components. Other papers either used Mann’s approach without using the flawed method of applying principal components, or used a different approach altogether, and it was on these papers that assessment of the last 1000 years was based. Many of the critical comments were directed at the Mann et al (98) paper which was acknowledged to be problematic. Nevertheless, before the second meeting, I noted the reviewers’ concerns to both the convening lead authors and the other review editor before the Bergen meeting to make sure they were aware of the issues being raised, and to ensure they were discussed in the wider group.

As you can see, Mitchell’s answer was mostly unresponsive to the question. The key phrase is the one that I bolded:

There were over a thousand comments in each review, and over sixty pages in the final chapter, so it was not practical or the best use of time for everyone to go over every part of the chapter line by line.

The answer, in short, is that the Review Editors did not in fact bother going over the chapter line by line because it wasn’t “practical”.

The implication for the Wahl insertion is quite clear. It was entirely possible for “individual team members to ensure inclusion of material in a report without oversight or challenge by other team members”.

Muir Russell
Now let’s look at Muir Russell’s coverage of this long-standing and central dispute – the dispute at the heart of the delete-all-emails request.

Muir Russell stated:

Wahl did not write any of the main text, though he did make some suggestions for very minor edits. p 82

From the perspective of the Hockey Stick dispute that was the most contentious issue of this chapter, Wahl’s suggestions were not “very minor”. They altered the assessment of the dispute from being “unclear” to being “very small”. This is not a “very minor” change.

Although Muir Russell quoted at length (p 82) from Mitchell’s answer, Mitchell’s answer to question 3 is not quoted. Instead, Muir Russell reported:

The evidence of the Review Editor suggests that no one person in the writing team could have overridden the team responsibility for the text. (p. 83)

and on page 76:

24. Irrespective of the above comments on issues i) and ii), the evidence of the Review Editor underlines the team responsibility for the text, and the unlikelihood that a single voice could dominate on an important issue. p 76

In making this assessment, Muir Russell did not quote directly from Mitchell and his direct evidence (see here) not only does not support Muir Russell’s finding, but concedes that the final draft wasn’t gone over “line by line” and that it was entirely possible that Briffa could have inserted Wahl’s changed assessment into the IPCC report without “oversight or challenge by other team members”. (Whether or not the Lead Authors were or were not jointly “responsible” for the text is a different issue.)