Climate Audit

Last summer, after Paul Thacker published a critical article about me in Environmental Science & Technology (also try here), I contacted three of the people prominently quoted in the article – Mahlman, Trenberth and Famiglietti – to obtain confirmation of what they said. As you will see below, their responses are extraordinarily lame. None of them would defend the quotations. Here’s excerpts from Thacker on each of them:

But what began as an interview, Mahlman explains, quickly evolved into a spirited debate. Whenever he pointed out the importance of Mann’s work, Regalado would try to shift the discussion back to McIntyre and McKitrick. “I told him that as far as I know they’re quacks. That kinda riled him. …

Oddly, the McIntyre incident is not an anomaly, according to Kevin Trenberth, head of the climate analysis section at the National Center for Atmospheric Research. “There have been several examples of people who have come into the field of climate change and done incredibly stupid things by applying statistics in ways that are inappropriate for the data,”‘? he says. …

“It is a concern if there is a group that thinks that this one paper is the most important to come out on climate change,”‘? says Jay Famiglietti, an associate professor in earth system science at the University of California, Irvine, and editor-in-chief of GRL, the journal in which McIntyre published his study. “If I had a student come to me and say, “I found this one paper that proves that climate change is hogwash,”‘? I’d say, “Well, that’s one paper out of how many? In science, you never look at [only] one paper.”‘?… Famiglietti, editor-in-chief of GRL, says that because the McIntyre paper generated a total of four letters, an abnormally high number, he will personally supervise their acceptance. He says that the letters differ in their specific criticisms and adds that he is ignoring the political controversy and focusing on the science.

When I contacted the three scientists, none of them were able to provide a coherent justification of their remarks to Thacker, but none of the three scientists has, in the mean time, corrected the public record, requiring me once again to rely on email correspondence to evidence just how mealy-mouthed they were.

MAHLMAN

Once again, here’s what Thacker said:

But what began as an interview, Mahlman explains, quickly evolved into a spirited debate. Whenever he pointed out the importance of Mann’s work, Regalado would try to shift the discussion back to McIntyre and McKitrick. “I told him that as far as I know they’re quacks. That kinda riled him. …

In Nov. 2005, I emailed Mahlman as follows:

In an interview with Environmental Science & Technology, you made highly derogatory personal comments about me, including the statement that Dr McKitrick and I were "quacks". Please provide me with the basis for this claim. Thank you for your attention. Stephen McIntyre

On Nov. 18, 2005, Mahlman wrote back:

I apologize for the delay in my response to your note provided below.

I have been recently totally immersed in reviewing the review draft of Working Group 1’s contribution to IPCC’s Fourth Assessment Review. I admit that I do not recall referring to you and Dr. McKitrick in any form of derogatory slang, nor can I even imagine myself making such a statement, either on or off the record.. If I actually did refer to you in any form of of such slang, I offer my deepest apology. It is certainly possible that I may have suggested that the two of you probably have not made either your scientific conclusions or your motives particularly clear. Indeed, I have not examined your paper in sufficient detail to make a definitive evaluation of its accurary or its significance to the rather odd "debate" about the Mann, et. al. reconstructions of paleoclimate proxy data over the past 1000 years Actually, I participated in a group review of their methodologies over five years ago and we found them to be solid.

At that time, I did point out to Mike that these paleo time series do not in any convincing way alter our inferences on the widespread conclustion that human-caused climate warming is now well underway, and is obvious in the data over the past century. It is clear that the upcoming Fourth IPCC Climate will come out very loud and clear about this inexorable fact.

I do value your candor in writing to me about this, and I wish you success in your future research activities.

Regards.
Jerry Mahlman

On Nov. 19, 2005, I replied to him:

A partial transcript of your interview with Paul Thacker is posted on the Internet here . In it, you are quoted as saying the following: "I said that as far as I know they’re quacks. " Perhaps this may refresh your recollection. While you have communicated an apology here to me, this is really insufficient under the circumstances and I request that you communicate an appropriate correction to Environmental Science & Technology.

You say that you told reporters that we have "not made either our scientific conclusions or our motives particularly clear". Again these are very critical statements; I see no basis for either claim and I would appreciate some support for these statements. Given that you go on to say that you have not "examined our paper in sufficient detail to make a definitive evaluation of its accuracy", there seems to be a certain recklessness in these statements, which is surely inappropriate.

You also say that you participated in a group review of MBH methodologies over 5 years ago and found them to be solid. May I ask what the occasion of the group review was, who was involved in the group review and the forms of due diligence that you carried out to determine that the methodologies were "solid". Your conclusion is completely contrary to my findings and I would like to know the basis of your opposite conclusions.

I do not disagree with your observation that the hockey stick and its cousin series do not, of themselves, have a great deal to say about the great issue of the impact of 2xCO2, although there may be knock-on effects, e.g. tuning of so-called detection and attribution studies. If it is your view that these paleo series are unimportant for the larger policy issue, then it is regrettable that you did not express this view more strongly at the time of IPCC TAR so that this particular graphic might have been de-emphasized both by IPCC and by subsequent users of IPCC, such as the Canadian government, who relied heavily on the hockey stick in their presentation of Kyoto policy.

Regards, Steve McIntyre

Mahlman did not reply. So much for this source.

TRENBERTH

On Sept. 1, 2005, I emailed Trenberth as follows:

In a recent article in ES&T, you are quoted as saying that "There have been several examples of people who have come into the field of climate change and done incredibly stupid things by applying statistics in ways that are inappropriate for the data.”‘?

The context suggests that your comments were directed as criticism against some of my work, rather than that of Mann et al. If so, I would appreciate it if you could give me some particulars on any "incredibly stupid" or "inappropriate" things that you believe that we did, in order that I might respond.Yours truly, Steve McIntyre

On Sept. 2, 2005, Trenberth replied:

I am in Germany, little access to email. I DO NOT KNOW THE ARTICLE YOU REFER TO: It seems like a general statement and it has been true. Such as attempts to analyze data without taking account of inhomogeneities and dealing with full quantities rather than anomalies from annual cycle. Proper appreciation of the data and its limitations is always needed. Kevin Trenberth

Shortly after receiving his reply, also on Sept. 2, 2005, I replied:

The ES&T article quoting you is here http://pubs.acs.org/subscribe/journals/esthag-w/2005/aug/business/pt_wsj.html

In our articles published in GRL and E&E in 2005, the issues of taking account of inhomogeneities or dealing with fulll quantities rather than anomalies from annual cycles did not arise. I agree that "proper appreciation of the data and its limitations is always needed". Our articles were entirely critical in nature and focused entirely on such issues.

Again I fail to understand the basis of your criticism of our work and I request particulars on the "incredibly stupid" and "inappropriate" things" that you claim that we did, in order that we may respond to this criticism.

Your truly, Stephen McIntyre

On Sept 6, 2005 10.51 am Trenberth replied:

Please note I was one of the authors on the following

Mann, M. E., C. Ammann, R. S. Bradley, K. Briffa, T. J. Crowley, M. Hughes, P. D. Jones, T. Osborn. M. Oppenheimer, J. Overpeck, S. Rutherford, K. E. Trenberth, T. M. L. Wigley, 2003: On past temperatures and anomalous late 20th Century warmth. Eos, 84, No. 27, 8 July 2003, 256″€œ257.
Mann, M. E., C. Ammann, R. S. Bradley, K. Briffa, T. J. Crowley, M. Hughes, P. D. Jones, T. Osborn. M. Oppenheimer, J. Overpeck, S. Rutherford, K. E. Trenberth, T. M. L. Wigley, 2003: Response to comment “On past temperatures and anomalous late 20th Century warmth.”‘? Eos, 84, No. 44, 4 Nov 2003, 473,476

These articles were published before I had published any climate articles. I pointed this out to Trenberth later that day as follows:

I remain puzzled. These references have nothing whatever to do with anything that I’ve been involved in.

Did you talk to Paul Thacker of ES&T about our articles? Did you use terms "incredibly stupid" or "inappropriate" in connection with our articles? Thank you, Stephen McIntyre

A little while later I emailed again:

Thacker says that he talked to you 2-3 weeks ago and has referred me to you for an explanation of your comments. These articles pertain to criticisms by Mann of Soon and Baliunas, not to criticisms by us of Mann. Could you please explain exactly what you said to Thacker. Thank you, Stephen McIntyre

Leter that day, Trenberth replied:

I do not know. I talk with media people every day, especially recently with Katrina. It was more than 3 weeks ago. Kevin Trenberth

So I replied again on the same day:

Maybe you don’t know what you said to Thacker. As matters stand today, is it your opinion that we did "incredibly stupid" things in our article and if so, what are they so that we can respond?

If you do not hold this opinion, would you please send a communication to Thacker either withdrawing your prior comment or correcting his statements in respect to your views on our article. Regards, Steve McIntyre

Once more, Trenberth replied promptly as follows:

I suggest you look at the comments and criticisms of your articles on http://www.realclimate.org/index.php?p=98 or http://www.realclimate.org/index.php?p=8 and eslewhere. I have been singularly unimpressed and unconvinced by your work in several respects. e.g.
"we demonstrate that the reconstruction of MM resulted, instead, from their selective censoring of key indicators from the MBH98 proxy dataset. Indeed, we are able to reproduce the MM reconstruction of anomalous 15th century warmth when the entire ITRDB North American data set (and the ‘Queen Anne’ series) are censored from the proxy network (Figure 4). These data (in fact, 70% of all of the proxy data used by MBH98 prior to AD 1600) were unjustifiably censored from the MBH98 dataset by McIntyre and McKitrick (2003) in their original analysis (see Jones and Mann, 2004, and Rutherford et al, 2004 for a discussion)."

You might like to respond to those comments
Kevin Trenberth

On Sept. 12, 2005, I replied to Trenberth:

You have suggested that I ought to look at the realclimate postings and that I “might like to respond”‘? to them. I am completely familiar with everything that has been written about our work at realclimate and I am surprised you do not know that I have already responded to these arguments on many occasions. I have posted extensive replies at my own site, for instance:
http://www.climateaudit.org/?p=81
http://www.climateaudit.org/index.php?p=62
http://www.climateaudit.org/index.php?p=53
http://www.climateaudit.org/index.php?p=52
http://www.climateaudit.org/index.php?p=22
etc.

The arguments that you cite have been rejected by two journals. They were made (inter alia) in a submission by Mann et. al. to Climatic Change in late 2003, to which the Editor asked me to respond. Based on my response, the article was rejected in 2004.Related arguments were recently repeated by Wahl and Amman in submissions to Geophysical Research Letters and Climatic Change and touted in an NCAR press release in May 2005. I was asked to respond to both these submissions. Based on our replies, the GRL submission was rejected. Surprisingly, UCAR has to this date not issued any notice that the GRL submission was rejected and the abstract to the submission remains on the UCAR website http://www.cgd.ucar.edu/ccr/ammann/millennium/refs/AmmannWahl_GRL2005.html. In connection with their CC submission, I pointed out that the authors had withheld key cross-validation statistics that were adverse to their conclusions and were relying on unpublished documents as supporting citations, both of which I requested from the Editor. They were promptly requested by the Editor to divulge the materials in question, which they refused to do. They even tried to pierce the slight veil of anonymity that supposedly existed, saying that the reviewer should request this from them personally. Eventually after the authors remained obdurate in their refusals, I submitted a detailed report outlining the mistakes, misrepresentations and flaws in the paper. I have not heard any further about the matter but to this point the paper has not been published.

You consider yourself sufficiently well-read on the subject to judge my work “singularly unimpressive”‘? and to make disparaging comments about it to a national reporter. But from your comments to date I can only conclude that you have not actually read either of our recent papers, nor have you followed the ensuing debate, and what little you know about the issue is based entirely upon second-hand comments and unreviewed commentary from the opposing side.
But, should you aspire at this point to a more balanced understanding of the issue, here are some detailed points in response to the specific items you quoted.

First, we have stated explicitly and on several occasions that we do not present any reconstruction as a positive account of past climate, as we do not endorse MBH98 methodology. Our principal conclusion in McIntyre and McKitrick [2003] was not a positive reconstruction, as sometimes claimed, but a negative conclusion:

More generally, the extent of errors and defects in the MBH98 data means that the indexes computed from it are unreliable and cannot be used for comparisons between the current climate and that of past centuries.

In our 2003 article, we stated explicitly that the often cited figure was presented

without endorsing the MBH98 methodology or choice of source data.

which would obviously have been necessary to present a positive reconstruction based on MBH98-type methodology. While we had thought that the above and other statements in the article itself had been sufficiently clear to avoid misunderstanding, in order to ensure that there was no misunderstanding, we issued an FAQ shortly after publication of McIntyre and McKitrick [2003] in which we stated:

Your graph seems to show that the 15th Century was warmer than today’s climate: is this what you are claiming?

No. We’re saying that Mann et al., based on their methodology and corrected data, cannot claim that the 20th century is warmer than the 15th century “€œ the nuance is a little different. To make a positive claim that the 15th century was warmer than the late 20th century would require an endorsement of both the methodology and the common interpretation of the results which we are neither qualified nor inclined to offer.
(http://www.uoguelph.ca/~rmckitri/research/trcqa.html).

That we are not arguing for a warm medieval era is even acknowledged at realclimate; e.g. here http://www.realclimate.org/index.php?p=167#comment-2688

It is puzzling that critics of M&M (McIntyre and McKitrick) repeatedly refer to their work as an alternative reconstruction. I base this on two ideas: 1. M&M say they are not offering an alternative reconstruction”⤮

[Response: I wouldn’t characterise the WSJ as ‘critics’ and yet they clearly think that M&M are offering an alternate reconstruction. It’s just one more thing they get wrong… -gavin]

Some misrepresentations of our position are particularly egregious. For example, Crowley requested a digital version of “our”‘? reconstruction. I replied to him that there was no such thing as “our”‘? reconstruction. He then asked for the digital version of one of our figures, noting that some people are “misinterpreting”‘? us. I sent him this data together with a repeated caveat. In EOS in July 2005, after previously receiving and acknowledging these caveats, Crowley astonishingly then criticized “our”‘? reconstruction as differing from other multiproxy reconstructions.
The conclusions in our 2005 articles were limited to statistical issues, such as bias, robustness, statistical significance and proxy quality.

We demonstrated the non-robustness of MBH98 results by discussing, in our 2005 Energy and Environment paper, a reconstruction in which a centered principal components calculations was used rather than the decentered MBH method, and a reconstruction in which bristlecone pines were excluded”€?an experiment MB&H themselves conducted but did not report. These variations each result in high 15th century values. We did not put these forward as representations of past climate, but as demonstrations of the non-robustness of MBH98. The phrases you quoted from realclimate misrepresent our findings for the purpose of creating a straw man. It is true that you get high 15th century values if you don’t use any 15th century North American tree ring data because, a fortiori, you aren’t using the bristlecones. But by stating it in terms of data deletion, realclimate falsely insinuates that we are trying to sell results based on a truncated data set. It is easy to verify that this is untrue, and that we used the same 15th century tree ring data as MBH98.

As we explained in our GRL paper, MBH98 PC methodology mines for hockey stick shaped series (either intentionally or unintentionally) and specifically promoted a low-level signal (driven by the bristlecones) from PC4 to PC1, inflating the associated eigenvalue in the process. If you use a centered PC methodology (without necessarily agreeing that this is a valid methodology either) and 2 PCs for the North American network you get high 15th century results. If you peer below realclimate rhetoric, you will see that they concede this. Their fallback position is that 5 PCs should be used in the North American network under Preisendorfer’s Rule N. The argument is ad hoc in the sense that it was not applied to determine proxy rosters in other tree ring networks, but more importantly, Rule N is necessary but not sufficient”€?it in no way overcomes the substantial problems with relying on bristlecones as the sole determinant of ones’ results. MBH acknowledged this problem in their 1999 paper and proposed an adjustment to remove CO2 fertilization effects, but conveniently only applied the adjustment to the pre-1400 portion. Had they applied an adjustment to the entire bristlecone series there would be high 15th century values regardless of whether you use 2 or 5 PCs.

Thus, the realclimate posting that you cite does not refute any of our claims.

One of the reasons for our interest in robustness is that MBH previously claimed that their methodology was robust to the presence/absence of dendroclimatic indicators altogether “€œ a claim which has been influential in the widespread use of this reconstruction. And yet even their defence against us abandons this claim “€œ obviously if their reconstruction is robust to presence/absence of all dendroclimatic indicators, it should be robust to the presence/absence of bristlecones (or for that matter, the presence/absence of 15th century North American tree ring proxies).

You made reference to the realclimate rhetoric about statistical significance. We showed in our GRL paper that the MBH98 reconstruction abysmally fails cross-validation R2 tests. This claim has not been refuted. It is now clear that the cross-validation R2 test was calculated at the time by Mann et al but not reported for steps prior to their AD1820 step. We also showed in our GRL paper that their seemingly significant RE statistic was spurious, as their benchmarking computation failed to account for the influence of their non-standard PC methodology.

Thus far, in responding to my complaints, you first claimed you could not remember making the comments, then when this position proved untenable you cited Eos commentaries pertaining to an unrelated study to back up your claim, revealing that you didn’t even know what articles were being discussed. Then you dug up old postings on realclimate, apparently in the belief that I am unaware of them, when in fact I have thrice refuted them at journals and debunked them at length on my web site.

Before you opine to the press about “incredibly stupid”‘? things that I and my coauthor are supposed to have done, I believe that you have an obligation to actually read our articles. It is highly inappropriate for you to rely on one-sided, unpublished commentaries on our articles by those who are publicly opposed to our work, and even if you had read the articles it would be highly inappropriate to make the sort of demeaning, derisive comments that you did.

The ES&T article purported to discuss reaction to our 2005 articles in GRL and E&E. Accordingly, I renew my request that you provide a quotation from either article containing an “incredibly stupid”‘? claim, and if you are unable to do so, I request that you publicly retract the statement you made to Environmental Science & Technology and issue a full and unreserved apology to me and my coauthor.
Yours truly,
Stephen McIntyre

Trenberth did not reply to this letter nor did he correct the public record.

FAMIGLIETTI
On Sept. 27, 2005, we wrote to Famiglietti in connection with the Ritson Comment, about being notified that the Comment had been accepted without consulting a Reply, in breach of AGU policies. The following is an excerpt:

We are particularly concerned over this disregard for editorial policies in light of some recent remarks attributed to you by Environmental Science & Technology. In the article, you are quoted as saying (in a discussion of our GRL paper):

"If I had a student come to me and say, "I found this one paper that proves that climate change is hogwash," I’d say, "Well, that’s one paper out of how many? In science, you never look at [only] one paper."

Our paper does not state that "climate change is hogwash" nor does it express any views on climate change. Our paper was limited to statistical issues. We are accordingly concerned that you may hold certain ideas about our work that are inconsistent with what we have actually published.

The ES&T article also stated you had decided to take over editorial oversight of our file. We all rely on an impartial refereeing process in order to ensure that the publication record is as sound and objective as possible. It is still in the public and professional interest that the rules of the process be followed. We are concerned that you have circumvented these rules in this case, to our detriment, and your public comments on our work suggest that you do not approach this file with impartiality. Therefore we request that you assign another Editor to handle the file.

Yours truly
Stephen McIntyre and Ross McKitrick

Famiglietti immediately contacted us and asked to speak to us on the telephone. We arranged a conference call, including Ross McKitrick. Famiglietti insisted that the call be off the record. That’s fine. But we also asked Famiglietti to provide an on-the-record response, which he refused to do, other than agreeing to us makeing the following comment in respect to his comments to ES&T:

We have spoken with Dr. Famiglietti about this. He has assured us that he was not expressing any views on our paper in particular. His point was a general one, namely that an individual paper should not be evaluated in isolation but in the context of the literature of which it is a part. He did not imply any disagreement with the original decision of GRL to publish it.

Later, of course, GRL rejected both the Ritson and Ammann comments.

In summary, this is surely a very unimpressive defence by each of the three scientists of what Thacker represented them as saying.