Michael Mann’s new book, The Hockey Stick and the Climate Wars, discusses my comment [SMc note – this post is by Hu McCulloch] , “Irreproducible Results in Thompson et al., ‘Abrupt Tropical Climate Change: Past and Present’ (PNAS 2006),” that was published in 2009 in Energy & Environment. My comment pertained to an article by Lonnie Thompson and eight coauthors that had appeared in the Proceedings of the National Academy of Sciences back in 2006. As soon as my comment was accepted for publication, I posted a less technical summary on Climate Audit, entitled “Irreproducible Results in PNAS” (4/24/09).
Dr. Mann’s discussion of my comment is the first published feedback to it by any of Thompson’s associates, and hence I am very grateful for the attention he has drawn to it. However, I beg to disagree with Mann’s appraisal of it.
In my comment, I show that the 2000-year tropical ice core d18O composite index shown in Thompson’s Fig. 6 and tabulated in his supplementary Data Set 3 cannot be constructed as a linear combination of the seven individual ice core series on which it was supposed to have been based, as shown and tabulated (back to A.D. 1600 only) in his Fig. 5 and Data Set 2.
Mann replies,
I knew McCulloch’s claim that the tropical ice core composite was “irreproducible” was false, as I was able to reproduce Thompson et al.’s results easily from their raw data. I was also able to identify McCulloch’s error — an incorrect assumption that the amplitude of variation in a series of measurements must be constant in time — in about a half hour of work. [fn. 60] (pp. 205-6)
While, as I noted already in a 5/1/09 update to my CA post, I don’t doubt that Mann can construct the 400-year ice core index shown (as 5-year averages) in Thompson’s Fig. 5 and tabulated in his Data Set 2, the bottom-line decadal 2000-year composite in Thompson’s Figure 6 and Data set 3 does not correspond to decadal averages of the illustrative 400-year series shown in Figure 5 and Data Set 2. Mann may have private access to different data than was archived with Thompson’s article, and perhaps this private data will generate the 2000-year composite, but I demonstrate that this cannot be done with the data provided to the public in Data Set 2, even for its truncated time period. Until Thompson provides this data as requested in my comment, his results in PNAS are irreproducible, as claimed.
Mann’s footnote 60 (on pp. 333-4) continues,
[60] For those who are interested in the technical details, the matter involved the statistical concept of standardized Z scores. These are the series of numbers that result from taking an original dataset, subtracting off the average, and dividing by the standard deviation. This yields a convenient new version of the dataset whose average is zero and standard deviation is equal to one. That latter property only holds for the full dataset. If one takes some subset of the data, the average will not in general be zero and the standard deviation will not in general be one. McCulloch’s error essentially amounts to having assumed that the original Z scores defined by Thompson and colleagues, over the full data interval A.D. 0-2000, would still have average zero and standard deviation of one over a much shorter interval of A.D. 1610-1970. Because of this error, McCulloch ended up using a different weighted average of ice core data than the simple uniform weighted average used in the original Thompson et al. paper, which was the only reason he was unable to reproduce the Thompson et al. result. The error was all McCulloch’s.
In fact, in my comment, I did find with a least squares regression that the tropical composite Z-score index in Data Set 3 is simply the average of the two regional composite Z-scores (for the Andes and Himalayas) in the same Data Set, to within rounding error. This led me to suspect that the regional composites were similarly computed by averaging Z-scores for the individual sites in the two regions, as I noted on p. 371 of my comment. Other possibilities were that he more logically averaged the raw d18O series before taking regional Z-scores, or that he weighted them using multivariate Ordinary Least Squares, multivariate Classical Calibration Estimation, or even Partial Least Squares as in Thompson’s recent article with Kinnard et al. in Nature (24 Nov. 2011). But all these methods produce fixed weights that would have shown up as a virtually exact fit in my regressions.
Although I did not have Mann’s access to the full 2000-year source data, and therefore could not compute the full-sample means or standard deviations as he did, the net coefficients should have been readily recoverable with an almost exactly fitting least squares regression using the 37 decades for which all cores had data in Data Set 2. In fact, the regression standard deviations were 100 times the expected rounding error for the Himalayan region, and 30 times the expected rounding error for the Andean region, using the best fitting coefficients. If Dr. Mann can find any exact fit to within rounding error, I hope he will share it with us.
Most of the sites in question have multiple cores, so in all likelihood Thompson simply used a different combination of cores for the 400-year series in Figure 5 than for the 2000-year series in Figure 6. Undoubtedly these cores vary in quality, and hence Thompson may have had valid reasons not to include all of them. However, if the subset used in Figure 5 is correct, then the index in Figure 6 is wrong, and vice-versa. Either way, the bottom-line 2000-year index in Figure 6 cannot be replicated, even over its last 400 years, from the data represented as its source in Figure 5.
If all of Thompson’s cores had been fully archived at the NCDC’s website, as is customary for this sort of data, it would be feasible to back out which ones he used by a least squares regression on decadal averages of all of them, region by region. The excluded ones would then simply show a zero coefficient. However, with the notable exception of Quelccaya, Thompson has kept most of this primarily NSF-funded data to himself (and a few select friends, apparently), for one, two, or even three decades after it was collected. If he does not release definitive versions of this valuable data soon, it may be lost to science forever.
Mann states that
in 2009 … an economist and climate change contrarian named Hu McCulloch alleged that he could not reproduce a tropical ice core record produced by Lonnie Thompson and colleagues, implicitly claiming either ineptitude or, worse, malfeasance. (p. 205)
Please note that “ineptitude” and “malfeasance” are Mann’s words, not mine. I merely claimed that Thompson’s results were irreproducible. Thompson and his colleagues could have easily corrected this problem, simply by providing the data from which his results were actually calculated and explaining how the calculation was done.
Mann concludes his footnote 60 as follows:
How McCulloch and all reviewers of the paper could have missed something as basic as this is rather bewildering, even more so since McCulloch could have simply walked over to the other side of campus to ask Thompson; they’re both faculty members at Ohio State University.
I naturally first attempted to contact Thompson directly in order to clear up the inconsistency in his data. After two voicemail messages went unanswered, I e-mailed Thompson, together with most of his coauthors, on Jan. 23, Jan. 26, and Feb. 6, 2008, but received no reply from any of them, as Mann would be aware if he had read p. 368 of my comment. Thompson himself undoubtedly has a very busy schedule and might have been traveling. However, the odds that all five of the addressees were out of town simultaneously were very slim. Pending a reply, I proceeded to send a one-paragraph letter to PNAS, with a link to my full paper as Supplementary Information.
According to Mann,
… McCulloch was unable to marshal a credible eough argument to publish a comment in the original journal of record (in this case, Proceedings of the National Academy of Sciences [PNAS]. (p. 205)
In fact, as I noted on p. 368 of my published comment, my letter “was rejected on the grounds that PNAS does not publish corrections to articles more than three months old.” In other words, any errors published in PNAS that have gone unchallenged for more than three months become part of the scientific canon as far as PNAS is concerned! Two climate journals then did reject my comment on the reasonable grounds that they did not publish the original paper. In the end, Energy and Environment kindly agreed to publish it.
Mann curiously refers to me in his discussion of my E&E comment by the informal screen name I used in my popularized Climate Audit post, rather than by the more formal J. Huston McCulloch that appeared on my complete E&E paper. Even in citing my comment in his note 59, he gives only the initial H. rather than J.H. While it is commendable that Mann reads Climate Audit, this gives one the impression that perhaps he did not actually refer to the published version of my comment, but only to the CA post, which omitted the actual equations I estimated and which were the heart of my technical argument. This may be why he did not understand my procedure.
Mann remarks (p. 205),
[Thompson’s] tropical ice core data provided independent support for the conclusion that modern warming was unprecedented for the past two thousand years and were featured in An Inconvenient Truth, making them a particularly tasty target for deniers.
In fact, Mann should know that Thompson's ice core data did not actually appear in Gore's AIT: The series Gore identified as “Dr. Thompson’s Thermometer” was in fact Mann’s own Hockey Stick, spliced together with the recent instrumental record so as to make them appear to be a single series. See Al Gore and “Dr thompson’s Thermometer” #2 (CA 11/10/07). During the question period after an OSU seminar on Jan 11, 2008, Thompson, who had been an official Scientific Advisor on the AIT project, admitted this error. I then challenged Thompson to correct this error with a publicly accessible statement, but he still has not done so to the best of my knowledge. See “Gore Scientific “Adviser” says that he has “no responsibility” for AIT errors”.
It should be mentioned that even if AIT had presented actual Thompson ice core data, it would not truly have been independent of the Hockey Stick as claimed by Mann, since one of the 12 proxies used in the crucial AD1000-1400 segment of the Hockey Stick was Thompson’s Quelccaya d18O record. This record happens to be one of the strongest of these 12 contributors to the Hockey Stick shape, after the disputed treering-based PC1.
Furthermore, Thompson has never actually calibrated his ice core indices to temperature except by eyeball, so that there is no statistical basis for Mann’s claim that they provide evidence of warming. I have attempted to correct this deficiency in a working paper I discuss in my 12/10/09 CA post, Calibrating “Dr. Thompson’s Thermometer”. I conclude, however,
It may be seen from Fig. 4 that “Dr. Thompson’s Thermometer” is in fact completely uninformative about the existence or absence of a Medieval Warm Period (MWP), Al Gore to the contrary notwithstanding. Temperatures throughout the period 1000-1990 could have been as high as 1.2° C warmer than 1961-90 or as low as 1.8° C colder. The estimates for the 1990s are considerably tighter because of the highly significant slope coefficient for LCZ4, but even that decade has a 95% CI of (-0.32, 1.75) °C.
The Steig et al. Corrigendum Affair
Mann’s new book goes on to call renewed attention to another 2009 controversy: In January 2009, Eric Steig and 5 co-authors, including Dr. Mann himself, wrote an article in Nature on Antarctic warming that attracted considerable attention on Climate Audit. On Feb. 26, 2009, I wrote a post here, entitled “Steig 2009’s Non-correction for Serial Correlation”, which showed that Steig and coauthors had failed to correct the standard errors of their temperature trend lines for serial correlation, and that when this was done their results were greatly weakened, though not reversed.
Since Nature has a policy of publishing corrections and comments “if and only if the author provides compelling evidence that a major claim of the original paper was incorrect,” and this correction did not in itself overturn their key result, I did not submit my comment to Nature, and only published it informally on CA instead. On Feb. 28, I alerted Steig and all of his coauthors by email of my post, inviting them to participate in the discussion.
None of the authors ever replied or participated in the CA thread, but on Aug. 6, 2009, they published a “Corrigendum” in Nature making essentially the same point I had made several months before in my CA post. See The Steig Corrigendum for discussion. A graph there by Roman Mureika shows that the portion of the continent that shows significant warming is greatly reduced when the correction is made.
On Aug. 6, I e-mailed the editors of Nature, reminding them that according to their Editorial Policies, “Plagiarism is when an author attempts to pass off someone else’s work as his or her own,” and requesting that they retract the Steig et al. Corrigendum and replace it with the short e-mail I had sent them, which contained the URL of my CA post.
Steig then wrote to Nature editor Michael White that he was in the field in Antarctica and not receiving e-mail when I had written him, that he was unaware of my post, and that he had not read it. White accepted Steig’s explanation, so I withdrew my complaint. See “Steig Professes Ignorance”.
Mann now writes (n. 61, p. 334),
McCulloch complained that Steig had appropriated his own finding. Yet it is self-evident that Steig et al. were aware of the need for the autocorrelation correction, since the paper explicitly stated (albeit, it turns out, in error) that it had been made. Had McCulloch notified Steig of the error when he first discovered it, or had he submitted a formal comment to Nature identifying the error, he would have received credit and acknowledgement. He chose, however, to do neither of these things. [Emphasis added]
In fact, I had attempted to notify not only Steig, but all 5 of his co-authors, including Dr. Mann himself, of my CA post. Mann himself is off the hook, assuming he does not read CA, because I received an automated e-mail from his computer indicating he was out of town and unlikely to read any e-mails that were not resent after his return. Steig’s computer did send me an automatic e-mail indicating he was in Antarctica until mid-March with no regular e-mail access. However, the message gave no indication that he would not read messages on his return. I have no reason to believe that any of the other four authors — David Schneider, Scott Rutherford, Josefino Comiso, or Drew Shindell — did not receive my message. All were co-authors of the Corrigendum, and hence all were responsible for its content. Even if all overlooked my e-mail, it is difficult to believe that none of their colleagues mentioned the CA post to them.
Furthermore, I did not submit my correction to Nature because of their explicit policy that they would not publish a correction that did not materially alter the conclusions of the original article. Evidently Nature relaxed this rule for Dr. Mann and his colleagues, but I naturally would rather this publication in Nature, however brief, had appeared on my vita rather than theirs.
In the end, the autocorrelation issue turned out to be the least of the original paper’s problems: Ryan O’Donnell, Micholas Lewis, Steve McIntyre and Jeff Condon (J. Climate April 2011, 24:2099-2115) have shown that the main results of the paper are dependent on oversmoothing that results from retaining too few principal components of the satellite covariance matrix. They find Antarctic warming to be concentrated in the Peninsula rather than spread throught West Antarctica as in the Nature paper. Furthermore, average trends are less than half what Steig et al. found for the entire continent, East Antarctica, and West Antarctica, yet were much stronger than what they found in the Peninsula itself. See O’Donnell et al 2010 Refutes Steig et al 2009.
Although Mann has much to say in his new book, he neglects to make any mention at all, that I can find, of the O’Donnell et al. refutation of his well-publicized paper with Steig et al.
Update 4/29/12:
As noted in “Thompson gets new NSF grant”, Thompson has in the last 8 months archived decadal data for Guliya and two Puruogangri cores back to 0AD, along with some previously unarchived annual data for Dasuopu.
However, in order to replicate the 2000-year index in Figure 6 of the PNAS paper, it would still be necessary to have the Sajama and Husacaran decadal data that was used back to 0AD. In addition, Thompson’s decadal data on either Dasuopu or Dunde must go back at least to 450 AD, yet both have only been archived back to 1000 AD.
With the newly archived Puruogangri data, plus a privately circulated spreadsheet with decadal data for Sajama and Huascaran back to 1000AD, it may now be possible to replicate at least the last 1000 years of the PNAS 06 2000-year index, though I have not yet had time to attempt that.
Climate Audit 
189 Comments
Thanks, Dr. M, for all the details
They’re familiar to readers but scattered throughout
Amazing, that Mann would deny the emails
And phone calls, and throw accusations about
The Hilary write-up and Brandon’s review
Are quite nicely done, on misstatements and such
But this sort of info provided by you
Add technical background, and helps very much
===|==============/ Keith DeHavelle
Moderated? I’m lumped in with cads! ];-)
I would change the last “Add” here to “Adds”
===|==============/ Keith DeHavelle
Thanks for the explanation, Hu. Turns out that once again, if Mann’s lips are moving, he’s telling porkies …
AGW supporters keep wondering why their earnest appeals so often fall on deaf ears. As exemplified by this case, I say that the reason is they lie to people. It’s not all that complex. When you lie to people, they tend to disbelieve you.
The problem is exacerbated by the repeatedly demonstrated unwillingness of other climate scientists to “kick butts and take names”. Thomson is a serial non-archiver. Many of us have complained, see my open letter to Subra Suresh of NSF for one example among many.
But until the other climate scientists take notice and more importantly take action, this kind of malfeasance will continue. The only good news in the deal is that the alarmists keep shooting themselves in the foot … which at least keeps the public from believing them, plus it provides great spectacles of them hopping around on the other foot and swearing …
And that is good in itself, but is bad for science in general.
w.
Willis,
You said “The only good news in the deal is that the alarmists keep shooting themselves in the foot …”.
It is even better than that. Their foot is invariably in their own mouth when they pull the trigger!
I would love to see a time when a matter is simply settled as being ‘error (acknowledged/corrected)’ or ‘no error (complaint retracted)’
I just have the feeling that (for example) Michael Mann and Eric Steig are reading this and already have some sort of semantic or moving-pea argument that allows wiggle room for each side to think the other is prejudicially incorrect. It would be nice to see explicitly where those battle-lines are drawn.
You have my sympathies though on being called out on certain ‘house-keeping’ or ‘fraternity’ protocol with regard to discussing the material with the other authors or publishing comments. It seems you did indeed to a good-faith effort in this regard. I guess since it is they that are publishing and you that are responding they feel no obligation to extend such courtesies (for example, wondering if Mann contacted you with regard to all the points where he is discussing you in his book before he published it).
Hu:
Nicely done. Your restraint is admirable.
Sounds like another “Eats, shoots, and leaves” situation on the part of Dr. Mann.
Even given that Steig was in Antarctica and out of regular email access (is there irregular email?), surely he did eventually come back to civilization to co-author the corrigendum, and actually read his email. Though it’s more likely he filters anything from CA regulars so they go directly to the spam folder. Problem solved.
Jeff Alberts: Though it’s more likely he filters anything from CA regulars so they go directly to the spam folder. Problem solved.
Oh, I think the team hates reading anything from critical source like CA, but they do at least look with one eye, in order not to miss their biggest blunders, if only to react and write up something to defuse any potential time-bomb they planted before it blows up in their peer group. And McCulloch showed it here that they do react to critique written here.
“Even given that Steig was in Antarctica and out of regular email access (is there irregular email?)…”
I have been to Antarctica. Twice. Most recently in 2000. ‘Way back then, whippersnapper, we had regular, daily access to the internet and email at McMurdo. I was in a hut on the ice, several miles away from McMurdo, and we had a T1 drop to a hub. I had good, fast internet via ethernet.
I am NOT saying Stieg was lying. He could have been at a camp for days or weeks for his coring. It is, however, inconceivable that he did not regularly transport his cores back to McMurdo for analysis, recording, and storage. EACH TIME he was back at McMurdo he would have, most likely, stayed overnight for a shower and a hot meal. AND. AND. AND to check his email.
I wish I had been a regular reader when Hu posted his OC and Stieg’s responses.
The great thing about Mann is that you can always depend on him to get it wrong.
He doesn’t ‘get it wrong’, he lies because he’s a liar.
read the first sentence of his prologue. wrong
Indeed. This was the first thing I had noticed – and discussed in my post. There are now three different “gospels” regarding the events of Nov. 17: Schmidt’s, Jones’s and now Mann’s!
Sunday shocker: Michael Mann misrepresents … again
But then accuracy has rarely been a feature in any of Mann’s … uh … “reconstructions”, has it?!
If you follow his timeline as stated
One conclusion cannot be denied
(And access to RC is related)
Mann is FOIA on the side!
===|==============/ Keith DeHavelle
Funny. SteveMc and I were talking one day and he asked if I read Mann’s book.
I hadnt, so I downloaded it and wham. mistake in the first sentence. Either that or mann is FOIA. We had a good chuckle over that mistake. Now to some such a slip is small potatoes, but where is the dedication to proper scholarship.
SteveMo:
Are you going to write a review? I am concerned that so far there have been no substantive critical reviews on Amazon. I am waiting for my copy of the book from the local library.
I probably wont. I tend to dive completely into things and not come up for air. If I read Mann’s book completely I would have to dive into all the details that steveMc, Brandon, Amac, and others have more command of. too many other fish to fry
Brandon has only scratched the surface of the disinformation in Mann’s book. I’m weary of picking the same old spitballs off the wall.
Even on something as simple as my very first contact with him. Mann reiterates his lie that I asked for an excel spreadsheet, a lie that we rebutted by producing the emails. Even the CRU Climategaters recognized that Mann’s claim on this point was untrue, but never spoke out. Mann inexplicably repeated this lie to the Penn State Investigation Committee, which accepted his evidence without crosschecking. And, unchecked within the “community”, continued to repeat this lie in the book.
By only mentioning this example in this comment, I don’t imply that there are not numerous other misrepresentations in the book. There are so many that it’s hard to keep track of them.
But Brandon’s once over is a good place to start IMO.
Click to access manns_book_review.pdf
Thanks for the link. It is an interesting if somewhat long review. Have you found any other detailed reviews of the book?
Thanks, John. Although unsigned and on Hilary Ostrov’s server, the 15 page review is by Brian Shollenberger, as Hilary notes in her 3/4 post at http://hro001.wordpress.com/
Brian’s review discusses several inconsistencies in Mann’s book. He even devotes a page to the Steig et al affair, picking up on the points I make above. One thing he overlooks, however, was that Mann himself was one of the co-authors of the Steig et al paper and Corrigendum, and therefore co-responsible for their contents!
In the interest of truth in posting, Hu, this 15 page review was written by Brandon Shollenberger – not Brian 🙂
Thanks, Hilary. I stand corrigendumed.
That’s okay Hu McCulloch. I’ve had plenty of family members make the same mistake while talking to me, so you’re not doing anything new. I’ve never understood how it happens, but I’m used to it.
By the way, thanks for pointing out Mann was one of the co-authors. I completely forgot about that. I don’t think it changes much, but it is amusing. Also, thanks for posting about the Thompson 2006 issue. I paused when I read it in Mann’s book, but I decided I wasn’t familiar enough with the topic to write about it.
Of course, even if I had wanted to, it never would have made it into that document given the fact it requires discussing (somewhat) technical details. If I do write a follow-up document which covers technical issues, I’ll try to bring it up there.
Hilary — Brandon’s review makes several good points, and should at least be published as a stand-alone post on one of the blogs. Do you know what is plans are for it?
Re: Hu McCulloch (Mar 4 21:50),
I’m sure Anthony would allow on WUWT if asked.
Hu, I agree. I don’t know what Brandon’s plans are, but I have just sent him an E-mail … letting him know what I think they should be. Certainly it deserves more attention and wider circulation than my quiet little corner of the blogosphere is likely to garner 🙂
Judith Curry had indicated her approbation of his work on this, and I have suggested that he drop her a line.
Hu McCulloch, I don’t have any particular plans for it. Given the length of it, I don’t think it’d work as a blog post. Of course, some of the individual points could be excerpted for posting on a blog, as you suggest. However, I’m not a blogger, and I don’t have much interaction with those who are. Given that, I don’t know which points would be best to focus on, nor where they ought to be published.
That said, people are welcome to use it however they’d like, and they don’t need my permission. I’d be happy to talk to them if they’d like my input on things, but I’m perfectly fine with it if they don’t even notify me.
Anyway, my e-mail address is very obvious (I use gmail, and there’s just a dot separating my name), so if anyone would like to get in touch with me, it’s easy to do. I’d be happy to help however I can, though I’m not good with self-promotion.
Is there a word limit for Amazon reviews?
Hu:
I could not find a statement at Amazon saying that there is a limit on reviews – except that it has to be more than 20 words. Brandon’s review is roughly 5700 words and I just pasted into a review box – though I did not submit it – without hitting a limit. As “observer” I have done quite a few reviews of around 1000 words that have been generally pretty well received – including a shorter one of Donna’s book.
Brandon should submit it himself if the whole thing is submitted.
However, others should feel free to submit short reviews commending Brandon’s review with a link to Hilary’s copy of it.
Hu:
I believe they don’t allow live links from reviews.
Hu McCulloch, I don’t know of a word limit, but I’d lose all my formatting if I tried to just copy it to Amazon. Maybe I could add different formatting to make it legible, but the amount of effort would be significant. And as bernie1815 says, I don’t think they’d allow me to post a link to the document.
If not for that, I’d have already posted it there. Granted, since my review would be a one-star review, it would probably just get ignored as another mindless one-star review (there are quite a few of those).
I think your review would move rapidly to the top as the most helpful critical review and be read by anyone with an open mind.
If I just Google “Climate Audit Mann on Irreproducible Results”, this post is the first hit, so it is not really necessary to imbed a live link in a 1-paragraph Amazon review for people to find a blog copy of your review. 1 paragraph is about the attention span of most Amazon review readers.
WUWT would be an obvious choice — Anthony has a submission page at http://wattsupwiththat.com/submit-story/ . It would have to be reformatted as HTML, but only a few tags would be necessary — blockquote, i (for italics), b (for bold), and p (for end of paragraph). Probably someone could help you with it.
Speaking of Amazon, I should insert a mention there of this post!
I just added Review #92 to Mann’s book on Amazon, with an inanimate “go-search” link to this post for details. I give it 2 stars, since I give Mann credit for at least mentioning my article. And that makes it stand out from the other reviews, most of which are a bipolar 5 or 1.
On further thought, Brandon, I think you’re best just to leave your review as a PDF, since that’s what you’re comfortable with and it would be a lot of work to reformat it for HTML. A brief blog post would then link to the PDF, with an invitation to add comments on it. Since your review is not the sole topic of Hilary’s post, that mention is not adequate. A short post on WUWT would of course draw much more attention.
Please be sure to put your name on your review, so that people who download it will know who wrote it. A date would also be helpful, in case you revise it later.
Hu:
Your review is very polite and professorial. It will be interesting to see the reaction to your concise statement. There are some strange birds out there!
Hu:
So far so good – 8 out of 12. I am sure MM will make you the subject of a twitter or perhaps Scott Mandia will send his “minions” to stuff the ballot box. Your verified purchase is going to slow them down a bit.
A commenter on my Amazon review objected that it was too hard to search for “Climate Audit, Mann on Irreproducible Results”, so I tried just pasting the URL into a reply. It doesn’t come out as a hot link, but at least the URL can be copied directly into a browser.
This will simplify linking to Brandon’s review when he’s ready to turn it loose!
Hu:
Well done. You have made an impact.
Brandon,
Publish the review as a self published book. Talk to Tom Fuller, he can help you.
Lets drive the review of the book sales to the top of amazon.
$1.99
Haha.
A book? Even with a second half added on, I doubt it is going to break 40 pages. I can’t imagine publishing a book that short, much less charging people two bucks for it.
By the way, I find it mildly amusing I’ve been told to get in touch with several different people, yet I’ve no contact information for any of them. For example, I don’t know the e-mail address of you or Tom Fuller.
Moshersteven at
gmail
40 pages is fine.
I sent you an e-mail, and after looking around some, I think you might be right about 40 pages being enough. Of course, selling something as a book after making it freely available for download doesn’t seem practical.
But eh, I don’t know. I’ve never dealt with publishing.
steven mosher:
Pretty cheap for such good consultancy 🙂
make it into a wiki or blog.
or serialize it on WUWT. ask anthony for a mannday. A given day in the week. you have
5700 words: break it into 7 blog posts. run a section every Mannday ( monday is a good choice )
That makes it last 7 weeks or 8
Breaking it up would be a bad idea, since people couldn’t just go to it. My post is 2844 words, so Brandon’s review would be about twice as long, but I don’t think there’s any limit on blog length (other than house rules).
As for Bernie’s comment that they don’t allow live links from reviews, an inanimate link might be feasible, eg the name of the blog and part of the title of the post should make it easily Googleable.
Subtle is SteveMo, “Monday is a good choice”, hilarious!
Steven Mosher, I actually thought it’d be a great idea to make a web site using a wiki type format which would cover the various details of the hockey stick debate. It wouldn’t be open to just anyone to edit. Instead, a small amount of people would have editing privileges, and everyone else would just have the ability to comment on a discussion page for each article. That, plus a roadmap/outline seems like it could provide a useful source.
But it’d be lot of effort for just a book review, as would a blog. Not only would the amount of effort be more than it seems worth, the lack of new material would mean the site would stagnate. After all, there is only so much you can write about a single book.
I do like the idea of serializing it on a preexisting blog though. That solves the effort problem, as well as most of the “advertising” problem. I might need to talk to Anthony about that possibility (though I’d have to get his e-mail address).
But first, I want to try to get the second part finished. I’m hoping to have it done sometime this weekend, so it’s not like there’d be much of a delay.
cool.
Just work with Anthony or Charles. Send a mail to me and I’ll forward it. Ive already pointed Anthony
at it,
Brandon.
Didn’t Andrew Montford (Bishop Hill) just publish a short piece which could be purchased for about a dollar? Could be Kindle or PDF.
To wit: http://www.bishop-hill.net/conspiracy-in-green/
Brandon –
Bishop Hill has posted a link to your review.
http://www.bishop-hill.net/blog/2012/3/7/brandon-shollenbergers-review-of-the-hockey-stick-and-climat.html
The attack dogs are already starting to foam – Frank o’Dwyer seems to be the designated debunker of your post. Why can’t the “consensus” leave this issue alone? Don’t they realise that everything they say or do undermines their credibility further?
Do you want to see something funny? Check the Twitter accounts for him and Michael Mann. Frank o’Dwyer wants Mann to say I’m wrong about a point, but Mann refuses. Instead, he says:
Some of the #s on p 18 of #HSCW not quite right. EU target is 2C tot rel. to pre-indust (approx 450ppm). To be fixed next printing.
@fodwyer Wording is ambiguous, but should be 2C total warming/450ppm for consistency. Everything else in the book consistent on this point.
I disagree about the wording being “ambiguous” (it’s clear, just clearly wrong), but otherwise, it seems Mann and I agree on this point while Frank o’Dwyer disagrees!
it’s like his criticism of the way you handle the Wegman review – Mann gets off the hook of being dishonest, and maybe defamatory, only if other evidence comes along (and we all knlow that it exists) to demonstrate that Wegman was concealing things. Truly extraordianry and mind-expanding. Nick Stokes could learn a lesson from Frank O’Dwyer
Just in case they might want to read my post and/or reply here, I sent the following e-mail to Mann, Thompson, Steig, and such of their coauthors as I could find e-mail addresses for.
I immediately received back the following automated reply from Mann’s computer. I’ll try to remember to resend to him after the 19th!
Who in the world deletes all email that comes in while they are traveling? Does anyone know someone who does this? Wow.
He doesn’t. I’s just an excuse.
Mark
I have run into automated messages like that several times–mostly from Academics.
Mann is tweeting incessantly. See http://twitter.com/#!/michaelemann. It’s hard to believe that he isn’t receiving email.
Hu, did you consider the possibility that he’s placed your email on a “bad” list and that you’re getting a blow-off?
I don’t think it was anything personal — it’s essentially the same message I got back in 2009 when I tried to notify him of my CA comment on his 2009 article with Steig:
I notice from his tweets that he will be signing his book in Toronto 3/12. Maybe he’s reaching out to you, Steve! 😉
Wonder what would happen if Steve walked in with a copy of the book to sign, with a big old SEG.
Fly on the wall time.
Maybe he always keeps that message on his public email, and people he really wants to correspond with have a different address for him.
I can well imagine a Mannian ploy would be to keep the Team contact line clear by using another email address for insiders. As in, “private” email.
Yup. What are the chances that Gavin Schmidt and Mann correspond using the publicly posted institutional emails? Zero.
Jeff and Lucy —
That’s occurred to me, but if he chooses not to read his PSU e-mail account, there’s not much I can do about it.
In fact, given all the interest by Cuccinnelli et al in FOIing his university e-mails, he may have switched entirely over to a private gmail account or some such.
you can tweet directly to him
Why not have someone you know go and pin/tape a note to his door?
send a mail from heartlandinsider@hotmail and pretend you are peter Gleick ( who used gmail )
“Mann is tweeting incessantly.”
I think the correct phrase is “tweeting furiously”. 🙂
Is he on a computer or sending smoke signals?
The bigger question, in this day and age, who doesn’t check their email several times a day no matter if they’re traveling or not? I think it’s safe to assume Dr. Mann has a laptop with a wireless card, and probably a smart phone.
Didn’t we recently learn that Mann never deletes emails…even when asked to?
Ouch. That’s gonna leave a mark.. 🙂
[Snip — Reference to the recent Penn State football scandal is out of line – HM]
Typically, plagiarism can be asserted, even if the offending author did not know about the prior art. May not seem fair all of the time, but it is the generally accepted way, and is one of the reasons for copyright registration, in some contexts, in the first place.
In courts, it’s easier to assert copyright infringement, or at least, the damages are greater, if the offending author can be shown to have known of your work before they published it. I have some experience with copyright infringement – plagiarism issues. One as an expert witness in such a court case, and one which played on on the pages of IEEE Computer magazine.
The IEEE case may be relevant to your concern. In that case, I accused some authors whose work was featured in a cover article, of plagiarism. I cited about a dozen offending items between their article and a patent of mine. I contacted multiple editors up the chain of that journal with my complaint. I also was able to prove that the authors were aware of my patent. In the end, the authors published a mild retraction in that journal, acknowledging my priority on the issues that concerned me.
In that view, the Nature authors remain open to the charge of plagiarism, and I believe you still have a more than worthy case.
best of luck
I withdrew my complaint after White accepted Steig’s statement that he was unaware of my CA post, so the issue is dead so far as I am concerned.
It is only because Mann has now revived the issue in his book, claiming that I did not notify Steig et al., and could have published my point in Nature myself, that I mention it at all.
Understood. However, just as Mann took to the option to revive it, you also have the perogative to revive your complaint to the editors of Nature.
Whatever else could be said of hockey sticks and the like, I think it’s practically a crime to plagiarize. Your original contributions were not acknowledged where it is most important to do so, namely in Nature. It’s not too late, especially given the new context. Of course that’s totally your call.
There’s another plagiarism incident around the same time that I’d meant to document and more serious. A longstanding Climate Audit technique has been to illustrate proxy weights by plotting circles with area proportional to the weights. I’d observed that proxy weights could be extracted from the linear algebra, even for RegEM. The points had been revived in CA posts at this period concerning Steig and I reported that I’d tweaked the RegEM algorithm to extract the weights. The same technique was applied in Mann et al 2009 (Science); Mann’s language closely tracked the CA language. But more importantly the concept was identical to the concept written up at CA, both in the extraction of weights and the illustration in a graphic.
Re this one Steve, and Hu’s, it seems Nature and Science both need to soul search about plagiarism, and about climate.
==============
I noticed that proxy weights (PLS B coefficients) had been plotted in a CA-reminiscent manner in Figure S5 of the SI to Kinnard et al (Nature 2011), as well. Is this the same issue?
Climate science inspires Dilbert?
http://dilbert.com/2012-03-06/
🙂
That was a wonderful ‘toon, Hu. There’s also a very timely message for The Team in his blog entry http://dilbert.com/blog/entry/the_unaware/
He begins:
Phil <I do many of my [peer] reviews on travel. I have a feel for whether something is wrong – call it intuition. If analyses don’t seem right, look right or feel right, I say so.>> Jones, in particular, could learn a lot from Dilbert’s creator. 😉
Hilary: Well spotted. It is a wise person who checks to determine his or her own limits. I always recommend to young would be consultants Chris Argyris’ HBR article, Teaching Smart People How to Learn. Interestingly much of what Steve McIntyre advocates at CA with respect to open access to data and readiness to discuss findings also underpins Argyris’ notion of effective problem solving in organizations.
Mann – “He chose, however, to do neither of these things.”
I presume Mann never interviewed you before writing his book. A note to the publisher of this deliberate misrepresentation and his failure to confirm the matter with you is actionable and requires acknowledgement of misrepresentation and a commitment to removal of the words in any subsequent edition. And an apology.
Who would ever take Mann’s word at face value, or value for anything for that matter. Certainly no me. Rather: If he said it in public, I would assume it to be falsehood or at least misleading and diverting …
His revival almost rises to the level of libel given that due notice was given and acknowledged by Nature and Steig. He is quite hypocritical that Mann, isn’t he?
Mark
I wonder if maybe Nick Stokes has the time and desire to try to reproduce Thompson et al 06?
“These are the series of numbers that result from taking an original dataset, subtracting off the average, and dividing by the standard deviation’
But not if one has a non-linear or non-random sampling rate. If the sampling rate is different across time, as in ice cores, then the mean is not the sum of the values divided n.
True, but Thompson had already converted depth into age, somehow, and had tabulated his results in tidy decades before computing his index.
For a couple of his ice cores, notably Quelccaya, there are nice annual layers that enable him to read off age directly. But for the others, one one can only establish a few points (say by dust from known volcanic eruptions) and then interpolate on the assumption of a constant accumulation rate. Since others might want to go back and revisit these assumptions, that is one more reason why Thompson should archive his full data, not just by inferred age, but by actual depth (with his inferred age).
Because of exponential thinning, even small imprecision in the thinning parameter can lead to enormous errors in dates as one goes back. The problem in Thompson’s tropical cores is MUCH more severe than for Greenland or Antarctic cores – an issue that is unerstudied and underdiscussed. In part, because of Thompson’s refusal to archive critical data, some of it from 25 years ago.
Reblogged this on Watts Up With That? and commented:
The machinations of Mann continue, here we see a shining example of the “team ethic”.
UEA defends its record on cheating:
http://www.eveningnews24.co.uk/news/uea_defends_its_record_on_cheating_1_1227357
Either Mann is incapable of understanding that he lies, or he thinks he can bluff through the lies, or he’s taken leave of his senses.
Not my call, but in the light of the lies about you, Hu, I think this should be taken up again with Nature. We need to figure out which of the possibilities above apply. For his sake and for ours.
===================
What’s most bizarre of all is Mann’s attitude that anyone who criticises his technical work is a ‘global warming denier’. It comes across to the reader as idealogical rhetoric or possibly even delusional. Certainly, a red flag and a major turn off in the credibility stakes.
Mann states that Hu’s error was “an incorrect assumption that the amplitude of variation in a series of measurements must be constant in time” but what does this mean? It may relate to the strange (in my eyes) practice of building the time series in pieces or steps. Using this procedure, the weight given to a series could change in different epochs (intervals, centuries) rather than having a constant weighting over the entire period. This is bizarre in my mind because it means that a given location is more or less an indicator of global climate depending on the centuries in question. That is obviously why I only used series that existed over the entire interval. I am guessing this was done to take advantage of the fact that there are lots of series in recent centuries but fewer into the past, but I still feel uneasy about it.
Craig — That statement puzzled me as well. But I think now that he just meant that the standard deviation of the input series over the last 400 years in the publicly available Data Set 2 is not necessarily the same as that over the full 2000 years that Mann evidently has private access to.
In fact, I was aware that Thompson might have been averaging Z-scores computed from the hidden full 2000-year source data, and mentioned this in my article. In this case, my regression coefficients would not have been exactly 1/3 and 1/4 times the 400-year standard deviaitons in the two regions, but would at least have been some set of constants in this neighborhood.
The fact that no set of constants gives an exact fit to within rounding error demonstrates that the illustrative data in Data Set 2 is not a truncated version of the data that was really used to generate the 2000-year indices in Data Set 3.
Mann is tweeting today. Maybe you should send him a tweet and tell him to check his email.
He’s an unmanly fellow, Micky Mann.
Brilliant. I wish I could understand it.
Hi, Hu,
Do you still have copies of the emails you sent? It would make a stronger case if you published them, all in one place, as supporting docs for this article.
You have to “leak” the e-mails then look for a perpetrator.
RDCII — Per your request (e-mails redacted):
Date: Wed, 06 Feb 2008 11:17:56 -0500
From: Hu McCulloch
Subject: Your inconsistent PNAS data sets
To: [Lonnie Thompson, Ellen Mosley-Thompson, Henry Brecher, Mary Davis, Ping-Nan Lin]
X-Mailer: QUALCOMM Windows Eudora Version 7.0.1.0
X-MIMEDefang-Relay-89167b66339720c294cd81d33948afd6488b114f: 71.74.56.123
X-Spam-Score: 0.30 () [Tag at 4.50] HTML_MESSAGE,MAILTO_TO_SPAM_ADDR
X-CanItPRO-Stream: 11_tagonly_no_subject
X-Canit-Stats-ID: Bayes signature not available
X-Scanned-By: CanIt (www . roaringpenguin . com) on 128.146.216.131
X-Authentication-warning: defang17.it.ohio-state.edu: defang set sender to
using -f
Dear Dr. Thompson and co-authors,
I haven’t heard back from any of you yet on my e-mails of 1/26 or 1/23, copied below. However,
I think that I may have come across what might be the source of the inconsistency of your
2006 PNAS Data Set 3 Low Latitude Composite Z-Score (LCZ) index. and in particular
its Himalayan component, with the ice core data in PNAS Data Set 2 on which it
supposedly is based.
When the 5-year averages for 1600-1999 for Guliya, Dunde and Dasuopu reported
in PNAS Data Set 2 at http://www.pnas.org/cgi/content/full/0603900103/DC1 are aggregated
into 10-year averages as in the attached file, the numbers for Dasuopu agree with the 10-year
averages for it from your 2003 Climatic Change article archived at ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/trop/ , to within the .005 permissible
rounding error, with only a 1-year discrepancy of dating. For example, the -17.40 and
-17.94 that PNAS Data Set 2 gives for 1970-74 and 1975-79 average to 17.67, exactly
the figure your file dasuopu-d18o.txt gives for the decade whose top is at 1980. Since
all the decades match exactly to within rounding error, it looks like the dating on one
or the other series is just off by 1 year.
However, the figures for Guliya and Dunde may as well be for different glaciers.
PNAS Data Set 2 implies -14.98 for Guliya 1930-39, versus -13.46 for 1931-40 in the NCDC
version. PNAS implies -10.595 for Dunde1930-39, versus -9.43 for 1931-40 in NCDC,
etc, etc. These discrepancies are much too large to be due to a mere 1-year
offset.
I recognize that some of these sites have multiple cores, and may require judgmental
age-calibration, so that perhaps you changed your minds between 2003 and 2006 as to
how these sites should best be interpreted. However, if PNAS Data Set 2 was correct and up-to-date
at the time of publication, then Data Set 3 was either wrong or out-of-date at that time, and vice-versa.
Either way, your PNAS data are inconsistent.
I would appreciate it, therefore, if you could tell me which of these series for Guliya and Dunde is
your preferred series (as of your 2006 PNAS article), and if the PNAS versions are
preferred, what the correspondingly corrected version of PNAS Data Set 3 would be.
Although the Andean data is not nearly as inconsistent as the Himalayan data,
there are also discrepancies with it that far exceed rounding error, as I noted earlier.
These also should be corrected.
My ultimate goal is to calibrate your ice core data to instrumental temperature
along the lines of P.J. Brown’s article on “Multivariate Calibration” in the 1982
Proceedings of the Royal Statistical Society B, in order to see what, if anything,
your composite Z-score and/or the underlying series tell us about pre-instrumental
temperature, if only back to the 1600 starting date of PNAS Data Set 2. If you would
archive consistent semidecadal or even decadal data on the 7 sites, this could easily
be extended back to the early part of the first millennium, with standard errors that
of course would increase as the number of series declined.
In my e-mails below, I stated that if your composites were any fixed linear combination
of the component series, the coefficients could be backed out exactly, to within the
.005 rounding error in the composites, from the whole sample or any subsample. I should have
mentioned that the subsamples would of course have to have at least as many observations
as there are coefficients to be inferred for this to work. Thus, the Himalayan series should
match for any subsample of length 5 or greater (4 slopes plus 1 intercept), while the Andean
series should match for any subsample of length 4 or greater. The subsamples I used, of
length 9 or 10, should all give essentially the same coefficients, as well as a confirmation of the
precision of the fit.
Of course, if you used some kind of nonlinear aggregation of the cores to find your
composite series, the linear fit would not be exact. However, it should still be
similar for subperiods, and there should be no subperiod like 1800-1889 for which
there is no correlation at all between the composite and component series.
Sincerely yours,
Hu McCulloch
Date: Sat, 26 Jan 2008 14:59:23 -0500
To:[Lonnie Thompson, Ellen Mosley-Thompson, Mary Davis, Henry Brecher, Ping-Nan Lin]
From: Hu McCulloch
Subject: More on your 2003 and 2006 ice core composites
Dear Dr. Thompson and co-authors,
In my 1/17 e-mail to most of you and some other OSU colleagues, I
reported what appeared to be a significant correlation between the
7-core Low-Latitude (or “Tropical”) Composite Z-score series (LCZ)
from your 2006 PNAS article and decadal averages of the HadCRUT3
global instrumental temperature index (t = 3.37 with 13 DOF, p = .0051).
The similar 6-core composite series from your 2003 article has not been
publicly archived, but presumably would have a similar correlation.
However, in order for this t-statistic and the underlying standard error to
be valid, it is important for the two regional components of LCZ to have been
computed using a fixed rule that makes no use of the instrumental record,
either directly or indirectly. If the coefficients were calibrated to the instrumental
record, there might still be a significant relationship, but the method of calibration
would have to be taken into account in order to obtain valid standard errors and
confidence intervals.
In a further attempt to determine the nature of the coefficients that were used
to compute your LCZ, I have now tried looking separately at the Andean Composite
Z-Score (ACZ) series and Himalayan Composite Z-Score (HCZ) series given
in PNAS 2006 Data Set 3 alongside the 7-core LCZ series. Regressions using
these disaggregated series should give more precise estimates of the component weights
than were obtained in my e-mail of Jan. 23, copied below, using LCZ alone.
(I neglected to mention in that e-mail that I of course included a constant
in the regression.)
I found that LCZ = (ACZ + HCZ)/2 to within the possible .005 rounding error
on the two-decimal place LCZ series, so that these three series are
indeed computationally consistent, and were combined using a fixed
rule that does not depend on instrumental temperature. It is not clear why one
would ever want to average z-scores in this manner, but still it is a fixed rule that
does not depend on the instrumental data.
I then tried regressing HCZ and ACZ separately on decadal averages of
their respective d18O series for the 37 decades for which complete ice core data
is available in PNAS 2006 Data Set 2, namely the 1610s through 1970s. If these
composites are any fixed linear combination of the underlying d18O series,
these regressions should be an exact fit, to within the .005 rounding error in the
reported values of each, using data for the whole period or any subperiod thereof.
(Since the relative rounding error in each of the d18O series is smaller than
that in composite Z-score series, it should be relatively unimportant in determining
the regression errors.)
For the Andean composite ACZ, the coefficients, with their se’s in parentheses
and slope t-stats in square brackets are as follows:
ACZ on
(constant) 22.4351 (.4844)
Quelccaya .4830 (.0277) [17.44]
Huascaran .5414 (.0184) [29.47]
Sajama .2321 (.0146) [15.92]
s^2 .0106
s .1032
R^2 .9871
DW 2.0668
Since this should be an exact fit to within rounding error, the standard deviation of
the residuals should be .005 or less, as noted above. The actual value of s is 20 times this,
so that there is a very perceptible inconsistency here between Data Set 2 and Data Set 3
that remains to be explained.
Nevertheless, the fit is quite tight, and shows that Quelccaya and Huascaran received
approximately equal weights, while Sajama received about half that weight. The t-statistics
are all off-scale for significance. Dividing the entire 37-period sample into 4 subsamples of
10, 9, 9, and 9 decades each increases the standard errors and reduces the t-stats
as is to be expected, but gives essentially the same coefficients as above, so that
there is no evidence here of time-varying coefficients. However, it is still important to know whether
the conspicuously unequal weight on Sajama was determined with reference to its correlation (or lack
thereof) with the instrumental series, or if it was determined by a fixed (albeit irrelevant) rule
such as averaging z-scores.
Unfortunately, the results for the Himalayan composite HCZ were much less coherent. The results for the full sample of 37 decades are:
HCZ on
(constant) 11.8343 (1.4764)
Guliya .0201 (.0495) [0.41]
Puruogangri .2173 (.0660) [3.29]
Dunde .1137 (.0998) [1.14]
Dasuopu .3436 (.0498) [6.90]
s^2 .0840
s .2897
R^2 .7673
DW 1.7942
Here, the standard deviation of the residuals is almost 60 times the permissible
rounding error of .005. It is almost half the standard deviation of HCZ itself during
this 37-decade period (0.566).
Only two of the coefficients, on Puruogangri and Dasuopu, are significantly
different from 0. Guliya and Dunde are therefore doing essentially nothing to improve
the fit and may as well have 0 coefficients. On examining the four subsamples,
the coefficient on Puruogangri is insignificant except in the last period (1890s – 1970s),
so that these observations are what is driving its significance in the full sample.
In the third subsample (1800s – 1880s), none of the coefficients is individually significant,
nor are the 4 coefficients even collectively significant: the regression F-statistic for the hypothesis
that all four are 0 is F(4,4) = 1.313, with p = .399. Thus even Dasuopu is contributing to
the fit in only 3 out of 4 of the subperiods.
In summary, although HCZ does, with the exception of the 19th century, have some
relation to your reported values for Puruogangri and/or Dasuopu, if not to Guliya or Dunde,
it is not even clear what the coefficients are from which it was computed, let alone
how they were determined. Even the relatively coherent ACZseries has a residual that is
inexplicably 20 times larger than it could be if it had really been computed from the tablulated
Andean ice core series.
Am I somehow mistaken here, or is there a fundamental disconnect here between
the summary PNAS Data Set 3 and its supposed ice-core source, Data Set 2?
Sincerely yours,
Hu McCulloch
PS: The addresses ——-, ——,
——-, and —— listed for some of your
colleagues on on the Byrd Center website, and used in my 1/23 e-mail below, are no longer active.
Date: Wed, 23 Jan 2008 15:53:36 -0500
From: Hu McCulloch
Subject: Your 2003 and 2006 ice core composites
To: [Lonnie Thompson, Ellen Mosley-Thompson, Mary Davis, Henry Brecher, Ping-Nan Lin, Tracy Mashiotta, Keith Henderson, Keith Mountain]
X-Mailer: QUALCOMM Windows Eudora Version 7.1.0.9
X-MIMEDefang-Relay-89167b66339720c294cd81d33948afd6488b114f: 128.146.216.78
X-Spam-Score: 0.00 () [Tag at 4.50]
X-CanItPRO-Stream: 11_tagonly_no_subject
X-Canit-Stats-ID: Bayes signature not available
X-Scanned-By: CanIt (www . roaringpenguin . com) on 128.146.216.133
X-Authentication-warning: defang19.it.ohio-state.edu: defang set sender to
using -f
Dear Dr. Thompson and Co-authors,
In order to understand the composite ice core indices in your
Climatic Change (2003) and PNAS (2006) articles better, I tried
regressing the PNAS 7-core version made available in PNAS
data set #3 on decadal averages of the underlying dO18 series,
computed from the 5-year averages in PNAS data set #2. If the
composite index is any fixed linear combination of the 7 underlying
series, this regression should be an exact fit to within rounding error.
Since the 2006 tropical composite is simply the average of the
Andean and Himalayan composite z-scores, I assumed that
the two sub-composites would likewise have been the simple average
of the available component series, either raw or as z-scores.
In the former case each Andean series would have equal
weight, as would each Himalayan series. In the latter case the
weights would be similar though not exactly equal, since the standard
deviations are all near unity. (Standard deviations range from
0.7073 for Dunde up to 1.5922 for Sajama, at least for 1600-2000).
The fit for 1610 – 1979, when all 7 series are complete, was
indeed very tight, though not as tight as I would have expected.
The coefficients and their estimated standard errors were
Guliya .0234 (.0275)
Puruogangri .0724 (.0366)
Dunde .0526 (.0542)
Dasuopu .1626 (.0311)
Quelccaya .3356 (.0512)
Huascaran .2526 (.0294)
Sajama .0933 (.0222)
Two of the series, Guliya and Dunde have essentially 0
coefficients (t < 1). Quelccaya, Huascaran and Dasuopu
are getting far more weight than could be accounted for
by averaging z-scores with different standard deviations.
In fact, if you had averaged z-scores, Dunde with its smallest
s.d. would have gotten the biggest weight rather than
the second smallest.
I would appreciate it if one of you could tell me how you selected
these particular weights. In the Abstract of the 2003 Climatic Change paper
(with the similar but unarchived 6-core composite), you say that you
computed the composite and then compared it to the Mann et al 1999
reconstruction and the Jones et al 1999 instrumental series. Did you
in fact compute the weights by calibrating them to one or both of
these series? Is there a working paper or some such that explains the
calculation? Also, if the weights are not being computed by a fixed rule
such as simple averaging of raw data or z-scores, how were the weights
adjusted when fewer than 7 cores are complete, as in the 1980's (6 cores),
1990's (2 cores), and before 1000 (4-6 cores)?
A second puzzle is that if the composite was a fixed linear combination
of the underlying series, the standard error of the regression should have been
no larger than .01, the rounding error in the two decimal place composite index
as reported. Yet the actual standard deviation of the residuals was
0.154. The R^2 was high (0.9526), but should have been almost 1.
Was the PNAS composite in fact based on different versions of the
underlying series than those in PNAS data set #2? Or were the weights
not constant, even during 1610-1979? Or was the formula in fact nonlinear?
Again, any clarification would be appreciated.
— Hu McCulloch
PS: You might be interested in the correction to Craig Loehle's non-tree ring
multiproxy reconstruction that just came out in Energy and Environment, with
panel standard errors that I contributed as a co-author. See
http://www.econ.ohio-state.edu/jhm/AGW/Loehle for links to the paper and
supplementary info.
J. Huston McCulloch —-
Economics Dept. voice (—–
Ohio State Univ. FAX —–,
1945 N. High St.
Columbus, OH 43210
URL: http://econ.ohio-state.edu/jhm/jhm.html
Re: Hu McCulloch (Mar 5 00:10), one imagines that a fly on the wall would have had an interesting view of a flurry of emails in response to these, whose upshot was: do nothing.
Trying to put myself in the place of a climate scientist, I can think of two reasons why “do nothing” is the response of choice.
(i) they think the skeptics are engaging in pedantry. Skeptics know enough to spot a mistake, but not enough to make a meaningful contribution to the “debate”. (Regular readers of CA would dispute this.)
(ii) It doesn’t look good to the climate scientist’s peers to be corrected by an amateur. The fewer insiders who know that they’ve been corrected by an amateur, the better. Therefore, such corrections as are necessary are sourced anonymously, as are methodological improvements. (This notion is undermined if they all read CA, of course…)
@JIT: With respect, what do you actually mean by ‘amateurs’?
In the years I’ve been following the AGW argument through the blogs and books of Steve McIntyre, Anthony Watts and Andrew Montford, I have learned that there is no truly ‘professional’ ‘climate scientist’. What we have here are well-qualified – and in some cases, professional – statisticians. It is in the statistics that the arguments are won or lost: I prefer to stand with the gentlemen of SM and Hu et al, than the thugs of Mann and Schmidt.
Re: Snotrocket (Mar 5 08:44),
Non-climate scientists. They may have more statistical competence than the “pros”, but that’s a different matter – I’m talking about the attitude of “Team” players.
The British troops who were evacuated from Dunkirk, in June 1940 were professionals.
The British troops who landed in Normandy, in June 1944 were amateurs.
Hu:
It is hard for me to understand why your email would not have provoked some response from at least one of the authors. It certainly does not strike me as personally confrontational. Were you already on an “enemies list” at the time?
he would have been known to them as a Climate Audit contributor.
Well, it may have something to do with the fact that just 12 days earlier I had publicly suggested that Thompson issue a statement correcting the claim in Gore’s AIT (on which he was an official scientific advisor) that the graph identified as “Dr. Thompson’s Thermometer” and used to corroborate Mann’s HS graph was based on his ice core data. In fact, as he admitted verbally, it was just Mann’s HS itself, spliced together with the instrumental record as if they were a single series.
Then 2 days later, CA ran a post reporting my account of the event
at https://climateaudit.org/2008/01/13/sticking-thermometers-in-places-they-dont-belong/ (Note that Steve’s first impulse for a title remained in the automatically generated URL, even though he changed it in the published version.)
So it didn’t exactly surprise me that he didn’t return my voicemail messages. I thought he might feel more comfortable with an e-mail, and that his coauthors might pitch in on the details and/or compose a reply, but no such luck.
It is reported that Al Gore’s new climate science training program no longer makes any reference to the graph previously trumpeted as “Dr. Thompson’s Thermometer.” Yet, no corrections were ever issued by Thompson, Mann, or Gore. The whole matter is just being quietly dropped as though they hope no one (that they care about) will notice.
Meanwhile, I also noticed that Lonnie Thompson was implicated in the conveyance of poor information related to the IPCC’s “Himalayagate” scandal:
does Lonnie Thompson correct his own mistakes in the scientific record?
Anyone know if Thompson ever corrected himself in the scientific or public record?
Why is it that I never seem to see any reference to the Chinese Tibetan tree-ring data as recorded on Jo Nov’s blog in December last?
Something about story you tell sounds familiar to me Hu. Didn’t something similar once happen to Steve? Where he found an error and posted something here, and then Gavin/the team then acted like they found the error instead, and proceeded to attack Steve about it?
Somebody straighten me out if I’m on the wrong track here.
Steve: yes, Gavin plagiarized a Climate Audit observation about an error at the Harry station in Antarctic (unrelated to the harry readme), submitted this to British Antarctic Survey who scrubbed the record the next day (complicating unwary reconciliation). Gavin pretended that a Mystery Man had done this “independently”, but the BAS published a statement acknowledging Gavin, whereupon he had to confess that he was the secret Mystery Man. This was in February 2009 at almost the same time as Hu’s autocorrelation incident. It was an odd sort of plagiarism as the purpose was not to puff Gavin’s resume with a fairly trivial point, but to disparage Climate Audit.
Hu:
I’m generally about 80% sympathetic to you on all things and about 90% on Mann doing you a dirty, here. Certainly you play more fair than McI and at least put things into a form (papers) that they can be engaged with. and I can follow you (not a meandering snark full of mistake-laden graphs, written in mystery story style, to the extent I can’t even tell what the assertion is). In other words, you are much more straightforward and academic.
Few issues though:
1. I would avoid “plagiarism” as a loaded word and as not precisely correct. They failed to credit you for an idea. This is different than using prose from soemone else. It was a failure to acknowledge/cite. McI repeats the same error here again wrt “Harry”. I have pointed this out before…
2. It is over the top to call Odonnel et al a “refutation” of Steig. I would be more measured (for example, the paper is more measured). It is probably halfway between a total refutation and a minor improvement in the same vein (as alleged by some of the Steigers). Maybe significant upgrade or some such word (not sure). But definitely not “refutation”. Talking that way is kind of stooping into the Michael Mann (the error is all theirs) type of rhetoric, which is a little gamey debatey stuff. not real science/math logic/disaggregation.
3. I advise avoiding EnE for publications. Do Climate of the Past. Even a very crappy paper is basically exposed and published as a submission. Even if you don’t get accepted, it is still displayed. You will not get a good peer review at EnE (look at the fellow you helped who had to do a total rewrite of a paper…how did that thing sail out, when I could see all kinds of issues just from being a blog-reader).
If Thompson and Mann were working at CERN, neutrinos would still be faster than light.
These indirect exchanges with Mann provide no satisfaction because in my view you cannot address Mann, the scientist. It is invariably Mann, the advocate, that voices his opinions and views of reality. He spins this exchange with McCulloch down to McCulloch being completely wrong in his analysis, his accusing Thompson of malfeasance and being a denier of climate change.
Notice the difference between Mann’s and McCulloch’s approaches to these issues; Mann is all or nothing and black and white on the issue leaving nothing to be discussed as would be the case in normal scientific discourse, while McCulloch, in my mind, is searching for a more complete explanation the data and evidence that has appeared in a published document and he concedes that perhaps there are extenuating circumstances for the data manipulations and data that he has not been able to view. Does Mann deal with these very pertinent issues and in detail that are raised by McCulloch, and particularly the issue of making data available to the public? No, and I do not recall where he really ever has.
Of course, on the other hand we have Thompson et al. who do not respond at all. I find this behavior of Mann’s and Thompson et al. as being a puzzlement given the expected response of scientists to questions concerning their works. My only hint is that these responses, or lack thereof, come from advocates/scientists and therein lies the linkage. What puzzles and concerns me more is the how other climate scientists react to these “communication” problems.
I have recently been in contact with scientists who work with the major temperature data sets and I have received good and timely responses from these people. Some have even referred favorably to the citizen-scientist. I would not expect the less advocacy involved climate scientist to speak out on the personal choices of other climate scientists and how they might communicate. Perhaps we should make a case for those scientists where they do communicate in a manner expected of scientists.
Mann: What a [self snip]. What I find especially vexing is his “Oh, he should have said something” claim. Mann et al. *do* read CA, as is showed by the use of the informal name of McCulloch.
Mann: “Had McCulloch notified Steig of the error when he first discovered it, or had he submitted a formal comment to Nature identifying the error, he would have received credit and acknowledgement. He chose, however, to do neither of these things.”
And Mann could have written something along the lines “As Steig had not received any information about McCullach blog-post and McCulloch had not submitted a formal comment to Nature identifying the error, we were unaware of his work – otherwise we would have given him due credit and acknowledgement.”
Which would still have been borderline to being non-factual (as the team does seem to read CA) but it would at least avoid pinning the blame on McCulloch for not writing emails – something McCulloch claims to be able to show to be untrue.
What I see in another field of sciene is that when people make things up, it usually is the best (for the one perpetuating the fabrication/falsification, that is) when they say as little as possible, leaving as much as possible to interpretation – there are so many ways one can get entangled with the known or discoverable facts that is best not to hand out more rope to your “enemies”. For finding the truth on the other hand, it is good if people doing untruthful thing if people produce many claims.
I may not understand many of the more technical claims, but usually “Falsus in uno, falsus in omnibus” is a policy that works very well – if people claim simple things that can be shown to be false, then it is sufficient basis to doubt all claims by them. “Web of lies” comes to my mind.
Another thing: People who aren’t a tad bit self critical and do not at least try to see if they may have contributed to an situation (and Mann clearly does not think in that direction) are suspect to me.
Kaufman credited Hu in his corrigendum to the Arctic warming paper, and Steve only got the ‘and others’.
TCO/scientist is apparently on moderation here. He cross-posted some critical remarks at my blog that are civil and topical.
AMac said:
Posted Mar 5, 2012 at 1:54 PM
From your site, AMac, here’s a few of TCO’s “civil and topical” remarks from the page you linked to:
and
and
Meanwhile, “scientist” said civilly and topically:
and
AMac, if you find those comments “civil and topical”, I truly don’t know what to say, except to note that that’s the last time you’ll ever sucker me into reading your blog. I’ve mucked out cesspits that smelled better than that …
w.
Willis,
As far as tone and vocabulary, I judge that the text that TCO apparently submitted to Hu’s post is not complimentary, but it does qualify as civil. This is even more true if considered in the context of the sorts of commentary that routinely pass moderation at high-traffic climate blogs.
The paragraphs you quote above are from other comments by TCO, which are less polite. The (few) folks who comment at my blog are expressing their own opinions, not mine. I tend to offer more leeway to people who disagree with me.
More importantly, TCO brought up three substantive issues, about which reasonable people can agree or disagree.
I am somebody who has been blacklisted at RealClimate.org (still am, AFAIK). This despite or because of the fact that I’ve never submitted anything that’s off-topic or uncivil. One of the factors that makes constructive dialog about AGW so difficult is the readiness of some people to attempt to muzzle contrary points of view (if I thought this was a major issue here, I wouldn’t comment).
AMac, you say:
My apologies. All of the comments were from the very page that you cited, so I assumed you were referring to them.
Let me add that I, like you, am banned from a couple of the climate websites. Not banned at RealClimate, I just quit posting there after being routinely censored.
So I can only agree with you that muzzling contrary points of view is anathema to me. That said, there have been some people who I have been happy to see 86’ed from one or another site. Generally, it is because they are not there to either lead or follow, but only to muddy the waters and insult the participants. It’s a difficult call, and a judgement call, and one that I am very, very hesitant to make.
That said, if Anthony Watts or our good host here bounces someone out of the site, absent any evidence either way I’m going to guess that they had good reason for doing so … a blog is a kind of literary salon, a place to encourage interesting discussion. So when a guy walks in with hobnail boots and starts lashing out at any and everyone within range, yeah, sometimes I have no problem with escorting him to the door …
w.
TCO, the donkey, also has virtues
Steven Mosher (Mar 5, 2012 at 3:26 PM)
> TCO, the donkey, also has virtues
Nice use of commas! 🙂
Thx buddy
Mann (or his minions) seems to read CA:
‘Mann changed the online code within hours of this being posted at CA to the following ..’
‘***Correction (29 Nov 2008): clihybrid.m added to directory.’
“Although Mann has much to say in his new book, he neglects to make any mention at all, that I can find, of the O’Donnell et al. refutation of his well-publicized paper with Steig et al.”
Interestingly, while the Guardian is in overdrive promoting Micheal Mann the man and his book, they’ve also published a brief FAQ by Carbon Brief on warming and ice in Antarctica. The last link is to Steig et al ‘pre-corrigendumed’ 2009 paper. It is still the canon.
http://www.guardian.co.uk/environment/2012/mar/02/is-antarctica-getting-warmer
Why doesn’t someone just list the lies that Mann has allegedly made and call him out as a liar?
Publish it and invite Professor Mann to sue for libel.
A judical examination of the facts should shed some light.
AMac, I do not know whether you were around to know the history on TCO, but he has a personal vendetta against SteveM and anyone who might agree with SteveM. He rather thinks of himself as someone who while not having the background to understand all the technical issues involved in the blog discussions has some innate ability to make judgments on how they are discussed. He will to his credit ask lots of question but frequently comes up short in his understanding of the issues, then gets frustrated and can be rather personally abusive. His criticisms of blog participants or even authors of published papers are never to the point or detailed but rather generalized and personal. He often just reiterates, on going from blog to blog, in rather vague terms how he pointed out some error to someone long ago. I have been involved in a number of those discussions and often TCO’s view of things is tainted by his lack of understanding of the finer points of the discussion. If I had a blog I would ban very few people, but TCO would surely be one I would.
the “pick top 1%” mistake in the Wegman report.
Actually, I have been curious about the details of this. I’ve seen it charged several times, but I’ve never seen it explained, refuted, or even brushed off as ridiculous. Which certainly doesn’t mean that it hasn’t been. More likely that I don’t understand the issue well enough.
Still, I would appreciate a reference to a specific technical discusssion, if anyone can provide such.
“It is over the top to call Odonnel et al a “refutation” of Steig. I would be more measured (for example, the paper is more measured). It is probably halfway between a total refutation and a minor improvement in the same vein (as alleged by some of the Steigers). Maybe significant upgrade or some such word (not sure). But definitely not “refutation”. Talking that way is kind of stooping into the Michael Mann (the error is all theirs) type of rhetoric, which is a little gamey debatey stuff. not real science/math logic/disaggregation.”
TCO must have been in a snit here over the fact that O’Donnell’s paper was ever published. He is the one who was always pestering those critiquing others papers that they had to publish to be credible. He appeared to want to be acknowledged in some blog discussions as an elder statesman and one who could make clear the uttering of others. TCO claims to have a great success rate at publishing papers, yet when he posts on blogs one would have to guess that he has a very different style and language for technical writing. I have asked him to show us papers that he coauthored (since he refuses to identify who he is) and he would not do it. I strongly suspect that TCO is not really sure who TCO is.
Anyway anyone who has had the privilege of discussing the use of the satellite and surface station data and methods to estimate Antarctica temperature trends with Ryan O’Donnell, and in particularly reading the replies to Steig in defense of the O(10) paper and understands the technical matters would well know that a TCO could not even carry O’Donnell’s ipad. O’Donnell is a standout when it comes to explaining technical matters.
Next time you have a conversation with TCO at your blog, AMac, please ask him to explain the details of what it is he is contending and with what it is he specifically disagrees. Otherwise he sounds way too much like the know-it-all uncle at family gatherings who expounds in general terms what is the matter with the world but in terms so general that you cannot generate an intelligent discussion. He is usually put up with because he is part of the family, but in the end he is usually ignored.
O(10) showed irrefutably that the Steig methodology transferred warming from the Peninsula to West Antarctica. It also pointed to the problem of the satellite data being temporally unstable while at the same time being an innovative tool to use a spatial reference to be used back in time and space to fill in data for the sparse surface station data. O(10) provided some innovative and more objective criteria for selecting the numbers of PCs to use. O(10) will become an important test case for AR5 with regards to what the IPCC contends about Antarctica warming.
There was discussion on the blogs that went beyond what O(10) could get by reviewers to put in the paper and to stay within the stated context of the purpose of their paper. The trends that show significant warming in the Antarctica and particularly West Antarctica are very sensitive to start dates. Ice core data indicates strongly that the temperatures trends in Antarctica tend to be cyclical which makes trends sensitive to start and end dates. Also I think a case could be made for demonstrating that the satellite data was not consistent with UH satellite data nor ice cores and was susceptible to changes in instrumentation.
“O(10) showed irrefutably that the Steig methodology transferred warming from the Peninsula to West Antarctica.”
Let’s see. The Steig paper was thought important (and trumpeted around the world) precisely because it showed the warming was not confined to the peninsula, but was widespread across West Antarctica. That is what the cover on Nature was all aflutter about. It was also the only meaningful result from Steig, was it not? Was Steig important in pioneering new statistical techniques? Showing how to properly coordinate satellite and land data? Demonstrating the right way to use PC’s? Didn’t think so. No, Steig was important and remains important in some minds because it showed warming across West Antarctica.
Sounds like a refutation to me. Oh, sure, one doesn’t use those kinds of fighting words in a journal. And one has to be careful not to fall into the same trap of claiming more than one can support. I agree with all that. But the uncluttered story can be summarized as follows:
Steig(09): “In contrast to previous understanding, we’ve now proven that warming is happening across West Antarctica!”
O(10): “Uh, no.”
I should have better noted in my post above that the satellite data used by Steig(09) and O’Donnell(10) is the Advanced Very High Resolution Radiometer or AVHRR and the UH to which I referred above should have been the UAH satellite data. A good short summary of the differences between Steig (09) and O’Donnell (10) was presented at the Air Vent by Ryan O’Donnell and is linked below:
http://noconsensus.wordpress.com/2009/07/06/tav-to-realclimate
Brandon has unleashed the first 15 page installment of his review of Mann’s book over on WUWT:
Please discuss the details over there, in order to keep them all in one place.
Great job, Brandon! Please post a link to it on Amazon — although they don’t do live links, you can just paste the URL into a short review to refer readers to WUWT.
I posted something like this on WUWT. I post it also here because I think it is important to come up with a metric to assess the level of confidence on a prior temperature reconstruction, sorry for the double post. I am tired of battling back and forth on other blogs (politics.ie, etc) about: ‘the paper I linked to is better than the paper you linked to’, ‘my researcher is better than your researcher’, ‘there was no Medieval Warm period, yes there was!’. For the general public it has become impossible to differentiate one analyis or reconstruction from the other, other than by ‘faith’. The discussion has become a religious experience, with high priests and acolytes, and faith unvariable leads to repression of dissent. This is wrong, we are on a dwonwards slope to the middle ages if we keep going like this. We need an unbiased metric.
The problem I see with temperature recosntructions (be it tree rings, or othewr proxies) that if random brownian noise is used as input in the principal component component analysis one always gets a hockey stick. It is a characteristic signature of using uncorrelated random data (or really really bad temperature proxies).
I propose creating a metric and applying it to all temperature reconstructions that will give an indication of the ‘goodness of fit’, or if its really ‘garbage in/garbage out’. It is basically a measure of the expected deviation of all the runs used against the reconstructed average:
Take a running ‘window’ average over a ‘straight’ portion of the temperature reference, so that the local deviation against the local mean is calculated (for our temperature record one could simplify by detrending the record from 1940′s to 2000′s and calculating a deviation). Lets call that the expected deviation from a ‘true’ temperature record. Now, perform your temperature reconstruction with the proxies of your choice (tree rings, stalagmites, bore holes etc.) use the analysis of choice (principal component analysis, weighted average, etc) and reconstruct a mean value that extrapolates temperatures backwards. Then take the running window deviation method to calculate the deviation of all the weighted proxies runs from the mean and compare it to the one from the reference temperature series. The expectation would be that if the proxies were ‘perfect’, the running deviation would be the same or slightly larger than the temperature record. When the proxies give the same running deviation as the reference temperature lets give it a score of 1.0
Now take large number of gaussian random number runs. Integrate each series to create ‘red’ or brownian uncorrelated noise runs. Call them tree rings or whatever, scale them so they are in the appropriate range, and run them through the same exact analysis as the actual temperature proxies. The result will be the typical hockey stick where the stick portion is constant by being the average of random variation. The blade will match very well the temperature reference (of course, since principal component analysis gives a weighted average, weighing more those that most closely resemble the temperature record).
Then after creating a mean calculate the running window deviation of all the weighted ‘fake proxy’ runs. That deviation is the one expected from completely random, uncorrelated data, garbage. Lets give it a score of zero, 0.0
The running deviation of the real proxies is expected to be much lower than this, in fact it should be bracketed by it, and it it starts approaching the deviation of the random data it should be given a score of zero and discarded as not useful data.
Notice the the real proxies and the random data runs probably match the temperature reference record pretty well for the reference temperature time period and are not very distinguishable form one another. Where the difference lies, and where one can segregate the wheat from the caff is in the extrapolated portion of the data. As a rule of thumb the deviation of random garbage data should be 1/3 the range of temperature anomalies or 1/3(0.4- -0.2) ~ 0.2 degrees.
Any body with the skills is up for it?
How else could we determine if a hockey stick is true or bad data or bad analysis?
I think there is another component of uncertainty that can’t be captured, no matter how this proxy data is organized, processed, filtered, weighted, interpolated, extrapolated, degaussed, principal componentized, or whatever.
It’s the epistemic issue; knowing what we don’t know.
That goes for skeptics as well as those who promote certainty in CAGW. Over time, with accumulation of proxy information, and advancement of phenomenological models, we might be able to reduce epistemic uncertainty, but it seems to me that the collective body of climate change researchers (and skeptics) are a long way from that.
On a related issue, I get the impression that the modeling component of climate change science is still very decoupled from the data interpretation component. That’s why one doesn’t see much on paleo model reconstructions, yet, which in an ideal attempt to match proxy reconstructions, might help to inform future proxy reconstructions and the like. It’s a messy business, in any case, and I really doubt there will be any ‘one-size-fits-all’ template of data or model evaluation. Debates about how to filter data will always go with the territory.
To be fair to Steig et al they did use a novel concept by utilizing the spatial relationships of the satellite data to enhance the coverage, in effect, of the ground stations and do it back in time prior to the use of satellite measurements. The authors of O(10) have always given Steig credit for that part of the analysis. One large difference was that Steig used the satellite data temporally as well as spatially in the second half of the time period studied while O(10) used the satellite data limited to spatial relationships.
O(10) took the basic concept of S(09) and executed it in the manner that they judged was more correct. The exchanges between Steig, the reviewer of O(10), and the authors of O(10) was in my view very revealing of what the differences in approaches was between the papers and the amount of background work that went into the two papers.
Andrejs Vanags (Mar 12, 2012 at 9:46 AM)
> if random brownian noise is used as input in the principal component component analysis one always gets a hockey stick. It is a characteristic signature of using uncorrelated random data…
I think that assertion is incorrect. See this June 2011 discussion at Nick Stokes’ blog of the Wegman report’s analysis of McIntyre & McKittrick’s analysis of MBH’s hockey stick results.
Do you have a link or reference that supports what you’re claiming?
.
More generally, the idea of moving beyond “validating” candidate proxies simply by seeking correlation in the instrumental period has a lot of merit, I think. Some thoughts along this line came up during a discussion of RL Smith’s 2010 paper on reconstructions.
AMac, why do you think that assertion is correct? The discussion you link to doesn’t dispute it. All it does is show the visual impact of a figure was heightened by picking visually appealing series. It doesn’t have any bearing on Figure 2 of the MM GRL paper (reproduced as Wegman’s Figure 4.2) which makes the exact same point.
Admittedly, it’s a slight exaggeration to say “one always gets a hockey stick.” In truth, you get one more than 99% of the time, but not quite 100%.
That should obviously have been “incorrect,” not “correct.”
AMac:
I think I read it at The Air Vent. It seemed simple enough to check so I created hundreds of ‘proxy time series’ from random data and was able to recreate the hockey stick shape every time. I used Matlab. I dont know how to upload files here but if you use Matlab I could make it available to you.
I have proved the effect to my own satisfaction. The hockey sticks from the random data even have the slight slope downwards for the stick portion before going up on the blade.
What I lack and I was hoping for was for some on this blog or elsewhere to run the running window analysis on the Mann time series (which I dont have), they are likely bad analysis/bad data due to the flatness of the ‘stick’. One could also do it on the time series used other reconstructions which seem more believable as they have a ‘shape’ or a non-flat stick.
Andrejs Vanags (Mar 12, 2012 at 3:14 PM)
Brandon Shollenberger (Mar 12, 2012 at 3:50 PM)
At the linked post at Nick Stokes’ blog, he and CA commenter OneUniverse seemed to agree that the distinctive hockey stick shape that is shown in Wegman’s Fig. 4.4 is largely an artefact of the “de-centered PCA” that MBH used (AFAIK, that is not a procedure that the MBH authors ever justified using). Nick says that the traces that are shown in Fig. 4.4 are not representative of all traces, but are “selected” for their “hockey stick index”.
If centered PCA is used, and no hockey stick index selection step is included, then the traces that result do not seem to demonstrate a hockey-stick-like character.
The prior paragraph refers to the last figure in Nick’s post, preceded by the text,
“Finally, here is the corresponding randomly chosen centered version. There is essentially no HS tendency.”
In the comments thread, OneUniverse (largely aligned with SMcI) and Nick (largely aligned against SMcI) seem to be in accord on the technical aspects of their respective findings, and even their interpretations aren’t all that dissimilar.
Andrejs, you doing the work to create random series and re-creating hockey-stick shapes trumps my blog reading. Are your interpretations consistent with what Nick (and OneUniverse) reported?
I just realized Andrejs Vanags said principal component analysis (PCA), not decentered PCA. I guess I misread it because what he said is exactly true for decentered PCA, but it isn’t true for standard implementations of PCA.
Sorry for the mix-up. As Nick Stokes and oneuniverse both said, decentered PCA creates hockey sticks, and the way MM and Wegman selected which samples to display gave a stronger visual impact than a random sample would.
The problem is not in the decentering. It is in the mining.
If you take something like say the HadCRUT temperature record, it basically rises from beginning to end. So if you are looking for say 1,000 year proxies that correlate in recent years to the temperature record, even if you use regular (not decentered) PCA, it will still mine for hockeysticks.
w.
Brandon, we observed that Mannian short-centered PCA “mined” data sets for HS-shaped series. This is 100% correct. We illustrated the strength of the data mining tendency in a variety of ways, in several articles as properties of this highly inappropriate method emerged. In our GRL article, we stated that it nearly always produced a HS-shaped PC1: distribution of “hockey stick index” in Figure 2 of that article shows the distribution over all simulations from red noise modeled to emulate tree ring persistence according to the methods described in the article. And that the first eigenvalue was overweighted. We observed that “some of which [PC1s from red noise] bore a quite remarkable similarity to the actual MBH98 temperature reconstruction – as shown by the example in Figure 1.” By saying that “some” of them had a MBH-type appearance, we meant no more than that. We did not say that “all” of them looked precisely like the MBH-stick. We observed that 1-sigma (by our definition) stick occurred over 99% of the time, 1.5 sigma 21% and 2 sigma 0.2% of the time. The higher sigma sticks e.g. the top panel of Figure 1 occurred only some of the time. Our illustration was not chosen from a “random” draw because Mann had claimed that his reconstruction was “99% significant”.
Other defects of Mannian principal components were discussed in connection with distribution of RE statistics; in connection with “perverse” flipping of data to match the bristlecones (MM 2005-EE) and the contamination of an actual non-HS “signal” with even a single HS series (Reply to VZ) – the latter a very strong point that’s been overlooke and which we had not considered at the time of our first article.
I’ve never seen any one successfully contradict our observation that Mannian principal components is biased towards selection of HS-shaped series. Critics seldom (Mann – never) quote directly from our articles and typically paraphrase what we said and then criticize a different point.
In the case of the North American network, it’s not that the PC method “created” the stick. The stick is inherent in the bristlecones. Mann’s method overweighted the bristlecones and made him think that it was the “dominant” pattern in the data (as Mann said in his reply to MM2003) when, in fact, it was a secondary pattern limited to the bristlecones. As we eventually discovered, Mann knew that the pattern was limited to the bristlecones when he made this criticism of MM2003 as he had done an analysis without the bristlecones in his CENSORED directory though it was by no means easy figuring this out.
Steve McIntyre, I believe my view is completely in line with what you describe here, though I do have one addition to make. According to Mann’s new book, he knew that pattern was limited as you describe prior to him publishing his 1999 paper. That means he knew his conclusions depended entirely upon those bristlecones long before you said a word to him.
It’s more than strange to read his 1999 paper, and the IPCC report, with that in mind.
Brandon : “[..] the way MM and Wegman selected which samples to display gave a stronger visual impact than a random sample would.”
I agree with your criticism of the Wegman Report. Their undisclosed decision to sample from a selection sorted by HS index struck me as unfortunate and puzzling, since the biased nature of the MBH decentered PCA is well-established by multiple independent agents (MM, Huybers, Wegman, the NRC panel), and is surely easy enough to demonstrate visually.
I don’t see how the same criticism can apply to MM, though. The only relevant figure/result I’m aware of is fig.1 in MM05 (GRL), which plots a single simulated PC1, chosen for its resemblance to the MBH temperature reconstruction. The fact that the PC1 was chosen because of its resemblance is disclosed :
The figure could perhaps be improved, though : the quoted sentence above is a statement about the PC1s as a group, therefore fig.1 needs to be a visual representation of the group, instead of a single member of the group, if it is to demonstrate the truth of the statement (instead of only providing an example).
I see I wasn’t clear enough. Let me try again.
Suppose you generate 10,000 red-noise records of 1,000 year pseudoproxies. You select from among them all of those whose last 150 years are correlated to the HadCRUT 150 year temperature records.
If you then average those records together you get a hockeystick. Why? Because in all of the selected records the last 150 years trend up, but the rest is purely random. So when you average them, the first 850 years averages out to basically a straight line while the last 150 years rises until the end … in other words, a hockeystick.
Note that this happens with plain vanilla averaging. It gets worse, as Steve points out, when you start using a weighted average.
w.
Quite so.
This point has been noted (more or less independently) at several blogs including CA, Lubos Motl, Jeff Id, David Stockwell. I first commented on this at CA in connection with the Jacoby tree ring reconstruction in northern Canada (which, curiously, rears its head again in Mann’s book.)
Jacoby had selected 10 of 33 or series that they’d collected in Canada as the “most temperature sensitive” and then added in the very HS-shaped and idiosyncratic cedar series from Gaspe. If one simulates 33 red noise series of similar persistence, selects the 10 most temperature sensitive and then add in one cherry picked series from say another 33, the somewhat HS-shaped Jacoby series is more or less median in its HS-ness.
This is a different but related point to Mannian PCs.
The failure of specialists to concede these obvious points seems little more than wilful obtuseness.
Just a quick question in this festival of curve deconstruction:
What would your positions concerning anthropogenic global warming be if there really was a hockey stick?
To me it wouldn’t matter much, for simple reason that there were likely many natural mean global temp increases in the past that were at least as severe (over the equivalent century time scale) as the one proposed by Mann. I’d be obligated to prove it or find the proof. But given the known history of late Pleistocene, I think it’s a good bet that whatever hockey stick was used to support IPCC, it still fails to override the null hypothesis.
You don’t have to stop deconstructing on my account, because it is very interesting to follow your thinking, and I get the strong sense that in this case, there is no hockey stick. And I don’t mean to under-rate the importance of that revelation. Still, I’m just trying to develop some perspective.
Michael G. Wallace, I don’t think the hockey stick has much scientific relevance to the global warming issue. It got a lot of attention as a PR piece, and it still matters a lot as a PR piece, but that’s it. In other words, it matters for social and political reasons, not scientific ones.
Of course, if defenders of the hockey stick aren’t completely right (and they aren’t), then the hockey stick controversy casts a bad light on the global warming movement. If that movement uncritically accepted and promoted something which was completely without merit (they did), the speaks to their credibility. Problems with the hockey stick obviously can’t overturn radiative physics and things like that, but they can give reason to doubt the reliability of the “consensus.”
As a side note, like Steve McIntyre, I’ve heard some people say if the hockey stick isn’t real, global warming is worse than we thought. They believe if natural causes could cause as much warming as we’ve seen in modern times, climate sensitivity must be far higher than expected. If that’s true, “breaking” the hockey stick means things are worse than we thought, and we should all be grateful for McIntyre’s efforts for warning us of the severity of global warming!
The hockey stick’s protected
For several different reasons
The science should reject it
For obvious malfeasance
But core folks rally round
“Some ethics rule? We’ll bend it!”
In private, they have found
It bad — but still defend it
And it is that dichotomy
That shows that there is trouble
Good science, like it oughta be
Is falling into rubble
As for the implications
That if it’s broke, it’s worse
That’s only true if stations
Go through CRU first
And NOAA, HCN
And others, and “adjusted”
The work here shows again
That some of this is busted
The well-known US drop
From 30s to the 70s
That threatened every crop
And needed desperate remedies
Has been just wiped away
In processing by BEST
Mid-70s hot, they say:
Let’s put that to the test!
===|==============/ Keith DeHavelle
Temprchur litchurchur
Besame, besame smooch.
Waddya wan’ now?
==========
Michael G. Wallace (Mar 13, 2012 at 4:15 PM)
My layperson’s view is that it is important to understand the climate of the past few millenia, because this would provide useful context for the likely or possible warming that the Earth may experience in the next century or two.
If it was as warm as the current decade back in medieval times, that suggests that current temperatures are not so high as to represent unprecedented challenges to species, ecosystems, and biomes. Conversely, if it hasn’t been this hot since the Holocene Climate Optimum (9,000 BP to 5,000 BP), that suggests that this warm period is rather exceptional.
I’d make a similar argument concerning the rate of climate change. Species and ecosystems would likely be more stressed by warming that is more rapid than what has been experienced in the past.
Unfortunately, multi-century reconstructions with decadal resolution seem to be having a prolonged Will Rogers moment. “It isn’t what we don’t know that gives us trouble, it’s what we know that ain’t so.”
I don’t see much reason to trust the conclusions of mainstream-acclaimed paleoclimate studies. And there are many lines of evidence that suggest that such reconstructions should be viewed with great suspicion. For example, it ought to have been nigh-impossible for Brandon Shollenberger to find so much wrong with Prof Mann’s recent first-person account of his work in this field. If paleoclimate studies were a robust, fully-functional scientific discipline, colleagues would have rapidly corrected most errors. The worst of them wouldn’t be around to critique. But that is not the case, at present.
AMac, Your last paragraph is profound in its implications and every person working in the field of climate, as well as every person who supports their work uncritically, should be made aware of it.
You should be in the running for quote of the week over at Anthony’s.
Buy the book, it’s an amazing document.
========
Willis Eschenbach, perhaps I’m mistaken, but I don’t believe what you describe is a feature of PCA. My understanding is PCA, if centered over the entire record, shouldn’t (effectively) screen series based on their correlation to the temperature record.
Mind you, I don’t disagree about the effect you describe. It’s the basic problem with a number of different methods used by Mann and others. I just don’t remember ever seeing that it happens with centered PCA.
In other words, it is my understanding the decentering is what causes the mining.
Willis Eschenbach (Mar 12, 2012 at 10:27 PM)
This is a very helpful and concise explanation of the context of what MM05, Wegman, Nick, and OneUniverse were variously considering as the “selection” step (e.g. how the panels in Wegman Fig. 4.4 were chosen from the pool of all possible pseudoproxies).
It provides what, for me, has been the missing piece: why a hockey stick would tend to have a blade of +/- a century, as opposed to one of, say, 400 years or 10 years.
The answer, then, is that the post-Little Ice Age instrumental record “trains” the selection method to highlight those data series that… will tend to yield a hockey stick with a blade of about a century.
Other steps in the calculation process can add emphasis to this trend. One is the process of double division that Nick describes, which will produce a smoother blade. Another is the use of de-centered rather than centered PCA.
Steve: Amac, Willis is talking about a different tho related phenomenon. You have to take care before assuming that it’s the same thing as at issue in the earlier papers. This particular issue affects the papers with a “small” number of proxies where they know the proxies ahead of time and the picks are contaminated e.g. repetitive use of Yamal. The issues in MBH are different.
Mannian PCA was centered on 1902-1980 and therefore assigned higher PCA weights to series with high deltas between 1902-1980 and 1400-1901. It’s a really BAD method.
The next step in MBH in effect weighted series by their correlation to 20th century temperatures.
The net result was to heavily overweight the bristlecones. There are old posts at CA showing the weighted averages of each proxy class other than bristlecones and their sum. It’s nothing more than low-order low-amplitude red noise.
The only people that have trouble understanding these simple points are climate scientists – just as they are the only ones who have trouble understanding upside-down contaminated Tiljander sediments.
Steve McIntyre, thanks for your inline response to AMac. You say:
The next step in MBH in effect weighted series by their correlation to 20th century temperatures.
This was my understanding, hence my responses to Willis Eschenbach. Decentered PCA obviously mines for hockey sticks, but it’s a later step which checks correlation of series to the modern temperature record. That step is what Eschenbach was describing, not PCA.
But as he says, if you screen, or weight, by a pattern in one period, you’ll deflate the variance of the rest of the record. That will overemphasize whatever pattern you’re comparing to.
Willis — As Steve has pointed out, there are two ways to generate a spurious HS.
The method you discuss requires picking the series that best correlate with temperature over the last 100 years. This was the method described by David Stockwell in the Australian geosciences newsleter, and apparently used by Jacoby (search CA for “A few good trees”) and D’Arrigo (“If you want to make cherry pie you have to pick a few cherries”) for their NAm treeline series.
But the MBH short-centered and short-scaled PC approach discovered by M&M is much more subtle, since it doesn’t even look at the temperature series to get a HS PC1. This is sort of like a 1-dimensional Chladni pattern: with data whose correlations that decay geometrically with distance, the first few PCs tend to look like successive polynomials — PC1 is linear, PC2 is quadratic, etc., in the short-centering period. Outisde that period, there is not much pattern, and so they become flat in the “shaft”. But then when the same short-centering period is used for calibration, and T happens to be uptrending during the calibration period, PC1 will have a very close correlation with T, resulting in a HS reconstruction.
Of course, it also helped a lot that the MBH PC1 included the stripbark bristlecones, which already had a HS that was spurious for biological rather than statistical ones.
In order to get either type of statistical spurious HS, it is important that the series be highly serially correlated. This is loosely called “red noise” in a lot of this discussion, but apparently EE’s reserve this term as a synonym for Brownian motion, ie a random walk with 1/f^2 power spectrum, which is much more than is required. See Wiki “colors of noise” at http://en.wikipedia.org/wiki/Colors_of_noise . All that is required is that the series be AR(1) with persistence that lasts several decades, but this doesn’t even rank as “pink noise”, which the article says refers to 1/f long memory fractionally integrated noise (FI(.5)). On this scale, AR(1) is barely “peachy”!
Hu McCulloch :”But the MBH short-centered and short-scaled PC approach discovered by M&M is much more subtle, since it doesn’t even look at the temperature series to get a HS PC1. ”
Not only is the biased effect of the MBH algorithm subtle, it’s also highly sensitive to the presence of hockey-stick shapes, as noted by MM eg. in their reply to von Storch and Zorita. In the AR1 simulations I ran off the MM code, the MBH method is still tending to create hockey-stick shaped PC1s even when the AR1 coefficients approaches zero.
oneuniverse (Mar 13, 2012 at 5:17 PM)
Off-topic, perhaps, but it was very refreshing to belatedly encounter the June 2011 post at Nick Stokes’ blog (see this thread’s comment “Posted Mar 12, 2012 at 4:48 PM” for the link). Nick followed up on pseudoproxy analyses relating to the Wegman Report that you conducted and then reported on at Climate Audit; while critical, he was not dismissive or insulting. You entered his comment thread and contributed much valuable insight. In the end, the two of you were within hailing distance on the major issues, notwithstanding that many other informed parties of all persuasions were participating in the conversation. I learned a lot — even if not as much as I might have (ref. SMcI inline at “Posted Mar 13, 2012 at 9:02 AM”).
If only more back-and-forths could be conducted as you and Nick managed to do!
Thanks AMac, funnily enough my approach to the discussion was in part inspired by your calm and inclusive exploration of the scientific issue of the Tiljander proxies.
Thank you all for discussing this.
What I am looking for is an index 0 to 1.
0 would indicate a time reconstruction which is consistent or indistinguishable with one in which red noise was used as the input instead of the actual proxies (using the same data analysis as the in actual reconstruction)
1 would indicate one to be expected if the temperature proxies were ‘perfect’ (I take that to mean that the deviation between proxy runs will not exceed the local deviation of the single reference run) Of course 1 will not be achievable but it bounds the problem.
Then one can compare various reconstructions by various authors or methods (cones, boreholes, pollen, etc) and have a gauge as to how much to believe in it. The index is a measure of confidence that the reconstruction is not garbage. One could even go as far as apply it to each series from each author on the IPCC ‘spaghetti graph’ and make something useful out of it. My suspicion is that the Mann series will have very low score. Others, who knows? may fare better. That’s the whole point, we are clueless as to which reconstruction to pick.
Other wise it becomes a matter of which author has a better name recognition, which author’s reconstruction fits your world view or your predefined expected results (bad, bad)
P.S. I think you all are getting to wrapped up on the specifics of the analysis. ANY analysis that will do a weighted average where more weight is given to those runs that come close to matching the last 150 years or so of the temp record and less weight (or zero) to those that don’t will result in a hockey stick shape (I’m talking about random inputs here). You can give the weight of 1 or zero a priori just based on if its raising in the last 150 years and then do the average (like Willis Eisenbach points out), or base the weight on the deviation or other metric of the fit for the last 150, and then average all (the non matching runs will have so little weight they drop out). That’s why the methodology is not that important. That’s also why, to avoid this type of discussion the index or score for each reconstruction should be based using the same analysis procedure they used for the proxy data, that way one compares apples to apples. One type of analysis may be more sensitive or more robust to noise inputs.
http://www.nature.com/ngeo/journal/v5/n3/full/ngeo1394.html
“…Radiative forcing of that magnitude is expected to result in a climate cooling of about 2 °C (refs 2, 3, 4, 5). This effect, however, is largely absent from tree-ring reconstructions of temperature6, 7, 8, and is muted in reconstructions that employ a mix of tree-rings and other proxy data9, 10. This discrepancy has called into question the climate impact of the eruption2, 5, 11. Here we use a tree-growth model driven by simulated temperature variations to show that the discrepancy between expected and reconstructed temperatures is probably an artefact caused by a reduced sensitivity to cooling in trees that grow near the treeline.”
Seems like this totally unkown author M.Mann wants to shed some doubts on very well known past temperature reconstructions done with tree rings named “hockey stick” by science super hero M.Mann….
as usually, journalists will not see any connections to AGW, hey there it´s volcanoes not that devil molecule…
Any thoughts on the implications on T-reconstructions??
Reporter: Emma Alberici Australian Broadcasting Commission – Lateline 10.50PM March 15, 2012.. Thought you would/could be interested!
Climatologist and director of the Earth System Science Centre in Pennsylvania State University Michael Mann joins Lateline.
Transcript
EMMA ALBERICI, PRESENTER: The Climate Commission’s latest report says global average temperatures have continued to rise over the last decade. It’s all part of the research that started more than 20 years ago in the United States. The lead climate scientist in much of that work was Michael Mann. Mann says he’s the central object of attack in what some have characterised as the best funded, most carefully orchestrated assault on science the world has known.
Michael Mann joins us now from Washington. He’s the director of the Earth System Science Centre at Pennsylvania State University.
Michael Mann, thank you very much for being there for us.
MICHAEL MANN, CLIMATOLOGIST, PENN. STATE UNI.: Thanks. It’s a pleasure to be here.
EMMA ALBERICI: Much of the modern debate around climate change can be traced back to your 1998 graph, what was known as “the hockey stick”. Now, that was made famous in Al Gore’s movie An Inconvenient Truth. It showed temperatures dating back 1,000 years. In your book you concede that thermometer readings don’t reach much further back in time than a century, so how reliable is this graph as a measure of climate trends?
MICHAEL MANN: Well in fact it’s the fact that thermometer measurements don’t go further back than about 100 to 150 years that leads scientists to use what we call proxy data, indirect measures of how the climate change in the more distant past, like tree rings and corals and ice cores, to attempt to reconstruct how climate change farther back in time. And so back in the late 1990s, my co-authors and I published a graph that came to be known as “the hockey stick” which attempted to estimate temperatures, the average temperature of the Northern Hemisphere back a thousand years using these sorts of data and what it showed was that the recent warming is indeed unprecedented as far back as those estimates go, about 1,000 years. In the more than decade period since we first published our work, many other researchers using independent approaches, different types of data have actually come to the same conclusion and even extended that conclusion a little bit farther back in time. So we do now know that not only has the globe warmed by a little less than a degree Celsius over the past century, that warming does appear to be unusual in a longer-term context.
EMMA ALBERICI: Back in 1998 a television interviewer asked you if your research proved that humans were responsible for global warming. Your answer was that it was highly suggestive of that conclusion, but you wouldn’t go further than that back then. At what point were you finally convinced that that link did exist?
MICHAEL MANN: Right, so, you know, all we could conclude with our work was that the recent warming was unusual in a long-term context. That alone doesn’t mean it’s due to us. It’s due to fossil fuel emissions and rising greenhouse gas concentrations. That conclusion has actually been established by taking models, theoretical models of the climate and subjecting them both to natural factors like volcanic eruptions and changes and the behaviour of the sun and the human factors of increasing greenhouse gas concentrations and what those more recent studies show is that you can’t explain that anomalous recent warming from natural factors. We can only explain it when we include the effect of humans on the climate.
EMMA ALBERICI: Your critics refer to what’s known as the medieval warming period of around 1,000 years ago when there were no coal-fired power stations, no motor vehicles and other modern phenomena that could explain the temperature rises as you suggest, and yet the planet was going through an extended heat spell between that period of 11th and 14th centuries. How do you explain that?
MICHAEL MANN: Yeah, so both in our estimates and many other estimates of how temperatures have changed in the past, we do see about 1,000 years, a period of relative warmth. Not warmth that actually rivals the most recent decades. The most recent warming takes us outside of that range of the past warmth. But there was an interval of time about 1,000 years ago that was little warmer than the colder interval that we call the Little Ice Age which took place several centuries later. Now we can actually explain that period of moderate temperatures 1,000 years ago based on natural factors. A fairly high amount of solar activity, so the sun was a little bit brighter; there were relatively few volcanic eruptions, which are a cooling influence on the climate. So when we put those natural factors into the climate, we can actually explain that relatively warm medieval period. Now it turns out another element of that medieval period of climate is that certain regions like Europe appear to have been warm while other regions like parts of the tropical Pacific were cold. And it turns out a lot of that regional variation in temperature has to do with things like El Nino. And so it’s pretty complex when you start to look at past climate changes and the different factors that can explain them. But the bottom line is that the recent warming is unusual in at least 1,000 years and other studies using climate models tell us we can only explain it from human activity.
EMMA ALBERICI: Now you’ve just published a book called The Hockey Stick and the Climate Wars and I have to say it’s a book that reads much more like a thriller than a scientific textbook. You’ve had death threats and charges that you misappropriated funds. On one occasion you went to work and were greeted by the FBI. Tell us what happened there.
MICHAEL MANN: Oh, well, the FBI actually came in at my – when I reported to them the fact that I had received a letter, an envelope that had white powder in it. And initially I had assumed the worst, but the FBI sent it off to their lab, they checked it out, it turns out it was a false alarm. Nonetheless, as you allude to, I have been subject to all sorts of personal attacks, threats to my safety, my life, threats to my family, and it’s not just me, it’s dozens of climate scientists in the US, in Australia and many other regions of the world where our findings are finding that climate change is real and potentially poses a threat to civilisation if we don’t confront that challenge. That represents a threat to certain vested interests and they’ve tried hard to discredit the science, often by discrediting and intimidating the scientists. Unfortunately it’s not all that new a tactic. We saw the same thing back in the 1970s, 1980s with tobacco, with the tobacco industry trying to discredit research establishing adverse health impacts of their product. It’s an old tactic and it’s now being used to try to discredit climate science, mainly coming from vested interests who don’t want to see us shift away from our current reliance on fossil fuels because they – understandably, they profit greatly from our current addiction to fossil fuels.
EMMA ALBERICI: Who are these vested interest groups?
MICHAEL MANN: Well I actually talk about this in some detail in the book and I refer to some other books that have been written on this topic that actually trace much of the attacks against climate science and climate scientists to various organisations and front groups that derive most of their funding from the fossil fuel industry and what they often do is issue press releases attacking mainstream science. They publish – they have folks publish op.’ eds attacking climate scientists. They sort of create what some have called an echo chamber of climate change denial that permeates the airwaves and our media and it’s been a real challenge for scientists, for the scientific community to try to communicate the very real nature of the climate change threat in the face of this fairly massive disinformation campaign.
EMMA ALBERICI: Now recount for us the events on that most pivotal day in November, 2009.
MICHAEL MANN: Right. I recount this in the prologue of my book. I try to take readers back to this point in time where I had woken up to find out that thousands of emails had been stolen from a server in the UK. These were thousands of private email messages between various climate scientists, including myself. This all happened, I believe, not coincidentally in the lead-up to the Copenhagen Climate Summit of December, 2009. And what happened were these stolen emails were mined for words and phrases that could be taken outta context to try to make it sound like scientists were fudging the data, like – as if climate change was an elaborate hoax. And there was a massive …
EMMA ALBERICI: Can I just pick you up on that because of course it became known, as we know, as “Climate-gate” and it’s been picked over extensively. Much has turned on the word “trick” – Mike’s nature trick, which we assume you were the Mike that’s being referred to in that particular email from the University of East Anglia in the UK. In anyone’s language, trick implies some kind of deception.
MICHAEL MANN: Well, no, in fact in the lingo of scientists – and this is something that’s well known to those who work in science and math and mathematics – the term trick is used by scientists to denote a clever approach to solving a vexing problem and that’s well known within sort of scientific lingo, that that’s what that means. And there are many examples where words that have completely innocuous meanings that are used – were being used in these private correspondences between scientists who understand the lingo of science, were intentionally taken outta context to misrepresent what scientists were actually saying, and in that particular email what the scientist was referring to was a clever way of comparing two different data sets. And in fact the world’s most foremost journal, scientific journal, the most austere scientific journal in the world, Nature, issued an editorial about this and they were very specific about the fact that the use of the word trick was entirely harmless. It was denoting a legitimate approach to comparing two data sets and what the – that editorial really blasted those who have tried to distort the use of technical jargon.
EMMA ALBERICI: If I can just pick you up in that same passage where “Mike’s nature trick” was used, there was also the unfortunate term “hide the decline”, which many people have assumed meant hide the decline in temperatures when you were trying to advance a thesis that temperatures were rising.
MICHAEL MANN: Yeah, so it actually had nothing to do with my work at all. That “hide the decline” was referring to a specific tree ring study by a set of scientists in the UK. In fact they had originally published that study in 1998 in Nature and what their study was about was something that’s known as the divergence problem. Those particular tree ring data that they were working with tracked temperatures very well up through about 1960, and for reasons that scientists are still investigating, and it may have to do with pollution and other factors, those trees stopped tracking temperatures after 1960. And so in their original paper, the main emphasis of that paper was on this problem with those type of tree ring data, this so-called divergence problem, and they were very specific in that paper about how those tree ring data should not be used after 1960 because of that decline in the response to temperature. So in that particular email, Phil Jones was talking about how he didn’t want to show the bad part of that curve, the after-1960 part of that particular tree ring data set, because it’s misleading, because it’s well known that it doesn’t represent temperature after 1960.
And these are just two very good examples of a much more general phenomenon. It’s very easy to pick through personal emails, take words and phrases out of context and completely misrepresent what people are talking about. There have now been I think nine different investigations in the US and the UK and in every case the findings of those investigations have been that there was no impropriety at all on the part of the scientists, that instead what was involved here was a very unfortunate smear campaign against scientists to try to scuttle that all-important Copenhagen Climate Summit 2009 and the cost of that attempt to sorta sabotage any progress on dealing with the climate change threat is – it’s several years later and we haven’t made any progress and we’re committing ourselves to higher and higher greenhouse gas concentrations and more threatening climate changes in the future because we still haven’t acted on the problem.
EMMA ALBERICI: Now given your research shows a recent increase of almost one degree Celsius across the globe, a rise unprecedented, as you say, during at least the last thousand years, what do you think are the implications of your research for Australia in particular of doing nothing to stop carbon emissions that are linked to those rising temperatures?
MICHAEL MANN: Well, you know, Australia faces many of the same threats from climate change that we face here in the US: rising sea level and loss of coastal settlement and erosion of our coastlines. In the case of Australia, a very real threat to the health of one of the world’s great natural wonders, the Great Barrier Reef, increasing drought in certain regions and increasing flooding in other regions. What we’re seeing play out in Australia and in the US and around the world are the very scenarios that we predicted decades ago, that the models told us we would be seeing: increasing heat, more frequent extreme heat in all the major continents of the world and we’re seeing all of this play out. So, if you actually take the projections of future climate change under an assumption that we don’t do anything to deal with the problem, so-called business as usual, then what we see by the middle of this century – my colleague James Hanson has referred to as a scenario where the Earth will be a fundamentally different planet from the one that we grew up on, and it has deeply ethical consequences. The decisions we’re making today about our fossil fuel emissions are gonna determine the world that we leave our children and grandchildren and there isn’t a whole lotta time to act if we are going to avert some of the more damaging impacts of future climate change in Australia, in the US and in the rest of the world.
EMMA ALBERICI: Michael Mann, thank you very much for your time this evening.
MICHAEL MANN: Thank you.
Do you have a comment or a story idea? Get in touch with the Lateline team by clicking here.
“increasing drought in certain regions and increasing flooding in other regions.”
Huuu, i read the same kind of Bernum-Forer effect in the horoscope.
http://en.wikipedia.org/wiki/Forer_effect
On March 4, I sent Mann (as well as Lonnie Thompson and Eric Steig and their co-authors) an e-mail informing them of this post and inviting them to participate in the discussion.
I received an automatic reply from Mann’s computer stating that he would be away from his office until March 19, and that any email received before then was likely to be unread and discarded. Accordingly, I resent my e-mail today, but received the same message back, with March 19 replaced by March 26:
I’ll try to remember to resend on March 27.
One wonders if the algorithm for the auto-reply adds a pseudo-random number of days from the date of receipt, if the sender is not a known member of the Team.
Hu, in one of the Climategate emails, Mann suggests to the others that they set their emails to reject emails from the “usual suspects”, so that they give up.
Hu, Mann is on the Penn State Campus at this moment. He is to be at a book signing at the campus bookstore at 1:00PM.
But still ‘away from my email’. There’s more intentionality in that phrase that originally met the eye.
Well, I tried e-mailing Mann again this morning to inform him of this thread, and sure enough within 60 seconds I received a reply back that
However, I noticed that the automated replies were coming from an account identified with him at meteo.psu.edu, instead of from the @psu.edu address that appears on his webpage and that I had been trying to teach him by. So I tried the second address, and so far (4 hours later) still have not received an automated rejection.
So perhaps he does read, or at least eventually catch up on, the meteo account, if not the more public account.
Hu
It looks like you will have to send the communication by recorded delivery or some such snail-mail method.
I’m satisfied that I have done a reasonable job of extending him the courtesy of notifying him of my comment on his book, since there is no indication thatthe meteo account did not work.
It’s not unusual for celebrities to have unlisted phone numbers, so I’m that sure Penn State would go along with their superstar having an unlisted e-mail account known only to his colleagues and students, and perhaps changed periodically. But then his formerly active public e-mail address should connect to a message that it is no longer active, not one that perpetually says, in effect, “Sorry I missed you, please try again next week.”
I find it particularly disingenuous that he faults me in his book, as quoted above, for not notifying the Steig team (which included himself) of my autocorrelation post, when he himself apparently already had his fingers in his ears!
Mike’s E-mail Trick to give the run-around to critics, so to speak!
Mike’s trick to hide his hotline?
I suspect he follows CA regularly, so I’m sure your comments have been heard (even if you don’t receive the benefit of acknowledgement).
As far as a reply is concerned, I do believe I’ve seen a number of different places that he’s ‘blocked’ various scientists with these sorts of auto-replies (and perhaps even specific email addresses too). One perhaps can only imagine the kind of mail he’s going to be receiving from around the world at a publicly available email account. In that regard, having alternate email accounts seems reasonable.
On the other hand, I bet one day we’ll find out that in order for ‘scientific discourse’ to ‘reasonably occur’, universities have determined they will have to provide scientists with clandestine email accounts that are unpublished and unknown to anyone besides those that use them “because of the debilitating spam and volume of irrelevant email arising out of public access/misuse”. When/If that happens, I don’t see a university having to come forward with such email accounts until getting caught with them– but they’ll just change to new ones, and so it goes…
the last resort would be a politely worded complaint in writing to his employers to the effect that he had not accorded you reasonable academic courtesy, Hu. Obviously, it is entirely up to you whether you want to go down that route – I would not, simply because it is already now all over the internet that Michael Mann is afraid to answer his critics. And we all know that the Team read this site and WUWT assiduously, if only to gripe at them and claim intellectual superiority.
One wonders if the intellectual superiority meme – superiority over Hu McCulloch for example – is coming across as much as required to anyone reading this thread.
But please don’t answer that – leave me an automated message.
Well, this certainly seems to be the ‘Trick’ de jour for Mann and other members of The Cause. Claim he was never contacted when, in fact, he rejects or ingnores attempts to contact him.
Did he get his degree in a clown college?
I’m still a little irked at the Stieg explanation… he was in Antarctica with no REGULAR email. I’ve been there myself. All permanent bases have internet connectivity. If he went out to a camp for coring, he would still have opportunity to check his email often. By allowing this explanation to stand, I find Stieg guilty of a lie of omission.
sorta like Richard Drake suggests, He’s hiding the declines again.
I guess I’m a little oblivious. I had read those posts, then completely ignored them when I made my comment. Mea culpa.
OTOH, I don’t think it can be overstated that this tactic of evasion followed by accusations of deception by omission seems to be widely adopted. Stieg used it as a defense against allegations of plagarism. A defense a high school history teacher would not have accepted!
I have no doubt in my mind that Stieg, had he cared to, could have received and answered email and, AND accessed this website. Not daily, perhaps, but often enough.
Mann has established his own reputation with respect to ethics and the adherance to the Scientific Method. It’s just one more crooked arrow in his already full quiver of tricks.
Zits on climate science:
AGU blogger says Mann’s Hockey Stick extensively verified:
AGU blog by Dan Satterfield says this about the Hockey Stick today:
“…You may have heard about this image, it’s called the “Hockey Stick”, and let me say right here, that no matter what you may read on the internet, this image has stood the test of dozens of reviews and investigations. Every scientific peer group that has looked at it says it’s good science and if anyone tells you differently, they are giving you political propaganda.
To put it in plain words- the graph is correct, get over it.”
[his emphasis]
hmm I meant to post this info and link about Mann’s course utilizing Gore’s “An Inconvenient Truth” here on this thread, as it is the more recent one, but here is the CA link:
5 Trackbacks
[…] example of the “revisionist scholarship” that can be found in Mann’s opus: Mann on Irreproducible Results in Thompson (PNAS 2006) Share this:ShareEmailFacebookDiggRedditStumbleUponPrintTwitterLike this:LikeBe the first to like […]
[…] to Michael Mann’s book (Click here to see prior post)…I suggest reading this post (click here), where Hu McCulloch responds to the assertions made by Michael Mann. This is certainly an […]
[…] finally had time to read Hu McCulloch’s guest post at CA on Mann’s fancy new book. His reveiw shows more of that Mannian honesty we’ve all come to know an love. My reading of it brings up some of my favorite history at […]
[…] Mann on Irreproducible Results in Thompson (PNAS 2006) (climateaudit.org) GA_googleFillSlot("wpcom_sharethrough_viplite"); Rate this: Share this:TwitterFacebookStumbleUponRedditDiggEmailLike this:LikeBe the first to like this post. This entry was posted in AMO, Michael E. Mann and tagged AMO, Atlantic Multidecadal Oscillation, climate change, Geophysical Research Letters, JunkScience.com, Mann, Michael Mann, Richard Kerr. Bookmark the permalink. ← Climate Change Impacts In The USA are Already [NOT] Happening […]
[…] Under previous NSF grants, Thompson has collected a truly amazing amount of valuable data on ice cores. However, he has also been notoriously lax about providing definitive archived versions of his measurements. See, for example, IPCC and the Dunde Variations, Juckes, Yang, Thompson and PNAS: Guliya, Gleanings on Bona Churchill, and Mann on Irreproducible Results in Thompson (PNAS 2006). […]