Potential Academic Misconduct by the Euro Team

I raise here a troubling issue pertaining to potential academic misconduct by the Euro Hockey Team. They stated:

The code used by MM2005 [the GRL article] is not, at the time of writing, available, ….

First, this statement is obviously untrue. The code used in McIntyre and McKitrick 2005 is available at the SI to the article which is cited in the publication. It is at ftp://ftp.agu.org/apend/gl/2004GL021750. (The code for MM05 (EE) is at http://www.climate2003.com/scripts/MM05_EE/ee2005.backup.txt as is the code for MM03). The Wegman Report specifically notes that they examined availability of our source code and confirmed that it was functional. Previously, Huybers examined our code as did Wahl and Ammann, neither of whom required any assistance from me. Huybers annotated the code in his Supplementary Information. So there is no possible question that this particular statement is false.

Given the fact that I’ve emphasized the importance of source code to determine what authors actually do, this untrue observation about supposed unavailability of code disparages Ross and myself.

We are the only authors that they criticize on this count. They didn’t comment on the unavailability of data or code from co-authors Briffa, Esper and Moberg or on Mann’s code and data. Esper never responded to any requests for data. It practically took a federal case with Science to get any data from him and it is still incomplete. Esper failed to provide any coherent explanation of his methodology, let alone source code. Briffa has never provided source code. Brifffa has even failed to even identify the sites in Briffa et al 2001 (which have been used in 6 other studies.) At this point, despite all the publicity, Mann has still not archived the actual results for his AD1400 step through to 1980 and the Euro Team didn’t collect it.

Standard definitions of academic misconduct include the following definitions:

1. Fabrication is making up data or results and recording or reporting them.
2. Falsification is manipulating research materials, equipment, or processes, or changing or omitting data or results such that the research is not accurately represented in the research record.

It appears possible to me that that (a) their false allegation that the source code for McIntyre and McKitrick was unavailable at the time of writing; and/or (b) their failure to report that data and code from Briffa, Esper, Moberg, etc was unavailable at the time of writing, are, either jointly or severally, fabrication and/or falsification within standard definitions of academic misconduct.

"Honest error" or "difference of opinion" are defences to a charge of academic misconduct. However, the availability of source code is not a "difference of opinion" and, given their isolation of us for criticism, it is hard for me to picture how they could put forward a defence that the error was an "honest error" or one that could have been made with even the most cursory due diligence at the SI.

Is it plausible that none of the Euro Hockey Team co-authors were aware of this "inaccurate representation of the research record"? I list them below with their affiliations:

M. N. Juckes, British Atmospheric Data Centre, Rutherford Appleton Laboratory, UK
M. R. Allen, Atmospheric, Oceanic and Planetary Physics, University of Oxford, UK,
K. R. Briffa, Climatic Research Unit, University of East Anglia, UK,
J. Esper, Swiss Federal Research Institute for Forest, Snow and Landscape, Bern University, Switzerland
G. C. Hegerl, Dept. Earth and Ocean Sciences, Duke University, NC, USA,
A. Moberg, Department of Meteorology, Stockholm University, Sweden,
T. J. Osborn, Climatic Research Unit, University of East Anglia, UK
S. L. Weber, Royal Netherlands Meteorological Institute (KNMI), De Bilt, The Netherlands
E. Zorita,GKSS Research Centre, Geesthacht, Germany

I sent the following email today to Martin Juckes, copy to the other coauthors:

Dear Dr Juckes,

In your recent submission to Climate of the Past, now available for comment on the Internet, you make the following untrue and defamatory statement:

"The code used by MM2005 [the GRL article] is not, at the time of writing, available, …. "

As you either know or should know, the code used in McIntyre and McKitrick 2005 is available at the Supplementary Information to the article at ftp://ftp.agu.org/apend/gl/2004GL021750, as is made clear in the article itself. (The code for MM05 (EE) and MM03 (EE) are at http://www.climate2003.com/scripts/MM05_EE and http://www.climate2003.com/scripts/MM03 respectively). The Wegman Report specifically noted that they verified availability of our source code at the time of their report last summer. Previously, both Huybers and Wahl and Ammann had examined the source code, neither of whom required any assistance from me. Huybers annotated the code in his Supplementary Information.

Given our emphasis on making source code available, your purpose in making this untrue statement was to disparage us, especially when you isolated us for criticism while omitting to report the actual unavailability of data and code from authors in the field, including your coauthors.

I request that you withdraw this allegation from your submission to Climates of the Past on or before November 2, 2006 or I will take such other steps as I see fit.

Yours truly,

Stephen McIntyre

146 Comments

  1. Posted Oct 29, 2006 at 8:08 AM | Permalink

    I guess there are only so many lies you can take, Steve?

  2. S. Hales
    Posted Oct 29, 2006 at 8:18 AM | Permalink

    Steve,

    I am sorry that this continues to happen. I have seen professors dismissed and disciplined for plagiarism. It ruined their careers. To use the work product of another is one thing but to knowingly lie about another’s work product is beneath criticism. I hope that your “other steps” will include a libel suit. Best of luck. Steve.

  3. bender
    Posted Oct 29, 2006 at 8:58 AM | Permalink

    What about that little caveat “…at the time of writing”?:

    The code used by MM2005 [the GRL article] is not, at the time of writing, available …

    Presumably it is easy to show that the code was available before “the time of writing” (which was when?)?

    At any rate, the statement doesn’t appear to add anything to the substance of the paper, so I think it is fair to conclude it is a deliberate attempt to dicredit.

  4. KevinUK
    Posted Oct 29, 2006 at 9:01 AM | Permalink

    Steve,

    Under normal circumstances I wouldn’t recommend this course of action but these are not normal circumstamces. If I were you I wouldn’t just write to this guy but I’d also write to his boss

    “Dr Bryan Lawrence
    British Atmospheric Data Centre
    Space Science and Technology Department
    R25 – Room 2.118
    CCLRC Rutherford Appleton Laboratory
    Fermi Avenue
    Chilton, nr Didcot
    Oxfordshire
    OX11 0QX
    UK

    Tel: +44 (0)1235 44 50 12
    Fax: +44 (0)1235 44 63 14

    E-mail: B.N.Lawrence@rl.ac.uk

    Further contacts at BADC

    He has after all written this article as a representative of the BADC and so you have every right to point out this clear slur on you character to the head of this goverment funded organistion (which as a UK taxpayer concerned with how my taxes are spent I feel strongly you should do), so that it can be retracted and the record therefore set straight. Indeed if you would like me to send a letter also, then I am happy to, just let me know.

    Or perhaps yourself and I could also write to the Director of NCAS who fund BADC, Prof. Stephen Mobbs at

    “Environment Building
    School of Earth and Environment
    University of Leeds
    Leeds LS2 9JT
    United Kingdom

    Tel: +44 (0)113 343 5158
    Fax: +44(0)113 343 6716

    E-mail: stephen@env.leeds.ac.uk

    KevinUK

  5. jae
    Posted Oct 29, 2006 at 9:13 AM | Permalink

    Good grief, this is almost unbelievable!

  6. welikerocks
    Posted Oct 29, 2006 at 9:16 AM | Permalink

    What freakin lovely lot these climate scientists are turning out to be. Their perpetual guilt fest upon the character of mankind is not only affecting their brain cells but their idea of “this is a good thing to say” in their publications.

    And they can’t read maps , don’t know the statistics they use, they hide data and code, throw away the MWP , and now pretend they are clueless and don’t know how to use the internet apparently.

  7. per
    Posted Oct 29, 2006 at 11:57 AM | Permalink

    I suspect you might not get far by relying on “academic misconduct”, since it is difficult to prove, and institutions are normally reluctant to criticise their own.

    However, UK libel law is much more enthusiastic. The authors would be forced to prove the truth of their claims, or to argue that the implication is not defamatory. Institutions are profoundly reluctant to engage in libel actions, because they rightly fear getting stiffed with an enormous bill. Academic journals would likewise face liability for making public the liability.

    hmmm.
    per

  8. MarkR
    Posted Oct 29, 2006 at 12:06 PM | Permalink

    SteveM

    What they are saying is a libel and is legally acionable in the UK (not sure about other countries.

    If they continue after fair warning that makes the consequences for them even worse.

    Go get em.

  9. Steve McIntyre
    Posted Oct 29, 2006 at 12:10 PM | Permalink

    #7. Well, they would be unable to prove the truth of this particular claim, because it is false. It is also false in an obvious way that anyone can understand and verify for themselves.

  10. MarkR
    Posted Oct 29, 2006 at 12:17 PM | Permalink

    Ps I would contact the editors as well.

    From their website:

    General Obligations for Authors

    The following general guidelines for authors have been adopted from the “Guidelines to Publication of Geophysical Research” of the American Geophysical Union (revised version of October 2000):

    An author’s central obligation is to present a concise, accurate account of the research performed as well as an objective discussion of its significance.

    A paper should contain sufficient detail and references to public sources of information to permit the author’s peers to repeat the work.

    An author should cite those publications that have been influential in determining the nature of the reported work and that will guide the reader quickly to the earlier work that is essential for understanding the present investigation. Information obtained privately, as in conversation, correspondence, or discussion with third parties, should not be used or reported in the author’s work without explicit permission from the investigator with whom the information originated. Information obtained in the course of confidential services, such as refereeing manuscripts or grant applications, should be treated similarly.

    Fragmentation of research papers should be avoided. A scientist who has done extensive work on a system or group of related systems should organize publication so that each paper gives a complete account of a particular aspect of the general study.

    It is inappropriate for an author to submit manuscripts describing essentially the same research to more than one journal of primary publication.

    A criticism of a published paper may sometimes be justified; however, in no case is personal criticism considered to be appropriate.

    To protect the integrity of authorship, only persons who have significantly contributed to the research and paper preparation should be listed as authors. The corresponding author attests to the fact that any others named as authors have seen the final version of the paper and have agreed to its submission for publication. Deceased persons who meet the criterion for co-authorship should be included, with a footnote reporting date of death. No fictitious names should be listed as authors or co-authors. The author who submits a manuscript for publication accepts the responsibility of having included as co-authors all persons that are appropriate and none that are inappropriate.

    General Obligations for Authors

  11. GeoS
    Posted Oct 29, 2006 at 12:55 PM | Permalink

    Steve…How about starting a fighting fund? You’re going to need money. Where should we send our two bits?
    GeoS

  12. MarkR
    Posted Oct 29, 2006 at 1:00 PM | Permalink

    I think that in the UK those that contribute to a libel action become parties to the loss, if loss there is.

  13. Steve McIntyre
    Posted Oct 29, 2006 at 1:50 PM | Permalink

    I’m assuming that they will behave responsibly and promptly retract the remark. I would prefer that people didn’t indulge in legal speculations.

  14. geove
    Posted Oct 29, 2006 at 1:54 PM | Permalink

    Interesting read on the fraud and greed of Eric Poehlman a reseacher
    in the academic community by Jeneen Interlandi in the NY times

  15. John Lish
    Posted Oct 29, 2006 at 1:56 PM | Permalink

    #13 I hope so too Steve. The statement doesn’t put Martin Juckes in a good light whatever interpretation you put on the motivations for its inclusion. Withdrawing it seems to be the only sensible course of action.

  16. Demesure
    Posted Oct 29, 2006 at 4:24 PM | Permalink

    This is sickening case of stupidity and incompetence. Those guys don’t know how to retrieve a public link even my 10 year old son can do with a simple browser and they pretend to be reconstructing past climates ?
    Man (no pun), they would have been fired long ago had they work in the private sector !

  17. Jeff Weffer
    Posted Oct 29, 2006 at 5:00 PM | Permalink

    The Team sure likes their tree rings.

    There has got to be a reason why they keep digging this up and attempting to reconstruct the reconstruction of the reconstruction of MBH. It can’t just be because they want resurrect the Hockey Stick over and over and over again.

    Is it desperation? The global warming alarm is just not resonating with the public any more. They want to defend Mann so often and with so many papers that Mann gets his reputation back again? They like the fact that they can recycle old analysis of old tree ring data (little effort required) and they get a paper published?

    There just seems to be more to it than a simple, Mann was right. It seems more complicated than circling the wagons.

  18. bender
    Posted Oct 29, 2006 at 7:22 PM | Permalink

    Oh, come now. They honestly think they are reconstructing temperature. And to some extent, they are. It’s just that their reconstruction is not nearly as good as it needs to be to support their claims. To dismiss dendroclimatology altogether just makes one look silly. If you want to have any kind of credibility in this debate you can’t be arguing in the black or white. It’s more complicated than that.

  19. BradH
    Posted Oct 29, 2006 at 9:27 PM | Permalink

    I’m surprised that Eduardo Zorita has attached his name to such a statement. He seemed quite reasonable in his posts here.

  20. Pat Frank
    Posted Oct 29, 2006 at 11:16 PM | Permalink

    #18 — I tend to dismiss the possibility that anyone can get defensible temperature numbers, as opposed to qualitative judgments, from tree ring widths or densities. Using principle component analysis on tree rings to extract temperatures is like using a silk dress to extract manners from a pig. It wrecks the dress, angers the pig, and makes a mockery of good manners.

    It may be possible to use carbon kinetic isotope effects to get at a tree ring temperature, but I’ve not seen anyone thinking along those lines.

  21. Demesure
    Posted Oct 29, 2006 at 11:52 PM | Permalink

    Pat Frank wrote: “Using principle component analysis on tree rings to extract temperatures is like using a silk dress to extract manners from a pig. It wrecks the dress, angers the pig, and makes a mockery of good manners.”

    Lol, problem is these guys have made a living with it. You would want them to gore the pig.
    Something like “Using models to predict 2100 temperatures is like using a silk dress to extract manners from a pig. It wrecks the dress, angers the pig, and makes a mockery of good manners.”

  22. Posted Oct 30, 2006 at 6:55 AM | Permalink

    It may be possible to use carbon kinetic isotope effects to get at a tree ring temperature, but I’ve not seen anyone thinking along those lines

    Pat, can you please expand on that sentence?

  23. welikerocks
    Posted Oct 30, 2006 at 8:10 AM | Permalink

    Now it behooves me, of course, to tell you what they’re missing. … It’s a kind of scientific integrity, a principle of scientific thought that corresponds to a kind of utter honesty–a kind of leaning over backwards. For example, if you’re doing an experiment, you should report everything that you think might make it invalid–not only what you think is right about it: other causes that could possibly explain your results; and things you thought of that you’ve eliminated by some other experiment, and how they worked–to make sure the other fellow can tell they have been eliminated.
    … In summary, the idea is to try to give all of the information to help others to judge the value of your contribution; not just the information that leads to judgment in one particular direction or another.

    –Richard P. Feynman, “Surely You’re Joking, Mr. Feynman!” (1985)

    Table 1: Langmuir’s Symptoms of Pathological Science
    1.The maximum effect that is observed is produced by a causative agent of barely detectable intensity, and the magnitude of the effect is substantially independent of the intensity of the cause.

    2.The effect is of a magnitude that remains close to the limit of detectability or, many measurements are necessary because of the very low statistical significance of the results.

    3.There are claims of great accuracy.

    4.Fantastic theories contrary to experience are suggested.

    5.Criticisms are met by ad hoc excuses thought up on the spur of the moment.

    6.The ratio of supporters to critics rises up to somewhere near 50% and then falls gradually to oblivion.

    DOCTORAL COLLOQUIUM KEYNOTE ADDRESS
    CONDUCT, MISCONDUCT, AND CARGO CULT SCIENCE
    James R. Wilson
    Department of Industrial Engineering
    North Carolina State University
    Raleigh, North Carolina 27695, U.S.A.
    (June 23, 1997 — 6:00 P.M.)

    http://www.ise.ncsu.edu/jwilson/colloq.html

  24. Posted Oct 30, 2006 at 8:38 AM | Permalink

    When I think of scientific conduct I think of the medical field. When someone does an experiment or conducts a survey with opposite results than another they don’t seem to get so defensive. It seems to be accepted that there are inherent limitations to surveys, statistical methods, experimenter error and biases. Things are only accepted after a long time circling around and gathering support and then can still be overturned. Many other aspects of conduct in NIH that are built-in are missing from other sciences.

    As to misconduct, I remember famous case of an expert who was convicted I believe simply for giving testimony in court that cot-deaths were independent events. Apparently evidence shows cot-death is not independent in families. Imagine if the standard of assuming independence falsly were a convictable offence in climate science!

    I am sure there are many instances of medical science that are less than perfect. But its a relative thing.

  25. Paul Penrose
    Posted Oct 30, 2006 at 8:39 AM | Permalink

    Now let’s not make more out of this than there is. I’m sure the authors will quietly retract that one statement saying something like “Oops, sorry about that, old chap. It was just a silly mistake, that’s all. No harm done.” and move on. Of course, some harm will have been done because not everyone that read the paper will see the retraction, but that’s the way of it. Let’s get on with frying some of those bigger fish out there, eh?

  26. Posted Oct 30, 2006 at 8:47 AM | Permalink

    Dear Steve,

    it’s unfortunate when someone not only produces untrue statements but even gets them published. These people have apparently no respect for the truth.

    I don’t quite believe that they could be ignorant about the fact that your sources with Ross are among the most available ones in that field.

    This is almost certainly a deliberate lie meant to help a certain agenda.

    Don’t panic, enjoy your moral superiority, and have a nice week

    Lubos

  27. Louis Hissink
    Posted Oct 30, 2006 at 9:40 AM | Permalink

    Steve,

    Whoa

    Take a step back please.

    Beeeeeeeg period of silencio.

    OK,

    now do the big enchilada

  28. Steve McIntyre
    Posted Oct 30, 2006 at 9:51 AM | Permalink

    Huh??

  29. Nick
    Posted Oct 30, 2006 at 11:09 AM | Permalink

    #25, 26 and others, the paper hasn’t been published yet. As far as I can see, the draft is up for public discussion at the website of an online journal. Not sure why McIntyre preferes to complain about it here rather than go and point out the mistake at the journal. Look here: http://www.copernicus.org/EGU/cp/cpd/recent_papers.html

    Cheers,

    Nick.

  30. Steve Sadlov
    Posted Oct 30, 2006 at 11:18 AM | Permalink

    Not so fast with this talk of libel suits. Once the rubicon is crossed, you can be that this blog will be watched 24 X 7 X 365 and anything even borderline will be pounced upon. Don’t go there yet, would be my own advice. This can be handled without involving lawyers.

  31. Steve McIntyre
    Posted Oct 30, 2006 at 11:21 AM | Permalink

    #29. There are lots of points about principal components, bristlecones etc, about which reasonable people can differ and I’m prepared to debate those things. However, there’s nothing reasonable about making false statements with an attempt to disparage. It is the obligation of the authors to correct such statements, regardless of whether the editor chooses to get involved and I have requested that they do so immediately. I complained directly to the authors by email and that is what I expect them to respond to, not the discussion here.

  32. Steve McIntyre
    Posted Oct 30, 2006 at 11:24 AM | Permalink

    #30. Steve, as I mentioned before, I fully expect the authors to behave responsibly and retract this particular assertion and have asked people not to indulge in legal fancies or speculations.

  33. per
    Posted Oct 30, 2006 at 11:44 AM | Permalink

    I remember famous case of an expert who was convicted I believe simply for giving testimony in court that cot-deaths were independent events.

    If you are alluding to the case of professor sir roy meadow, he was initally found guilty of serious professional misconduct by the general medical council’s fitness to practice panel, i.e. he was not convicted. The decision was quashed by a UK court, and a recent appeal upheld that he should not be found guilty of serious professional misconduct. The recent appeal also established that experts should be held accountable by their relevant professional body for evidence that they give in court.

    yours
    per

  34. Nick
    Posted Oct 30, 2006 at 12:51 PM | Permalink

    #31, “there’s nothing reasonable about making false statements with an attempt to disparage” – McIntyre

    “This is almost certainly a deliberate lie meant to help a certain agenda” – Lubos in #26

    Pot. Kettle. Black.

    Cheers,

    Nick.

  35. fFreddy
    Posted Oct 30, 2006 at 12:56 PM | Permalink

    Nick, would you care to cite specifics ?

  36. Mark T
    Posted Oct 30, 2006 at 1:06 PM | Permalink

    Lots of things can have different meanings when taken out of context, Nick. Steve made it clear within the entire post (#31) what he meant. He has also complained directly to the authors, not just on the blog, so your previous comment (#29) is also without merit.

    Mark

  37. MarkR
    Posted Oct 30, 2006 at 1:59 PM | Permalink

    Hi SteveM

    Re #37 Someone is trying to get you into trouble with this post.
    Unless you have reason to believe the allegation is true, I should delete it.

  38. Posted Oct 30, 2006 at 2:02 PM | Permalink

    I’ve put the strange comment in the moderation queue for Steve to peruse.

  39. Steve McIntyre
    Posted Oct 30, 2006 at 2:05 PM | Permalink

    #37, 38. Thanks. I’ve purged the comment which has nothing to do with any topics here.

  40. MarkR
    Posted Oct 30, 2006 at 2:09 PM | Permalink

    SteveM

    I think dirty tricks are afoot. Do you have a disclaimer on your site to distance your good self from any possible third party malicious libellous posting?

  41. Steve McIntyre
    Posted Oct 30, 2006 at 2:14 PM | Permalink

    #41. Do you have some suggested disclaimer language?

  42. MarkR
    Posted Oct 30, 2006 at 2:22 PM | Permalink

    For example: “Opinions expressed on Climate Audit are those of the individual posters and do not necessarily reflect the opinions of Free Republic or its operators.”, or similar, most forums/blogs have a standard form of wording on them.

  43. MarkR
    Posted Oct 30, 2006 at 2:28 PM | Permalink

    Or “Messages posted on the Democratic Underground Discussion Forums are the opinions of the individuals who post them, and do not necessarily represent the opinions of Democratic Underground, LLC.”

  44. Paul Penrose
    Posted Oct 30, 2006 at 3:42 PM | Permalink

    Nick,
    If you don’t follow up with detailed accounts to back up your assertions, you will be guilty of the very thing you are accusing others of. This will, of course, render your credibility zero. Put up or shut-up.

  45. Nick
    Posted Oct 30, 2006 at 5:16 PM | Permalink

    #44 Paul,

    I am not making any assertions and I don’t care if posters on CA think I have credibility or not – why should anyone care about something so inconsequential? I was just making the point that McIntyre spouts off about people “making false statements on the web” whilst simultaneously running a web site in which he, his co-host and regular posters impute all sorts of underhand behaviour to pretty much anyone who works in climate science. Read the posts in this thread for a typical example. I particularly liked #8, where it was suggested that McIntyre sue for libel the authors of a draft paper opened for discussion on a science journal website, and #26 where we read that the authors of the paper are “propagating a deliberate lie meant to serve a particular agenda”. I wonder how long they sat together plotting and scheming to come up with a lie to serve their agenda? Do you think they had secret meetings where everyone had to come in disguise or just did it by email? Great stuff and very entertaining to read.

    More seriously, these sort of bizarre posts make McIntyre look ridiculous by association and make it easy to dismiss him as a crank with a persecution complex, whereas in fact he probably has a few interesting points to make.

    Cheers,

    Nick.

  46. bender
    Posted Oct 30, 2006 at 5:21 PM | Permalink

    “ridiculous by association” is a logical fallacy. Careful or Judith Curry is going to be on your case.

  47. Steve Sadlov
    Posted Oct 30, 2006 at 5:27 PM | Permalink

    Nick – if you are a climate scientist, or even an environmentalist, you are not winning friends or influencing people. Good day!

  48. Steve McIntyre
    Posted Oct 30, 2006 at 5:56 PM | Permalink

    Nick, I don’t take responsibility for the views expressed here other than my own.

    Where did I “spout off” about people making “false statements on the web”. In this case, I objected to a disparaging statement, easily shown to be false, made by a group of no fewer than 8 authors in the field – none of whom seemed equal to the challenge of identifying the error – in a formal submission to an academic journal after a lengthy period of preparation.

    In most cases, I do not speculate as to motives of climate scientists for doing what they do; however, in this case, there was an obvious intent to disparage me as a hypocrite for on the one hand insisting on provision of code and on the other hand supposedly not living up to my counsel.

    I asked the authors to withdraw the statement as it was untrue. I agree that I have made many critical comments about the quality of multiproxy studies. I believe such comments to be true and you are welcome to contradict any such comments.

    In this particular case, if the roles were reversed, and I had made a false statement about, for example, the availability of Esper’s data, I would correct the statement forthwith.

  49. Paul Penrose
    Posted Oct 30, 2006 at 6:17 PM | Permalink

    Nick,
    It should now be obvious to everyone that you are not interested in rational debate or polite discussion. You are here in an attempt to smear Steve McIntyre using logical fallacies such as guilt by association. You are now on my ignore list. I will no longer intentionally read any of your postings nor respond to any that I accidentially do read. I suggest others here do the same, however all are free to do as they see fit, of course.

  50. welikerocks
    Posted Oct 30, 2006 at 6:29 PM | Permalink

    Hey Nick,
    So if I see you at the Free Speech rally later, I’ll know you’re just there for the chicks. ;)

    Hang in there Steve McIntyre!

  51. john lichtenstein
    Posted Oct 30, 2006 at 8:00 PM | Permalink

    SteveM I wonder if it’s possible that this isn’t a thinko where the author is writing the exact opposite of what he means. I make that mistake sometimes. Maybe he is trying to say that it’s only the MM2005 article for which the code is available.

  52. bender
    Posted Oct 30, 2006 at 8:06 PM | Permalink

    #51 LOL

  53. john lichtenstein
    Posted Oct 30, 2006 at 10:35 PM | Permalink

    You laugh Bender but sometimes it takes a while for it to be funny.

  54. BradH
    Posted Oct 31, 2006 at 3:03 AM | Permalink

    RE:#51

    Well, John, the phrasing used by the article would seem to mitigate against that.

    The code used by MM2005 is not, at the time of writing, available, but the code fragments included in their text imply that their calculation used data which had been centred (mean removed) but had not been normalised to unit variance (standardised).

    [See note 9 at 1019 here.]

  55. BradH
    Posted Oct 31, 2006 at 3:06 AM | Permalink

    My bad – it’s lines 6-8 of 1019. Not note 9.

  56. per
    Posted Oct 31, 2006 at 4:15 AM | Permalink

    I particularly liked #8, where it was suggested that McIntyre sue for libel the authors of a draft paper opened for discussion on a science journal website…

    when you comment in a public forum, you have a responsibility for the words you use. I have seen previous examples of referees losing their temper and publishing amazing content on the “Climate of the Past” website, and I have to say that I do think that the open nature of this forum is a distinct disadvantage for the journal. If you think that laws of defamation do not operate just because you are a scientist, or just because it is a peer-reviewed journal, I have to tell you that I am certain that the journal will find out if this is true or not.

    yours
    per

  57. BradH
    Posted Oct 31, 2006 at 5:17 AM | Permalink

    I’ll put in my two cent’s worth here, as a lawyer.

    Steve’s approach – condemning the authors and requiring a retraction, while also questioning whether it could have been an honest mistake – is entirely appropriate and no less than one would expect in such circumstances.

    Talk of action for libel was not at Steve’s initiative, but an urging from posters here. Steve raised the issue of academic misconduct – not libel.

    For those who are urging promptly threatening or actually lodging such a claim, I can tell you that even the British courts would take quite a dim view of such an action.

    There are two aspects to actions for defamation (a joint term which includes both libel and slander). The first is to show that a defamation has occurred.

    Presuming the first is proven, the second element is to assess the appropriate award. Contrary to the belief of many, the award is not always monetary compensation. It’s often an undertaking to make a retraction.

    Even where monetary awards are recommended, they are not always large (I’ve read cases where judges have awarded $1.00 compensation). Many factors are considered, including the extent of the damage incurred (eg. reputation, future earning capacity, etc.) and how widespread the publication has been (“publication” can range from saying something to a couple of people in a bar, to worldwide, syndicated broadcasts).

    The judge will also often take into consideration whether or not the person defamed has sought a retraction and whether the defamer has, in fact, made such a retraction.

    So to all those champing at the bit for Steve to march into the Old Bailey demanding justice – cool it!

    He’s doing what’s appropriate and reasonable, in the circumstances. If he can’t obtain a retraction (and an apology, if that’s what he seeks), then and only then would he need to consider how important it is to him that the record be corrected and what he needs to do, in order to ensure it.

  58. Nick
    Posted Oct 31, 2006 at 5:50 AM | Permalink

    #48 Where did I “spout off” about people making “false statements on the web”? – McIntyre.

    You did it here. The paper is open for public discussion and comment as part of the review process. You could have simply posted up a comment at the journal’s discussion/review site, stating that the authors are in error and that the code used actually *is* available. Instead, you emailed the lead author saying that if he didn’t do what you want by Thursday you will “take such other steps as I see fit”. An unusual way to contribute to the review process of a scientific paper and one that sounds slightly threatening. On top of that you put up a topic here headed “Potential Academic Misconduct”. Nice use of the word “potential”. Personally, I’d call this spouting off.

    Cheers,

    Nick.

  59. welikerocks
    Posted Oct 31, 2006 at 7:32 AM | Permalink

    Nick, IMHO your concerns are simply unfair, lame and wimpy in the grand scope of things.

  60. Steve McIntyre
    Posted Oct 31, 2006 at 8:16 AM | Permalink

    #58. Nick, I’m used to processes involved in offering securities to the public. The obligation to make full, true and plain disclosure rests with the company, not the securities commission. I believe that responsiblity for publication accuracy lies with the authors, not with the journal review process. That’s why I approached the authors and did not engage the journal review process on this matter. It should be the obligation of the authors to deal with the matter, not the journal.

    In terms of your point, I’m not objecting here to “false statements on the web” which I take to include statements at usenet groups, blogs etc. Once again, I’m complaining about a false and disparaging statement submitted to an academic journal by a group of 8 prominent climate scientists familiar with the field, who were financed by a Dutch public agency.

  61. per
    Posted Oct 31, 2006 at 3:01 PM | Permalink

    On top of that you put up a topic here headed “Potential Academic Misconduct”

    well, here’s a couple of questions for you nick. What do you call it when someone publishes an untrue statement ? What do you call it when that statement is in an academic journal ?

    If the statement in question is easily verifiable as false, what do you call that ?

    I am interested in hearing your reply.
    per

  62. Steve Sadlov
    Posted Oct 31, 2006 at 4:41 PM | Permalink

    I am interested in in obtaining Eduardo Zorita’s statement about the subject matter of this thread.

  63. Ken Fritsch
    Posted Oct 31, 2006 at 5:22 PM | Permalink

    I am not making any assertions and I don’t care if posters on CA think I have credibility or not – why should anyone care about something so inconsequential? I was just making the point that McIntyre spouts off about people “making false statements on the web” whilst simultaneously running a web site in which he, his co-host and regular posters impute all sorts of underhand behaviour to pretty much anyone who works in climate science. Read the posts in this thread for a typical example.

    Easy, Nick, I am sure that you have your reasons for being upset by what Steve M has revealed in the Euro team paper submission, but if you were actually worried about spotting off why would you come on here and, well, spot off — or is there bad spotting off and let’s say well intended spotting off.

    I not only see problems with the obvious inaccuracy in the submission’s statement about code availability from MM, but with the general patronizing tone of references to MM papers in the submission. I agree with Ross M calling some of the paper submission’s reactions to the work of MM juvenile. Since your posts have lacked content about the issues raised in this thread, I cannot determine whether you have sufficient background information to make any judgments about the criticisms of the submission. For all I know you could be a disgruntled fan of the Euro team. If you have something to say content-wise I think now would be the time to engage.

  64. Pat Frank
    Posted Oct 31, 2006 at 8:49 PM | Permalink

    #22 — I’ve not had a chance to do that yet, John, but promise something. However, after some literature searching, I’ve found some papers that address a C-13 temperature correlation in tree rings. So, although I had not seen anyone addressing the idea, there have been people thinking about it nevertheless. I.e., I was wrong. :-) But I have to familiarize myself more with the literature before discussing this further. I’ll get back to you.

    However, in the course of the lit search, I did find this, which looks worth sharing:

    K. V. Kremenetski ea. 2004Medieval climate warming and aridity as indicated by multiproxy evidence from the Kola Peninsula, Russia” Palaeogeography, Palaeoclimatology, Palaeoecology 209, 113-125.

    Abstract: Data obtained from the low-elevation Khibiny Mountains (ca. 67–68°N; 33–34°E) on the Kola Peninsula, northwest Russia, indicate a period of exceptionally warm and dry conditions commenced at ca. AD 600 and was most pronounced between ca. AD 1000 and 1200. Warmer summer temperatures during this period (coeval with the “Medieval Warm Period” observed in other parts of Europe) are evident in a 100–140 m upward shift in the pine (Pinus sylvestris L.) limit in the Khibiny Mountains. On average, the cellulose of pine trees that grew between ca. AD 1000 and 1300 is enriched by àƒÅ½à‚ⲱ3C values of around 1″‚° compared to the modern trees from the region, further suggesting warmer summer climate than at present. The Medieval Warm Period was also accompanied by a steady decline in avalanche activity and the resulting formation of soils on the current avalanche cones in the Khibiny Mountains, suggesting lower winter precipitation and thinner snow cover. Lower precipitation is also evident by currently submerged tree stumps dating to the medieval period that indicate lower lake levels on the Kola Peninsula. In the middle of the peninsula at about AD 1000, the level of small closed-basin lakes was ~1 m lower than the modern time at some sites. Drier conditions may be attributable to decreased cyclonic activity. The medieval warm and dry episode was followed at ca. AD 1300 by the development of a colder climate with increased precipitation resulting in a decline in the alpine pine limits, increased avalanche activity, and higher lake levels. That phase corresponds to the modern aeolian episode reconstructed in subarctic Finland. Our results indicate that the Medieval Warm Period on the Kola Peninsula experienced notably warm and dry conditions. Hence, this period of warming extends to northwestern Russia as well as other parts of Europe.

    It appears that the Medieval Warm Period, which certain people seem interested in restricting to Europe, extended all the way across Asia to the Kola Peninsula. And, no doubt amazing to all, the MWP was warmer there than the modern climate.

  65. jae
    Posted Oct 31, 2006 at 9:06 PM | Permalink

    18, bender:

    To dismiss dendroclimatology altogether just makes one look silly. If you want to have any kind of credibility in this debate you can’t be arguing in the black or white. It’s more complicated than that.

    Color me silly. The dendroclimatologists have NOT made any credible case that their science applies to temperature proxies. You hedge at everything. Make a stand, man!

  66. Curt
    Posted Oct 31, 2006 at 9:08 PM | Permalink

    #64 Pat — The Kola Peninsula is in northwestern Russia, next to Finland. Perhaps you were thinking of the Kamchatka Peninsula on the Pacific? Regardless, it’s yet another piece of evidence for a significant MWP. Is anybody logging these in one place?

  67. bender
    Posted Oct 31, 2006 at 9:19 PM | Permalink

    Re #65
    I “hedge” because I understand the gray of the matter. Tree rings are a weak proxy for temperature. Put it in perspective. No one thinks they’re a five-star proxy. The debate is whether they’re 0, 1, or 2 stars. You’re a 0. I’m a 1. Rob Wilson & Eduardo Zorita are a 2. The Team is a whopping 3. A 4 is what we need to support the kinds of claims they’re trying to make.

    So I’m taking a stand. I’m standing on my “1”.

  68. Steve McIntyre
    Posted Oct 31, 2006 at 9:48 PM | Permalink

    #64. Pat a C13 series from Lipp et al was used in Crowley and Lowery 2000. See reference there.

    The medieval treelines in this area were further north than at present. They were also further at Yamal, which I’ve been meaning to post about.

  69. Dave Dardinger
    Posted Oct 31, 2006 at 9:51 PM | Permalink

    Curt,

    Have you looked at Co2 Science Magazine ? The Idso’s have started a compilation of evidence of the MWP worldwide.

  70. Dave Dardinger
    Posted Oct 31, 2006 at 10:01 PM | Permalink

    Actually that last link doesn’t seem to work anymore. Try this link.

  71. Pat Frank
    Posted Oct 31, 2006 at 11:58 PM | Permalink

    #66 — Oops, sorry and thanks. You’re right about my misplacement of Kola out to Kamchatka. Must have been those k’s. :-)

  72. Pat Frank
    Posted Nov 1, 2006 at 12:00 AM | Permalink

    #68 — Thanks, Steve, I’ll check it out.

  73. Steve McIntyre
    Posted Nov 1, 2006 at 8:06 AM | Permalink

    I’ve received a reply from Martin Juckes which I plan to respond to today.

  74. Steve Sadlov
    Posted Nov 1, 2006 at 10:59 AM | Permalink

    RE: #67 – they are likely a temperature proxy in the various Continental, Subarctic, and Arctic climate zones. In other climate zones, they are probably a moisture proxy if any.

  75. Steve Sadlov
    Posted Nov 1, 2006 at 11:24 AM | Permalink

    RE: #74 – But Bristlecones may fit in more with the non Continental (and colder) climate trees. So, in a way, even though I debated Dano about B’cone areas not being Mediterrannean (he insisted they were Mediterrannean Highland trees in California sites) I will concede that at least in California sites, they may well be moisture, not temperature, limited.

  76. bender
    Posted Nov 1, 2006 at 12:55 PM | Permalink

    Re #75
    Why don’t you give us your review of Salzer & Kipfmueller (2005)? They purport to do a reconstruction of both T and P from bcps.

  77. Steve McIntyre
    Posted Nov 1, 2006 at 3:18 PM | Permalink

    #64. It’s funny that you mention Kamchatka because there’s an article this week on proxies form Kamchatka at Climate of the Past http://www.copernicus.org/EGU/cp/cpd/2/1051/cpd-2-1051_p.pdf

  78. Steve McIntyre
    Posted Nov 1, 2006 at 11:31 PM | Permalink

    Martin Juckes wrote back saying that ” the citation at that point in our article should
    be to the Energy and Environment article (which we refer to as MM2005c) and
    not MM2005 as it stands. I’m sorry for this mistake….” I wrote back to him advising him that merely substituting an untrue statement about MM05(EE) code for an untrue statement about MM05 (GRL) code was unacceptable to me. He also pointed out that the script contained a reference to a function NHbeta, which wasn’t archived. (I’d archived a slightly later version NHgamma. I’ve uploaded the earlier version and edited the script slightly to ensure that it is operating in a turnkey basis. It’s at http://www.climate2003.com/scripts/MM05_EE.

    Here’s the full text of my response to Juckes and his coauthors:

    Your statement, as it stands, is untrue about source code for either article – MM05(GRL) or MM05(EE). Source code for MM05 (GRL) was archived in January 2005. Source code for MM05 (EE) was archived on March 1, 2006, shortly after your original inquiry, but long before the submission of your article to Climates of the Past. If you propose to merely substitute an untrue statement about MM05(EE) code for an untrue statement about MM05(GRL) code, that is unacceptable to me. Further, given that your allegation occurs in the context of a discussion of principal components — an issue treated in MM05 (GRL) to which your text currently refers, your proposed substitution is opportunistic as well as untrue.

    In response to your query about the function NHbeta, I have uploaded this function. I have also made a couple of slight edits to the backup script so that references to local directories are replaced by references to online directories to enhance the script so that it is as turnkey as possible. (I have continued to archive the original scripts.) In March 2006, I inadvertently uploaded a file containing the function NHgamma, a slight modification of NHbeta. I apologize for this oversight. However, the differences between NHgamma and NHbeta are immaterial to the issues at hand, as the differences only pertain to the handling of undocumented MBH re-scaling. These modifications were made in response to new information on MBH methodology, which became available during 2005.

    Your paper purports to carry out an intercomparison and evaluation of millennial paleoclimate reconstructions. To our knowledge, ours were the first studies in the field which provided source code so that results could be directly tracked and, to date, are the only studies in this field in which figures and statistical results can be replicated from an online script. (Wahl and Ammann, following our example, have provided considerable documentation, but fall well short of providing an equivalent script that generates their figures and statistics.) None of your coauthors have archived code for any of their articles; in some cases, they have even failed to archive their data as used or to provide accurate data citations for their articles and have refused or failed to provide this information upon request.

    If you wish to intercompare and evaluate availability of source code for paleoclimate studies, I strongly welcome the effort. However, any attempt to disparage our efforts in providing source code through untrue or misleading statements or any caviling about these efforts while omitting to discuss the record of your coauthors can only be interpreted as an intentional effort by you and your coauthors to provide a misleading representation of the research record.

    By objecting to this particular allegation, please do not consider that this constitutes acquiescence with any other statements in the article. I have made an immediate objection to this particular allegation because, given our emphasis on code availability, your false claims are both disparaging and defamatory and well outside the bounds of academic give-and-take. I intend to respond to scientific and statistical issues in other venues over the next few weeks.

    I re-iterate that I am not prepared to acquiesce in your presentation of untrue and defamatory statements about source code availability or to acquiesce in your presentation of a misleading representation of the research record on availability of paleoclimate source code. I re-iterate my previous notice requiring that you amend your submission so that it contains no such untrue and defamatory statements.

    Yours truly

    Stephen McIntyre

  79. Dave Dardinger
    Posted Nov 2, 2006 at 12:22 AM | Permalink

    re: #78

    What does Martin Jukes think he’s doing? The whole point of archiving code is so that people can see how it works and ask questions if necessary. Assuming they’d found a missing function the proper thing to do would have been to notify you at the point that happened, not now. As is, it sounds like, assuming they even examined the code at the time, they decided that a missing piece of code was too useful to have rectified.

  80. per
    Posted Nov 2, 2006 at 4:53 AM | Permalink

    what Dave Dardinger said !
    Even so, it appears somewhat difficult to reconcile, “The code … is not available”, with the idea that one small part of the code is wrong/ missing.

    it is difficult to avoid speculating on the intention behind such an error.
    per

  81. MarkR
    Posted Nov 2, 2006 at 6:59 AM | Permalink

    If they were genuine, they would’ve contacted SteveM, and asked him. That would have been the polite, and correct thing to do.

  82. bender
    Posted Nov 2, 2006 at 8:50 AM | Permalink

    Cute, Mr. Juckes, cute. We’ll remember that one.

  83. Steve McIntyre
    Posted Nov 2, 2006 at 9:17 AM | Permalink

    Here’s Juckes response today:

    Dear Mr McIntyre,

    Thank you for that additional information — I had suspected there was a close relation between NHbeta and NHgamma, but didn’t want to get drawn into speculation when a clear
    answer was just an email away.

    I notice that your website still contains the statement: "The computer script used to generate the figures and statistics in the E&E here will be located here [in a couple of days]" which may lead many people to think that the code is not yet available.

    I appreciate that you have disclosed more code than is normal in this and other fields of research. Our paper is an intercomparison of reconstructions: the statement about unavailability of your code was only there because there is uncertainty about the interpretation of your results which would be easily resolved if the code had been available. As many people would assume, on the basis of your many writings on this issue, that the code was available I thought it was worth pointing out that it was not (though we could of course have found it if we had happened upon the web page you have now disclosed to us).

    Returning to this code, I notice that that Figure 2 of your 2005 Energy and Environment paper describes a line which clearly has non-zero mean as being a pincipal component of centred data: this cannot be true because the principal components of centred data will all have zero mean. The code you have provided at http://www.climate2003.com/scripts/MM05_EE/ uses the "svd" function to generate the data for figure 2. As the "svd" function does not do any centering of the data, could you confirm that the description of the line in figure 2 of that paper is in error?

    I appreciate that providing software in the fully transparent form you advocate takes considerable efforts and appreciate the efforts you are making to move in this direction,

    sincerely,
    Martin Juckes

    One of the reasons to archive code upon publication is because you forget to do it afterwards. Here’s an example against myself. While I had archvied code for MM03 and MM05 (GRL) at the time of publication, unfortunately I didn’t archive the MM05b at the time of publication and then forgot to do it. Shame on me. Juckes inquired about the code in early 2006; I replied to him that I would archive the code forthwith and let him know. I got distracted by preparing for the NAS hearings in Feb 2006 and archived the code on March 1, 2006 and forgot to notify Juckes. In similar circumstances at some point between January 2006 and publishing an article, had I been in Juckes’ shoes, I would have refreshed the inquiry prior to submission especially when the respondent had indicated an intent to cooperate. However Juckes made no further inquiries. Worse, I spent several hours with coauthor Weber at KNMI (the study was financed in Holland) and Weber did not mention any issues in respect to the code. If Weber and Juckes choose to continue in this spirit, this will not be the last that she hears of this episode.

    A cursory inspection of the climate2003.com would have disclosed that the webpage was stale. For Juckes to rely on staleness of information there is pretty opportunistic. I’m pretty easy to locate at climateaudit and I think that even a climate scientist could have located me here.

    Juckes raises an issue with Figure 2 of MM05 (EE) Figure 2. The point made in the figure is valid, but, in retrospect, I’d preferred that we’d used a somewhat different illustration. MM05 (EE) Figure 2 attempted to illustrate a perverse feature of Mann’s method by which an arbitrary increase in non-BCP proxies in the period 1400-1450 resulted in a decrease in the Mannian "PC1". In my usual calculations with centered PCs, I used prcomp or princomp, which operate on the covariance or correlation matrix, but which yield series with a difference scale than svd applied to a data matrix from the Mann method. In this one Figure, in order to scale directly to Mann’s "PCs", I used svd on the data matrix as is pre-Mann de-centering (in which all series had a mean of 1000).

    In retrospect, I should have used prcomp for both calculations. I’ve re-done Figure 2 using prcomp, as shown below, and have added text to my archived backup script documenting this figure. The perverse property of MBH methodology is nicely demonstrated using prcomp.

    Adding values to the beginning of non-BCP series increases the early part of the PC1 using a conventional methodology; while adding value to the non-BCP series at the start results in a lower reconstruction under Mann’s methodology. What’s the over/under on Juckes criticizing this perverse property of the Mann method as opposed to him expressing outrage about our Figure 2?

    Re-Stated MM05b Figure 2 using prcomp. Top – effect of adding arbitrary amount to non-BCP proxies fropm 1400-145o. Top Effect on centered (covariance) PC1. Bottom – effect on Mannian "PC1". Increasing non-BCP values for 50 series reduces early 15th century values of the Mannian PC1.

  84. bender
    Posted Nov 2, 2006 at 9:43 AM | Permalink

    In similar circumstances at some point between January 2006 and publishing an article, had I been in Juckes’ shoes, I would have refreshed the inquiry prior to submission especially when the respondent had indicated an intent to cooperate. However Juckes made no further inquiries.

    Yes, that would have been the “collegial” (Curry’s admonition) thing to do. Tsk tsk. Glad to see he’s come to his senses.

  85. Steve Sadlov
    Posted Nov 2, 2006 at 9:48 AM | Permalink

    RE: #83 – You’ve been burned by dirty pool. Your adversaries are street fighters in academic garb. Assume they will always fight dirty and be anal retentive about everything. Too bad it had to come to this. They have made you stronger.

  86. fFreddy
    Posted Nov 2, 2006 at 10:42 AM | Permalink

    Oh, well, if Steve McIntyre is only 99.9 – 99.99 % right, then maybe we should allow the warmers to carry on being 0 – 5 % right …

  87. Mark H
    Posted Nov 2, 2006 at 10:50 AM | Permalink

    RE: 83

    1. “In similar circumstances at some point between January 2006 and publishing an article, had I been in Juckes’ shoes, I would have refreshed the inquiry prior to submission especially when the respondent had indicated an intent to cooperate. However Juckes made no further inquiries.” Some know this as ‘passive aggressive’ behavior, others as the typical ploy used in organizational politics wherein competing members ‘barely’ hold to the letter of the rule and wait for a minor miscommunications (and therefore the opportune gotcha’s) of their percieved opponents. Good managers/employees and reasonable people in a healthy organization follow the rule of ‘lavish communications’ to avoid such problems…not that climate science has such people.

    2. Aside from the blame game, what is the bottom line? Will the record be corrected?

  88. Steve McIntyre
    Posted Nov 2, 2006 at 11:25 AM | Permalink

    #87. Juckes has somewhat mis-stepped in this strategy by entering the fray with a manifestly untrue claim. Now that I’ve responded, he certainly doesn’t seem interested in a gracious correction but in finding whatever wart he can.

  89. Mark H
    Posted Nov 2, 2006 at 11:57 AM | Permalink

    RE: 88

    Well the bottom line is confessing that a serious error regarding M&M source code availability has been put in print by him (and he agrees in writing, claiming innocence), and that he will correct his statement. If he does not correct it then it destroys his claim of innocence and supports it a claim of malicious intent and damage with a lie – i.e. liable.

    In the UK, in particular, he is seriously exposed; he would have to prove each of his contentions.

    So why not write a polite but direct inquiry – will the record be corrected in the appropriate manner? Any attorney would tell you to do so.

  90. Mark H
    Posted Nov 2, 2006 at 11:59 AM | Permalink

    A PS – Any attorney would tell “Juckes” to do so, but it always seems that it requires an attorney to get someone to do the right thing…

  91. bender
    Posted Nov 2, 2006 at 12:01 PM | Permalink

    Re #88 You mean he’s NOT going to retract the statement?! That would be rather unwise on his part.

  92. jae
    Posted Nov 2, 2006 at 12:08 PM | Permalink

    Of course he will correct it; it would be unbelievably foolish not to do so.

  93. Barney Frank
    Posted Nov 2, 2006 at 2:35 PM | Permalink

    If he does not correct it then it destroys his claim of innocence and supports it a claim of malicious intent and damage with a lie – i.e. liable.

    Probably just a typo but the term is “libel”.
    Now, if he actually commited libel he could be liable for damages.

  94. Steve McIntyre
    Posted Nov 2, 2006 at 2:58 PM | Permalink

    Here’s exactly what Juckes said that he would do, which prompted my letter in #78:

    Thank you for the correction: The citation at that point in our article should be to the Energy and Environment article (which we refer to as MM2005c) and not MM2005 as it stands. I’m sorry for this mistake. I had an email from you at the start of this year saying the code used by MM2005c was not then available (and that you would inform me when it was), but since the situation has been remedied I will correct the text accordingly.

    It sounds to me like he merely plans to substitute MM05 (EE) for MM05 (GRL) in the above statement – which is what I objected to. His recent letter gives me little comfort on this count.

    Another relevant point in this discussion is the fact that I spent many hours at KNMI with coauthor Nanne Weber, who never mentioned any problems with source code. Plus as all of you know, coauthor Zorita and I actually exchanged code and toy data sets to ensure that we were accurately replicating one another’s understanding of Mannian methodology. Eduardo never mentioned that coauthor Juckes had any “uncertainties” about our methodology.

    In addition, while Juckes now says that there were supposedly “uncertainties” in our method, Juckes himself never expressed any “uncertainties” to me. As Juckes says in his own email – such uncertainties can be readily resolved with an email. When he encountered these supposed “uncertainties”, he took no steps to resolve them.

  95. bender
    Posted Nov 2, 2006 at 3:34 PM | Permalink

    It’s clear that he ought to remove the incorrect, offensive statement. If he doesn’t it can only be because he’s getting some childish satisfaction from digging his heels in. His colleagues should shame him for that, not reward him.

  96. MarkR
    Posted Nov 2, 2006 at 4:17 PM | Permalink

    Now Juckes is in full posession of the facts, I trust he will do the right thing.

  97. Posted Nov 3, 2006 at 8:57 AM | Permalink

    See http://www.cosis.net/members/journals/df/article.php?PHPSESSID=fb8156e80bb4fef23d9f9e1c22527f86&a_id=4661 for further details of the correction to our paper. I haven’t responded on this page previously because Stephen McIntyre only drew my attention to the related discussion on http://www.climateaudit.org/?p=886.

    Regarding comment 83 above, the centred PC would also show what Stephen McIntyre calls `this perverse property of the Mann method’ if Stephen didn’t have a “-” in his code inverting the centred PC. I am sure this is an oversight.

  98. Stan Palmer
    Posted Nov 3, 2006 at 11:29 AM | Permalink

    re 97 — perverse property

    A question from a layman.

    Doesn’t figure 2 indicate that increasing ring widths in the 15th century will decrease PC1 in the 15th century while the Mannian increase in the 20th is preserved. How is this an oversight or a mistake?

  99. BradH
    Posted Nov 4, 2006 at 2:58 AM | Permalink

    RE:#98

    As we’d say in Australia, “Jeez, mate, you’ve got balls.”

    Anyone who goads Steve and Ross in such an obvious fashion must either be extremely sure of themselves or…well, something else.

    Michael Mann shows a tendency to “move on”. But, then, so do Steve and Ross. I hope your study is watertight.

  100. Mark H
    Posted Nov 5, 2006 at 12:38 PM | Permalink

    RE:93

    Thanks for the generous assessment that ‘liable’ was a typo for ‘libel’ – actually it was the synaptic malfunction of an aging brain. That’s the best that can be said for a homophonic spelling error wherein one knows better.

    I like the pun…

    Mark H.

  101. BradH
    Posted Nov 5, 2006 at 10:48 PM | Permalink

    Re: # 99

    Just for the sake of clarity, my #99 should refer to Martin Juckes’#97, not to Stan Palmer’s #98.

    Sorry for the confusion.

  102. Posted Nov 6, 2006 at 4:32 AM | Permalink

    Re #99: I hope so too. That is not to say that it can’t be improved upon, of course. The biggest uncertainty is in the estimate of the temperatures of the 15th to 17th centuries. I hope our uncertainty about our estimates (which are significantly cooler than those of Mann et al. 1998, but still warmer than those of Huang et al. 2000) is properly reflected in our manuscript.

  103. Stan Palmer
    Posted Nov 6, 2006 at 6:23 AM | Permalink

    re 102

    Prof Jukes

    Did you notice my comment/question in 98. Will you please provide more context for your “perverse property” comment in 97. Will you please explain the oversight that you were referring to. Your comment in 102 seems to be referring to something quite different.

  104. Posted Nov 6, 2006 at 6:57 AM | Permalink

    Re #98/103: I was referring to the statement ‘a perverse feature of Mann’s method by which an arbitrary increase in non-BCP proxies in the period 1400-1450 resulted in a decrease in the Mannian “PC1″‘ — if the principal components are plotted as they come out of the relevant R functions then the McIntyrean PC1 also has this property. Stephen, elsewhere in this blog (I’m not sure where, on one of the other pages he has started up), now says that he is not concerned about the difference between the Mannian and the McIntyrean PCs, but rather in the issue you mention: i.e. the fact that a positive perturbation in some tree-ring series produces a negative perturbation in the index which is supposed to be related to temperature. I didn’t address this point in my previous comment. I suspect the answer is related to the fact that some of the trees in the Mann et al. collection are more sensitive to rainfall than to temperature. The rationale for including these is explained in Mann et al.’s paper. Whether you accept that rationale or not, a consequence is that some series may have weak or negative correlation with temperature.

  105. Steve McIntyre
    Posted Nov 6, 2006 at 7:10 AM | Permalink

    104. Martin’s comment here is, of course, completely wrong.

    The Mannian PC1 has different properties here than a centered PC1, despite MArtin’s efforts to camouflage and evade the point.

    Stephen, elsewhere in this blog (I’m not sure where, on one of the other pages he has started up), now says that he is not concerned about the difference between the Mannian and the McIntyrean PCs, but rather in the issue you mention

    There are multiple things wrong with Mannian methods. There’s no such thing as a McIntyrean PC. Mannian PC errors are not the only problem with 1000-year reconstructions, but I’ve never suggested that I’m not concerned with the differences between Mannian and centered PCs. Martin, I realize that there are many pages of this blog, but try to at least provide a quotation for such a comment.

    I suspect the answer is related to the fact that some of the trees in the Mann et al. collection are more sensitive to rainfall than to temperature. The rationale for including these is explained in Mann et al.’s paper. Whether you accept that rationale or not, a consequence is that some series may have weak or negative correlation with temperature.

    Martin, your surmise is again, of course, completely wrong. BTW I would have thought that an “evaluation” of 1000-year reconstructions would examine the use of proxies with a weak or negative correlation with temperature and how these perform in Mannian data mining.

  106. Stan Palmer
    Posted Nov 6, 2006 at 9:47 AM | Permalink

    Prof Jukes

    Thank you for your reply in 104 to my comment of post 98 abut the “perverse property” issue. However would you please supply some more context.

    In what way is there an error or oversight in fig 2 from McIntyre’s paper. Your post 97 describes one. The point, that I took from figure 2, was that an increase in tree ring widths in the 15th century resulted in an apparent decrease in 15th century PC1 (with obvious implications for temperature) while maintaining the Mannian increase in the 20th century. There was no suggestion that the McIntyre PC was “better” that the Mannian PC or that some proxies are sensitive to precipitation.

    In 97 you mentioned a “-” sign as the reason for oversight but this was not mentioned in 104. I understand that both the Mann and McIntyre PCs were inverted to provide a clearer link to an implied temperature.
    century

  107. Martin Juckes
    Posted Nov 6, 2006 at 3:32 PM | Permalink

    Stan, the error in McIntyre and McKitrick’s figure 2 was to say that their PC was based on centred data when it was not.

  108. Martin Juckes
    Posted Nov 6, 2006 at 3:41 PM | Permalink

    Re 105: Glad you’ve noticed the proliferation of pages on your blog Stephen.

  109. Steve McIntyre
    Posted Nov 6, 2006 at 3:44 PM | Permalink

    #108. A blog has a different function than a referee report. I’ve set up a Juckes Omnibus thread, where I’d invite you to localize your welcome participation here.

  110. Pat Frank
    Posted Nov 14, 2006 at 8:54 PM | Permalink

    #22 — John, before replying, I wanted to go look at Lipp’s paper (1991 Tellus B 43, 322) after Steve M. cited it, to see what he’d done. But the paper isn’t available on line and I’ve not had a chance to collect a hard copy. So, I’ll outline what I have in mind, and perhaps someone can comment about whether it’s practical.

    I’m not conversant with the equation editor some of you folks use, so please bear with me.

    Essentially, the rate of any reaction is dependent on the atomic weight of the isotope that is involved in the rate-limiting step. If reactants are otherwise identical, except that one molecule has a 12C and another a 13C at the appropriate site, the 12C reaction rate will proceed faster. The product molecules will be enriched in 12C atoms, and the remaining reactant will be enriched in 13C-atoms, relative to the starting mixture.

    Typically, the 12C/13C ratios in the product molecules are determined by the entry of the mass term into the energetics part of the reaction rate equation. The equations for this are well-known, and I don’t need to reproduce them here.

    The ratio of 12C to 13C in the product is referenced to the same ratio in the reactant and = 12R/13R = the “isotope effect” = exp[(deltaEa)/RT], where Ea is a reaction activation energy, R = the gas constant, and T = temperature in Kelvins. This ratio is usually only slightly larger than 1 because the masses of 12C and 13C are so similar.

    This is a very simple and very well-established relationship.

    Virtually all of the cellulose in a tree is produced from glucose produced from the photosynthetic fixation of atmospheric CO2. So, in calculating a tree isotope effect, the atmospheric 12CO2/13CO2 must be known. As 13C is a stable isotope, its ratio in atmospheric CO2 has probably been pretty constant over geological time.

    In applying to tree rings, one would have to know the activation energy, Ea, of the reactions leading from CO2 to cellulose or lignin. However, one can readily do a greenhouse experiment, growing trees under conditions of known temperature all the while monitoring the atmospheric 12CO2/13CO2 ratio. The 12C/13C in the tree rings can eventually be measured and from that ratio, and the known atmospheric 13CO2/12CO2 while the tree was growing, the effective 12R/13R that typifies the rings can be calculated.

    From the equation above, and the known 12R/13R, and the known T, one can then calculate an effective deltaEa for that tree under the known growth conditions. One has thereby determined the effective deltaEa that governs tree ring production under all the various conditions of growth that were employed. Now that one has the deltaEa, it can be applied to rings taken from trees in the field.

    Turning now to a tree ring from the field, one can measure the 12C/13C ratio in a ring, calculate the tree ring 12R/13R using the known atmospheric 13CO2 ratio, apply the known deltaEa determined from the greenhouse studies, and then back-calculate the temperature at which the cellulose and/or lignin was produced by the tree.

    This would be the temperature in the living wood at the time of synthesis, which should be directly convertable to an air temperature at the bark-air interface. This could be done for every single ring in a sample core, and an analytically valid termperature series would result.

    This temperature series would be free from all the usual uncertainties that plague the proxy reconstructions; uncertainties that arise, as we all know, from the qualitiative nature of the growth-limitation arguments. In contrast, a PCA applied to a temperature series derived from a measured isotope effect would reflect . . . . wait for it . . . a real temperature!

    Of course, real systems are more complicated than I’ve let on til now. From my reading so far, the major complication is that trees respire during the night. This will complicate a simple relationship between the 12CO2/13CO2 ratio in the atmosphere with 12C/13C in tree glucose, because in respiration free glucose is turned back into CO2. The 12C-containing glucose will again react faster, and so the tree will enrich itself in 13C glucose during the night. This will affect the 12C/13C ratio that subsequently appears in the tree ring.

    Another problem is that stress can affect the metabolic pathways in a tree, and this will again change the 12C/13C ratios that end up in a ring.

    However, all of this can be tested by greenhouse experiments, in which a standard tree is subjected to temperature, water, and nutrient stresses, and the 12C/13C ratios of the rings monitored. Greenhouse trees also respire at night, and so an “effective” 12C/13C ratio for tree rings may be available that inherently accounts for the respiratory back-reaction at all the different stress inputs.

    I don’t know what these experiments would show, obviously, but I think the experiments are worth doing because the potential payoff is large. It may even be possible that empirical relations may emerge between tree-ring 12C/13C ratios and the tested variables that will indicate which sort of stress a tree experienced during growth. This may especially be true if it were possible to measure cellulosic and ligninic ratios seperately (more information is always better).

    I’d guess that about 5 years of moderate-scale greenhouse experiments would be enough to explore the parameter space, so that the results could be applied. Twenty years would really nail it down. Anyone want to write a proposal? Rob Wilson?

    If it all worked out, you’d have a set of empirical parameters, referenced to a given sort of tree under given conditions applicable to the field, and all based in very well-understood reaction rate theory. These should allow one to calculate the real temperatures experienced by the tree during ring-wood synthesis.

    That’s about where my thinking stands now. It may be that people in the field have already considered this idea, and rejected it for good reasons. Someone may comment. Dave Dardinger?

  111. Steve McIntyre
    Posted Nov 14, 2006 at 9:11 PM | Permalink

    Pat, did you notice discussion of dC13 in bristlecones (and references) here http://www.climateaudit.org/?p=329?

  112. Pat Frank
    Posted Nov 14, 2006 at 10:03 PM | Permalink

    #111 — I didn’t notice that, Steve, thanks. I was able to get copies of all the references there except Tang, Glob. Clim. Chng., 1999. Scanning them, none seem to use the isotope effect to derive temperatures. There may be two reasons, one anticipated. First, one supposes that no one knows the effective “deltaEa.” But also, it appears that the atmospheric ratio of 13CO2 has changed with time. This complicates matters and may explain why no one has exploited the temperature dependence of the isotope effect to generate a T-time series.

    Thinking about it, there may be a strong feedback between the biological isotope effect and the atmospheric 13CO2 content. More or less growth could bounce the ratio around, if the flux of total CO2 through the global photosynthetic cycle was significant with respect to the total atmospheric concentration.

    I wonder if it would be possible to reconstruct a long atmospheric 13CO2 time series by sampling the inorganically sedimented calcite. Likely that would be hard, because the weathering of old limestones would modify the dissolved CO2 concentration. Details. It’s never easy. :-)

  113. Posted Nov 15, 2006 at 4:16 AM | Permalink

    Re #112

    Pat, the change in 13C due to fossil fuel burning is quite huge, e.g. in the past 20 years d13C changed from -7.5 to -8.1 per mil (data and graphs from CDIAC). Fossil fuel (and biomass in general) is 13C depleted, compared to the current levels in air.

    For the past 150 years, the relative stable d13C (some -6.2 to -6.5 pro mil in the period 1400-1800 AD, based on ice cores/air measurements, see Bàƒ⵨m ea. for purchase, comparing air/seawater trends of d13C) reduced to the above values.

    Thus the whole d13C story as temperature indicator is complicated by the large change in d13C from fossil fuel burning, which is complicated itself by the increase of 12C prefered biogenic uptake (counteracting the d13C decrease), while the decay of old plant material is coming from material made with d13C levels of year(s) ago… Further complication is the exchange with the upper ocean layer (which is richer in 13C than the atmosphere, but follows the atmospheric trend). And last but not least, local d13C levels may be much different in spring (maximum leaf/branches growth) vs. summer/autumn for forested places (larger seasonal differences) than for a place somewhere in the middle of the oceans…

    All together, I suppose that this makes it near impossible to use d13C as temperature indicator. An alternative may be the 18O/16O ratio in tree ring cellulose, but that seems to be more a precipitation indicator than a temperature indicator.

  114. Paul Dennis
    Posted Nov 15, 2006 at 6:42 AM | Permalink

    Re #22, 110-113

    Pat, that was a very elegant discussion of how a potential 13-C geothermometer may work. If only science were so simple.

    I think most trees utilise the C3 photosynthetic pathway, though I stand to be corrected here. In this path CO2 is converted to phosphoglycerate via the enzyme RuBisCO (Ribulose biphosphate carboxylase/oxygenase). Farquhar et al. (Aust. J. Plant Physiol., volume 9, 121-137, 1982) have produced a quantitative model to describe the carbon isotopic composition of plants. The robust nature of the model has been well documented.

    delta (plant) = delta (atm) – a -(b-a)*(p(i)/p(a))

    where delta (plant) is the carbon isotope composition of the plant tissue, delta (atm) is the carbon isotopic composition of atmospheric CO2, a is the isotopic fractionation resulting from differences in the diffusion rates of 13CO2 and 12CO2 in air (4.4 per mille), b is the isotopic fractionation due to the primary carbon fixing enzyme RuBisCO (30 per mille), p(i) is the intercellular CO2 concentration and P(a) is the partial pressure of CO2 in the atmosphere.

    The major source of variability in the isotopic composition of C3 plants is the term p(i)/p(a). Thus all the other terms which should have a temperature dependent term, can be treated as constant.

    Factors which affect the ratio p(i)/p(a) can be both environmental and biological and include irradiance, soil moisture availability, humidity, salinity, soiul fertility, altitude, landscape position, atmospheric P(CO2) (environmental) and photosynthetic capacity, development status, genetic variation, competition, growth form, life span (biological).

    Given the point that Ferdinand raises about the variability of atmospheric carbon isotope composition, and the multiplicity of factors, not all orthogonal, that are implicit in the Farquhar model I think it is probably a forlorn hope to try and develop a C-13 thermometer for tree ring cellulose

  115. Steve McIntyre
    Posted Nov 15, 2006 at 7:03 AM | Permalink

    Hi, Paul. Nice to hear from you again.

  116. Paul Dennis
    Posted Nov 15, 2006 at 7:08 AM | Permalink

    Thank you Steve.

    I have been following the site regularly but just been too busy to make constructive contributions.

    I’m very impressed with the high standard of science and high level of debate that you host here.

    Keep up the good work.

  117. Willis Eschenbach
    Posted Nov 15, 2006 at 3:51 PM | Permalink

    Re #114, Paul, thank you for a very clear post. In particular, your list of factors that affect the p(i)/p(a) ratio was quite comprehensive. Looking at them, I realized, by and large, these are the factors that affect ring width. With some additions to your list, I would say that the list of factors affecting ring width are:

    Environmental

    “‚⠠ temperature

    “‚⠠ irradiance

    “‚⠠ soil moisture availability

    “‚⠠ wind

    “‚⠠ humidity

    “‚⠠ cloud amount

    “‚⠠ mist amount (think mist forests raking moisture from the air, which all trees do to some extent)

    “‚⠠ salinity

    “‚⠠ soil fertility

    “‚⠠ altitude

    “‚⠠ landscape position

    “‚⠠ atmospheric p(CO2)

    Biological

    “‚⠠ photosynthetic capacity

    “‚⠠ water storage capacity

    “‚⠠ development status

    “‚⠠ leaf stomata density

    “‚⠠ genetic variation

    “‚⠠ competition

    “‚⠠ tree health (insects, parasites, etc.)

    “‚⠠ growth form

    “‚⠠ life span

    While some of these do not vary much for a given stand of trees, others vary greatly. Also, there are large differences between (and within) stands (e.g., those with more exposure are more sensitive to clouds and wind), and between different species (e.g., needle-leafed trees respond differently to lack of moisture than broad-leafed).

    While this does not rule out the use of tree climate response as a possible proxy, the onus is on the researchers to show that they have

    1) accounted properly for these confounding variables, and

    2) once those variables are allowed for, accounted properly for the “u-shaped” relationship between growth and temperature.

    I’m not saying that can’t be done … just that, to my knowledge, it hasn’t ever been done. Showing a fairly low correlation between ring width and temperature in a correlation period, particularly without using a verification period (as Juckes has done), doesn’t pass the test. This is particularly true given the short length of the instrumental record.

    For example, tree ring width for a particular species may be the same during a warmer, drier century as it is during a cooler, wetter century. How can we possibly extract a temperature signal for say the years 1300-1400 without knowing how wet the century was compared to our correlation period? And that’s just one of the many confounding variables.

    w.

  118. jae
    Posted Nov 15, 2006 at 4:25 PM | Permalink

    Hey, I think I’ve got a new way for the Team to do reconstructions, taking advantage of the divergence problem. Just assume that the divergence is because temperatures have now become so high that the trees are now on the “right-hand” side of the growth/temperature quadratic. That would justify “flipping” the curves after, say, about 1980. Presto, a great hockey stick! Since the Team has embraced avant-garde science and statistical methods, this methodology should be acceptable.

  119. Pat Frank
    Posted Nov 16, 2006 at 6:27 PM | Permalink

    #113, 114 — Ferdinand and Paul, thanks for your critical comments. They always help one’s thinking. Ferdinand, if the change in the modern atmospheric 13C ratio due to fossil fuel burning is known, then this variation can be factored into the calculation and should not pose too much of a problem. The real question is what the atmospheric 13C ratio has done over the last several thousand years. If it’s varied significantly, then I agree that using it to calculate a temperature will be difficult, if not impossible. Perhaps ice cores could help set the long-term ratio, though I doubt it.

    Paul, thanks for the list of variables. I couldn’t imagine more, although Willis did in #117. :-) The thing is that several of these variables can be explored in greenhouse studies. One would also have to choose a standard tree species for each climate. Within that species, some of the effects of genetic variation will show up, as different trees of the same standard species react slightly differently in the greenhouse to the same set of conditions.

    With a large-scale set up, one could test the effects of temperature, soil fertility, altitude, and water availability on these trees. One would then obtain a deltaEa for each of these variables that could be averaged over the genetic variability. Greenhouse 13C would also be monitored continuously, perhaps using a relatively inexpensive time-of-flight mass spectrometer. If each of, say, four greenhouses had four subdivisions, one would be able to test all four variables in concert, in 16 concurrent experiments lasting 5-20 years. Obviously, a deltaEa averaged over the genetic variablility would produce an uncertainty in the derived temperature from a wild-type tree.

    What one ends up with in such a case is a kind of semi-empirical model, in which the deltaEa includes all the factors that have not been explicitly controlled. For example, suppose you tested soil nutrients in terms of phosphate (probably a ubiquitously limiting factor) nitrate, and alkalinity. The effect of, e.g., insufficient iron or potassium would not be tested, and would therefore show up as a variation in deltaEa. The deltaEa’s would thus never be a true net energy, but would be the energy multiplied by some unknown factor that includes and accounts for all the uncontrolled variables. The question is how much the semi-empirical deltaEa changes with, e.g., variations in e.g., iron and potassium availability. If these effects are second order, then only small uncertainties would be introduced into the derived temperature by their unknown contribution in the field.

    In practice, some of these imponderables, including some of those Willis listed, such as landscape position, irradiance, crowding, and perhaps others, could be accounted for in the field if researchers are very careful in choosing the trees to be cored. They wouldn’t choose a tree in the middle of a crowded stand, or one that is shaded most of the day, etc. That is, in the field one would choose trees that came as close as possible to the known and standardized greenhouse conditions. I think judicious ecological choices coupled with a good greenhouse experimental program could go a long way toward discovering whether the system can work.

    It may end up being too problematic, or it may not. On the positive side, it may turn out that detailed 13C ratioing of parts of tree rings, using something like laser desorption mass spectrometry to look a submillimeter spots, could detail the metabolic responses of the tree to tested variables as they show up in different parts of the wood. These details may allow certain effects to be deconvolved. If so, then the same microscopic tests of cores from wild trees may allow some sort of reconstruction of some of the past ecological variables it experienced. I don’t know, and no one will know unless the experiments are done.

    In any case, except for the potential problem of knowing the millennial atmospheric 13C ratio, just now I don’t see any fatal problem stopping a research program. But then, I’m neither a tree biologist, nor a dendrochronologist.

  120. bender
    Posted Nov 16, 2006 at 7:39 PM | Permalink

    “judicious ecological choices” => D’Arrigo’s cherry pie

  121. Pat Frank
    Posted Nov 16, 2006 at 8:53 PM | Permalink

    #120 — not really. The ecological criteria would be objectively determined by the physical greenhouse conditions, and the trees would be critically chosen according to whether their specific locale matched those physical conditions; otherwise, the derived deltaEa’s would be inapplicable. The choices, in other words, would have no connection whatever to the ‘global average temperature’ of Jones & co., or to anyone else’s dendro reconstruction.

    In principle, presuming the method worked, a modern tree core temperature record derived from 12C/13C would test the local accuracy of Jones’ network stations.

  122. bender
    Posted Nov 16, 2006 at 9:17 PM | Permalink

    Fair enough. Useful distinction.

  123. jae
    Posted Nov 16, 2006 at 9:25 PM | Permalink

    119: How are you going to deal with the upside down quadratic relationship between temperature and growth? If temperatures get to approx. 25 degrees, growth slows, regardless of all the other factors. How do we know what was going on 300 years ago? There are many better temperature proxies than tree rings, and it is useless to dwell on them. They are very good for droughts (maybe that can be related to temperatures??).

  124. jae
    Posted Nov 16, 2006 at 9:35 PM | Permalink

    WAY too much noise in tree ring proxies to make them useful. Tree lines, sediments, solar cycles, stalagmites–much more information there, methinks.

  125. Nicholas
    Posted Nov 16, 2006 at 10:10 PM | Permalink

    Pat: Doesn’t carbon dating rely on the carbon isotope ratio in the atmosphere being more or less steady over time?

    It might be that the carbon dating errors are caused by this assumption being not-quite-valid. (Some dates are off by a bit; that guy whose paper was linked recently, which was criticizing the grape harvest proxy, had another paper on his site about carbon dating errors.) But, I think you might find quite a bit of research into this issue among scientists who practice carbon dating techniques. They may have already come up with a way to plot the changes over time, since it would help them calibrate their results.

  126. bender
    Posted Nov 16, 2006 at 11:04 PM | Permalink

    The carbon isn’t used for dating, just temperature calibration. The annual tree rings themselves provide the annually resolved dates.

  127. Dave Dardinger
    Posted Nov 16, 2006 at 11:07 PM | Permalink

    re: #125

    Oh yes, the C14 dating anomalies have been know for decades and are routinely used in calculating dates. I think they’ve been calculated for tree ring series for most all of this interglacial. IOW, if you have a charcoal sample and you date it as 4129 BP based on the known halflife of C14 and the present (or at least recent) C14 level you can go to a book or online table and convert to a true date. This is one thing treering series have been very useful for. Of course there can be some anomalies where there might be two different actual dates which give the same C14 date, but except for the very recent past where atomic tests confuse things the dates are still going to be close. But there are still things which have to be considered such as water containing recent CO2 contaminating old samples and making them seem much younger than they are or old carbon from say dissolved limestone doing the opposite. I’m sure the people who measure the actual ages like to have as much detail as they can so they can check for such possible contamination.

  128. Willis Eschenbach
    Posted Nov 16, 2006 at 11:39 PM | Permalink

    Re #119, Pat F., thank you for a thoughtful quote. In it, you say:

    In practice, some of these imponderables, including some of those Willis listed, such as landscape position, irradiance, crowding, and perhaps others, could be accounted for in the field if researchers are very careful in choosing the trees to be cored. They wouldn’t choose a tree in the middle of a crowded stand, or one that is shaded most of the day, etc.

    While we know if the tree is crowded or shaded now … was it crowded or shaded in say the year 1566?

    w.

  129. Paul Dennis
    Posted Nov 17, 2006 at 4:06 AM | Permalink

    Re #119

    Pat, and others who are interested here are some references that relate to using 13C/12C in tree rings:

    1) as a palaeothermometer

    Wilson, A.T. and Grinsted, 1977, 12C/13C in cellulose and lignin as palaeothermometers. Nature, v265, 133-135

    2) as a measure of humidity

    Leavitt, S.W. and Long, 1989, Drought indicated in carbon-13/carbon-12 ratios in south western tree rings. Water. Resour.\ Bull., v 25, 341-347

    McNulty, S.G. and Swank, W.T., 1995, Wood d13C as a measure of annual basal growth and soil-water stress in a Pinus Strobus forest. Ecology, v 76, 1581-1586

    3)Solar radiation

    Hanba, Y.T. et al., 1996, Solar radiation affects modern tree ring delta 13-C:observations at a cool temperate forest in Japan. Isotopes environmental health Studies. v 32, 55-62

    4) Water use effiency

    Pate, J and Arthur, D., 1998, d13C analysis of phloem sap carbon: novel means of evaluating seasonal water stress and interpreting carbon isotope signatures of foliage and trunk wood of Eucalyptus globulus. Oecologia, v 17, 301-311

    5) Laser ablation isotope ratio measurements at 40 micron resolution

    Schulze. B. et al., 2004, Laser ablation-combustion-GC-IRMS – a new method for online analysis of intra-annual variation on d13C in tree rings. Tree Physiology, v 24, 1193-1201

    Theres a long history of people trying to use isotopes in tree rings.

    Pat, your outline research programme is interesting and I’m sure some groups are beginning to think and work along these lines. Farquhar’s group at ANU is probably a very good place to start.

  130. Pat Frank
    Posted Nov 17, 2006 at 2:43 PM | Permalink

    Interesting comments all, thanks. #113, Jae, the isotope effect method, if it worked, would be entirely free from the quadratic response growth effect because it doesn’t depend at all on ring-widths or densities.

    #125 – Nicholas, as Bender noted in 126, we’re talking about stable isotopes, whereas carbon dating uses 14C, which is radioactive. You’re right, though, that the 13C content of the atmosphere would have to be known, or estimatable, across most of the Holocene for the method to be fully useful.

    #128 – Willis, you raise a good point, as usual. It would be hard to know the stand environment of a tree in its distant past. I’d guess, though, that if a tree had been crowded in the past, there would be stumps or snags of those previously living trees close by, which could be evaluated. This presumes an undisturbed forest. In any case, I’d expect dendroclimatologists to be sensitive to these issues, and work out protocols as best they could to minimize such problems, or at least render estimates. Any forest has outlier trees. Perhaps these could be selectively chosen for study, with surveys done for close-by snags of now dead, once-living trees. All of this would require, of course, a committment by dendroclimo people to analytically reproducible methods; something that has apparently escaped them up til now.

    #129 – Paul those are excellent references, thanks. The Wilson 1977 paper in particular shows thinking along the same lines as proposed here already 30 years ago, and they found some interesting lignin-cellulose differences. They also show that the isotope ratios reflected leaf temperatures rather than bark temperatures, which may make the method both more accurate and more revealing of metabolic responses. Wilson’s results, plus the 2004 paper of Schulze, ea, demonstrating a laser desorption-GC-mass spec method with 40 micron resolution, beautifully demonstrates that using 12C/13C to determine a growth temperature is at least analytically possible, if not empirically so. Something I don’t understand just now, is that Wilson in 1977 reported a sd (+/- 0.02 %%) ten times smaller than Schulze reported in 2004 (+/- 0.24%%). Anyway, following the information you provided, the stage is at least set to do the baseline greenhouse experiments to see whether the isotope effect approach to temperature recovery is possible.

  131. Paul Dennis
    Posted Nov 17, 2006 at 4:35 PM | Permalink

    test message

  132. Paul Dennis
    Posted Nov 17, 2006 at 4:56 PM | Permalink

    Sorry folks for the test message. I was able to submit posts from my office, but not from home. The test was to see if I can now submit messages from home. I can!

    re #130

    Pat you rightly ask about the difference in precision between the Schulze work of 2004 and the earlier Wilson study which, despite being more than 30 years older had a factor of 10x better precision. This is related to the sample preparation method and the mass spectrometry technique used.

    In the Wilson work they used a technique known as dual inlet isotope ratio mass spectrometry and larger volume samples. Dual inlet techniques are the gold standard and able to produce measurements of quite stunning precision. Today +/-0.005 per mille would be achievable. To give some indication what this means the 13-C content of CO2 is about 1%, or 0.001. i.e. 1 in every 100 carbon atoms is 13-C. A precision of +/-0.005 per mille allows us to discriminate between two samples that differ by just 5 13-C atoms in 100 million carbon atoms!

    In the more recent Schulze article they were using a form of isotope ratio mass spectrometery known as continuous flow IRMS. This is inherently less precise than the classic dual-inlet techniques. However, the advantage is that you can more easily handle smaller samples (eg. those produced by laser ablation), and also measure very many more samples in a shorter time scale. Moreover the instrumentation is cheaper to buy, easier to run etc.

    Bsically it’s a trade off….sample throughput versus precision. In dealing with organic systems many have opted for continuous flow techniques arguing that the inherent sample to sample variability doesn’t require the ultimate in precision.

    Having said this I am still surprised by the precision quoted by Wilson. For this to be a good estimate of his measurement precision he needs to run replicates through the sample preparation and measuremnt procedure.

    My opinion, FWIW, is that there is a place for both techniques. My lab has 5 dual-inlet instruments and 3 continuous flow ones. You can do some other neat things with the continuous flow machines that you can’t with the dual-inlet. For example we’ve just installed this week a system that will allow us to do simultaneous C, N, S and H isotope analyses on the same sample!

    As a point of information the Schulze method is strictly speaking laser ablation. Using UV wavelengths you get ‘cold’ ablation of samples. In tree rings the wood is converted to a very fine aerosol which is then combusted.

    I hope this clarifies the precision question.

  133. Pat Frank
    Posted Nov 17, 2006 at 11:31 PM | Permalink

    #132 — Thanks, Paul, that answers my question. You sound very relevantly expert. You wouldn’t want to give it a try, would you? After all, an initial experiment might just be a preliminary proof-of-concept. :-)

  134. Posted Nov 18, 2006 at 6:14 AM | Permalink

    Re #119,

    Pat, As there are no direct air measurments from the pre-industrial past, the ice cores will give some indication of past atmospheric 13C/12C ratio changes, be it rather smoothed. The fastest accumulation ice core (Law Dome) still needs some 60 years to closure, thus any faster ratio change will go unnoticed or leveled off.

    Corraline sponges may give some more detailed ratio changes (5-10 years resolution), but are influenced by local ocean streams and show an offset between water and air ratios (and a varying d13C change of water vs. air of between 0.6 and 1), see the graph from Bàƒ⵨m ea.

    This gives some indication of pre-industrial variations. To be noticed that the LIA d13C is higher than in the previous period, which points towards more biogenic uptake (or less decay?) during the LIA, while one should expect the reverse…

    Other proxies may be better (more recent calcite sediments?) or worse (stalagmites – reflect very old local calcite deposits – with high d13C levels, together with more recent air d13C levels in CO2 which dissolved the calcite).

  135. Posted Nov 18, 2006 at 7:12 AM | Permalink

    Here is the d13C trend graph (air/water) of Bàƒ⵨m ea.:

    If it doesn’t show up:

  136. Paul Dennis
    Posted Nov 18, 2006 at 11:22 AM | Permalink

    re #134 and 135

    Ferdinand and Pat, I think that the ice and coralline data are probably the best we have for reconstructing past atmospheric CO2 13-C isotope compositions. This is a comparatively simple system as far as dissolved inorganic carbon is concerned. The surface water, in isotopic equilbrium with the atmosphere, will have a DIC composition that tracks that of the atmosphere. Neither of these archives are going to provide high temporal resolution.

    Working with terrestrial proxies is far more difficult. As you point out Ferdinand I don’t think there is any scope for using speleothems. The dissolved inorganic chemistry of groundwater is complicated. First, in the unsaturated zone of aquifers, CO2 partial pressures can be elevated by several tens to a hundred fold over the atmosphere due to plant respiration; isotope compositions are fixed by the dominant vegetation (C3, C4 or CAM). Secondly, in the aquifer there is dissolution of carbonate, usually with a heavy isotope signature (close to 0 per mille VPDB). The evolution of the groundwater carbon isotope composition will depend on wether or not dissolution occurs under open or closed system conditions. Finally, kinetic effects may dominate during speleothem precipitation with rapid degassing of CO2 from drip water in a cave. The final isotope signature is likely to be far removed from that of the atmosphere.

    There may be a possibility that some lake sediments might preserve a signal.

    What is needed is an archive of ancient air. Groups have looked at air collected from deep within old, stable sand dunes. Old scientific instruments such as aneroid barometers may provide some data but it will be patchy and data quality may be compromised though this can be checked by a suite of measurements on the other atmospheric gases.

  137. Pat Frank
    Posted Nov 18, 2006 at 2:02 PM | Permalink

    #136 — Paul would ancient air be trapped in old porous basalts? Would it be possible to polish down pieces of basalts looking for uncracked pore-spaces that could contain samples of atmosphere trapped when the basalt hardened? The basalts themselves could be dated using the usual radiometric methods, and that date applied to the trapped air.

    Or suppose one took some basalt and evacuated it to high vacuum, pumping out all the gasses that were in communication with the atmosphere by way of the crack network. Once it was at high vacuum, say 10^-10 torr or so, and perhaps heated to 250 C, one could suppose the removal of all the gasses that were in adventitious communication with external air over time. Then the basalt could be crushed. Any cryptic gas trapped in closed pores would thus be released. That gas should reflect the local atmosphere that was present when the magmatic pore closed and hardened. This gas could be swept into a GC-MS and the 12C/13C ratio in the CO2 fraction could be determined. Again, the basalt itself could be conventionally dated.

    There are probably enough modern basalts that hardened within the last 100 years or so, i.e., when the atmospheric 12C/13C ratio was known, to test the method. Clearly, the method would be time-wise spotty, but consistent results, that is, a constant or closely-varying ratio, might give some confidence that the 12C/13C ratio was fairly stable over millennial times. That finding could then be applied to ancient wood — appropriately chosen, pace Bender and Willis — to derive a temperature.

  138. Pat Frank
    Posted Nov 18, 2006 at 2:22 PM | Permalink

    #135 — Ferdinand, thank-you for that graph. First, am I wrong or is the mean-line the mean of all the data, and not just the atmospheric set? It looks to me that the mean of just the atmospheric data over the 1375-1850 baseline time would be displaced to slightly higher values of 13C (~ -6.3 %%). On the other hand, the mean seems to show a fairly stable 13C %% over the 500 years before the onset of industrialization; to within +/- 0.2%% 13C or so for the mean line as drawn. Obviously we’d have no idea how an sd of this magnitude would translate into a temperature uncertainty, presuming it typified the Holocene.

  139. Paul Dennis
    Posted Nov 18, 2006 at 2:44 PM | Permalink

    re #137

    It’s an interesting thought Pat. Are you an experimentalist and analyst by any chance. You have a good grasp of appropriate techniques. You wouldn’t like a job in my lab would you!? I think there is a potential problem with magma derived CO2. All magmas contain varying amounts of CO2, though much of this is outgassed at the time of the eruption. Some would potentially be included in gas vesicles and would interfere with the atmospheric signal. There are quite a large number of studies looking at gas released from crushed basalts and other rocks and minerals. Most of these are aimed at understanding the processes that occur in magma chambers etc.

    With respect to minerals I remember that Yapp and co-workers had carried out some interesting experiments on goethite (an iron hydroxide mineral) that is formed as a weathering product. On heating the goethite released CO2 with distinctive isotope signatures. I don’t believe that the compositions were ever related to atmospheric compositions and I’m not sure if this idea was ever developed very far.

    From time to time I’ve put my mind towards the problem of accurately measuring pre-industrial atmospheric compositions and now the debate here has spurred my interest again. There is a rich heritage of great scientists including Lavoisier, Priestley and others who have measured the composition of air. One wonders if any of their flasks still exist that have remained sealed. I once asked the Cavendish Laboratory in Cambridge if they had any. Othe rplaces might be the Royal Institute.

    In all seriousness such measurements would only provide a patchy record. What is required is an archive that has global distribution and a potentially good time resolution and certainly better than a decade.

  140. MarkR
    Posted Nov 18, 2006 at 4:52 PM | Permalink

    re #139 How about air samples from sealed coffins, crematoria ash containers, trapped air from shipwrecks.

  141. Paul Dennis
    Posted Nov 18, 2006 at 5:33 PM | Permalink

    re #140

    Mark, I think all these have some potential but one needs to be careful. If there is any organic material present then even small amounts of decay and oxidation will produce CO2.

  142. Posted Nov 18, 2006 at 6:07 PM | Permalink

    Re #138,

    It looks like they have used the shallow water sponge data to make a base line. The problem with the air data is that they needed to scale down the air d13C scale (as the d13C change in water is about 80% of the change in air) to make air/water changes comparable, but that the dd13Cwater/dd13Cair is not constant, thus the 1400-1850 AD avarage for air indeed may be somewhat higher.

    You are right about the relative small pre-industrial changes. That should be the result of some 1 K difference in temperature between MWP and LIA (although in this case the MWP is not covered). But, indeed that doesn’t say much about the changes in d13C one will find in tree rings covering the MWP-LIA period. Only that the influence of changing d13C levels in the pre-industrial period were much smaller than thereafter, and will have less influence on the temperature “calculation”…

    I have posted a new graph with a better resolution at the previous link. It seems that the previous code to present graphics on the blog doesn’t work anymore (John A!) and the reply box shows some strange behaviour (text dissapears at the extended right side of the box) in IE…

  143. Pat Frank
    Posted Nov 18, 2006 at 8:56 PM | Permalink

    #139 — Darn, I didn’t think of volcanic CO2. It’s copious, isn’t it. That puts paid to old basalt. :-)

    Thinking about the problem and the various difficulties — shallow fresh water getting infusions of old carbonate or decaying matter, etc. — it occurred to me to wonder whether there is a calcerous oceanic surface-water diatom one could use. If there was, then one could baseline its 12C/13C response versus the modern atmospheric values so as to get a calibration with respect to its metabolic isotope effect. Perhaps one could extend this across a century or so, using the most reliable 12C/13C atmospheric ratios, to see if the metabolic effect is reproducible and stable. One would then look for exactly that species of diatom in sedimentary benthic drill cores and do the 12C/13C ratio. So long as the calcerous morphology was unchanged in retro-time, one could reasonably argue for species identity and the same metabolic isotopic frationation as the modern species.

    I understand from previous discussions here that there may be a problem with carbonate exchange over time in sedimentary diatoms. That would, of course, progressively skew the pristine 12C/13C diatomaceous carbonate ratio. But I wonder whether one could experimentally account for that, by adapting a method from K/Ar dating. That is, one could argue that any carbonate exchange process would most strongly affect the surface layers of the diatomaceous carbonate, and this exchange would progressively attenuate deeper into the skeleton.

    So, suppose one baked the sample with a progressive series of temperatures, taking advantage of the thermal reaction: CaCO3 —-(heat)—-> CO2 + CaO. Each pulse of CO2, coming from deeper in the calcerous skeleton, would be swept into the GC-MS and the 12C/13C ratio measured. At some point, if a pristine layer is reached, the data would show a progressively sloped plot, followed by a plateau. The plateau should show the original 12C/13C ratio.

    Would that work, Paul? If it did, it also seems possible that the same diatom could be used to reconstruct ancient SST’s using the identical kinetic isotope method we’ve discussed above for tree rings. That is, one could derive a deltaEa for the modern diatom in the same way as might be derived for a tree species. It could be tested against slightly older diatoms for which the SSTs and atmospheric 12C/!3C ratios were known. Knowing the deltaEa derived from modern diatoms, would could apply it to the measured pristine 12C/13C ratio of ancient conspecific diatoms and back-calculate the ancient ambient sea surface temperature.

    Thanks for the job-offer by the way, even in jest. :-) As you deduced I am indeed, like you, an experimental scientist, happily employed at SLAC. If you ever need to speciate sulfur — sulphur, to you :-) — in complex materials, let me know. I had the pleasure of working on wood from the 17th century Swedish ship “Vasa” awhile back.

    #142 — Thanks for the better graph, Ferdinand, it’s much easier to read. I wonder, do you think the 80% scaling is due to a metabolic isotope effect of the sponge, that depletes 13C in its skeleton? I can’t think, just now, why the atmospheric 12C/13C ratio would change much because of dissolution in water.

  144. Posted Nov 19, 2006 at 8:10 AM | Permalink

    Re #143,

    Pat, Emiliania huxleyi, one of the abundant coccolithophores (phytoplankton with a calcite skeleton), may be a good candidate for d13C variation search. Ocean sediments are already used as (ocean) surface temperature proxy, see J. Eiràƒ⬫sson ea., where they used several types of proxy measurements. Interesting is that they detected several solar cycles and a (2 K) warmer MWP vs. a colder LIA in the near Icelandic sediment…

    The relative slower change of d13C of upper ocean waters vs. air may be a matter of CO2 storage capacity of shallow waters vs. the atmosphere in equilibrium. This may be about 1.25:1 (something that is alluded too in carbon cycle diagrams).
    Organic carbon in phytoplankton is preferentially build with 12C (as one of the oldest lifeforms, probably C3 cycle? Paul may give some better explanation here!), thus during build-up of the organics, the surrounding water will get enriched in 13C vs. the atmosphere. The shells are build with a direct (non-organic) chemical reaction, and reflect the 13C ratio of the surrounding water (at the time of formation, which varies with the seasons, and may be influenced by the inner organic buildup…).

    Thus even in the pre-industrial more or less dynamic equilibrium between air and upper ocean waters, the d13C in water is always higher than in air. Deep ocean water at the other side is richer in 13C (around 0 pro mil) than air, but poorer than the upper ocean layer. Any change in upwelling of the oceans (as good as any change in more or less downwelling of cold CO2-rich waters at the North Atlantic sinks) will have its influence on the equilibrium…

    Gruber, Keeling and Bates made a comprehensive overview of the difficulties involved in calculating the variability of North Atlantic carbon sinks over the years. The supplemental material is particularly interesting, and the equation includes a term for changes in mixed upper ocean volume over the seasons.

  145. Posted Nov 19, 2006 at 8:10 AM | Permalink

    Re #143,

    Pat, Emiliania huxleyi, one of the abundant coccolithophores (phytoplankton with a calcite skeleton), may be a good candidate for d13C variation search. Ocean sediments are already used as (ocean) surface temperature proxy, see J. Eiràƒ⬫sson ea., where they used several types of proxy measurements. Interesting is that they detected several solar cycles and a (2 K) warmer MWP vs. a colder LIA in the near Icelandic sediment…

    The relative slower change of d13C of upper ocean waters vs. air may be a matter of CO2 storage capacity of shallow waters vs. the atmosphere in equilibrium. This may be about 1.25:1 (something that is alluded too in carbon cycle diagrams).
    Organic carbon in phytoplankton is preferentially build with 12C (as one of the oldest lifeforms, probably C3 cycle? Paul may give some better explanation here!), thus during build-up of the organics, the surrounding water will get enriched in 13C vs. the atmosphere. The shells are build with a direct (non-organic) chemical reaction, and reflect the 13C ratio of the surrounding water (at the time of formation, which varies with the seasons, and may be influenced by the inner organic buildup…).

    Thus even in the pre-industrial more or less dynamic equilibrium between air and upper ocean waters, the d13C in water is always higher than in air. Deep ocean water at the other side is richer in 13C (around 0 pro mil) than air, but poorer than the upper ocean layer. Any change in upwelling of the oceans (as good as any change in more or less downwelling of cold CO2-rich waters at the North Atlantic sinks) will have its influence on the equilibrium…

    Gruber, Keeling and Bates made a comprehensive overview of the difficulties involved in calculating the variability of North Atlantic carbon sinks over the years. The supplemental material is particularly interesting, and the equation includes a term for changes in mixed upper ocean volume over the seasons.

  146. Pat Frank
    Posted Nov 19, 2006 at 11:20 PM | Permalink

    #145 — Good post, Ferdinand. I hadn’t thought about organic carbon build-up depleting surface waters of 12C, but if the CO2 used for photosynthesis is from the dissolved reservior, then that’s what will happen. Of course, night-time respiration will tend to restore some of the 12CO2. Deep water upwelling will further complcate the picture. The situation is rather like that of trees. In fact, Wilson’s 1977 paper, referenced above by Paul, mentions that respiration in roots diminishes the 13CO2 ratio in the air around young trees. Tree wood thus only reflects the atmospheric ratio when the trees are adult and the tree canopy rises well above the soil.

    It’s clear there are a lot of confounding factors in each case; tree rings and calcerous plankton both. Honestly, it will take a real professional, cognizant of all the subtlties, to sort it out and decide whether an isotope effect can produce a valid thermometer.

One Trackback

  1. [...] (Potential Academic Misconduct by the Euro Team) [...]

Follow

Get every new post delivered to your Inbox.

Join 3,304 other followers

%d bloggers like this: