New Mann Paper

Michael E. Mann, 2006, Climate Over the Past Two Millennia, Annu. Rev. Earth Planet. Sci. 2007. 35:111–36 is online here , No signs so far of Mann, M.E. et al, Robustness of proxy-based climate field reconstruction methods, 2006 (accepted), which was cited by “Anonymous Referee #2″ in the Burger-Cubasch review.

Funding generously provided by NSF here.


65 Comments

  1. Posted Dec 2, 2006 at 11:10 AM | Permalink

    Mann hasn’t lost his touch with the snipey comment:

    The first category comprises annually or perhaps decadally
    resolved high-resolution proxy records, such as tree rings, corals, ice cores, laminated
    sediments, and historical documentary proxy information. Such records can
    potentially be calibrated against the shorter available instrumental records to yield
    quantitative climate reconstructions [see, e.g., reviews by Bradley (1999) and Jones
    & Mann (2004)]. The second category includes records that cannot be explicitly calibrated
    in this manner because they are less well resolved in time, have less precise
    age models, or both. Nonetheless, the records can often provide meaningful insights
    into centennial-scale climate changes in the past. Examples are nonlaminated marine
    lake and marine sediments (Keigwin 1996…)

    I think Keigwin might give askance that his sediment record “cannot be explicitly calibrated”.

  2. Proxy
    Posted Dec 2, 2006 at 12:46 PM | Permalink

    Mann does not appear to distinguish between the qualitative nature of evidence from proxy climate data and that of theoretical climate model simulations.

  3. Gary
    Posted Dec 2, 2006 at 1:53 PM | Permalink

    So where are references to variation in solar insolation? How can a paper with such a grand title avoid this?

  4. Pat Frank
    Posted Dec 2, 2006 at 1:58 PM | Permalink

    I find it depressing that the HT is carrying on as though there were no M&M critique at all and as though their cherry-picking, data-snooping, PC4-elevating, bc pines backdoor-inveigling, r^2-censoring, data-occultating, methods-obfuscating, theory-dodging approach produced quantitatively valid millennial climate reconstructions, which are then accepted at face value by the press and loudly touted by the bloomeranians as yet more evidence for the bankruptcy of technical civilization.

    I find it further depressing that there are far more of them than there is of Steve, and they are having no trouble at all keeping ahead with their gray-weather sleet of publications. Steve is ever playing catch-up, showing the analytical poverty of the last HT paper all the while three more are in the pipeline. Steve, you’re winning every battle while we’re all losing the war. Somehow soomeone needs to boot Wegman to stop speechifying, and get him and some of his colleagues to start writing critical papers. Maybe Ross can give him an academically collegiate phone call.

  5. Jeff Weffer
    Posted Dec 2, 2006 at 2:10 PM | Permalink

    It doesn’t appear that Mann will reform. He is going to stick it out, getting more brazen every time, until the end.

    Many more hockey sticks in that paper, now going back 2,000 years.

    Have some fun with this one Steve.

  6. Rev Jackson
    Posted Dec 2, 2006 at 2:18 PM | Permalink

    Upton Sinclair explained it best way back in the last century: “It is difficult to get a man to understand a thing when his salary depends on his not understanding it.”

  7. Hans Erren
    Posted Dec 2, 2006 at 3:09 PM | Permalink

    I see mann refers to the Luterbacher Hockeystick:

    http://home.casema.nl/errenwijlens/co2/errenvsluterbacher.htm

  8. Brooks Hurd
    Posted Dec 2, 2006 at 6:56 PM | Permalink

    On page 123, Mann now claims that every reconstruction performed to date indicates anomalous 20th century warmth.

    Although there are significant differences between the various published NH mean temperature reconstructions (Figure 4), every reconstruction performed to date indicates that large-scale late-twentieth-century warmth is anomalous in the context of at least the past 1000–2000 years.

    Mann’s 4th Summary Point must be a classic.

    Tests with synthetic pseudoproxy networks derived from climate model simulations indicate that statistical methods used for reconstructing past climate from proxy data are likely to yield reliable reconstructions back at least 1000 years within estimated uncertainties, given the statistical properties estimated for actual proxy networks.

  9. Henry
    Posted Dec 2, 2006 at 7:13 PM | Permalink

    Look at future issue 2:

    Continued investigations of climate field reconstructions should focus on the use of teleconnected local climate responses and the stability of the associated long-term relationships

    What does “teleconnected” mean? Is it about how bristlecone pines may be more related to global than to local temperature?

  10. Willis Eschenbach
    Posted Dec 2, 2006 at 8:01 PM | Permalink

    Teleconnected means that local climates sometimes respond to distant climate events. Air temperature in Africa and the Indian Ocean is correlated with the El Niños in the Pacific, for example.

    However, “teleconnected” seems to be invoked by paleodendroclimatologists to imply that trees can respond to distant climate events without the local climate responding to those events, a phenomenon I refer to as plantelepathy …

    w.

  11. Jeremy
    Posted Dec 2, 2006 at 9:14 PM | Permalink

    Ah yes, that’s why I come here, for the high-brow humor.

    Plantelepathy, lol.

  12. Ken Fritsch
    Posted Dec 2, 2006 at 9:26 PM | Permalink

    Just finished reading the review and once again Mann according to Mann is the man, or more than the man if we assume that all men are imperfect. All his methods are correct and all criticisms of them are incorrect (even though that might be because he defaults all works as incorrect if it did not exactly replicate his less than totally revealed methods).

    I saw no comments about MM, but he had some strong criticisms of the criticisms of Von Storch, Zorita, Burger and Cubasch of Mann’s methods. It will be interesting to see what kind of replies his paper receives.

  13. Ken Fritsch
    Posted Dec 2, 2006 at 9:42 PM | Permalink

    However, “teleconnected” seems to be invoked by paleodendroclimatologists to imply that trees can respond to distant climate events without the local climate responding to those events, a phenomenon I refer to as plantelepathy …

    Would not you expect that Mann and cohorts at some point will attempt (and maybe even feel scientifically obligated) to theorize a teleconnection of local proxy responses to at least regional climate events — say through a secondary variable such as rainfall that in turn can be correlated to temperature anomalies? I guess one could also conjure up a rationalization that local temperatures are not as precisely measured as their averages would be as represented by say the NH average, although I would expect something more imaginative than that.

  14. jae
    Posted Dec 2, 2006 at 10:10 PM | Permalink

    LOL: give him enough rope and he will hang himself.

  15. Posted Dec 2, 2006 at 10:45 PM | Permalink

    The paper has nice colors. PCA is minimized, the hockey stick is recycled with a postmodern reduction, and National Academy of Science and Wegman panels, much like M&M, don’t exist and never existed. All reconstructions have always given a hockey stick and all people who have ever questioned the unlimited sharpness of the stick, e.g. von Storch et al., are heretics who must be punished.

  16. Posted Dec 2, 2006 at 11:37 PM | Permalink

    No et.al. on the paper? Where have all his friends gone?

  17. Pat Frank
    Posted Dec 3, 2006 at 2:30 AM | Permalink

    LuboÅ¡ – 15 – there’s a real career opportunity here for a man of your talents. You can use Higgs theoretical formalism to derive a theory of the climate scalar temperature field (Tiggs field) that permeates the globe. It’s caused by the self-energy of reflected radiation beating against itself and decays rapidly to zero by 300 feet (maximal tree height), which is why it hasn’t been detected in the troposphere. This field will provide the theoretical basis for claiming that all trees everywhere (except in the second order quadratic divergence zone) produce rings of the same width no matter the local conditions. The field will operate by way, of course, of the Jones propagator, which is dipolar and restricted to the imaginary plane and for which all trees are an antenna (by excitation of cellulolignonic resonances). You can have a grateful and influential Michael Mann, physicist, as your co-author. It’ll mean instant tenure. :-) Don’t worry about having to build a Tiggs field detector for the skeptics. The hockey stick blade proves the existence of the field, which proves the validity of the blade, which proves … well, you know; all of AGW climate science works like that.

  18. Pat Frank
    Posted Dec 3, 2006 at 2:32 AM | Permalink

    That question mark in “Lobo?” should have been an s-chupchik, but the new format didn’t recognize it.

  19. Hans Erren
    Posted Dec 3, 2006 at 4:10 AM | Permalink

    I also see that the low countries (figure 3, page 12) are truncated at 1251, whereas data is available back to 751

    http://www.knmi.nl/klimatologie/metadata/nederland_wi_zo.html

    A.F.V. van Engelen, J. Buisman en F. IJnsen,
    A Millennium Of Weather, Winds And Water In The Low Countries, in:
    History and Climate: Memories of the future?
    Edited by Jones et al., Kluwer Academic/Plenum Publishers, 2001
    data:

    http://www.knmi.nl/klimatologie/daggegevens/antieke_wrn/nederland_wi_zo.zip

  20. Demesure
    Posted Dec 3, 2006 at 4:59 AM | Permalink

    #16 This paper is just a meta-collection of HS so what is the need of co-authors ? Besides, grant by the NSF for this work is “just” $146,755. You wouldn’t want Mann to share such a meager pie. He badly needs money for baby-sitters.

  21. Hans Erren
    Posted Dec 3, 2006 at 5:44 AM | Permalink

    $146,755: that’s 1.5 man year in my calculations (including overhead and taxes).

    I don’t think this paper has 1.5 man years of (new, unfunded by other grants) work in it.

  22. Demesure
    Posted Dec 3, 2006 at 5:50 AM | Permalink

    $146,755 for a meta-study, that’s the price of celebrity, Hans!
    Maybe the Royal Society should have a closer look to the money spent for junk science while sermoning Exxon for funding skeptics.

  23. Posted Dec 3, 2006 at 7:10 AM | Permalink

    I am definitely not trying to defend Mann here, but please keep in mind that about 30% – 40% of the grant goes to the institution in overhead. Some portion is used to fund graduate assistants. Of what is left over, a portion goes to stuff like new computers etc. Then, the left over usually pays for a couple of months of the researcher’s summer salary.

  24. Posted Dec 3, 2006 at 8:32 AM | Permalink

    No citation of M&M. Emperor Mann has a new invisible suit.

  25. Steve McIntyre
    Posted Dec 3, 2006 at 9:20 AM | Permalink

    #23. He also got $100,000 for new computers separately – http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=0548962

  26. Al Gore (the real one)
    Posted Dec 3, 2006 at 9:31 AM | Permalink

    NSF regulations pertaining to research misconduct were updated and published in the Federal Register on 18 March 2002. They include the following:

    Research misconduct means fabrication, falsification, or plagiarism in proposing or performing research funded by NSF, reviewing research proposals submitted to NSF, or in reporting research results funded by NSF.

    Fabrication means making up data or results and recording or reporting them.

    A finding of research misconduct requires that”¢’‚¬?
    (1) There be a significant departure from accepted practices of the relevant research community; and
    (2) The research misconduct be committed intentionally, or knowingly, or recklessly; and
    (3) The allegation be proven by a preponderance of evidence.

  27. Paul Linsay
    Posted Dec 3, 2006 at 9:47 AM | Permalink

    #17, Brilliant! A second paper will discuss plantelepathy and the relationship to Gaia…

  28. Hans Erren
    Posted Dec 3, 2006 at 10:23 AM | Permalink

    the gross monthly salary for a university professor in Holland is ”€š⪠8161 or annually ”€š⪠140,000
    that’s high compared to us averages of $100,000

    http://www.pitt.edu/~galletta/2006sals.html

    So a grant of $146000 pays a fulltime research of 1.5 man years. However considering it is a grant for extra work, one could hire five graduate students for the same amount.

    BTW, Is there turnkey software available?

  29. bender
    Posted Dec 3, 2006 at 10:32 AM | Permalink

    Review articles such as this one serve an important function. They are designed to punctuate the literature, rather than advance the field. This paper signals the end of the first chapter in proxy-based temperature reconstruction. Hallelujah.

    Chapter two began this summer. And you all know who is at the beginning of chapter two.

    If you want chapter two to read like it should, it will probably be necessary for Steve M to somehow get on the granting gravy train, in order to get away from always having to play catch-up with the Team, and to start innovating.

  30. Posted Dec 3, 2006 at 10:34 AM | Permalink

    #20.

    This paper is just a meta-collection of HS so what is the need of co-authors?

    It appears the paper is a ‘position statement’ as you say, you might expect other authors who have the same position to make the statement together. It costs nothing to add coauthors, he wouldn’t have to share funds to get coauthors. It can be the other way as people generally want to get more publications. Its kind of an idle speculation but really but could he really be on the out? Lets hope so.

  31. bender
    Posted Dec 3, 2006 at 11:04 AM | Permalink

    No, David Stockwell, it is not a “position statement”. It is a review article. Authorship on a review article is not a matter of soliciting who buys in to the review. It is a matter of who did the writing. Reviews are very often single-author, invited papers. Mann is not alone on the view presented in this article. Mann is not “on the out”.

  32. Posted Dec 3, 2006 at 11:27 AM | Permalink

    Dear Pat Frank #15,16, your ideas are really great but still, I will leave these breakthroughs and the related pies to others. ;-)

  33. EP
    Posted Dec 3, 2006 at 11:32 AM | Permalink

    Has this new Mann paper been accepted? How can he continue using questionable proxies without defending them initially?

    He seens to have put an underline under something that hasn’t been fully answered.

    Isn’t it therefore the task of his peers to point this out?

  34. bender
    Posted Dec 3, 2006 at 11:46 AM | Permalink

    EP, It’s more than just been accepted. It’s been published. That’s how Steve M can provide the citation that opens this post.

  35. Steve McIntyre
    Posted Dec 3, 2006 at 11:55 AM | Permalink

    It’s pretty amazing to see Mann’s PC1 continue to be used and illustrated. Usage of Mann’s PC1 has actually increased in Team-world since problems surfaced. Prior to the controversy, no third party had actually used Mann’s PC1 in one of the data-snooping exercises. In 2006, it was used in Osborn and Briffa 2006 (and then crops up in a NAS panel illustration as “Western U.S.”; it was used in Hegerl et al 2006; it was used in Rutherford et al 2005; and now recurs in Mann 2006. It’s more popular than ever.

  36. EP
    Posted Dec 3, 2006 at 12:54 PM | Permalink

    This is either a case of the left hand not knowing what the right hand is doing or poor reviewing. How do scientists point out material that doesn’t justify using questionable data?

  37. bender
    Posted Dec 3, 2006 at 1:27 PM | Permalink

    Re #36 The flaws in this work simply haven’t been proven & exposed to a sufficient degree yet for reviewers to know any better. One M&M GRL article is not a take-down. NAS & Wegman are only a start.

    Change only comes from within. The ruling orthodoxy itself needs to fracture before any major reform happens. M&M can can act as a wedge, but they do not have enough force to drive their own wedge.

  38. Posted Dec 3, 2006 at 1:32 PM | Permalink

    No signs so far of Mann, M.E. et al, Robustness of proxy-based climate field reconstruction methods, 2006 (accepted)

    But now it is in press:

    Mann ME, Rutherford S, Wahl E, Ammann C. 2006a. Robustness of proxy-based climate ï⪧?eld reconstruction methods. J. Geophys. Res. In press

    Can’t wait.

    The true value of p can in fact be estimated from the proxy data themselves, and Mann et al. (2006a) have estimated the average value of p for the full network of 112 proxy multiproxy indicators used by MBH98 to be p = 0.29 +- 0.03.

    Lots of basic stuff in the paper (my emph):

    In the CFR approach, hemispheric or global means, as well as any climate indices of interest, are computed directly from the spatial reconstructions of the underlying spatial ï⪧?eld, just as they would be for, for example, modern gridded instrumental climate records. CFR methods do not require that a proxy indicator used in the reconstruction exhibit any local correlation with the climate ï⪧?eld of interest , but instead make use of both local and nonlocal information by relating predictors (i.e., the long-term proxy climate data) to the temporal variations in the large-scale patterns of the spatial ï⪧?eld.

    Continued investigations of climate ï⪧?eld reconstructions should focus on the use of teleconnected local climate responses and the stability of the associated long-term relationships.

    The authors incorrectly implemented the MBH98 procedure,

  39. Pat Frank
    Posted Dec 3, 2006 at 2:55 PM | Permalink

    #38 – UC quoted, “CFR methods do not require that a proxy indicator used in the reconstruction exhibit any local correlation with the climate field of interest…

    This gives them carte blanche to use any proxy they like, so long as it gives the desired answer. It also absolves them a priori from any experiment that tests whether a given proxy is a faithful reflector of temperature. “Good” proxies now need only to follow the accepted (and also uncritiqued) grid-cell average. I can’t imaingine how that statement could get past the reviewers and the journal editor.

  40. Posted Dec 3, 2006 at 3:05 PM | Permalink

    # 39

    And note the p refers to proxy noise autocorrelation. But what is proxy noise, given that there is no requirement for local correlation? Is it individual proxy minus NH Temperature? This is one reason I can’t wait to see Mann 2006a. Maybe, after reading that, we will finally understand what is going on..

  41. Posted Dec 3, 2006 at 3:34 PM | Permalink

    CFR methods do not require that a proxy indicator used in the reconstruction exhibit any local correlation with the climate ï⪧?eld of interest

    I’m pretty sure that James Randi has a $1 million prize for Mann if he can prove that trees react to the global climate field without responding to local climate.

    This is the 21st Century and I’m reading about voodoo in a science journal.

  42. nanny_govt_sucks
    Posted Dec 3, 2006 at 3:36 PM | Permalink

    CFR methods do not require that a proxy indicator used in the reconstruction exhibit any local correlation with the climate ï⪧?eld of interest

    So it’s OK to use baseball batting averages, or % of pirates in the population as proxy indicators as long as they show some degree of correlation with the instrumental temperature record over the calibration period, right?

  43. bender
    Posted Dec 3, 2006 at 5:28 PM | Permalink

    Re #38 I could not find this paper, but I have full access to JGR. Do you know the volume and page numbers?

  44. Dave Dardinger
    Posted Dec 3, 2006 at 5:44 PM | Permalink

    re; #38ff

    What about the second half of the statement,

    (CFR methods…) make use of both local and nonlocal information by relating predictors (i.e., the long-term proxy climate data) to the temporal variations in the large-scale patterns of the spatial ï⪧?eld.

    Could someone translate this into plain english? I understand all the individual words and phrases, but the meaning of the entire statement eludes me?

  45. Posted Dec 3, 2006 at 7:24 PM | Permalink

    There seems to be an expectation that advocates of disaster when proven to have been mistaken about their dire prophecy will simply amend their beliefs to be consistent with the demonstrated facts. That simply isn’t the way humans behave. Th phenomenon is called Cognitive Dissonance. See http://en.wikipedia.org/wiki/Cognitive_dissonance.

    Doom sayers don’t recant their beliefs when confrounted with disconfirming evidence. They redouble their faith.

    Mann will go to his grave believing in the Hockey Stick. No one can dissuade him. Don’t bother trying. Concentrate on those who havn’t made up their minds yet.

  46. Rev Jackson
    Posted Dec 3, 2006 at 8:23 PM | Permalink

    By their fruits shall ye know them!!

  47. David Smith
    Posted Dec 3, 2006 at 8:54 PM | Permalink

    Re #44 What a sentence. The writer must have had a prior career writing the tax code.

    I have two possible interpretations:

    1.”CFR methods determine how proxies are affected by changes in their environment. The environmental changes can be local or distant.” Or,

    2.”Local changes in taxidermy affect the nuances flowing through the iceberg”.

  48. Stevan Naylor
    Posted Dec 3, 2006 at 9:12 PM | Permalink

    I’ll take number 2, Dave; for $200.

  49. Willis Eschenbach
    Posted Dec 3, 2006 at 9:23 PM | Permalink

    Mann makes some interesting claims about the correlation “r” in his paper, viz:

    For the reconstruction of Mann & Jones (2003), the average decadal correlation between the eight proxies used and the closest available instrumental annual mean surface temperature gridpoint record during the twentieth century is r = 0.47.

    and

    Based on this criterion, the average value for the Mann et al. (1998, henceforth MBH98) network of 112 indicators is r = 0.41 at annual timescales (and higher, at decadal and longer timescales) (Mann et al. 2006a).

    These numbers seem high to me, particularly given that the correlation of the 18 proxies used by Juckes with local temperatures is not statistically different from zero. Anybody know whether these claims are true?

    w.

  50. Willis Eschenbach
    Posted Dec 3, 2006 at 9:49 PM | Permalink

    Some absolutely priceless reasoning. The new Mann paper confirms its findings that MBH99 is really, truly correct, we’re not kidding this time, by comparing it against models forced with changes in historical solar and volcanic forcing.

    The historical volcanic forcings are from the paper by Crowley.

    Crowley estimated the size of the historical volcanic forcings by comparing the eruption sizes against the temperature record of …

    yep, you guessed it …

    MBH99.

    I weep for the passing of science …

    w.

  51. Pat Frank
    Posted Dec 3, 2006 at 10:50 PM | Permalink

    #50 — Sounds like Tiggs field reasoning to me: 1 because 2 because 1. Buck up Willis. Someone is gratified. The post-modern cultural-relativists are thrilled, for example. Mann, by your recounting, has truly turned climate science into cultural text. He’s proven their case. Here’s a peculiarity that demonstrates how bizarre P-M thinking can be: When Mann is finally proven wrong by actual science, that disproof will prove the P-M case. That is, exactly by, and coincidentally with, the analytical disproof, his work’s status as cultural text will be proven. See? Contemplate the intellectual train-wreck of P-M and be glad you’re only worried about science.

  52. Posted Dec 3, 2006 at 11:07 PM | Permalink

    #43

    bender, I quoted the ‘Climate Over the Past Two Millenia Paper’, sorry for confusion. Mann refers to this Robustness of proxy-based .. ‘in press’ paper many times.

    Me:

    But what is proxy noise, given that there is no requirement for local correlation?

    Already forgot the Ritson’s method. They don’t need reference signal to determine noise properties from the observations..

  53. bender
    Posted Dec 4, 2006 at 1:54 AM | Permalink

    Re #52
    Thanks, I realize that. No confusion. My point: if it is “In Press”, and it is slated for 2006, it should be available. So maybe it is actually slated for 2007 now?

  54. Posted Dec 4, 2006 at 2:30 AM | Permalink

    #53

    OK, so it cannot be published anymore in 2006? The ‘In Press’ paper is cited 6 times, and IMO it is very essential (‘A number of problems with theVon Storch et al. (2004) study have now been identified’ etc..) Kind of funny to cite a paper that doesn’t exist.

  55. Rev Jackson
    Posted Dec 4, 2006 at 3:20 AM | Permalink

    Re All Above: By their fruits ye shall know them!!

  56. Willis Eschenbach
    Posted Dec 4, 2006 at 4:25 AM | Permalink

    Further oddities from the Mann paper:

    The true value of à?’? [which he calls the "temporal autocorrelation coefficient", which I assume is the lag-1 autocorrelation] can in fact be estimated from the proxy data themselves, and Mann et al. (2006a) have estimated the average value of à?’? for the full network of 112 proxy multiproxy indicators used by MBH98 to be à?’? = 0.29 ± 0.03.

    This seems very low. The average lag-1 autocorrelation of the 42 proxies listed by Juckes in his archive (“mitrie_proxies_01.csv”) is 0.58 ±0.05.

    Steve M., do you have a file of the MBH98 proxies? And were there really 112 of them?

    w.

  57. Posted Dec 4, 2006 at 4:41 AM | Permalink

    Willis, I think he refers to proxy noise autocorrelation. To estimate that you need a reference. i.e. Noise=Proxy-Temperature. But as we know, there is no need for correlation with local temperature. Thus, to obtain noise p 1) they use NH Temperature or 2) they use Ritson’s method or 3) something I can’t even imagine.

  58. Jean S
    Posted Dec 4, 2006 at 6:39 AM | Permalink

    re #44/#47/#51: I think it would be about time to write “The Mannism Generator” (ManGe) according to the well known “The Postmodernism Generator”:

    http://www.elsewhere.org/pomo

    Any volunteers? If someone is up to the task, I think we could have a thread here over CA, where people could post mannian phrases (with references) to be added to ManGe. If the thing works well, maybe one should submit a sample paper to, e.g., Climatic Change ;)

  59. Willis Eschenbach
    Posted Dec 4, 2006 at 6:40 AM | Permalink

    Well, it’s not the Northern Hemisphere residuals. The Mitrie NH-proxy residuals (42 proxies) have an average autocorrelation of 0.49 ±0.02.

    I gotta say that I don’t see how the Ritson method would apply. Ritson says:

    We [I guess the Royal We, as he is the only author] assume the proxy records to be closely approximated by random ‘noise’ superimposed on a slow fluctuating signal component with comparatively large excursions over multi-decadal periods. Below is a ‘differencing’ technique used to remove the large variance highly correlated slow component from consideration prior to determining the AR1 autocorrelation component.

    Consider a proxy-site of N individual tree-ring records each of T years duration. The recorded growth-amplitudes are Xi(j) where i the record number goes from 1 to N and j the index of year goes from 1 to T.

    From this, it seems to me that the Ritson method, if it is valid, is only valid for a number of proxies from a single area, a “proxy-site” in their words. I don’t see how it could be used for a group of widely separated proxies. For a single site, it is possible that their assumption of a “large variance highly correlated slow component” in all of the proxies would be true.

    But for a widely separated group of proxies, each of which is subject to a different slow local temperature swing, there is no “highly correlated slow component”. Otherwise, we could just use a low-pass filter to remove the high-frequency component (which they call “noise”) from each proxy, and the resulting curves would all be the same … which is clearly not true.

    Finally, am I missing something here? It seems that what they are doing is determining the autocorrelation of the high-frequency component of the signal … I’m not clear on what the purpose of this calculation might be.

    w.

  60. Jean S
    Posted Dec 4, 2006 at 6:44 AM | Permalink

    re #50: Perpetual-motion machine! I can hardly wait for Crowley et al (2008) estimating the size of the historical volcanic forcings by comparing the eruption sizes against the temperature record of Mann et al (2007)…

  61. Posted Dec 4, 2006 at 6:59 AM | Permalink

    #59

    From this, it seems to me that the Ritson method, if it is valid, is only valid for a number of proxies from a single area, a “proxy-site” in their words. I don’t see how it could be used for a group of widely separated proxies. For a single site, it is possible that their assumption of a “large variance highly correlated slow component” in all of the proxies would be true.

    If the ‘signal’ is local temperature, then Ritson method underestimates p. Or at least I think so, some time ago wrote this http://www.geocities.com/uc_edit/ar_again.pdf ( but it is not peer-reviewed, Read at Your Own Risk)

  62. Brooks Hurd
    Posted Dec 5, 2006 at 12:26 PM | Permalink

    Re; 50

    Willis,

    I weep for the passing of science …

    As do I.

    Perhaps you have discovered the true meaning of “C” in “GCM”. It really stands for circular reasoning.

  63. Steve McIntyre
    Posted Dec 5, 2006 at 12:53 PM | Permalink

    Steve M., do you have a file of the MBH98 proxies? And were there really 112 of them?

    Willis, you’re reminding me of the good old days. Remember in the wake of MM03, when Mann said that MM03 made the “grievous” error of not using 159 proxies – a figure never previously mentioned. When I asked for an identification of the 159 series from Mann, he refused; Nature said that the figure of 159 series had nothing to do with them. A listing of series that appear to have been used in MBH98 is at the Corrigendum SI for MBH by step. There are some small curiosities as some series are used in (say) the 1450 step but not the 1500 step – which would be hard for anyone to replicate.

    112 series were used in the AD1820 step – which is where this figure comes from. 22 series were used in the AD1400 step. The PC series are re-calculated in some steps but not others; and retained according to an unknown and non-replicable procedure.

    If you want to look at data, look at the Nature SI. I really need to write up my emulation of MBH – I have lots of interesting notes. If you look at the Category MBH98 – Replication, there are some old posts that are mostly still valid.

  64. Dan Hughes
    Posted Dec 6, 2006 at 12:38 PM | Permalink

    wow … go see what Piekle Jr is saying in this thread, under: That Didn’t Take Long — Misrepresenting Hurricane Science, about recent statement by Mann.

  65. Mark T
    Posted Dec 6, 2006 at 1:05 PM | Permalink

    On down the page, in the reference to the WMO report he says:

    In particular, it should now be completely unambiguous that those who are representing hurricane impacts as being related to greenhouse gas emissions, without acknowledging that this is not a widely shared perspective among scientists, are either cherry picking the relevant science or misrepresenting the community consensus.

    Seems I’ve heard that term before.

    Mark

Follow

Get every new post delivered to your Inbox.

Join 3,330 other followers

%d bloggers like this: