Briggs on McShane and Wyner

As usual, a good analysis from Matt Briggs here

51 Comments

  1. Third Party
    Posted Aug 17, 2010 at 8:45 AM | Permalink

    “McShane and Wyner emulate Howe by applying a forearm check to the throat to Mann’s proxy reconstruction of temperature, cracking his hockey stick irreparably, leaving his models sprawling on the ice. ”

    Will uu@W get stitched up in the dressing room and return to the ice soon? Will the team enforcers get a simultaneous shift to beat up the new guys in the league?

  2. Ron Cram
    Posted Aug 17, 2010 at 9:14 AM | Permalink

    Steve,
    Thanks for drawing our attention to this. Matt is a great statistician and a fun writer, second only to ClimateAudit.

  3. bender
    Posted Aug 17, 2010 at 9:24 AM | Permalink

    Look at Briggs’ yellow flatline. The LIA is barely, if at all, discernibly cooler than the MWP.

    • BillN
      Posted Aug 17, 2010 at 10:19 AM | Permalink

      It’s worse than we thought; the “year without a summer” was not discernibly cooler than 1998!!!

      -BillN

  4. Scott Brim
    Posted Aug 17, 2010 at 11:05 AM | Permalink

    Could someone knowledgeable about the Mann et al reconstruction give us some general insight as to the provenance of the calibration period temperature data, especially its spatial distribution? Thanks in advance.

    • AMac
      Posted Aug 17, 2010 at 11:46 AM | Permalink

      Re: Scott Brim (Aug 17 11:05),

      In Mann08, AFAIK, all reconstructions calibrate proxies to the historical temperature models maintained by Hadley (CRUTEM3v for land and HADCRUT for land + ocean). These both come up with a monthly average temperature anomaly for the years 1850 to the present (more-or-less). The surface of the globe is divided into gridcells, each 5 degrees of longitude by 5 degrees of latitude. For each gridcell, average temperature anomalies are calculated from weather station records and from neighboring gridcells. Obviously, the farther back in time, the sparser the instrumental records.

      Mann08’s EIV-method reconstructions use a broad set of annual averages, one for the northern hemisphere (NH) and one for the southern, IIRC. Their CPS method locates each proxy in its 5×5 gridcell, and uses that gridcell’s imputed annual temperature anomaly. (As Steve McI has pointed out in numerous posts, Mann08’s methods have some non-intuitive twists and turns concerning this point.)

      For the Tiljander proxies from Southern Finland, I pulled the annual temperature anomaly for its gridcell location, and graphed it against the varved data series. The results are here. The point is that I have no coding skills, and was able to extract this record from the archived zipped files by hand (it’s a pain, though).

      Hope that helps (corrections welcomed). For discussions of the reliability, accuracy, and meaning of the instrumental historical record ~1890 to the present, I would look to Zeke Hausfather’s posts on the subject at “Lucia’s Blackboard.”

      • Scott Brim
        Posted Aug 17, 2010 at 3:10 PM | Permalink

        Thanks, AMac. What would also be very useful — for those of us who are curious as to how, specifically, the Mann et al reconstruction is organized — would be for someone knowledgeable to develop an end-to-end process flow sheet, written in the form of a topical outline, covering the major steps of the reconstruction plus the titles and the subtitles of the first and second sub-tiers of the reconstruction’s subsidiary processes and steps.

        • AMac
          Posted Aug 17, 2010 at 3:48 PM | Permalink

          Re: Scott Brim (Aug 17 15:10),

          I imagine that your satisfaction will depend on the hourly rate that you pay for that project–and your RFQ may not get many respondents! AFAIK, very few people have ventured to look carefully at Mann08’s inner workings. I believe they are Steve McI, RomanM and UC, and Jeff Id. Among pro-AGW Consensus advocates and scientists, Tamino worked through some Mann08-type reconstructions in his recent book review. I don’t know of others who have reported doing so. Though the methods of Kaufman09 seem similar, as do other multiproxy efforts from Prof. Mann’s lab (obviously).

          Steve: Kaufman’s methods are different from Mann08 – it’s pretty much CPS in the style of Bradley and Jones 1995 – other than smooth before scale. Yamal is very influential in Kaufman.

        • Scott Brim
          Posted Aug 17, 2010 at 11:15 PM | Permalink

          AMac, you are quite correct in recognizing how much effort would be required to develop an end-to-end process flow sheet for the Mann et al temperature reconstruction.

          However, I would point out that it is not possible to integrate the various audit findings against Mann et al into a logical, cohesive analysis of the reconstruction as a whole without having such a process flow sheet in hand.

          Some have suggested that Steve & Company here at CA should begin publishing their own papers on various climate science topics. In my opinion, there is no point to this. Such papers would add no further value to the work that has already been done.

          The logical next step for the Climate Audit senior staff is to integrate their findings into a knowledge-managed topical framework which examines the Mann et al temperature reconstruction as a whole.

          A process flow sheet is a necessary starting point for that kind of effort.

        • Atomic Hairdryer
          Posted Aug 18, 2010 at 4:29 AM | Permalink

          What ‘senior staff’ do you think exist? The idea of a process flow sheet is nice, but you say

          “What would also be very useful — for those of us who are curious as to how, specifically, the Mann et al reconstruction is organized ”

          Unfortunately Mann et al did not provide that information, hence the decade or so effort to try to reconstruct, reproduce and validate. If you want to understand more of the history, try reading The Hockey Stick Illusion.

        • Scott Brim
          Posted Aug 18, 2010 at 9:31 AM | Permalink

          You are correct, Mann et al did not provide the information needed to create a step-by-step, end-to-end process flow sheet for their temperature reconstruction.

          In the nuclear industry, Mann et al could not have done such a poor job of documenting their work without being called to account for it.

          What I want to see here is a nuclear-grade auditing process, and that means moving to the next higher plane of auditing theory as it applies to scientific and technical projects.

          In the nuclear industry, quality assurance and quality assurance auditing are not merely administrative “check the box” types of processes.

          QA is a tool for communicating to management whether or not they are actually getting the quality they expect. In this case, “management” is the tax-paying public, since climate science is largely a government funded enterprise.

          The Hockey Stick Illusion has its place, certainly, but it does not represent the kind of systematized, knowledge-managed review process that is necessary for the auditors to build their own reputations as trustworthy auditing professionals.

        • Andrew Russell
          Posted Aug 18, 2010 at 10:54 AM | Permalink

          I believe WattsUpWithThat (or maybe this site) had a posting some months ago on how to re-create MBH98 using OpenOffice, R, and the proxies Steve M. dug out of the “CENSORED” direcotry. But I can’t seem to find that posting/process now – anybody know where it is?

          And to repeat, “The Hockey Stick Illusion” (which I’ve almost finished) is not an engineering process manual, but a history book that reads like a ‘whodonit’. But it reveals much about the refusal of ‘climate scientists’ to follow the Scientific Method, much less any engineering method.

        • Scott Brim
          Posted Aug 19, 2010 at 10:17 AM | Permalink

          Re: Andrew Russell (Aug 18 10:54),

          It is my opinion as a former QA auditor in the nuclear industry that a systematized, end-to-end documented review of Mann et al which is topically structured in direct alignment with the reconstruction’s process flow sheet would add a level of professionalism to the endeavor which is not possible to attain by writing competing research papers; by writing critical books about the hockey stick; or by writing a series of audit findings against various climate science sub-topics and posting these findings helter-skelter on the Internet.

          The auditing efforts of Steve & Company against Mann et al and The Hockey Team will not attain critical mass as a policy influencing endeavor unless the individual findings and their supporting information are integrated into a single product which presents a unified picture of the Mann et al temperature reconstruction — not only how, technically, the reconstruction is being done; but also how the various audit findings relate to each other and also to the temperature reconstruction as a whole.

          Such an auditing document could be implemented in the form of a series of hyperlinked web-pages, organized according to the process flow sheet, and with links to both the reconstruction’s technical detail and to the findings which relate to that particular technical detail.

          An integrated, well structured, professionally written auditing product done in the form of a topical outline which directly follows the temperature reconstruction’s process flow path will be more than the sum of its parts ….. much, much more, I think.

  5. RayG
    Posted Aug 17, 2010 at 11:09 AM | Permalink

    The articles cited in Briggs’ second footnote are also worth reading. The usual enforcers are circling their wagons (pardon the mixed metaphor) at RC and Tamino. They aren’t fast enough to get within reach of Old #9 and son.

    • Johan C
      Posted Aug 17, 2010 at 11:35 AM | Permalink

      Old #9 and son ???

      • David Jay
        Posted Aug 17, 2010 at 11:45 AM | Permalink

        Gordie Howe

        • Hoi Polloi
          Posted Aug 17, 2010 at 2:55 PM | Permalink

          Gordie Who? Sorry, never heard of the guy but outside Canada and maybe parts of US hockey is only known as field hockey. Gretzky rings a bell though. Never to old to learn something, though.

          ”Hockey would be a great game… if played in the mud.” – Jimmy Cannon

        • Steven Mosher
          Posted Aug 17, 2010 at 11:42 PM | Permalink

          Im from michigan and I have to say that the ‘great one’ was
          my favorite. even in LA.

        • Doug Badgero
          Posted Aug 18, 2010 at 10:20 PM | Permalink

          Also from MI – One word: Yzerman

      • Stan Plamer
        Posted Aug 17, 2010 at 11:48 AM | Permalink

        Gordie Howe and his son were both famous hockey players

        Howe’s number was 9

        Both he and Rocket Richard wore 9. This made it a famous number in hockey. Bobby Hull chose 9 to emulate them Wayne Gretzky chose 99. Many otehr players have done the same.

        Gretzky was The Great One

        Howe was The Greatest of Them All

        • Stan Plamer
          Posted Aug 17, 2010 at 11:51 AM | Permalink

          Briggs described the M&W paper with a hockey metaphor. Following that metaphor, it would be interesting to see Gavin’s Bulldog go into the corner with Howe. So M&W as Gordie Howe going into the statisical corner with Tamino

        • Tom
          Posted Aug 17, 2010 at 12:13 PM | Permalink

          Tamino won’t stray from RC and his blog. He would never dare get into a dog fight over this on a forum he cant control.

        • Chris S
          Posted Aug 17, 2010 at 12:15 PM | Permalink

          Unfortunately for most in the UK, the hockey metaphor makes the article harder to understand than the paper being discussed.

          And I thought only girls played hockey;)

        • Kevin Lohse
          Posted Aug 17, 2010 at 12:30 PM | Permalink

          Actually, there is strong support for Ice Hockey being invented by British Army officers snowed in in Quebec during the French Wars. Canuks generally ignore the tale.

        • Stan Plamer
          Posted Aug 17, 2010 at 12:48 PM | Permalink

          The hockey rink is surrounded by boards. The shape of the rink is a long rectangle with two corners at each end near the goal. One defensive strategy is to clear the puck to a corner since the boards restrict movement there. Forwards must go onto the corner to struggle for the puck with one of the defencemen. The defenceman cannot afford to let the forward get teh puck since he can then pass it in front of the goal. The straggle for the puck in the corner is thus very violent. It is considered to be the most violent part of the game. Many players will not go into the corner for fear of the struggle that they will be subjected to.

          Howe was a forward for Detroit. He had no fear of going into the corner. He is regarded as one of teh strongest players ever to play the game. The question is if Gavin’s Bulldog is prepared to go into the statistical corner with M&W.

        • Johan C
          Posted Aug 17, 2010 at 1:58 PM | Permalink

          Thank you! I admit the whole hockey analogy was lost on me.

        • RayG
          Posted Aug 17, 2010 at 2:34 PM | Permalink

          For the Brits out there, think of a combination of Bradman, Lara and Gooch which fits in with the “sticket wichet” metaphor used in describing the MW 2010 paper.

        • DBD
          Posted Aug 17, 2010 at 2:49 PM | Permalink

          Maybe witha little Botham too (if there is such a thing as a ‘little’ Botham).

        • SOI
          Posted Aug 17, 2010 at 3:07 PM | Permalink

          Gooch? Did you say Gooch? In the same sentence as Bradman and Lara? Bowlers still talk with fond memories about what they call the golden age of Gooch.

        • See - owe to Rich
          Posted Aug 17, 2010 at 3:59 PM | Permalink

          Look, don’t diss Gooch too much. There aren’t many batsmen who have scored 456 (333+123) in a single Test. The sheer poetry of those numbers places him in the pantheon.

          Rich.

        • RobT
          Posted Aug 17, 2010 at 4:02 PM | Permalink

          Re: Stan Plamer (Aug 17 11:48),
          Carrying on with the hockey metaphor, replace Belak with Mann, Tamino, or whoever you choose:

          http://www.hockeyfights.com/fights/93965

  6. Ursus maritimus
    Posted Aug 17, 2010 at 12:23 PM | Permalink

    I nominate Steve for the Lady Byng.
    Keep you stick on the ice :)

    • RayG
      Posted Aug 17, 2010 at 1:36 PM | Permalink

      Seconded with Honorable Mention to Andrew Montford and Lord Moncton.

  7. Ursus maritimus
    Posted Aug 17, 2010 at 12:25 PM | Permalink

    http://en.wikipedia.org/wiki/Lady_Byng_Memorial_Trophy

  8. Kenneth Fritsch
    Posted Aug 17, 2010 at 2:18 PM | Permalink

    For all you British bloggers out there, Gordie Howe was the Babe Ruth of ice hockey and ice hockey when it was played without masks and helmets. And I say that as Blackhawks fan. Bloody girls’ game indeed.

    • Speed
      Posted Aug 17, 2010 at 3:00 PM | Permalink

      “For all you British bloggers out there,” Gordie Howe was ice hockey as WG Grace was to Cricket.

      • GrantB
        Posted Aug 17, 2010 at 10:34 PM | Permalink

        Dr WG Grace was bowled out on the first ball of a charity match, but continued to play, exclaiming “They came to see me bat, not to see you umpire”.

        Somewhat like the team’s approach to scrutiny.

      • Bernie
        Posted Aug 18, 2010 at 6:38 AM | Permalink

        Dr Grace was a gentleman – largely. Gordie Howe was the Billy Bremner or Nobby Stiles or, for you younger folks, Roy Keane of Hockey.

    • Gary
      Posted Aug 17, 2010 at 3:38 PM | Permalink

      Gordie Howe ( http://en.wikipedia.org/wiki/Gordie_Howe ) played in five (!) different decades. Unprecedented.

      Steve M. reminds me a little of Bobby Orr – the first great offensive-minded defenseman – who could play both ends of the rink with ease.

    • James Evans
      Posted Aug 18, 2010 at 1:45 AM | Permalink

      Who’s Babe Ruth?

      • PhilJourdan
        Posted Aug 18, 2010 at 11:41 AM | Permalink

        Did you audition for the movie “The Sandlot”? LOL

      • TGSG
        Posted Aug 18, 2010 at 4:10 PM | Permalink

        ok, intentional or not, that was funny right there.

  9. RayG
    Posted Aug 17, 2010 at 2:35 PM | Permalink

    Oops, meant “wicket.”

  10. RayG
    Posted Aug 17, 2010 at 2:37 PM | Permalink

    trying again…”sticky wicket.” Don;t think that I will ever truly get the connection between brain and fingers on keyboard correctly. Of couse if I did and it was plotted, it would look much like a hockey stick.

  11. PhilJourdan
    Posted Aug 17, 2010 at 3:20 PM | Permalink

    Howe was before my time, but I followed Gretzky! He will always be the “great one” when I hear Hockey

  12. Salamano
    Posted Aug 17, 2010 at 3:33 PM | Permalink

    It appears this study is going like wild-fire through the usual suspects’ websites.

    And…speaking about ‘Usual Suspects’…Joe Romm has something up on it (if y’all don’t already know)…

    http://climateprogress.org/2010/08/16/hockey-stick-paper-mcshane-and-wyner-statisticians/

    For all his talk about non-climate scientists butting into climate science, he seems to be making some pretty bold statements as a non-statistician wading into deep statistical waters.

  13. hunter
    Posted Aug 17, 2010 at 5:33 PM | Permalink

    a question has been raised:
    why did MW not use a fourier transformation to analyze the data?
    If not, why not?
    The secondary question for this would be:
    Did Mann use the FT, and if not why not?
    Any insights on this will be greatly appreciated.

    • Eddieo
      Posted Aug 18, 2010 at 11:33 AM | Permalink

      I don’t know what this would add as FTs are usually used to identify periodic signals not Hockey Sticks. It could be useful for identifying periodic fluctuations in Ice Core records etc. or even the instrumental record.

  14. Doug Badgero
    Posted Aug 17, 2010 at 6:48 PM | Permalink

    Hunter,

    While it may be an interesting exercise I am not sure it would add anything to the debate since the underlying data is not periodic. We would all just be arguing about what the output of the domain shift “meant”.

  15. Earle Williams
    Posted Aug 18, 2010 at 2:47 AM | Permalink

    hunter,

    A better question to ask is why M&W or anyone would perform a Fourier transform on the proxy data. Whomever raised it apparently doesn’t understand frequency domain signal processing. The question is somewhat akin to the proverbial Star Trek query of “Did you try reversing the polarity?”

    I’m loathe to attempt to offer insights, as the depth of topic is far beyond a blog comment. Just google the wiki page for Fourier transform and read any three of the references for a taste of the topic. But relevance to M&W? None.

    • Andrew Russell
      Posted Aug 18, 2010 at 11:51 PM | Permalink

      Re: Earle Williams (Aug 18 02:47),

      Why perform Fourier transforms? As one of my favorite programming books(“Real-Time Programming: Neglected Topics”)says about doing those on any time series, “With a sufficiently warped mind, any signal is repetitive”. I’m being facetious, but being weak in statistical methods, the description of principle components in “The Hockey Stick Illusion” did make me think they sounded a bit like FFTs.

2 Trackbacks

  1. [...] shows that the public are no longer supporting the scare. And though the science has been shown to be [...]

  2. By Top Posts — WordPress.com on Aug 17, 2010 at 7:14 PM

    [...] Briggs on McShane and Wyner As usual, a good analysis from Matt Briggs here [...]

Follow

Get every new post delivered to your Inbox.

Join 3,295 other followers

%d bloggers like this: