The Boulton Hockey Stick

Following (below left) is the Hockey Stick diagram endorsed by Geoffrey Boulton, General Secretary of the Royal Society of Edinburgh, in their December 2009 (post-Climategate) Policy Advice statement Climate Change and the U.N Copenhagen Summit here. On the right is a confirmation plot from data archived in late 2009 and February 2010 by Boulton’s associate and 2007 hire at the University of Edinburgh, Gabrielle Hegerl (showing that I’ve located the precise data version for the Boulton hockey stick.)

Boulton Figure 3 Caption: Estimates of mean decadal temperatures over the land areas between 300 and 900 in the northern hemisphere during the last 1500 years. Prior to the instrumental record of the last 150 years (shown in red), temperatures are deduced from tree-rings, lake sediments and ice cores. The dashed lines show the range of higher frequency variability in the data. The record shows an early mediaeval cool period from prior to about 950AD, a mediaeval warm period until about 1200AD, the so-called Little Ice Age from about 1450 to 1850AD and the very strong late 20th Century warming. Temperatures in sub-surface rocks can be used to deduce long-termchanges in surface temperature that naturally smooth out inter-annual variations to show long-term trends. Temperature records from 631 boreholes have been used in this way to show how distinctive the 20th Century warming has been compared with the preceding 400 years. (From: Hegerl, G.C. and others. 2007. Detection of human influences on a new, validated 1500-year temperature reconstruction; Journal of Climate, 20 (4): 650-666.)

Boulton’s Policy Advice assured his readers that this reconstruction was “independent of the University of East Anglia reconstruction, about which there has recently been much controversy”:

Several independent estimations have now been made of the global or hemispheric average temperatures for the last two millennia. Figure 3 is one of these, and shows that the late 20th Century warming has been rapid and large compared with earlier periods (note that this is independent of the University of East Anglia reconstruction, about which there has recently been much controversy). If we look in more detail at the 20th Century warming however (Figure 3),we see that the pattern of climate change has been much more complex than the smoothly accelerating pattern of greenhouse gas concentration

Hegerl et al (J Clim 2007) describes a couple of different versions of her reconstruction: one step starts in 1505 (12 records) ; one in 1251; one in 946 (7 records) and one in 558 (5 records). Hegerl did not provide accurate digital or even paper citations for the records; the archive is smoothed data. The task of identifying the provenance of each of the series is further complicated by the fact that the only information about the smoothing is that it is “decadally smoothed” – the exact filter is not reported.

The version shown in the Boulton Policy Paper is the 5-record version starting in 558. I’m now in a position to identify the provenance of each of these 5 records – something that takes a lot of patience.

The first record in this group is described by Hegerl as follows:

Western United States: This time series uses an RCS processed tree-ring composite used in Mann et al. (1999), and kindly provided by M. Hughes, and two sites generated by Lloyd and Graumlich (1997), analyzed by Esper et al. (Boreal and Upper Wright), and provided by E. Cook. The Esper analyses were first averaged. Although there are a number of broad similarities between the Esper and Hughes reconstructions, the correlation is only 0.66. The two composites were averaged.

CA readers, but apparently not rigorous Journal of Climate peer reviewers, know that MBH99 goes back only to 1000 and that there is no candidate “RCS processed tree ring composite” in MBH99. Needless to say, this is Mann’s infamous PC1 (from Mann and Jones 2003, not MBH99)- accepted without demur by rigorous Journal of Climate peer reviewers even though Mann’s PC1 and the use of strip bark had been sharply criticized in the NAS panel report. In the CH5 reconstruction, only the PC1 is used (the strip bark foxtails are not averaged in.)

Next is a series described as follows:

Northern Sweden: This is from Grudd et al. (2002) by way of Esper.

This is Tornetrask – a site used in every multiproxy reconstruction that I know of – including, for example, MBH99, Jones et al 1998 and Briffa 2000. Esper’s RCS chronology is only slightly different than the RCS chronology from Briffa (2000). The measurement data used in Briffa 2000 wasn’t archived, but based on the measurement data in Briffa 2008, it looks like the Briffa 2000 and Esper measurement datasets matched. RCS methods are pretty trivial mathematically and thus there isn’t a lot of difference between the two chronologies. This is a ring width chronology (not an MXD density chronology) and, in this case, wasn’t bodged.

The third record is also familiar to CA readers:

Taimyr Peninsula: This is from Naurzbaev et al. (2002) by way of Esper.

Taymir, like Tornetrask, is a Briffa 2000 site and used in most recent multiproxy reconstructions. The Esper version of measurement data is smaller than the Briffa 2008 version, which pulls in some Schweingruber sites (the Briffa 2000 version has never been archived, but is presumably fairly similar to the Esper version.)

The fourth record is a West Greenland isotope series from Fisher et al 1996. This is also used in virtually every multiproxy study: MBH99, Jones et al 1998, Mann and Jones 2003, Moberg 2005, etc.

The 5th record is a Chinese composite from Yang et al 2002, used in many studies. This is used in many recent multiproxy studies as well.

East Asia: This is the high-resolution record (10-yr average) from Yang et al. (2002).

Although this supposedly has 10-year interval, it used the Thompson Dunde ice core previously smoothed to 50-year intervals.

The serial re-use of these proxies is very familiar to CA readers – a point confirmed by Wegman et al (2006).

Far from the Boulton hockey stick – a composite of the Mann PC1, Tornetrask, Taymir, West Greenland isotopes and the Yang composite – being “independent” of the controversial East Anglia reconstructions (regardless of whatever Boulton had in mind here precisely), the Boulton hockey stick is not “independent”.

Boulton and the Team that can’t shoot straight.

See CA category here for prior posts on Hegerl.


  1. jae
    Posted Feb 14, 2010 at 6:04 PM | Permalink

    Unbelievable! Will they ever stop publishing new studies based on the same tired “specially selected” proxies? Does this make Boulton part of the Team?

  2. bob
    Posted Feb 14, 2010 at 6:23 PM | Permalink

    These proxies have turned into pure gold, furnishing a never ending revenue stream for lots of hungry scientists.

  3. Posted Feb 14, 2010 at 6:23 PM | Permalink

    Something not right –

    I’m not able to view either of these images correctly on this page:

    • Posted Feb 14, 2010 at 6:25 PM | Permalink


      when I click on the links in my above reply, they both show perfectly.

  4. Posted Feb 14, 2010 at 6:25 PM | Permalink

    I think there is a fundamental disconnect between what Boulton perceives is our understanding of the ‘ancestry’ of such data sets and his understanding. Either he is under rating our understanding – or he is over rating his.

    Its very hard to use smoke and mirrors when the winds of change are blowing and people are throwing well aimed stones..

    Its getting all a bit fur coat and no knickers for these guys..

  5. Henry
    Posted Feb 14, 2010 at 6:25 PM | Permalink

    There is another coincidence. I see 14 proxy series and 2 instrumental series in the archived data (table titled “Scaling pr#9D9C3”). The instrumental series run from 1861 to 1995/96. The starting points of the proxy series vary from 513 to 1505, but 13 of them end in 1960 (the remaining one in 1975). This seems odd for a 2007 paper, whether or not there is a divergence problem. It is difficult to see what is happening in Figure 3 of the Royal Society of Edinburgh briefing because of overlapping lines, but I would guess that the blue line (which starts around 550: 558 according to the title) stops in 1960 too, hidden under the red and black lines.

    I seem to recall that the practice of ending proxy series in 1960, and then covering ends of lines with other lines on charts, was something CRU did.

    Steve: Hegerl did indeed cut off the series in 1960. I don’t know why in her case. She’s got MAnn’s PC1, Yamal, the Yang composite, Luterbacher – and doesn’t have the large density data set. It’s not as though this data set turns down. I don’t know why she did it right now.

  6. Posted Feb 14, 2010 at 6:36 PM | Permalink

    Independent means I fed the same garbage in and got out slightly different garbage. Independent means we used the same garbage, but we published at different times. It means we used the same garbage, but have different names for that garbage.

    • Bernie
      Posted Feb 14, 2010 at 10:01 PM | Permalink

      I was watching Casablanca this evening. That immortal line by Claude Rains as Captain Louis Renault seems to fit rather neatly – “round up the usual suspects!”

  7. John G. Bell
    Posted Feb 14, 2010 at 6:40 PM | Permalink

    Let us say that climate science starts to mend itself and two years from now this is among the retracted papers. Would you have any sympathy to attaching reviewers names to the retractions? Only serious and obvious flaws would would reflect badly on the reviewer so reputations would be protected among peers.

    As you know, I am not a scientist. Would you say that the errors you found were obvious? Perhaps you are more familiar with the material and issues than most reviewers?


    • Posted Feb 14, 2010 at 6:56 PM | Permalink

      They certainly should be obvious to the author(s) of the papers. Not sure how obvious they should have been to the reviewers. But, since it doesn’t seem like any of the reviewers requested archived data, one can wonder at how rigorous they could have been.

      Steve: Please look at the Hegerl category (left frame) for prior posts.

  8. Barclay E MacDonald
    Posted Feb 14, 2010 at 7:16 PM | Permalink

    There are very few people who could put this together. Thank you for your continued efforts to clarify what few of us would ever be able to understand.

    Your argument regarding lack of independence is persuasive. In addition some pertinent links to previous analysis of the 5 series would of course be helpful to better understand Dr. Boulton’s analytical skills in choosing the combined results of these specific series to present to Copenhagen and the World. Such a review may further help us to understand why Dr. Boulton is so convinced of the necessity of his placement on the Muir Russell panel. Exemplary work!

  9. ZT
    Posted Feb 14, 2010 at 7:49 PM | Permalink

    Boulton’s December 2009 post-climategate ‘Policy’ Advice statement contains a few Pachauri-isms


    The probability that impacts elsewhere will have
    major indirect consequences for Scotland can be
    assessed by exploring the continuation of
    well-established present trends and by simulations
    of future climate. By 2020, yields from rain-fed
    agriculture (the dominant method) in some African
    countries could be reduced by up to 50%. The global
    trend of decay of glaciers and snow cover is already
    reducing melt water from major mountain ranges.
    Glacier melting in the Himalayan & associated
    mountain chains currently supplies most of the
    dry-season flow of the great rivers ofAsia (Indus,
    Ganges,Brahmaputra,Mekong,Yangtze,Yellow) on
    which 20% of humanity depends for much food and
    energy production. It is estimated that continuation
    of current trends could reduce dry-season flow of
    those rivers by more than 60% by 2050.More than
    20 million people were displaced by sudden
    climate-related disasters in 2008 alone.An estimated
    200 million people could be displaced as a result of
    climate impacts by 2050.


    So Boulton, the expert glaciologist, was actually behind the Independent who reported that the Himalayan melting scare was bogus in November 2009


  10. MikeN
    Posted Feb 14, 2010 at 8:44 PM | Permalink

    You say Briffa 2000 wasn’t archived. Didn’t Briffa 2008 use the same proxies, making the archived data inherited to Briffa 2000?

    There were three 2008 datasets: Yamal, Tornetrask-Finland and Taymir. Yamal was the same; the Finland addition in 2008 was distinct; it’s non-trivial trying to detect the 2000 Taymir layer in the expanded 2008 dataset.

  11. Jimchip
    Posted Feb 14, 2010 at 8:54 PM | Permalink

    Mr. McIntyre,
    I have one specific question: Did anyone like Hughes, Cook, or Esper kindly provide you data, assuming you asked, ala “and kindly provided by M. Hughes”.

    My comment on “Independent” is similar to a previous comment on Jonesisms. I’ll call these ‘Teamisms’. Independent means that, although the underlying data is nearly identical (within some choices of exact datasets used), there was a (slightly) different value-added method used which makes the results independent.

    There is a small chance that independent meant “in this case, wasn’t bodged.”

  12. Shallow Climate
    Posted Feb 14, 2010 at 9:25 PM | Permalink

    Here’s an idea that will never fly, but wouldn’t it be A-OK if it did: Instead of saying “independent”, we say “___% independent” (again, not that anyone on the Team would ever say it). So, in the present case, Boulton/Hegerl might say, “this finding/reconstruction is 0.25% independent of others previously published”. I mean, ya know, this is alleged to be science and all that, so what’s with the use of words in some generic, quasi-poetic sense, when science would dictate that numerical citations always be used? Let’s do it with “independent” and also with (I’ve grown to disdain this word and I’m going to throw up right after I force myself to say it) “robust”. Again, I am grateful to CA.

    • WillR
      Posted Feb 16, 2010 at 12:23 AM | Permalink

      Re: Jimchip (Feb 15 02:06),

      A couple of interesting links in this context…

      Salami Slicing — What is salami slicing? Simply put, this involves publishing separate, but similar articles that rely on the same set of data. Researchers may slice the salami to increase their publishing output and drugmakers may view this as a way to promote useful findings for their meds.

      Study Design and Ethics

      Good research should be well justified, well planned, and appropriately designed, so that it can properly address the research question.

      • Posted Feb 16, 2010 at 2:46 AM | Permalink

        All the abuses I ever saw had to do with design of experiments.

        • WillR
          Posted Feb 16, 2010 at 6:53 PM | Permalink

          Re: thecrutapeletters (Feb 16 02:46),

          My first statistics prof said essentially the same thing. He turned red and spit while he explained that you should decide your statistical analysis tactics first — while you decided what data should be collected and what you wanted to prove. He made it clear that “wanting to prove” and “proving” were independent issues — which is why you had to state your tests up front — so you could decide “success or failure” by reviewing the results — not by deciding which results to present to prove your views. He was a really nice fellow and a good prof — but he had no respect for people who tried to “rescue” projects with statistics — nor did you get approval for “data torture”. He was a good prof, a good researcher, a good designer and — a very nice person. Funny how it all goes together…

    • SteveGinIL
      Posted Feb 16, 2010 at 7:28 PM | Permalink

      Re: Jimchip (Feb 15 02:06),

      One possibility is that the team was necessary so they could data recycle amongst themselves, appearing to be independent, and not be caught individually for self-plagiarism.

      The Team’s history is one of them being allowed to peer-review each other and claiming independent review, so they were data recycling and then circle-jerking the world with their circle-reviewing. Round and round they go.

      Does it ever stop? It is the gift that keeps giving.

  13. Al
    Posted Feb 15, 2010 at 12:37 AM | Permalink

    The talking point counter to the recent climate ‘science’ revelations goes; ‘the conclusions are supported by MANY other independent sources.’ Is this really the case? Is it possible to establish that there aren’t ‘many’ other sources and, like the Boulton rehash, these confirmations are actually just new renderings of the same data?

    I’m reminded of this[1] diagram that attempts to illustrate the history of a family of related operating systems. I don’t think this specific format would work, but my intent should be clear; challenge the assertion that there are many independent confirmations of AGW by illustrating the dependency of these ‘results’ on the same few warmed over data series.


  14. Steve McIntyre
    Posted Feb 15, 2010 at 9:02 AM | Permalink

    Jimchip, interesting reference to Data Recycling in medicine – this is exactly on point.

    The references place the highly unusual referencing of Wahl and Ammann into a sharper framework. Wahl and Ammann re-cycled arguments originally advanced by Mann in his 2004 Nature Response and later at realclimate; they didn’t cite the earlier analyses nor even acknowledge Mann. The purpose was (unusually) to appear more independent than they really were, rather than usual purpose in a case where plagiarism is alleged.

    • Jimchip
      Posted Feb 16, 2010 at 5:47 PM | Permalink

      Re: Steve McIntyre (Feb 15 09:02),

      This is a database search engine that I first saw referred to in a Science awhile ago. It was an article discussing self-plagiarism. I only used it to it to look up friends and associates at the time 🙂

      eTBLAST 3.0:
      a similarity-based search engine

    • WillR
      Posted Feb 16, 2010 at 6:47 PM | Permalink

      Re: Steve McIntyre (Feb 15 09:02),

      I really prefer the more Colourful “Salami Slicing” term — kinda makes you want to sit down and have a lunch and a beer with your “fellow colleagues”. 🙂 It is much better than the “Least Publishable Unit”…

      “Salami publication,” or “salami slicing,” is a variant of the smallest-publishable-unit strategy. In salami slicing, data gathered by one research project are separately reported (wholly or in part) in multiple end publications. Salami slicing, apparently named by analogy with the thin slices made from larger pieces of salami meat, is generally considered questionable when not explicitly labeled, as it may lead to the same data being counted multiple times as apparently independent results in aggregate studies

      But probably the LPU is the closest… though Google has a fair bit on the Salami

      To each their own.

      • gimply
        Posted Feb 16, 2010 at 7:32 PM | Permalink

        The long list of Weaver pubs should provide at least a little accompaniment for a dollop of dijon – haven’t looked but there are just *so many* of them.

      • WillR
        Posted Feb 16, 2010 at 10:05 PM | Permalink

        Re: Jimchip (Feb 16 19:15),

        Maybe we should be looking at it this way… we have an initial (or master study) — the bookends. Then we slip a new slim volume between the bookends, amongst the other volumes and call it a whole new library.

  15. John Murphy
    Posted Feb 16, 2010 at 2:31 AM | Permalink


    Hang on. How does the Boulton hockeystick sit with Jones’ admission to teh BBC that there has been no warming in the last 15 years?

    It surely can’t. Are they both lying? Only one? Which one?Why?

  16. Posted Feb 16, 2010 at 6:04 PM | Permalink

    A thread brought me over to this blog site. Scrolling down I found this discussion of the Boulton Hockey Stick. Our local paper in Abuquerque yesterday has an op-ed on the The Real Cliamte Controversy. The essence of the authors piece is that the Wegman report to Congress was a result of plagaried material taken from atextbook written by one of the authors of the Hockey Stick. In addition, the author claimed that an independent investigator, John Mashey, has a report that proves that the Wegman committee had committed a crime by falsely and misleading congress. I am puzzled by this accusaton by Mashey now since the original report was issued in 2006. Why now?
    I haven’t seen any of this controversy mentioned before on any of the AGW skeptic blogs. Does anyone know if there is a crimal investigation into the Wegman committee?

    Steve: This is stupid nonsense. You can find some discussion of it over at Lucia’s – you can discuss it there if you want.

    • Dave Dardinger
      Posted Feb 16, 2010 at 6:49 PM | Permalink

      Re: mandolinjon (Feb 16 18:04),

      I haven’t seen any allegations, but it sounds silly to me. The purpose of Wegman’s report was to show that Mann was or was not using proper statistical analysis. The easiest way to do this is to cite proper (and know improper) methods. The way you do this is to present material from textbooks and journal papers. If Wegman didn’t actually provide the textbook or paper used that’s sloppy, but it would take away from what he was trying to do, so I rather doubt he did so. It may be that some versions didn’t include the bibliography provided, but that would need to be shown.

      As for this Mashey report I’ve never heard of it. A citation, if available would be useful.

      • RomanM
        Posted Feb 16, 2010 at 8:24 PM | Permalink

        These spurious allegations of “plagiarism” seem to have originated at DeepClimate and were picked up by several other AGW blogs.

        The claims were made regarding Wegman’s paraphrasing of some verbally obtuse descriptions of the principles behind the use of tree rings in paloeclimatology from a book by Raymond Bradley. Anybody with a brain larger than that of a chipmunk could recognize that it had zero impact on the statistical analysis, however, the proprietor of the DC blog who is weak on science and strong on sophistry seemed to think otherwise. Whether or not a reference was given for the material is completely irrelevant (although the source was in fact referred to in another part of Wegman’s report).

        It is a classic example of distraction from the meat of the report to an inconsequential side issue and indicates just how short on decent arguments people such as DeepClimate are. I am surprised that anybody else would have paid attention to it.

      • Posted Feb 16, 2010 at 11:02 PM | Permalink

        the page for John Mashey is we Thanks for your help. I think I have enough ammunition to call the editors to task.

2 Trackbacks

  1. By Climategate, what is going on? - EcoWho on Feb 15, 2010 at 3:45 AM

    […] The Boulton Hockey Stick […]

  2. […] at a Royal Society of Edinburgh program, in a RSE Policy Advice paper on Copenhagen in which he presented a supposedly “independent” hockey stick and even on the issue of Himalayan […]

%d bloggers like this: