One of the ongoing Team mantras has been that the Mann hockey stick has been supported by a “dozen independent studies”. Obviously, I’ve disputed the claim that these studies are “independent” in any non-cargo cult use of the term “independent”. A new article by Jones and multiple coauthors (Holocene 2009) comments on this issue.
Let’s review the bidding a little on this topic. The viewpoint expressed here (and which is pretty obvious upon merely inspecting the proxies used in the reconstructions) is that the studies are not independent either in authorship or in proxy selection – with certain stereotyped proxies (Graybill bristlecones, Briffa Yamal) being used over and over again, thereby contaminating any supposed independence. This lack of independence is readily verified. And yet the opposite claim is repeatedly asserted.
The Wikipedia article on the HS controversy reports (citing Mann’s 2003 Senate testimony):
More than a dozen independent research groups have now reconstructed the average temperature of the northern hemisphere in past centuries.
In response to our original article, Wigley told USA Today in 2003, using language similar to Mann’s 2003 testimony, that a “dozen independent studies” had yielded similar results,
In addition to the Nature paper [MBH98], about a dozen independent studies suggest the 20th century was warmer than normal, Wigley points out.
Mann’s pre-emptive strike against our 2006 articles at realclimate again invoked “multiple independent studies”:
the basic conclusions of Mann et al (1998,1999) are affirmed in multiple independent studies.
Mann’s reply to questions from the House Energy and Commerce Committee in 2005 made the same claim about a “dozen independent studies”:
Recent work since the TAR has provided further support for this conclusion, which is now common to more than a dozen independent studies published in the peer-reviewed scientific literature.
The various supposedly “independent” reconstructions are not in fact independent either in authorship or proxy selection. There are important defects in each such study individually with proxy quality and robustness with respect to outlier results.
This prompted the following inline response by William Connolley:
[Response: At the moment, this looks like wild assertion / mud slinging. Given that the various reconstructions are the same on the important points, it seeems that the major conclusions are robust. Asserting that everyone else is wrong and only you are right is implausible – William]
Realclimate, showing a highly questionable commitment to its claim that “questions, clarifications and serious rebuttals and discussions are welcomed”, then censored my attempt to support my argument.
That the authors are not independent can be seen merely by inspecting the names of the coauthors of the Team studies in the usual spaghetti graph. Briffa et al  with coauthor Jones is obviously not “independent” in authorship from Jones et al  with coauthor Briffa. Jones and Mann  and Mann and Jones  are not independent of Briffa (Jones) et al 2001 or Jones (Briffa) et al (1998). MBH (Mann, Bradley and Hughes [1998, 1999]) is not independent of Bradley and Jones , which in turn is not independent of Hughes and Diaz  or Bradley, Hughes and Diaz , etc etc. To say that these supposedly “independent research groups” are not in fact “independent” in any sense familiar to non-climate scientists is hardly “wild assertion/mudslinging” as Connolley claimed.
Wegman touched on this in 2006, but his “social network” analysis didn’t really capture the essence of the issue. Unfortunately he focused on Mann’s coauthorships, rather than Jones and Bradley, who are much more central figures. He also failed to consider the canonical Stick studies as a group – which would have yielded a more germane result than the one actually produced.
The issue of proxy re-use is just as serious (or more serious). Wegman et al 2006 had an instructive illustration of this, showing proxy usage by study drawing on information that I provided in a spreadsheet. Wegman observed of this figure:
It is clear that many of the proxies are re-used in most of the papers. It is not surprising that the papers would obtain similar results and so cannot really claim to be independent verifications.
Without acknowledging a source, IPCC AR4 conceded that the reconstructions shown in its new spaghetti graph were “not entirely independent” [carefully not mentioning bristlecones or Yamal]:
As with the original TAR series, these new records are not entirely independent reconstructions inasmuch as there are some predictors (most often tree ring data and particularly in the early centuries) that are common between them, but in general, they represent some expansion in the length and geographical coverage of the previously available data (Figures 6.10 and 6.11)
Citing IPCC AR4, Jones et al 2009 now concedes that “several chronologies” have been utilized in “virtually all published studies”, again carefully not mentioning bristlecones or Yamal:
Several chronologies extending over a longer time span, with variability displaying a strong and direct association with changing local temperatures, have been utilized in virtually all published studies aimed at reconstructing Northern Hemisphere (NH) or global average surface temperature changes during the millennium leading up to the present (Jansen et al., 2007[IPCC AR4]).
Indeed, nowhere in their entire review dp Jones et al 2009 mention the word “bristlecone”.
Note that Jones et al 2009 substantially embellished the actual IPCC AR4 wording. IPCC AR4 did not state that these repetitively used chronologies had “a strong and direct association with changing local temperatures”. Indeed, Review Comments are against this claim. As an IPCC AR4 reviewer, I contested the use of bristlecone/foxtails in a spaghetti graph of proxies supposedly qualified against local temperature, observing (See SOD Review Comment 6-1144) in respect of a series from Osborn and Briffa 2006 (which claimed a correlation of 0.19 – itself hardly “strong and direct”:
I checked the correlation of this data to HadCRU2 gridcell temperature and only obtained an insignificant correlation of 0.04. The authors said that they had cited the temperature data incorrectly, that they had actually used CRUTEM2 yielding a correlation of 0.19 and that HadCRU2 data was spurious in its early portion (1870-1887) because there was no station data. However there is station data at GHCN going back to the data in HadCRU2. D’Arrigo et al 2006 considered using foxtails and rejected the use of this data because it did not meet standards of being correlated to gridcell temperature, expressed in very similar terms to Osborn and Briffa 2006. The contrasting views of D’Arrigo et al 2006 certainly establish that the relationship is “ambiguous” and that this proxy should not be used on multiple grounds.
Notwithstanding the fact that the actual correlation was 0.04 (as could be easily calculated), IPCC rejected my call that they not be shown in a spaghetti graph of temperature proxies as follows:
Some of what the reviewer says may be true, but is as yet unpublished and the current review is based on multiple strands of evidence, among which the results of Mann and colleagues remains relevant.
Whether or not I had published the incorrectness of the claimed Osborn and Briffa correlation is irrelevant. My calculation was correct and the authors either knew it or should have known it. (And, needless to say, IPCC did not require that some point be published if the opinion was adverse to us e.g. Ammann’s secret emails to Briffa, withheld from the IPCC Review Process.)
From this weak and contested gruel – an actual correlation of 0.04 to local temperature – Jones and the extended Team have ratcheted the argument in favor of the non-independent proxies to a claim of a “strong and direct association with changing local climates” for these non-independent proxies.
And they wonder why they have a PR Challenge.
In fact, I’m not sure that anyone presently claims that bristlecone/foxtail chronologies have a “strong and direct association with changing local climates”. When confronted with opposing evidence, Mann’s defence was “teleconnection” – that bristlecone chronologies teleconnected with world temperatures.