Phil Jones’ first instinct on learning about Climategate was that it was linked to the Yamal controversy that was in the air in the weeks leading up to Climategate. I had speculated that CRU must have done calculations for Yamal along the lines of the regional chronology for Taimyr published in Briffa et al 2008. CRU was offended and issued sweeping denials, but my surmise was confirmed by an email in the Climategate dossier. Unfortunately neither Muir Russell nor Oxburgh investigated the circumstances of the withheld regional chronology, despite my submission drawing attention to this battleground issue.
I subsequently submitted an FOI request for the Yamal-Urals regional chronology and a simple list of sites used in the regional chronology. Both requests were refused by the University of East Anglia. I appealed to the Information Commissioner (ICO).
A week ago, the Information Commissioner notified the University of East Anglia that he would be ruling against them on my longstanding FOI request for the list of sites used in the Yamal-Urals regional chronology referred to in a 2006 Climategate email. East Anglia accordingly sent me a list of the 17 sites used in the Yamal-Urals regional chronology (see here). A decision on the chronology itself is pending. In the absence of the chronology itself, I’ve done an RCS calculation, the results of which do not yield a Hockey Stick.
In today’s post, I’ll also show that important past statements and evidence to Muir Russell by CRU on the topic have been either untruthful or deceptive.
The Relevance of Yamal
The Yamal chronology is relevant both because, since its introduction in 2000, it has been used in virtually all of the supposedly “independent” IPCC multiproxy studies (see an October 2009 discussion here) and because it is particularly influential in contributing an HS-shape to the studies that do not use bristlecones.
IPCC AR4 Box 6.4 showed the eight proxies which have been used the most repetitively (this wasn’t its intent.) Of these eight proxies, Briffa’s Yamal (labelled “NW Russia”) is shown with the biggest HS blade, larger even than Mann’s PC1 (labelled here as “W USA”). See here) and tag yamal.
Figure 1. Yamal Chronology in IPCC AR4 Box 6.4. Labelled as “NW Russia”
In previous posts, I’ve satirized the “addiction” of paleoclimatologists to bristlecones and Yamal as, respectively, heroin and cocaine for climatologists. (In pharmacological terms, upside-down Tiljander would be, I guess, LSD, as the psychedelic Mann et al 2008 is indifferent as to whether proxies are used upside-down or not (cue Jefferson Airplane‘s insightful critique of Mannian statistics.)
Although Yamal and Polar Urals had been long-standing topics at Climate Audit, they first attracted wide attention in late September 2009, when measurement data became available for the three “regional chronologies” of Briffa et al 2008 (Taimyr-Avam, Tornetrask-Finland and Yamal).
The 2008 Taimyr-Avam and Tornetrask-Finland networks were dramatic expansions of the corresponding networks of Briffa (2000), but the Yamal network, which was already much smaller than the other two networks, remained unchanged. Analysis of the previously unavailable Taimyr data showed that Briffa had added measurement data from several Schweingruber sites into the Taimyr-Avam regional chronology (a point not mentioned in the article itself.) Since there were a number of Schweingruber sites (including Polar Urals) in a similarly sized region around Yamal, it seemed almost certain that CRU would have done a corresponding regional chronology calculation at Yamal.
This raised the obvious question of why Briffa et al 2008 had not similarly included Schweingruber data in a Yamal “regional” chronology. Ross posed the question in a contemporary op ed as follows:
Combining data from different samples would not have been an unusual step. Briffa added data from another Schweingruber site to a different composite, from the Taimyr Peninsula. The additional data were gathered more than 400 km away from the primary site. And in that case the primary site had three or four times as many cores to begin with as the Yamal site. Why did he not fill out the Yamal data with the readily-available data from his own coauthor? Why did Briffa seek out additional data for the already well-represented Taimyr site and not for the inadequate Yamal site?
The question applied not just to the Khadyta River site in the original CA post, but to Polar Urals and other nearby sites. These questions resulted in considerable controversy at the time. CRU protested their innocence and posted a lengthy response on October 29, 2009, denying that they had ever even “considered” use of the Schweingruber Khadyta River site, discussed in contemporary Climate Audit posts. In a submission to Muir Russell, they later denied ever re-appraising their Polar Urals chronology.
The Climategate dossier was released in November 2009, a few weeks after the Yamal controversy. As Fred Pearce observed in The Climate Files, the Climategate dossier begins with Yamal and ends with Yamal. Pearce also observed that the word “Yamal” occurs more often than any other “totem” of the disputes, even more than “hockey stick”. Nearly all Climategate documents with unbleached dates were copied after my Yamal posts and Yamal measurement data dominated the earliest documents.
The Climategate dossier revealed that CRU had, after all, calculated a Yamal-Urals regional chronology as early as April 2006. (CG1 – 684. 1146252894.txt). The present FOI request referred to this email.
The Yamal-Urals Regional Chronology
The FOI list consists of 17 larch datasets in western Siberia between 50 and 75E, all but one north of 64N. It includes: (1) Yamal of Briffa (2000); (2) Polar Urals used in Briffa (1995); (3) additional Polar Urals data from 1999 (the “update”); (4) the Schweingruber Khadyta River site; (5) 7 other Schweingruber larch data sets; (6) two Shiyatov data sets collected in the 1960s; (7) four Vaganov data sets collected in 1992. All have been publicly archived except the Vaganov data, which, although used as early as MBH98, was first disclosed in the Climategate documents. (CRU placed a Vaganov subset online at their website in April 2012 as part of their FOI response.)
The Schweingruber Khadya River site in the FOI list (russ035w) is exactly the same site that I had discussed at Climate Audit in September 2009. Site russ176w in the FOI list is the Polar Urals “update” discussed in many Climate Audit posts.
The figure below compares core counts for the withheld regional chronology (salmon) to core counts for Yamal (cyan). Modern core counts for the regional chronology are about 20 times higher than core counts in the reported Yamal chronology, reaching nearly 400 cores in the 1960s. In the 1980s, core counts are still around 300, as compared to 12 in the Yamal chronology.
CA readers will doubtless recall CRU’s statements that the “best” indication of regional ring widths requires use of “all the data”).
So what is the “best” indication of relative ring-width changes in this Yamal region? One approach is to judge this by making use of all the data to hand.
The next figure shows a Yamals-Urals regional chronology (script shown in first comment). In this figure, I’ve calculated the chronology after allowing for inter-site differences. If inter-site differences are ignored, a “method” potentially used in other Briffa et al 2008 regions, 20th century levels are a little higher but not exceptional. (I’ll post up an attachment showing the differences.)
Figure 3. Yamal-Urals regional chronology. In this regional calculation, inter-site differences in mean ring width are allowed for by scaling post-1500 mean ring width at each site to equal the overall post-1500 mean ring width.
Although this chronology seems to look a lot different than the canonical Yamal hockey stick, prior to the 20th century, they are extremely similar (800-1900 correlation of 0.91.) Briffa’s Yamal chronology diverges high in the 20th century and especially the late 20th century, as shown in both figures below. In the left figure, both series are scaled on 800-1900 to emphasize the 20th century high divergence of the Briffa Yamal chronology; on the right is a scatter plot of Briffa’s Yamal (800-1996) against the regional chronology, with values after 1850 colored, also showing the high divergence.
Figure 4. Left – Yamal and regional chronology (scaled on 800-1900). right – scatter plot. Click to enlarge.
CRU Statements and Evidence
The question for CRU defenders is to justify their preference of such a small core count, when they had already calculated a regional chronology with an order of magnitude more cores. Since the original criticism in September 2009, CRU has given a variety of different responses, but none, in my opinion, answer the question. Indeed, none of their responses to date have even admitted or disclosed their prior calculation of a regional chronology, let alone explain why they didn’t report it, preferring instead to attack their critics.
CRU’s first reaction to Climate Audit questioning of their inconsistent handling of the Taimyr and Yamal regional chronologies was ridicule – a reaction quickly endorsed by realclimate and other advocacy blogs.
In The Climate Files, Fred Pearce described this response as follows:
For Briffa, as a tree ring technician, what McIntyre was doing was farce. McIntyre simply didn’t know what he was dealing with, he argued. The new data, Briffa said with evident disdain for the ignoramus on his trail, was estimating temperature from “maximum late wood density” measurements. These were quite different from the tree ring width measurements used in the Yamal series. Scientifically they were not comparable. And the data McIntyre had chosen only went back 100 years, which was useless for the longer time series Briffa was interested in.
Realclimate, known from Climategate emails to have been coordinating with CRU on this matter, immediately accused me of “randomly” adding in data found on the internet:
McIntyre has based his ‘critique’ on a test conducted by randomly adding in one set of data from another location in Yamal that he found on the internet. People have written theses about how to construct tree ring chronologies in order to avoid end-member effects and preserve as much of the climate signal as possible. Curiously no-one has ever suggested simply grabbing one set of data, deleting the trees you have a political objection to and replacing them with another set that you found lying around on the web.
These criticisms were widely accepted in the advocacy “community”. However, the incorrectness of the derision is eloquently rebutted by the inclusion of the Khadyta River dataset in the FOI list. Not that I expect any of them to withdraw these criticisms.
Khadyta “Not Considered”
A few weeks later (October 29, 2009), CRU took a different and inconsistent line in an article on the CRU website. This time, instead of ridiculing the inclusion of Khadyta River, they conceded that the site met their criteria, but claimed that they had “simply not considered” it at the time:
Our current practice when selecting data to incorporate in a regional chronology, is to include data exhibiting high levels of common high-frequency variability (i.e. on the basis of high inter-site correlations, where these are calculated using high-pass filtered data). Judged according to this criterion it is entirely appropriate to include the data from the KHAD site (used in McIntyre’s sensitivity test) when constructing a regional chronology for the area. However, we simply did not consider these data at the time, focussing only on the data used in the companion study by Hantemirov and Shiyatov and supplied to us by them.
They re-iterated this claim in both their March 1 and June 17 submissions to Muir Russell, stating, for example, in the latter (p. 8):
“McKitrick is implying that we considered and deliberately excluded data from our Yamal chronology. The data that he is referring to were never considered at the time because the purpose of the work reported in Briffa (2000) and Briffa et al. (2008) was to reprocess the existing dataset of Hantemirov and Shiyatov (2002).
Again, CRU’s claims not to have “considered” inclusion of Khadyta River (and other similar Schweingruber sites) is refuted by the FOI list. The untruthfulness of this evidence was not commented on by Muir Russell.
The “Purpose” of Briffa et al 2008
According to CRU’s website statement and submissions to Muir Russell, the reason why they had failed to “consider” the Khadyta River data was that the “purpose” of Briffa 2000 and Briffa et al 2008 was simply to “reprocess” Hantemirov and Shiyatov 2002. (See quotations above.)
As always with CRU, one has to watch the pea.
One of the purposes of Briffa (2000) was clearly to demonstrate the effect of RCS methodology on the Hantemirov and Shiyatov 2002 dataset. I have no objection to CRU claiming this “purpose” for Briffa (2000).
But, by 2008, this was no longer their “purpose”. Indeed, one doubts whether the editors of Phil Trans B would have accepted a 2008 paper with such a mundane purpose. The actual “purpose” of Briffa et al 2008 is stated quite clearly and was entirely different: it introduced and discussed “regional” chronologies. In accordance with that purpose, the datasets for the Taimyr and Tornetrask regions were dramatically expanded from 2000.
CRU’s evidence to Muir Russell on the “purpose” of Briffa et al (2008) was deceptive and/or untruthful. [Update May 11 - Gavin Schmidt agreed that CRU's evidence here is a "mis-statement" but characterizes the "mis-statement" as only "slight" and rejects the idea that CRU would intentionally mislead Muir Russell. Be that as it may, the evidence as it stands was untrue on this point and has not been corrected.]
Nor was CRU consistent with their excuses, even within individual submissions to Muir Russell.
In their June 17 submission to Muir Russell, after previously making the implausible and untrue claim about reprocessing, CRU admitted that the “purpose” of Briffa et al 2008 was to present regional chronologies and that they had, indeed, planned to do one for the Yamal-Urals region. However, they claimed that they didn’t have enough time to complete it and therefore reverted to the small Yamal dataset:
When we later received a request to submit a paper to a planned themed issue of the Philosophical Transactions of the Royal Society about ‘The boreal forest and global change’, Briffa and colleagues decided to use some of the material to hand in preparing a draft. It was intended that this should describe 3 continuous 2000-year ring-width series, each originally planned to represent the integration of a large-regional data set of subfossil and living tree data. The focus was to be on representing large-regional growth signals and initial comparisons with equivalent regional temperature data. The western, ‘Fennoscandia’, series would incorporate near tree-line pine data from northern Sweden and Finland; the Avam-Taimyr series would integrate larch data from near the Taimyr peninsula tree-line region. Between these we had intended to explore an integrated Polar Urals/Yamal larch series but it was felt that this work could not be completed in time and Briffa made the decision to reprocess the Yamal ring-width data to hand, using improved standardization techniques, and include this series in the submitted paper.
The “explanation” of insufficient time is completely implausible. They had already calculated a Yamal-Urals regional chronology: they could have used the one at hand? And why did they tell Muir Russell that couldn’t “complete” the chronology in time, when they had already done the calculation. If they felt that there were technical issues that disqualified the regional chronology calculated in 2006, why didn’t they report these problems (along with the regional chronology itself) in the article itself and/or to Muir Russell.
Ironically, in East Anglia’s FOI refusal, they argued that release and use of “incomplete” data could result in “incorrect or misapplied conclusions” (apparently disregarding the implications for the admittedly incomplete Yamal chronology applied in Briffa et al 2008):
There is little public interest in the release of unfinished or incomplete data – i.e. does not contain a description of how it was created or why the “selected” methods were chosen – and so does not reflect the full breadth of academic rigour and thought applied to it. The information may well be incorrect, untested, unreviewed and may not accurately reflect the proper outcome of the research. Incorrect or misapplied conclusions could be drawn from the publication of unfinished data and any assessment of the merit of the work should be based upon a final, approved version of the data.
In any event CRU’s evidence that they couldn’t complete the regional chronology in time is not credible on its face. The question of the regional chronology was raised in my submission and Muir Russell’s negligent failure to examine this topic means that questions that should have resolved continue to linger.
Instead of disclosing to Muir Russell that they had already calculated a Yamal-Polar Urals regional chronology (and explaining their reasons for withholding it), CRU spent much of their energy in their submissions attacking my supposed endorsement of an update to the Polar Urals chronology.
The issue in older Climate Audit posts was the failure to report the updated Polar Urals chronology, given both the lack of 1000-year chronologies and the discrepancy between the two chronologies in the modern period.
However, after the Climategate dossier proved the existence of a Yamal-Urals regional chronology, the principal issue was the failure to use the Yamal-Urals regional chronology in Briffa et al 2008 or to otherwise report it (not a beauty contest between Yamal and Polar Urals). As noted above, CRU had noted their use of “common high-frequency variability” as their test whether to include data in a regional chronology. Polar Urals and Yamal have strong common high-frequency variability, expressed by Hantemirov and Shiyatov (2002) as follows:
Initially, a 1250-year chronologyfrom the Polar Ural mountains (Shiyatov, 1995) was used as a ‘master’ dating series as well. There is a high degree of similarity in the annual variability of radial tree growth between the Yamal and Polar Urals areas because of their proximity (about 200 km).
CRU was well aware of the common high-frequency variability between the two sites: this was presupposed in their development and analysis of the “Yamal-Urals” regional chronology. Astonishingly, instead of reporting on these results in their October 2009 website article or in their submissions to Muir Russell, CRU claimed that they had “never undertaken any reanalysis” of Polar Urals: (June 17, p.7):
We had never undertaken any reanalysis of the Polar Urals temperature reconstruction subsequent to its publication in 1995.”
Obviously this assertion is refuted by the presence of both the original Polar Urals data and the “update” in the FOI list.
“All the Data”
As shown above, there are obvious and visible difference between the regional and Yamal chronologies. However, in their website response to Climate Audit in October 2009 and to their submissions to Muir Russell, CRU claimed that there was little difference between their Yamal chronology and a chronology using “all the data”. How can these contradictory claims be reconciled?
The graphic below shows compares core counts of CRU’s “all the data” (violet) to the core counts of the Yamal-Urals regional chronology that CRU had calculated in 2006. Evidently, contrary to their representations, CRU did not use “all the data” after all.
In their original statement on Yamal, CRU stated:
We would never select or manipulate data in order to arrive at some preconceived or regionally unrepresentative result.
Undoubtedly this is how they think of themselves. But their history shows that they have had a strong sense of what their results “should” look like and have, on other occasions, selected and manipulated data so that their results accord with “preconceived” results.
The “Briffa bodge” was a completely arbitrary “adjustment” of Tornetrask MXD data so that the answer made ‘sense’. The “Briffa bodge” was the predecessor to the “very artificial adjustments” described in Climategate source code documents.
Likewise, CRU’s decision to “hide the decline” by deleting MXD data after 1960 was evidently done so that the MXD temperature reconstruction accorded with preconceived ideas. This was done by CRU themselves and was a different manipulation of data than “Mike’s Nature trick (as described in more length in previous CA posts.)
Had their Yamal-Urals regional chronology had been in accordance with their previous results, I am completely convinced that they would have used it in Briffa et al 2008 and/or their October 2009 online article without a second thought. My surmise is that the apparent failure of the (still withheld) Yamal-Urals regional chronology to accord with their expectations caused CRU not to use it in Briffa et al 2008. I realize that this is a harsh statement, but it’s what I think. Muir Russell had an opportunity and obligation to investigate, but unfortunately failed to do so.
The response of IPCC defenders is always that none of this “matters”, that they can “get” a Stick using other data and that CRU’s withholding of the Yamal-Urals chronology is no more than punctuation errors.
Climategate 2 emails contain an interesting vignette between Mann and Andy Revkin on this point. Revkin relied on the “without tree rings” reconstruction of Mann et al 2008 to re-assure himself that the picture was “solid”
needless to say, seems the 2008 pnas paper showing that without tree rings still solid picture of unusual recent warmth, but McIntyre is getting wide play for his statements about Yamal data-set selectivity.
As CA readers know, Mann’s “solid” reconstruction without tree rings was a mirage: it depended on Mann’s use of contaminated Tiljander data, used upside down. In the SI to Mann et al 2009, Mann conceded that his no-dendro reconstructions did not validate without contaminated data (but, unfortunately, did not notify PNAS or retract the earlier paper). Nor did Mann notify Revkin of the then pending admissions, instead allowing Revkin to continue to believe that the Mann et al 2008 no-dendro reconstruction was “solid”. In his recent book, Mann made no reference to the apparent concessions on the invalidity of Mann et al no-dendro reconstructions, instead claiming that Mann et al 2008 had used “objective” methods to validate the contaminated data.