A Belated SI for D’Arrigo et al 2006

The other day, I noticed that the long dormant WDCP supplementary information (and here) for D’Arrigo et al 2006, of which Rob Wilson is a coauthor, had been updated on April 30, 2012. In 2005, D’Arrigo et al (then under review at JGR) had been cited by IPCC AR4. At the time, as an IPCC reviewer, I attempted to obtain both very rudimentary information about the sites used and unarchived measurement data from the authors, from the IPCC and from the journal (JGR, which was theoretically subject to AGU policies requiring data archiving.) My efforts were totally rebuffed. I was even threatened with expulsion as an IPCC reviewer for asking for data. I tried again in October 2009 and was once again rebuffed. So what accounted for this belated update nearly seven years later? The backstory proved interesting. The new SI is an improvement but still unsatisfactory and, unfortunately, contained errors on the long contentious Polar Urals data set.

On September 20, 2005, after accessing then unpublished D’Arrigo et al from the IPCC WG1 website for unpublished and cited articles, I had sent the following request to Rob Wilson as a D’Arrigo coauthor (Wilson, despite savage criticism from the Team, has tried to be cordial with me):

Dear Rob, I was asked to be an IPCC 4AR reviewer and accepted. I’ve looked at your interesting D’Arrigo et al. 2005, submitted. I noticed that you used a Polar Urals data set of 155 cores. The SChweingruber larch data set russ021 archived at WDCP has 93 cores. What accounts for the difference? Who did you get the data from? Can you send me a copy of the measurement dataset that you used? Thanks, Steve

Wilson’s more senior coauthors refused to allow me to have the measurement data. I therefore asked IPCC and the journal to require them to supply me with measurement data. My follow up requests were forwarded to Phil Jones and Tin Osborn at CRU (though he had no connection to the data) and therefore crop up inn Climategate emails as my request was quickly forwarded to CRU for their support. The incident is referred to 2005 CG2 emails e.g. CG2- 2590, where D’Arrigo wrote Osborn and Wilson urging that I be fired as an IPCC reviewer, while also expressing the need for caution in their emails as “Lord V will stop at nothing”:

they should fire him [McIntyre] as a reviewer of IPCC – i cant believe they included him in the first place! So, please email him back and tell him that he should as he says take it up with the ipcc authors and see whether it is still appropriate to include him as a reviewer.

we should however be very cautious about our emails, lord v will stop at nothing (this is sort of fun in a harry potter way)…

D’Arrigo had earlier removed a reference to McIntyre and McKitrick in a earlier version of their article stating “i dont want to validate macintyre and mcitrick by referencing their work”. Jones contacted IPCC Chair Susan Solomon, who then threatened to expel me as an IPCC reviewer if I persisted in efforts to obtain supporting data for unpublished articles cited by IPCC. See earlier chronology here.

Although Wilson was prevented from supplying me with measurement data, he confirmed that other nearby sites had been used in their Polar Urals chronology (mentioning russ022w and russ001.) I couldn’t reconcile the core counts to the reported 157 and therefore looked for other datasets, quickly finding the Schweingruber POLURULA measurement data (corresponding to the russ176w chronology) and observed that core counts from this data set (rather than russ001), in combination with the other two data sets, exactly matched. (This surmise has proven correct as the core IDs in the actual D’Arrigo et al data set include these core IDs and not russ001.)

The POLURULA (russ176w) data set is what I’ve called the “Polar Urals Update”. On Sep 26, 2005 at CA here, I did a standard RCS calculation on the combined larch data from russ021 and russ176, observing with considerable consternation that the additional data resulted in a dramatic change: instead of Polar Urals standing as evidence of a cold 11th century, as Briffa had reported, it now showed warmth (consistent with the Shiyatov treeline data discussed at CA.)


Figure 1. September 26, 2005 Polar Urals re-calculations.

The effect of the additional POLURULA measurement data on the Polar Urals chronology (and CRU’s apparent disinterest in this additional data) then became a long standing issue at Climate Audit and was raised directly in my submissions to the SciTech Committee and Muir Russell (but ignored or fobbed off.)

I tried again in October 2009 for the measurement data, a request that once again was passed to Phil Jones, who expressed confidence that AGU would not require D’Arrigo and Wilson to provide data as follows:

I doubt JGR/AGU will follow this up.

It was therefore rather a surprise to see that the authors of D’Arrigo et al had belatedly begun to provide Supplementary Information on April 30, 2012 – seven years later. While the SI remains unsatisfactory and is even incorrect on some points, it provides long overdue clarification on others.

So why now?

The answer lies in the seemingly unconnected publication by Mann et al of an article in Nature Geoscience entitled “Underestimation of volcanic cooling in tree-ring-based reconstructions of hemispheric temperatures – pdf here in February 2012. Mann observed that the D’Arrigo et al 2006 reconstruction did not show the projected impact of major past volcanic events:

Both models were driven with estimated forcings for the past millennium, during which several volcanic events dwarf any modern eruptions. The main features of the two simulations compare favourably (Fig. 1b). We then compared the simulations to a state-of-the-art tree-ring reconstruction of Northern Hemisphere temperatures based on a network of annual treering thickness records from over 60 high-elevation and boreal treeline sites (consisting of 20 regional series) across North American and Eurasia [7 - D'Arrigo et al 2006].

Given the uncertainties in both the radiative-forcing estimates used to drive the climate models and the temperature reconstruction itself, the overall level of agreement is striking (Fig. 1b). Both simulations capture much of the low-frequency variation in the reconstruction and the residual variability (Supplementary Information) is within the estimated range of internally generated climate noise13. Yet there is one glaring inconsistency: the response to the three largest tropical eruptionsad 1258/1259, ad 1452/1453, and the 1809+1815 double pulse of eruptions is
sharply reduced in the reconstruction. Both models predict a drop of 2 C following the 1258/1259 eruption, whereas the reconstruction shows a decrease of only 0:6 C. A similar pattern holds for the two other largest eruptions. Regardless of the predicted cooling, in no case does the reconstruction show more than ~1-2 C cooling relative to the modern base period; that is, there seems to be a 1 C floor on the cooling recorded by the reconstruction.

Although trees at treeline have long been held by dendros to be sensitive to temperature, Mann hypothesized that trees at treeline had “reduced sensitivity to cooling”:

the discrepancy between expected and reconstructed temperatures is probably an artefact caused by a reduced sensitivity to cooling in trees that grow near the treeline. This effect is compounded by the secondary effects of chronological errors due to missing growth rings and volcanically induced alterations of diffuse light….

Trees growing near the latitudinal or elevational treeline are typically selected for use in reconstructing past temperature changes7,8 because their growth is primarily limited by temperature. One unintended consequence is that these same trees may not document the full extent of past cooling events, and in particular the response to the immense AD1258/1259 eruption

In March, the D’Arrigo authors submitted a Comment on Mann’s article. Apparently Mann then told Nature Geoscience that he could not respond to the D’Arrigo et al comment unless he had full access to all the data used in D’Arrigo et al 2006. My correspondent observed somewhat laconically:

I am sure you will appreciate the irony of this.

Indeed.

Thus, in order to respond to Mann’s article, they had no alternative other than to finally provide supporting data for D’Arrigo et al 2012. Unfortunately, while the SI is immeasurably improved, some data sets continue to remain unarchived and some errors persist in the SI.

Regional Chronologies

A major improvement is that the new SI finally provides the 19 regional chronologies (both STD and RCS) as calculated in D’Arrigo et al. Previously only one regional chronology had been archived (the Gulf of Alaska chronology used in Kaufman et al 2009 and which was forced out in connection with the publication of Kaufman et al 2009 – see CA at the time.)

Measurement Data Still Missing

The new SI reports that the following measurement data used by D’Arrigo et al 2006 remains unarchived:
– Graumlich (1997) and Giddings (1948) data used in the Seward chronology. (contact Wilson)
– Icefields (Jasper), Alberta data from Luckman and Wilson 2005. The SI states:

The data used by Luckman and Wilson will soon be archived in the ITRDB. For now, a data request can be sent to Brian Luckman (lxxx@uwo.ca)

I’ve requested this data from Luckman a number of years ago and was blown off.
– Coastal Alaska data about which they say:

Sub-fossil material: Data not archived and continually being updated.
Relevant contact is Greg Wiles (gxxx@wooster.edu) – primary generator
of the data – and Rob Wilson (rxxx@st-andrews.ac.uk) who has original
2006 version used.

- data used in Alps/Tirol STD from Wilson, R.J.S. and Topham, J. 2004. Violins and Climate. Theoretical and Applied Climatology, 77: 9-24. Wilson now states:

The Topham data are not archived. For these data please contact Rob Wilson (rxxx @st-andrews.ac.uk).

- data from Nicolussi and Schliessling 2001 used in their Alps RCS reconstruction. No contact information was given for this.
– various sites collated by Esper (Jaemtland, Taymir, Alps/Tirol STD) about which the authors say (with apparent frustration):

Although all the Esper et al. (2002) data are archived in several locations (e.g. ITDRB, WSL, Science), it may not be clear exactly which data-sets were used. Please contact Jan Esper (Jxxr@geo.uni-mainz.de) for the specific data-sets used in this study.

CA readers are of course aware that Esper is a determined non-archiver and has never answered a single email from me.
-
Chronologies with Site Lists

A number of measurement data sets have been archived since D’Arrigo et al 2006. For the following chronologies, there is now a list of component sites with archived measurement data: Seward (mostly), NWNA Alaska, Yukon (2 – though TTHH is obsolete at ITRDB and may not correspond), Central Alaska (2), Wrangells(10), Coastal Alaska (10), Central NWT (4), Manitoba (1), Labrador (2 of 5 archived), Gaspe (long missing Gaspe update archived earlier this year), Yakutia (8), Mongolia (1).

Sites remaining incomplete are: Seward, Labrador, Alps/Tirol (STD), Alps RCS.

Sites remaining under Esper confusion are the classic sites of Tornetrask, Taymir and Polar Urals.

Polar Urals
The fresh SI for Polar Urals is a total confusion. The D’Arrigo authors stated:

STD version
Historic (ITRDB RUSS021) and Living (ITRDB RUSS021) NB. Historic data are LASI while living data are PCOB – for this reason RCS detrending was not possible as the variance changes between the two species.

RCS version. Keith Briffa’s Yamal chronology was used. Briffa, K. R. 2000. Annual climate variability in the Holocene: interpreting the message of ancient trees. Quaternary Science Reviews 19:87-105.

Raw data and more explanation can be found at these links:

http://www.cru.uea.ac.uk/cru/people/briffa/yamal2000/

http://www.cru.uea.ac.uk/cru/people/briffa/yamal2009/

NB also, that the ITRDB has much more data available now [my bold] for this region:
ITRDB RUSS220
ITRDB RUSS176
ITRDB RUSS001
ITRDB RUSS002
ITRDB RUSS006

This commentary is inaccurate on a number of points – some surprisingly.

1) The STD reconstruction was not derived only from RUSS021 (which only had 91 cores). The STD dataset used in D’Arrigo et al 2006 (which I obtained today) was the same as that used in Esper et al 2002 (which I obtained six years ago. In addition to russ021w (LASI), both used the russ176w-POLURULA (LASI) and russ022w (PCOB) datasets. This can be proven by matching ID numbers and cores.

2) Contra the new SI, RUSS021w did not use both LASI and PCOB. The living cores in RUSS021 were LASI; the Polar Urals PCOB cores were in RUSS022. Schweingruber included both LASI and PCOB cores in the same identification sequence 86201, 86202, 86203,… but there are no duplications between russ021w and russ022w. The D’Arrigo SI is simply wrong on this point.

3) The assertion (citing russ001, russ002, russ006, russ176 and russ220) that “the ITRDB has much more data available now” is flatly untrue. None of the data in any of these five data sets was collected subsequent to 1990 – indeed russ001, russ002 and russ006 date back much earlier, even to the 1960s. russ176 was not only available at the time of D’Arrigo et al 2006: it was used in it.

4) russ220 is a bizarre and stupid dataset that should be removed from WDCP/ITRDB. I’ve heard through the grapevine that NSF has required Jacoby and D’Arrigo to tidy up their unarchived data, accounting for much recent long overdue archiving. (Needless to say, they got more money to do what should have been done long ago.) None of the data in russ220 appears to have been collected by D’Arrigo and Jacoby. It includes the old Schweingruber Polar Urals ring width data discussed above (russ021w, russ022w and russ176w). It also comingles density data e.g. old Schweingruber MXD data from russ021x. In any event, there is nothing new in russ220.

The present SI contains a sly reference to a longstanding error in D’Arrigo et al 2006, which its authors have been aware of, but refused to correct. D’Arrigo et al 2006 showed a RCS reconstruction labelled “Polar Urals”, but it looked nothing like my calculation: it was very Hockey Stick shaped. It turned out that their figure was not from Polar Urals after all – they had used the Yamal chronology instead, mistakenly labeling it as “Polar Urals”. I pointed this out to them and suggested that the authors issue a corrigendum on this point but they refused. Now, in Mannian style, they have issued the following very sly reference to the error in the seven-years-late SI, giving them some ongoing cover, I guess (but notably not conceding that the legend in the article itself was wrong):

RCS version. Keith Briffa’s Yamal chronology was used.

The SI also contains a very coy reference to a longstanding skirmish between Rob Wilson and me. The Yamal chronology (with YAD061, the most influential tree in the world), has a huge Stick and is very popular, while the Polar Urals “updated” chronology has an elevated MWP and, other than its one use in Esper et al 2002, has been ignored. Wilson has argued that there was a “good” reason for preferring Yamal arising from variance changes; I’ve challenged his explanation at CA. (No concern about “variance changes” is reported in the article itself. Further, a variance change criterion, if applicable, should be applied consistently; my surmise is that any criterion that rejected RCS for Polar Urals would also reject other chronologies. D’Arrigo et al have almost total algorithmic obscurity on this point.)
The D’Arrigo SI appears to allude to this longstanding dispute as follows:

Historic data are LASI while living data are PCOB – for this reason RCS detrending was not possible as the variance changes between the two species.

As noted above, this “explanation” is wrong in its premise that the living data in russ021w was PCOB. The authors have no evidence that Schweingruber, an eminent and experienced dendro, had incorrectly allocated PCOB data to his LASI russ021w dataset (rather than grouping it with other PCOB measurement data in russ022w.) The error on the part of the D’Arrigo authors appears to have arisen many years ago when they incorrectly added a prefix P (for PCOB) to some russ021w cores as well as to russ022w cores. I noticed this today within minutes of finally getting the measurement data as used in D’Arrigo et al 2006.

32 Comments

  1. Posted Oct 20, 2012 at 4:09 PM | Permalink | Reply

    Obfuscation, as shown here, is sad
    Seems they must hide the science that’s bad
    How much better to fix
    (And avoid years of tricks)
    This is science, not some private fad

    ===|==============/ Keith DeHavelle

  2. Jeff Alberts
    Posted Oct 20, 2012 at 4:13 PM | Permalink | Reply

    I remain gobsmacked.

  3. Posted Oct 20, 2012 at 4:14 PM | Permalink | Reply

    Well, better (very) late than never, I suppose.

    An interesting – but no doubt coincidental – aside (from my perspective anyway). You had noted and quoted:

    The new SI reports that the following measurement data used by D’Arrigo et al 2006 remains unarchived:

    Icefields (Jasper), Alberta data from Luckman and Wilson 2005. The SI states:

    The data used by Luckman and Wilson will soon be archived in the ITRDB. For now, a data request can be sent to Brian Luckman (lxxx@uwo.ca)

    I’ve requested this data from Luckman a number of years ago and was blown off.

    Hadn’t realized previously that Luckman was at my alma mater (UWO). But, more importantly, the last time I saw “Luckman” was in the (now famous Nov. 2009) “Miracles and Stip Bark Standardization” post:

    http://climateaudit.org/2009/11/16/luckman-at-the-canadian-society-for-petroleum-geologists/

    P.S. Noticed a few typos in the headpost:

    “My follow up requests were forwarded to Phil Jones and Tin Osborn at CRU [...] crop up inn Climategate emails [...] The incident is referred to 2005 CG2 emails …”

  4. Craig Loehle
    Posted Oct 20, 2012 at 4:27 PM | Permalink | Reply

    In science the details matter. If you don’t want to obsess over details, there are lots of other professions that are appropriate. The idea that it “doesn’t matter” when data are labeled as the wrong species, or the wrong site, or key data left out because you don’t like them, is horrifying. And stonewalling is not a “clever way of solving a problem” but it sure is a trick.

    • Craig Loehle
      Posted Oct 20, 2012 at 5:47 PM | Permalink | Reply

      Actually, I was wrong. There are NOT many professions where the details don’t matter. The painters who did my house did not get a speck on the carpets and I did not have to call them back for a do-over. Construction workers can make a mess while working but when done it better look nice. The dentist better drill the right tooth. In many professions big mistakes get you fired. Except climate science (TM).

      • Duster
        Posted Oct 22, 2012 at 3:04 AM | Permalink | Reply

        Craig, in post-normal science details don’t really matter because you can make them up. You’ll note that a recent study found that 2/3s of withdrawn articles were due to outright falsified data. And that was I believe mostly in biology and medicine. That latter is chilling.

  5. Skiphil
    Posted Oct 20, 2012 at 6:18 PM | Permalink | Reply

    Free the data, free the data! Every time I read one of these accounts I am sickened by the careless, neglectful attitudes and behaviors of too many climate scientists. In the 21st century there is no reasonable excuse for not promptly archiving such data.

    If scientists want to keep data private indefinitely they should (1) not accept public funding, (2) not submit for publication articles relying upon private data, and (3) not allow the IPCC or other public policy reports to be based upon such private data. Once they choose to operate in the public domain with OUR money (of whichever country’s taxpayers), the rules are (should be) different than what they do in their private personal time or for one lab or university only.

    Of course, countless truly private activities exist, just don’t ask the public to pay for them or accept them as the basis for policy. Publicly funded and/or publicly promoted data used to further PUBLIC policy recommendations cannot be allowed to remain private. Publicly accessible reliable archiving should be required in all fields which receive public funds, and most definitely in areas of climate science used to promote major political, economic, and legal policies.

    • Geoff Sherrington
      Posted Oct 21, 2012 at 4:22 AM | Permalink | Reply

      Skiphil, You are so correct that public data (especially major data that influences puublic policy) should be completely open. There are no hafl-way points. The selective release of information, or the demanding of a large fee for public information, merely sets a stage for corruption.
      As Craig notes before you, in many employments a bad mistake can get one fired. In the atypical climate scange arena, mistakes are howled down or hushed up and indeed sometimes followed by a scheme to award an honour or a medal or more funding.
      The key word is accountability. Poor scientists can play fast and loose with wrong information if they know that they do not have a punishment awaiting them. In this contect, readers might like to visit a long post that I wrote at http://joannenova.com.au/2012/10/gergis-hockey-stick-withdrawn-this-is-what-95-certainty-looks-like-in-climate-science/ It is dated October 19, 2012 at 7:25 pm and is long because it includes several emails.
      In essence, the particular government funding agency declines to answer if rebates should be made by authors if or when papers are withdrawn by the authors.

  6. Posted Oct 20, 2012 at 7:09 PM | Permalink | Reply

    Utterly awesome Steve.

    The error on the part of the D’Arrigo authors appears to have arisen many years ago when they incorrectly added a prefix P (for PCOB) to some russ021w cores as well as to russ022w cores. I noticed this today within minutes of finally getting the measurement data as used in D’Arrigo et al 2006.

    They call him a troublemaker – but it’s they and they alone that cause the trouble to themselves and everyone else, the world’s policy makers included, by not doing what any scientist should have done in the first place.

  7. theduke
    Posted Oct 20, 2012 at 10:02 PM | Permalink | Reply

    Steve writes: “Apparently Mann then told Nature Geoscience that he could not respond to the D’Arrigo et al comment unless he had full access to all the data used in D’Arrigo et al 2006. My correspondent observed somewhat laconically:”

    I am sure you will appreciate the irony of this.

    “Indeed.”

    That is priceless. Seems Mann is doing some ClimateAuditing himself now. Maybe he should apply for a job at climateaudit.org as an associate.

    ..

    • Skiphil
      Posted Oct 20, 2012 at 10:22 PM | Permalink | Reply

      Re: theduke (Oct 20 22:02),

      Yes, but considering how little Mann seems to have learned over the years he will not even qualify as an Intern.

    • Not Sure
      Posted Oct 21, 2012 at 11:57 AM | Permalink | Reply

      Who is Steve’s correspondent? I re-read the post a couple of times, but I couldn’t figure it out.

      • Posted Oct 21, 2012 at 12:10 PM | Permalink | Reply

        The correspondent emerged without introduction, then withdrew to the shadows. Given the context I’m not expecting Steve to announce the name.

  8. Rob Wilson
    Posted Oct 21, 2012 at 2:23 AM | Permalink | Reply

    Just to defend Jan Esper a little, much, if not all the data of the data generated by the WSL over the last decade is archived here:

    http://www.wsl.ch/dendro/dendrodb.html

    Rob

    Steve: To my knowledge, this entire data set is old Schweingruber data. Can you point to a single data set at this site that provides fresh Esper data. Nothing from Pakistan; nothing from Morocco. The site codes match Schweingruber site codes. Rob, it’s one thing to defend Esper but factual defences work better. I note for the record that Esper has recently archived Pakistan data at WDCP. HOwever, he has failed to archive his new Russian data – data that bears on an important controversy. He has refused to provide it on request as well. And refuses to acknowledge emails.

    BTW I haven’t commented on the recent Esper et al 2012 paper except for passing comments observing that I found it very interesting. In particular, it made a couple of points that very much interested and surprised me. I regard it in some ways as a post-Hockey Stick paper as it looks at the data as it arrives, rather than framing it as a HS confirmation.

    • Posted Oct 21, 2012 at 11:41 AM | Permalink | Reply

      Sounds like there could yet be hope for Esper.

    • Frank
      Posted Oct 21, 2012 at 3:35 PM | Permalink | Reply

      Rob: Thanks for taking the time to communicate with the skeptic community. I hope that Steve will a correction for “non-archiver” above, that accurately reflects for both your comment and his reply. However, the standards for disclosure of data should be the same no matter what Esper’s work concludes about the MWP. Even small efforts like yours provide some hope that the politicization of climate science won’t tarnish the integrity of science indefinitely.

      May I suggest that you look to the rules that have been adopted for archiving the results from clinical trials in response to the suppression of data showing that drugs don’t work?

      http://grants.nih.gov/ClinicalTrials_fdaaa/at-a-glance.htm

      Steve: Rob has not given any reason to change my statement. Rob arm-waved to a site at which Esper has, to my knowledge, not archived a single data set.

      • Tom Gray
        Posted Oct 22, 2012 at 8:45 AM | Permalink | Reply

        From the NIH site, the applicable compliance requirement:

        t this time, FDAAA requires the responsible party to submit summary results information (including adverse events) no later than 1 year after the primary completion date (see Definition) for registered applicable clinical trials involving drugs that are approved under section 505 of the Food, Drug and Cosmetic Act (FDCA) or licensed under section 351 of the PHS Act, biologics, or of devices that are cleared under section 510k of FDCA. However, NIH encourages results reporting for all NIH supported clinical trials registered in ClinicalTrials.gov, regardless of whether or not they are required to do so under FDAAA

        This would seem applicable with modifications for papers that are used in major reports such as the tPCC ARs. If authors wish their paper to be considered for assessment within an AR, this would seem to be a desirable prerequisite.

        • Tom Gray
          Posted Oct 22, 2012 at 8:50 AM | Permalink

          Perhaps, it is also pertinent to note that for patents, an inventor must submit all prior art to which he/she is aware. This includes prior art that comes to their attention after the patent has been submitted. Failure to do this is very serious. It is deemed “inequitable conduct” and will result in the patent being unenforceable or, to all extents and purposes, cancelled.

  9. Posted Oct 21, 2012 at 3:52 AM | Permalink | Reply

    Rob Wilson, since no one’s responded
    May I offer “Welcome to CA!”
    And thanks! While some data’s absconded
    This will add one more source to the fray

    You may get some grief from other quarters
    For temerity, showing up here
    But from this group, these gentle sporters
    Simply seek to make this science clear

    ===|==============/ Keith DeHavelle

  10. Don Keiller
    Posted Oct 21, 2012 at 6:48 AM | Permalink | Reply

    Obfuscation, cherry-picking and plausible deniability.
    Key “Team” training parameters.

  11. Solomon Green
    Posted Oct 21, 2012 at 7:55 AM | Permalink | Reply

    Being a relatively newcomer to your site I was unaware of your correspondence with Dr. Solomon. In particular to your very reasonable explanation for your request, to which you received no reply.

    “In my opinion, examining the underlying data is an important part of reviewing materials. This opinion is obviously widely shared as the provision of supplementary information is standard for most paleoclimate articles (though not as widely shared as I would like) and many journals have policies requiring the archiving of data used in articles (although the policy is not always upheld.) Unavailability of underlying data would be a point that I would raise in my planned review.”

    The fact that seven years after that correspondence some of the data which you requested has come to light provides no excuse for Dr. Solomon’s appalling manners and lack of integrity. No one should agree to chair a review knowing that the documents necessary for that review would be withheld from the reviewers. But worse, when informed that the data were not being made available, to abet in their concealement rather than to order their release showa that, however competent an scientist, she was not a fit and proper person to chair the panel.

  12. doug
    Posted Oct 21, 2012 at 10:23 AM | Permalink | Reply

    “The error on the part of the D’Arrigo authors appears to have arisen many years ago when they incorrectly added a prefix P (for PCOB) to some russ021w cores as well as to russ022w cores. I noticed this today within minutes of finally getting the measurement data as used in D’Arrigo et al 2006.”

    Steve, you are one heck of an auditor! Please steer clear of any job offers from the tax people!

  13. KnR
    Posted Oct 21, 2012 at 11:13 AM | Permalink | Reply

    Information is power , therefore control of information is always desired by those that seek power . When you add in laziness and a ‘ivory tower ‘complex that regards the world outside of their own acedmic realms to be filled by fools , you can see how things get like this .

  14. Betapug
    Posted Oct 21, 2012 at 12:06 PM | Permalink | Reply

    “….the discrepancy between expected and reconstructed temperatures is probably an artefact. This effect is compounded by the secondary effects of chronological errors…”

    From speculation to certainty in adjacent sentences!

    One small step for a Mann, one giant leap for Mannkind??

  15. Robin Edwards
    Posted Oct 21, 2012 at 2:00 PM | Permalink | Reply

    I have noted with interest the comments in this thread regarding Jan Esper.

    My very recent experience (last week) when I emailed him to ask tentatively for a lead to the data he used in his 2012 paper “Variability and extremes of northern Scandinavia temperatures…” was an instant reply with attached files of data, description and an Excel file (that I have not yet been able to read due to the age of my computer!). The main data file is about 640K bytes, and on initial perusal consists of numerous individual sets of observations with, as expected, a one year resolution. I wrote immediately to thank him. In due course I hope to be able to send him my comments on any information I can deduce from these data.

    I also think, Steve that his own conclusions are very striking to say the least.

    This seems to be an example of a very courteous and friendly attitude, which might serve as an excellent example to other scientists.

    Robin

    • theduke
      Posted Oct 21, 2012 at 2:48 PM | Permalink | Reply

      Perhaps he supplied the data because he sensed “your aim is to [not] try and find something wrong with it?”

      When you say he’s “an excellent example to other scientists,” are you implying that other scientists should rudely ignore Steve’s requests for data?

    • Manfred
      Posted Oct 21, 2012 at 4:10 PM | Permalink | Reply

      Providing information only to those who are supposed to be in one’s “camp” or not known to be critical is, in my view, more disturbing than not giving information to anyone.

      What would such behaviour imply, when it comes to reviewing another person’s work, or deciding about employment etc. ?

  16. Robin Edwards
    Posted Oct 21, 2012 at 5:33 PM | Permalink | Reply

    What I wrote was “This seems to be an example of a very courteous and friendly attitude…” refering to the specific response that I received from JE. Can’t see anything wrong with it so far.

    Actually, the data contain some oddities in their formating, with what I take to be some missing separators and the occasional use of “labels” that seem to change case for no apparent reason. Not a big deal, of course. What is more trouble for me is the names of the data groups, which are numbers with an alpha suffix – usually a or b, but which is in some case missing altogether. Some of the last 50 or so lines (out of around 8500) also have c or d as their suffix. What this means I have not been able to work out yet. Perhaps it will all become clear when I can read the Excel file.

  17. tty
    Posted Oct 22, 2012 at 7:44 AM | Permalink | Reply

    Re Mann’s views on ”underestimation of volcanic cooling” there is an interesting paper in the latest issue of Climates of the Past with the somewhat cryptic title “Constraining the temperature history of the past millennium using early instrumental observations” (http://www.clim-past.net/8/1551/2012/cp-8-1551-2012.pdf). It is actually about the digitalization of weather data from 900 English East India Company ships logs from 1790 through 1830 which give rather good coverage of the tropical and temperate parts of the Atlantic and Indian Oceans. The most interesting result is probably that the extremely large eruptions of 1809 (unknown) and 1815 (Tambora) apparently had a quite subdued effect over the oceans (on the order of 0.5 degrees).
    Otherwise oceanic temperatures in 1790-1830 seem to have been very slightly warmer than in 1961-1990 and barometric pressure significantly and mysteriously lower.

  18. Skiphil
    Posted Oct 22, 2012 at 12:26 PM | Permalink | Reply

    Not specific to D’Arrigo but I recommend to all this prior CA thread on ‘flaccid’ journal reviewing and policies related to SIs:

    http://climateaudit.org/2008/04/11/supplementary-information-and-flaccid-peer-reviewing/

  19. Robin Edwards
    Posted Oct 23, 2012 at 7:42 AM | Permalink | Reply

    I hope Manfred does not think that I am “not critical” or “are supposed to be in one’s “camp”. By this no doubt he means believing in the tenets of ardent AGW believers. I am most certainly not. I am very strongly critical of the pseudo-arguments put forward by such deluded people. In general I am sceptical of climate analyses until I have made my own analyses of any data that my software can handle, which I have done since 1992 on a very regular basis. I have examined literally more than a thousand time series of many different types, and have no doubt at all that “the consensus” has been sadly but effectively conned by self-seeking climatologists.

    Robin

One Trackback

  1. [...] mention, though, the fierce hatred of anyone else attempting to do that double-checking. Here’s an example, of both the corruption of certain scientists and of serious problems with the data — some of [...]

Post a Comment

Required fields are marked *

*
*

Follow

Get every new post delivered to your Inbox.

Join 3,112 other followers

%d bloggers like this: