Climate Audit

Crowley’s comments to The Chronicle of Higher Education were as follows:

According to Mr. Crowley, the Duke professor, he received repeated e-mail messages from Mr. McIntyre demanding data and documentation, which grew increasingly threatening. "I’m usually happy to send people some stuff," said Mr. Crowley. However, he added, "McIntyre comes back time and again. He could take up a huge amount of time. It’s like you have nothing better to do in your life than answer questions from Steven McIntyre."

Obviously, Crowley’s comments have waived any expectation of privacy. So here’s the actual correspondence. It speaks for itself and cannot be construed as "increasingly threatening". If Crowley was happy to "send people stuff", then why did it take so long to obtain information? Crowley also requested data from me, which I sent promptly.

Dec. 12, 2003
Dear Dr. Crowley, I am interested in examining the actual proxy data used in Crowley-Lowery 2000, which was referenced by IPCC. I have been unable to locate the data, as used, at the World Data Center for Paleoclimatology. Can you direct me to an FTP location where you have archived this data or otherwise make the data available. Thank you for your attention. Yours truly, Stephen McIntyre

Mar. 29, 2004
Dear Dr. Crowley, I would appreciate it if you could respond to the request below [Dec. 12, 2003], which is consistent with most journal policies. Thanks for your consideration, Steve McIntyre

Apr. 15, 2004
I am looking for an FTP or email version of the proxy series that you used in Crowley-Lowery 2000. Could you send this to me. Thanks, Steve McIntyre

Apr. 15, 2004
Dear Dr. Crowley,Is there some problem with supplying this data? Regards, Steve McIntyre

April 15, 2004
Dear Ms. Kessler, [Editor, Ambio] I have received no reply from Thomas Crowley in respect to supplying the data requested below [Mar. 29, 2004 and predecessor emails] which is not publicly archived. Most journals require authors to publicly archive or supply data upon rquest and I am sure that Ambio is no different. I would appreciate it if you would take some steps to inspire Dr Crowley to supply the data requested. Thanks for your attention, Steve McIntyre

April 16, 2004
Dear Dr. Crowley, In this article, you state that "The new composite time series were converted to mean annual temperature in the following manner. The two composites were scaled to agree with the Jones et al. (31) instrumental record for the Northern Hemisphere over the intervals 1856-1880 and 1920-1965 (too few of the proxies record information after this date)."What exactly is this transformation? Did you make the mean of the CL index equal to the mean of CRU in both periods? Or is it some other transformation? Regards, Stephen McIntyre

June 29, 2004 [responding to December 12, 2003 email]
Dr. McIntyre: as I am soon to go on a long trip, the data will be placed in the WDC by the end of August. For your information, this project was not funded by any federal grant and I only used records already published, which could be obtained by anyone. In return I would like to have a copy of your own publications, plus any of the data you used to draw your conclusions. Thomas Crowley

July 6, 2004
Dear Dr. Crowley,I realize that the data which you used was published in paper form, but many of the series are not publicly archived and, in any event, the version for some series is sometimes ambiguous, so I do not agree that the data is presently “available to anyone”‘?. I have listed the series below, along with some specific problems I have encountered in trying to reconstruct your data base.

You mention that your Ambio 2000 article did not use federal funding, although you have received federal funding for other articles. Both Lonnie Thompson and Phil Jones – the stated sources for your Ambio 2000 data – have been supported by U.S. federal funding and obligations might well ensue from obtaining the data from these authors. In any event, there is an additional scientific obligation to disclose your data, which would not be avoided by a possible legalistic technicality of merely not receiving federal funding for this article.

This matter would appear to be moot, since, after 6 months of not acknowledging any communications from me on this topic, you plan to finally archive the data. When you do so, I suggest that you consider AGU policies for data citation and, for the several series which are already in the public domain, you might include cross-references using AGU citation policies.Regards, Steve McIntyre

White Mt, Sierra Nevada tree ring: Did you use Sheep Mt or Campito Mt? Which version? The WDCP version of Sheep Mt (ca533) appears to incorporate later data than the referenced article.

Central Colorado tree ring: There are 3 series in the underlying article: Almagre, Mt Evans and Hermit Lake. Which did you use or is it a combination? Two series co071-Almagre and co072 “€œ Mt Evans archived by Lamarche.

Jasper, Alberta tree ring: There are no contributions by Luckman at WDCP. There is a contribution by Schweingruber labelled Athabaska “€œ historisch”‘? at cana170w_crns.crn : is the series that you used?

Marion Lake, Michigan pollen: The WDCP series ftp://ftp.ngdc.noaa.gov/paleo/pollen/asciifiles/fossil/ascfiles/gpd/lake27.txt is the only related data set, but does not seem to have the data.

Sargasso Sea O18: This appears to be: GISP2 O18: This appears to be data at: Iceland documentary: I can’t locate digital version of this.

C England temperature (Lamb): The data at Hadley Center goes back to 1659. I haven’t been able to locate digital version of this series back to 1000.

N Sweden tree ring: I’ve located digital version of series used in Briffa et al. (1992), but this refers to older temperature reconstruction used in Briffa et al (1990), which I’ve been unable to locate.

French Alps tree ring: I presume that this is the series fran010.crn at WDCP Black Forest, Germany tree ring. I’ve been unable to locate digital version of this.

N Urals tree ring: This temperature reconstruction is not archived at WDCP, but I presume that this appears to be the version archived at Briffa’s website.

Qilian Shan Mts tree ring: I’ve been unable to locate digital version of this. Dunde Ice cap O18: A few weeks ago, responding to my inquiries to Climatic Change, Thompson posted up 10-year averages of this data. Did you use the data in this form?

Chinese documentary (Chu): I’ve been unable to locate a digital version of this.

July 6, 2004
Dr McIntyre,while quoting chapter and verse on my obligations as a U.S. citizen to a non-national, you did not mention anything about sending me copies of your own papers and analyses, which I requested, plus the data you used in your analyses. please send all that material on to me. tom crowley

July 6, 2004 McIntyre to Crowley
I’m happy to do send this to you, but request that, pending the archiving of Crowley-Lowery data with WDCP, you first send me the data that I requested 6 months ago.

July 7, 2004 Crowley to McIntyre
sorry, if you want the data you should be first willing to share your own, tc

July 12, 2004
In fact, the data and methods for our article on Mann et al. were posted up last year at http://www.climate2003.com see http://www.climate2003.com/SI.MM03.htm . I have attempted to make this transparent and if you have any inquiries, I’d be happy to answer them. Back to the matter at hand – could you now please fulfil your end of the bargain by forwarding the CL2000 data. Thanks, Steve McIntyre

July 20, 2004
Once again, I would appreciate a copy of the data which you used in CL00. I really don’t understand your apparent reluctance to provide this information. This has been going on for 6 months. Yours truly, Stephen McIntyre

July 24, 2004 Crowley to McIntyre
hello, I am in europe and in very infrequent email contact. I cannot post the ambio data until I get back. please remind me august 12, tc
July 24, 2004
Will do. Thanks. Steve Mc.

Sep. 15, 2004
Dear Dr. Crowley, I have been unable to locate this information at WDCP. Did you archive it yet as you promised? Regards, Steve McIntyre

sep. 15, 2004
Steve, sorry I did not, I can send you a file that has the information. I am away from the office today (sick) and will take care of it by week’s end tom

Sep. 15, 2004 McIntyre to Crowley
Thanks. Would appreciate it.

Sep. 21, 2004 McIntyre to Crowley
Any progress on this?

Sep. 21, 2004 Crowley to McIntyre
Steve, believe it or not I am still sick, honest. very slow recovery but will attend to it as soon as I am back, tom

Sep. 21, 2004 McIntyre to Crowley
It’s no fun getting older, when you have to recover from illness.

Oct. 10, 2004
McIntyre to Crowley Any progress on this?

Oct. 15, 2004
Dear Dr Crowley, you have promised on several occasions to provide this data, but it is still not available. I am really quite puzzled. Regards, Stephen McIntyre

Oct. 15, 2004 [with Excel spreadsheet attached]
Stephen, My sincere apologies for not getting this to you sooner. Attached is an excel file of the scaled data we used in the Ambio composite. Note that there were two composites we illustrate – one with low resolution records in it, the other without. The better estimate is the one without the low resolution data because the latter simply required too much interpolation to make estimates reliable. I also include the full composite with the low resolution data in it – you can derive the subset composite by just excluding the low resolution data. the numbers in the composite seem represent the arithmetic mean of the data included in the composite (not all sites went throughout the record). the composite is therefore not a temperature composite but just the mean value of the individual scaled time series. Separately I scaled that to NH temperature for the purpose of comparing with the Mann et al. figure. I hope this is sufficient and again I apologize for being so slow about getting this to you. Tom Crowley note also that the purpose of the ambio

Oct. 15, 2004 McIntyre to Crowley
Thanks for this. While you have the file open, do you have the source versions that you used before your scaling?

Oct. 15, 2004
No, those where done sometime before — we got these from various published sources when I was at texas a&m. with moving and everything I am not sure where they are, or if I even have them. tom

Oct. 18, 2004
Dear Dr Crowley, Do you recall where you got the digital versions of the proxy series in the first place? Where I have been able to locate other versions of the series, I’m having trouble reconciling the your scaled versions to these other versions (after applying smoothing and scaling). I notice a reference to Phil Jones and Lonnie Thompson in acknowledgements in connection with data: did they send you the data? Did you scan any of it?

I’m curious about the reasoning in connection with a few series. The Zhu (Chu) 1973 phenological series was criticized by Zhang (Climatic Change 1994) as having incorrect dates from lunar calendar conversions in the 13th century, which then affected the interpretation of the phenology. Why did you use the Zhu version rather than the Zhang version?

Also for Tornetrask, you used the Briffa (1990) version rather than the Briffa et al (CLim Dyn 1992) version, which is the version used in other multiproxy studies. Why was that?

The Lipp German C13 data looks like the 100-year smoothed version rather than the annual data – was this scanned? Some of the series seem to be connected to precipitation rather than temperature (e.g. the French Alps tree ring which has very low correlation to other series). How did you decide to use the French tree ring data? Regards, Steve McIntyre

Oct. 18, 2004
off the top of my head I can answer only some of this; 1) some were acquired from others, some were digitzed (eg, C13 record) – I presume Phil and Lonnie gave me some of the data if I acknowledged them – cannot recall specifically what Phil sent. 2) note that there are differences sometimes between raw data and processed data because one has to adjust raw data for growth effects – some of the differences you see may be that 3) with respect to Tornetrask quite simply I cannot recall why – I have subsequently compared some of the Fennoscandian data with Esper’s work – there are differences but I seem to recall the general trends are pretty similar. note that the purpose of our study was not to produce the best estimate of past temperature change, but to assess the stability of a conclusion about the Medieval Warm Period by including information that discussed this period — even if it was not the best information. this was a sensitivity study that focused on three things — what happens when you just have a near-time invariant data set? what do alll the data look like in the time domain (Fig 1 — by comparison Seong and Baliunas do not show ANY data), and what does the record look like if you just brutally average all the data without any weighting for temperature etc. the remarkable conclusion — which I did not expect — was that even with the above limitations the results looked a like Jones and Mann. sure there were differences, but that is to be expected due to need for screening of questionable or low resolution records, or weighting the record.) but the main point is that the general conclusion about the relative magnitude of Medieval warmth appeared to be robust. Please note that I cannot answer all the specific questions you address — I have many many obligations and quite simply cannot go back and dig up some material that was analyzed five years ago, at another institution. tom crowley

Jan. 12, 2005
Dear Dr Crowley, I realize that your records and documentation in this matter are inadequate and that you may not be able to answer this question, but perhaps you can: when I compared the graphic for Lamarche [1974] to the digital version of the White Mtns series that you provided me, I was unable to reconcile the two: the peak in Lamarche [1974] was in the late 12th century, while the peak in your version was in the 20th century. Do you recall where exactly you obtained the digital data used for this series? Regards, Steve McIntyre [no response]

Mar. 19, 2005
Hello, may I get a copy of your time series showing the prominent warming in 1400s? tom crowley

Mar. 19, 2005 [with attachments]
Enclosed are our recent articles. If you read the articles themselves, including the E&E article, you will see that our position is that the MBH98 reconstruction lacks statistical significance – we estimate the R2 in the 15th century step in their version to be 0.0 and we show that their RE statistic is spurious. Mann et al. have refused to disclose their R2 statistic or the digital version of their 15th century step and have only replied with a critique of the R2 statistic, so it’s pretty clear that our estimation of their R2 statistic is pretty close. We also show that their PC method mines for hockey stick shaped series, putting the controversial bristlecone pine series in dominaiton of their PC1. We show that, contrary to their claims of robustness, their results are not robust to the presence/absence of bristlecone pines (showing the effect of the bristlecones in the PC4 if the PC results are calculated correctly.) We show that the high 15th century values result from MBH98-type calculations using the archived version of the Gasp” series (rather than the editing in MBH98), correct principal components calculations and the roster of retained PCs used in MBH98. We do not ourselves argue that 15th century temperatures were high – only that Mann et al. are not entitled to claim under their methods that 20th century values are robustly higher than 15th century temperatures. Mann et al have argued that they can retain 5 PCs in their calculation under a variation of Preisendorfer’s Rule N applied to tree ring networks, which was not mentioned in MBH98 and which cannot be reconciled with retained PC rosters in other networks. Be that as it may, their own current position shows the lack of robustness which we point out. I hope that this clarifies your question. Regards, Steve McIntyre

Mar. 21, 2005
Steve, thanks for the clarification. would it still be possible to obtain the temperature reconstruction showing the large warming – it is my understanding that some people are misinterpreting that series and i would like to have a copy for my records, tom ps it was not clear to me from your description whether the R2 statistic using your approach is different than 0.0 for the 15th c. – can you clarify? tom crowley

Apr. 6, 2005 [with attachment of data]
Dear Tom, Here is a text file for the Figure in our Energy and Environment article showing the alternative reconstruction using centered PC calculations, the PC retention roster of MBH98 and the archived version of the Gasp” series. We get similar results under other variations e.g. without the bristlecone pines (which arguably are contaminated by a fertilization or precipitation effect) or without the PC4 representing the bristlecone pines. Our point on statistical significance was that our calculations showed that the MBH98 reconstruction had an R2 of 0.0 and that the RE statistic was inaccurately benchmarked and the apparent "significance" claimed by Mann et al. was spurious. If an actual climate "signal" is being recovered, then the reconstruction should obviously have significance in both RE and R2 statistics. Obviously Mann et al. withheld the adverse R2 statistic (and continue to refuse to produce it). We are not putting forward the attached calculations as "our approach" as we do not endorse MBH98 methods. The R2 statistic was insignificant in both versions. Let me know if you have any further questions. Regards, Steve McIntyre

Apr. 8, 2005
Dear Dr Crowley, I’ve been informed that you recently made a presentation in Washington referring to the evidence of raw proxy data. In October, you told me that you had "misplaced" your raw proxy data and were then able to only provide a smoothed and transformed version. I take it that you have subsequently located the "raw data" in question and, if so, I would appreciate a copy of the data. Thanks, Steve McIntyre

Apr. 8, 2005
Steve, I think I said something to the effect that is the closest I can get to the raw data – showing the standard deviations of the individual time series — with respect to tree rings, these data have been conservatively processed for growth rate effects during the early years of growth — some of the data I pulled off from the web or from authors that you know for the growth processed data, some I did myself from published data available on web sites — that is why I said this was the closest I can get — due to lack of time (~15 minute presentation) one cannot go into every detail about that preprocessng of growth rate – what I sent you before for the ambio paper for tree rings was the post-processed growth rate data — my conservative processing just cuts off the first part of the tree ring record that shows anomalous growth – one can determine that from looking at longer sites from teh same area that show no "bulge" of growth over the same interval as a longer tree. authors process them differently, but I don’t necessarily agree with those processing methods. that is the long and the short of it. I still would like to have a time series of your computation of the 1400’s warming, for whether you say it or not, others do cite the record as an indication of the warming – I will however make it cleaer in any presentation that this is what has happened with respect to your calculation….. tom

Apr. 8, 2005
Tom, I did send you the file you requested. Could you check your inbox again. If you can’t find it, I’ll re-transmit. I don’t understand the first paragraph below [re Apr 8, 2005 from Crowley] – is the same data as C&L [2000] or different datasets that you were discussing? Thanks, Steve McIntyre

Apr. 8, 2005
Steve, it has been very hectic this week with visitors and university obligations, maybe we can talk sometime midweek, ok? tom

UPDATE Oct 2009:
Shortly after this post was published, Crowley published an editorial in EOS, a widely distributed journal, repeating and amplifying the untrue allegations rebutted above and adding further untrue allegations about previous correspondence with Mann. I’ve placed Crowley’s EOS editorial online here. Despite our explicit statements to Crowley that we were showing the effect of different alternatives (e.g. not using the disputed bristlecones) and were not ourselves offering an alternative reconstruction, Crowley nonetheless criticized our demonstration of the instability of the Mann reconstruction as being an alternative reconstruction of the type that he is now challenging us to make. Crowley:

The McIntyre and McKitrick paper (referred to here as MM) showed an unexpectedly large warming in the 1400s, apparently calling into question Mann’s analysis and the uniqueness of the late twentieth century warming. However, the consensus among climate scientists most familiar with the data is that the MM warming in the 1400s is due to an error in the MM analysis method; it can also not be supported by an examination of the data.

There was no “error” in our analysis method. Our analyses were examined in 2006 by both the Wegman Panel and North Panel and neither panel identified any error in our analyses. The issue lay with the non-robustness of the MBH reconstruction to the bristlecones, as we pointed out.

Crowley’s untrue statements about my conduct contributed to third party scientists holding an adverse impression of how I handled myself. For example, Marcel Crok, a Dutch reporter, gave the following account of an interview with Nanne Weber, a Dutch climate scientist:

Here she started to talk about the tone of the debate or the scientific soap. According to her the tone in the messages of Mcintyre is not so friendly. She gave me a copy of the EOS paper of July 12 2005 in which Crowley gives his opinion about this. I told her I have read a lot of the correspondence between McIntyre and different parties involved and that according to me he is nowhere not friendly.

I submitted a Comment to EOS responding to Crowley’s untrue statements, which I have placed online here

I don’t know whether Crowley’s editorial was “peer reviewed”. EOS took over 8 months to review my Reply. I didn’t hear back from them until June 2006. The reviewer recommended an “Encourage resubmission”, noting that I had raised “valid concerns”, but suggested that the material be merged into one comment, which he felt would not be too hard to do.

Reviewer #1 Evaluations:
Recommendation: Reject and encourage resubmission
Improve Grammar: No
Annotated Manuscript: No
Anonymous: You may reveal my name to the Author(s)
Referrals: No

Manuscripts #1272 and #1273 need to be considered together. They are the author’s comments concerning a recent forum by Crowley.

The author is raising what appear to be some valid concerns. However, the author also requests that Eos publish both of these manuscripts, which together total 1370 words in length. Because the author is responding to one forum, I do not believe it is appropriate for there to be two separate comments about that forum from the author.

The author instead should be encouraged to resubmit one comment (which would take the place of manuscripts 1272 and 1273), with a maximum of 750 words in length. In reviewing these two manuscripts, it appears that it would not be too difficult for the author to boil down these two comments into one comment that is 750 words or less.

In the last paragraph of manuscript #1272, the author notes, “Both correspondences are posted online at http://www.climate.org/correspondence.htm” I was not able to locate these correspondences at this web site, and this particular link did not appear to be working. However, if these correspondences indeed have been posted at this web site, this may bring up a question concerning potential dual publication. It would be appreciated if the author could clarify as to whether, where, and when there may have been any other publication of these comments.

Please note that for #1273, I have simply recommend that the manuscript be rejected. The comment for #1273 should be read along with this comment for #1272.

Notwithstanding the reviewer’s “Encourage resubmit”, the EOS editor rejected re-submission on the basis that the matter was no longer “timely” – despite the fact that EOS itself had delayed a review for 8 months.

Thank you for submitting your articles “Comment on Crowley” and “Comment on Crowley 2”. I offer my sincere apologies for the delay in delivering this decision. Unfortunately, these no longer appear to be timely. Therefore I am not able to accept them for publication at this time.

Thank you for the opportunity to examine this work.

Sincerely,
Fred Spilhaus
Editor, Eos

I followed up with EOS, observing that they were still responsible for publishing untrue comments and asking them to correct the record. They said that they would “look into” it. I never heard from them again.

In July 2006, Crowley and I met at the House Energy and Commerce Committee meetings. Following the long day of hearings, I asked him and others to go for a beer. We chatted about basketball among other things – he was at Duke at the time and I follow the Raptors (Vince Carter had come from Duke rival, North Carolina.)