In my last post, I observed that Ellen Mosley-Thompson’s archiving record was even worse than that that of her husband, Lonnie Thompson, whose failure to adequately archive his ice core measurements has long been a subject of criticism at Climate Audit. In particular, I observed that Ellen had archived nothing from the 15 expeditions to Greenland and Antarctica that, according to her CV, she had led.
This post has occasioned fresh commentary on Lonnie Thompson’s archiving, which I will review in today’s post. In particular, I will assess Thompson’s statement in an email to a CA reader stating:
…our ice core data are archived at the World Data Center NOAA Paleoclimate data base in Boulder Colorado…
Despite Thompson’s claim, no data whatever is archived for many ice cores. For other cores, my issue is that the Thompson archive is completely inadequate, as I’ll discuss below. I remain mystified by Thompson’s intransigence in establishing a comprehensive and meticulous archive of his measurement data, as, in my opinion, he should regard the establishment of such an archive as an essential part of his scientific legacy and his #1 priority given his age and health.
Cores With No Archive Whatever
As noted in my recent post, according to her CV, Ellen has led “nine expeditions to Antarctica and six to Greenland to retrieve ice cores”. Antarctic sites include Plateau Remote, Dyer Plateau and Siple Station; Greenland sites include those in the PARCA program e.g. GITS, D2, D3, Raven, Tunu. Despite Lonnie Thompson’s claim that “our ice core data are archived at the World Data Center NOAA Paleoclimate”, no data from any of these Ellen-led expeditions has been archived at the NOAA Paleo website.
In my earlier post, I noted that one Greenland data set associated with Ellen had been archived, but pointed out that this came from a much earlier expedition that she had not led and that the data had been transcribed by third parties. For further clarification of this, in 1966, the first long ice core was drilled in Greenland at Camp Century (1387.4 m to bedrock). Lonnie Thompson studied dust concentrations in this core as part of his 1977 thesis entitled Microparticles, Ice Sheets and Climate (Ohio State University Institute of Polar Studies Report 64). These results were published in an academic journal in 1981 by Lonnie and his wife, Ellen Mosley-Thomspon as Temporal variability of microparticle properties in polar ice sheets (Journal of Volcanology and Geothermal Research 11). In 1990, this data was transcribed at WDC/NSIDC and eventually archived at NOAA (with the most recent deposit dated in 2007). This is the only archived Greenland/Antarctica associated with either Thompson and, as I stated in the earlier post, does not come from one of the 15 expeditions to Antarctica/Greenland led by Ellen Mosley-Thompson.
Subsequent to this early consideration of dust in the Camp Century, Greenland ice core, Lonnie’s career, as is well known, has been devoted to high-altitude tropical ice cores. In the late 1980s and 1990s, Thompson drilled a series of tropical ice cores: Quelccaya 1984, Dunde 1987, Guliya 1992, Huascaran 1993, Sajama 1997, Dasuopu 1997, Kilimanjaro 2000 and Puruogangri 2000. Each expedition was punctuated with a short article in Science, none of which, in my opinion, can really be considered as a comprehensive technical report on the ice cores.
Since 2000, Thompson has conducted expeditions to Bona-Churchill (two cores 2002 – NSF award here), Quelccaya and Coropuna (four cores 2003 – NSF award here), Puncak Jaya, New Guinea and Nevado Hualcán, Peru (four cores 2007 – see NSF award here and most recently Alto dell’Ortles, Italy (see NSF award here). See Thompson’s NSF award history
Nothing has been archived at NOAA from any of these expeditions despite Thompson’s assertion that “our ice core data are archived at the World Data Center NOAA”.
NSF funding for Puncak Jaya, Hualcan and Ortles has not expired (it continues in each case to March 2013), but NSF funding has expired for the other three programs (Bona-Churchill – May 2006; Quelccaya/Coropuna – June 2007; Naimona’nyi – March 2010). Nor has there been any publication of the Bona-Churchill ice core, a lacuna noted at CA for a number of years – see here.
Cores with Very Defective Archives
Let me now turn to the archival situation for the early cores, where Thompson has uniformaly failed to provide anything like a comprehensive or definitive archive of measurement data, but has (in some cases, grudgingly) provided digital versions of a single figure in an article, which Thompson has then attempted to pass off as an adequate archive.
Thompson’s NSF award for Quelccaya/Coropuna contains a relevant description of the major component of an adequate archive (though, needless to say, Thompson has not established such an archive for these cores):
This award will help obtain and analyze four ice cores to bedrock from the Quelccaya and Coropuna ice caps in the Peruvian Andes. These new cores, covering the last 10,000 years will provide a more robust and detailed history of Pacific sea surface temperature variability than has been possible with previous cores. Approximately 6,000 samples will be obtained from each core for stable isotopes of oxygen and deuterium, insoluble dust concentration and size distribution, and soluble aerosol chemistry analyses. These and other proxy measurements will be used to reconstruct a regional temperature and precipitation history and produce a high-resolution record of El Nino Southern Oscillation (ENSO) variations
This is precisely in accordance with the sort of definitive archive that I’ve advocated for a number of years. Each Thompson ice core has several thousand samples, on which various measurements are typically taken: O18, dust, chemistry, etc. If Thompson’s ice cores as important as represented, then they deserve nothing less than a definitive archive i.e. the measurements for all 6000 or so samples for each core. These are not “large” archives. I presume that much of the data is already in digital form somewhere at Ohio State.
Other information that is required for a definitive archive is detailed information on layer thickness (including photographs of ice core) used to date the core. The dating of rapidly thinning ice cores is subject to a variety of uncertainties. Subsequent investigators should be able to re-examine Thompson’s dating and, if they disagree, they should have the information that enables them to reassess the isotope history. (An example of such a re-consideration is Vinther’s recent reassessment of the 1970s and 1980s vintage Agassiz cores in light of information from more recent Greenland drilling.)
Instead of this sort of definitive archive, Thompson’s archive at NOAA for Dunde, Guliya and Dasuopu were, until a few months ago, nothing more than decadal O18 chronologies for the period 1000-1990 with no detailed sample information, no ancillary dust, chemistry, nothing on deeper portions of the core.
Aside from mere craftsmanship, there is an important additional reason why Thompson should be expected to archive all measurement data for these cores. In multiproxy studies, the raw measurement data (by sample) is combined with the dating assigned to the samples to produce (for example) isotope “chronologies” (to borrow the term used by dendros). As discussed at CA on numerous occasions – see, for example, here, Thompson has published or distributed inconsistent chronologies (see the figure below.) Some of the inconsistencies are hard to understand. The Dunde version of Thompson et al (PNAS 2006) is inconsistent with the version in Yao et al (Ann Glac 2006), of which Thompson was coauthor. The Dunde version of Thompson et al (PNAS 2006) is consistent with the version used in Mann et al (1998), but not with the intermediate Thompson et al (Clim Chg 2004).
Dunde Versions – see here.
Correspondence with Thompson and Journals
Until 2003, Thompson had archived nothing from any of his Himalaya cores (Dunde, Guliya, Dasuopu), all of which have been important in the paleoclimate discussion. In October 2003, shortly prior to publication of our 2003 article on MBH, I requested data on these cores from the Thompsons, but got nowhere.
In 2004, I was asked to review a response by Mann et al to our 2003 article. In that capacity, I asked to see supporting data for their submission (data that they had refused to supply to me late in 2003.) Stephen Schneider said that, in his 28 years as editor, no reviewer had ever requested supporting data. I replied that times change and that I wanted to see the supporting data. Schneider said that this would require approval by their editorial board. I asked that this be obtained. This incident is documented in 2004 Climategate emails. As a result, Climatic Change adopted a policy requiring authors to supply supporting data, but not code. I then insisted that Mann supply the requested supporting data; Mann appears to have withdrawn the submission rather than comply with the new data policy.
In March 2004, I requested supporting data for Thompson’s 2003 article under the new policy, a request that was supported by Climatic Change. However, instead of providing a comprehensive and definitive archive, Thompson simply archived decadal O18 chronologies for Dunde core 3, Guliya and Dasuopu for the top part of the cores from AD1000 to 1990.
Thompson has subsequently pointed to this absurdly inadequate archive as evidence that he has “archived his data”. However, for the various reasons set out above, archiving the published decadal O18 histories for these sites is only one small component of a definitive archive.
Over the years, I’ve attempted to persuade both Thompson himself and the various journals that a definitive archive is both required and long overdue. I’ve collated this correspondence here. In 2004 and 2005, I tried to persuade Climatic Change. In 2005 and 2006, I tried to persuade Science to require Thompson to provide a definitive archive of cores that Thompson had published in Science. This resulted in them supplying sample information for two Kilimanjaro cores that were interesting (see contemporary CA discussion here) but not relevant to my request for sample data from Dunde, Guliya and Dasuopu. I then had a lengthy correspondence with Science re-iterating my request for details on these three cores; the editors said that they corresponded with Thompson on the matter, but the requests were stymied and ultimately the Science editors stopped answering without ever answering or resolving the matter.
In 2007, I tried to obtain measurement data for the three cores once again with both the editor of PNAS and then Ralph Cicerone, the president of NAS. Once again, the correspondence went nowhere. I asked for an archive of sample data, carefully explaining the difference between sample data and a decadal chronology. PNAS responded that Thompson had told them that he had already provided what I asked for and that I should simply go to the NOAA site – a site that I was obviously well aware of and which I had distinguished in my initial response. And which obviously didn;t contain the sample data that I had asked for. Both Thompson’s unresponsive answer and PNAS’ seemingly wilful obtuseness to the problem (think Gavin Schmidt or William Connolley on upside-down Tiljander) are all too characteristic of the climate community and have contributed to the present adverse atmosphere.
Rather than enforce journal policies, Cicerone told me to write Thompson, who had refused for years to provide this data. So I wrote to Thompson one more time, this time copying Cicerone, Gerald North, Brooks Hanson of Science. Once again, Thompson ignored my email.
In comments on the prior thread, CA reader Roger has criticized me for being insufficiently diligent in trying to obtain measurement data from Lonnie Thompson. I believe that the attached correspondence completely refutes his criticism.
As I observed above, it remains a source of great puzzlement to me why Lonnie Thompson does not regard the establishment of definitive archives of his data as an integral part of his scientific legacy. At this point, it is evident that it will be difficult, if not impossible, for him to carry out the definitive statistical analyses of his data. Over the past 10 years, he has fallen further and further behind. The logical solution to his quandary is simply to provide a comprehensive and definitive archive and thereby let any interested scientist carry out the analyses that Thompson himself will either never carry out or not carry out for many years. Nor in my opinion should Thompson, at this age and stage of his career, spend his time haring off to yet another drill site. No vice president of exploration in a mineral exploration company would personally feel obliged to sit on a drill rig. Surely there are other people capable of supervising the next drill program. Thompson should let them do their job.