The recent success in getting at least some data from Phil Jones – which he had obstructed since my original request in 2003 – has caused me to refresh my attempts to get Lonnie Thompson to archive his data so that the scandalous inconsistencies between different versions can finally be appraised. Last year, he published an article drawing on seven tropical ice cores in PNAS, which has a data policy that provides inter alia:
Unique Materials: Authors must make Unique Materials (e.g., cloned DNAs; antibodies; bacterial, animal, or plant cells; viruses; and computer programs) promptly available on request by qualified researchers for their own use.
and:
Databases: Before publication, authors must deposit large data sets (including microarray data, protein or nucleic acid sequences, and atomic coordinates for macromolecular structures) in an approved database and provide an accession number for inclusion in the published paper. When no public repository exists, authors must provide the data as Supporting Information online or, in special circumstances when this is not possible, on the author’s institutional web site, provided that a copy of the data is provided to PNAS.
These policies seem a little better on paper than some other journal policies, In addition, their webpage invites people experiencing problems to write to them. So I did so. Here’s my letter:
Dear Sirs,
Last year, I was invited to make a presentation to the National Academy of Sciences Panel on Surface Temperature Reconstructions on millennial temperature reconstructions and have published several peer-reviewed articles in the field, which were cited by the above panel in their report last year.
I am writing in connection pursuant to your policies for availability of unique materials and databases http://www.pnas.org/misc/iforc.shtml#submission in connection with Thompson et al 2006, Abrupt tropical climate change: Past and present, PNAS 103, 10536-10543
Your policy statement says that:
Unique Materials: Authors must make Unique Materials (e.g., cloned DNAs; antibodies; bacterial, animal, or plant cells; viruses; and computer programs) promptly available on request by qualified researchers for their own use.
and:
Databases: Before publication, authors must deposit large data sets (including microarray data, protein or nucleic acid sequences, and atomic coordinates for macromolecular structures) in an approved database and provide an accession number for inclusion in the published paper. When no public repository exists, authors must provide the data as Supporting Information online or, in special circumstances when this is not possible, on the author’s institutional web site, provided that a copy of the data is provided to PNAS.
Thompson et al 2006 describe results from ice cores drilled at Dunde, Guliya, Dasuopu, Puruogangri, Quelccaya, Huascaran and Sajama. For each core, several thousand samples were taken and analyses on a sample-by-sample basis made for isotopes, chemistry and other indicators. The information for each core constitutes a large data set within the meaning of your policies. There is an excellent public repository for ice core data at the World Data Center for Paleoclimatology, which satisfies your definition of a public repository. Under your policies, Thompson et al had an obligation to archive this data as a condition of publication, but this appears to have been overlooked. Although Thompson et al provided a highly abbreviated summary of isotope information as Supplementary Information, the Supplementary Information is incomplete and not compliant with journal policies.
There is a pressing need to ensure compliance with journal data policies, because numerous inconsistent summaries are in gray and peer-reviewed circulation. For example, the figure below illustrates substantial differences between Dunde δO18 data summaries. These discrepancies can only be reconciled through examination of the underlying large data sets, which should have been archived prior to publication had journal policies been followed.
Dunde Versions. Heavy black — Yao et al 2006 (3 year rolling average); thin black – MBH98 (annual); red – PNAS 2006 (5-year averages); blue – Clim Chg 2003 (10-year averages); purple – Yang et al 2002 (values in 50 -year intervals); green – Crowley and Lowery 2000 (original in standardized format, re-fitted here for display by regression fit to MBH98).
I request that you ensure that Thompson et al comply with your data policy by forthwith archiving the large datasets used in the PNAS article for each individual ice core (Dunde, Dasuopu, Guliya, Puruoganri, Quelccaya, Sajama, Huascaran) and for the entire suite of isotopes and chemistry. In addition, because the discrepancies may result from changing algorithms for dating the ice cores, I further request that the dating procedure for each core be made available under your Unique Materials policy.
Thank you for your attention.
Stephen McIntyre
We’ll see what happens. BTW the National Academy panel on Data Integrity, promised in the wake of the Surface Temperature Panel, has been empanelled and held its first hearings this week. Gerry North was the first speaker. He’s sent me a copy of his presentation which I’ll post on some day.
Climate Audit 