CORRESPONDENCE WITH MANNNovember 11, 2003
Professor
Michael E. Mann School
of Earth Sciences University
of Virginia Dear
Professor Mann, We
apologize for not sending you a copy of our recent paper (“MM”) in Energy
and Environment for comment, as we understood from your email of September
25, 2003 that time constraints prevented you from considering our material. We
notice that you seem to have subsequently changed your mind and hope that you
will both be able to clarify some points for us and to rectify the public record
on other points. 1)
You have claimed that we used the wrong data and the wrong computational
methodology. We would like to reconcile our results to actual data and
methodology used in MBH98. We would therefore appreciate copies of the
computer programs you actually used to read in data (the 159 data series
referred to in your recent comments) and construct the temperature index shown
in Nature (1998) (“MBH98”), either through email or, preferably through
public FTP or web posting. 2)
In some recent comments, you are reported as stating that we requested an Excel
file and that you instead directed us to an FTP site for the MBH98 data. You are
also reported as saying that despite having pointed us to the FTP site, you
and your colleague took trouble to prepare an Excel spreadsheet, but
inadvertently introduced some collation errors at that time. In fact, as you no
doubt recall, we did not request an Excel spreadsheet, but specifically asked
for an FTP location, which you were unable or unwilling to provide. Nor was an
Excel spreadsheet ever supplied to us; instead we were given a text file, pcproxy.txt.
Nor was this file created in April 2003. After we learned on October 29, 2003
that the pertinent data was reported to be located on your FTP site ftp://holocene.evsc.virginia.edu/pub
(and that we were being faulted for not getting it from there), we examined this
site and found it contains the exact same file (pcproxy.txt) as the one we received, bearing a date of creation of August 8,
2002. On October 29, 2003, your FTP site also contained the file pcproxy.mat, a Matlab file, the header to which read: “MATLAB 5.0 MAT-file,
Platform: SOL2, Created on: Thu Aug 8
10:18:19 2002.” Both files contain identical data to the file pcproxy.txt
emailed to one of us (McIntyre) in April 2003, including all collation errors,
fills and other problems identified in MM. It is therefore clear that the file pcproxy.txt
as sent to us was not prepared in April 2003 in response to our requests, nor
was it prepared as an Excel spreadsheet, but in fact it was prepared many months
earlier with Matlab. It is also clear that, had we gone to your FTP site
earlier, we would simply have found the same data collation as we received from
Scott Rutherford. Would you please forthwith issue a statement withdrawing and
correcting your earlier comments. 3)
In reported comments, you also claimed that we overlooked the collation errors
in pcproxy.txt and “slid” the incorrect data into our calculations,
a statement which is untrue and made without a reasonable basis. In MM, we
described numerous errors including, but not limited to, the collation errors,
indicating quite obviously that we noticed the data problems. We then describe
how we “firewalled” our data from the errors contained in the data you
provided us, by re-collating tree ring proxy data from original sources and
carrying out fresh principal component calculations. We request that you
forthwith withdraw the claim that we deliberately used data we knew to be
in error. 4)
On November 8, 2003, when we re-visited your FTP site, we noticed the following
changes since October 29, 2003: (1) the file pcproxy.mat had been deleted
from your FTP site; (2) the file pcproxy.txt no longer was displayed
under the /sdr directory, where it had previously been located, although it
could still be retrieved through an exact call if one previously knew the exact
file name; (3) without any notice, a new file named “mbhfilled.mat”
prepared on November 4, 2003 had been inserted into the directory. Obviously,
the files pcproxy.mat and pcproxy.txt are pertinent to the
comments referred to above and we view the deletion of pcproxy.mat
from the archival record under the current circumstances as unjustifiable. Would
you please restore these files to your FTP site, together with an annotated text
file documenting the dates of their deletion and restoration. 5)
We note that the new file mbhfilled.mat is an array of dimension
381x2016. Could you state whether this file has any connection to MBH98, and, if
so, please explain the purpose of this file, why it has been posted now and why
it was not previously available at the FTP site. 6)
Can you advise us whether the directory MBH98 has been a subdirectory within the
folder “pub” since July 30, 2002 or whether it was transferred from another
(possibly private) directory at a date after July 30, 2002? If the latter, could
you advise on the date of such transfer. We have prepared a 3-part response to your reply to MM. The first, which we have released publicly, goes over some of the matters raised in points #2-#5 above. The second is undergoing review. It deals with additional issues of data quality and disclosure, resulting from inspection of your FTP site since October 29, 2003. The third part will consider the points made in your response, both in terms of data and methodology, and will attempt a careful reconciliation of our calculation methods, hence the necessity of our request in point #1. Thank you for your attention.
Yours
truly, Stephen
McIntyre
Ross McKitrick cc: Timothy Osborn |