ORIGINAL: A little while ago I mentioned that the original SI at Nature for MBH98 had been deleted. It contained some information which is not in the Corrigendum SI, such as RE statistics for the various steps and the original (incorrect) data listing. Now Mann has either deleted the UMass SI for MBH98 and MBH99 previously located at ftp://eclogite.geo.umass.edu/pub/mann/ONLINE-PREPRINTS/MultiProxy and at ftp://eclogite.geo.umass.edu/pub/mann/ONLINE-PREPRINTS/Millennium, or he has put a block on my access (I’m blocked from the UVirginia server). Some, but not all, of this information formerly at UMass is at the WDCP archive. Mann had previously attempted to delete the UMass dataset in the aftermath of MM03, but the late John Daly protested to UMass and it was restored.
UPDATE: To add a little context to the deletion comment for recent followers of this debate, the original Mann deletion, which got a little publicity in 2003, was the deletion of a dataset showing MBH98 data from Mann’s website, the URL to which had been provided in response to my original request for MBH98. MM03 discussed many problems in this dataset. Mann’s response to MM03 was that this was the "wrong" dataset and was not the dataset used in MBH98, even though it was on his website, was the one to which I had been directed and on which I had sought specific re-confirmation prior to publication of MM03 that it was the dataset actually used in MBH98. Although Mann has sought to argue that we used the "wrong" dataset and failed to notice the errors in it (as recently as Rutherford et al ) , we had obviously noticed errors and re-collated over 300 tree ring series to avoid the problems in the principal component series. See MM03 Scorecard.
As an excuse for errors in this dataset, Mann falsely said that this data collation at his FTP site had been prepared especially for me, but it had actually been posted up in August 2002, about one week after the ftp://holocene.evsc.virginia.edu website was started, and long before my original request. The deletion removed this evidence, but I’d looked up the date (which I had not originally paid attention to) about 1 day before the deletion. (He also falsely said that I’d asked for an Excel spreadsheet, which was manifestly untrue, since I’d asked for an FTP location.) Nothing ultimately turns on this bizarre turn of events, but it was a huge surprise to me at the time and an interesting introduction into the world of academic politics. In business, you would never be able to delete a dataset in controversy (particularly when the date is an issue), but climate scientists didn’t seem to care. They mostly thought that Mann had taught me a lesson.
That’s why "Mann deletions" is something that concerns me.
UPDATE: 7 pm EDT April 8. I can access the UMass FTP site now. I’m still blocked from the UVA site and the original Nature SI is still deleted.
UPDATE: April 10 10 am EDT. There are some advantages to the amount of traffic on this website. After putting up the post below on Friday, my access to the UVA website has suddenly changed and today (Sunday) I can access the ftp://holocene.evsc.virginia.edu/pub site for the first time in over a year.
UPDATE: My access to ftp://holocene.evsc.virginia.edu/pub is blocked again.