cc: Scott Rutherford <srutherfordatXYZxyz.uri.edu>, Zhang <email@example.com>, firstname.lastname@example.org, Tim Osborne <t.osbornatXYZxyz.ac.uk>, Keith Briffa <k.briffaatXYZxyz.ac.uk>, Irina Fast <f14atXYZxyzat.fu-berlin.de>, email@example.com, rbradleyatXYZxyz.umass.edu
date: Wed, 05 Feb 2003 13:19:29 -0500
from: "Michael E. Mann" <mannatXYZxyztiproxy.evsc.virginia.edu>
subject: Re: program code
The code we used in Mann/Bradley/Hughes 1998 was not changed or "improved", but there may
be different versions of the code floating around, and in a previous email to Uli Cubasch,
I indicated that I was not sure the version you have (from Tim Osborn), is identifical to
the version we used in our original paper (it would require some work on my part to insure
it gives precisely the same results, and I don't have the time to do that). I suspect,
however, that the code is the same as the one we used in our paper and any differences, if
they exist, should be minor (as long as the code compiles and runs correctly on the
platform you have--the possible platform-dependence of fortran is a potential cause for
Numerous people have coded up our method independently, including Ed Zorita, w/ whom I
believe your group has a close collaboration, and my graduate student Zhang has
successfully coded this up independently in Matlab (its a short script, which didn't take
Zhang long to write anyway). I'm copying this message to Zhang, so that he can provide you
with his matlab version of the code if you are interested. Because Zhang's version is in
Matlab, it should run correctly, independently of the particular platform (an advantage
over the fortran code) [As an aside, on a pedagogical note, I would still encourage you to
code this up yourself].
As I indicated in a previous email to Uli, the selection of the optimal subset of EOFS to
retain is not automated in the code, and you need to do that yourself...The methodology we
used is described in detail in our publications.
We have tested this method against the approach our group now uses for climate field
reconstruction (Schneider RegEM approach), and find that the results are similar, but the
cross-validation statistics improve slightly w/ the RegEM approach, which we now favor and
use in place of the old, Mann et al approach.
Details of this latter approach are described in these two manuscripts (as well as the
original paper by Schneider referenced within):
Mann, M.E., Rutherford, S., Climate Reconstruction Using 'Pseudoproxies', Geophysical
Research Letters, 29 (10), 1501, doi: 10.1029/2001GL014554, 2002.
Rutherford, S., Mann, M.E., Delworth, T.L., Stouffer, R., Climate Field Reconstruction
Under Stationary and Nonstationary Forcing, Journal of Climate, 16, 462-479, 2003.
The RegEM code is available over the web, and Scott Rutherford can provide you with the ftp
side if you are interested. It, too, is available only in matlab.
I hope you find this information of help.
Best of luck w/ your research,
At 06:10 PM 2/5/03 +0100, Irina Fast wrote:
I believe that you have not heard about me as yet. My name is Irina Fast.
Since the January 2003 I am a PhD student at the Free University in Berlin in
the framework of the EU-Project SOAP. My supervisor is Ulrich Cubasch.
At the SOAP's start-up meeting it was proposed to use your multiproxy
calibration method (published in 1998) for the joint analysis of model
simulations and proxydata.
Because your method was essential improved since 1998 I would like to know if
you kann provide us with your program code.
We could try to code your approach ourselves, but we do not know if this kind
of analysis will success in our case. In the case of failure we will have to
search for other analyses methodes. And the timespan for the data
processing is rather short. Naturally you will not miss our gratitude and
I apologise for my mistakes in this letter.
Freie Universit�t Berlin
Institut f�r Meteorologie
phone: +49 (0)30 838 711 22
fax: +49 (0)30 838 711 60
Professor Michael E. Mann
Department of Environmental Sciences, Clark Hall
University of Virginia
Charlottesville, VA 22903
e-mail: mannatXYZxyzginia.edu Phone: (434) 924-7770 FAX: (434) 982-2137