from: "Michael E. Mann" <mannatXYZxyzginia.edu>
subject: Re: revised manuscript
to: Tim Osborn <t.osbornatXYZxyz.ac.uk>, Scott Rutherford <srutherfordatXYZxyz.edu>, Bradley Raymond <rbradleyatXYZxyz.umass.edu>, Hughes Malcolm <mhughesatXYZxyzr.arizona.edu>, Briffa Keith <k.briffaatXYZxyz.ac.uk>, Jones Phil <p.jonesatXYZxyz.ac.uk>, mannatXYZxyzginia.edu
Thanks--your suggestions (w/ Keith and Phil) all are very helpful, and improve the
manuscript significantly. This was exactly what was needed!
A couple minor clarifications on some points that were raised:
1. In discussing how spatial reconstructions are averaged to get hemispheric means, we just
meant that, not that these spatial reconstructions were the results of CFR approaches. I
believe it is true that Briffa et al, 1998a does average gridpoint reconstructions to get a
hemispheric mean, right? We should clarify the wording to make sure that this is clear. The
issue of different approaches (CFR vs local calibration) is discussed later in the paper...
2. Re, McINtyre/McKitrick: they didn't identify any "errors" at all. They (a) used an
incorrect spreadsheet representation of the data (which Scott prepared based on a
misunderstanding of precisely what they were looking for, since the predictors cannot be
included in a single matrix because of the stepwise nature) and (b) used different versions
of certain series that were on the NGDC site. There were some minor errors in the
supplementary information, and we have a corrigendum in press in "Nature" that makes note
of these. I agree that we should add a sentence clarifying how PCs were calculated in the
PC/proxy network at the beginning of that section.
3. You're right that RE is not necessarily less than r^2 in calibration. That statement is
4. I believe that including weights in the hybrid approach that are proportional to the
fraction of variance resolved by the indicator is essential. In fact, we know that we get
completely wild results if we don't do that. The reason is that an indicator that has
essentially no low-frequency variability left (say, an already pre-whitened tree-ring
chronology) has essentially only noise left in the low-frequency band. Now, if you use the
weighting convention we use, it gets almost no weight. On the other hand, if you give it
full weight in that band, you magnify that tiny bit of residual noise up to the full
amplitude of all other indicators, and now you've got completely noise competing w/ other
series that have much more signal. I think a sentence needs to be added to clarify this
point, and also a sentence added that indicates that highly unconstrained results are
possible if this is not done. The example of weights of 0 and 1 is an extreme limiting
case, which is probably never realized (Scott?).
5., Re the ECHO-G simulations, yes we can certainly downplay the discussion of this in our
response. However, I stand by the claim that they appear to use a solar forcing that is
about twice that of Crowley's! In the talk they gave in Nice (Phil was there), they showed
their solar constant varying between 1362 and 1368 W/m^2 over the past millennium! That is
much larger than Crowley. As no details are provided in the Gonzalez-Rouco paper, this is
very difficult to discern. However, the huge Medieval peak, and the claim that 40% of 20th
century warming is solar-driven (they make both claims) is ONLY consistent w/ a much larger
solar forcing then anyone else has used. Christiensen's group has used Crowley's forcings
on the same exact model, and gets nothing near the amplitude of variability they got (like
the GKSS paper, I believe this paper is in submission). There is something very odd in what
the GKSS group has done, and I think Phil agrees w/ this. It would be unfortunate if the
issue of precisely how large a solar forcing they used isn't examined closely before they
publish their results in more detail.
But I agree we don't need to get into this in the paper or the response. We can make the
point, as you suggested, that as shown in Jones and Mann, the reconstructions are broadly
consistent with the vast majority of simulations that have been done...
I don't think it should be a problem waiting about a week for Scott to resubmit (in time
for the Osborn et al '04 paper to have been submitted). We still need to give Malcolm and
Ray a chance to get back w/ any additional comments, if they have them...
At 07:00 AM 6/16/2004, Tim Osborn wrote:
Dear Scott et al.
the manuscript reads well - the results section in particular does a good job at
presenting the sequence of many experiments in a logical and structured way.
We've combined all our (me, Keith and Phil) comments together, and put them in the
attached files (one for the manuscript, one for the response to the editor).
Many are simple changes. Some are questions/comments for you, that I've put in
CAPITALS. I've also written an improved description of the MXD data.
And the "Osborn et al. (in preparation)" paper that describes the gridding and
low-frequency aspects of the MXD data (as well as our "local" calibration of it) is now
referred to as Osborn et al. (2004, submitted to Global and Planetary Change). It is
not, in fact, submitted as I write, but should be very soon - a full draft is complete
and I have comments from Phil and Fritz, and Keith is halfway through providing his
comments. I'll let you know as soon as its submitted - it would be preferable if you
could hold off returning your revised manuscript until we have actually sent off Osborn
et al. - it should be done by the end of the week. Then "Osborn et al. (submitted)"
will be true!
At 19:15 28/05/2004, Scott Rutherford wrote:
Attached is a revision of the Northern Hemisphere comparison for J. Climate. Also
attached is a reply to reviewers. (both microsoft word format)
I've been getting extensions for the submission of the revision and would like to try to
get it in around June 15. If you want to make suggestions/comments directly in the text,
please use Microsoft Word's "track changes" feature if possible. There are a few
highlighted parts that need particular attention.
Please let me know if you have question or difficulty with the files.
Dept. of Natural Sciences
Roger Williams University
phone: (401) 254-3208
One Ferry Road
Bristol, RI 02809
Dr Timothy J Osborn
Climatic Research Unit
School of Environmental Sciences, University of East Anglia
Norwich NR4 7TJ, UK
phone: +44 1603 592089
fax: +44 1603 507784
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