Friday, May 25, 2012


date: Tue, 24 Oct 2000 18:37:31 -0700
from: Ben Santer <>
subject: Figures for revised version of paper

Dear All,

Sorry that it has taken me so long to revise our paper. As I mentioned in a
previous email, I've had to repeat most of the calculations using an improved
estimate of the pre-eruption reference level temperature (Tref). I've also had
to look at the sensitivity of our results to uncertainties in Tref. I'd like to
thank Tom for prompting me to take a critical look at this issue - it's an
important one. I'd also like to thank the rest of you for all the comments that
you've sent me. I hope I've addressed them adequately in the revised paper.

Another major change is that, rather than giving results are based on a variety
of different filtering options -- e.g., estimation of volcano parameters from
unfiltered data (too noisy) and highly smoothed data (13-term Gaussian filter
leads to underestimate of volcanically-induced cooling) -- we now only give
results for our "best guess" filtering option,
a five-term binomial filter. We still discuss sensitivities to tau (the volcanic
signal decay time) and choice of ENSO index. Restricting attention to one
filtering option reduces the length of Tables, and hopefully improves the
clarity of the paper.

I've rewritten the discussion of the iterative method, and we now make it clear
that although this approach is automated, its implementation still involves a
number of subjective decisions (filter choice, choice of averaging period for
estimating pre-eruption reference temperature, choice of tau, etc.) Many of the
changes made here attempt to address useful comments that I received from Tom.

Lennart and Erich kindly provided me with the SLP data from the GSOP, GSO1 and
GSO2 integrations. Recall that we did not have this data previously, and so our
estimation of ENSO signals in GSO1 and GSO2 and of ENSO/volcano signals in GSOP
was based on simulated Nino 3.4 SSTs only. We've now also used the (simulated)
SOI to perform ENSO/volcano signal estimation.

Section 5 (discussion of ECHAM4/OPYC results) has been completely rewritten,
and the ordering of individual subsections should now be more logical. We
discuss the simulated Pinatubo signal first, then the "ENSO component" of
simulated temperature trends, and finally residual trends after the removal of
volcano and ENSO effects.

Today I'm sending you, as postscript attachments, the revised Figures for the
paper. To simplify things I've encoded the Figure number at the top of the
postscript file. I don't want to overload your mailboxes, so I'm sending the
Figures in two separate mail messages. There should be 11 Figures in total.
Tomorrow I'll send you the revised text of the paper and the Tables. Please let
me know if you have any problems printing these files. Note that all Figures
except Figure 7 are in color. Color is not essential for some of the Figures,
and in the next day or two I'll prepare black-and-white versions of Figues 3, 5,
6, 8, 9, 10 and 11. But for now I thought you might find it easier working with
the color versions.

I will be going in for surgery on November 1st, and am not sure how long it will
be until I get back to my office. I realize that it may not be feasible to
submit the paper before November 1st. But I'd really appreciate it if you could
send me comments before November 1st. These will keep me occupied while I'm
trying to get back on my feet!

With best regards,

Benjamin D. Santer
Program for Climate Model Diagnosis and Intercomparison
Lawrence Livermore National Laboratory
P.O. Box 808, Mail Stop L-264
Livermore, CA 94550, U.S.A.
Tel: (925) 422-7638
FAX: (925) 422-7675
Attachment Converted: "c:\eudora\attach\"

Attachment Converted: "c:\eudora\attach\"

Attachment Converted: "c:\eudora\attach\"

Attachment Converted: "c:\eudora\attach\"

Attachment Converted: "c:\eudora\attach\"

Attachment Converted: "c:\eudora\attach\"

No comments:

Post a Comment