Monday, December 12, 2011


From: Tim Osborn <>
To: "Michael E. Mann" <>, "Robert Matthews" <>
Subject: Re: Mann and Jones, climate of the last two millennia
Date: Fri Oct 3 09:56:06 2003
Cc: Phil Jones <>, Keith Briffa <>,,,,

Dear Mr. Matthews,
I have not read the criticism on the website you refer to, but will add to Mike Mann's
response in a small, but hopefully helpful, way.
Comparison of the Mann and Jones proxy-based reconstruction with instrumental temperature
data *is* a valid comparison to make, provided that the reconstruction is *calibrated* to
represent the instrumental record and provided that the *uncertainties* in the calibration
are taken into account when making the comparison.
That is, after all, the purpose of calibration - to allow two different data sets to be
As is clear from their article, Mann and Jones do undertake a careful calibration and only
make comparisons after the calibration, and their comparison figure includes their
estimated uncertainty range. Thus the conclusions they draw (regarding whether recent
warming is unprecedented) are valid and are supported by their analysis.
This does not mean that future work, perhaps using new proxy records or different methods
for calibration or for estimating calibration uncertainties, will not change those
conclusions. But it remains true that their conclusions are supported by their analysis.
As an example of a poor comparison, see the piece by Fred Pearce on page 5 of 12 July 2003
issue of New Scientist. This is a short news article about the Mann and Jones paper, and
it unfortunately shows a comparison figure without the associated calibration
uncertainties. That is not a good comparison. I mention this in case you were thinking of
including a diagram in your article, perhaps showing the Mann and Jones results. If you
do, then it will only be valid for comparing the recent instrumental temperatures with the
proxy-based reconstruction of earlier temperatures if the reconstruction uncertainties are
included. Try to avoid the mistake that Fred Pearce made.
At 21:11 02/10/2003, Michael E. Mann wrote:

Dear Mr. Matthews,
Unfortunately Phil Jones is travelling and will probably be unable to offer a separate
reply. Since your comments involve work that is his as well, I have therefore taken the
liberty of copying your inquiry and this reply to several of his British colleagues.
The comparisons made in our paper are well explained therein, and your statements belie
the clearly-stated qualifications in our conclusions with regard to separate analyses of
the Northern Hemisphere, Southern Hemisphere, and globe.
An objective reading of our manuscript would readily reveal that the comments you refer
to are scurrilous. These comments have not been made by scientists in the peer-reviewed
literature, but rather, on a website that, according to published accounts, is run by
individuals sponsored by ExxonMobile corportation, hardly an objective source of
Owing to pressures on my time, I will not be able to respond to any further inquiries
from you. Given your extremely poor past record of reporting on climate change issues,
however, I will leave you with some final words. Professional journalists I am used to
dealing with do not rely upon un-peer-reviewed claims off internet sites for their
sources of information. They rely instead on peer-reviewed scientific research, and
mainstream, rather than fringe, scientific opinion.
Michael E. Mann
At 08:30 PM 10/2/2003 +0100, Robert Matthews wrote:

Dear Professor Mann

I'm putting together a piece on global warming, and I'll be making reference to your
paper in Geophysical Research Letters
with Prof Jones on "Global surface temperatures over the past two millennia".

When the paper came out, some critics argued that the paper actually showed that there
have been three periods in the last 2000 years which were warmer than today (one just
prior to AD 700, one just after, and one just prior to AD 1000). They also claimed that
the paper could only conclude that current temperatures were warmer if one compared the
proxy data with other data sets. (For an example of these arguments, see:

I'd be very interested to include your rebuttals to these arguments in the piece I'm
doing. I must admit to being confused by why proxy data should be compared to
instrumental data for the last part of the data-set. Shouldn't the comparison be a
consistent one throughout ?

With many thanks for your patience with this
Robert Matthews



No comments:

Post a Comment