date: Tue, 03 Aug 2004 07:55:09 -0600
from: Tom Wigley <wigley@cgd.ucar.edu>
subject: question etc.
to: Sarah Raper <sraper@awi-bremerhaven.de>

Sarah,

In a paper by Robert Nicholls he says "the 9-88cm range in the
IPCC TAR does not include the additional uncertainties from
changes in terrestrial water storage". True -- but for the AR4?
Any idea?

Did you see Gregory's comments on our volcano paper?

Thanx for the notes from Paris. Some comments ...

(1) I pointed out the DT2x vs Qaer thing years ago when I reviewed
the first Had2 paper claiming a good 20th century fit. It is pretty
obvious, but often ignored. I believe James Murphy put in a sentence
to cover this in the above-mentioned paper -- but I still judge the
modellers (including Jerry Meehl and others at NCAR)as being
dishonest here.

(2) 1% runs. Collins' number is irrelevant -- but there are uncerts
in Q2x that we could look at.

(3) Re Myles, what is the 'Allen, Raper Mitchell' policy piece?

(4) I first estimated XKLO from the seasonal cycle. On this basis
it cannot be too high. We could put in a finite land heat capacity.
I do not think the numbers have to be 'realistic', whatever that
means. It is just a tuning  parameter that has to be balanced
in a statistically sensible way against RLO.

(5) We will need to make an effort to include every  AOGCM and
avoid upsetting anyone.

(6) No problem re convergence of AOGCM DT2x. We didn't
use AOGCMs in our Science paper -- and the reason (I hope)
that you stuck to 1.7-4.2 in the TAR was because this was a
fair estimate of the range based on other evidence. At the high
DT2x end the sensitivity to sensitivity is less.

It is the 'full evidence' range that must be covered. (It not, the AR4
will be strongly criticized.) To base a DT2x range on AOGCM
results alone would be crazy.

(7) We really need to push for forcing information.

----------

We have a lot to discuss when I visit soon. It is possible that Eirik
will come to Norwich then too -- altho he will stay with James
and Camilla.

Tom.

