Thursday, April 12, 2012


date: Fri, 6 Jun 2008 16:05:08 +0100
from: John Haslett <>
subject: SUPRAnet Session on Palaeoclimate modelling and statistical methods
to: Caitlin Buck <>, Edouard Bard <>, Paula Reimer <>, Mark Bateman <>, David Richards <>, Mel Knight <>, James Annan <>, Michel Crucifix <>, Julia Hargreaves <>, Gabi Hegerl <>, Michael Evans <>, Stephen Brooks <>, Nicki Whitehouse <>, Regine R�thlisberger <>, Atte Korhola <>, Barbara Wohlfarth <>, Keith Bennett <>, Maarten Blaauw <>, "Stephen T. Jackson" <>, Brian Huntley <>, Christoph Spoetl <>, Dan Charman <>, Lasse Holmstr�m <>, Clive Anderson <>, Caitlin Buck <>, Michael Goldstein <>, Andrew Millard <>, Andrew Parnell <>, Jonathan Rougier <>, Marian Scott <>, Eleanor Stillman <>, Rob Wilson <>, David Lowe <>, Keith Briffa <>, Tamsin Edwards <>, Peter Challenor <>, Philip Brohan <>, Katy Klauenberg <>, Richard Wilkinson <>, John-Paul Gosling <>, Julie M Jones <>, Hanna Sundqvist <>, Jo Brown <>, Oliver Heiri <>, Bo Li <>, William Austin <>, Lindsay Collins <>, Luke Skinner <>, Gerrit Lohmann <>, Michael Salter-Townshend <>

Dear all

This mail concerns the Session Wed 14:30 - 16:30
Palaeoclimate modelling and statistical methods

I'm mailing it to all, even though some of you will have interests focussed
on other parts of the meeting. I'm not sure that I have an up-to-date sub-list.

Jonty Rougier and I have been charged with structuring this part of the
workshop. The following are proposals.

We (JH, JR) would appreciate constructive criticism. We anticipate some
informal discussion on the Tuesday evening where we can finalise these
We make these proposals in the context of the overall workshop commitment
to deliver 'product'; see below.

We note that this will be the last of the formal sessions. It will have
been preceded by several sessions on 'evidence from' various sources. But
here the focus will be more explicitly on the workshop theme "uncertainty
in palaeoclimate reconstruction". There will however have been almost no
previous discussion on modelling per se.

We propose thus to structure the session by using the first hour to provide
'tutorials' on some basic issues, and then to use the remaining hour to
identify SOME of the open issues. We would be interested to hear from
people who might be able to play a role here, recalling that we do want to
finish the week with concrete next steps

As I understand it, there remains quite a lot of flexibility in how we use
our time, especially on the Thursday


Overall, it seems to us that there are two big inter-related questions for
THIS session:

Why is it so hard to compare proxy based reconstructions with
those from physical models as in GCMs?

How and where can we best use GCMs and proxies jointly to clarify
and reduce uncertainties about the palaeoclimate?


1 20 min tutorial on GCMs, barely touching on uncertainty
2 20 min tutorial on modelling 'system uncertainty', given several
sources of
information, each with its own uncertainty, including GCMs
3 10 min on current work at the interface of proxy reconstructions
and GCMs.

This will be followed by a less formal, but nevertheless structured,
discussion on 'open questions' and 'the way forward'.
For this latter part of the session, we hope to identify a VERY few people
who will be prepared to lead discussion
and to contribute to 'product'. Such people might each have prepared a VERY
short (<5 mins) presentation.

Open questions include:


Recall that two forms of 'product' have already been identified:

1) a position paper on quantifying and
modelling uncertainty in palaeoclimate reconstruction intended for
a major journal such as Nature Geoscience and

2) plans for several (c. 4) funding proposals to be written over
the 6-9 months following the meeting (for which we have extra
travel funds).

John Haslett E-mail
Professor Phone +353 1 8961114 (direct)
Department of Statistics +353 1 8961767 (sec)
School of Computer Science Fax +353 1 6770711
and Statistics Room 146
Trinity College
Dublin 2,Ireland



No comments:

Post a Comment