cc: "'Burgess Jacquelin Prof \(ENV\)'" <Jacquie.Burgess@uea.ac.uk>, <t.osborn@uea.ac.uk>
date: Mon, 14 Jan 2008 20:08:57 -0000
from: "Tim Lenton" <T.Lenton@uea.ac.uk>
subject: RE: UKCIP08
to: "'Phil Jones'" <p.jones@uea.ac.uk>, <a.j.watson@uea.ac.uk>

Phil, Tim,

Thanks for the update.

I wasn't criticizing UKCIP08, or at least I wasn't thinking of it at all in
my comments, and I was probably over-stating my case. I was criticizing what
Hadley Centre and others have claimed they can offer in terms of regional
predictions.

I do have serious doubts about the merits of down-scaling model output in
regions where the pattern of change and therefore the boundary conditions
for the regional model are not conserved across an ensemble of global model
runs. The key question then is, is the UK one of these regions? (From what I
have seen, it looks to be.)

In such circumstances I think a probabilistic approach has to be the way to
go.

Tim

-----Original Message-----
From: Phil Jones [mailto:p.jones@uea.ac.uk] 
Sent: 12 December 2007 13:52
To: t.lenton@uea.ac.uk; a.j.watson@uea.ac.uk
Cc: Burgess Jacquelin Prof (ENV); t.osborn@uea.ac.uk
Subject: UKCIP08

  Tim, Andy,
      Tim Osborn mentioned to me this morning that you were a little
skeptical
  about what UKCIP08 is trying to do.  Rather than knock it without knowing,
I
  would be happy to explain it to you. I'm attaching a press release about
the first of
  the reports (one what has happened to climate up to the present).
     Future climate change is uncertain, and this will be fully expressed in
the
  remaining reports which will come out with the launch in October 08.
However,
  if organizations of whatever kind are to react or do anything, they need
to
  know what is most likely. UKCIP have discussed extensively with user
  communities across the country and in most cases UKCIP08 will be giving
  them the best information currently available (with our current range of
  uncertainties).
      You will note I've referred to UKCIP and UKCIP08. Between us these are
  two different things. UKCIP doesn't fully understand what UKCIP08 will
provide
  and this is a problem. UKCIP are going around the country finding out what
is
  wanted and saying what UKCIP08 will provide. UKCIP08 is developing what
  it will provide trying to incorporate as much as the users want, but if we
know
  we will have no confidence to say anything we won't.
     UKCIP will have to educate the user community on how to use the UKCIP08
  portal and more importantly how the 'output' should be used. No-one at
UKCIP
  realises this.
      In essence, users want spatially detailed (5km) and high-temporal
(sometimes hourly)
  information that users claim they need. We do ask the obvious questions on
  what are they using now!  The HC produce this at 25km (monthly) from RCM
runs, then emulated
  in the way David Sexton presented at a seminar in ENV back in October.
This gives
  pdfs of variables which the Weather Generator (which CRU and Newcastle are
  developing) transforms into future weather. Users will then get 100
possible 30
  year realizations (they won't be able to get less than this) of the
percentile range
  they choose (e.g. 90th percentile of the single pdf for precip or a joint
one
  with another variable, such as T).
     So users should then put these data through their impact model (say a
crop
  model for wheat) to develop a pdf of their result (say yield).  So from
one pdf (precip/temp
  - depending on emissions scenario) to another pdf (of the effect on
yield).
This is the bit users will and are finding difficult.

     There are users, big and small in loads of sectors. With this approach,
they
  will all get consistent scenarios (termed probabilistic projections - but
not
  probabilistic is sense of dice throwing). They need something for planning
purposes - it might as well
  be done by groups best able to do it and explain it. UKCIP08 is only 4
groups
  and only 4 people who know how and why it is being done the way it is. It
is
  a quantum leap in scenario provision.

     You're not alone in knocking UKCIP08. The second attachment really is
in
  confidence. I hope you won't be going down to DEFRA to complain about what
  UKCIP08 is doing. I spent two fruitless hours explaining what UKCIP08
  will do to Lenny Smith in November. He clearly didn't listen.  Maybe he
thinks
  his expertise will be sidelined by what UKCIP08 will supply, but we need
  people to interpret to the impacts community what we've done (we don't
  have much written so far) rather than gripe and say they'll go public if
they
  don't get a piece of the cake!

  Cheers
  Phil


Prof. Phil Jones
Climatic Research Unit        Telephone +44 (0) 1603 592090
School of Environmental Sciences    Fax +44 (0) 1603 507784
University of East Anglia
Norwich                          Email    p.jones@uea.ac.uk
NR4 7TJ
UK 
----------------------------------------------------------------------------


