date: Mon Oct  4 16:25:07 2004
from: Tim Osborn <t.osborn@uea.ac.uk>
subject: Re: past 1000 yr
to: Tom Wigley <wigley@cgd.ucar.edu>

   Hi Tom - I'd be happy to contribute if I have something worth contributing!  I'm a bit
   rushed today and away tomorrow, but can respond to further emails later in the week.
   At 14:31 03/10/2004, Tom Wigley wrote:

     Caspar Ammann and I plan to publish some MAGICC
     results for the past 100 years.

   Presume you mean 1000 years, hence relevance of ECHO-H/von Storch.

      Part of the reason is the new
     solar forcing, as in my Science note with Peter Foukal.

   Yes I saw that.  With a brief scan I didn't realise that you were presenting a new forcing
   history, just discussing reasons why long-term changes may be lower than previously
   estimated.  But presumably you can use such reasoning to develop a new forcing history -
   or, better, a range or even a PDF of such histories.  And then extend it using 14-C or
   10-Be, or a combination?

     So we
     address both forcing and senstivity uncertainties. In
     addition, the drift due to incorrect initialization is an issue.

   Surely not so in MAGICC?  But yes, it is in GCMs and particularly so in ECHO-G.

     I have not yet read the Storch paper or your comment -- but
     did you mention this problem?

   We said that ECHO-G had a redder spectrum than other model simulations (there was no room
   to say that it showed greater fluctuations, but we cited the Jones/Mann paper which has an
   intercomparison figure in it).  We didn't talk about the reasons for this (drift early on,
   strong solar forcing throughout and no tropospheric aerosols to mitigate recent warming)
   because we'd already said that the simulation didn't necessarily represent real climate
   history.

     Also, can you remind me just what was done with the ECHO
     run?

   Main problem in terms of introducing "drift" (or "adjustment") was that they used a control
   run with present day CO2 as initial conditions.  Although they allowed a 70-year spin-up
   (prior to AD 1000) to adjust back to pre-industrial CO2, this doesn't look long enough and
   the adjustment probably goes on for the first 400 years of the run - i.e. there is
   gradually disappearing cooling trend over this period.  All based on MAGICC runs, but still
   fairly convincing (including non-zero heat flux out of the ocean in ECHO-G itself).

     If you have something to add on this, you can join as a co-author.

   I'm not quite sure what you plan, nor the input you need, but hopefully I can help.
   Cheers
   Tim
