cc: <bane@cs.man.ac.uk>
date: Tue, 9 Dec 2003 20:02:19 -0000
from: "Rachel Warren" <R.Warren@uea.ac.uk>
subject: new versions of MAGICC/SCENGEN
to: "christopher barton" <c.barton@uea.ac.uk>, <m.hulme@uea.ac.uk>

   Chris,  Mike,



   FYI : the new versions of MAGICC and SCENGEN have been provided to me as executables.

   So far, Tom Wigley has declined to give me the source code!   Apparently he is not giving
   the

   source code of the new SCENGEN to anyone at all.  Tim Mitchell thinks it's no great loss
   and that

   he could easily create an alternative, which he plans to do anyway as a part of Nigel
   Arnell's Tyndall round 3 project.



   However, the more important question is the new version of MAGICC itself.

   Tom has offered to help convert my existing MAGICC code to a version which will be
   consistent with the new code.



   The reason for this is that the new version of SCENGEN, which he is not allowing anyone to
   have, requires different inputs from MAGICC to the orginal version of SCENGEN (so the new
   version of MAGICC won't go with the old version of SCENGEN).



   However Tom has taken on board my point that our results should be consistent, and offered
   to take my magicc.tar

   code and modify it so that it is scientifically the same as the new version.  There were
   some significant changes (scientifically) between the version I was given and the new
   version, connected with ice melt and tropospheric ozone forcing to name a couple.  He and I
   are having an ongoing e-mail conversation about the conversion ... I will let you know when
   I have a new version of code consistent with the latest version.



   I don't anticipate this creating any work for the softIAM team since inputs and outputs
   remain the same.

   I will simply provide revised code for MAGICC (SCM) when the time comes.

   Then we will just need to do a test case within and without the softIAM framework.



   When Tim Mitchell creates a new SCENGEN version next summer, the softIAM team (possibly
   just myself by that time)

   will need to talk to him about how this affects the interface within softIAM if we want to
   incoporate the new version.



   Rachel
