date: Fri Sep  7 14:53:20 2001
from: Keith Briffa <k.briffa@uea.ac.uk>
subject: Fwd: 
to: t.osborn@uea.ac.uk

     X-Sender: berninge@mm.mappi.helsinki.fi
     X-Mailer: QUALCOMM Windows Eudora Version 5.1
     Date: Fri, 07 Sep 2001 16:42:50 +0300
     To: Keith Briffa <k.briffa@uea.ac.uk>, Frank.berninger@Helsinki.fi
     From: Frank Berninger <Frank.Berninger@Helsinki.fi>
     Subject:
     Cc: John Grace <jgrace@srv0.bio.ed.ac.uk>
     Dear Keith,
     After our telephone discussion.
     An additional remark: We will not only have to downscale the GCM.s but also upscale the
     growth models. There are different schools for that and I am in the school that requires
     less climate data to do the job.
     If we look at GCM downscaling I think we will run on two levels:
     I am not totally shure what to say about the GCM stuff.... On the one hand GCM:s are too
     bad (=do not agree with each other sufficiently) to justify a complicated effort to
     scale them down, on the other hand we would require weather data that is good enough (or
     our model results will be flawed.... I am remembering with horror a project where
     climate change was imposed on temperature but not absolute humdity and the trees got
     unhappy in dry desert air....) So I would give your free hands to decide.
     1) For a few sample sites (where models will be run at greater detail: Most models
     require:
     Temperature (i.e. daily max and min) , Irradiance, Air humidity, wind speed and
     precipitation
     For our systems the sensitivity to wind speed is very small. (i.e unless there are
     dramatic changes predicted by GCM:s forget about it. Precipitation is probably not so
     important on timberline.
     The locations will be
     3 Northern Scandinavia
     1 Scottland
     3 in the alps
     1 in the Southern Apeninnes
     1 in the Pyrenees
     2) For the rest of the area probably changes in temperature and irradiance at the ground
     on a monthly basis would be ok. (Depending on the upscaling scheme from the process
     models). Precipitation in the mountains is tricky and models will not be very sensitive
     to it.
     Tree ring modelling:
     I would like to see the work of Tom Melvin continued. I have made a trial to interpret
     chronologies using a kind of limitation hypothesis and found interesting results (see
     attachment (that is at the moment in the hands of a few referees of Nature). I think
     that this could be associated with more long term modelling approaches (like the ones of
     Tom) in different ways.,
     I would try to phone you again (probably on wednesday thursday)
     Frank
     Frank Berninger
     PhD, Docent
     Department of Forest Ecology
     POBOX 27
     00014 University of Helsinki / Finland
     Tel +358 9 191 58134
     Fax + 358 9 191 58100

   --
   Professor Keith Briffa,
   Climatic Research Unit
   University of East Anglia
   Norwich, NR4 7TJ, U.K.

   Phone: +44-1603-593909
   Fax: +44-1603-507784
   [1]http://www.cru.uea.ac.uk/cru/people/briffa[2]/

