cc: Jonathan Overpeck <jto@u.arizona.edu>, Stefan Rahmstorf <rahmstorf@ozean-klima.de>, Anders Levermann <levermann@pik-potsdam.de>, Eva Bauer <eva.bauer@pik-potsdam.de>, Eystein Jansen <eystein.jansen@geo.uib.no>, Keith Briffa <k.briffa@uea.ac.uk>,oyvind.paasche@bjerknes.uib.no
date: Wed, 15 Feb 2006 16:42:33 +0000
from: Tim Osborn <t.osborn@uea.ac.uk>
subject: Re: Fwd: Re: Millennium Simulations
to: Fortunat Joos <joos@climate.unibe.ch>,plattner@climate.unibe.ch

<x-flowed>
Dear all,

I have redone all plots for the two alternative baseline periods - 
see attached PDF.  Please look at them on the screen as well as 
printing them, because with some printers you can hardly see the 
paler greys and the yellow lines whereas they seem quite bright on my screen.

Page 1 is 1500-1899.  Page 2 is 1961-1990.

Mine and Keith's opinion:

1500-1899 looks better for panels (a), (c) and (d).  They look 
equally as good for panel (c).  For panel (e) we are unsure and below 
are some arguments for and against.

[Obviously I can tidy up the 1961-1990 version a bit more, in terms 
of labelling etc., though it is clearly going to be tricky to find 
gaps for the key and titles and the vertical scale of panel (d) would 
need to be changed/extended a bit and then wouldn't match the scale I 
used for panel (e).  So 1961-1990 is a bit harder to get everything 
looking good and consistent.]

At first we thought that the new 1961-1990 version of panel (e) 
looked better for the reason that there is clearer separation between 
the "all forcings" (thick lines) and the "natural-only forcings" 
(thin lines) in the early 20th century.

On closer inspection, however, we then were swayed back to the 
1500-1899 version of panel (e), because the reason for the clearer 
separation of the "Nat" and "All" runs in 1961-1990 version is that 
the stronger solar forcing runs (dark and pale blue and green) are 
pushed downwards.  But pushing them downwards means that during the 
"Little Ice Age" period these runs (especially the dark and pale blue 
ones) are clearly in the bottom part of the range of the 
reconstructions relative to 1961-1990 - and the question is why 
should we say that the "Nat" runs cannot capture the first phase of 
20th century warming when we have started them from cooler 
conditions, purely on the basis of the amount of warming achieved in 
the other runs by the time the reference period is reached.  This 
seems harder to defend.  It relates back to my earlier comments about 
(1) using as long a reference period as possible; and (2) thinking 
about climate changing from the relative stable period, rather than 
going backwards from the present period with its strong transient changes.

Views please?

If the decision is made to go with 1961-1990, then Keith suggests 
sticking with 1500-1899 for panels (a)-(d) as before, and make the 
new EMIC runs (currently panel (e)) into a stand-alone figure with 
1961-1990 baseline.

Views required urgently!

Cheers

Tim

</x-flowed>

Attachment Converted: "c:\eudora\attach\modelsA-E_2versions.pdf"
<x-flowed>
Dr Timothy J Osborn
Climatic Research Unit
School of Environmental Sciences, University of East Anglia
Norwich  NR4 7TJ, UK

e-mail:   t.osborn@uea.ac.uk
phone:    +44 1603 592089
fax:      +44 1603 507784
web:      http://www.cru.uea.ac.uk/~timo/
sunclock: http://www.cru.uea.ac.uk/~timo/sunclock.htm
</x-flowed>
