from: Hugues Goosse <hgsatXYZxyzr.ucl.ac.be>
subject: Re: IPCC AR4 model runs of last 1000 years
to: Tim Osborn <t.osbornatXYZxyz.ac.uk>
I have an additional comment compared to our phone call yesterday. The
forcing that I have send you are global mean, but aerosols and
deforestation forcing are mainly occuring in the northern Hemisphere. If we
use forcing for NH only, we would get a decrease of 0.5 W m-2 at least
compared to global mean. This means that our forcing would be lower than
the one of the GKSS model that only includes global forcings (Solar, GHG,
no latitude change in volcanos). This does not change the fact that our
model has a low climate sensitivity but this probably helps to analyse the
figure more precisely.
I have read your paper about ECHO-G and MAGICC. To my point of view, it
brings some very useful information compared to the paper we have written
on a similar subject. In particular, it brings some quantitative
information that were out of the scope of our brief study. The first point
(which was the starting point of our experiments) is that we agree that the
differences between ERIK simulation and other model results is not due to
the forcing used except during the last 100 years, as the forcing of ECHO-G
does not include aerosols forcing.
We also agree that the drift (or experimental design) is responsible for a
significant part of the differences between models. Nevertheless, you
suggest that the experimental design is the main source of the differences
while, using a much simpler method (too simple ?), we were more in favour
of a larger role of climate sensitivity after the first 250 years of the
simulations. This first depends on the models that are compared. For
ECBILT-CLIO that has a low sensitivity, the different climate sensitivity
certainly plays a large role when results are compared to those of ECHO-G.
The role of climate sensitivity is probaby weaker when ECHO-G is compared
to models that have a climate snesitivity in the range 2.5-3 K.
This of course also depends on the climate sensitivity of ECHO-G itself.
Your analysis suggest a climate sensitivity around 2.4 while Stuber et
al.'s give about 3. If I have understand well your figure 6, despite your
correction, the agreement between MAGICC and adjusted ECHO-G is not perfect
for the period 1050-1300. With a higher sensitivy like the one proposed by
Stuber et al., you would get probably higher temperature during this period
in MAGICC (and thus a better agreement with corrected ECHO-G) but then a
worse agreement during some other parts of the simulation. This is more or
less what we get : a reasonable scaling for low frequency variations during
the pre-industrial period was not working well for the last 100 years (we
were not hoping to match high frequency changes like the ones associated
with volcanos). This might also be related to a different sensitivity of
the ECHO-G to different forcings (like shown for instance by Gregory et al.
2004 for the HADCM3).
This illustrates that indirect comparisons are difficult and uncertainties
will certainly remain after any indirect comparison. Nevertheless, We
probably agree that those comparisons can provide some useful information
when it is not possible to use the original models because of cpu time
All the best
At 15:07 13/06/2005 +0100, you wrote:
>please see attached figure. Still in progress - more lines to add, plus
>legend is incomplete.
>You were right that Crowley et al. (2003) use SO4 aerosol forcing that
>reaches about -1 W/m2 by 2000. The lower value I told you comes from some
>of their newer simulations!
>Your data are a dirty yellow colour. (c) shows "all other forcings" which
>should include everything except volcanic and solar irradiance. You will
>see that ECHO-G (red) is highest from 1850 to late 1900s, as expected
>because of no SO4 cooling. But your data catch up and reach ECHO-G by
>1990, despite SO4 cooling and land-use cooling. Hence your GHGs are
>stronger (ECHO-G GHG was +1.8 W/m2 in 1990). They include only CO2, CH4
>and N2O. Maybe that explains it?
>Anyway, I think (hope) that everything is correct, but I would be please
>if you could check that the "Goosse2005" series look right.
>At 11:18 13/06/2005, you wrote:
>>Yes, it is the direct forcing applied in the model (TSI changes) not the
>>forcing at the tropopause.
>>At 09:55 13/06/2005 +0100, you wrote:
>>>At 08:28 02/06/2005, you wrote:
>>>>I have attached the file I send to Pascale. It includes Solar forcing,
>>>a quick question - is the solar forcing the irradiance change, and thus
>>>should I multiply by 0.25*(1-albedo) to convert to a forcing comparable
>>>to GHG forcing?
>>>Dr Timothy J Osborn
>>>Climatic Research Unit
>>>School of Environmental Sciences, University of East Anglia
>>>Norwich NR4 7TJ, UK
>>>phone: +44 1603 592089
>>>fax: +44 1603 507784
>>Institut dAstronomie et de G�ophysique G. Lema�tre
>>Universit� catholique de Louvain , Chemin du cyclotron, 2
>>1348 Louvain-la-Neuve, Belgium
>Dr Timothy J Osborn
>Climatic Research Unit
>School of Environmental Sciences, University of East Anglia
>Norwich NR4 7TJ, UK
>phone: +44 1603 592089
>fax: +44 1603 507784
Institut dAstronomie et de G�ophysique G. Lema�tre
Universit� catholique de Louvain , Chemin du cyclotron, 2
1348 Louvain-la-Neuve, Belgium