Monday, June 18, 2012


cc: Ronald Stouffer <>
date: Fri, 19 Sep 2008 04:01:15 -0700
from: Karl Taylor <>
subject: Proposed experiment design for CMIP5 (revised)
to:, Curtis Covey <>, Jerry Meehl <>, "Mitchell, John FB (Chief Scientist)" <>,,, Andreas Hense <>, Asgeir Sorteberg <>, Erich Roeckner <>, Evgeny Volodin <>, "Gary L. Russell" <>, Gavin Schmidt <>,, Greg Flato <>, Helge Drange <>, Jason Lowe <>, Jean-Francois Royer <>, Jean-Louis Dufresne <>, Jozef Syktus <>, Julia Slingo <>, Kimoto Masahide <>, Peter Gent <>, Qingquan Li <>, Seita Emori <>, Seung-Ki Min <>, Shan Sun <>, Shoji Kusunoki <>, Shuting Yang <>, Silvio Gualdi <>, Stephanie Legutke <>, Tongwen Wu <>, Tony Hirst <>, Toru Nozawa <>, Wilhelm May <>, Won-Tae Kwon <>, Ying Xu <>, Yong Luo <>, Yongqiang Yu <>, Kamal Puri <>, Tim Stockdale <>, Gabi Hegerl <>, James Murphy <>, Marco Giorgetta <>, George Boer <>, Myles Allen <>, claudia tebaldi <>, Ben Santer <>, Tim Barnett <>, Nathan Gillett <>, Phil Jones <>, David Karoly <>, D�ith� Stone <>, "Stott, Peter" <>, Francis Zwiers <>, Toru Nozawa <>, Ken Sperber <>, Dave Bader <>, Mike MacCracken <>,, Stephen Klein <>, "A. Pier Siebesma" <>, William Rossow <>, Chris Bretherton <>, George Tselioudis <>, Mark Webb <>, Sandrine Bony <>, James Hack <>, Martin Miller <>, Ken Kunkel <>, Christian Jakob <>, Kathy Hibbard <>, "Eyring, Veronika" <>,,,,,, Peter Cox <>, Pierre Friedlingstein <>, Olivier Boucher <>, Bala Govindasamy <>, Jonathan Gregory <>, Chris Jones <>, "Jones, Gareth S" <>, David Lobell <>, peter gleckler <>, Cath Senior <>, Keith Williams <>, "stephen e. schwartz" <>, David Easterling <>, Inez Fung <>, Duane Waliser <>, William Collins <>, Ken Caldeira <>, Dave Randall <>, Joyce Penner <Penner@UMICH.EDU>, Anna Pirani <>, Bjorn Stevens <>, "V. Ramaswamy" <>,, Bin Wang <>, Tianjun Zhou <>, Thomas Stocker <>

Content-Type: text/plain; charset=windows-1252; format=flowed
X-MIME-Autoconverted: from 8bit to quoted-printable by id m8JB2EeK011729

Dear all,

As you know, the experiment design for CMIP5 has been under review.
Thanks to those of you who have commented on the draft of July 21?,
2008. In response to input from individuals and groups, I have modified
the document except when a suggestion seemed controversial or had
previously been ruled out by consensus. The revised experiment design
summary is attached (Taylor_CMIP5_expts7.pdf), and below is a brief
summary of the major changes made and the major issues raised in
response to the earlier draft. These issues will need to be discussed at
the WGCM meeting next week.

I apologize for not having time to revise the document earlier. I also
have failed to fully synthesize a clear consensus for the experiment
design. After the WGCM addresses the major outstanding issues, I will
revise the document once more and try to incorporate any input I have
already received that is consistent with the agreed upon design.

I have attached three other documents prepared by interested groups:

1. A document (OzoneDatabase_InputCMIP5_V2.pdf) from SPARC CCMVal
describing their plan to �provide stratospheric ozone boundary
conditions suitable for long-term global climate model simulations such
as the � CMIP5 experiments 1.2 and 2.1-2.4.�

2. A summary (comments_IDAG.pdf) from the �International Detection &
Attribution Group,� providing a perspective on a possible shift of
priorities for some of the CMIP5 experiments, and contributing text that
provides further motivation for performing the detection and attribution
experiments summarized in Table 6.

3. A letter (CFMIP2_WGCM_Sept2008_summary.pdf) from the CFMIP
coordination committee (with the endorsement of WGNE and of the GEWEX
Scientific Steering Group) suggesting that the CMIP5 experiment design
should include a few additional (short) experiments and output useful
for diagnosing cloud responses in models.

Important changes made to the CMIP5 experiment design document include
the following:

1. The document now reflects the view that the near-term and long-term
experiments are equally important, but modeling groups are not required
to contribute results for both types of simulations. In the earlier
draft, the fact that none of the near-term experiments were designated
as mandatory made it appear that those experiments were not of as much

2. Several changes were made to Table 2, which summarizes the future
projection runs with prescribed concentrations of long-lived greenhouse
gases according to various representative concentration pathways
(RCP�s), including the following:

- the priorities of RCP2.7 and RCP4.5 were switched, with RCP4.5 now
considered mandatory (but not RCP2.7). This change was made because
RCP4.5 will be imposed in expt. 3.3, the emission-driven coupled carbon
climate model run (instead of RCP8.5, which was so designated in the
previous draft). RCP4.5 will now also be specified in the near-term

- likewise, the highest priority for extending an RCP run beyond the
21^st century has been switched to the RCP4.5 simulation.

- The first priority for each RCP will be to simulate through the end of
the 21^st century, rather than to year 2150 as previously specified.

3. The possibility that the WGCM will decide that the analysis of carbon
climate feedbacks should be done for a realistic RCP, rather than for
the idealized 1%/yr CO2 increase, was allowed for by including expt.
3.4. In this case expts. 4.2a,b would be eliminated.

4. The priority of one run of Expt. 4.2 has been downgraded to �very
high�, and Expt. 4.3c, which would have allowed us to analyze �fast�
responses of the carbon cycle to changes in atmospheric concentrations,
was eliminated.

5. Following the advice of the detection and attribution community, the
priority of expt. 6.2 has been increased to �very high� and the
simulation has been switched to GHG-only from anthro-only. The highest
priority ensemble simulation (expt. 6.4) has been switched to
natural-only from GHG-only.

6. Sections 7 and 8 were switched, so now all the long-term expts. are
described before the near-term experiments.

7. Section 9 (additional remarks) was eliminated.

There are a number of issues that must be resolved.


1. Should additional experiments be performed that aim to isolate in
each model the effects of aerosols. In particular should special
diagnostic runs be performed to calculate aerosol forcing, just as
special CO2-only runs are a part of the experiment design?
2. Should the strength of carbon/climate feedbacks be evaluated in
the historical and RCP4.5 simulations or in the idealized 1%/yr
3. Should the priority of the detection and attribution experiments
be higher?
4. Should the CFMIP recommendation, which calls for some additional
experiments to be performed that would improve understanding of
the range of cloud feedbacks in models, be accepted? It can be
argued that the consequences of uncertainty in cloud feedbacks are
at least as important as the consequences of uncertainty in
carbon/climate feedbacks, which receive considerable attention
under the current experiment design.
5. Is it important to isolate the impact of land use changes on
historical and future climate change? The current experiment
design will not provide this information.
6. The ending date for the historical runs and the start date for
future runs need to be agreed upon. It has been proposed to extend
the historical runs to very near present (say 2009), but start the
projection runs earlier (say 2005). Whatever is decided, the
historical concentrations (or emissions) should smoothly
transition into the future.
7. Is it absolutely necessary for models with a carbon cycle
component to run two control simulations, one with prescribed
pre-industrial concentrations and one fully coupled and free to

I hope those of you who will be at the WGCM meeting next week will have
a moment to think about these issues, and I look forward to the discussion.

With best regards,

Karl E. Taylor


Attachment Converted: "c:\eudora\attach\Taylor_CMIP5_expts7.pdf"

Attachment Converted: "c:\eudora\attach\OzoneDatabase_InputCMIP5_V2.pdf"

Attachment Converted: "c:\eudora\attach\CFMIP2_WGCM_Sept2008_summary.pdf"

Attachment Converted: "c:\eudora\attach\comments_IDAG.pdf"

Attachment Converted: "c:\eudora\attach\preWGCM_letter.pdf"

No comments:

Post a Comment