Agenda -- Quaas 11:59, 30 October 2007 (EDT)
Would we want a dedicated simulation for this? If so, the following could be recommended:
2x15 months starting in October 2005 (1x present-day, 1x pre-industrial aerosol emissions); submit output for Jan - Dec 2006
The pre-industrial run would be needed for the forcing estimates. If no nudging option exists, 5-year runs would be needed
If we don't use the Calipso/Cloudsat simulator, any year different from 2006 could be used equally well.
If several groups decide this can be done at a reasonable cost in both man-power and computing time, the advantage of carrying out separate simulations would be to potentially get some initial results quickly.
Wait for CICCS simulator?
The CFMIP ISCCP/Calipso/Cloudsat simulator is scheduled to become available as a beta-release in spring 2008. Should we opt to become test users and implement the beta-version prior to running any experiment?
Re: Wait for CICCS simulator? -- Quaas 12:48, 30 October 2007 (EDT)
Yves asked about the complexity of the simulator. As perhaps some clue on this, it took me a couple of days to implement and test the ISCCP simulator in ECHAM5, and the interface shouldn't be too different. The ISCCP simulator uses less than 2% of the model's CPU time. I think, however, that the risk is that the CICCS simulator won't be really mature in the beta release.