Minutes of the CCSM Advisory Board, Scientific Steering Committee,
and CCSM Working Group Co-Chairs
Thursday, 8 July 2004
Eldorado Hotel, Sunset Room

CAB Attendees: Ed Sarachik (Chair), Isaac Held, Max Suarez, Eric Sundquist, John Drake, Michele Rienecker, and Tim Killeen

SSC Attendees: Bill Collins, Chris Bretherton, Bill Large, Scott Doney, Gordon Bonan, Jim Hurrell, Jay Fein, Maurice Blackmon, Tom Henderson, Danny McKenna, Dave Bader, Phil Merilees, and Lydia Shiver

Working Group Co-Chair Attendees: Matthew Huber, Steve Jay, Clara Deser, Bette Otto-Bliesner, Pat Worley, Cecelia DeLuca, Phil Rasch, Natalie Mahowald, Jim Randerson, and Inez Fung

Guests: Don Anderson, NASA; Tom Bettge, NCAR SCD; and George Carr, CCSM SE Group

1.Welcome. Sarachik welcomed everyone and asked all attendees to introduce themselves. He stated that the CAB advises Killeen, Anthes, Collins, and Fein regarding the CCSM project. Sarachik is rotating off the CAB and the new CAB chair is Daniel Jacob.

2.Discussion of Annual CAB Letter. The CAB discussed its comments to be included in the annual letter, stating that a tremendous amount of work has been done, including revectorization of the code so that it will run on the Earth Simulator and the IPCC runs that are currently under way. Sarachik suggested that the tropical biases are the most important to be corrected and solved.

The CAB stated that the CCSM development and release cycles are unsustainable without additional interaction with the broader community. Sundquist suggested that more positive steps for collaborations between the science community and the assessments community might be helpful. Collins also stated that he thinks the CCSM project has been conservative in its research, and it needs to change that and look for interesting science issues to explore. Drake suggested a greater university connection to help entrain new scientists and support staff. Collins suggested that the working groups could be restructured, that more joint working group meetings could be held, and the annual workshop could be restructured to focus on specific science topics and biases. Fung suggested tiger teams as an approach to solving individual biases, and Large agreed with the suggestion but said the reporting responsibilities would need to be avoided.

Killeen stated that the CAB should make very clear suggestions to UCAR/NCAR management on how to redirect the staff and funding of the CCSM project.

Fung suggested that a graduate fellowship program could help with burnout and also could direct young people to work on special topics, such as solving biases.

The CAB recommended that the SSC state its highest priority be to significantly reduce the biases in the model over the next 3 years.

The CAB requested details on what has been accomplished from the CCSM Business Plan, what resources appeared, what resources did not appear, and what science cannot be done due to lack of resources.

The format of next year's annual workshop was discussed. Some suggestions were:
-have additional meetings on focused topics
-use plenary session for focused topics
-use plenary presentation time for other things
-take advantage of large attendance to discuss coupled system issues
-use poster session more to show individual work and set aside time for formal
poster session
-all working groups have short business meeting at the same time, and the rest
of the workshop be spent on science discussions on the coupled system
-use workshop for more coordination of coupled model

3.State of Computer Resources for IPCC Runs. Collins reported that there are three phases for IPCC FAR: pre-industrial (1870), 20th century (1870-2000), and emissions scenarios, along with the SRES scenarios. He estimated that CCSM would use 7 million CPU hours at T85x1 for all the IPCC scenarios. He also reported that a DOE climate end station is under way as an additional resource, and Washington, Drake, SciDAC staff, Collins, and Hack are involved.

4.Update on Earth Simulator. Fein requested an update on the status of the MOU and the collaboration with CRIEPI.

5.ESMF and CCSM. Collins reported that there is an active effort starting for the stand-alone CAM with ESMF and with ice and ocean models. The benefits will be a single method of communication, infrastructure to create a unified CCSM (all use coupler and single executable), interoperability with assimilation packages, and methodology for hierarchical modeling (nested grids using assimilation and nested process models). The SEWG is writing acceptance criteria for CCSM using ESMF.