Minutes of the Joint Meeting of the
CCSM Scientific Steering Committee
CCSM Advisory Board
Friday, 29 June 2001
The Village at Breckenridge
Attendees: Jeff Kiehl (Chair, SSC), Ed Sarachik (Chair, CAB), Eric Sundquist, Max Suarez, Isaac Held, Susan Solomon, Tim Killeen, David Schimel, Rick Anthes, Francis Bretherton, Peter Gent, Danny McKenna, David Randall, Maurice Blackmon, Ping Chang, Dick Moritz, Bob Dickinson, Jay Fein, Anjuli Bamzai, Dave Bader, Lydia Harper
1. Welcome. Anthes welcomed everyone and then he read a letter from Killeen, Fein, and himself thanking Blackmon for his service and leadership to the CCSM project.
2. "Climate" in Washington." Fein discussed the Bush speech of 11 June 2001 with directives to the Secretary of the Department of Commerce to define the role of coordination between federal agencies in climate research.
Fein informed the SSC and CAB that Anjuli Bamzai would be leaving NSF at the end of the year, and the SSC and CAB members thanked her for her efforts on behalf of the CCSM project.
Killeen discussed the upgrading of software engineering capabilities at NCAR, the hiring of 11 new scientists, and the draft NCAR Strategic Plan that includes many initiatives that CCSM can participate in. Killeen encouraged all SSC and CAB members to review the NCAR Strategic Plan and send their comments to him. To obtain a copy of the Plan, please send an e-mail to Lydia Harper at firstname.lastname@example.org. He also said that the computer upgrade was ongoing but no definite details could be discussed.
Blackmon thanked Fein for all his support for the CCSM project that has made a huge difference in many areas, such as the hiring of Mark Stevens and Sylvia Murphy to develop diagnostics tools and the funds (with partial DOE support) to hire a Manager of Software Engineering (Tony Craig).
Blackmon also thanked the collaborative efforts of staff at PCMDI for their dedication and objective analysis of several model configurations to achieve CCSM2.
3. Report on Organization of CCSM. Kiehl reported that he accepted the position of Chair of the SSC, that he will hire a coordinator, and that Tony Craig will report to him as Manager of the Software Engineering Group within CGD. The CCSM Program Office will include Kiehl, the new coordinator, Harper as administrator, and Craig as the SE Manager. Kiehl will work closely with the NCAR co-chairs of the Atmosphere Model Working Group (Hack) and the Ocean Model Working Group (Gent) to direct the software engineering group. Kiehl stated that his personal goals as Chair of the SSC are improving planning of CCSM development, improving communications with the community, exploring boundary issues, converting the CCSM Plan (2001-2005) into a "Living CCSM Plan," and starting outreach activities. The "Living CCSM Plan" will be a web-based document that defines the CCSM development process, defines the current model structure, defines the model release policy, defines the near-term model evolution, defines the long-term model evolution, and provides material for outreach activities. The outreach activities will include establishing greater visibility in national and international arenas, providing presentation materials for members of the SSC, CAB, and NSF, exploring the establishment of graduate and postdoctoral CCSM programs, and exploring the establishment of topical workshops. Kiehl also stated that he hopes to have a better defined model development process, have flexible diagnostic capabilities, and focus the CCSM project around problem areas, e.g., double ITCZ, etc. Some of the software engineering issues are improving model efficiency, maintaining and improving model portability, maintaining online status of model components, and defining linkages to other programs, such as DOE's Avant Garde, and NASA's CAN. Some future CCSM activities are to move to a more coherent support structure; better define links to other programs; address persistent biases; hold topical workshops; define new components (e.g., glacial processes, atmospheric chemistry, etc.); generalize diagnostic web pages for the community; and establish links to other development activities in the U.S.
Schimel suggested that Kiehl involve McKenna in the software engineering decisions, along with Hack and Gent, or maybe Rasch instead of McKenna for atmospheric transport involving biogeochemical processes.
Fein reiterated to the SSC and CAB members that proposals for Small Grants for Exploratory Research (SGERs) can come to him via individual PIs and from the SSC members.
4. Comments on the Workshop and CCSM2 and CAB Perspective. Sarachik said he was impressed with the diagnostics web pages and that a web page that lists problems would be helpful. Held said that topical workshops would be very beneficial, and that future workshops would be the ideal opportunity to discuss scientific hypotheses. Fein suggested science teams be developed that would focus on high priority issues that plague climate models. Fein suggested a meeting in Washington, DC, of the CAB and U.S. CLIVAR SSC members and others to have a dialog on modeling issues. Bretherton stated that the invited plenary speakers brought another group's perspective on issues to the workshop participants and that the plenary speakers should be continued. Randall suggested implementing tutorials surrounding the model, and Kiehl stated that the SSC and working group co-chairs discussed this earlier in the week. Kiehl will send an e-mail questionnaire to professors that teach modeling to gain information about what they think their students need in the way of tutorials. Schimel suggested plenary speakers on software engineering issues and strategies next year. Moritz suggested that more science/hypotheses discussion should be held in the working group meetings.
At the next SSC meeting, the design of next year's workshop will be discussed. Suggestions were that working groups have joint meetings if appropriate, results of the coupled model be shown and discussed, discuss development of the next generation model, plenary talks be shorter with ten minutes for speakers to discuss specific ideas in relation to CCSM, hold a formal poster session, develop new leadership by having postdocs have major roles that are visible, have an abstract volume, hold topical crosscutting sessions with less time for working group meetings, give working group co-chairs more direction on content and purpose of working group meetings, have time to discuss new ideas, end sessions later (6 p.m. instead of 5 p.m.), and continue the open forum session.
Randall suggested that the PCMDI analysis package be available at NCAR or on people's individual computers, Kiehl suggested expanding the diagnostics package for regional purposes. Suarez stated that the component model selection process should be more transparent to the community and information on the process better communicated. Kiehl and Randall pointed out that selection of the current atmospheric model was an open process that took place in two AMWG meetings over the winter months. Sundquist suggested we invite a plenary speaker to present information on hydrologic observations to help continue to bridge collaborations with observationalists.
The CAB members discussed the contents of their annual letter that will include advances made since last year, more information on selection process by each working group posted on the WWW, formulate a code policy, progress seen on management and science issues, and that the CAB will comment on lack of adequate computing resources and the unstable computing environment. Fein stated he could discuss inadequate computing resources with other USGCRP partners in the CSL. Kiehl recommended that CCSM have a dedicated computer, Sundquist suggested a formal agreement with DOE on computing, and Killeen requested information on CCSM computational needs in one year and three years. Kiehl stated that when the next NCAR computer architecture is known, more information could be given to Killeen about computational needs in the next 1 to 3 years.
The CAB will encourage the SSC to compile a list of development runs, production runs, and application runs that could be done and factor in any new components that will be developed, including the number of years for each simulation that needs to be run and what science value will be shown.
The CAB and U.S. CLIVAR SSC will meet in Washington, DC to have a dialog about CCSM and other U.S. modeling activities, discuss current generic problems, discuss need for national infrastructure that CCSM could address, etc.
Other suggestions from attendees were to start a coupled model working group, hold an NCAR Summer Colloquium on Coupled Modeling, and demonstrate the SCD Viz Lab at the next workshop.
Anthes proposed a draft schedule for the next CAB meeting. The proposed schedule was the night before the meeting begins have an informal dinner with CAB members and guests to discuss sensitive and confidential information and the climate in Washington. On the first half day, invite people from federal agencies, congressional staffers, and science advisors, and have people mostly not from NCAR give an overview of CCSM and CCSM related topics and discuss CCSM participation in national and international arenas. Next the CAB could have lunch together and discuss other business. The second half of the day would be the meeting with the U.S. CLIVAR SSC members. The second day would be the CAB Executive Session and outreach activities. Harper should let the SSC and Working Group Co-chairs know the dates of the next CAB meeting.