Minutes of the
CCSM Advisory Board Meeting
8 and 9 January 2002
NSF, Room 340
Attendees: Ed Sarachik (Chair), Francis Bretherton, Michael Ghil, Isaac Held, Tony Hollingsworth, John Drake, Susan Solomon, Max Suarez, Eric Sundquist, Michele Rienecker, Rick Anthes, Tim Killeen, Maurice Blackmon, Jay Fein, Jeff Kiehl, and Lydia Shiver
Guests: Margaret Leinen, Eric Itsweire, Steve Mecham, David Legler, Mike McCracken, Bob Weller, Ming Ji, Cecilia Bitz, Peter Gent, Jim Hack, Dave Bader, Ants Leetmaa, Eric Lindstrom, Louis Uccellini, Gene Rasmussen, Chris Bretherton, Paul Schopf, Peter Schultz, Vaughan Turekian, Tsengdar Lee, J. Shukla, John L. Hayes, Paul Anastas, Richard Moss, Jim Todd, and Ann Marie Schmoltner
1. Welcome and Introductions. Ed Sarachik welcomed everyone and thanked them for attending.
2. NSF Program Manager and CAB Chairman Remarks. Jay Fein discussed the role of the CCSM Advisory Board (CAB) as being an advisory panel for the Community Climate System Model (CCSM). It is unique in its involvement of all four major agencies: DOE, NASA, NOAA, and NSF. Ed Sarachik discussed CAB's role in facilitating the integration and cooperation between agencies and how the Washington, DC meeting of the CAB is designed so that many agency and program managers can attend.
Margaret Leinen also welcomed everyone and thanked them for all the effort to make CCSM a success. She discussed the future plans of the U.S. Global Change Research Program (USGCRP), NSF's long-range plans, and NSF's plans with the Office of Science Technology and Policy (OSTP) and the Office of Management and Budget (OMB).
3. CCSM Overview, Status, Progress. Jeff Kiehl, Chair of the CCSM Scientific Steering Committee (SSC), gave a presentation on the status of the CCSM. The model algorithms were frozen at the end of December, 2001. Control simulations are underway, and these, along with the model, will be released to the community in May, 2002. The SSC is planning to hold a topical workshop on tropical climate coupled models. The SSC is also collecting information about data requirements and computing plans from all the CCSM working groups. Plans for the next version of CCSM include higher atmosphere model resolution, interactive carbon cycle model, and elimination of tropical biases.
4. Agency Perspective on Climate Modeling and Related Activities, Needs, Issues
a. Dave Bader, DOE, reported that DOE's strategy is the same as it was during the last CAB meeting. The SciDac initiative was announced and $8 million in funding was competed for climate modeling with a focus on software engineering and computer issues. DOE has completed a review of the Los Alamos National Laboratory and they plan to review PCMDI. The new five-year DOE cooperative agreement with NCAR should be finalized in March.
b. Ants Leetmaa, NOAA/GFDL, discussed how GFDL fits into the NOAA climate services area. He stated that climate and climate observations need to be better organized. NOAA will be involved in operational climate services (seasonal-to-interannual to decadal), and he intends to have operational services that will be able to do scenarios of climate change. Leetmaa stated he is seeking to enhance collaborations with other institutions and that GFDL is revamping all their model codes to a flexible modeling system (FMS). GFDL will release its model framework, atmosphere model component, and ocean model component in the spring.
c. Jay Fein, NSF, reported that as a result of the merger of the CSM and PCM into CCSM-2, the CCSM-2 would be used to perform IPCC runs for assessments. The U.S. should have more than one model for this purpose, and it is hoped that NOAA/GFDL will provide a second model and that NOAA gets the resources and builds capacity to become the "operational" leg of the planned "two-center" approach. The research "center" would include CCSM and GFDL and its collaborators. Both would serve as pipelines to operations. Scientist-to-scientist interactions are the most productive way to bring that about. A change of attitude was begun with the community idea used to develop CCSM and that can be extended to more than one research center. This is critical so that GFDL's research is not the only pipeline to operations in the future.
d. Eric Lindstrom, NASA, stated that assimilation of satellite data sets into models is being done at NASA's Data Assimilation Office and joint NASA/NOAA centers, NASA's Goddard Space Flight Center is doing seasonal-to-interannual research, and a joint ocean modeling program is underway with NASA's JPL, Scripps Institution of Oceanography, and MIT. He said the NASA Earth System Modeling Framework project would help to develop climate models.
5. CAB Business
a. ECMWF Computer Acquisition. Sarachik requested that Tony Hollingsworth give an update on the recent computer acquisition at ECMWF. Hollingsworth reported that IBM was the chosen contractor and that the new hardware will be installed between October 2002 and 2007. He said the main production codes were used as benchmarks, additional hardware would be installed to meet performance commitments on each of the main production codes, and the main migration issues would be on the control, O/S, operations, and applications codes.
b. Hadley Centre Review. The Hadley Centre is block funded at $30 million per year and is used as a resource for government and policymakers. Their model development is becoming more modular, they make small improvements to a stable model, they have a deliverable of producing IPCC reports, and they have links with the community but they are not direct.
c. NASA Earth System Modeling Framework (ESMF) Project. This activity is a collaborative project between NCAR, MIT, NASA DAO, and GFDL. It includes the usage of CCSM and other models, such as WRF, in its testbed. It is an open system for models and will include software and technical aspects only; no parameters are included. A prototype should be ready in 18 months, and workshops will be held to demonstrate it.
d. NAS Report: Improving the Effectiveness of U.S. Climate Modeling. This study states the key issues and concerns are: the lack of adequate access to high-end computing by the climate modeling community; the scarcity of human resources applied to computational and scientific research problems; the difficulty of matching the financial rewards offered by private industry; the lack of appropriate software available to optimize performance on the new generations of massively parallel computers; the lack of software standards and protocols for building different climate models and the absence of uniform computer and observational data-archiving standards, which inhibits exchange of useful information; the need for uniform criteria with which to judge climate models; the need for widely available standard software tools to diagnose and compare climate model output; and the need for a strong interaction between observations of the climate system, research into fundamental climate processes, and integrative climate modeling.
6. CLIVAR Presentation. David Legler, Chris Bretherton, Bob Weller, and Paul Schopf attended the CAB meeting to discuss the notion of "climate process teams" (CPTs). A CPT would be a group of scientists, both modelers and process researchers, who would focus on addressing an important deficiency in climate models, e.g., convection, marine stratus, ocean mixing, etc. They asked the CAB for its opinion of the idea. The CAB's response was positive, and it agreed to write a letter to Anthes, Killeen, Fein, Blackmon, and Kiehl relating how the CPT project(s) might benefit the CCSM, as well as high-end modeling in general. They thought the idea had merit and potential for agency representatives to take to interagency groups to promote an announcement of opportunity. The CAB also encouraged Kiehl to discuss ideas with the SSC on collaborative research between the CCSM and CLIVAR participants.