[an error occurred while processing this directive] [an error occurred while processing this directive]


Minutes of the CCSM Advisory Committee Meeting 05 November 2003

CCSM Advisory Committee Meeting

5 and 6 November 2003

NSF, Arlington, VA

Attendees: Ed Sarachik (Chair), Michael Ghil, Isaac Held, Tony Hollingsworth, John Drake, Max Suarez, Eric Sundquist, David Griggs, Michele Rienecker, Tom Crowley, Francis Zwiers, Steve Pacala, Rick Anthes, Tim Killeen, Maurice Blackmon, Jay Fein, Steve Reid, Dave Bader, Jeff Amthor, Bill Collins, Phil Merilees, and Lydia Shiver

Guests: Scott Doney, Bette Otto-Bliesner, Cecilia Bitz, Rick Smith, Jim Hack, Ants Leetmaa, Jim Mahoney, Steve Lord, Steve Meacham, Margaret Leinin, Tsengdar Lee, Anjuli Bamzai, Don Johnson, David Halpern, David Legler, Eric Itsweir, Pam Stephens, Neil Swanberg, Jerry Elwood, Dave Vernado, Tom Spence, Lydia Gates, Peter Milne, Peter Backlund, Rita Colwell, and Ahsha Tribble

1. Welcome. Sarachik welcomed Bill Collins, new Chair of the SSC, David Griggs, Hadley Centre, and Jeff Amthor, new DOE Program Manager. The Terms of Reference were read, and everyone introduced themselves.

Anthes welcomed everyone and explained that a special award was being given to Francis Bretherton, who was unable to attend due to weather in Madison, as his term was finished on the CAB. He requested everyone sign Francis' letter, and Rita Colwell stopped by the meeting to present the award to Francis, in abstentia, and a picture of the group was taken (http://www.cesm.ucar.edu/management/CAB/).

Killeen welcomed everyone and stated that there was a lot going on, that CCSM was close to being frozen for IPCC, new computer nodes were being added to support IPCC at NCAR, Bill Collins was appointed new SSC Chair, vectorization of CCSM was being done so that it could be run on the Earth Simulator, and ESMF was going forward and using CCSM as a testbed.

Fein welcomed everyone to NSF, and he stated that the CAB's advice was very helpful to him and NSF and NSF appreciated the CAB members' time and efforts.

2. CAB Membership. Sarachik will rotate off as Chair of the CAB at the end of December 2004. It was decided that CAB membership would be discussed in the Executive Session. Also, nominations for the next CAB Chair were requested and should be sent to Anthes, Killeen, and Sarachik.

3. SSC Membership. Collins discussed the membership of the SSC and recommended the addition of Gordon Bonan and Ben Santer, because the land and climate change areas would be very high priority in the short- and long-term. Collins reported that Ping Chang, Texas A&M, was rotating off the SSC and that Jim Carton (University of Maryland) would replace him. Collins stated that the institutional balance would be maintained.

4. Annual CCSM Workshop. The 9th annual CCSM workshop will be held in July 2004, IPCC results will be shown, and future directions for the CCSM project will be discussed. The theme will be biogeochemistry. David Archer, Inez Fung, and Colin Prentice are confirmed plenary speakers, and Ralph Keeling has been invited to speak. Topics at the plenary sessions and working group meetings will be the release of CCSM3, preliminary results from IPCC, experiments with the new model (climate variability studies, paleoclimate studies), and progress on the coupled chemistry and climate component, including biogeochemistry, aerosols, interactive tropical chemistry, and coupled carbon cycle experiments. There will be a SSC meeting on Tuesday, and a combined CAB, SSC, and Working Group Co-chair meeting on Friday afternoon.

The workshop will be held at the Eldorado Hotel in Santa Fe, New Mexico. Santa Fe is at a lower altitude than Breckenridge and an historical setting. The hotel is located near the Plaza and the Santa Fe Opera season will be held during this time also. CCSM agreed to move the dates of the annual workshop so that CLIVAR could hold its first international meeting during the third week of June.

Sarachik stated he wanted to have a CAB meeting on Thursday evening and invited the SSC and working group co-chairs to attend that meeting also.

5. Status of IPCC Version of CCSM. Collins reported that the next release of CCSM would be configured using CAM 3.0, Land CLM2.2, POP 1.4.3, CSIM4, and coupler 6. CCSM will have multiple dynamical core options (Eulerian and finite volume), and a T42 and T85 version would be available. We also hope to have a T31 version. Collins explained that the Eulerian dycore only would be used for IPCC runs because the finite-volume dycore would not be ready in time. Suarez stated that both dycore versions must be well documented, and Griggs stated that the T42 and T85 versions should be labeled differently rather than just as CCSM.

The improvements in CCSM3 are the improved winter land surface temperatures, tropical tropopause temperatures, cloud forcing responsiveness to tropical SSTs, east Pacific surface energy budget, double ITCZ, and tropical variability. The atmosphere model has dramatic changes in the physics, the land model is new, changes to the ocean model have been made in the KPP boundary layer and solar absorption (chlorophyll), and a new coupler will be ready.

The release of CCSM3 will include the new model physics; model data sets and control runs for T31, T42, and T85, and all technical documentation on the web. The data sets will be mirrored at PCMDI also. Another Journal of Climate special issue is being discussed with the SSC and Dave Randall. CCSM3 will be run on the Earth Simulator, hopefully starting in December 2003. An IPCC computer run schedule was shown also.

In summary, Collins stated that near-future issues, long-term issues, and new scientists and computing resources will need to be discussed and decided (Collins' IPCC presentation).

Sarachik summarized the CAB's recommendation that almost immediate and complete release of data sets would be best. Collins agreed to check on if regional data sets would be released. Killeen stated that NCAR is working with CRIEPI/MEXT on use of the Earth Simulator, and discussions with NSF are under way.

6. CCSM Implementation Plan. Merilees distributed a draft Implementation Plan. In Phase I (Sep. 03-Aug 04), a need for increased funding was relevant but it was not included in the budget plan. CCSM has added an associate scientist, funded by Killeen, to work on the atmosphere model component, DOE has funded a new associate scientist for two years to work with the Climate Change Working Group on IPCC, Killeen and NSF are funding a new software engineer to work on vectorization and flexibility of the model, an associate scientist has been reassigned within CGD to work with the Biogeochemistry Working Group, and a new associate scientist is being sought to work with the Paleoclimate Working Group. New scientists are being sought using the Scientist I and ASP programs (Merilees' future plan and presentation).

Fein stated that Margaret Leinen would discuss the budget outlook with the CAB the next day, and he stated that it is highly unlikely that new funds would be found. Any new funds for CCSM would be done by reprogramming. Killeen stated that NCAR has purchased additional computing resources using its equipment fund, and he has agreed to hire a new oceanographer and has already hired a new scientist I for paleoclimate work (Caspar Ammann).

Held stated that the need for more scientists to work on core atmosphere/ocean science research should be stressed because the modeling efforts are very dependent on those core sciences.

7. Computing Needs. Sarachik chaired this session to discuss the past, current, and future computing needs of CCSM and climate modeling. OSTP has a high-end computing revitalization task force, and it should be monitored. Hollingsworth, ECMWF, stated that they are clear of vector based machines and that they are using scalar machines. It was stated that sensible software engineering coding would make vector and scalar machines usable across platforms. Collins stated that CCSM will maximize the flexibility and portability of CCSM.

8. Update from GFDL. Leetmaa discussed GFDL's ongoing FMS model development, decisions about which code will be used for IPCC will be done by mid-November, and IPCC runs will be started in December. Future GFDL plans include adding the terrestrial carbon cycle, oceanic ecosystem modeling, coupled terrestrial and ocean carbon cycle, and fully interactive atmospheric chemistry (Leetmaa's GFDL presentation).

Several CLIVAR Climate Process Teams (CPTs) include collaborations between NCAR, GFDL, and NASA, and will closely link field programs to modeling efforts.

GFDL and NCAR have a collaborative agreement and ongoing collaborations to identify and overcome common shortcomings of both of their models, identify and explain major differences focusing on climate sensitivity, and establish common standards for model diagnostics and modeling infrastructure with DOE and NASA.

Update from GFDL/NCAR Collaboration. Held reported that the most recent focus of the collaboration has been on the atmosphere models and that the ocean modeling group is beginning to form. Both groups decide their paths differently and use different metrics. The collaboration project includes looking at global climate sensitivity and cloud feedbacks, regional climate responses, similarities and differences in both of the model's biases, and sharing diagnostic tools and data, which has become very valuable (Held's GFDL/NCAR presentation).

Update from GSFC. Rienecker stated that the main goals of the program at Goddard are to use models and assimilation tools to advance the understanding of climate variability and change, support science goals, maximize the impact of satellite observations, transition satellite assimilation capabilities to operations, contribute to the science of observational systems, advance chemistry constituents/aerosol modeling and assimilation, and advance subseasonal-to-decadal climate prediction, which is the ultimate test of models used for climate change.

GSFC is broadening the Data Assimilation Office (DAO) to include a single modeling system with a finite-volume dynamical core and evolving physics, modeling for weather and climate to improve moist processes, how to use cloud resolving models in the next generation model, and developing and validating in collaboration with others and the general community including CPTs and GMAO Science Team. The GMAO interaction is in collaborations with CCSM and GFDL on the CPTs in atmospheric physics, the ESMF partnership, and finite-volume dynamical core research. They are also involved in the data assimilation area in collaboration with WACCM on chemistry-climate coupling and chemistry assimilation (Rienecker's GMAO presentation).

9. CCSP. Jim Mahoney joined the meeting to discuss CCSP. He stated that the CCSP Strategic Plan was on the web (www.climatescience.gov) and would evolve as plans evolve and that everyone should check for periodic updates on the web. The CCSP Synthesis and Assessment Reports will be government documents, which mean that they will be CCSP documents and part of the CCSP program. He stated that the process needs to be transparent and that the proposed report guidelines would be published on the web soon.

He outlined the Lautenbacher Initiative as a basic approach to develop synthesis and assessment documents that will begin with an outline, avoid duplication of work already done, and be about shorter term predictability in that seasonal science topics are chosen and a meeting or workshop is held to discuss lessons learned, critical problems, and what works well and what does not work well. He stressed that regional climate modeling is very important for future funding, and regarding the IPCC scenarios, they will be used to do comparisons to new models, but CCSP will do comparisons and then make changes to see the results of the changes. Mahoney stated that there is no point to modeling if time is not taken to analyze and validate and ask science questions.

Mahoney agreed to talk to Kathy Olson, OSTP, about computing needs for climate modeling.

Killeen and Leetmaa will try to translate CCSP research requirements and distribute them.

10. Additional Updates-NCEP. Steve Lord gave a presentation about the NCEP program and stated that NCEP has a new global coupled atmosphere/ocean model with no tuning that has better SST skill. They also are working on the NCEP Global Ocean Data Assimilation System (GODAS) that will be a real-time global ocean database (Lord's NCEP presentation).

ESMF. The Earth System Modeling Framework project is a three-year project at NCAR, funded by NASA, with many collaborators that has developed applications that will need to be incorporated into codes and data assimilation systems to make it very easy to work across any computing platform. The project will end in Spring 2005 and will have a product. A white paper on where ESMF should go after the three year project is complete has been written and is available from Killeen. The white paper states that plans need to be made to continue ESMF, or at the very least to maintain it, and the software engineering needs to continue to evolve, more broad based funding is needed, and a grander vision is hoped for rather than only maintaining ESMF. The white paper will be presented to NASA management soon and funds will be requested so that NASA is involved in the continuation of ESMF. ESMF is also collaborating with PRISM, which is the European project similar to ESMF. To obtain further information about ESMF, please go to www.esmf.ucar.edu.

Collins stated that CCSM has not developed a plan to work with ESMF and that the SSC would be discussing this soon.

Cyberinfrastructure. Steve Meacham discussed this NSF project that will advance scientific research and education, will speed up new technology into science, will handle more complex systems and problems, and provide more data and more diverse types of data. He stated that NCSA in Pittsburgh could make time available for a large project such as CCSM, and CISE and GEO are available to help facilitate this. Killeen requested help with data storage and archiving due to the large data volume of IPCC scenarios and Earth Simulator experiments, and Meacham recommended that we contact the PACI centers. He encouraged NCAR to work with other NSF supercomputing centers (Meacham's cyberinfrastructure presentation).

NSF Update. Margaret Leinen joined the meeting, and she welcomed the CAB members and thanked them because the CAB's role is so important for community input to issues. She discussed the FY04 budget, which is in the continuing resolution stage now. There is $1.3 billion in the VA, HUD, and Independent Agencies bill that is still undecided. She stressed that if GEO gets an increase the highest priority will be in increasing staff and salaries in GEO. She encouraged CCSM to leverage every capability for high-end computing resources from PACI and other supercomputing centers and NCAR. Margaret asked that we put together a computing needs statement as early as possible to help NSF find funds to do "compelling science," which is laid out clearly and is interdisciplinary in nature.

DOE Update. Jeff Amthor said that a new position would be opened soon at DOE for a manager of climate modeling, that DOE is funding a CCSM Climate Change Working Group liaison position at NCAR and a project at Colorado State University, and will continue funding for SciDAC. He stated that biogeochemistry and atmospheric chemistry would be added to the SciDAC project and who would do what and what the milestones would be are being discussed. He also asked specifically how DOE could help the CCSM project through the SciDAC project. Another DOE focus is the aerosol-climate connection. He also stated that DOE would support ESMF (Amthor's DOE presentation).

Blackmon thanked DOE for their funding and collaborations at Argonne, Oak Ridge, Los Alamos, etc.

Drake stated that DOE's computing resources are ready to help with IPCC at NERSC, Berkeley, and with the Cray X1 when it comes online.

NASA Update. T. Lee updated the CAB on how NASA will support the new GMAO at Goddard that will make the best use of satellite measurements to improve prediction capabilities. NASA strongly supports the ESMF project and its continuation.

NOAA Update. Anjuli Bamzai discussed the Applied Research Centers (ARCs), which will take other's models and customize them for seasonal-to-interannual research. The ARCs are located at CDC, FSU, COLA, University of Washington CSES, and ECPC at Scripps (Bamzai's CDEP presentation).

11. CCSP Synthesis and Assessment Reports. Amthor stated that by October 2005 that this will be parallel with IPCC. The general principles are that the 21 products are comprehensive but not necessarily big, should include a summary of science conclusions that are explicit references to confidence levels, and be prepared in an open and transparent fashion (Amthor's CCSP presentation).

12. CCSM Working Group Future Research/Development Directions. Hack summarized the Atmosphere Model Working Group future directions as understanding, improving, and evaluating key processes in the physical climate system; attacking the problems; gaining acceptance for incremental improvements; focusing on developing a better understanding of similar performance, and identifying and filling holes in the research. The AMWG hopes to extend its capabilities in aerosol modeling, atmospheric chemistry, offline transport modeling capabilities by replacing MOZART and MATCH with a CAM transport version, and including WACCM and biogeochemistry (Hack's CAM presentation).

Rick Smith, LANL, summarized the Ocean Model Working Group future directions as completing the ocean component configuration for the IPCC runs, continuing the CPTs collaborations, and having a working group meeting in January. After IPCC the OMWG will begin using POP2.0, which contains new processes for partial bottom cells, bottom boundary layer, tripole grid, enhancement of the Gent-McWilliams scheme to anisotropic G-M, vertical mixing over rough topography, tracer advection scheme, and natural freshwater flux boundary conditions. They also are planning to add a x3 coupled model CCSM version, an upper ocean model, embedded coastal models, data assimilation, and biogeochemistry and ecosystem models. Plans in the science area are to address the biases in the coupled model, integrate ocean-only and ocean-ice integrations, and coupled ice. This group also wants to pursue high-resolution ocean modeling (Smith's OMWG presentation).

Cecilia Bitz, University of Washington, summarized the Polar Climate Working Group future direction as working toward understanding the model. She stated that help is needed in obtaining funding for polar climate research by changing the perspective of NSF Program Managers; reviewers are looking for models that look like data and not for use for science (Bitz's presentation).

Scott Doney, WHOI, summarized the Biogeochemistry Working Group future directions as using the CCSM1 with carbon and interactive land for IPCC, documenting CCSM2 with carbon, and developing CCSM3 with carbon by porting CASA/OCMIP to CCSM3 physics and adding Thornton's new land model and a more advanced ocean model with marine ecosystem dynamics. This also will be submitted to IPCC. After IPCC the group plans to develop a CCSM3 with carbon plus model, including dust, land-atmosphere coupling and active chemistry, and isotopes (Doney's BGCWG presentation).

Bette Otto-Bliesner summarized the Paleoclimate Working Group future directions as trying to understand the climate system, its sensitivities, and stability by taking advantage of the new features that have been added to CCSM3. She stated that new components in the near future are carbon cycle, isotopes, and atmospheric chemistry. In the future the group will try to resolve ice sheet and topography LGM at high resolution (Otto-Bliesner's Paleo-WG presentation).

The meeting was adjourned and the CAB Executive Session began.

The CAB discussed the CCSM Data Policy and its content regarding when data sets would be publicly released and agreed that the policy would be adhered to. Shiver will send the CCSM Data Policy URL to the CAB members.

The CAB discussed the need for short, concise statements of priorities for all climate modeling (CCSM, DOE, NASA, NOAA) that can be given to Halpern, OSTP, Marburger, etc., and used to gain funding. Killeen said that Leinen asked for an advocacy document or approach for the next generation of climate modelers, "sustainable" computer resources, Earth Simulator-type plans, and plans beyond IPCC. Sarachik called for a conference call to continue this discussion.