|
CCSM Software Engineering Working Group Report |
||||||
| Bill Collins discussed the CCSM3 release
schedule, current CCSM3 control runs, and the special issue of the Journal
of Climate. Tony Craig summarized the status of software engineering in CCSM. He briefly discussed several topics including the June 2 CCSM3 release, the status of CCSM3 porting, new CCSM3 scripts, new CCSM3 test capabilities, the status of CCSM3 on the Earth Simulator, and long term priorities. He declared success with regard to the vectorization, although it is not quite complete. Priorities for the next 6 to 24 months include implementing chemistry and biogeochemistry in the coupled model, ESMF, the CCSM/CAM unification process and moving to single executable, continued refactoring of all models, and continued improvement in CCSM SE infrastructure. Rob Jacob presented the overview of SciDAC progress. He briefly summarized the SciDAC CCSM vectorization activities, and then continued with descriptions of SciDAC development activities for each CCSM component. CAM activities included participation in the Performance Tiger Team, performance analyses on the Cray and IBM systems, and performance optimizations in the physics, Eulerian dynamical core, and the finite volume dynamical core. A status report on the implementation of the subgrid orography scheme in CAM followed. Ocean development activities focused on POP, HYPOP, and ocean biogeochemistry. The CLM2 restructuring was briefly summarized. Coupler activities included both performance and software engineering improvements to MCT, while MPH activities included redesigned support for both single and multiple executable versions of CCSM and runtime invocation of data model, dead-model, and real-model options in CCSM. Cecelia Deluca and Erik Kluzek presented a status update from ESMF. Cecelia indicated that ESMF was interacting with the DoD and NASA to get continued support for the project. She indicated that work was progressing well. Erik summarized his work on CAM related to subroutinization and the ESMF interoperability milestones. The CAM/SSI interoperability experiment is complete and the CAM/MIT interoperability experiment is under development. Over the next year, CCSM efforts will focus on prototyping ESMF as coupler superstructure and infrastructure, subroutinization of component models, and incorporation of high-level redesign of CAM into the CAM model. Tony Craig led a discussion of CCSM/CAM unification. The goal was to have an open discussion about some of the goals, needs, and desires as the unification project planning begins. Many good ideas were generated. Notes from the discussion are available at: http://www.cesm.ucar.edu/working_groups/Software/presentations/040211.various/CCSM_uni_plan.txt In the afternoon, we had a discussion of design process. Several people summarized their experiences. In the end, the SEWG came out with several recommendations about processes to try. |
||||||
|
||||||
|
||||||
Community Climate System Model
http://www.cesm.ucar.edu