Progress of Interactions Among CCSM and Other Modeling Efforts

1
Progress of Interactions Among CCSM and Other
Modeling Efforts

• Bill Collins
• http//www.cgd.ucar.edu/wcollins
• National Center for Atmospheric Research
• Boulder, Colorado

• Earth System Modeling Framework (ESMF)
• National Centers for Env. Prediction (NCEP)
• GFDL and GSFC (other attendees)

2
Interactions with ESMF Project

• Clarification of CCSMs reasons for adopting
  ESMF
• Enable incorporation of assimilation capabilities
• Simplify construction of complex, hierarchical
  models
• Facilitate development of single-executable CCSM
• CCSM has adopted a thorough set of acceptance
  tests
• Test plan has been reviewed by SSC and WG
  co-chairs.
• The first phase starts on Jan. 1, 2005, ends in
  Sep. 2005.
• CCSM and SCD are seeking to support ESMF
• We submitted a NASA MAP proposal in Nov. 2004.
• The proposal would support dedicated ESMF SE in
  CSEG.

3
Acceptance Test Plan for ESMF

• Phases
• Test ESMF in stand-alone CAM CLM.
• Test ESMF in full CCSM.
• Rationale for two-phase approach
• All ESMF functionality needed for CCSM is not
  yet ready.
• Risk of resource commitment is minimized if we
  just test two components.
• Stand-alone CAM has a much larger external user
  community than CCSM.
• CAM is better instrumented for performance
  evaluation.

4
Components of the Phase 1 Plan

• Quantitative acceptance metrics for
• Computational performance
• Solution separation
• Memory requirements
• Qualitative survey metrics
• Evaluation of ESMF-enabled code by
• Atmospheric Model Working Group
• Land Model Working Group
• Software Engineering Working Group

5
Other Criteria for Phase 1 Acceptance

• Commitment to customer support
• Long-term software produce support
• Stability of software interfaces
• Support on CCSM production platforms

6
Implications of a Successful Phase 1

• ESMF will be incorporated into the CAM and CLM
  base code.
• CSEG will begin phase 2 acceptance tests.
• External CAM and CLM users will need to install
  ESMF.

7
Interactions with NCEP

• Bill Collins visited NCEP in May 2004 for
• Two-day briefing on NCEPs modeling activities
• Discussions of possible collaborations
• Results
• Agreement to use Don Johnsons tests of dynamics
• Agreement to explore climate model test-bed
  concept
• Follow-up activities
• Phil Rasch (AMWG co-chair) attended conference on
  isentropic coordinates organized by Johnson and
  Randall.
• Phil is working with Don on applying tests to FV
  dynamics.

8
Interactions with GSFC and GFDL

• GSFC comments from Michele Rienecker
• GFDL presentation by Isaac Held (215)