-
Notifications
You must be signed in to change notification settings - Fork 64
CCPP Framework Meeting Minutes 2021 07 13
Dom Heinzeller edited this page Jul 15, 2021
·
2 revisions
- Transition to capgen.py
- Standardized way to read tabulated data in CCPP schemes?
- Update on standard names
- Status of capgen in MPAS:
- September 30 - Dave wants to run CCPP physics running in MPAS, this used to be June and slipped
- CCPP-compliant physics are B4B with existing physics
- Dave has a branch of MPAS that implements the ccpp-framework and uses it to call at least one physics scheme
- Simple suite, hardcoded to use CCPP (simple = several schemes and a subcycle, no fancy automation and conversations)
- In the end, MPAS needs to be able to run both CCPP and non-CCPP physics
- WRF will call CCPP-compliant physics directly w/o the framework
- Need to ask Dave about missing capabilities of the framework for use in MPAS
- Steve:
- No requests from MMM on additional framework capabilities, but plenty of work to do for CESM
- Cheryl built the musica box model a while back, which uses capgen
- feature/capgen branch had major updates on March 18, smaller updates Apr 10 and May 21
- Cheryl uses a branch in Steve's fork that was updated in April, too - thus, feature/capgen is up to date
- Steve currently working on DDT type issue (inconsistency in how we describe DDTs in the metadata)
- Next: actions (var_action class), which is required for Dom to work on unit conversions and for Steve to work on diagnostics; a lot of other development depends on the var_action capability, too
- UFS:
- Deliverable to transition to capgen at the end of last year ... slipped
- Additional deliverables that build on capgen
- Issues/Questions:
- There is no management to coordinate the work between NOAA end NCAR
- Is there a gap in funding and personal to get capgen working in the UFS with the required capabilities?
- Dom can start working on the code from end of July (after CIRES ASA season)
- Need to wait until Andrew shows up again
- What is left to do in order to merge
- Host model PRs:
- ccpp-physics: https://github.com/NCAR/ccpp-physics/pull/687 (will be merged with fv3atm PR)
- UFS/fv3atm: https://github.com/NOAA-EMC/fv3atm/pull/337 (will be added to commit queue for next 2-3 weeks)
- only few comments from Moorthi so far
- SCM: https://github.com/NCAR/ccpp-scm/pull/269 (can be merged anytime after ccpp-physics PR)
- Status of current PRs for ESCOMP:
- Cheryl is ok with the current names in the PRs, will look at the rules soon
- Grant: ready to go (caveat: this will need updates in the future, and it only covers 2/3 of the existing standard names)
- Status of related PRs for UFS/SCM:
- Little engagement from NOAA-EMC, therefore do not expect much pushback if NCAR requests more changes
- Moorthi's comments: Exner function ("dimensionless" is redundant), x/y vs zonal/meridional wind
- MPAS interpolates to zonal/meridional winds co-located with temperature etc and sends those to the physics, therefore ok
- PRs: everyone who opened comments to check and resolve conversations, then approve if satisfied