CCPP Framework Meeting Minutes 2019 11 14

Physics unification MPAS, CESM, FV3

• only a handful of schemes that have been ported for FV3 are of interest to NCAR
• radiation (RRTMG)
  • only pre/post/entry points should be different, core code should be the same
  • for MPAS, Jimy Dudhia is working on rrtmg lw, Laura Fowler rrtmg sw
  • this would be a good target to unify
• governance: people interesting at using a scheme within a certain suite; for the above example:
  • Jimy and Laura from NCAR
  • Steve or another physics person from CESM?
  • someone from NOAA (DTC?)
  • must fit with the ongoing ccpp-physics governance discussion at NOAA
  • discuss at PI call?
• where should the authoritative repository for a common pool of physics be?
  • NCAR/ccpp-physics already taken for NOAA physics
  • escomp not owned by an organization, hence this would be the best place

Proposal for adding information on how to obtain/checkout code from Steve

• add <source>..</source> property (optional) for groups, subcycles, schemes, ...
• in general, replacing <scheme>name</scheme> with <scheme name="..."> other attributes go here </scheme> would give us more flexibility for adding attributes (such as tuning parameters, optional arguments, ...)
• information that is purely scheme-specific should got into metadata file, all group/suite/host dependent information should go in SDF
• concerns are that adding more information to the SDF will make it unreadable and too complex for users
  • can syntax highlighting help with reading it?
  • develop a suite definition file viewer/explorer (with much more capabilities, e.g. tracing variables throughout a suite)?