CCPP Framework Meeting Minutes 2020 04 23

Update from last week:

• removal of CCPP dynamic build from NCAR master branches - happening as we speak
• delayed from July to October: UFS Short-Range Weather Application public release, we are working on the CCPP aspect of it (which physics will be supported); most likely, it will still use ccpp_prebuild.py (no dynamic CCPP build anymore)

Work on cap_gen.py:

• reports generated by cap_gen.py
• comes with two kinds of tests:
  • from the command line (tests shell interface)
  • Python script (tests Python interface)
• the main issue to work on is https://github.com/NCAR/ccpp-framework/issues/282
  • how does the host model query the additional attributes in the accompanying properties array?
  • no solution yet, need to discuss with NRL once we have a proposal
  • another (related?) issue - with a proposal - is wrt water species (dry, wet, ...): separate arrays?
    • can we leverage this idea for issue 282?
    • would GFDL be ok with having an advected-tracers array, and a non-advected tracers array?
    • inside the physics: can we also make use of this idea for the vertically diffused tracer array?
    • but inside physics we can also use the information in the properties array (different ways forward)
    • an auto-generated interstitial scheme by the CCPP framework to assemble these vertically diffused tracer arrays seems to be the easiest/best way