[main] Followup for #1187 for NOAA models

[grantfirl]

Description of Changes:

These changes allow NOAA models to compile/run with the changes from #1187.

• change calls to subroutines in NOAA-specific interstitials to reflect changes in interfaces for the added ccpp_bcast calls
• remove ccpp_bcast calls from two subroutines in aerinterp.F90 due to hanging UFS RTs; if reinstated in the future, the overloaded ccpp_bcast subroutine will need to add interfaces for 4D and 5D integer arrays, plus the entire arrays of aer_pres and aer_in should be allocated at once so that sending/receiving MPI ranks can effectively broadcast; there may also be an issue with different size slices of the array being broadcast from different ranks.

Tests Conducted:

SCM RTs - see NCAR/ccpp-scm#682
UFS RTs - see ufs-community/ufs-weather-model#3190 (tested along with many other accumulated physics changes, but B4B for all tests that actually finished)

Dependencies:

None

Documentation:

None

Issue (optional):

None

[climbfuji]

Looks as expected. Regarding the sending of slices, that would need to be implemented using other MPI functionality than bcast.

Let me note here that there are still schemes that read with all ranks instead of reading with rank zero and broadcasting (e.g. photochemistry). These should be looked at in the future, too.

[grantfirl]

Looks as expected. Regarding the sending of slices, that would need to be implemented using other MPI functionality than bcast.

Let me note here that there are still schemes that read with all ranks instead of reading with rank zero and broadcasting (e.g. photochemistry). These should be looked at in the future, too.

@climbfuji Agreed. I didn't want to extend the time to get this merged in and clean up some of the other code management tasks that need to happen after this is taken care of. Plus, the sync from NCAR/main to ufs/dev was already pretty large and I wanted to just get this initial capability in as-is before extending it.

Added #1210