Bug fixes for RRFS 32-bit physics support, and minor feature additions

[SamuelTrahanNOAA]

Description

Fixes two bugs and adds features:

1. aerinterp.F90 in CCPP stepped all over memory when kind_phys=4 because it read 8-byte reals into that array.
2. aerinterp.F90 did not report NetCDF errors up the call tree via errmsg & errflg
3. add regression tests for rrfs conus13km 32-bit physics and -DFASTER=ON
4. Initialize some uninitialized variables in GFS_typedefs.F90
5. Changes the COMPILE lines to use names instead of numbers so we don't have to renumber the entire rt.conf when adding a compilation line

Input data additions/changes

• No changes are expected to input data.

Anticipated changes to regression tests:

• Changes are expected to the following tests
  New regression tests for 32-bit physics in RRFS

Subcomponents involved:

• AQM
• CDEPS
• CICE
• CMEPS
• CMakeModules
• FV3
• GOCART
• HYCOM
• MOM6
• NOAHMP
• WW3
• stochastic_physics
• none

Combined with PR's (If Applicable):

Commit Queue Checklist:

• Link PR's from all sub-components involved in section below
• Confirm reviews completed in ALL sub-component PR's
• Add all appropriate labels to this PR.
• Run full RT suite on either Hera/Cheyenne AND attach log to a PR comment.
• Add list of any failed regression tests to "Anticipated changes to regression tests" section.

Linked PR's and Issues:

• fixes Add compilation lines without updating every COMPILE in the rt.conf #1796
• depends on fv3atm Fix memory errors: uninitialized variables, and NetCDF writes to arrays that are too small NOAA-EMC/fv3atm#668
• depends of ccpp-physics Fix bug in aerinterp read routines that break 32-bit physics and pass NetCDF errors up through errflg&errmsg ccpp-physics#85
• Uninitialized Grid variables in GFS_typedefs.F90 grid_create NOAA-EMC/fv3atm#667
• 32-bit physics crashes in aerinterp due to type mismatch in NetCDF calls. ccpp-physics#83
• aerinterp routines do not pass NetCDF errors up through errmsg and errflg ccpp-physics#84

Testing Day Checklist:

• This PR is up-to-date with the top of all sub-component repositories except for those sub-components which are the subject of this PR.
• Move new/updated input data on RDHPCS Hera and propagate input data changes to all supported systems.

Testing Log (for CM's):

• RDHPCS
  • Hera
  • Orion
  • Jet
  • Gaea
  • Cheyenne
• WCOSS2
  • Dogwood/Cactus (IKE Day, skipping)
  • Acorn
• CI
  • Completed
• opnReqTest
  • N/A
  • Log attached to comment

[DeniseWorthen]

@BrianCurtis-NOAA I had the same issue, and found hashing out lines 135 through 138 in rt.sh worked. However this is probably not the preferred fix. Otherwise I think you can export ACCNR prior to running.

Are you editing code during testing to get it working?

If you are, it needs to be brought back into the PR, discussed, or not changed at all.

We need to understand---are the RTs being tested with uncommitted changes?

[jkbk2004]

I agree with forcing along with line 216 in getopts block.

[zach1221]

@DeniseWorthen No, sorry. To clarify you can use the export command, for example "export ACCNR=${ACCNR:-epic}", on Orion, before running ./rt.sh. It doesn't require any changes to the rt.sh script. This is how I ran the tests, but I thought I'd offer the other workaround as well, while acknowledging it's likely not preferred.

[DeniseWorthen]

@zach1221 @jkbk2004 Have any RT logs been posted for cases where the PR was run w/o committed changes?

Edit: Thanks for the clarification.

[SamuelTrahanNOAA]

Do people want me to make changes or not? If so, what are the changes? I'm seeing too many conflicting comments.

[jkbk2004]

Do people want me to make changes or not? If so, what are the changes? I'm seeing too many conflicting comments.

@SamuelTrahanNOAA Please, go ahead to add fix. Sounds like option to force somewhere inside getopts block

[BrianCurtis-NOAA]

Do people want me to make changes or not? If so, what are the changes? I'm seeing too many conflicting comments.

Move lines: +134 to +141 to 216 (after the while getopts section)

[SamuelTrahanNOAA]

Move lines: +134 to +141 to 216 (after the while getopts section)

I tested that fix, and it works. With -a, the workflow uses the account I give it. Without -a, it aborts with the expected message.

The fix is in my branch now.

[BrianCurtis-NOAA]

@jkbk2004 @zach1221 other than Cheyenne, give it a look over and I assume you can start the merging.

[zach1221]

Thanks, @BrianCurtis-NOAA . Yes, since we're skipping Cheyenne, testing is complete and we can begin the merging process.

[dustinswales]

@zach1221 Physics has been merged.
@SamuelTrahanNOAA You can go ahead and update your FV3 submodule pointer.

[SamuelTrahanNOAA]

You can go ahead and update your FV3 submodule pointer.

I updated the fv3atm's submodule pointer to ccpp-physics. It now points to the authoritative ufs/dev branch. You can merge the fv3atm PR.

[zach1221]

FV3 sub-pr is merged. Please update the submodule pointer and revert the .gitmodules for FV3. @SamuelTrahanNOAA

[SamuelTrahanNOAA]

FV3 sub-pr is merged. Please update the submodule pointer and revert the .gitmodules for FV3

The deed is done. We are ready for final approvals and a merge.