Misc fix: 3-D input field read with dycore enabled; cam_in/cam_out from physics data & Add FADIAB test

[jimmielin]

Tag name (required for release branches):
Originator(s): @jimmielin

Description (include the issue title, and the keyword ['closes', 'fixes', 'resolves'] followed by the issue number):

This PR fixes a set of miscellaneous bugs that were preventing a standalone case to run with the SE dycore - namely, the FADIAB compset.

• Fixes input field read error (on dynamics INI grid) for all 3-D variables
  Previously the model would crash when reading any 3-D field in ncdata with the SE dycore

INFLD_REAL8_3D: field = U, grid = INI
INFLD_REAL8_3D: field(1:16,0:-1,1:3), file(488,30,1)
 INFLD_REAL8_3D: Mismatch between array bounds and field size for U, dimension 2

the vertical dimension was not being handled correctly in infld_real8_3d. Additionally, a dimensions name check was missing the right index offset.

• Removes pointer attribute from cam_in and cam_out throughout cam_comp.F90 as this structure no longer has chunks in SIMA.
• Updates NUOPC atm component to use cam_in and cam_out directly as allocated in physics_types.F90 via the CAM registry. Otherwise there would be

forrtl: severe (408): fort: (7): Attempt to use pointer CAM_OUT when it is not associated with a target

• Adds FADIAB test on SE dycore. The CAM test uses analytic IC, but I decided to make it read a default ncdata here to have test coverage for this functionality as it was fixed by this PR.

Describe any changes made to build system: N/A

Describe any changes made to the namelist: N/A

List any changes to the defaults for the input datasets (e.g. boundary datasets): N/A

List all files eliminated and why: N/A

List all files added and what they do: N/A

List all existing files that have been modified, and describe the changes:
(Helpful git command: git diff --name-status development...<your_branch_name>)

M       src/control/cam_comp.F90
  - remove pointer from dechunkized cam_in/cam_out

M       src/cpl/nuopc/atm_comp_nuopc.F90
  - use cam_in/cam_out from physics_types, not inline

M       src/utils/cam_field_read.F90
  - fixes to infld_real8_3d

M       cime_config/testdefs/testlist_cam.xml
  - add FADIAB test with read initial condition (for coverage of dycore-enabled ncdata read)

If there are new failures (compared to the test/existing-test-failures.txt file),
have them OK'd by the gatekeeper, note them here, and add them to the file.
If there are baseline differences, include the test and the reason for the
diff. What is the nature of the change? Roundoff?

B4B except new tests

derecho/intel/aux_sima:
SMS_Ln9.ne3pg3_ne3pg3_mg37.FADIAB.derecho_intel.cam-outfrq_se_cslam (Overall: DIFF): new baseline

derecho/gnu/aux_sima:
SMS_Ln9.ne3pg3_ne3pg3_mg37.FADIAB.derecho_gnu.cam-outfrq_se_cslam (Overall: DIFF): new baseline

If this changes climate describe any run(s) done to evaluate the new
climate in enough detail that it(they) could be reproduced:

CAM-SIMA date used for the baseline comparison tests if different than latest:

[nusbaume]

Thanks for the bug fixes @jimmielin! I have some questions and cleanup requests, but otherwise it looks good!

[nusbaume]

Should these be declared here, or should they be coming from physics_types? Maybe a better question is are these variables needed here at all? It looks like they are only used in a commented-out subroutine call, and my guess is that subroutine could just use cam_in and cam_out directly from physics_types itself if it wanted to.

[jimmielin]

It looks confusing here I think these should be inout arguments since they're passed in from the nuopc coupler. I'll update that part.

I don't know whether they're really needed - there's commented out code as you mentioned and I am not sure how CAM-SIMA will handle restarts (or does it already handle restarts?) -- this and the subsequent various phases have cam_in and cam_out passed as inout but I'm not sure what kind of role they play in CAM-SIMA, it just seems to exactly replicate the subroutine signatures for CAM.

[jimmielin]

All arguments removed or switched to use from physics_types!

[nusbaume]

Are these arguments actually needed anymore? It doesn't look like they are being used anywhere.

[cacraigucar]

Looking at the ESCOMP/CAM version, it appears that there may be uses for these during future development of code which hasn't been ported yet. Your call on whether we keep them or not

[jimmielin]

Thanks, I've removed all of the arguments. There is one that is passed down to the dycore, and I've changed that to a use statement that works.

[nusbaume]

Are these arguments actually needed? They don't appear to be used anywhere in the subroutine.

[cacraigucar]

In ESCOMP/CAM, these variable are used in cam_write_restart. Do we have writing restart implemented in CAM-SIMA yet? If not, and we think this is where restart will happen, then I propose we leave them in

[jimmielin]

A use statement from physics_types will work, so I've removed all the arguments.

[nusbaume]

Do we need to bring in cam_out_t and cam_in_t if we are no longer declaring the variables in this module?

[jimmielin]

Thanks, removed!

[jimmielin]

This is an example that will work instead of passing in through the subroutine. In the end they're all the same object.