Grid and fix file generation for the coupled model

[DeniseWorthen]

DESCRIPTION OF CHANGES:

Adds a new utility cpld_gridgen which creates the following files required by the coupled model

• The CICE6 grid and mask file
• The mesh file for the desired OCN/ICE resolution, which is identical for MOM6 and CICE6. Also used by WAV.
• The lat,lon,bathymetry and land mask files required to generate the mod_def file for WW3 on the tripole grid.
• The mapped ocean mask on the FV3 tiles
• The ESMF regridding weights required to create the CICE6 IC from CPC (SIS2) reanalysis.
• The ESMF regridding weights required to remap the CICE6 or MOM6 output from tripole grid to a rectilinear grid (optional).

TESTS CONDUCTED:

A new regression test script is added for the utility.

cd fix
./link_fixdirs.sh emc $machine
cd ../reg_tests/cpld_gridgen
./rt.sh -b

Baselines have been created on hera,orion, and jet.

DOCUMENTATION:

Documentation has been added in sorc/cpld_gridgen.fd/docs

ISSUE:

Resolves Issue #143

CONTRIBUTORS (optional):

@MinsukJi-NOAA

[DeniseWorthen]

@GeorgeGayno-NOAA I believe the RTs are now functioning on all platforms. After 615a9fa I was able to verify several times on hera, orion and jet. I merged in the latest develop and then realized I had not updated this branch w/ the unit test which reads a netcdf file. That is where the current workflow jobs are failing so I will work on that now.

[MinsukJi-NOAA]

@GeorgeGayno-NOAA @DeniseWorthen The latest commit failed to build on all 3 platforms.

-- Found sigio: /apps/contrib/NCEP/libs/hpc-stack/intel-2022.1.2/sigio/2.3.2/lib/libsigio.a (found version "2.3.0")
-- Configuring done
CMake Error at sorc/global_cycle.fd/CMakeLists.txt:27 (add_library):
  Cannot find source file:

    ../../ccpp-physics/physics/sfcsub.F

  Tried extensions .c .C .c++ .cc .cpp .cxx .cu .m .M .mm .h .hh .h++ .hm 
  .hpp .hxx .in .txx


CMake Error at sorc/global_cycle.fd/CMakeLists.txt:27 (add_library):
  No SOURCES given to target: global_cycle_lib


CMake Generate step failed.  Build files cannot be regenerated correctly.

[DeniseWorthen]

That's the new ccpp submodule. I thought I had it working.

[MinsukJi-NOAA]

That's the new ccpp submodule. I thought I had it working.

I didn't know ufs_utils now has a submodule. after git submodule update, it is building.

[DeniseWorthen]

I've now gotten an error during runtime though:

./cpld_gridgen.sh: line 109: ncks: command not found

I think we need an NCO module loaded.

[GeorgeGayno-NOAA]

That's the new ccpp submodule. I thought I had it working.

I didn't know ufs_utils now has a submodule. after git submodule update, it is building.

Yes, a recent update to 'develop' now points to the CCPP physics repo.

[MinsukJi-NOAA]

Got the same ncks error. NCO module was added in one of my commits. It looks like they were deleted here:
DeniseWorthen@412f458

[DeniseWorthen]

Ah, sorry. I was merging develop and didn't realize. Let me fix it.

[MinsukJi-NOAA]

a0595bf passed all regression tests on hera and orion. Looks like jet needs a new baseline.

[GeorgeGayno-NOAA]

a0595bf passed all regression tests on hera and orion. Looks like jet needs a new baseline.

If you create the new baseline, I will copy it to the proper directory. Just let me know where the files are.

[DeniseWorthen]

I have one I was using earlier. Let me test it.

[GeorgeGayno-NOAA]

Thanks, @GeorgeGayno-NOAA I'll add an abort if someone tries more than one task. Thinking back, I'm pretty sure I played w/ running on more than one task for the 1/12 deg case (which we're not currently implementing). But I was doing that in an interactive job.

I can also push a change that deallocates some variables sooner, but in my test on Jet I still got out-of-memory errors with that change.

I have been able to run 3 times against the baseline I created on jet using the change I showed above.

I tried your branch at a0595bf with 4 mpi tasks. It failed with an error message as expected. Thanks. That will help users (and save us questions!).

[DeniseWorthen]

I have a jet baseline here /lfs4/HFIP/h-nems/Denise.Worthen/CPLD_GRIDGEN/BASELINE

[GeorgeGayno-NOAA]

I have a jet baseline here /lfs4/HFIP/h-nems/Denise.Worthen/CPLD_GRIDGEN/BASELINE

Thanks. I placed the data under the official directory.

[MinsukJi-NOAA]

I have a jet baseline here /lfs4/HFIP/h-nems/Denise.Worthen/CPLD_GRIDGEN/BASELINE

Thanks. I placed the data under the official directory.

Jet regression tests passed.

[GeorgeGayno-NOAA]

@DeniseWorthen and @MinsukJi-NOAA - is everything working now? I can do some final tests today, then merge to 'develop'.

[MinsukJi-NOAA]

@DeniseWorthen and @MinsukJi-NOAA - is everything working now? I can do some final tests today, then merge to 'develop'.

@GeorgeGayno-NOAA It looks good to me.

[DeniseWorthen]

@GeorgeGayno-NOAA I agree. I think this is ready now.

[GeorgeGayno-NOAA]

@DeniseWorthen I have a minor fix so it builds with GNU on Hera. I will check it into your branch.

[GeorgeGayno-NOAA]

The branch at eff6549 was successfully compiled on Jet, Orion, Hera (Intel and GNU), WCOSS-Cray and WCOSS-Dell. The cpld_gridgen regression tests were successfully run on Jet, Orion and Hera. Ready to merge.

[DeniseWorthen]

Thanks so much @MinsukJi-NOAA and @GeorgeGayno-NOAA for helping get this done.