Set flags properly for rotated tripolar grids

[Hallberg-NOAA]

Added code to tht two versions of clone_MD_to_MD in the FMS1 and FMS2 versions of MOM_domain_infra.F90 to properly change the flags when a tripolar grid is rotated, so that it does not lead to misleadingly incorrect answers. With the current versions of FMS, doing grid rotation testing with a tripolar grid will lead to error messages about the incomplete implementation of FMS, but these failures are preferable to the model silently working incorrectly. All answers are bitwise identical in cases that worked correctly before.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 37.12%. Comparing base (2b59089) to head (d9bd123).

Additional details and impacted files

@@             Coverage Diff              @@
##           dev/gfdl     #581      +/-   ##
============================================
- Coverage     37.13%   37.12%   -0.01%     
============================================
  Files           271      271              
  Lines         80779    80794      +15     
  Branches      15068    15082      +14     
============================================
- Hits          29999    29998       -1     
- Misses        45181    45195      +14     
- Partials       5599     5601       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[marshallward]

I am confused about the interpretation here. Generally turns is used to indicate a clockwise rotation from input grid to model grid. If MD_in has its north Y_AXIS folded and we rotate once, then the MOM_dom fold should be on the east face, i.e. MOM_dom%X_AXIS = FOLD_EAST_EDGE should have the fold on the east side. But here we set X_AXIS to FOLD_WEST_AXIS. Or is the operation backwards here?

The two-turn case looks correct to me.

Also, there are if modulo(qturns, 2) blocks preceding these if modulo(qturns, 4) = 1, 2, 3 blocks. These statements could be divided up and moved inside of the modulo(qturns, 2) blocks.

[Hallberg-NOAA]

The main point of this PR was to get FMS to issue a message indicating that it has not yet fully implemented rotated tripolar grids, rather than just silently doing completely the wrong thing when we tried to put in turns with a tripolar grid. You would be better positioned to describe the sense of the turns, @marshallward, than I am, and I am sure that you are right about the changes we need to make, and we should make them now. The true test, though, will be when FMS finally supports southern-, eastern- and western- tripolar grids, and then we can actually see what gives the right answers!

[Hallberg-NOAA]

This PR has been updated to make the changes suggested in the review.

[marshallward]

Looks good!

[marshallward]

Gaea regression: https://gitlab.gfdl.noaa.gov/ogrp/MOM6/-/pipelines/22871 ✔️