Adjust ny for baroclinic channel to give 500 km domain

[xylar]

This is a port of https://github.com/MPAS-Dev/compass/pull/605

The following description is copied from there.

This PR adjusts the number of cells used in meshes for the baroclinic channel test case to ensure the 'standard' 160x500km domain is defined. Previously, the channel has had a shorter y-length of sqrt(3)/2 * 500 => 433km.

• When using the make_planar_hex_mesh routine, the number of cells along the y-axis should be scaled by 2/sqrt(3) to account for hexagonal geometry.

The channel is a standard community test case, and use of the 160x500km domain enables model intercomparison, e.g.:

• Ilıcak, M., Adcroft, A.J., Griffies, S.M. and Hallberg, R.W., 2012. Spurious dianeutral mixing and the role of momentum closure. Ocean Modelling, 45, pp.37-58.

• Petersen, M.R., Jacobsen, D.W., Ringler, T.D., Hecht, M.W. and Maltrud, M.E., 2015. Evaluation of the arbitrary Lagrangian–Eulerian vertical coordinate method in the MPAS-Ocean model. Ocean Modelling, 86, pp.93-113.

Checklist

• [x] User's Guide has been updated
• [x] Documentation has been built locally and changes look as expected
• [x] Testing comment in the PR documents testing used to verify the changes

[xylar]

Actually, it looks like only the spherical meshes are described in developers_guide/framework/mesh so we should probably add a new planar mesh section, here or in a separate PR.

I think that should be a separate PR. At the moment developers_guide/framework/mesh describes what's literally in the framework (polaris.mesh) but I think it would be okay to expand it to cover functions from mpas_tools like make_planar_hex_mesh that could typically be used in test cases, and what the proper usage is (including this issue).

[xylar]

Hmm, thinking about it some more, I think it belongs in a "Mesh" section in "Organization".

[xylar]

Testing

I successfully ran all baroclinic_channel test cases on Chrysalis with this branch. The results can be found at:

/lcrc/group/e3sm/ac.xylar/polaris_0.1/chrysalis/test_20230415/bcn_chan_fix_ny

[xylar]

@dengwirda, could you have a look at the code change here and see if you agree that they are equivalent to https://github.com/MPAS-Dev/compass/pull/605?

I verified that I get exactly the same values for ny as you have prescribed in compass for the 3 standard resolutions.

[xylar]

Thanks @alicebarthel! I'll fix that here.

[xylar]

Thanks @alicebarthel, @dengwirda and @cbegeman!

I'll merge in just a moment when CI finishes.

[xylar]

Retroactive testing (bad form!)

I changed the resolution to 0.1 for the test cases that are currently at 10 km. Then, I did polaris list and go:

   0: ocean/baroclinic_channel/100m/default
   1: ocean/baroclinic_channel/100m/decomp_test
   2: ocean/baroclinic_channel/100m/restart_test
   3: ocean/baroclinic_channel/100m/threads_test

as expected.