Diagnostics bug fix + de-hardcoding refRatio in makeFineMask calls

[mbkuhn]

Summary

Fixing the use of the mask array in PrintMaxMACVelLocations to avoid going out-of-bounds Correcting places where the refinement ratio was hard-coded in makeFineMask calls

Pull request type

• [x] Bugfix
• [ ] Feature
• [ ] Code style update (formatting, renaming)
• [ ] Refactoring (no functional changes, no api changes)
• [ ] Build related changes
• [ ] Documentation content changes
• [ ] Other (please describe):

Checklist

The following is included:

• [x] new unit-test(s)
• [ ] new regression test(s)
• [ ] documentation for new capability

This PR was tested by running:

• the unit tests
  • [ ] on GPU
  • [x] on CPU
• the regression tests
  • [ ] on GPU
  • [ ] on CPU

Issue Number: #941 #1008

[mbkuhn]

Other details: I added the unit test (commit 406c00c), but it still passed without the bug fix. However, in debug mode, the test doesn't pass because it can properly detect the out-of-bounds index. Is that good enough? Any suggestions?

With the full PR, the bug is fixed, and it does pass the test in debug.

[marchdf]

I think it's fine. We only test OOB in debug mode. So the real thing is when we turn on the nightly test harness we have debug and release runs. Which we used to do and we will do again ;)

[mbkuhn]

This PR does change a lot of different files, but it's not well suited to being tested with regression tests, because the changes are purely in error measurements going to text or netCDF outputs, not in anything that would change the values of the plt files. However, I think the fact that the regression test run fine and the code compiles fine are good indications that those widespread changes are nothing to worry about. The one bug fix that does change the functionality is addressed with a unit test.