Updates to ODA driver suggested by @Hallberg-NOAA Issue#277

[MJHarrison-GFDL]

These changes follow suggestions from @Hallberg-NOAA for clarity in documentation and code readability. These modifications should not impact existing applications (e.g. SPEAR) which are reliant on this code.

[codecov]

Codecov Report

Merging #297 (561bdd3) into dev/gfdl (3f57d75) will increase coverage by 0.00%.
The diff coverage is 0.00%.

❗ Current head 561bdd3 differs from pull request most recent head 6984f9a. Consider uploading reports for the commit 6984f9a to get more accurate results

@@            Coverage Diff            @@
##           dev/gfdl     #297   +/-   ##
=========================================
  Coverage     37.10%   37.10%           
=========================================
  Files           263      263           
  Lines         73791    73789    -2     
  Branches      13759    13756    -3     
=========================================
  Hits          27381    27381           
+ Misses        41348    41346    -2     
  Partials       5062     5062           

Impacted Files	Coverage Δ
src/core/MOM.F90	51.38% <0.00%> (ø)
src/ocean_data_assim/MOM_oda_driver.F90	0.00% <0.00%> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[Hallberg-NOAA]

Thank you very much for these updates, @MJHarrison-GFDL. They are a significant improvement to the code!

I have a couple of other minor, specific tweaks to suggest that I think would make this even better and avoid the need for another revision to comply with an imminent change to the optional arguments for the get_param interface for real variables.

The other changes that we might consider as a part of this would be:

1. Change the units of the dt argument to apply_oda_tracer_increments() from [s] to [T_to_s],
2. Correct the units of T_bias and S_bias at about line 570 in get_bias_correction_tracer(), including changing the scale factor in the call to horiz_interp_and_extrap_tracer() for temperature from scale=US%degC_to_C to scale=US%degC_to_C*US%s_to_T, as well as the floor on T_bias at about line 592, and similarly for salinity.

[MJHarrison-GFDL]

Thank you @Hallberg-NOAA . I have addressed your latest comments. I modified the masking in the bias_adjustment routine so that it is based on the masking flag returned from horiz_interp_and_extrap_tracer, rather than the arbitrary ceiling value used in the previous version. This change should not impact answers for MOM6_version>=20190101 .

[Hallberg-NOAA]

Thanks for the updates to address the overall comments @MJHarrison-GFDL , but the specific comments relating to the get_param calls also need to be dealt with. As written, this code will no longer compile once PR #300 is merged into dev/gfdl.

[marshallward]

Gaea regression: https://gitlab.gfdl.noaa.gov/ogrp/MOM6/-/pipelines/17927 ✔️