Fix data read for on-grid interpolation

[angus-g]

The wrong MOM_read_data interface was being used: a 2D slice of a 3D field was expected, but the interface for a 2D field was being called.

See mom-ocean#1567.

[codecov]

Codecov Report

Merging #124 (3ff06be) into dev/gfdl (a965005) will decrease coverage by 4.70%.
The diff coverage is 0.00%.

❗ Current head 3ff06be differs from pull request most recent head b31ec83. Consider uploading reports for the commit b31ec83 to get more accurate results

@@             Coverage Diff              @@
##           dev/gfdl     #124      +/-   ##
============================================
- Coverage     33.48%   28.78%   -4.71%     
============================================
  Files           262      249      -13     
  Lines         71362    72991    +1629     
  Branches      13310        0   -13310     
============================================
- Hits          23898    21010    -2888     
- Misses        42992    51981    +8989     
+ Partials       4472        0    -4472     

Impacted Files	Coverage Δ
src/framework/MOM_horizontal_regridding.F90	36.53% <0.00%> (-5.63%)	⬇️
src/framework/posix.F90	0.00% <0.00%> (-55.27%)	⬇️
src/user/benchmark_initialization.F90	53.78% <0.00%> (-42.01%)	⬇️
...rameterizations/vertical/MOM_entrain_diffusive.F90	27.99% <0.00%> (-38.68%)	⬇️
src/parameterizations/vertical/MOM_kappa_shear.F90	34.40% <0.00%> (-36.64%)	⬇️
src/tracer/MOM_tracer_hor_diff.F90	35.22% <0.00%> (-36.16%)	⬇️
src/core/MOM_continuity_PPM.F90	39.46% <0.00%> (-31.03%)	⬇️
src/ALE/PCM_functions.F90	50.00% <0.00%> (-30.00%)	⬇️
...ameterizations/lateral/MOM_mixed_layer_restrat.F90	44.23% <0.00%> (-28.71%)	⬇️
src/ALE/PPM_functions.F90	56.75% <0.00%> (-26.58%)	⬇️
... and 234 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a965005...b31ec83. Read the comment docs.

[Hallberg-NOAA]

I have looked into this code and the underlying routines that are called, and I agree that it makes sense. Assuming that it passes the TC and pipeline testing, this PR should be accepted once it reaches the front of the queue.

[Hallberg-NOAA]

This has passed pipeline testing at https://gitlab.gfdl.noaa.gov/ogrp/MOM6/-/pipelines/15618.