Add optional argument to FMS2 version of get_field_size

[MJHarrison-GFDL]

• Omitting this argument returns field size using fms1 format, e.g. (Nx,Ny,1,Nt) for data with no vertical extent
  Otherwise, if fms1_format=.false., then return (Nx,Ny,Nt,0)

[codecov]

Codecov Report

Merging #64 (2358142) into dev/gfdl (fc5253f) will not change coverage.
The diff coverage is n/a.

@@            Coverage Diff            @@
##           dev/gfdl      #64   +/-   ##
=========================================
  Coverage     29.02%   29.02%           
=========================================
  Files           244      244           
  Lines         71854    71854           
=========================================
  Hits          20854    20854           
  Misses        51000    51000           

---

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 fc5253f...2358142. Read the comment docs.

[marshallward]

I agree with the proposal to remove the fms1_format flag and return a consistent format regardless of backend FMS.

[marshallward]

Gaea regression: https://gitlab.gfdl.noaa.gov/ogrp/MOM6/-/pipelines/14790 ✔️