gdas-validation naming convention for diagnostic files

[RussTreadon-NOAA]

The finalize method in g-w ush/python/pygfs/task/atm_analysis.py assumes fv3jedi_var.x netcdf diagnostic files end with the suffix .nc4.

         # get list of diag files to put in tarball
        diags = glob.glob(os.path.join(self.task_config.DATA, 'diags', 'diag*nc4'))

This is not the case in gdas-validation. Our yamls use .nc as the suffix for netcdf diagnostic files. For example,

satwind_goes-16.yaml:      obsfile: $(DATA)/diags/diag_satwind_abi_goes-16_{{ current_cycle | to_YMDH }}.nc
satwind_goes-17.yaml:      obsfile: $(DATA)/diags/diag_satwind_abi_goes-17_{{ current_cycle | to_YMDH }}.nc
scatwind_metop-a.yaml:      obsfile: $(DATA)/diags/diag_ascatw_ascat_metop-a_{{ current_cycle | to_YMDH }}.nc
scatwind_metop-b.yaml:      obsfile: $(DATA)/diags/diag_ascatw_ascat_metop-b_{{ current_cycle | to_YMDH }}.nc

As a result gdasatmanlfinal creates an empty atmstat file (diagnostic file tarball).

We need to either

1. change the diagnostic file suffix to .nc4 in the observation yamls, or
2. update g-w atm_analysis.py to look for diagnostic files with the .nc suffix

This gets to the larger question of what suffix to use for netcdf files as we transition to JEDI. Currently the GFS uses .nc even though GFS .nc files are netCDF-4.

[CoryMartin-NOAA]

Building off of this discussion, the Unidata netCDF4 developers seem to suggest that .nc is their preferred suffix (see https://docs.unidata.ucar.edu/netcdf-c/current/faq.html#why-arent-different-extensions-used and https://docs.unidata.ucar.edu/netcdf-c/current/faq.html#What-convention-should-be-used-for-the-names-of-netCDF-files)

But we should discuss the convention of filenames in order to ensure that all the IODA observation file names are consistently named, we can iterate and comment on this Google doc (anyone with NOAA credentials can access): https://docs.google.com/document/d/1QLq5a_uVlTkcUkY8CfLOHCRzcLj5yvptigaiQ8B2LtY/edit?usp=sharing

[RussTreadon-NOAA]

Partially addressed by PR #937