Add JEDI-based analysis calculation

[DavidNew-NOAA]

This PR is a companion to GDASApp PR #1293 and Global-Workflow PR #2949.

It does two things:

1. It adds the YAML for the fv3jedi_analcalc utility, introduced in the above GDASApp PR.
2. It prepares the variational and ensemble DA FV3 increment converters to write increments using the FMS non-restart format created in FV3-JEDI PR #1251. The variational analysis job will also now write using this format for its intermediate increment before conversion to the FV3-readable format.

[DavidNew-NOAA]

@danholdaway Now that I've added the atmosphere_ prefix to analysis_variables in the atmosphere test template, the client testing fails. I assume it's using the ````developbranchparm/atm/jcb-base.yaml.j2``` which isn't yet updated, and that's why it's failing?

[danholdaway]

Actually you've stumbled upon something that isn't ideal. The key analysis_variables also needs to be a generic one because it is used here: https://github.com/NOAA-EMC/jcb-algorithms/blob/43d8ff6ba14baf9402ee58d0f3351a143c21211b/3dvar.yaml.j2#L19

So you need both analysis_variables and atmosphere_analysis_variables. Just use the YAML anchor in the testing YAML and JCB base in GDASapp. You can't add a model specific algorithm with a generic key and you can't use a model specific key in a generic algorithm. 🙃

[DavidNew-NOAA]

I would say this PR is complete, but DO NOT MERGE, because I'm holding off on the associated GW and GDASApp PRs until I have clarity on GW PR #2992 and ensemble recentering, both of which I hope to discuss with @aerorahul next week.

[DavidNew-NOAA]

Closing for now while I overhaul this set of PRs