Add an option to point to the "enkfgdas" backgrounds instead of the pre-generated 30 member ensemble

[guillaumevernieres]

Work done

Mainly what the title says. The default ctest is now assuming that DOHYBVAR=YES, I don't really think it's worth it to cover the DOHYBVAR=NO in the ctest. I may change my mind, but in both cases, soca is configured to use the hybrid envar solver, the only difference is the origin of the ensemble members.

Work left to be done

We need to define an ensemble "COM_OCEAN/ICE_HISTORY_PREV" environment variable in the prep jjob (in the g-w). Flagged as a TODO in the code.

[AndrewEichmann-NOAA]

no complaints

[ShastriPaturi]

admissible

[emcbot]

Automated Global-Workflow GDASApp Testing Results:
Machine: hera

Start: Tue Nov 28 14:49:18 UTC 2023 on hfe07
---------------------------------------------------
Build:                                 *SUCCESS*
Build: Completed at Tue Nov 28 15:44:06 UTC 2023
---------------------------------------------------
Tests:                                 *SUCCESS*
Tests: Completed at Tue Nov 28 16:05:24 UTC 2023
Tests: 100% tests passed, 0 tests failed out of 53