Use conditional values for the MOM5 "type" variant.

[penguian]

See #116

Each of access-esm1p5, access-esm1p6 and access-om2-bgc currently uses different type values with different versions of mom5,as follows.

Package	mom5 version	type
access-esm1p5	access-esm1.5	ACCESS-CM
access-esm1p6	access-esm1.5	ACCESS-CM
access-esm1p6	access-esm1.6	ACCESS-ESM
access-om2-bgc	default	ACCESS-OM-BGC

As per #116 , use conditional values for the type variant in mom5 to reflect this.

[dougiesquire]

@penguian, why are there two entries in the table for access-esm1p6?

[anton-seaice]

Is the suggestion to use the type variant instead of version for building from the same codebase for different deployments ?

[harshula]

Note, The version determines what source code is used.

[dougiesquire]

@harshula and I discussed removing the type variant and just setting the type in the package based on the version. Currently I think it’s possible to have incompatible combinations of version and type. It should also be similarly possible to remove the access-gtracers variant.

I think our versions are:

version	branch	type	access-gtracers
access-om2	master	ACCESS-OM	T
access-esm1.5	access-esm1.5	ACCESS-CM	F
access-esm1.6	master	ACCESS-ESM	T

A couple of questions:

• Is the plan to ever include an access-cm2 version in this package? I think ACCESS-CM2 uses the CM2-0.1 MOM5 tag. From a quick glance, it looks like we could move to using our master branch, but we’d have to check properly.
• Do we want to retain support in the SPD for ACCESS-OM2 with the old WOMBAT?

[harshula]

Hi @aidanheerdegen ^

[anton-seaice]

• Do we want to retain support in the SPD for ACCESS-OM2 with the old WOMBAT?

Oh - I forgot to reply about this.

This is the only config which has sea-ice bgc coupling (as you know), whilst it doesn't get much use, it does get occasional use and I think is "nice to have" still

[aidanheerdegen]

Is the plan to ever include an access-cm2 version in this package?

Yes. My understanding was that CM2 was using the MOM5 master branch (or a tag on that branch), but we would need someone to confirm. Maybe @atteggiani or @penguian?

Re: old wombat and sea-ice bgc coupling

Is there any plan/desire to add sea-ice bgc coupling to the generic tracer BGC implementation?

[dougiesquire]

Is there any plan/desire to add sea-ice bgc coupling to the generic tracer BGC implementation?

Yes, though probably not for a while since this got more complicated with the move to generic tracers

[aidanheerdegen]

I reckon add support for building with legacy wombat when it's required, i.e. someone needs sea-ice bgc with an updated version of the code. Until then they can use legacy builds, no?

[atteggiani]

Is the plan to ever include an access-cm2 version in this package?

Yes. My understanding was that CM2 was using the MOM5 master branch (or a tag on that branch), but we would need someone to confirm. Maybe @atteggiani or @penguian?

Re: old wombat and sea-ice bgc coupling

Is there any plan/desire to add sea-ice bgc coupling to the generic tracer BGC implementation?

As mentioned by Dougie, the suite u-cy339 uses CM2-0.1 (https://code.metoffice.gov.uk/trac/roses-u/browser/c/y/3/3/9/trunk/suite.rc#L285 - MOSRS account needed).
But I'm unsure if there is a more "updated" version of ACCESS-CM2 suite that uses a different branch.