Should Model be renamed to IncompressibleModel or BoussinesqModel?

[ali-ramadhan]

I guess BoussinesqModel is a subset of IncompressibleModel? Maybe BoussinesqModel is only true if you're running with tracers. If you're running with zero tracers then there's no buoyancy per se.

[glwagner]

Are we anticipating having more than one model?

There’s a few possible designs. For example, we could abstract the “equation”, rather than the model.

[glwagner]

If we have Compressible than Incompressible makes sense. But if we have QGModel, then Boussinesq makes more sense. Our choice depends on the models available.

[johncmarshall54]

I would say incompressible. Essential assumption is divv=0. We can have a non-Boussinesq incompressible model by going in to pressure coordinates. This would be one version of a family of incompressible models. Ali also has an incompressible version. So to me a compressible/incompressible naming fork makes sense.

…

On Thu, Dec 12, 2019, 9:25 AM Gregory L. Wagner ***@***.***> wrote: If we have Compressible than Incompressible makes sense. But if we have QGModel, then Boussinesq makes more sense. Our choice depends on the models available. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#566>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKXUEQRYWL2UJ3KNBCOZYADQYJCVPANCNFSM4JZ6UGIA> .

[glwagner]

I think @johncmarshall54's stance is sensible. It's more likely we will have non-Boussinesq and compressible models than a QGModel.

[ali-ramadhan]

Sorry for the slow reply. Indeed I was thinking of introducing a CompressibleModel type but will open another issue about it.