Support for new model grids

[ThomasHaine]

Oceanspy supports structured Arakawa C-grid depth-level models, like MITgcm. Currently, full functionality and testing has focused on MITgcm.

• @malmans2 has some experience using Oceanspy with NEMO, which is also on a C-grid.
• @Mikejmnez has experimented with HYCOM, but the hybrid vertical coordinate is already interpolated to depth levels.
• MOM6 is on a C-grid too, although its arbitrary Lagrangian Eulerian (ALE) vertical coordinate differs from $z^*$ used by MITgcm.

What models (grid topologies) should we prioritize for future Oceanspy support? What technical steps are involved in extending Oceanspy to support these models? E.g., what's required to support NEMO and/or MOM6?

Let's discuss!

[ThomasHaine]

NEMO is a C-grid model, but the convention on the layout of the velocity and pressure points is different to MITgcm. This is easy to handle.
Vertical coordinates and topography require some more thought.

See also this chat.

[ThomasHaine]

CESM (POP ocean model) uses a B-grid but it also has a tripolar grid with 2 poles in the northern hemisphere. See this link.

Also, xgcm already has some functionality to handle POP output. See POP-tools.