Deal with any vertical coordinates

[briochemc]

Future CESM, GFDL, and ACCESS models, to name a few, will likely use MOM6+, which allows them to use very different vertical coordinates. It would be good to allow for these grids at some point.

[briochemc]

I think this is just a case of using cell thicknesses instead of a 1D depth vector. Although if velocities are used (#3), I'm unsure if they need to be projected along the horizontal coordinates first before being transformed into mass transports?