Deal with velocity inputs (instead of just water mass transport)

[briochemc]

Right now I use umo and vmo (mass transport in kg/s IIRC). Sometimes CMIP output only include uo and vo (velocities in m/s). It would be good to be able to deal with these too.

I think this just needs a conversion to mass transport using seawater density and cell face areas?

[briochemc]

Draft implemented in 7ac2be1. Should add better tests and visualize difference between using (umo,vmo) vs (uo,vo) locally as well.

[briochemc]

Visual tests show it's not working. I need to "regrid" the B-grid uo,vo to the C-grid u,v positions to match the C-grid umo/vmo positions of the tests.

[briochemc]

Related discussion on Pangeo discourse: https://discourse.pangeo.io/t/interpolate-onto-different-arakawa-grid-positions/4560/5

[briochemc]

Partial solution in 05f604c for ACCESS at least (uo/vo on B-grid vs umo/vmo on C-grid). There is still some difference, but I'm unsure how to deal with it. Maybe it has to do with the orientation where instead of the area derived from the cell thickness I should instead use the upstream cell thickness, or use the mean or the minimum of both directions, or maybe it's non-rectangular vertical face shapes, or density? The right way to check would be to dive into the MOM code... 😨