computation of vorticity terms at the boundary

[raphaeldussin]

This is currently discussed in some emails but sedimented into mailboxes.

Do we want to have pycnal computed those terms ?
I think we should be able to compute them outside of this regridding tool.

PyCNAL is able to process as many scalars variables (T,S,ssh) and couple of vector components (u,v), (uice,vice),... If we compute (dv/dx) and (du/dy) fields before passing them to PYCNAL, they
can be treated as either scalar or vector variables.

I think that since u,v and x,y co-rotate, there is no need to rotate these fields. If I am wrong, the option to do it is to pass those field as a new vector variable and use the options :

rotate_velocities=True, rotation_angle='you read it from the ocean_hgrid.nc', Arakawa='B'

NB: since the resulting gradient is on the q-point, both fields are colocated so we want to make
sure that PyCNAL does not re-colocate them (coded for Arakawa='C')

please try different options and let me know what works.