[CF-metadata] axis attribute

[Jonathan Gregory]

Dear Richard

> In this case, I suspect it would still work if the "x" and "y" indices 
> are transposed, so as long as we can identify the two dimensions which 
> are horizontal.
Right. I agree with that, in my alternative argument of today i.e. X and Y
are useful to identify the horizontal coords. If X and Y are attached to
1D coordinate variables, then any 2D auxiliary coordinate variable which
has those dimensions must be in the horizontal. That is a bit more work, but
it seems more natural to me that X and Y should belong with 1D axes.

> But there could be cases where the parity flip would 
> matter, e.g. where axial quantities such as vorticity are involved. So I 
> think it is useful to allow the axis = X and Y attribute for coordinate 
> variables that are a continuous transformation from cartesian 
> coordinates or lat/lon coordinates (without a parity inversion).
I would say that the standard name is the right way to identify lon and lat
specifically. Maybe we could have a convention that X-Y-up is a right-handed
coordinate system - what do you think?

Best wishes

Jonathan