[CF-metadata] proposed new vertical coordinate

[Jonathan Gregory]

Dear Ben

I agree that the proposal is good - thanks. I would comment that the quantity
ln(p0/p) could have a standard name which says what it is explicitly, rather
than "atmosphere_log_pressure_coordinate". I suppose that it could be called
"natural_logarithm_of_reciprocal_air_pressure", adding a note to the
stdname definition about the reference pressure, since logs can only be taken
of dimensionless quantities anyway. The other dimensionless coordinates have
got stdnames of such-and-such coordinate because those quantities have no other
purpose and they can't so easily be described. What do you think? Would that
be awkward?

Best wishes

Jonathan