[CF-metadata] Dimensionless vertical coordinate values

From: Jonathan Gregory <j.m.gregory>
Date: Sun, 30 Sep 2012 09:32:05 +0100

Dear Richard

I've just looked through the definitions again while thinking about Rich's
ticket 93, and I agree with you (again) that this needs clarification. In
some cases it could be made clearer which is the quantity that you would
expect to be used as a vertical coordinate, with the standard name that is
given in the relevant definition of Appendix D.

Hybrid height is a peculiar case. It says, "There is no dimensionless hybrid
height coordinate. The hybrid height is best approximated as a(k) if a
level-dependent constant is needed." There are two unusual things about this.

* It seems inevitable that the vertical coordinate is indeed one of the
terms, as you said. There must be a 1D monotonic vertical coordinate variable,
and I can't see what it could be other than a(k), the height part of the
hybrid coordinate. Perhaps it is this quantity, therefore, which has the
standard_name of atmosphere_hybrid_height_coordinate, even though it is not
itself really a hybrid. What do you think?

* It's not dimensionless. Although Appendix D was introduced for coordinates
which are dimensionless, I don't think that's a necessary restriction, because
formula terms can have a more general function of computing a multidimensional
vertical coordinate from various 1D terms, which might be either dimensional
or dimensionless, one of which could be the 1D monotonic vertical coordinate
variable.

So I'm disagreeing with what I said before. The vertical coord var might have
to be one of the terms. If you agree and have time to draft a defect ticket,
or a ticket to modify the convention if it seems more substantial, that would
be interesting and useful.

Best wishes and thanks

Jonathan
Received on Sun Sep 30 2012 - 02:32:05 BST