[CF-metadata] Re: CF-1.0-beta5: grid mappings

From: Jonathan Gregory <jonathan.gregory>
Date: Tue, 11 Mar 2003 19:12:41 +0000

Dear Brian

Thanks for putting this into the standard. I agree that with this addition we
should say we have a proper (non-beta) release of the CF standard.

I have a few small comments below.

Best wishes

Jonathan


End of 2nd para of 5.6. "The mapping coordinates should have units which are
consistent with their standard_names."

Appendix A. "The attributes listed here are not applicable to grid mapping
variables. See Appendix E for the attributes to be used on grid mapping
variables."

Albers, Lambert: Do you mean if there are two standard parallels they are
supplied as an ordered 2-element vector?

Polar stereographic: I would suggest straight_vertical instead of
straight-vertical. Although this is less like proj, I think the unusual
appearance of a hyphen instead of an underscore would be a potential nuisance.

units: What would we do if a grid mapping parameter had dimensions which didn't
correspond to either the coordinate variables or true lat and lon? This is not
the case for any of the mappings currently in Appendix E, but I raise it in
case it would be a structural problem in the future. I don't think it would be;
I would suggest that we could get round it by having two attributes, both a
grid_mapping_parameter (numerical) and a grid_mapping_parameter_units (string).
Received on Tue Mar 11 2003 - 12:12:41 GMT