[CF-metadata] physical vs dimensional units

From: Steve Emmerson <steve>
Date: Mon, 04 Apr 2011 09:56:19 -0600

Martin,

On 04/04/2011 09:37 AM, Schultz, Martin wrote:
> if this is the direction to take, then we will need several "expressed_as" standard_names for atmospheric chemistry. Typical example: volatile organic compounds are often "expressed as" mass mixing ratio based on either their molecular mass or in "grams C" which counts only the mass of the C atoms (because this is what counts in the sense that it ultimately produces CO2 at the end of the reaction chain). Hence, one would need things like
> "mass_mixing_ratio_of_ethane_in_air_expressed_as_carbon_mass" and many similar ones. You can do the same with nitogen for species like NO, NO2, HNO3, N2O, etc.

Not necessarily. If you have an ambiguous physical quantity (i.e., one
that can be expressed in multiple -- but incompatible -- units), then
you should be able to use its "unit" attribute to refine its meaning.

Note that the "unit" attribute could still be "clean" (i.e., without
attached information).

> Still, there may then be a need to define "readable units" which would be obtained from the combination of standard_name and units. I do sympathize with people who prefer to read "nmol mol-1" on a plot rather than "1" (or "1.e-9" in this case).

Indeed. In my opinion, every display application should allow the user
to specify non-default output units.

Regards,
Steve Emmerson
Received on Mon Apr 04 2011 - 09:56:19 BST