Hello Steve,
On Tue, Mar 19, 2013 at 6:19 PM, Steve Hankin <steven.c.hankin at noaa.gov> wrote:
> A question to debate in your trac ticket. Per the CF documentation, the
> definition of the standard_name is "The name used to identify the physical
> quantity"
> (http://cf-pcmdi.llnl.gov/documents/cf-conventions/1.6/cf-conventions.html#standard-name).
> It is the 'units' string (together with the netCDF data type) that describe
> how that physical quantity is represented as a value in the binary file.
Thanks for pointing this out. I find standard names so useful that I
overlooked this aspect of the convention. Why only physical
quantities? It does seem a bit outdated to me now that software is
increasingly doing examination of netCDF file content.
> A standard_name for the physical quantity time already exists, of course --
> it is "time".
The way it is used in the CF convention "duration" would be a more
appropriate term for me. Just saying...
> It is listed as such in the CF standards names table. So
> should "datetime_iso8601" be a proposal for a new and recognized value of
> the time 'units' attribute, rather than a new standard_name? This isn't so
> clear. To start answering it ...
>
> What is the reason that standard names exist? Arguably they are to guide
> generic software (including metadata search engines) in their ability to
> locate physical quantities in a CF file. In that case it is standard_name
> ="time" that should guide the software to the variable; and
> units="datetime_iso8601" that will instruct the software on how to interpret
> the values it finds there.
Yes, it makes sense to me but "datetime_iso8601" as units value does
not seem to be a winning proposal on this list.
-Aleksandar
Received on Wed Mar 20 2013 - 20:59:03 GMT
