Just to stick my oar in, I'd like to support the URN idea, mainly for
the reason that the presence of a fully-qualified CRS (or any other
attribute) concisely conveys the *intent* of the data provider. Even
if the implementation is faulty (e.g. a typo in one of the attributes)
the presence of an authoritative CRS shows what was meant, and a tool
can flag any detected inconsistency. Mistakes in NetCDF headers are
so common in my experience that this would be very valuable from the
point of view of someone (like me) who tries to develop tools, even if
it's sometimes redundant. I think I've just restated in different
words what others have posted before but I thought it was worth
showing my support!
For example, take files produced by some variant of the NEMO tripolar
ocean model. There is no analytical form of the numerical grid so
it's a hard CRS to work with, and very unlikely to be defined by EPSG
(yet). It would be very difficult for a tool to detect the presence
of a NEMO grid from the CF axis definitions alone - it would be
recognised simply as a generic curvilinear grid. However, if there
were a globally unique identifier for the CRS, a tool could recognise
this and behave appropriately (for example the WMS implementation I'm
working on will recognise a NEMO grid and can deal with it more
efficiently than a generic grid, but I'm having to check for the
presence of an ad-hoc attribute at the moment, yuck).
(There are of course several variants of the NEMO grid but each could
have its own URN.)
Jon
On 10/3/07, Benno Blumenthal <benno at iri.columbia.edu> wrote:
> This is an example which illustrates the usefulness of namespaces with
> attributes in particular and the semantic web in general: alternate
> representations of the same information. I have been working on an
> ontology for CF and I intend to crosswalk between CF representations of
> projections and OpenGIS representations of projections: the Semantic
> Web framework would allow writing rules for generating one set of
> metadata from the other, and all representations would be simultaneously
> available.
>
> A writer of software could then choose which representation to use,
> independent of which representation was used in the dataset metadata.
>
> Benno
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
>
--
--------------------------------------------------------------
Dr Jon Blower Tel: +44 118 378 5213 (direct line)
Technical Director Tel: +44 118 378 8741 (ESSC)
Reading e-Science Centre Fax: +44 118 378 6413
ESSC Email: jdb at mail.nerc-essc.ac.uk
University of Reading
3 Earley Gate
Reading RG6 6AL, UK
--------------------------------------------------------------
Received on Wed Oct 03 2007 - 11:59:18 BST
