Hi Jonathan
I agree that GCMD aren't precise. I would argue the way forward is to try to
get them to sub-class their list appropriately so we can use our appropriate
sub-class. If they don't, I think we should, and put some mechanism in place
to semi-automate the maintenance so when their list changes, we know, and
make some concious decisions, and where possible we change our list by
changing theirs first ...
(We also should put the list of constraints you list below in the CF
document).
The thing I worry *a lot* about is the proliferation of standard naming
schema; there is a strong risk of communities becoming self-isolating ...
It should be relatively trivial to put something in place to watch the GGMD
list automatically; what would be nicer is to have a "machine-understandable"
set of rules about what to do with the constraints you have below, and that's
what I would want to get GCMD (or us) to address via a set of attributes for
the standard names ... I'll get back to the list on this.
(Why dont you want country as a geographical region ... for example: it's a
perfectly valid thing to want to integrate the pollution arriving into and
leaving a country?)
Cheers,
Bryan
On Thursday 01 April 2004 8:26 am, Jonathan Gregory wrote:
> Dear Bryan
>
> The GCMD list hasn't grown, I don't think; we made a subset of it. We
> omitted
>
> * countries i.e. we included only geographical regions.
>
> * regions that could be specified by coordinate ranges, because that's what
> you'd do in CF. If we included "Western Hemisphere", for instance, it might
> conflict with a longitude range.
>
> * ill-defined regions, such as "Mid-latitude" (which if precise could
> anyway be specified by a coordinate range).
>
> * names that correspond to elements of standard_names, such as "Surface"
> and "Land surface".
>
> * names which aren't geographical regions, such as "Mantle".
>
> Hence I don't think consistency can be maintained entirely automatically.
> The GCMD list apparently serves a rather less precise function than the CF
> list. I wish it could be done automatically, because I agree that we want
> to keep in step, and in this case I agree we should ask GCMD to add
> "Indo-Pacific Ocean" to their list.
>
> Could we semi-automate it, by being made aware of when GCMD update their
> list and have some software suggest possible additions to the CF list?
>
> Best wishes
>
> Jonathan
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata
--
Bryan Lawrence, Head NCAS/British Atmospheric Data Centre
web: www.badc.nerc.ac.uk phone: +44 1235 445012
CLRC: Rutherford Appleton Laboratory, Chilton, Didcot, OX110QX, UK
Received on Thu Apr 01 2004 - 00:45:51 BST
