Hi Cecelia,
The CF group is very interested in this interaction and is involved in
a similar interaction with the PRISM group. The existing standard name
table, linked to the CF home page at
http://www.cgd.ucar.edu/cms/eaton/cf-metadata/index.html, is pretty much a
first draft. It is our highest priority to advance this to a stable
version. Any input we receive from ESMF would be most welcome.
Brian
On Wed, 9 Oct 2002, Cecelia Deluca wrote:
>
> Hi all,
>
> I just wanted to add a few comments related to conventions for field
> names.
>
> We asked ESMF application groups to list all the input fields and other
> data that are required to run a model, and data that the model can export.
> We also asked application groups to identify the data items in those
> lists that have entries in the CF conventions. Where CF names don't
> exist, we said we'd interact with the CF group and see if they would be
> willing to augment the convention. *We have no plans to create a new
> ESMF convention or standard.*
>
> During the recent meeting we also discussed supporting multiple,
> domain-specific conventions, just as you suggested, so that one would not
> be restricted to CF.
>
> In any case, ESMF will not require that groups use standard field names,
> though it's certainly helpful to follow a convention. And we did
> agree to use CF for the interoperability experiments (which I interpreted
> as meaning 1) in the preparation of import/export lists as described in
> the first paragraph above and 2) within import/export state data
> structures to be utilized for coupling.
>
> If anyone remembers things differently, please feel free to comment!
>
> Cecelia
>
>
>
> On Wed, 9 Oct 2002, Erik Kluzek wrote:
>
> > All
> >
> > We have been discussing the proposal for
> > standard-name adoption by ESMF and the creation of an
> > approved list of fieldnames. At the ESMF meeting it was
> > suggested that we extend the CF standard to come up
> > with a list of all the fields we would need in our work.
> > So if CF had a given field, we would use the CF name,
> > but we would also include in this list new fields not
> > part of the CF standard.
> >
> > One of the problems with the above is that if you are
> > using ESMF for something not covered by the CF convention --
> > (say modeling magma flow or space weather) you are hosed.
> > It also requires that ESMF be constantly in the business
> > of maintaining a standard. Maintaining a set of standard-names
> > isn't something that we really want to be doing.
> >
> > So here's a suggestion for an alternative. If the
> > interface allows you to both plug in the standard name
> > AND the standard that is followed -- then you allow more
> > flexibility for usage. You also release standards development
> > to the community. At the same time we can state that CF
> > is the assumed standard for work with ESMF. This way we
> > can make the assumption that most models using ESMF will
> > use the CF standard. So all inter-operability experiments
> > would be done with the CF standard. Brian Eaton also let me
> > know that CF is interested in getting a complete list of
> > fields for CF. So if ESMF finds a need for additional fields
> > they can be readily adopted for the CF standard.
> >
> > This could look something like...
> >
> > SetupImportState( "CF-1.0", "air_pressure" )
> >
> > If you wanted to get real fancy the Framework could
> > "know" about different versions of the CF standard and
> > be able to map names across different versions. But, minimally
> > it could check that both the Import and Export states use
> > the same standard. There obviously would have to be a strict
> > convention about how to name the standard as well (so that
> > if one model called it "cf-1.0" and another "CF1.0" it wouldn't
> > choke).
> >
> > Erik Kluzek, (CGD at NCAR)
> > National Center for Atmospheric Research
> > Boulder CO, (off) (303)497-1326 (fax) (303)497-1324
> > --------- Home page and public PGP key---------------
> > http://www.cgd.ucar.edu/~erik
> > !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> >
> > _______________________________________________
> > Esmf_tech mailing list
> > Esmf_tech at ucar.edu
> > http://mailman.ucar.edu/mailman/listinfo/esmf_tech
> >
>
>
>
> _______________________________________________
> ccsm-progs mailing list
> ccsm-progs at cgd.ucar.edu
>
> For a list of the membership/archive of message and how to subscribe/unsubscrbe...
> http://www.cgd.ucar.edu/mailman/listinfo/ccsm-progs
>
-----------------------------------------------------------------
Brian Eaton | email: eaton at ucar.edu
Climate Modeling Section |
National Center for Atmospheric Research |
P.O. Box 3000, Boulder, CO 80307 |
Received on Thu Oct 10 2002 - 11:56:29 BST
