This may be complicating the issue more than resolving it, but how about in
some way applying the ensemble dimension to the CF global attributes for source
and institution, rather than transplanting the information into auxiliary
coordinates?
It could also be done for the global history attribute, which carries a lot of
provenance information. If we're talking about ensemble files with data
compiled from many different sources, it would be nice to be able to
incorporate the history associated with each member of the ensemble, rather
than just having data about the creation of the ensemble itself.
It may not be technically feasible to apply a dimension to a global attribute
per se, but I think something that aims in that general direction would have a
lot of benefits.
--Seth
Seth McGinnis
Associate Scientist
NARCCAP Data Manager
ISSE / IMAGe / NCAR
On Tue, 23 Mar 2010 16:12:13 -0700
John Graybeal <jbgraybeal at mindspring.com> wrote:
>based on the definitions it would be nice to see source_institution (or
>institution_where_produced) and source_method (or source_provenance), in
>order to avoid collisions. just a thought, I know it's not CF-ish to imagine
>far-flung-future problems so feel free to ignore it!
>
>john
>
>On Mar 23, 2010, at 12:11, Jonathan Gregory wrote:
>
>> Dear John
>>
>>> it sounds like a modest proposal is to recommend either of
>>> these attributes to identify the ensemble coordinate variable:
>>>
>>> axis = "ensemble";
>>> standard_name = "realization";
>>>
>>> with the idea of keeping it simple.
>>
>> For auxiliary coordinate variables of the ensemble axis,
>>> other standard names suggested are:
>>>
>>> experiment_id, source, forcing, experiment, model, institution
>>
>> To keep it simple again, I'd vote for just source and institution, which
>> correspond to these existing global attributes of CF:
>> institution
>> Specifies where the original data was produced.
>> source
>> The method of production of the original data. If it was model-
>generated,
>> source should name the model and its version, as specifically as could
>be
>> useful.
>> But that doesn't have to be part of the trac proposal, as these are
> proposals
>> for new standard names. The trac proposal could just say that the ensemble
>> axis could have auxiliary coordinate variables supplying information about
>the
>> ensemble members. When they come from different models, aux coord variables
>> are probably a better way to describe than a coord variable of
> "realization",
>> I suppose.
>>
>_______________________________________________
>CF-metadata mailing list
>CF-metadata at cgd.ucar.edu
>http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
Received on Tue Mar 23 2010 - 17:59:56 GMT
