Dear Jonathan,
Thanks for the comments on the units of the variable "leadtime". I'll
change the units to "days" or "months" without any fixed time reference.
Simon Wood's comments are very useful too, as is his example. As I
understand it, the use of "reference time" only requires one of the
variables "forecast period" and "validity time" to identify the forecast
data. I also wondered if Simon meant (ensemble, time, level, latitude,
longitude) to dimension the dataset. At least, that would the case with
my example.
> In your example, I think that experiment_id, source, realization and
> institution should be listed in the coordinates attribute of the data
> variable, since they are auxiliary coordinate variables.
Do you mean something like this?
float geopotential(ensemble, time, level, latitude, longitude) ;
geopotential:coordinates = "reftime leadtime experiment_id source
realization institution" ;
Finally, following Alison's summary posting and in agreement to Jamie's
comment, if the standard name option is not going to be used to describe
the metadata of the forecast systems and "ensembles" as a standard name
issue is already closed, where do we stand now? Do we need to propose a
list of names for the "standard_metadata" attribute? There are no
external vocabularies available in the medium-range, monthly, seasonal
and interannual forecasting communities.
Regards,
Paco
Jonathan Gregory wrote:
> Dear Paco
>
>> What is not so clear to me is whether attributes such as "institution"
>> or "source" or the one mentioned by Jamie "experiment_id" would be
>> allowed as auxiliary variables, "realization" being for me the
>> coordinate variable. Their inclusion in the list of accepted standard
>> names would be the best for me
>
> I think that would be fine, myself. Bryan has a reservation about calling
> them standard names, and I can see that this is a bit different from many
> standard-named quantities, but on the whole I feel we don't need a distinction
> since they are formally similar.
>
>> However, the use of external
>> dictionaries poses certain problems, as discussed.
>
> I would suppose that external dictionaries would be a further optional
> feature. That is, you can supply that attribute to point to a list of
> standardised values for the particular quantity (such as institution), but
> otherwise its values are not standardised. For the sake of self-description,
> the values of the quantity should be as intelligible as possible without
> an external dictionary.
>
> In your example, I think that experiment_id, source, realization and
> institution should be listed in the coordinates attribute of the data
> variable, since they are auxiliary coordinate variables. Also, I would
> suggest that the units of leadtime are just days, not "days since ...",
> because forecast period is an elapsed time, rather than an encoded time/date,
> isn't it?
>
> Best wishes
>
> Jonathan
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata
--
________________________________________
Francisco J. Doblas-Reyes
European Centre for Medium-Range
Weather Forecasting (ECMWF)
Shinfield Park, RG2 9AX
Reading, UK
Tel: +44 (0)118 9499 655
Fax: +44 (0)118 9869 450
f.doblas-reyes at ecmwf.int
_______________________________________
Received on Mon Dec 11 2006 - 08:27:46 GMT
