Hi all,
ECMWF produces hindcasts as part of our operational system. These
hindcasts refer to a specific forecast run. For example, today's
forecast has associated a set of hindcasts. Tomorrow's forecast will
have a different set. So our hindcasts have 2 dates:
- the base date of the hindcast run (we call it DATE)
- the base date of the forecast they refer to (we call it REFDATE)
We have discovered that such naming convention was a mistake, because it
requires REFDATE to have the same treatment as DATE. We plan to migrate
to the following terms:
DATE=base date of the forecast they refer to
HINDCASTDATE=base date of the hindcast run
This way, HINDCASTDATE doesn't need to be treated as a DATE, it can just
be yet another attribute.
Hope this is clear and useful.
Manuel
Gross, Tom wrote:
> If we wanted to contain in a single file the results of a model run from
> the past, through the present and into the future, I would use a single
> variable, "time" to give the time coordinate to all time slices, valid
> times. Presumably the application wants to use all of these as one
> continuous run, or they would have been split into different files. The
> extra information of when those different segments were launched should
> be contained in extra variables which represent the times of launching
> the runs. These datums, like "forecast_reference_time" would contain a
> single date. But to introduce three time coordinate variables would be
> a real headache.
> Opps, there I go again argueing that datums are a necessary part of
> metadata, in the same sense as units.
> Tom Gross
>
>
> *From:* ] *On Behalf Of *CJ Beegle-Krause
> *Sent:* Wednesday, October 04, 2006 6:18 PM
> *To:* John Caron
> *Cc:* Ed Hartnett; cf-metadata at cgd.ucar.edu
> *Subject:* Re: [CF-metadata] question and time and standard_name...
>
> There is a need to distinguish whole or parts of model runs, for example
> hindcast, nowcast and forecast could be sections of the same individual
> model run. It would also be good to agree on what each of those terms
> could mean.
>
> John Caron wrote:
>
>> another one is "forecast_reference_time", the base time a model
>> run was made from.
>>
>> Ed Hartnett wrote:
>>> Howdy all!
>>>
>>> The CF document says that standard_name may be used for time
>>> coordinate variables.
>>>
>>> But the standard name table gives only one valid standard_name for a
>>> time variable: "time".
>>>
>>> Can I assume then that there are no other standard names for time
>>> coordinate variables?
>>>
>>> Thanks!
>>>
>>> Ed
>>>
>> _______________________________________________
>> CF-metadata mailing list
>> CF-metadata at cgd.ucar.edu
>> http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata
>
>
> --
> CJ Beegle-Krause, Ph.D.
> NOAA/NOS/ORR/Hazmat
> 7600 Sand Point Way NE
> Seattle, WA 98115
> voice: (206) 526-6961
> fax: (206) 526-6329
>
> <:}}}}}><
> <:}}}}}><
>
> \!/ \!/ \!/ >^<**>^< \!/ \!/ \!/
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata
Received on Thu Oct 05 2006 - 09:06:35 BST
