[Go-essp-tech] syntax on the DRS ID

[Sylvia Murphy]

Hi Bob,
The goal here is to connect a model metadata simulation instance to the data COLLECTION within ESG.  As such, we will need to parse this ID to exclude product, time_frequency, and cmor_table, which I believe are at the atomic dataset level.

Metafor reported today they plan to try and put this ID into the output of the questionnaire.  I am cc'ing them on this email so they can see the syntax and can chime in about the granularity of information.  
Thanks,
Sylvia
On Jan 28, 2011, at 12:42 PM, Robert S. Drach wrote:

> Hi Sylvia,
> 
> The DRS publication-level dataset ID - the IDs you will have available to match against - is defined in the DRS document:
> 
> http://cmip-pcmdi.llnl.gov/cmip5/docs/cmip5_data_reference_syntax.pdf
> 
> See section 3.4: Publication-level dataset_id encoding. The form of the ID is:
> 
> cmip5.product.institute.model.experiment.time_frequency.realm.cmor_table.ensemble
> 
> Please note there are a few caveats at this point:
> 
> - Although the CMIP5 best practice recommendation is to follow the DRS specification for dataset IDs, it is possible that the drs_id attribute may differ, if the person doing the publisher configures esgpublish to do so. In contrast the drs_id form is hardwired into the esgpublish code. The question then is which one to match against if they differ. (I would suggest the drs_id, but this remains open for discussion).
> - Determination of the product field is relatively complicated in some cases. In some cases you may need file-level information such as time ranges to generate the product.
> - Depending on  the  particular requirements for matching  metafor information to  datasets, it might make sense to  just match individual fields. All the above fields (product, institute, etc.) are defined individually in the dataset metadata on the gateway.
> 
> --Bob
> 
> Sylvia Murphy wrote:
>> Hi Karl,
>> I have been talking to the metafor group about the DRS_ID you mentioned on the call this week.  They don't think they are going to have it output directly in their file but I may be able to put it together from various pieces.  I wanted to double check with you on what pieces are going into the ID since it may be different from my recollections.  At one point the syntax was going to be:
>> 
>> institution_model_simulationname_ripvalue + other stuff 
>> is that still valid?  The "other stuff" (e.g. variable names) will be more granularity than I'll need.  
>> If you have some examples of actual ids that you have seen generated by data providers that would help as well.
>> 
>> Thanks,
>> Sylvia
>> ***********************************
>> Sylvia Murphy
>> NESII/CIRES/NOAA Earth System Research Laboratory
>> 325 Broadway, Boulder CO 80305
>> Email: sylvia.murphy at noaa.gov <mailto:sylvia.murphy at noaa.gov>
>> Phone: 303-497-7753
>> Fax: 303-497-7649
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> ------------------------------------------------------------------------
>> 
>> _______________________________________________
>> GO-ESSP-TECH mailing list
>> GO-ESSP-TECH at ucar.edu
>> http://mailman.ucar.edu/mailman/listinfo/go-essp-tech
>>  
> 

***********************************
Sylvia Murphy
NESII/CIRES/NOAA Earth System Research Laboratory
325 Broadway, Boulder CO 80305
Email: sylvia.murphy at noaa.gov
Phone: 303-497-7753
Fax: 303-497-7649