[ESIP-AQ] granularity for GCI?
Stefano Nativi
stefano.nativi at cnr.it
Tue Jan 31 03:27:04 EST 2012
Dear Martin,
In my opinion you raised a good point.
As for an infrastructure full of data, I would suggest to access the
GEOSS Common Infrastructure which had million of resources (not only
datasets !).
If you are interested you may use either the GEO portal
(http://www.geoportal.org/web/guest/geo_home) or the broker interface,
if you are also interested in semantic discovery
(http://www.eurogeoss-broker.eu/).
Best regards,
Stefano
> Dear all,
>
> granularity is indeed an important discussion. Have you checked GENESI-DEC (http://portal.genesi-dec.eu/search/) as one implementation that is already populated with a lot of data? It may be that I am too stupid to operate that tool correctly (or I don't take the time to read the documentation carefully), but my (little) experience with this project is that you either get too little data or far too much. In my view the right level of granularity should really be discussed on a broad international level, because this is the historic chance to link data sets together, but if we mess up by making it close to impossible to find them (for example due to granularity variations), then a lot of damage will be done as it will become impossible to convince people of the value of interoperability.
>
> This was the political answer. My technical 2 cents are that finer granularity will always give you more flexibility, but then the burden of organizing the data lies with the search engine/web application. This would then play the ball back into the field of the semantic metadata descriptions, where we should continue the harmonization process which was installed in the AQ community catalogue and re-discussed at the Solta meeting. Do we find a volunteer to lead this discussion?
>
> Best regards,
>
> Martin
>
>
>
>> -----Original Message-----
>> From: esip-aqcluster-bounces at lists.esipfed.org [mailto:esip-aqcluster-
>> bounces at lists.esipfed.org] On Behalf Of Yang, Wenli (GSFC-610.2)[GEORGE
>> MASON UNIVERSITY]
>> Sent: Monday, January 30, 2012 11:51 PM
>> To: Lynnes, Christopher S. (GSFC-6102); esip-aqcluster at rtpnet.org
>> Cc: Johnson, James E. (GSFC-610.2)[R S INFORMATION SYSTEMS INC]; Ted
>> Habermann; Zhao, Peisheng (GSFC-610.0)[GEORGE MASON UNIVERSITY]
>> Subject: Re: [ESIP-AQ] granularity for GCI?
>>
>> A coverage itself is very flexible, as is the ISO dataset. A coverage (dataset)
>> can include any number of "related" variables. Theoretically the "related"
>> variables can be spread in different physical files although we perhaps don't
>> want to advertise multiple variables from multiple data files in one coverage.
>>
>> Currently the limitation is in WCS version 1.0, which limits one variable in one
>> coverage. For other services or for higher versions of WCS, multiple variables
>> can be advertised (actually for ftp service, one has to get all variables). For
>> metadata catalog, the ISO scheme provides very flexible way. Both
>> data_series and dataset concepts are flexible, i.e., the ISO metadata scheme
>> can deal with single/multiple variables, single/multiple datasets, and single
>> multiple data_series.
>>
>> -----Original Message-----
>> From: esip-aqcluster-bounces at lists.esipfed.org [mailto:esip-aqcluster-
>> bounces at lists.esipfed.org] On Behalf Of Lynnes, Christopher S. (GSFC-6102)
>> Sent: Monday, January 30, 2012 5:05 PM
>> To: esip-aqcluster at rtpnet.org
>> Cc: Zhao, Peisheng (GSFC-610.0)[GEORGE MASON UNIVERSITY]; Ted
>> Habermann; Johnson, James E. (GSFC-610.2)[R S INFORMATION SYSTEMS
>> INC]
>> Subject: Re: [ESIP-AQ] granularity for GCI?
>>
>> My concern with the registration of individual coverages in the GCI is that this
>> approach works fine in the early, heady days of prototyping, but does not
>> scale well once we get serious about populating it. With everything. The
>> database may be able to handle it, but the user interfaces are then left to
>> deal with the results, which comprise many, many records of vastly different
>> scale/aggregation, from entire systems to individual variables.
>>
>> However, having registered that concern, if the community consensus is to
>> go that way, then we will proceed with all due expedience to start populating
>> in that vein.
>>
>> On Jan 30, 2012, at 4:44 PM, Zhao, Peisheng (GSFC-610.0)[GEORGE MASON
>> UNIVERSITY] wrote:
>>
>>
>>> It is noted the WCS 1.0 restricts the range to a single variable, i.e. one
>>>
>> coverage only includes one variable. The WCS 1.1 or later remove this
>> limitation.
>>
>>> From: Bagwell, Ross E. (GSFC-580.0)[COLUMBUS TECHNOLOGIES AND
>>>
>> SERVICES
>>
>>> INC]
>>> Sent: Monday, January 30, 2012 4:09 PM
>>> To: Erin Robinson; Lynnes, Christopher S. (GSFC-6102)
>>> Cc: Zhao, Peisheng (GSFC-610.0)[GEORGE MASON UNIVERSITY]; Johnson,
>>> James E. (GSFC-610.2)[R S INFORMATION SYSTEMS INC];
>>> esip-aqcluster at rtpnet.org; Ted Habermann
>>> Subject: RE: [ESIP-AQ] granularity for GCI?
>>>
>>> Erin,
>>>
>>> I agree with classifying coverages as datasets, as the terminology is more in
>>>
>> line with the geospatial world. A coverage is not an attribute as much as it is a
>> prescribed set of data, having a number of feature attribute tables with one-
>> to-one related records.
>>
>>> Ross
>>>
>>>
>>>
>>> From: esip-aqcluster-bounces at lists.esipfed.org
>>> [mailto:esip-aqcluster-bounces at lists.esipfed.org] On Behalf Of Erin
>>> Robinson
>>> Sent: Monday, January 30, 2012 4:01 PM
>>> To: Lynnes, Christopher S. (GSFC-6102)
>>> Cc: Zhao, Peisheng (GSFC-610.0)[GEORGE MASON UNIVERSITY]; Johnson,
>>> James E. (GSFC-610.2)[R S INFORMATION SYSTEMS INC];
>>> esip-aqcluster at rtpnet.org; Ted Habermann
>>> Subject: Re: [ESIP-AQ] granularity for GCI?
>>>
>>> Hi Stefan, All -
>>>
>>> The way we left things with the AQ Community Catalog was that individual
>>>
>> WCS coverages were registered. In ISO 19115 this was hierarchy
>> level=attribute. The scope codes are described here on the NOAA ISO wiki.
>> Each coverage level record has a service identification object that describes
>> the WCS service and how to access. Theoretically a smart search client could
>> pull either the dataset or the service.
>>
>>> Since the ESIP Winter Meeting I have gone back to the AQ Community ISO
>>>
>> record to make some updates and bring it more inline with changes that have
>> occurred within ISO in the last 18 months or so. There are several things that
>> the community needs to discuss. The granularity of the records should go on
>> this list. I think classifying coverages as attributes may not be the right thing -
>> maybe coverages should be datasets? Would be a good addition to the ESIP
>> AQ agenda to discuss the record, in general.
>>
>>> Erin
>>>
>>> Erin Robinson
>>> Information and Virtual Community Director Foundation for Earth
>>> Science | 314.369.9954 | erinrobinson at esipfed.org www.esipfed.org
>>>
>>>
>>> On Mon, Jan 30, 2012 at 3:23 PM, Lynnes, Christopher S. (GSFC-6102)
>>>
>> <christopher.s.lynnes at nasa.gov> wrote:
>>
>>> On Jan 30, 2012, at 3:15 PM, Stefan Falke wrote:
>>>
>>>
>>>> Chris,
>>>>
>>>> I don't think we did finalize the approach to submitting/registering
>>>>
>> services with GCI. It may be good to expand the question to include other
>> catalogs/registries to understand how they work. For example, at what level
>> are resources registered in GCMD?
>>
>>> Data collections aka Dataset_Series in ISO-speak.
>>>
>>>
>>>> The AQ Community Catalog could be another example for defining the
>>>>
>> approach to registering data services. I don't recall whether items were
>> registered at the WCS level or at the specific individual coverages level. I've
>> copied Rudy and Erin to solicit their thoughts.
>>
>>>> The Community Catalog brings up another dimension to registering
>>>>
>> resources into GCI because the AQ Community Catalog itself was registered
>> in the GCI - not its individual services directly. The GCI Clearinghouse was
>> able to read the Web Accessible Folder of the AQ Community Catalog to get
>> the individual services. But I think a question was how well the services and
>> their metadata were captured in the GCI so that a user could do the type of
>> 'deep' queries needed to get the specific services identified in the return of a
>> search.
>>
>>>> In any case, another great reason for picking up the metadata group
>>>>
>> discussions. I talked with Glynis Lough (copied) last week and she agreed that
>> metadata makes for an excellent focus topic for re-energizing the ESIP AQ
>> Workgroup discussions. Maybe this could be incorporated in the next
>> agenda.
>>
>>>> Stefan
>>>>
>>>> On Mon, Jan 30, 2012 at 1:36 PM, Lynnes, Christopher S. (GSFC-6102)
>>>>
>> <christopher.s.lynnes at nasa.gov> wrote:
>>
>>>> Stefan,
>>>> Did we ever come to an agreement on what granularity to submit ACP
>>>>
>> resources to GCI? One item for each coverage would be a bit much; one for
>> the whole ACP could not give much detail on what's available. Is there
>> something in the middle?
>>
>>>> --
>>>> Dr. Christopher Lynnes NASA/GSFC, Code 610.2 phone: 301-614-5185
>>>>
>>>>
>>>>
>>>>
>>> --
>>> Dr. Christopher Lynnes NASA/GSFC, Code 610.2 phone: 301-614-5185
>>>
>>>
>>>
>> --
>> Dr. Christopher Lynnes NASA/GSFC, Code 610.2 phone: 301-614-5185
>>
>>
>> _______________________________________________
>> ESIP-AQcluster mailing list
>> ESIP-AQcluster at lists.esipfed.org
>> http://www.lists.esipfed.org/mailman/listinfo/esip-aqcluster
>> _______________________________________________
>> ESIP-AQcluster mailing list
>> ESIP-AQcluster at lists.esipfed.org
>> http://www.lists.esipfed.org/mailman/listinfo/esip-aqcluster
>>
>
> -------------------------------------------------------------------------------
> -------------------------------------------------------------------------------
> Forschungszentrum Juelich GmbH
> 52425 Juelich
> Sitz der Gesellschaft: Juelich
> Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
> Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
> Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender),
> Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
> Prof. Dr. Sebastian M. Schmidt
> -------------------------------------------------------------------------------
> -------------------------------------------------------------------------------
>
> Kennen Sie schon unsere app? http://www.fz-juelich.de/app
> _______________________________________________
> ESIP-AQcluster mailing list
> ESIP-AQcluster at lists.esipfed.org
> http://www.lists.esipfed.org/mailman/listinfo/esip-aqcluster
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.esipfed.org/pipermail/esip-aqcluster/attachments/20120131/102b1b42/attachment.html>
More information about the ESIP-AQcluster
mailing list