[Esip-documentation] Related data files in NetCDF
Nan Galbraith
ngalbraith at whoi.edu
Tue Jun 12 15:19:53 EDT 2018
Thanks for the replies! The DataCite relationType list is a little ...
robust ... for this
application, but we could use IsSupplementTo; the ISO codeList also
seems OK for our
purposes, too.
We just want to provide a link to data that can be downloaded elsewhere,
whose
relationship to the NetCDF file is (usually) just that it was collected
at the same time
and place. I suppose that if the related data is in a NetCDF file on
line, too, that will
be obvious from the URL.
I'll take another look at the esip wiki page - lots of info there.
Thanks again -
Nan
On 6/7/18 4:15 PM, Ted Habermann wrote:
> Nan,
>
> An important element of the DIF model is the pair of Type and Subtype that come from lists managed by GCMD (which is now part of CMR). That could be good or bad.
>
> This model is very similar to the ISO model for onlineResource shown below. Most of these elements are optional. This model adds a title to the descriptive material (nice) and also specifies the codeList location which allows you to specify the codeList you are using.
>
> <cit:onlineResource>
> <cit:CI_OnlineResource>
> <cit:linkage>
> <gco:CharacterString>URL</gco:CharacterString>
> </cit:linkage>
> <cit:protocol>
> <gco:CharacterString>usually http</gco:CharacterString>
> </cit:protocol>
> <cit:applicationProfile>
> <gco:CharacterString>description of applications that might use the resource</gco:CharacterString>
> </cit:applicationProfile>
> <cit:name>
> <gco:CharacterString>title</gco:CharacterString>
> </cit:name>
> <cit:description>
> <gco:CharacterString>description</gco:CharacterString>
> </cit:description>
> <cit:function>
> <cit:CI_OnLineFunctionCode codeList="http://www.isotc211.org/2005/resources/Codelist/gmxCodelists.xml#CI_OnLineFunctionCode" codeListValue="download">download</cit:CI_OnLineFunctionCode>
> </cit:function>
> </cit:CI_OnlineResource>
>
> The is very short (download, information, offlineAccess, order, search, completeMetadata, browseGraphic, upload, emailService, browsing, fileAccess) and not well designed for describing relations.
>
> I think the DataCite relationType list might be a good one for this application: IsCitedBy, Cites, IsSupplementTo, IsSupplementedBy, IsContinuedBy, Continues, IsNewVersionOf, IsPreviousVersionOf, IsPartOf, HasPart, IsReferencedBy, References, IsDocumentedBy, Documents, IsCompiledBy, Compiles, IsVariantFormOf, IsOriginalFormOf, IsIdenticalTo, HasMetadata, IsMetadataFor, Reviews, IsReviewedBy, IsDerivedFrom, IsSourceOf, IsDescribedBy, Describes, IsVersionOf, HasVersion, IsRequiredBy, Requires.
>
> This list is used in DataCite as an attribute in the relatedIdentifier which serves a similar function as what you want but is focused on DOI's that have defined URLs for each identifier.
>
> There is a bit more information on this general topic at http://wiki.esipfed.org/index.php/Online_Resource_Documentation
>
> Ted
>
> ==== Ted Habermann ========================
> Director of Earth Science, The HDF Group
> On 6/7/18, 12:09 PM, "Esip-documentation on behalf of Nan Galbraith via Esip-documentation" <esip-documentation-bounces at lists.esipfed.org on behalf of esip-documentation at lists.esipfed.org> wrote:
>
> The OceanSITES data management team would like to formalize a
> method to include pointers to 'related data' files in our NetCDF files.
> We are serving long-term, open ocean data sets, and for some of these
> there may be deployment/recovery cruise data, short term (single
> deployment) data from 'guest' instruments, or ... some other kinds of
> data that aren't in our collection but are available elsewhere.
>
> The GCMD DIF has a Related_URL concept that would probably
> work for us:
>
> <Related_URL>
> <URL_Content_Type>
> <Type>Related URL Type keyword</Type>
> <Subtype>Related URL Subtype keyword</Subtype>
> </URL_Content_Type>
> <URL>URL up to 600 characters</URL>
> <Description>Text (may be multi-line)</Description>
> </Related_URL>
>
> I'm wondering if there are other standards that might also work, and if
> there are standard ways to express something like this in NetCDF. I
> was just looking at some ways to group attributes, but I haven't found
> anything that's widely accepted.
>
> In the ACDD, we 'references' is one of the suggested attributes, taken
> from CF; however it doesn't have the descriptive fields that the DIF
> provides.
>
> Any suggestions would be appreciated!
>
> Cheers - Nan
>
>
>
--
*******************************************************
* Nan Galbraith Information Systems Specialist *
* Upper Ocean Processes Group Mail Stop 29 *
* Woods Hole Oceanographic Institution *
* Woods Hole, MA 02543 (508) 289-2444 *
*******************************************************
More information about the Esip-documentation
mailing list