 |
| Titel |
Concept for Future Data Services at the Long-Term Archive of WDCC combining DOIs with common PIDs |
| VerfasserIn |
Martina Stockhause, Tobias Weigel, Frank Toussaint, Heinke Höck, Hannes Thiemann, Michael Lautenschlager |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250075550
|
|
|
|
|
|
| Zusammenfassung |
| The World Data Center for Climate (WDCC) hosted at the German Climate Computing
Center (DKRZ) maintains a long-term archive (LTA) of climate model data as well as
observational data. WDCC distinguishes between two types of LTA data:
Structured data: Data output of an instrument or of a climate model run consists
of numerous, highly structured individual datasets in a uniform format. Part of
these data is also published on an ESGF (Earth System Grid Federation) data
node. Detailed metadata is available allowing for fine-grained user-defined data
access.
Unstructured data: LTA data of finished scientific projects are in general
unstructured and consist of datasets of different formats, different sizes, and
different contents. For these data compact metadata is available as content
information.
The structured data is suitable for WDCC’s DataCite DOI process, the project data only in
exceptional cases. The DOI process includes a thorough quality control process of technical
as well as scientific aspects by the publication agent and the data creator. DOIs
are assigned to data collections appropriate to be cited in scientific publications,
like a simulation run. The data collection is defined in agreement with the data
creator.
At the moment there is no possibility to identify and cite individual datasets within this
DOI data collection analogous to the citation of chapters in a book. Also missing is a compact
citation regulation for a user-specified collection of data.
WDCC therefore complements its existing LTA/DOI concept by Persistent Identifier
(PID) assignment to datasets using Handles. In addition to data identification for internal and
external use, the concept of PIDs allows to define relations among PIDs. Such structural
information is stored as key-value pair directly in the handles. Thus, relations provide basic
provenance or lineage information, even if part of the data like intermediate results are lost.
WDCC intends to use additional PIDs on metadata entities with a relation to the
data PID(s). These add background information on the data creation process (e.g.
descriptions of experiment, model, model set-up, and platform for the model run etc.) to
the data. These pieces of additional information increase the re-usability of the
archived model data, significantly. Other valuable additional information for scientific
collaboration could be added by the same mechanism, like quality information and
annotations.
Apart from relations among data and metadata entities, PIDs on collections are
advantageous for model data:
Collections allow for persistent references to single datasets or subsets of data
assigned a DOI,
Data objects and additional information objects can be consistently connected
via relations (provenance, creation, quality information for data), |
|
|
|
|
|
|