Data Model

The GCIS maintains two concurrent models of the data : a relational model, and a semantic model. Entities represented by the GCIS are refered to as resources, as defined in
RFC 2396. They are assigned Uniform Resource Identifiers, or URIs.

Resources

The terminology below describes the resources represented in the GCIS :

Report

A report, such as the nca3, or the nca2, may have chapters, figures, tables, findings, and references associated with it.

Chapter

A chapter of a report, such as our-changing-climate, has a unique mnemonic identifier, and possibly, a number. Some chapters, for instance an appendix, do not have a number. Chapters may also have figures, tables,, findings, and references associated with them.

Figure

A figure in a report, such as observed-us-temperature-change, may be composed of one or more images.

Finding

A finding in a report, such as global-climate-is-changing, has a statement as well as phrases and bibliographic references describing the confidence level, uncertainties, and evidence for that finding.

Table

A table, like a figure, is embedded in a report, and possibly in a chapter in a report.

Reference

A reference is a bibliographic entry or citation. It is uniquely identified with a UUID, like e679d754-46b3-4d62-a7dd-4a7f0c727ebe. A reference in a report may appear in a chapter, figure, finding or table of that report. Moreover, a reference refers to a publication such as an article, web page, or book.

Publication

A publication in the GCIS may be a report, a report-specific resource -- a chapter, figure, finding, table -- or a non-report-specific resource : a journal, article, image, web page, book, or dataset.

Image

An image may be associated with one or more figures. For instance, this image is a part of the figure described above.

Book

A book has an ISBN, but is uniquely identified within GCIS using a UUID.

Journals

A journal may have a print and online ISSN, and is uniquely identified within GCIS using a mnemonic identifier, like climatic-change .

Article

An article in a journal has a DOI (if available), like 10.1002/grl.50527, which is used to identify it uniquely in the GCIS.

Web Page

A web page is assigned a UUID, like 26625ddf-dd19-4dd1-a35d-33c68c5b2d6e, to identify its state at a particular point in time.

Contributor

A contributor to a publication is an organization, a role, and optionally a person.

Organization

An organization may be related to other organizations, and is identified uniquely by a mnemonic identifier.

Person

A person is given a numeric id. If a person has an orcid, it is used to uniquely identify the person within the GCIS. The only associations between people and organizations is via their affiliations when contributing to GCIS publications.

Dataset

A dataset is given a unique mnemonic identifier, such as nca3-ncep-ncar-r1. Datasets are considered a type of publication, as mentioned above. As such, they may be generated by an activity.

Platform

A platform, such as jason-1 may be associated with one more or instruments.

Instrument

An instrument, such as poseidon-2 may be associated with one or more platforms via instrument instances.

Instrument Instance

An instrument instance, such as jason-1/poseidon-2 is a platform/instrument combination, and may be associated with one or more datasets.

Activity

An activity, such as 063fd83f-maurer-process, is used to describe the process of going from one publication to another. See below.

Project

A project, such as cmip3 may be associated with many models.

Model

A model, such as ccsm3 is used for a model run.

Scenario

A scenario, such as sres_b1 is also used for a model run.

Model Run

A model run is uniquely determined by a model, scenario, time range, time resolution and sequence number. Model runs may be associated with activities.

Identifiers

Resources in GCIS are identified by URIs. These URIs have been designed to be Cool URIs for the Semantic Web. The URIs may contain mnemonic identifiers, which are ASCII strings composed of lower case letters, numbers, underscores and dashes. UUIDs are sometimes used, as are Digital Object Identifiers (DOIs), when they exist. DOIs are standardized as ISO 26324.

GCIS identifiers (GCIDs) may be composed of identifiers for individual resources. For instance, the chapter identifier, our-changing-climate, is part of the GCID /report/nca3/chapter/our-changing-climate. Because of this, resource identifiers for chapters, figures, findings, and tables may not be globally unique.

Relational Model

The relational model used by GCIS captures one-to-many, many-to-many, and many-to-one relationships between the resources above. Journals have many articles, reports have many figures, findings, tables, and chapters. The relationship between images and figures is many to many, as is the relationship between publications (of any type) and contributors. Publications may also be related to each other, and in this case, the relationship between two publications can be annoted with a term from a semantic vocabulary. Furthermore, the relationship may have an activity associated with it.

Semantic Model

All GCIS resources have representations in turtle. The GCIS vocabulary is defined in the GCIS ontology. Many other ontologies are used including, most notably PROV. The entity-activity-agent model of PROV has been applied to the GCIS through the use of resources, activities, and contributors.