A Semantically Enabled Metadata Repository for Solar Irradiance Data Products

The Laboratory for Atmospheric and Space Physics, LASP, has been conducting research in Atmospheric and Space science for over 60 years, and providing the associated data products to the public. LASP has a long history, in particular, of making space-based measurements of the solar irradiance, which serves as crucial input to several areas of scientific research, including solar-terrestrial interactions, atmospheric, and climate. LISIRD, the LASP Interactive Solar Irradiance Data Center, serves these datasets to the public, including solar spectral irradiance (SSI) and total solar irradiance (TSI) data.

The LASP extended metadata repository, LEMR, is a database of information about the datasets served by LASP, such as parameters, uncertainties, temporal and spectral ranges, current version, alerts, etc. It serves as the definitive, single source of truth for that information. The database is populated with information garnered via web forms and automated processes. Dataset owners keep the information current and verified for datasets under their purview.

This information can be pulled dynamically for many purposes. Web sites such as LISIRD can include this information in web page content as it is rendered, ensuring users get current, accurate information. It can also be pulled to create metadata records in various metadata formats, such as SPASE (for heliophysics) and ISO 19115. Once these records are be made available to the appropriate registries, our data will be discoverable by users coming in via those organizations.

The database is implemented as a RDF triplestore, a collection of instances of subject-object-predicate data entities identifiable with a URI. This capability coupled with SPARQL over HTTP read access enables semantic queries over the repository contents.

To create the repository we leveraged VIVO, an open source semantic web application, to manage and create new ontologies and populate repository content. A variety of ontologies were used in creating the triplestore, including ontologies that came with VIVO such as FOAF. Also, the W3C DCAT ontology was integrated and extended to describe properties of our data products that we needed to capture, such as spectral range.

The presentation will describe the architecture, ontology issues, and tools used to create LEMR and plans for its evolution.