Challenges and potential solutions for the connection of ESPAS, IUGONET and ISDC

ESPAS and IUGONET are collaborative e-science infrastructure projects in the space weather and adjacent science domains. Whereas the ESPAS project - founded by the European Union (EU) - is bringing together more than 20 EU partner institutions, IUGONET is a Japanese only project based on the collaboration of 5 universities (7 organizations). ISDC is the data management system at GFZ German Research Centre for Geosciences for archiving and distribution of satellite and ground-based geodetic, geophysical and atmospheric data. Major objectives for the organization and realization of data management workflows and appropriate components are almost identical for the 3 systems. Centralized catalogs are filled via harvesting mechanism, and data search and provision is based on categorized metadata. Even though the data models are different, major entities and appropriate relationships are similar and used in the same way. The challenges in the connection of the systems are coming from the different technical realizations of the systems. The IUGONET system is based on DSPACE software using the SPASE data model. ESPAS is using ISO/OGC standards realized by D-NET system components. And for the development of a new ISDC system a semantic Web based approach using the ISDC ontology network consisting of merged domain and terminological ontologies is used. Proposals for the connection of the different systems have to take into account the operational status of the systems. The IUGONET system is up and running, the ESPAS system is in a realization phase, and the new ISDC system is still in a proof of concept phase. Interoperability can be realized step by step in different ways. A promising approach is the common development and use of a harmonized vocabulary for the annotation of data and context data. Standardized SKOS modeled terms derived from a combined thesauri (SPASE, GCMD, ...) can be used to connect identical or similar measurement data, and also data about used instruments, platforms, observations, applications, people, etc. In that way, the quality and quantity of result sets of a search request can be significantly improved. Another approach for the connection of the systems, also based on semantic Web technology, can be realized by the generation and use of semantic relationships of classes and properties related to the different data models after a transformation of XML schemas and appropriate instances into RDF respectively OWL. This approach finally would lead into a distributed but confederate system integrating data, context data and applications in a strong coupled way. Queries from a single entry point covering the different data sources in ESPAS, IUGONET and ISDC as well as in addition - via linked data connected - resources would be possible.