Implementing a Federated Data Archive with Asynchronous Data Query, Gathering and Analysis Capabilities

Software architectures for data archives must be easily extensible to meet the requirements of new space science missions of increasing complexity and scope. Simultaneously, they must ensure that data is discoverable and easily usable by the research community. They must also respect the independence and rights of data providers, who may have digital rights management issues to enforce. The Canadian Space Science Data Portal (CSSDP) is facing these challenges by adopting a federated, portal-based, virtual observatory model, founded on Space Physics Archive Search and Extract (SPASE) metadata. The approach allows for the separation of the search, discovery and retrieval of pertinent data from its actual storage location, and the insulation of the researcher from the technical details of such data search, discovery, retrieval and usage. CSSDP implements a federation approach that allows data storage locations and methods to evolve separately from data access, and for data analysis capabilities to be offered through a common interface. This allows further independence of (and changes to) any archive in the federation, which is achieved without any change to CSSDP itself and without any impact on individual researchers. As CSSDP has successfully addressed core data storage and access challenges, it is now moving toward supporting improved data manipulation and analysis. CSSDP currently supports researcher collaboration and researcher "tagging" of relevant data, along with basic analytics; in the future, CSSDP will also allow researchers to insert scientific workflows and data analysis tools into their workspace, This paper will discuss how relevant features were implemented for CSSDP, the knowledge gained, the value delivered to the space science research community, and how to position data archives to achieve the same result in related disciplines.