EOL Arctic Data Connects: A Prototype based on VIVO for Improved Discovery and Access to the Bering Sea Project Archive

The Bering Sea Project was a research program from 2007 through 2012 that sought to understand the impacts of climate change and dynamic sea ice cover on the eastern Bering Sea ecosystem. Over the six-year period of the program hundreds of multidisciplinary datasets coming from a variety of instrumentation and measurement platforms within thirty-one categories of research were processed and curated by the National Center for Atmospheric Research (NCAR) Earth Observing Laboratory (EOL). For the investigator proposing a field project, the researcher performing synthesis, or the modeler seeking data for verification, the easy discovery and access to relevant data is of prime importance. The heterogeneous products of multidisciplinary field programs such as the Bering Sea Project challenge the ability of researchers to identify which datasets, people, or tools might be relevant to their research, and to understand how certain data, instruments, or methods were used to produce particular results.
 
EOL, as a partner in the NSF funded EarthCollab project, is using linked open data to permit the direct interlinking of information and data across platforms and projects. We are leveraging VIVO, the open-source semantic web application, to address connectivity gaps across distributed networks of researchers and resources and identify relevant content, independent of location.  During the Bering Sea Project more than 100 scientists engaged in field data collection, original research, and ecosystem modeling to connect climate, physical oceanography, plankton, fishes, seabirds, marine mammals, humans, traditional knowledge and economic outcomes.  In the first round of development of EOL Arctic Data Connects (http://vivo.eol.ucar.edu) we transformed ISO standard metadata in the EOL production database to RDF triples. In the second round of development we are working on connecting to sources beyond our relational database in order to link hundreds of publications and authors to grants and datasets. The out-of-box VIVO implementation was adapted to model these scientific datasets, utilizing unique identifiers for publications (i.e. publication DOIs), data (i.e. data DOIs) and other connected objects to enable interoperability between resource types, and to yield the most stable and effective connections.
 
The poster will present our approach in connecting ontologies and integrating them within the VIVO system, using the Bering Sea Project datasets as a case study, and will provide insight into how the geosciences can leverage linked data to produce more coherent methods of information and data discovery across large multi-disciplinary projects.