VANILLA: a dynamic data schema for a generic scientific database

Scientists from widely varying communities are frequently called upon to work together, sharing their data and their expertise to investigate a common issue. Difficulties frequently arise in sharing data because each community, and sometimes each scientist, has their own conventions for structuring the data. This results in a data schema that is incompatible with the other scientific communities working on the investigation. The paper presents VANILLA, a generalized data schema developed to address these issues, whose prototype shares data among the varied communities of forest canopy science. VANILLA uses techniques from semantic and dimensional databases in a federated approach to manage the sociological and technological issues in scientific data integration. All data is organized according to a thesaurus so that new data schemas can be rapidly built incorporating terminology and semantics of new scientific domains or methods. This data schema has been successfully used to store and analyze micro meteorological forest data and stem map data.