NASA's Earth Observatory Natural Event Tracker: Curating Metadata for Linking Data and Images to Natural Events

Abstract:

On any given date, there are multiple natural events occurring on our planet. Storms, wildfires, volcanoes and algal blooms can be analyzed and represented using multiple dataset parameters. These parameters, in turn, may be visualized in multiple ways and disseminated via multiple web services. Given these multiple-to-multiple relationships, we already have the makings of a microverse of linked data.

In an attempt to begin putting this microverse to practical use, NASA's Earth Observatory Group has developed a prototype system called the Earth Observatory Natural Event Tracker (EONET). EONET is a metadata-driven service that is exploring digital curation as a means to adding value to the intersection of natural event-related data and existing web service-enabled visualization systems. A curated natural events database maps specific events to topical groups (e.g., storms, fires, volcanoes), from those groups to related web service visualization systems and, eventually, to the source data products themselves.

I will discuss the complexities that arise from attempting to map event types to dataset parameters, and the issues of granularity that come from trying to define exactly what is, and what constrains, a single natural event, particularly in a system where one of the end goals is to provide a group-curated database.