GeoWS: enabling data discovery within the geosciences Abstract


  • Multiple data centers exist within the geosciences community, archiving diverse data types. For instance UNAVCO hosts GPS, strain and soon InSAR and LIDAR data, the Marine Geoscience Data System at LDEO hosts a wide variety of data acquired during oceanographic cruises including seismic reflection, bathymetry, sampling and station data, and the IRIS Data Management Center archives seismic data, gravimeter data, meteorological data, strain data and many other geo-time series data sets along with associated metadata and data products. The increasing demand for tools that enable cross-disciplinary data discovery has spurred these three groups to work together to facilitate data discovery across discipline-specific data centers through adoption of compatible Web services interfaces, called GeoWS. The capabilities of these compatible services are demonstrated in the GeoWS tool. Currently available as a prototype, this tool allows discovery from all three data centers using a simple map interface along with a few fields. Users can search by latitude/longitude ranges; elevation and time range criteria will be added in future releases. Returned information includes data products and instrument locations from all three data centers that fit those search criteria. The underlying web services are all OGC compliant, enabling easy incorporation into community-driven mapping services. The three participating organizations will host mirrored UDDI-compliant service registries to allow discovery of GeoWS compliant web services; any organization will be able to register their services. While initially simple, use of web services enables extension of GeoWS to incorporate search criteria to other types of metadata. The design of GeoWS is purposefully extensible, intended to be a common interface that other data centers can adopt. This will provide an easy access mechanism for data centers to offer data discovery from their peer organizations and provide a means for researchers to discover a broad range of data types from a single application.


publication date

  • 2006

presented at event