Data Services Newsletter

Volume 8 : No 3 : September 2006

GeoWS: A Partnership for Interoperable Data Exchange

The IRIS DMC is initiating development of a web service architecture to provide state of the art, programmatic components for client-server data access. This coincides with other organizations’ efforts all across the scientific landscape that have set into motion initiatives to make their unique data repositories available in a standard way such that any scientist, educator, or student can have a wealth of data discovery at their fingertips, online and on demand. This ‘dream’ of unified online data availability is referred to as ‘cyberinfrastructure’ (CI) for science and engineering. IRIS is working to be a part of this larger effort.

As web service planning at IRIS began to take shape, an opportunity was realized among cooperative partner organizations that a team should be formed to build a unified method of data interchange between data centers. Members of UNAVCO, Lamont-Doherty Marine Geology and Geophysics (MG & G), and IRIS gathered in August to discuss the prospects of being able to view data from all three data management systems simultaneously. Agreeing on a method for data exchange via web services, the group quickly put together a set of desired functionality and sketched out an implementation plan. At the end of the meeting, we agreed upon a name for our team and the project: GeoWS.

GeoWS logo
Figure 1: GeoWS logo

In short, the idea is to be able to present an interface that allows the user to draw upon data ‘layers’ from all three organizations’ data centers. Lamont will provide ocean cruise track, dive track, and core sample information. UNAVCO will provide GPS station locations and data points. Finally, IRIS will present seismic station, earthquake, and data product information. This represents just the beginning of what each of these organizations can ultimately provide to the GeoWS project.

To tie all these data together, a client program will be used in the form of a mapping interface that can pull back a reference map from a freely available web mapping service (WMS) and facilitate latitude and longitude selections from the map itself. The GeoWS data providers will then independently make available online web services known as ‘web feature services’ (WFS). A ‘feature’ is a data point on the map that indicates location and basic identification as well as a link to more information if the user selects the item. The WFS can provide a series of these feature points, which are then overlaid on a base map as a ‘layer’. Users can selectively turn data layers on and off so they only see the information that they want to.

GeoWS contributors
Figure 2: GeoWS contributors

The interaction described above represents the first step in realizing data interoperability. When success in this endeavor is achieved and can be demonstrated, the group can then address more complex features and draw upon other sources of data. The key to the success of this venture is that the group is small, making coordination manageable, and the adoption of web service standards, which helps to overcome the problems gathering different types of data from disparate data systems.

This cooperative effort benefits everyone involved. The three organizations of GeoWS now have a springboard for providing web service features in the form of a highly visible representation of cross-disciplinary data interoperability. Proper leveraging of emerging web service standards as well as use of existing web service resources will allow each of our data centers to propel forward into a new realm of data accessibility. GeoWS is the road we are paving to reach out to the cyberinfrastructure community and to all the users who will benefit from it.

GeoWS participants

  • IRIS DMC
    • Tim Ahern
    • Rob Casey
    • Mari Francissen
    • Linus Kamb
    • Joanna Muench
    • Bruce Weertman
  • MG&G
    • Bob Arko
    • Suzanne Carbotte
  • UNAVCO
    • Greg Anderson
    • Fran Boler
    • Ben Hoyt
    • James Matykiewicz
    • Chris Stolte

by Rob Casey (IRIS Data Management Center)

Page built 19:58:18 | v.7276e3af