Over the past decade, desktop image analysis and geographic information system (GIS) software have matured into the dominant tool for geoprocessing. Desktop solutions usually incorporate a broad range of geospatial processing functionality together with locally maintained data in order to provide a tightly coupled and largely autonomous environment for carrying out operational geospatial activity. This approach has also been adopted by the research community. Researchers typically bring together at the local level both the necessary analytical tools and the data for the research activity at hand.

There is now, however, a move away from such desktop technology. A loosely coupled service-oriented architecture, based on the deployment of Web services developed and maintained by a dispersed community, is now seen as a more powerful and flexible approach. Likewise, with this approach, the data may also be distributed and accessed directly from databases maintained by the collection agencies rather than being duplicated at the site where analysis is being undertaken. This new approach, based on established and emerging standards for geospatial interoperability, has many advantages, which are discussed herein.

This chapter describes the challenges associated with the more dispersed and collaborative nature of the operational and research programs that are based on such an architectural approach. The need for and benefits of a persistent interoperability test bed for geosciences research and education are discussed, as is the question of ways in which to facilitate the move of such an architectural approach into routine operational use.