Remote thin-client access to spatial database systems

Numerous federal agencies produce official statistics that are made accessible to ordinary citizens for searching and data retrieval. This is often done via the Internet through a web browser interface. If this data is presented in textual format, it can often be searched and retrieved by such attributes as topic, responsible agency, keywords, or press release. However, if the data is of spatial nature, e.g., in the form of a map, then using text-based queries is often too cumbersome for the intended audience. We propose to use the capabilities of the SAND Spatial Browser to provide more power to users of these databases. Using the SAND Spatial Browser allows users to define the spatial region of interest with greater specificity, instead of forcing them to retrieve data just for a particular location or a region with a predefined boundary. They can also make use of ranking which is the ability to retrieve data in the order of distance from other instances of the data or aggregates of data that are user-defined. Work is distributed between the SAND server and the individual clients for query evaluation, data visualization and data management. This enables the minimization of the necessary requirements for system resources on the client side while maximizing the number of connections one server can handle concurrently. Concrete experience with interfacing the SAND system with FedStats data is also discussed.