Design of an efficient distributed GIS application

Abstract

In distributed GIS (geographical information system), all locations that store data are networked together and queries gain their results from any number of locations. The queries, which can be both spatial, and a spatial can be fired from any location in the network and the result is obtained by combining disseminated information across various repositories of data [K. Buchler and L. McKee, 1996]. The proposed model of the distributed GIS will demonstrate many of the important user requirements, design concepts and relevant technologies for real world, remote archives and access capabilities and how these can be integrated into a multi-level user environment. This paper discusses the issues involved in implementing a distributed system that provides a computational infrastructure for the development of decision support and research applications requiring access to and manipulation of large geospatial data sets such as satellite imagery, from remote servers. We give an overview of some of the components of the prototype systems we have developed. The approach used here is particularly useful if the amount of data to be processed is very large (for example multiple channels of a satellite image) but the final result is relatively small (for example a processed satellite image of a localized area or perhaps just a few numbers such as average sea temperature or percentage cloud cover). This project integrates high performance computers and mass data stores to provide useful services to Web based clients for decision support applications including environmental analysis for agricultural planning and land management, analysis of satellite and photoreconnaissance images for defense command and control etc. The end users will thus be able to make the transition from traditional DSSs (decision support systems) based on manual processing of non electronic data such as maps to information age decision systems based on automated or semi automated processing of digital data from large online data archives. There is an emerging need for distributed GIS. The objective of this work is to provide models, methods, tools and frameworks for the development of open distributed geographical information systems and to apply them in at least one test case, to demonstrate and validate the usability and viability of the results. The commercial objective is to decrease the cost of geodata management and further increase the ROI (return on investment) of geodata collection and establishment. The approach is to apply the emerging reference model of open distributed processing standard in combination with existing distributed object technology in the application domain of distributed GIS. The results from this project will be a validation and demonstration of open distributed processing, based on distributed object technology, applied to GIS. The results will be used to increase the share of research and development projects in the area of GIS.