Service oriented architecture for geographic information systems supporting real time data grids

Advances in distributed systems and universal adoption of the Internet helped academia, governments and industry to gain access to substantial amount of data, including geospatial data. Geographic Information Systems which are traditionally known to be highly data-centric are one of the important consumers of these data products. However several significant problems such as different types of distributed geophysical applications, and distributed nature of the geospatial data sources cause severe limitations for seamless integration of data and applications. Additionally proliferation of sensors introduces unique challenges for Geographic Information Systems because the observations and measurements obtained from a wide variety of sensors are of great interest to these systems.; In this dissertation we propose and evaluate a Service Oriented Architecture for seamless coupling of real-time and non-real time geographic data with scientific geospatial applications. We describe necessary principles and service features for accessing both data types and outline Data Grid architecture to address these requirements. We describe various methods for improving the performance of our Web Services such as adoption of Binary XML Frameworks and integration of a message oriented middleware for streaming the responses. We introduce a novel approach to map Geography Markup Language (GML) based XML queries to relational database queries. We evaluate the performance and usefulness of our approach in several real-world applications such as earthquake science tools and real-time GPS streams.