A SOA-based Decision Support Geographic Information System for Storm Disaster Assessment

SOA (Service-Oriented Architecture) is an architecture of software design suitable for changing environment, in which the difficulty in the sharing of data and operations. In this paper a SOA-based decision support system (DSS) was built for assessing the short-term risk of storm disaster at coastal area of Shanghai. This was done by integrating a hydrodynamic model and a disaster assessment model into geographical information system (ArcGIS Server 9.3) as a web service using ArcObjects in ArcGIS and Java. Storm flood was simulated by hydrodynamic model at the Yangtze estuary. Population, house area, important units and submerge water depth were considered as impact factors to build storm-tide evaluation index system, with which the influence of storm flood was acquired by quantitative evaluation model using fuzzy method. Geographic information system was investigated to show storm flood process and disaster assessment results in two and three dimension map. According to the assessment results, Optimization Algorithm was employed to calculate afflicted population assigned to available refugee settlements. And optimal retreat routes were provided by Shortest Path Algorithm according to the traffic information of Shanghai. To meet the business demands, the two algorithms were published as web services and integrated into DSS through web services composition(WSC). With the support for 3D geospatial visualization capabilities of Skyline software, this system can provide convenient interfaces for rendering the output in 3D display. The results show that SOA is an effective way to solve information isolated island, integrate heterogeneous computing systems and realize data and services sharing. The system demonstrates the feasibility of implementing information systems that are interoperable, low-cost, web-based, and which have a high evolution capacity through web services. This DSS can help disaster prevention department to analyze and visualize (charts, maps) the possible effects of storm events on both the immediate and long-term risks of flood damage at different regional level, which supplied convenient approach for making scientific policy decision of the flood control and disaster alleviation.