| GIS: GIS Component Software for Oil Spill Response
A Rhode Island company develops a Windows-based oil spill response system that fully integrates GIS, mapping and database capabilities, model interfaces and links to real-time data.
By Eoin Howlett and Mark R. Bradstreet Introduction
There has recently been a considerable increase in the use of computer-based systems for responding to both man-made and natural disasters such as oil spills, atmospheric releases, earthquakes, and floods. The dramatic advances in hardware and operating system design means that many predictive numerical models that were previously only used by universities and research institutes, are now being packaged as a component of integrated GIS-based decision support applications. Common examples of these are oil spill model prediction systems, atmospheric dispersion models, dredge material disposal models, and a variety of hydrodynamic and water quality models.
Applied Science Associates of Narragansett, Rhode Island is leading the new wave of component based GIS application development. ASA builds applications in near real-time for their clients real-time decision support using solid component architecture, GeoViewª, provided by Blue Marble Geographics of Gardiner, Maine. Case Study
Consider the following scenario. Oil company Mega Oil has its corporate offices in Dallas, Texas, its affiliated shipping company Mega Shipping in London, and production and distribution facilities all over the world. A spill occurs from one of their vessels which has had a stop in Philadelphia, on the Delaware River. The spill response advance team leaves within three hours from London to go to the spill site. Dallas is the command site for the worldwide spills, and sets its spill response plan in motion. A local area network inside the Dallas office allows all of the responders to use a GIS to view their existing maps and response/protection plans for the area. Included are geo-referenced data for stockpiled equipment, pre-planned boom sites, and the locations of sensitive environmental areas. As soon as the team from London arrives at the spill site, Mega Shipping sets up an on-scene control center and satellite links allowing on-scene responders to view all the database information stored in Dallas.
The response teams get to work. After the first day at the command site separate groups are set up to manage the several aspects of the clean up response. The Operations group begins coordination of recovery and shoreline clean up efforts. The Logistics group tries to acquire all the necessary equipment and personnel within "reach." The Planning group predicts where the spill will go in the next 24-48 hours and prepares plans for the subsequent days' activities. A Finance group tracks the costs as response assets are enabled and assigned. The Legal group tries to keep track of every activity so that they have all the necessary data for insurance claim and potential litigation use.
On day two, experts from a local university are brought in as consultants and promptly point out several instances in which the environmental data displayed is incorrect or out of date. Response contractors begin arriving from several nearby states charging Mega Shipping for response equipment and personnel that are difficult to track. Information from several sources in the field yields contradictory data on the size and location of the spilled product. Real-Time Solutions
The above presents a formidable problem for any data management system. The scenario requires the system to manage all of the static pre-prepared planning data; keep track of response resources as they are deployed; allow the planning group to manipulate data playing "what-if" scenarios with predictive models; flag all of the logistics requests for additional equipment and inform planning when and if they will become available; and store all of the transactions for cost and legal purposes.
The above data must be displayed on a map-based presentation so that the emergency responders and various agency representatives in a unified command structure receive their information of interest in a common display format. The display format should be simple and consistent, so that responders can quickly identify with well-known landmarks. This may be as straightforward as including local highways, roads, coastal features, river mile markers, or significant marine navigation points on the map presentation. Although this may appear to be a trivial point, it has proved to be a very important feature of printed status maps, especially when information is to be faxed, or described to responders over radio or mobile phone.
So is integrating a GIS the solution? Our experience has been that integrating fully functional GIS packages designed to address a huge variety of geographic data is not feasible for response situations. A complete GIS package, although very powerful, provides so much analytical functionality that only the GIS expert can really use it in a time effective manner. In an emergency situation the ability to respond quickly is paramount. In addition, most GIS' are developed to manage relatively static information, and are not linked to time-changing processes. The type of GIS functionality needed for emergency response problems is to access data from master centrally located databases, and also integrate data that may be changing constantly at the accident site. Fast and Accurate Assessment
Although the newest PC-based GIS applications are actually "application frameworks" that are customizable, their limitations in functionality, accepted data formats, various coordinate system support, user interface design, and development utilities make them impractical for developing true "subject-oriented" applications.
IntegratingBuilding on the GeoView™ software component, as opposed to using a GIS application framework, proved to be the ideal solution for developing the oil spill response application. The GIS software component provided all of the GIS-related functions required such as handling map projections, georeferencing raster and vector data, layer and object classification. Eliminating the need for converting data to and from various formats is a real time saver. But the overriding advantage in using a software component is that full control remains with the software developer.
We were able to use the standard programming language of our choice like Visual Basic, and Visual C++ to create sophisticated user interface components and incorporate a variety of other third-party components to develop a fully integrated oil spill response system. Using a Windows-specific development environment allowed us to access the Windows Application Programming Interface (API) directly. Use of the Windows API made animation of results and full control over windows messages possible and allowed for high speed display of complex objects. Future developments for Windows NT will allow us to access the OpenGL directly and allow the 3D display of numerical model results. Conclusion
Using GeoView™, Applied Science Associates has rapidly developed a Windows-based oil spill response system that fully integrates a user friendly GIS, powerful mapping and database capabilities, model interfaces, environmental data tools, numerical models, and linkages to real-time monitoring data. This system allows users to maintain environmental and geographic databases for use by a number of different prediction models.
Information in existing GIS databases may be imported, displayed, edited, and used interactively with the models, or exported to other GIS software formats. Near real-time data acquisition directly into the GIS is accomplished through links to remote GPS and environmental sensor hardware. Spatial and logical query selection tools provide access to the attribute information within the integrated GIS database. Groups of GIS objects may be created by spatial selection, or by logical arguments keyed into a simple form. The integrated GIS enhances data capture and visualization, facilitating faster assimilation and display of information in a more efficient modeling environment. The numerical models provide the predictive capabilities needed for environmental planning and management. The ability of models to communicate with remote-sensed and real-time changing GIS data allows modelers to rapidly adjust model predictions based on real-time observations.
By taking advantage of the Blue Marble Geographics suite and other component-based software, a customized GIS application can focus on rapidly changing data for use in numerical model and emergency response situations. This off-the-shelf GIS component library can be incorporated as DLL or OCX component in a 16 or 32 bit Windows environment plus allows the use of standard development tools such as Visual C++ and Visual Basic.
The system is currently in use by government agencies and oil companies in countries throughout the world for oil spill response, contingency planning, regulation compliance and emergency preparedness training. About the Authors:
Eoin Howlett is the Director of Software Development at Applied Science Associates. His primary interests are integrating GIS technology with marine-based numerical modeling and real-time environmental monitoring. He may be reached at: [email protected] Mark R. Bradstreet is the director of sales and service at Blue Marble Geographics. He provides assistance to application developers searching for appropriate software components to accelerate development of their GIS applications. He may be reached at 207-582-6747 (phone) or through e-mail: [email protected] Back |
