GIs in Transportation

Geographic information system (GIS) technology is often used in improving transportation. Transportation applications of GIS, also referred to as GIS-T, are popular and extensive in scope. Transportation problems range from day-to-day operations (in-vehicle global positioning system [GPS] navigation, the North American E-911 emergency response system) to large-scale efforts (prioritizing policy alternatives, long-range transportation planning) by metropolitan planning organizations. This entry reviews the policy contexts and market forces that influence the way institutions approach transportation problems. Then, data models and operations (with a focus on analysis and modeling) unique to GIS-T are reviewed.

Policy-wise, the span of policy goals has stretched from improving efficiency and mobility (e.g., the interstate highways) to embracing sustainable development, including mitigating negative externalities (e.g., air pollution, safety, congestion), as indicated in the passage of legislation such as the Clean Air Act, the Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA), and the Transportation Equity Act for the 21st Century (TEA-21). Along with the changing policy context (as well as advances in geotechnology), GIS-T in the public sector has widened from the inventory of infrastructure and travel demand forecasting to Web-based public participation tools and decision support systems.

Market forces driven by advances in technology (e.g., the Internet, positioning technology, wireless sensor networks) are reshaping the way GIS handles transportation problems. They not only change the way geographic information is serviced, or disseminated, but also change the way geospatial data are captured. The era of what some call geonomadism or geosurveillance has already arrived. GIS-T in the private sector encompasses location analysis, logistics, and location-based services.

To understand how GIS in transportation works, it is necessary to have a grasp of basic data models and operations in GIS-T. Data models and operations are dependent on application needs, and they are interrelated. For example, spatial mapping of the fixed transportation infrastructure, monitoring the pavement conditions of sections of highway, and finding the best route for ambulance dispatch require different specifications of data models and operations.

A universally adopted approach to representing a transportation network is to break down the network to links and nodes where links correspond to linear segments and nodes correspond to intersections. Links and nodes are different from lines and points in that they are topological constructs (e.g., they possess connectivity). As many events occur along linear features, it is common to identify their location using a linear referencing system (LRS). LRS uses offset distance (how far) from a reference point along the route. For example, police officers can locate accidents on the highway using an LRS. The link-node model conveniently (but redundantly) stores transportation networks in files, but it is not amenable to representing something of the dynamic nature that is referenced using an LRS (e.g., accidents, pavement conditions, bus routes) and that does not [Page 1312]necessarily begin or end at the node. GIS has capabilities to dynamically generate those events tied to links based on their attributes (dynamic segmentation). Origin-destination matrices are used to store spatial interaction data (e.g., traffic flows among traffic analysis zones), although this is not well supported in generic GIS software due to the dominance of relational databases.

...