Linear Referencing

Determine Application, Representation, and Topology

There are fundamental differences in the structure of networks for different applications. Road and river networks, for example, do not have similar topological structures. The attributes and the analytical methods [Page 262]associated with different network types require different network representations. Therefore, the first step in a linear referencing process is to define which network data sets and spatial representations are to be employed for the application at hand.

Determining Route Structure

The next step is to determine the route structure. The term route in this context is the largest individual feature that can be uniquely identified and to which events can be linearly referenced. Any linear feature can become the underlying element defining routes, but, generally speaking, a route should be longer than the events to be referenced so that event segmentation is minimized. For example, if streets are the target network for linear referencing, one may want to define the routes as single entities that represent the entire northbound and southbound directions of travel along the street, even though there are many underlying features (different blocks of the street between intersections) in the network data set. Routes may be further divided if the street name or prefix changes somewhere along the length of the route. Figure 1 shows the definition of four routes along an arterial road, based on direction of travel, street name, and street prefix.

...