Generalized model for linear referencing

Geographic Information Systems utilize a twoor threedimensional spatial reference system for geo-locating features. Aspatial attributes about the feature can then be given a spatial context by joining the aspatially attributed feature with its spatial representation. In Transportation, a significant amount of information is maintained about linear features, such as roadways or rail lines. Because attribute values may vary along the linear feature, a one-dimensional linear, rather than spatial, referencing system is typically utilized. Without prescribed standards for linear referencing, the potential for linear locations to enable integration of information across the life cycle of a transportation facility, joining of information across multiple databases, interoperability across heterogeneous GIS, and support for data mining and warehousing is not realized. Even within a single institution, multiple Linear Referencing Methods (LRM) are often employed, restricting their ability to share information. A Generalized Model for Linear Referencing is proposed as a logical model for simplifying and implementing the US Transportation Research Board NCHRP 20-27(2) Linear Referencing System conceptual model. It abstracts the similarities which exist for all LRMs and presents these as a system of linear elements, measurement methods, and location, distance, and translation expressions. An LRM is characterized as a method of measuring (e.g., mile point, address, station) along linear elements, such as routes, streets or alignments. A location expression is introduced to specify event locations. It consists of a linear element and a distance expression. The syntax and semantics of the distance expression are dictated by the LRM. A generalized form of translation enables the calibration of distances along a linear element, the registration of linear elements with one another, and the transformation of location expressions between LRMs. The vast amounts of historical data stored in legacy systems against multiple LRMs can be integrated without having to standardize on a single LRM.