A Comprehensive Process for Linear Referencing

INTRODUCTION The primary objective of this paper is to identify, demonstrate, and analyze the necessary and sufficient requirements for exploiting linear referencing. A linear referencing process is developed and presented that expands on the extant linear referencing data models, methods, and systems that have appeared in the literature. This process is intended to provide a framework for the implementation of linear referencing among an expanding group of Geographic Information System (GIS) users. Linear referencing can be used by many organizations, industries, and institutions that work with linear features, such as road-management organizations, transit organizations, oil and gas exploration industries, and water-resources managers, to name only a few. The common element among these industries is their use of linear features and the need to reference a position or measure along those features. As GIS becomes more prevalent among an increasingly diverse and rapidly growing set of users, including small to midsize municipalities, government agencies at every level, and private businesses, there is an increasing demand for more sophisticated approaches to data management. When the network databases that have long been modeled in GIS (Curtin 2007a) are an important element of the analyses undertaken by these groups, the need to successfully implement linear referencing becomes an important issue, and a process for linear referencing is essential. This research presents a comprehensive process for linear referencing. In the following section, linear referencing is formally defined and its advantages outlined. This is followed by a comprehensive literature review that discusses both the applied use of linear referencing--particularly in GIS--and the theoretical models and methods that have been developed. Based on this review, a seven-step process for linear referencing is presented and its use is demonstrated through a case study of the city of Richardson, Texas. Conclusions regarding the potential use of this process and opportunities for future research are discussed. Linear Referencing Defined The term linear referencing emerged from engineering applications where it was preferable to locate a point along a linear feature (often roads) by referencing that location to some other well-defined location, rather than using classical geographic coordinate systems. The most familiar illustration of linear referencing is the mile markers along U.S. highways (Federal Highway Administration 2001, Federal Transit Administration 2003). Determining locations with linear referencing differs from traditional geographic coordinate and reference systems (latitude-longitude, Universal Transverse Mercator (UTM), state plane, etc.) in that the underlying entity used as a basis for measurement is not the earth, but is rather a linear feature or a set of linear features organized into a network. Just as there are myriad coordinate systems for the globe, there are multiple linear referencing systems. A common definition for a linear referencing system (LRS) is a support system for the storage and maintenance of information on events that occur along (or within) a transportation network. In this context, an LRS consists of an underlying transportation network that supplies the geographic backbone for the location of events, a set of objects with well-defined geographic locations (also known as a datum), one or more linear referencing methods (LRMs), and a set--or sets--of points or linear events that should be referenced to the underlying network. This paper will demonstrate that in many cases the underlying network and the datum are one and the same. An LRM can be defined as a mechanism for finding and stating the location of an unknown point along a network by referencing it to a known point (Vonderohe, Chou et al. 1997). More specifically, an LRM is a process for determining a previously unknown location based on (1) a defined path along the underlying transportation network, (2) a distance along that path measured from a known datum location, and (3) optionally an offset from the path. …