Building a Highway Linear Referencing System from Preexisting Reference Marker Measurements for Transportation Data Management

Article excerpt


In the United States, each state has a department of transportation that is tasked with managing, maintaining, and developing state highways. To manage events associated with the highways, transportation data systems have been developed to store relevant event information such as pavement types, construction zones, and motor vehicle collisions. The events are stored in a database with location values that are based on distances from measured points along the highway. However, before the advent of geographic information systems (GIS), the benefit of storing the event locations was limited to tabular analyses and paper maps. GIS now can take full advantage of the spatial location information, but significant work is required to match historical measurement systems on the highways to currently available digital street networks. Making this task even more difficult is the fact that many highways have undergone significant realignments over time and known reference markers for the same location on a highway now may have different measurement values. To account for these changes, a linear referencing system (LRS) can be developed from a digital street network.

Linear referencing is the process of storing geographic locations along a linear feature based on their positions relative to measured reference locations. On a highway, intersections and ramps can serve as reference locations to calculate distances to other geographic locations along the length of the highway. Typically, as changes occur to the roadway, new measurements are developed to account for the differences in the length of the route. For example, a bypass could be constructed, increasing the length of a highway that originally traversed through a city. New measure markers are placed along the bypass, but the entire length of the highway is not recalculated to maintain the consistency of previous event locations. In this situation, the beginning and end measurements of the highway remain the same, but the true distance of the route no longer would be equal to the original distance after the realignment. The ability to incorporate and represent these multiple measurements for the same route is fundamental to the concept of linear referencing and for effective transportation data management.

At the national level, the U.S. Federal Highway Administration (2011) maintains a highway inventory system known as the Highway Performance Monitoring System (HPMS) (http://www. HPMS contains information on the condition, extent, performance, use, and operating characteristics of the nation's roadways to accommodate a data-driven process for analysis, planning, and funding allocation purposes. States are required to submit roadway geometry information for all public roads with an associated LRS, but are at different stages of the submission process. The final long-term goal is to have a complete, standardized LRS accessible to the public for all roadways in the country. However, this is an ongoing task and many local agencies or private consulting firms have immediate needs for an LRS to locate different types of events on highways. There also may be a need to build an LRS on a newer or more accurate street network for various projects. Therefore, regardless of a national system, local systems frequently are necessary.

Numerous studies have been devoted to conceptual data modeling of linear referencing systems (Vonderohe and Hepworth 1998, Fletcher et al. 1998, Easa and Chan 1999, Adams et al. 2000, Adams et al. 2001, Scarponcini 2002, Curtin et al. 2007). Of these, most of the transportation-focused LRS literature outlines data models or best practices for developing new systems (Fletcher et al. 1998, Kiel et al. 1999, Scarponcini 2002, Steiner et al. 2002, Curtin et al. 2007, Zhang et al. 2010). They provide a comprehensive process for building an LRS from scratch and present guidelines for defining the base measuring system that can be used in new data models. …


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.