Accuracy of Linearly Referenced Data by Using Geographic Information Systems

Abstract

Distance accuracy issues are identified that should be considered when linear referencing procedures based on geographic information system (GIS) technology are applied to map objects and events to highway networks. One of the most pressing issues is that of consistency among GIS functions to compute surface distances. Several GIS packages are compared using two sets of points: National Geodetic Survey geodetic control points for which independent estimates of distance were available and virtual points for measuring the effect of increasing distance between points. Results show a wide discrepancy in distances measured, and tools are provided to identify two groups of GIS packages: those that use a variable radius ellipsoidal model of the surface of the earth and those that use a less flattened, possibly spherical model of the earth’s surface. The accuracy associated with distance measurements varies according to the model of the earth’s surface used. Thus, transportation agencies need to be aware of their distance accuracy needs and the distance accuracy their GIS package can provide. Guidelines are provided that transportation agencies could use to determine distance accuracy requirements and to evaluate GIS distance accuracy capabilities. These guidelines could also be used as a foundation for the development of standardized GIS distance-measuring functions to facilitate distance and linear referencing data exchange across hardware and software platforms.