Geographic Information System Platform for Road Accident Risk Modeling

Abstract

Road accident risk assessment requires a thorough understanding of both the vehicle accident involvement process and the severity of resultant injuries. A geographic information system (GIS) platform is especially suited to this type of problem because it provides an efficient system of linking a large number of disparate data bases, it provides a spatial referencing system for reporting output at different levels of aggregation, it allows input of both historical and statistical accident experience in estimating accident risk at different locations and times, and it allows controls on a myriad of risk factors explaining variations in accident involvement and injury severity. A GIS-based accident risk model developed for the Ontario highway network is described. The model provides estimates of accident risk at four levels of spatial aggregation as specified by the user: networkwide, route-specific, route-section-specific and site-specific.