Spatial and Temporal Distribution of Pedestrian Crashes

Abstract

In order to develop effective and targeted safety programs, the location and time-specific influences on vehicle-pedestrian crashes must be assessed. Therefore, spatial autocorrelation was applied to the examination of vehicle-pedestrian crashes in geographic information systems (GISs) to identify any dependency between time and location of these crashes. Spider plotting and kernel density estimation (KDE) were then used to determine the temporal and spatial patterns of vehicle-pedestrian crashes for different age groups and gender types. Temporal analysis shows that pedestrian age has a significant influence on the temporal distribution of vehicle-pedestrian crashes. Furthermore, men and women have different crash patterns. In addition, the results of the spatial analysis show that areas with high risk of vehicle-pedestrian crashes can vary during different times of the day for different age groups and gender types. For example, for the age group between 18 and 65, most vehicle-pedestrian crashes occur in the central business district (CBD) during the day, but between 7:00 pm and 6:00 am, crashes for this age group occur mostly around hotels, clubs, and bars. Therefore, specific safety measures should be implemented during times of high crash risk at different locations for different age groups and gender types, in order to increase the effectiveness of the countermeasures in preventing and reducing the vehicle-pedestrian crashes.