No abstract

No abstract

Examining the spatial distribution of bicycle accidents under different conditions and in different periods is an important issue for increasing cyclist safety. A point pattern analysis methodology of 1,437 bicycle accidents that resulted in injury or death in the city center of Vienna, Austria, between 2012 and 2014 is described. Network-based kernel density estimation was used to examine the hot spots of bicycle accidents, and the network-based nearest-neighbor distance was taken into account to check the significance of the hot spots. Moreover, the global cross nearest-neighbor distance was used to test the effect of urban components on the distribution of bicycle accidents. An understanding of the temporal and conditional differences was obtained by analyzing the accident data in terms of four classifications: all accident data and then the accident data classified according to season, light conditions, and precipitation conditions. It was concluded that the bicycle accident hot spots varied in space according to season, light, and precipitation conditions. Also, these detected hot spots were significant for the pattern of accidents, no matter what classification was used. Besides these points, at the .95 confidence level, bicycle accidents tended to cluster by signalized intersections, bus and tram stations, subway stations, and city bike stations. As a result, a systematic framework was proposed for spatiotemporal analysis of bicycle accidents for the built environment. The framework can serve as a guide to determine effective strategies for cyclist safety in urban areas.