Factors driving intersection pedestrian crash risk in concentrated urban environments
Concentrated urban environments are known to be pedestrian crash clusters. An in-depth understanding of pedestrian crash risk is a key to identifying the appropriate distribution and focus of safety interventions. This research focuses on the assessment of the risk of pedestrian crashes at the Melbourne Central Business District (CBD) intersections. Intersection pedestrian casualty crashes reported to police that occurred within the Melbourne CBD during a 10-year period from 2000 to 2009 were used. Pedestrian road crossing counts and Annual Average Daily Traffic (AADT) at intersections were used as the measures of exposure. The characteristics of the built environment (land use, roads, public transport and environmental features) and the socioeconomic environment were used as independent predictors. Count data regression methods (Poisson and Negative Binomial) were employed to investigate the association between pedestrian crashes and predictor variables during weekday daytime hours (07:00-18:59) and hours of darkness (19:00-06:59). The three most important categories of predictors of intersection pedestrian crashes during daytime hours were land uses surrounding the intersection (shops; entertainment areas), road factors (hierarchy of roads; divided vs. non-divided; percentage of left turn lanes) and public transport factors (bus stops; distance from railway station), respectively. Similarly, during hours of darkness, pedestrian crashes were highly correlated with the characteristics of surrounding land uses (entertainment; gaming; cinema/theatre; accommodation) and road factors (hierarchy of roads; divided vs. non-divided). The findings of this research provided a better understanding of the unique factors that are likely to influence pedestrian crash risk at intersections in concentrated urban environments.