Assessing pedestrian crash risk and injury severity in concentrated urban environments

Walking is an indispensable travel mode in the transport system; almost all road users are pedestrians for all or part of their trip. It is also among the most environmentally sustainable transport modes and its promotion has been rigorously advocated since the rise of concern about global warming and environmental degeneration. Moreover, as a form of physical activity, walking has considerable benefit for public and individual health. Notwithstanding the positive aspects of walking, there are a number of barriers preventing people from walking as a mode of travel, with the lack of safety chief amongst these. The focus of this thesis, therefore, was to address the pedestrian safety problem by identifying factors associated with pedestrian crash occurrence and with the severity of injuries sustained. The Central Business District (CBD) of Melbourne was selected as the case study for this research. This site has a high level of pedestrian activities and high pedestrian crash rate. The PhD research aims to address the following questions: 1. What is an appropriate design for a pedestrian road safety data system to collect the data necessary for pedestrian safety analysis in the Melbourne CBD? 2. Using the tools developed in 1, what are the risk factors for pedestrian crashes in the Melbourne CBD: a) What are the characteristics of the physical and socioeconomic environment and the transport system that influence the occurrence of pedestrian crashes? b) What are the characteristics of the physical and socioeconomic environment, the transport system, road users (drivers/pedestrians) and traffic that influence the severity of injury sustained by pedestrians in traffic crashes? A principal contribution of this research was the development of a conceptual framework for studying pedestrian safety. A review of the available pedestrian data systems revealed that data collection practices are non-systematic and irregular, and the existing data, particularly pedestrian exposure to risk, is not sufficient for pedestrian safety analyses. Thus, a methodology was developed to estimate pedestrian exposure at CBD intersections and midblocks using existing data sources and a set of complimentary data collections. Pedestrian crash risk at CBD intersections and midblocks were evaluated, separately: during daytime hours, or alternatively called working hours, (7-18), and hours of darkness, or alternatively called non-working hours, (0-6 and 19-23). Findings showed that pedestrian crash risk is multi-factorial in nature. Different sets of factors were found to be associated with risk across different space-time frameworks. For intersections, the three most powerful predictors of pedestrian collision rate during daytime hours were the characteristics of land-use, road and public transport, respectively. Similarly, during hours of darkness, the rate was highly correlated with the characteristics of land-use and road. For midblocks, the most powerful predictor of the frequency of pedestrian crashes during daytime hours was public transport, followed by land-use and road characteristics, respectively. In contrast, midblock pedestrian collision rates during hours of darkness were mainly predicted by the length of the midblock, followed by land-use and public transport characteristics, respectively. Another key finding of the pedestrian crash risk analyses was that different measures of pedestrian exposure influence pedestrian crashes at intersections compared with midblocks. While traffic volume (Annual Average Daily Traffic (AADT)) played a significant role in developing pedestrian crash risk models at intersections, the product of “pedestrian road crossing counts” and AADT (P×AADT) was the influential pedestrian exposure measure in the midblocks daytime-hours risk model. Results of the pedestrian injury severity analyses also revealed a diverse set of factors associated with pedestrian injury severity for different space-time frameworks. The correlates of major pedestrian trauma at intersections were time of day, vehicle movement, pedestrian age, vehicle colour, and land-use and public transport characteristics. For midblocks, influential factors were time of day, the characteristics of land-use, and the interaction of day of week and speed zone. The injury severity data reported by the police for pedestrian crashes in the Melbourne CBD (2000-2009) was found to be not consistent with hospital data. Thus, using the police data can lead to misleading conclusions about pedestrian injury risk factors. Findings of this research provide evidence-based insight into the pedestrian safety problem in the Melbourne CBD and have implications for other concentrated urban areas. This knowledge can be applied by road safety researchers and practitioners to modify existing design practices to achieve a safer pedestrian environment and to facilitate the Victorian Government’s strategic plan for a Safe System for all road users. In addition, six new Central Activity Districts and 26 principal activity centres are projected to be designed and implemented by 2030 in Melbourne. The outcomes of this research provide critical, new information to enhance the pedestrian safety design guidelines for these new developments.