Document Type : علمی - پژوهشی


1 Assistant Professor , Department of Geoscience Engineering, Arak University of Technology

2 Assistant Professor, Department of Geoscience Engineering, Arak University of Technology


Pedestrians are considered as one of the most vulnerable road users due to many reasons including the lack of protective shells or being less visible to drivers, especially in low-light conditions or adverse weather and their accidents lead to many casualties or irreparable injuries. Accordingly, in recent years, efforts have been carried out to provide methods to investigate the presence of organized patterns of pedestrian accidents and to evaluate and organize them. When dealing with spatio-temporal data such as traffic accidents, the effect of events located in the spatial and temporal neighborhood of the studied event cannot be neglected. In this research, pedestrian accidents during a five-year period of 2014-2018 in Mashhad have been examined. By aggregating traffic data over 253 traffic analyses zone (TAZ), the temporal pattern and the presence of temporal autocorrelation among monthly and hourly pedestrian accidents has been confirmed using time series analyses. Using Buishand’s homogeneity test, a sudden change of accident occurrence at different hours of the day (7:00-8:00 A.M.) and months of year (July and September) was identified. Spatio-temporal differential Moran's I index was applied for the first time in safety analyses to evaluate the spatial correlation of changes of pedestrian accidents between the two periods of the beginning (2014) and end of the analysis period (2018), and the locations with significant changes were identified. As an exploratory method, this research can provide traffic engineers and planners of urban transportation with a tool to identify areas showing high rate of accidents over a period of time and therefore, require more attention in terms of allocating budget or providing strategic approaches to effectively improve the pedestrian safety.