Macro-Level Modeling of Urban Transportation Safety: Case-Study of Mashhad (Iran)

Abstract

Transportation safety can be aimed at the planning stage in order to adopt safety management and evaluate the long-time policies. The main objective of this research was to make use of crash prediction models in urban transportation planning process. As such, it was attempted to gather data on the results of transportation master plan as well as Mashhad urban crash database. Two modelling method, generalized linear model with negative binomial distribution and geographically weighted regression, were considered as the methods used in this research. Trip variables, including trip by car, trip by bus, trip by bus services and trip by school services, were significant at 95%. The results indicated that both finalized models were competent in predicting urban crashes in Mashhad. Regarding to results urban transportation safety will be improved by changing the modal share for example from private car to bus. The application of the process presented in this study can improve the urban transportation safety management processes and lead to more accurate prediction in terms of crashes across urban traffic areas.

Keywords

• safety planning
• urban transportation
• macro-level modelling
• negative binomial
• geographically weighted regression