On the safety effects of off-peak hour speed characteristics of urban arterials

Among the factors related to traffic crash risk, the speed characteristics are crucial. Most studies on the safety effect of speed characteristics focused on highways and rural roads, whereas the investigations on urban roads are not comprehensive. Urban arterials operate at higher speeds during off-peak hours, which may possibly lead to more serious crashes. Hence, this study focuses on the correlation between speed characteristics and serious crash occurrence (i.e., injuries and fatalities) on urban arterials during off-peak hours, while considering the interaction between speed and road/traffic characteristics. The spatial autocorrelation and intrinsic correlation of injury and fatal crashes are analyzed by using multivariate conditional autoregressive model (MVCAR) from 12 urban arterials in a district in Ningbo, China. Research findings include: (1) speed characteristics, including the percentage of speeding vehicles, mean speed, speed standard deviation, speed skewness, were found as significant and the inclusion of interaction terms of speed characteristics improved the model fit; (2) the interaction terms of percentage of speeding vehicles with the presence of median and access density, speed skewness with access density showed significant effects; (3) the interaction term of mean speed and access density are positively correlated with crash risk; (4) Speed standard deviation is positively correlated with crash risk. The findings can provide guidance for improving urban speed management and safety.