Spatiotemporal Variation of Rear-End Conflicts at Signalized Intersections under Disordered Traffic Conditions

The present study explores the spatial and temporal variation of rear-end conflicts at signalized intersections under mixed traffic conditions with poor lane discipline (disordered traffic conditions). Vehicular trajectory data was extracted using the videographic field data of three signalized intersections in India. Due to disordered traffic conditions, a simple and robust method is proposed to identify the leader and follower vehicle pairs for identifying rear-end conflicts. Spatiotemporal variation of identified conflicts is investigated using two popular surrogate safety measures (SSMs). A significantly higher number of rear-end conflicts occur at the downstream of intersections compared to the upstream section. The curbside lanes with a higher proportion of smaller vehicles observe more rear-end conflicts than medianside lanes. Different vehicle classes with varying lateral movements affect vehicle following time, resulting in constantly changing leader-follower pairs and a high probability of rear-end conflicts. Dynamic parameters such as speed, acceleration, and spacing maintained by the following vehicles offer detailed insights into identifying conflicting interactions. Finally, rear-end conflict and crash probability at different study locations and vehicle classes is determined using the peak over threshold approach of extreme value theory.