Identification of Secondary Traffic Crashes and Recommended Countermeasures

Abstract

Secondary crashes (SCs) usually occur due to congestion or other prior incidents. SCs are increasingly spotted as a significant issue in traffic operations, leading to reduced capacity, extra traffic delays, increased fuel consumption, and additional emissions. SCs have substantial impacts on traffic management resource allocation. One of the challenges in the traffic safety area of the transportation industry is to determine an adequate method for identifying SCs. The specific objectives of this study are: identification of SCs using spatiotemporal criteria and exploring the contributing risk factors to the identified SCs. Two different approaches were explored to identify SCs. The first approach is based on a “static” method that employs a predefined 2 miles-2 hours fixed spatiotemporal threshold. Four-year (2011 to 2014) crash and traffic data from the Crash Analysis Reporting (CAR) system database were used. The linear referencing tool of Geographic Information Systems (GIS) was applied to identify crashes that fell within the threshold. About 1.49% of all crashes were identified as SCs. A Structural Equation Model (SEM) was developed to investigate the contributing risk factors to the occurrence and severity level of SCs. Model results revealed that a series of driver attributes contributed to the occurrence of SCs, including the influence of alcohol or drug, inattentive driving, fatigue or speeding. Other variables that might lead to higher probabilities of SCs include vehicle attributes (brake defects, motorcycles), roadway conditions (roadway surface, vision obstruction) and environmental factors (raining condition Given that about 40% of SCs were rear-end crashes, this study also examined contributing factors to severity levels of rear-end SCs. Results revealed that the presence of horizontal curves, presence of guardrail, and posted speed limit showed a significant influence on the severity level of SCs. Crash modification factors were also developed by considering the roadway and traffic characteristics. In contrast to the static method, the dynamic approach identifies a dynamic spatiotemporal impact area for each primary incident using the Speed Contour Plot method. This analysis was explored using the Regional Integrated Transportation Information System (RITIS) and the SunGuide™ database for the year of 2014-2017. This study further analyzed contributing risk factors to SCs on I-95 and found that SCs were more likely to occur if primary incident clearance times were longer. It also revealed that SCs were more severe at night and on weekends. It implies that timely emergency responses would have a significant effect on mitigating SCs. These findings point