Prediction of Seasonal Variation in Traffic Collisions on Urban Highway: A Case Study in the Province of British Columbia

Traffic collisions are one of the world's major problems. According to the World Health Organization (WHO), about 1.35 million people die each year in traffic collisions across the world and a further 20 - 50 million are injured or disabled. This number increases every year. Various methods/techniques were developed in order to assess and improve highway safety. Collision rates, linear regression and generalized linear regression methods have been used for safety analysis. Research has shown that there are limitations with such approaches due to the non-linear relationship between collision frequency and exposure. Collision prediction modeling is the recommended technique for estimating road safety in the American Association of State Highway and Transportation Officials (AASHTO) Highway Safety Manual (HSM). However, traffic seasonal variation and weather impacts were not considered in the prediction modeling as the annual average daily traffic (AADT) is one of main variables having a direct impact on safety. This study analyzed the seasonal variations of traffic and collisions on urban highways in British Columbia, Canada. Collision risks related to winter weather were investigated and assessed. Neural network regression models are developed to predict the seasonal variation of traffic collisions. The study concludes that traditional techniques of highway safety assessment on urban highways without the consideration of traffic seasonal variation and weather condition, especially impacts of winter condition, might result in underestimating/overestimating the collision risk. The proposed models could incorporate traffic seasonal variations, temperature and precipitations, address no-linear issue and provide over 85% accuracy of traffic collisions over a 5 year-period and about 10-25% more accurate estimate of traffic collisions comparing the prediction model developed by British Columbia Ministry of Transportation and Infrastructure (BCMOTI), Canada. The study suggests that the collisions can be more accurately predicted if seasonal variation of traffic volume and weather related factors can be incorporated in the prediction model on urban highways. The study also confirms the importance of winter maintenance to prevent traffic collisions. This will be very important for Canada and other countries with strong seasonal weather impacts (e.g. severe winter conditions) as the proposed models will be able to provide guidance for winter road maintenance and highway safety management policies in order to reduce traffic collisions and improve highway safety. Furthermore, it also suggests a need for an interdisciplinary approach in highway safety assessment in order to integrate all factors including human, vehicle, traffic and road/environment and provide a clear and comprehensive understanding of the causes of collisions.