Traffic crashes tend to occur at relatively greater frequencies at particular locations, at particular time periods, and for particular subsets of drivers and vehicles. It is well recognized among the road safety community that crash-risk is highly elevated when inclement weather conditions occur in the winter. To present, most of the road safety studies focus on event-based analysis or seasonal analysis and give little attention to explore high-risk conditions at the daily temporal scale. The purpose of the study is to advance our understanding of high-risk crash conditions at the daily level and their occurrences in Southern Ontario, Canada. The study explores different definitions of high-crash days, and quantifies the influences of weather conditions, risk exposure, months and timing of precipitation on the likelihood of a high-crash day occurring using binary logistic regression model. Additionally, an approach for estimating the relative risk exposure using available traffic count data has also been developed. The results of the study show a small proportion of high-crash days are responsible for a considerable amount of traffic crashes during the winter. The risk of traffic crash is twice as high on high-crash days in comparison to non-high-crash days. The modeling approach well-fits the data and shows that winter weather conditions have significant influence on high-crash days with results being mostly consistent across the four study areas, Toronto, the Area Surrounding Toronto, London and the Area Surrounding London. Low temperature, heavy snowfalls, high wind speeds, high traffic volumes, early winter months, occurrence of precipitation in both morning and evening increase the odds of high-crash days to a large extent. The results of study could help to pre-schedule traffic operation and enforcement, to effectively distribute road safety resources and personnel, and to create situational awareness among road users and other stakeholders.

Traffic crashes tend to occur at relatively greater frequencies at particular locations, at particular time periods, and for particular subsets of drivers and vehicles. It is well recognized among the road safety community that crash-risk is highly elevated when inclement weather conditions occur in the winter. To present, most of the road safety studies focus on event-based analysis or seasonal analysis and give little attention to explore high-risk conditions at the daily temporal scale. The purpose of the study is to advance our understanding of high-risk crash conditions at the daily level and their occurrences in Southern Ontario, Canada. The study explores different definitions of high-crash days, and quantifies the influences of weather conditions, risk exposure, months and timing of precipitation on the likelihood of a high-crash day occurring using binary logistic regression model. Additionally, an approach for estimating the relative risk exposure using available traffic count data has also been developed. The results of the study show a small proportion of high-crash days are responsible for a considerable amount of traffic crashes during the winter. The risk of traffic crash is twice as high on high-crash days in comparison to non-high-crash days. The modeling approach well-fits the data and shows that winter weather conditions have significant influence on high-crash days with results being mostly consistent across the four study areas, Toronto, the Area Surrounding Toronto, London and the Area Surrounding London. Low temperature, heavy snowfalls, high wind speeds, high traffic volumes, early winter months, occurrence of precipitation in both morning and evening increase the odds of high-crash days to a large extent. The results of study could help to pre-schedule traffic operation and enforcement, to effectively distribute road safety resources and personnel, and to create situational awareness among road users and other stakeholders.