Many regions across the United States experience a number of winter snow events every year, during which a multitude of vehicular crashes occur. The authors formally compare crashes and occupant injuries reported on Interstate highways in Iowa during winter snow event periods with those reported during equivalent winter nonsnow event periods. The comparison is fairly comprehensive because it includes detailed crash, weather, and roadway geometry information on seven different sections of the Iowa Interstate highway system. Further, the analysis utilizes traffic count data specific to snow event and comparable nonsnow event periods. The comparison was facilitated by using a geographic information system that merged data from a variety of sources, resulting in the creation of a unique data set with significant spatial and temporal coverage. After explicitly controlling for seasonal, monthly, weekly, hourly, and locational variations in crash occurrence, significant increase was observed when winter snow event injury and noninjury crash rates (crashes per million vehicle kilometers) were compared with equivalent winter nonsnow event injury and noninjury crash rates. The data were then analyzed for injury occurrence. Results of a logit model indicated that crash injury occurrence on Interstate highways in Iowa depended on traffic, road geometry, and number of vehicles involved in a crash. Another finding from the logit model was that crashes during snow events were less injurious compared with equivalent nonsnow event crashes. Snow event-specific crash data were then analyzed to study the effects of snow event elements (e.g., snowfall intensity) on injury occurrence in vehicular crashes.
ASJC Scopus subject areas
- Civil and Structural Engineering
- Mechanical Engineering