Analysis of severe and non-severe traffic crashes on wet and dry highways

Abstract

This study examined the effects of several variables associated with geometric, environmental, and drivers' demographic factors on the likelihood of occurrences of traffic crashes on two and four-lane urban and rural highways. Among the variables tested, the key factors contributing to the crash are used as input variables to Poisson and negative binomial regression models. Four separate datasets representing severe and non-severe crashes occurred on dry and wet pavement surfaces are used to develop their respective models. The results of the analysis showed that the common principal determinants influencing both severe and non-severe crashes occurred on dry and wet pavement surfaces on the selected highways are AADT (Annual Average Daily Traffic) and TADT (Truck Average Daily Traffic). The posted speed limits were found to be significant only for non-severe crashes occurred on wet pavement surfaces. The urban-rural designation of the segments was found to be a key factor for non-severe crashes occurred on dry pavement surfaces. Macrotexture and IRI (International Roughness Index) were critical determinants for non-severe dry pavement, and severe wet pavement crashes respectively.