Special lower speeds or longer headway maintained by

Special attention was given to
crash severities by vehicle types in several studies. Shankar and Mannering
(1996) used a multinomial logit model to predict five-level rider injury
severities from statewide single-vehicle motorcycle crashes in Washington. In their
study, pavement surface and weather interaction were significant factors to
increase the likelihood of being injured. Interestingly, the crash severity on
wet pavement without rainfall was limited only to property damage and possible
injury crashes. Similarly, wet pavement for single-vehicle motorcycle crashes
was more likely to result in no injury according to a study by Savolainen and
Mannering (2007). The authors argued that it could be a result of lower speeds
or longer headway maintained by riders in these conditions as they adjust for
the perceived higher risk. Khorashadi et al. (2005) also utilized the
multinomial logit model to analyze driver injury severities involving
large-truck accidents. Results in their study showed that rainy weather was significant
to increase injuries in urban area accidents. Ulfarsson and Mannering (2004)
showed that wet roads led to increasing relatively higher injury severities
involved in single sport utility vehicle (SUV)/minivan accidents using
multinomial logit model. In a study by Kim et al. (2007), bicyclist injury
severities in bicyclemotor vehicle accidents were predicted by the same method.
Inclement weather including rain, snow, and fog was found to increase the
probability of fatal injury approximately by 129%. The authors stated that the
inclement weather effect was largely due to increased slipperiness which
reduced both the vehicle’s and bicycle’s maneuverability and visibility. Pai
and Saleh (2008) used ordered logit model to assess motorcyclist injury severities.
In their study, on the contrary, fine weather was a statistically significant
factor to increase the most severe category of motorcyclist injury.

Rainfall-related effect on crash
severity outcomes has been also identified along with roadway characteristics.
Ordered probit models were used in a Abdel-Aty’s study (2003) to predict driver
injury severity in Central Florida, with crashes occurring in specific roadway
sections, signalized intersections and toll plazas in expressway systems. It
was found that crashes happening in signalized intersections with bad weather
and dark street lighting had a significantly higher probability of severe
injury. Abdel-Aty stated that an angle and turning collision in the adverse
weather and dark street light conditions was a possible reason to contribute
higher probability of injuries in signalized intersections. Donnell and Mason
(2004) employed ordinal and nominal logistic regression models to predict
interstate highway crash severity in cross-median and median-barrier
collisions, respectively. Researchers found wet or icy pavement surface to be
significant factor in decreasing crash severity. In contrast, in a study by Lee
and Mannering (2002), the nested logit analysis showed that wet roadway
surfaces increased the likelihood of evident and disabling injury/fatality in
run-offroadway accidents. The conflicting results from these two studies
suggest the need for a detailed analysis for weather-related crash severity by
collision type.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now