Related Experiment Videos
Explaining two-lane highway crash rates using land use and hourly exposure
J N Ivan1, C Wang, N R Bernardo
1University of Connecticut, Connecticut Transportation Institute, Civil and Environmental Engineering, Storrs 06269-2037, USA. john.ivan@uconn.edu
Accident; Analysis and Prevention
|September 20, 2000
Summary
This study models highway crash rates using Poisson regression, finding traffic intensity significantly impacts single and multi-vehicle incidents differently. Key factors include traffic volume, land use (driveways), and road geometry, with varying effects by crash type.
Area of Science:
- Transportation Engineering
- Traffic Safety Research
- Statistical Modeling
Background:
- Highway safety is influenced by numerous factors including traffic volume, land use, and environmental conditions.
- Predicting crash rates accurately is crucial for implementing effective safety interventions.
- Existing models often require refinement to differentiate between single and multi-vehicle crash dynamics.
Purpose of the Study:
- To estimate Poisson regression models for predicting single and multi-vehicle highway crash rates.
- To identify significant predictors of crash rates, including traffic density, land use, light, and time of day.
- To analyze how these factors differentially affect single-vehicle versus multi-vehicle crashes on rural, two-lane highways.
Main Methods:
- Utilized Poisson regression models to analyze crash data from seventeen rural, two-lane highway segments.
- Incorporated hourly traffic counts, land use data (driveways), traffic density (volume/capacity ratio, VMT), and light/time conditions.
- Developed separate models for single-vehicle and multi-vehicle crashes, examining variable significance and effect direction.
Main Results:
- For single-vehicle crashes, significant predictors included daytime (negative effect), log(V/C ratio) (negative), no passing zones (positive), shoulder width (positive), intersections (negative), and driveways (mixed).
- For multi-vehicle crashes, significant predictors were daylight (positive), intersections (negative), and driveways (positive).
- Traffic intensity was a key differentiator for crash rates, even when controlling for time and light, with distinct effects for single vs. multi-vehicle crashes.
Conclusions:
- Traffic intensity, land use, and road characteristics significantly influence highway crash rates.
- The factors affecting single-vehicle and multi-vehicle crashes differ substantially.
- Findings provide insights for targeted highway safety improvements and future research directions.