Weather impacts on transportation safety: rainfall in Texas

Traffic safety and rain-related crashes are a major concern for engineers, planners, and emergency officials. Texas has more roads than any other state; therefore, the study of the factors that cause crashes is crucial to reducing the risk. Traffic safety is significantly impacted by weather conditions. This study contributed to transportation safety by providing an in-depth analysis of rain-related crashes in Texas. The first study used fatal crash data and Geographic Information Systems (GIS) to explore the spatial and temporal distribution of rain-related fatal crashes. The Getis-Ord Gi* statistic was used to identify clear and distinct spatial clustering patterns of rain-related fatal crashes and their correlation with rainfall. Fatal crash data were obtained from the Fatality Analysis Reporting System (FARS) website maintained by the National Highway Traffic Safety Administration. The spatial statistical analysis revealed the spatial patterns of rain-related fatal crashes that are clustered in different counties depending upon rain conditions. The second study contributed to traffic safety by assessing the impact of rain and by conducting a statistical analysis of FARS crashes from 1982 to 2011. Logistic regression was used to identify the dominant factors associated with rain-related fatal crashes. The rain-related crash factors were categorized and examined at the state and county levels and were classified as environmental-related, roadway-related or driver-related. This study results can be used to aid the traffic safety professionals in the planning and design of roads. The final study adopted a matched-pair model to quantify the relative crash risk ratio during rainy conditions on two different areas in Texas: a segment of an interstate highway and an urban center from 2006 to 2013. Crash data were extracted from the Crash Records Information System database maintained by the Texas Department of Transportation. Hourly rainfall data was obtained from the National Weather Service's West Gulf River Forecast Center Multisensor Precipitation Estimates. Comparison of crashes during paired dry and rainy periods indicates that the relative risk ratio during rainy periods was greater than 1.0 for all road segments during the entire analysis period. The results indicate that rain conditions increase the crash risk when other factors are kept constant. The knowledge gained from this matched-pair study can be used to reduce rain-related crashes and further traffic safety countermeasures. The first method used in this research was aimed to study the spatial distribution of rain-related fatal crashes at the county level. The second method used in this study was aimed to use logistic regression to identify rain-related crash factors for Texas at the state and county levels. The third method used in this study was aimed to quantify relative crash risk at a fine resolution with a matched-pair method. The methodology developed in this research to gather and analyze crash data at the state, county, and road segment levels can be used to identify the causes of and reduce rain-related crashes. The benefits of this approach are discussed in more detail in the conclusions and recommendations.