The relationship between traffic accidents and road geometry in rural areas: The case of Zonguldak

This study comprehensively examines the relationship between fatal and injury traffic accidents and road geometry on rural highways in Zonguldak Province between 2016 and 2022. The primary aim of the research is to analyze the geometric features of roads at accident-prone sections to determine the impact of these factors on accident occurrence and to identify design-related issues. The traffic accident data used in the study were obtained from traffic accident reports provided by the Zonguldak Regional Traffic Inspection Branch Office, Devrek Regional Traffic Inspection Station Office, and the General Directorate of Security. Road geometric data were derived from highway projects acquired from the 15th Regional Directorate of Highways. The study utilized data 941 traffic accidents that occurred over a 7-year period between 2016 and 2022. Approximately 156 kilometers of highway projects related to four different state road segments in the vicinity of Zonguldak were analyzed as part of the study. Subsequently, the road geometric data corresponding to the accident segments were matched to form the core dataset of this study. The basic descriptive statistics of the road geometry variables have been determined, and models have been developed to analyze the effects arising from the combined influence of independent variables. The impact of road geometry on the frequency and severity of traffic accidents was examined based on combinations of horizontal and vertical geometric elements. Four primary geometric combinations were analyzed in terms of their effects on the annual number of accidents per kilometer: horizontal curve and vertical curve, horizontal curve and longitudinal grade, horizontal alignment and vertical curve, and horizontal alignment and longitudinal grade. In this context, models based on multiple regression analyses were developed to enhance the effects of design elements on accidents were quantified. The results of the study revealed the impact of road geometric features on the likelihood of traffic accidents. It was proven that horizontal and vertical alignments have a strong relationship with highway traffic accidents. By examining the number of accidents per kilometres, it was determined that horizontal curves are riskier than horizontal alignments, and vertical curves are riskier than sections with longitudinal grades. The section where horizontal and vertical curves coexist was identified as the most hazardous, with the highest annual accident rate of 9.2 accidents per kilometres. Multiple regression analyses established the relationships between the independent variables of the four different combinations and the number of accidents, leading to the development of predictive models. These findings identified the variables contributing to accidents in critical road segments and provided numerical insights that can be utilized to enhance the safety of existing roads and inform the planning of new roadways. This study offers numerical models based on a comprehensive dataset, addressing a gap in the literature and providing a unique contribution to understanding the impact of road geometry on accidents. The analysis of real-world data collected from the field delivers concrete, localized, and effective solutions for road safety studies, enabling a better understanding of the conditions and causes of traffic accidents. Therefore, the data and methods used in this study, as well as the results obtained, are considered valuable. This study is expected to provide practical benefits for engineering applications in addition to its academic significance and is intended to serve as a resource for decision-makers and road safety experts.