An Impact of Traffic Characteristics on Crash Frequency

Al-Nuaimi, Atheer; Jameel, Abeer K. · 2023 · Crossref

DOI: 10.1051/e3sconf/202342703040

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the impact of specific traffic characteristics on crash frequency, addressing the critical need for improved road safety in highway networks. Motivated by rising global and local road fatality rates, which are included in the United Nations’ Sustainable Development Goals, the research focuses on Iraq, where road fatalities have increased significantly. While various factors influence safety, this paper specifically examines the relationship between average daily traffic (ADT) and the percentage of heavy vehicles (HVP) and crash occurrences. The objective was to develop a safety performance function (SPF) that quantifies these effects for rural roads, specifically using the Old Baghdad–Baquba road as a case study. The methodology employed a Generalized Linear Model (GLM), specifically the Poisson Gamma regression, to account for over-dispersion in crash data. The study area, a 70 km two-lane, two-way rural road, was divided into 1 km segments. Two primary datasets were utilized: crash data collected from traffic departments in Baghdad and Diyala provinces for the years 2016–2019, and traffic volume data collected over eight hours across ten weekdays. Traffic volumes were converted to equivalent passenger cars to determine ADT. Before model development, the researchers checked for multicollinearity between independent variables using the Variance Inflation Factor (VIF), ensuring values remained below the acceptable threshold of 5. The analysis was conducted using The R Project for Statistical Computing. The results demonstrated statistically significant positive correlations between both traffic characteristics and crash frequency. The Poisson Gamma model indicated that ADT is a significant factor at a 95% confidence level, with a coefficient of 0.82 for the natural log of ADT. Similarly, HVP showed a significant positive association with crash frequency (coefficient of 0.17, p-value 0.000327). Diagnostic plots, including residuals versus fitted values and Q-Q normal plots, confirmed that the data followed the assumed Poisson Gamma distribution. The authors attribute the increased crash frequency associated with heavy vehicles to the higher number of overtaking maneuvers required, which leads to more head-on and run-off-road crashes on two-lane roads. The significance of this study lies in its provision of a validated safety performance function for rural roads in Iraq, offering practitioners a tool to understand how traffic volume and composition affect safety. The findings reinforce the notion that higher traffic volumes and heavy vehicle percentages increase crash risks on two-lane rural highways. The authors conclude by recommending the development of similar SPFs for other road networks to allow for comparative analysis across jurisdictions. They also suggest future research compare the Poisson Gamma model with other techniques, such as Gaussian regression, to further refine crash prediction capabilities.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• incidence prevalence

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: crash risk outcomes
• Methodological Resource: dataset resource