Modeling the Relationship Between Traffic Volume and Air Pollution Levels in the City of Zintan, Libya: A Seasonal Analytical Study

Abstract

Background: Traffic-related air pollution is a pressing environmental and public health concern in rapidly urbanizing Libyan cities. However, quantitative studies addressing seasonal variations, vehicle classification, and meteorological controls remain scarce, particularly for inland mountainous cities.
Objective: This study quantifies the statistical association between traffic volume---disaggregated by passenger cars and heavy trucks---and concentrations of PM₂.₅, NO₂, and CO in Zintan, Libya, across four seasons, while integrating meteorological variables.
Methods: Traffic and pollutant data were collected from six urban sites during morning, midday, and evening periods over 28 days (seven days per season), yielding 504 observations per variable. Meteorological data were measured using a portable weather station. Statistical analyses included descriptive statistics, Pearson correlation, multiple linear regression with 10 fold cross validation, two way ANOVA, and GIS mapping.
Results: Strong positive correlations were found between vehicle count and all three pollutants (PM₂.₅: r = 0.87, NO₂: r = 0.84, CO: r = 0.85; p < 0.01). The Pearson correlation coefficient between total vehicle count and PM₂.₅ (r = 0.87) aligns with the R² of the traffic-only regression model (R² = 0.76, as r² = 0.76), confirming internal consistency. Heavy trucks (≈9% of fleet) showed a statistical association approximately 12 times higher with PM₂.₅ concentration per vehicle compared to passenger cars. Important interpretive note: This estimated 12 fold difference may represent an upper bound, as this study could not disentangle truck traffic from correlated wintertime sources such as residential generators and waste burning. Furthermore, this ratio is derived from marginal regression coefficients under the assumption of ceteris paribus (all other variables held constant), and may not directly translate to real-world emission equivalence when traffic composition changes dynamically.* Winter recorded the highest pollution levels (PM₂.₅: 65–102 µg/m³) despite 14% lower traffic volume than summer, due to a shallow boundary layer (400–700 m) and thermal inversions. The final regression model (classified traffic + meteorology) explained 88% of the variance in PM₂.₅ (R² = 0.88, RMSE = 4.0 µg/m³). A significant interaction between season and vehicle type was identified (p < 0.001).
Conclusion: Traffic volume is strongly associated with pollution levels in Zintan, with heavy trucks showing a disproportionately large statistical association. Winter meteorological conditions exacerbate pollution levels. The study recommends conducting a focused truck-specific diurnal pattern study, and if peak hours are confirmed, then considering truck restrictions during winter peak hours, alongside developing a weather based early warning system