Evaluating the impact of urban traffic patterns on air pollution emissions in Dublin: a regression model using google project air view data and traffic data

Tafidis, Pavlos; Gholamnia, Mehdi; Sajadi, Payam; Vijayakrishnan, Sruthi Krishnan; Pilla, Francesco · 2024 · DOAJ

DOI: 10.1186/s12544-024-00671-z

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 complex relationship between urban traffic patterns and air pollution emissions in Dublin, Ireland, aiming to provide data-driven insights for urban planning and policy. Motivated by Dublin’s status as a highly congested city with air quality levels that frequently exceed World Health Organization guidelines despite meeting EU legal limits, the research seeks to bridge the gap in understanding the specific dynamics between traffic and pollution in this context. The authors argue that a multifaceted approach is necessary, as traffic is not the sole determinant of urban air quality. The methodology employs high-resolution datasets from May 2021 to April 2022, combining Google Project Air View (GPAV) street-level air quality measurements with hourly vehicle counts from Dublin City Council’s SCATS system. Meteorological data from Met Éireann was also integrated to account for weather variables. The study area focused on 29 intersections in Dublin’s inner and outer city sectors. To analyze the relationships, the researchers utilized Pearson’s correlation coefficients and two machine learning regression models: Support Vector Regression (SVR) and Gaussian Process Regression (GPR). The models were trained on traffic volume, meteorological factors, and air quality data, with performance evaluated using Root Mean Square Error (RMSE) and R-squared ($R^2$) values. Spatial averaging was applied at varying distances from the study center to account for spatial variability. The results revealed distinct seasonal patterns in pollutant concentrations, such as elevated nitrogen dioxide ($NO_2$) levels from September to December and increased ozone ($O_3$) during spring months. However, the anticipated direct correlation between vehicular traffic and air pollution was not pronounced. Pearson’s correlation analysis showed no significant linear relationships between traffic counts and pollutants. In model performance, GPR generally outperformed SVR, particularly for particulate matter ($PM_{2.5}$), achieving $R^2$ values between 0.40 and 0.55 at specific distances. Despite these moderate fits, the models left considerable variance unexplained, indicating that traffic and weather alone do not fully capture the complexity of urban emissions. The lack of a clear pattern in carbon dioxide ($CO_2$) emissions relative to traffic volumes further suggested the influence of non-vehicular sources, such as residential heating and industrial activities. The study concludes that urban air quality is influenced by a multitude of factors beyond traffic volume, necessitating comprehensive policymaking that addresses broader emission sources. The superior performance of GPR highlights the importance of accounting for spatial variability and uncertainty in air quality modeling. While the findings are specific to Dublin, the methodological framework offers theoretical support for traffic planning and control in other urban centers globally. The authors acknowledge limitations regarding the exclusion of other potential variables, such as public transport usage and industrial activities, suggesting future research should incorporate these factors to refine predictive models.

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 2026-06-25; verification: verified.

Topics

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

• regulatory evaluation