Air pollution remains a pressing challenge in urban environments, and understanding the impact of traffic is key to developing effective mitigation strategies. This new study explores the hourly relationship between traffic volume and the concentrations of two major air pollution indicators: nitrogen dioxide (NO₂) and particulate matter (PM₁₀).
Using high-resolution data from two monitoring stations in Helsinki, the researchers applied the Prophet time series model to forecast hourly traffic trends for 2024. These projections were then compared against annual average NO₂ and PM₁₀ levels. The study employed polynomial regression and cross-correlation analysis to identify temporal patterns and evaluate the strength and timing of the relationship.
The findings reveal a strong correlation between traffic volume and air pollution concentrations, particularly at the high-traffic Mäkelänkatu supersite, with minimal time delay observed. Root mean square error (RMSE) values and model fit assessments further validated the predictive power of traffic data in estimating NO₂ and PM₁₀ trends.
This work highlights the value of traffic-based proxy models as practical, data-driven tools for real-time air quality monitoring and for guiding targeted interventions to improve urban air quality.
We’re pleased to announce that the URBREATH project has been acknowledged in a recent preprint titled “A Proxy Model for Traffic Related Air Pollution Indicators Based on Traffic Count”.
Air pollution remains a pressing challenge in urban environments, and understanding the impact of traffic is key to developing effective mitigation strategies. This new study explores the hourly relationship between traffic volume and the concentrations of two major air pollution indicators: nitrogen dioxide (NO₂) and particulate matter (PM₁₀).
Using high-resolution data from two monitoring stations in Helsinki, the researchers applied the Prophet time series model to forecast hourly traffic trends for 2024. These projections were then compared against annual average NO₂ and PM₁₀ levels. The study employed polynomial regression and cross-correlation analysis to identify temporal patterns and evaluate the strength and timing of the relationship.
The findings reveal a strong correlation between traffic volume and air pollution concentrations, particularly at the high-traffic Mäkelänkatu supersite, with minimal time delay observed. Root mean square error (RMSE) values and model fit assessments further validated the predictive power of traffic data in estimating NO₂ and PM₁₀ trends.
This work highlights the value of traffic-based proxy models as practical, data-driven tools for real-time air quality monitoring and for guiding targeted interventions to improve urban air quality.
Read the full preprint here.