Evaluating traffic-related air pollution in urban areas: A case study of Warsaw using the ATMO-Street model chain

Urban air quality is a major concern due to high population density and pollutant levels. To develop effective mitigation strategies, distinguishing between local and external emission sources is crucial. In Polish cities, residential heating is the primary source of particulate matter emissions, with transportation also playing a significant role. This study assesses the impact of the transport sector on Warsaw’s air quality by analysing PM10, PM2.5, and NO2 concentrations. A comprehensive approach was applied, integrating emissions from traffic, road dust resuspension, and residential heating. These emissions were estimated using a bottom-up approach based on detailed road and point source data. The ATMO-Street model chain was employed, combining the GEM-AQ model for background concentrations with the IFDM bi-Gaussian dispersion model and the OSPM street-canyon module. Model validation using data from nine monitoring stations demonstrated that including the OSPM module significantly improved NO2 predictions at the traffic station, enhancing GEM-AQ and IFDM model performance by 62% and 46%, respectively. The results showed that traffic contributes 84% of NO2 concentrations at this location. Furthermore, incorporating the street-canyon effect and resuspension emissions—previously excluded—improved model accuracy for PM2.5 and PM10 by 34% and 55%, respectively, compared to GEM-AQ. The study revealed that traffic contributes up to 41% of PM2.5 and 42% of PM10 at the traffic station. These findings underscore the importance of including road dust resuspension and street-canyon effects in urban air quality modelling. • First-time coupling of GEM-AQ with the ATMO-Street (IFDM + OSPM) model chain. • Novel use of detailed, bottom-up emissions, explicitly separating road dust resuspension, traffic, and residential heating. • Quantified road dust resuspension as a key PM10 source in street canyons, resolving seasonal model underestimations. • First meticulous and detailed validation of ATMO-Street for PM10 and PM2.5, showing major accuracy improvements.