Vehicular exhaust impact simulated at microscale from traffic flow automatic surveys and emission factor evaluation

Vehicular emissions are a large NOx and CO source in Italian urban areas. In order to assess the impact of heavy traffic roads on local air quality a micro-scale simulation of pollutant concentration fields was produced. The investigated areas are in downtown of Reggio Emilia and Modena, two cities in central Po valley, Italy, and focused on high traffic intersections. An urban traffic station of the regional air quality monitoring network is present in both investigated areas, where traffic is expected to be the main local source of atmospheric pollutants. The simulation has been performed by the micro-scale model suite Micro-Swift-Spray (Aria Technologies, France and ARIANET, Italy) a Lagrangian particle dispersion model directly derived from the SPRAY code, able to account for buildings and obstacles. Simulated pollutants are NOx and CO, as main tracers of combustion emissions. Direct measurements of traffic flow have been continuously collected for 12 day survey periods (in Reggio Emilia from January 13 to 24, 2014 by a two channel doppler radar traffic counter and in Modena from October 28 to November 8, 2016 by four one channel doppler radar traffic counters) and used for the hourly modulation of vehicular emissions. Specific emission factors were obtained by the combination of radar counts with vehicular fleet composition for each municipality: these depend on vehicle type, fuel type, speed and EURO category and were calculated according to the EMEP/EEA guidelines for air pollutant emission inventory. Simulated concentration fields were evaluated over the period with direct traffic counts for the two studied areas: for both areas the results were compared to local air quality measurements collected at the traffic urban monitoring stations and also at the respective urban background stations. The simulated NOx hourly concentrations show a very large agreement with the observations, even if they result underestimated compared to the observed atmospheric concentrations at the traffic site. Simulated and observed concentrations show a fair agreement for CO. The results outline the representativeness of air quality stations in characterizing the sites for pollution level and for dominant pollutant sources.