COMPARISON OF AIR QUALITY MEASUREMENTS WITH MICROSCALE SIMULATION OF ATMOSPHERIC DISPERSION OF VEHICULAR EMISSIONS BASED ON AUTOMATIC TRAFFIC FLOW DATA

Abstract

Assessing vehicular traffic contribution to urban air pollution is a key information for population and policymakers. A microscale simulation of dispersion of vehicular traffic emissions was performed for a busy crossing within the inner ring road in central Reggio Emilia, a mid-size city in the central Po valley, Italy. The investigated area includes a station of the regional air quality monitoring network: the station is classified as “urban traffic” station, i.e. is expected to be mainly influenced by vehicular traffic emissions. Microscale simulation models are best suitable for dispersion modelling within urban areas, where concentration largely depend on air stagnation due to buildings and obstacles. Micro-Swift-Spray (Aria Technologies, Francia e Arianet, Italia), a Lagrangian particle dispersion model directly derived from the local-scale Spray model, has been used to simulate the dispersion of vehicular traffic emissions (as NOx and CO) accounting also for building and obstacles. A locally installed radar traffic counter provided vehicle speed and length continuously over 12 days (13-24 January 2014). These data have, along with fleet composition details, have been used to calculate tailored emission factors depending on vehicle type, fuel, speed and EURO emission standard. Simulated concentration fields were produced and results were compared to local air quality measurements: correlation between simulated and observed hourly NOx concentration resulted extremely satisfactory (r=0.86). Comparison with air quality measurements in urban background conditions allowed to better assess model’s performance and the amount of primary and secondary nitrogen dioxide at the investigated site