Study of the ADMS–URBAN model for assessing air pollution due to the vehicular traffic in Ravenna

The Environmental Regional Agency (Arpa) of the Emilia – Romagna Region has chosen the ADMS – Urban model as a tool to support the Air Quality Assessment. We used the model to calculate the atmospheric dispersion of pollution originated from vehicular traffic in Ravenna urban area. We input different emission factor’s, background levels’ and meteorological datasets in order to identify the model best performance conditions. This paper describes the comparisons of the CO and PM10 concentrations predicted using the modelling system with the data of the urban air quality monitoring network. We performed a statistical analysis to determine the agreement between predicted and measured time series of concentrations at four permanently located stations in the centre of the town over 2004 - 2005. The overall agreement between measured and predicted daily concentrations is good for CO (the index of agreement, the fractional bias, the normalized mean square error vary from 0.61 to 0.64, from -0.28 to -0.29, from 0.55 to 0.68, respectively, while the factor two values range from 62 to 63% at the stations mainly influenced by the vehicular traffic). The results are not so good as to PM10: the model underpredicts the measured PM10 concentrations at all the stations. We also performed a diagnostical analysis, studying the model performance’s dependence on meteorological parameters, as atmospheric stability and wind speed and direction. The model performance decreases at the highest wind speeds (u>4 m/s) because the model underpredicts the measured CO daily concentrations. For wind speeds higher than 2 m/s, the modelling system performance depends on the wind direction and this dependence can be connected with the orientation of the major streets located next to the monitoring stations. Any atmospheric stability dependence of the model performance is not observed.