The model SIRANE for atmospheric urban pollutant dispersion; PART II, validation of the model on a real case study

摘要

We analyse the performance of the model SIRANE by comparing its outputs to field data measured within an urban district. SIRANE is the first urban dispersion model based on the concept of street network, and contains specific parametrical law to explicitly simulate the main transfer mechanisms within the urban canopy. The model validation is performed by means of field data collected during a 15 days measurement campaign in an urban district in Lyon, France. The campaign provided information on traffic fluxes and cars emissions, meteorological conditions, background pollution levels and pollutant concentration in different location within the district. This data set, together with complementary modelling tools needed to estimate the spatial distribution of traffic fluxes, allowed us to estimate the input data required by the model. The data set provide also the information essential to evaluate the accuracy of the model outputs. Comparison between model predictions and field measurements was performed in two ways. By evaluate the reliability of the model in simulating the spatial distribution of the pollutant and of their time variability. The study includes a sensitivity analysis to identify the key input parameters influencing the performance of the model, namely the emissions rates and the wind velocity. The analysis focuses only on the influence of varying input parameters in the modelling chain in the model predictions and complements the analyses provided by wind tunnel studies focussing on the parameterisation implemented in the model. The study also elucidates the critical role of background concentrations that represent a significant contribution to local pollution levels. The overall model performance, measured using the Chang and Hanna (2004) criteria can be considered as 'good' except for NO and some of BTX species. The results suggest that improvements of the performances on NO require testing new photochemical models, whereas the improvement on BTX could be achieved by correcting their vehicular emissions factors.