Aerosol chemical and optical properties over the Paris area within ESQUIF project

Abstract. Aerosol chemical and optical properties are extensively investigated for the first time over the Paris Basin in July 2000 within the ESQUIF project. The measurement campaign offers an exceptional framework to evaluate the performances of the chemistry-transport model CHIMERE in simulating concentrations of gaseous and aerosol pollutants, as well as the aerosol-size distribution and composition in polluted urban environments against ground-based and airborne measurements. A detailed comparison of measured and simulated variables during the second half of July with particular focus on 19 and 31 pollution episodes reveals an overall good agreement for gas-species and aerosol components both at the ground level and along flight trajectories, and the absence of systematic biases in simulated meteorological variables such as wind speed, relative humidity and boundary layer height as computed by the MM5 model. A good consistency in ozone and NO concentrations demonstrates the ability of the model to reproduce the plume structure and location fairly well both on 19 and 31 July, despite an underestimation of the amplitude of ozone concentrations on 31 July. The spatial and vertical aerosol distributions are also examined by comparing simulated and observed lidar vertical profiles along flight trajectories on 31 July and confirm the model capacity to simulate the plume characteristics. The comparison of observed and modeled aerosol components in the southwest suburb of Paris during the second half of July indicates that the aerosol composition is rather correctly reproduced, although the total aerosol mass is underestimated by about 20%. The simulated Parisian aerosol is dominated by primary particulate matter that accounts for anthropogenic and biogenic primary particles (40%), and inorganic aerosol fraction (40%) including nitrate (8%), sulfate (22%) and ammonium (10%). The secondary organic aerosols (SOA) represent 12% of the total aerosol mass, while the mineral dust accounts for 8%. The comparison demonstrates the absence of systematic errors in the simulated sulfate, ammonium and nitrates total concentrations. However, for nitrates the observed partition between fine and coarse mode is not reproduced. In CHIMERE there is a clear lack of coarse-mode nitrates. This calls for additional parameterizations in order to account for the heterogeneous formation of nitrate onto dust particles. Larger discrepancies are obtained for the secondary organic aerosols due to both inconsistencies in the SOA formation processes in the model leading to an underestimation of their mass and large uncertainties in the determination of the measured aerosol organic fraction. The observed mass distribution of aerosols is not well reproduced, although no clear explanation can be given.