Aerosol modeling with CHIMERE -: preliminary evaluation at the continental scale

Aerosol modeling is a challenging scientific problem aimed at improving our knowledge in the many complex processes involved in multiphase chemistry and transport. Correct simulations of aerosols are also required in order to elaborate particle emission reduction strategies. The CHIMERE chemistry transport model (Atmos. Environ. 35 (2001) 6277) has been improved to account for particle transport, formation, deposition at the European scale. The aerosol model accounts both for inorganic (NO3-, SO42-, NH4+) and organic species of primary or secondary origin. Secondary organic aerosols from biogenic and anthropogenic gas precursors are partitioned into gas and particulate phases through a temperature dependent partition coefficient. The modeling approach is presented in this paper with preliminary simulation results over Europe. Comparisons with available data at background stations give acceptable results on PM10, with correlation coefficients usually exceeding 0.5 and normalized errors in the 30-80% range in many regions. However, results on sulfate, nitrate and ammonium species display less correct error statistics. Comparisons on sulfate concentrations give normalized errors in the range 30-80% in summer and less correct in winter. Temporal correlation coefficients usually range from 0.30 to 0.70. Nitrate concentrations are better simulated during winter than during summer. Difficulties in simulating heterogeneous and aqueous phase processes could explain model deficiencies. Moreover, temperature dependence of gas/particle partitioning processes for nitrate, ammonium and secondary organic species could mainly explain the seasonal variability of biases. Model deficiencies are observed in Southern countries, certainly due to natural dust emissions and resuspended particles. Finally, sea salts seems to have a quite significant influence on error statistics in coastal areas. (C) 2004 Elsevier Ltd. All rights reserved.