Atmospheric Environment 41 (2007) 173–188 Evaluation and intercomparison of Ozone and PM10 simulations by several chemistry transport models over four European cities within the CityDelta project R. Vautard a,b , P.H.J. Builtjes c , P. Thunis d , C. Cuvelier d,Ã , M. Bedogni e , B. Bessagnet f , C. Honore´ f , N. Moussiopoulos g , G. Pirovano h , M. Schaap c , R. Stern i , L. Tarrason j , P. Wind j a Laboratoire de Me´te´orologie Dynamique, IPSL, Ecole Polytechnique, 91128 Palaiseau, France b LSCE/IPSL laboratoire CEA/CNRS/UVSQ, Gif sur Yvette, 91191 Cedex, France c TNO Built Environment and Geosciences, Apeldoorn, The Netherlands d European Commission-DG Joint Research Centre, Institute for Environment and Sustainability, I-21020 Ispra, Italy e Mobility & Environment Agency, via Beccaria 19– 20122 Milano, Italy f INERIS, Parc Halata, Verneuil en Halatte, France g Laboratory of Heat Transfer and Environmental Engineering, Aristotle University, 54 124 Thessaloniki, Greece h CESI Ricerca S.p.A., via Rubattino 54– 20134 Milano, Italy i Free University Berlin, Inst. of Meteorology, Carl-Heinrich-Becker-Weg 6-10, D-12165 Berlin, Germany j Norwegian Air Pollution Institute, Air Pollution Section/Research Department, P.O. Box 43, Blindern, NO-0313 Oslo, Norway Received 7 March 2006; received in revised form 9 July 2006; accepted 19 July 2006 Abstract The CityDelta project Cuvelier et al. [2006. CityDelta: a model intercomparison study to explore the impact of emission reductions in European cities in 2010. Atmospheric Environment] is designed to evaluate the air quality response of several emission abatement scenarios for 2010 at the scale of the European continent, and specifically in the areas where most people live: the cities. Before evaluating this response, the model simulations in a control case must be evaluated against observations in order to understand their main strengths and weaknesses. In this article six different models are used to simulate a full year (1999) of air quality pollutant concentrations over domains encompassing a large area around four major European cities: Berlin, Milan, Paris and Prague. Three models are used both at large-scale (typically 50 km) and small-scale resolution (5 km). The intercomparison of the simulation results for ozone and particles smaller than 10 microns (PM10) leads to the following conclusions: (i) Models capture fairly well the mean, daily maxima and variability of ozone concentrations, as well as the time and intercity variability. However, a significant overestimation of ozone in city centres is found especially for large-scale models. (ii) PM10 simulation skill is generally poor, and large-scale models underestimate their mass. The difference between Milan (highly polluted) and the other cities is not reproduced. All models have difficulties in capturing the observed seasonal variations. ARTICLE IN PRESS www.elsevier.com/locate/atmosenv 1352-2310/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2006.07.039 Ã Corresponding author. Tel.: +39 0332 785075; fax: +39 0332 786291. E-mail address: kees.cuvelier@jrc.it (C. Cuvelier).