Page 377 PM10 LONG-TERM ASSESSMENT OF EMISSION REDUCTION SCENARIOS OVER NORTHERN ITALY Elisabetta Angelino 1 , Marco Bedogni 2 , Claudio Carnevale 3 , Enrico Minguzzi 4 , Edoardo Peroni 1 , Cesare Pertot 5 and Guido Pirovano 5 1 Regional Agency for Environmental Protection of Lombardia, Milan, Italy 2 Mobility and Environment Agency of Milan, Milan, Italy 3 DEA, University of Brescia, Italy 4 Regional Agency for Environmental Protection of Emilia-Romagna, Bologna, Italy 5 CESI Spa, Milan, Italy INTRODUCTION Air quality assessment and policies analysis show an increasing interest in long term simulations with Chemical Transport Models (CTMs). In this frame the CityDelta open model inter-comparison exercise (http://rea.ei.jrc.it/netshare/thunis/citydelta/) has been organized by the Joint Research Centre (JRC-IES) of Ispra, in collaboration with EMEP, IIASA and EUROTRAC, as a contribution to the modelling activities in the CAFE (Clean Air For Europe) Project (6 th Framework Programme). The aim of CityDelta exercise was to compare the results of different photochemical dispersion models in order to estimate air quality response to local and global emissions variations. The exercise was carried out by twenty scientific groups working on eight European domains. Starting in 2002, the second phase of CityDelta (2003-2004) was focused on estimations of PM 10 and PM 2.5 concentrations fields, with a special attention devoted to the partitioning of their main inorganic components (nitrates, sulphates and ammonium) and to the response to emission reduction scenarios. DOMAIN AND METHODOLOGY The computational domain (300 x 300 km 2 ), centred on Milan (North of Italy), includes a large flat area (all the central and most of the western Po Valley) surrounded by mountains along three sides. It is a complex test bench in which low wind speeds and intense solar radiation cause high ozone episodes during summer and critical PM concentration levels during winter. The modelling system employed for the simulations was composed by a common set of pre-processors for input data (emissions, boundary conditions and meteorological fields) and two different numerical CTM models. To perform the simulations, the domain has been horizontally divided into 5x5 km 2 cells and vertically in 11 varying level ranging from 20 to 3900 m. In order to keep the results homogeneous as much as possible, the models shared the same input fields, derived from the same data provided by JRC to all participants. More details concerning the computational domain and the modelling system are reported on Angelino, E. et al. (2005). The chemical and transport models used for the simulations were CAMx (ENVIRON Corp., 2004) and TCAM (Decanini, E. and M. Volta, 2003). The Comprehensive Air quality Model with extensions (CAMx) provides the option of using two different chemical mechanisms: SAPRC99 (Carter, W.P.L., 2000) and CBIV (Gery, M. et al., 1989) 1999 version, modified to model ozone and fine/coarse PM, using RADM (Chang, J.S. et al., 1987) mechanism for aqueous phase chemistry, ISORROPIA (Nenes, A. et al., 1998) for inorganic sulphate-nitrate- ammonium chemistry, SOAP (Strader, R. et al., 1999) semi-volatile scheme for secondary organic aerosols. CAMx user can choose the gas-phase chemistry solver as either IEH (Implicit-Explicit Hybrid) (Sun, P. et al., 1994) or CMC (ENVIRON Corp., 2004), based on an "adaptive-hybrid" approach. TCAM is a multi-phase Eulerian 3D model. It implements