Abstract Number: 46 Working Group: Aerosol Modeling
Abstract Secondary Organic Aerosols (SOA) account for a significant part of particulate matter (PM). Simulating their formation and fate remains challenging, since to date, air quality models usually underestimated them. SOA formation processes include multistep heterogeneous mechanisms, and their parametrizations involve several variables such as kinetic data and physical properties not well known. A better assessment of the SOA composition notably requires the understanding of the thermodynamic equilibrium of SOA compounds between the phases involved. This is usually done using either, theoretical, or experimental data. However, the validation of the modelling parametrizations developed must be done by comparison between measurement and modeling data. In this work, various key molecules, well known as SOA markers, have been modeled, using CHIMERE air quality model, through detailed formation pathways including the calculation of their gas/particle partitioning. Modelled concentration values have been compared to measurements of biogenic, (e.g. pinonic acid, pinic acid and MBTCA: α-pinene oxidation by-products), and anthropogenic (e.g. DHOPA: toluene SOA marker) SOA markers, performed on both, gaseous and particulate phases. Mechanisms have been developed based on data obtained from the Master Chemical Mechanism (NCAS, Universities of Leeds and York) and the scientific literature. The gas/particle partitioning has been calculated by using saturated vapor pressure and Henry’s constant values through the thermodynamic model SOAP. The gas phase chemistry has been simulated using MELCHIOR2. Biogenic emissions have been computed with the MEGAN 2.1 algorithm. The comparison has been done with SOA marker measurements performed for a one year period (2015) at the SIRTA sampling site (25 km SW from Paris city center). Filters and PUF (polyurethane foam) samples were collected every third day (24 h sampling) and then analysed by GC-MS after derivatization using native standards. The capacity of the model to reproduce seasonal variations of concentrations and the gas/particle partitioning (influenced by several parameters such as humidity and temperature) was evaluated.