Effective 2-D product classification for near-explicit gas-aerosol systems: Tools and applications for efficient modeling of complex aerosol systems
Student Seminar Series
Department of Atmospheric & Oceanic Sciences
presents
a talk by
Dalrin Ampritta Amaladhasan
PhD student
Effective 2-D product classification for near-explicit gas-aerosol systems: Tools and applications for efficient modeling of complex aerosol systems
Secondary organic aerosol (SOA) material makes up a sizable fraction of the total atmospheric aerosol mass. Gas–particle partitioning is a key process in the formation and evolution of secondary organic aerosol (SOA). Our work builds on a thermodynamic equilibrium gas–particle partitioning model based on AIOMFAC. In principle, this framework allows for predicting SOA formed for any well-characterized organic precursor when a system of representative surrogate components can be established. The objective of this study is the development of a flexible modular gas–aerosol box model with automatic product classification to facilitate the modeling of complex close-to-reality aerosol systems. This can be attained by the development of tools for automating aerosol property predictions used as inputs in the model. These tools use cheminformatics-based SMILES/SMARTS libraries library to process thousands to millions of compounds for functional group classification as well as pure component vapor pressure estimation for any precursor system and interfaced with the model; 2-dimensional (2-D) product lumping schemes to have an adjustable degree of species in the system are implemented as the next crucial part of improving the gas-aerosol model efficiency. We explore the capability of a different polarity vs. volatility classification coordinates including AIOMFAC-predicted activity coefficient ratios in distinct solvents as a source of physicochemical information on the mixing of system components. This metric can be used for representing SOA formed for any given aerosol system and hence in lumping / simplifying a complex multicomponent aerosol system into an adjustable number of species based on a selection of lumping schemes to evaluate the effect of each lumping scheme. classification schemes of varying complexity and resolution were tested and that we will discuss the differences and capabilities of those approaches using monoterpene-derived SOA systems. A discussion of the results showing the capability of the property prediction tool as well the 2-D product lumping scheme will be elaborated.
Wednesday Jan 15/ 2.30 PM/ Room 934 Burnside Hall