S2AS v1.0 and 2D polarity–volatility lumping framework v1.0: automated compound classification and scalable lumping for organic aerosol modelling

Abstract. Advancements in near-explicit chemical reaction mechanisms, such as the Master Chemical Mechanism (MCM) or the Generator of Explicit Chemistry and Kinetics of Organics in the Atmosphere (GECKO-A), have enabled highly detailed simulations of atmospheric chemistry. Such simulations offer a bottom-up approach to accompany and inform laboratory chamber experiments of organic aerosol formation or to model the complex chemistry of mixtures of volatile aerosol precursors for specific tropospheric conditions. These chemical reaction mechanisms, while comprehensive, generate hundreds to millions of organic components, creating computational challenges for subsequent applications in multiphase equilibrium gas–particle partitioning models to predict secondary organic aerosol (SOA) mass concentrations, phase compositions, and hygroscopicity. The wealth of simulated reactions and components also requires substantial simplifications for reduced-complexity representations in large-scale atmospheric models. This study introduces a suite of software tools to automate relevant pure-component property predictions as well as a 2-dimensional (2D) polarity–volatility lumping framework to systematically reduce the complexity of chemical mechanism outputs. We introduce a new polarity metric for use in the 2D framework, a ratio of a component's activity coefficients in water and an organic solvent (hexanediol). This ratio is computed using the Aerosol Inorganic–Organic Mixtures Functional groups Activity Coefficients (AIOMFAC) model. The 2D framework offers grid-based and cluster-based methods to select an adjustable number of surrogate species and offers flexibility in the choice of polarity axis. Our methods utilize the Simplified Molecular Input Line Entry System (SMILES) description of molecular structures. A new tool, SMILES to AIOMFAC subgroups (S2AS), is introduced to automatically generate AIOMFAC-model input files and to handle exception cases consistently. We demonstrate the application of our framework using systems of hundreds to thousands of components generated by near-explicit chemical mechanisms. The new framework enables tailored reduced-complexity representations of gas–particle systems.