Quantitative assessment of supercooled liquid water sensitivity to different aerosol field inputs over the Sichuan Basin

Abstract. Aerosol-cloud interactions profoundly influence the properties of supercooled liquid water, which in turn play a critical role in aircraft icing. However, accurately quantifying aerosol emission inventories and their spatiotemporal distributions remains a major challenge. In this study, the Thompson-Eidhammer aerosol-aware microphysics scheme is applied to an in-flight icing event over the high-aerosol-concentration environment of the Sichuan Basin, China. Three numerical experiments with different initial aerosol number concentrations are conducted: Default, Climatology, and Copernicus Atmosphere Monitoring Service reanalysis (CAMS), representing clean and polluted conditions. All three experiments successfully reproduce the synoptic-scale spatial distribution of supercooled liquid water. Compared with the clean environment, the polluted scenarios simulate higher supercooled liquid water mass mixing ratios, greater cloud droplet number concentrations, smaller median volume diameters, and longer cloud system lifetimes. The experiments also reveal that stronger auto-conversion process in clean conditions suppresses supercooled liquid water formation, whereas enhanced accretion process in polluted environments promotes supercooled liquid water depletion. Comparison with in situ aircraft observations indicates that, among the three numerical experiments, the CAMS experiment performs best in capturing high supercooled liquid water contents and large median volume diameters. These findings highlight the importance of real-time aerosol input for improving the simulation of aerosol-cloud interactions and supercooled liquid water characteristics.