Variability of submicron ice-nucleating particles at urban and rural sites in Japan related to aerosol chemical components

Ice-nucleating particles (INPs) play a pivotal role in cloud formation and precipitation, yet predicting their concentrations remains difficult due to limited understanding of their spatiotemporal variability and characteristics. Therefore, it is important to provide insights into how aerosol properties and meteorological parameters are related to INP concentrations. This study investigated INP concentrations, their heat sensitivity, particle chemical components, and meteorological parameters at urban (Yokohama) and rural (Noto) sites in Japan during Dust and Non-Dust periods. While no single key factor consistently explained INP concentrations, relatively strong associations were observed with factors implying particles such as mineral dust, heat-sensitive biological particles, and Pb-containing particles from coal and waste combustion, with variations depending on site and season. During the Dust period, INP concentrations at both sites were correlated with the mineral indicators ( R > 0.57), though INPs above −15 °C remained low at Noto site ( R = 0.49). In the Non-Dust period, mineral indicators cannot account for INP levels at either site ( R < −0.31). Heat treatment-induced deactivation of INPs above −15 °C in the urban area suggested a close association between biological INPs and mineral particles. Although submicron INPs account for only a small fraction of total INPs at the surface (e.g., < 6.5 % at −12.5 °C), the particle types that contributed to the variability in INP concentrations were nevertheless supported previous qualitative findings. Our findings also highlight the complexity of quantitatively linking environmental factors that vary by region and time to INP concentration variations to characterize aerosol–cloud interactions. • Submicron INPs were linked to some typical INP types, varying by site and period. • Submicron INPs were correlated by mineral components only during the dust period. • Only the urban site showed heat-sensitive INPs, indicating nearby biological sources. • Results highlight limitations of single predictors for INPs across sites and seasons.