Source apportionment and parameterization of ice nucleating particles observed at a high-altitude station in the north-eastern Mediterranean in autumn 2021 during the CALISHTO campaign

Aerosol source apportionment improves the understanding of aerosol-cloud interaction processes and benefits the parameterization of ice nucleating particles (INPs), which also contributes to the predictability of climate models for quantifying the impacts of aerosols on the changing climate. This study, which took place in the frame of the Cloud-Aerosol InteractionS in the Helmos background TropOsphere (CALISHTO) campaign, investigates the interactions between mixed-phased clouds and aerosol particles at Helmos Mt. in Peloponnese, Greece (north-eastern Mediterranean). The source apportionment of INPs originating from different aerosol sources is achieved by identifying exclusive characteristics of relevant air masses. A synergy of measurement techniques was employed, including in-situ measurements for INP number concentration and aerosol property characterization, remote sensing techniques for atmospheric condition observations, as well as modelling simulations for calculating aerosol particle footprints.

The number concentration of INPs was observed in the mixed-phase cloud regime (>−27°C) in both the planetary boundary layer (PBL) and the free troposphere (FT). The results show that one in a million of aerosol particles can serve as INPs under the background condition in FT. The presence of precipitation/clouds may enrich INPs by suspending biological particles from near ground sources or releasing cloud-processed particles when the observation site is above PBL. The intrusion of remotely transported air masses leads to increased INPs for conditions above PBL, suggesting the observed INPs are of both local and remote origins. In addition, the INP abundance of different sources spans a range of three orders of magnitude and increases following the order of marine aerosols, continental aerosols, and then dust plumes. Biological particles are approximate to INPs observed in continental and marine aerosols, whereas mineral dust particles dominate the observed INPs when dust plumes are present. Furthermore, a case study on a calendar day was performed to investigate the effects of precipitation/clouds on INP abundance in the PBL. In contrast observations above the PBL, the presence of precipitation/clouds may lead to wet removal of aerosol particles and thus, decreased INPs.

Statistical analysis suggests that INP concentration in the mixed-phase cloud regime is significantly correlated with fluorescent particles, including biological and non-biological particles such as dust particles associated with fluorescent materials. The ratio of fluorescent to nonfluorescent particles and the ratio of coarse (>1.0 μm) to fine (<1.0 μm) particles are also found to be significantly correlated with observed INPs from different aerosol sources. Such properties further constrain the ice formation ability of aerosol particles showing fluorescence and are then used to improve the parameterization of INPs as a function of temperature, particle number concentration and the fluorescent or coarse particle ratio. The adapted INP parameterizations are demonstrated to be able to predict >90% INP observations within an uncertainty range of a factor of 10. The improved predictabilities of the adapted INP parameterizations are demonstrated by comparisons to parameterizations reported in the literature, and the improvement will reduce the uncertainties in cloud physics simulations.