The Critical Role of Observations in Developing Numerical Representations of Ice-Nucleating Particles for Southern Ocean Mixed-Phased Clouds

The abundance of supercooled liquid clouds over the remote Southern Ocean (SO) region challenges the aerosol-cloud microphysical interactions that are currently represented in global climate models. In particular, most global climate models simulate SO clouds that are over-glaciated compared to satellite observations, leading to large radiative biases. Primary ice formation initiated by atmospheric ice nucleating particles (INPs) is a critical process in cloud glaciation and is augmented by additional processes, such as secondary ice production. Due to extremely high winds and large distances from land, the INP sources are likely dominated by locally produced sea spray aerosol (SSA) and long-range transported terrestrial aerosol (e.g., dust and pollution). However, a dearth of aerosol and INP observations in the remote SO have limited our ability to advance numerical representations of freezing processes until recent years. In this study, we demonstrate ways in which recent INP observations have improved our understanding of SO INP populations and how modeling studies can aid in developing hypotheses that can be tested by future observational studies.

Coastal and ship observations were used to 1) characterize remote INP populations and 2) develop and test numerical representations of INPs relevant for remote oceanic regions. These observations revealed that INPs associated with SSA are up to a factor of 1000 less ice nucleation active compared to mineral dust. Extremely low INP concentrations observed over the SO support the prevalence of supercooled liquid clouds known to dominate the SO region. In a series of nudged simulations with the Community Earth System Model, Versions 1 and 2, we demonstrated that observed INP number concentrations in the marine boundary layer are successfully predicted using an approach that considers only SSA and dust INPs. This predictive tool was further used to estimate INP populations at higher altitudes over the SO study region, revealing that while SSA dominates the INP population below 3-5 km, mineral dust INPs may be critical for INP populations present above 5 km. Finally, an evaluation of model-estimated INPs present at higher altitudes using recent airborne measurements of INPs from the 2018 Southern Ocean Clouds, Radiation, Aerosol Transport Experimental Study will be presented.