Progress and Challenges in Simulating Ice-Nucleating Particle Concentrations for Global Climate Models

Aerosols contribute globally not only to the direct radiative forcing of climate and impacts on warm clouds via cloud condensation nuclei, but also to populations of ice-nucleating particles (INPs) and their impacts on mixed-phase and ice clouds. Cloud and climate models increasingly aim to represent the complex interactions of INPs with cloud processes by implementing aerosol-aware parameterizations of ice nucleation.

While these developments are exciting, aerosol-aware cloud and climate models are also susceptible to biases in simulated atmospheric INP populations. Until recently, ice nucleation parameterizations and simulated INPs had not been systematically challenged with field observations.

Our group has addressed this need through both observational closure studies and evaluations of model skill in simulating INP concentrations and their aerosol sources. We have focused on immersion-mode ice nucleation due to its importance for freezing in mixed-phase clouds. In recent case studies, we show that observational closure was achieved for both dust and sea spray INPs, while errors in simulated INPs were primarily attributable to errors in simulated aerosol surface area and composition.

Important uncertainties in INP sources and parameterizations remain. Notably, warm-temperature INPs, often associated with biological or biogenic particles, are challenging to both observe and simulate. Nevertheless, at colder freezing temperatures (e.g., below -20°C), our results suggest that present-day INP parameterizations are likely adequate for simulating the first-order dependencies of INP concentrations on freezing temperature and aerosol amount. These results also highlight the importance of ensuring fidelity in simulations of natural aerosol sources, particularly dust, in models that link cloud ice nucleation to simulated aerosol properties.