Toward improving ice nucleation parameterization and aerosol effects on mixed-phase and ice clouds

Description

Ice nucleation is an important microphysical process in mixed-phase and ice clouds. However, our current understanding of nucleation mechanisms is still very poor. In order to include aerosol effects on ice-containing clouds, ice nucleation processes must be described in terms of aerosol properties. Classical Nucleation Theory (CNT) can be used to do that if its key parameters are properly constrained. In this study, the new laboratory data for deposition nucleation on mineral dust collected at the Pacific Northwest National Laboratory Atmospheric Measurements Laboratory (AML) together with published data from other laboratories are used to constrain the parameters in CNT, such as contact angle (a parameter in CNT to represent the ice-nucleating ability of aerosols). Lognormal probability distribution functions (PDF) of contact angle are derived for different dust types, temperatures, and dust particle sizes. Using cloud-resolving model simulations, we evaluate the sensitivity of cloud properties to changes in the PDF parameters (mean and standard deviation) for contact angle distribution. We find that with the derived lognormal distribution of contact angle, the observed freezing probability can be reasonably reproduced. Simulated cloud microphysical properties, cloud onset time, thickness, and lifetime are very sensitive to changes in contact angle distribution, which suggests the importance of improving the ice nucleation parameterizations.