Improved Parameterization of Heterogeneous Ice Nucleation by Natural Dust and Soot in the Community Atmospheric Model Version 5

We improve the classical-theory-based parameterization of heterogeneous ice nucleation (Hoose et al. 2010) by replacing their single contact angle (a) model with the probability density function of (a-PDF) model to better represent the ice nucleation behavior of dust and soot found in observations. We further update the classical theory formulation in Hoose et al (2010) by correcting the unphysical behavior of the original expression related to the calculation of the rate of heterogeneous nucleation (J_het). We re-fit the classical theory to derive the uncertain parameters (i.e., onset a and activation energy in the single a model; mean contact angle and standard deviation in the a-PDF model), using more recent observation datasets, including that for the Saharan natural dust. We implement the classical-theory-based formulation in the Community Atmospheric Model version 5 (CAM5), and investigate the impact of the new parameterization of heterogeneous ice nucleation on mixed-phase clouds and climate, and the roles of natural dust and soot. Our results show that ice crystals can form at lower altitudes (with warmer temperatures) simulated by the a-PDF model than by the single a model, which is attributed to the PDF distribution of contact angles in the a-PDF model. In the sensitivity experiments with the a-PDF model, we find that the change of mean of contact angle has larger impact on mixed-phase clouds than that of standard deviation. Consistent with the Hoose et al. (2010), our a-PDF model indicates that the immersion freezing of natural dust plays a dominant role in the heterogeneous nucleation in mixed-phase clouds.