Toward Improved Cloud-Phase Simulation with a Mineral Dust and Temperature-Dependent Parameterization for Ice Nucleation in Mixed-Phase Clouds

Abstract

Abstract Mixed-phase clouds are frequently observed in the atmosphere. Here we present a parameterization for ice crystal concentration and ice nucleation rate based on parcel model simulations for mixed-phase stratocumulus clouds, as a complement to a previous parameterization for stratus clouds. The parcel model uses a singular (time independent) description for deposition nucleation and a time-dependent description for condensation nucleation and immersion freezing on mineral dust particles. The mineral dust and temperature-dependent parameterizations have been implemented in the Geophysical Fluid Dynamics Laboratory atmosphere model, version 4.0 (AM4.0) (new), while the standard AM4.0 (original) uses a temperature-dependent parameterization. Model simulations with the new and original AM4.0 show significant changes in cloud properties and radiative effects. In comparison to measurements, cloud-phase (i.e., liquid and ice partitioning) simulation appears to be improved in the new AM4.0. More supercooled liquid cloud is predicted in the new model, it is sustained even at temperatures lower than −25°C unlike in the original model. A more accurate accounting of ice nucleating particles and ice crystals is essential for improved cloud-phase simulation in the global atmosphere.