Refining the Simulation of Mixed‐Phase Clouds Over Coastal Antarctica With a Tailored Ice Nucleating Particle Parameterization

Abstract

Climate models have substantial biases in downwelling shortwave radiation over the Southern Ocean (SO) and Antarctica, primarily due to their inability to accurately represent the frequent occurrence of low‐level mixed‐phase clouds. Building on the evaluation of the Unified Model (UM) Regional Atmosphere and Land version 3 (RAL3), this study investigates the sensitivity of RAL3 to changing cloud and aerosol properties to improve the simulation of clouds and radiation over coastal Antarctica. We evaluate the impact of several approaches to reducing these biases. The main findings suggest that using a monthly aerosol climatology and an aerosol aware cloud‐droplet activation parameterization to diagnose droplet number concentrations substantially underestimates aerosol concentrations at the Antarctic coast, leading to large biases in radiation fluxes compared to surface observations. Increasing the prescribed fixed cloud droplet number concentration or incorporating more accurate orographic data has a limited effect on improving cloud and radiation biases. Additionally, using an ice nucleating particle (INP) parameterization tailored to the pristine Antarctica environment significantly increases the liquid water path and reduces temperature and radiation biases relative to the control simulation. These results underscore the necessity of employing region‐specific INP parameterizations to accurately simulate mixed‐phase clouds over the SO and Antarctica.