Improved Representation of Low-Level Mixed-Phase Clouds in a Global Cloud-System-Resolving Simulation

Abstract

Low-level mixed-phase clouds are important for Earth's climate but are poorly represented in climate models. A one-moment microphysics scheme from Seiki and Roh (2020, https://doi.org/10.1175/JAS-D-19-0266.1) improves the representation of supercooled water and verifies it with a single-column model. We evaluate the performance of this scheme using a global cloud-system-resolving simulation. We show that the scheme has several major improvements over the original scheme on which it is based, which underestimated the generation of supercooled droplets. The new scheme suppresses the original scheme's tendency to overestimate the conversion of cloud water to rain, vapor to cloud ice, and cloud water to cloud ice. It greatly improves the previously underestimated production of low-level mixed-phase clouds at middle-to-high latitudes, particularly over the ocean at the middle latitudes of the Southern Hemisphere. It also increases the lifetime of liquid clouds, thus improving the simulation of low-level liquid clouds in western coastal regions of the tropics. The temperature dependency of the ratio of mass fraction of liquid cloud to the sum of ice and liquid clouds, F, reveals that mixed-phase clouds statistically develop in a much wider range of temperature (−30°C ∼ 0°C), which supports the development of more mixed-phase clouds in our simulation. The change to a wider range of F at given temperature is expected to be important, because it allows more complex feedback processes to arise from different cloud phase regimes. An improved simulation in seasonal variation of shortwave radiation and its cloud radiative effect are also identified.