Influence of the Joint Variability of Liquid Water and Cloud Droplet Concentration on Area-Mean Autoconversion Rates in Large-Eddy Simulations of Stratocumulus

Marine stratocumulus play an important role in the global radiative balance, yet their representation in Earth System Models (ESMs) remains a challenge. One aspect of representing low clouds in ESMs is how best to represent microphysical processes such as autoconversion, which occur at scales much smaller than the typical grid size of ESMs. For nonlinear rates like autoconversion, the neglect of subgrid-scale (SGS) variability can potentially yield large rate biases. Some ESMs attempt to represent the SGS variability through a so-called "enhancement factor" E, which is a simple multiplicative factor applied to the process rates. A recent version of the Energy Exascale Earth System Model (E3SM) uses a fixed autoconversion enhancement factor of 3.2.

Previous studies on the SGS variability of factors governing autoconversion have focused on cloud regime, ESM grid size, and vertical dependence of E, but have neglected the possible influence of the joint variability of cloud water and cloud-drop concentration. Here we use large-eddy simulation (LES) with size-resolving bin microphysics, constrained by observations and large-scale forcing obtained during the Aerosol and Cloud Experiments in the Eastern North Atlantic (ACE-ENA) Department of Energy (DOE) 2017 field campaign. From LES output we analyze the impact of the co-variation of cloud water mixing ratio qc and cloud droplet concentration Nc on the autoconversion enhancement factor. We find values of E smaller than the value of 3.2 used in E3SM. The inverse relative variance (IRV) of qc and Nc, a measure of horizontal homogeneity that is related to E, show that vertical changes throughout the lower 2/3 of the cloud layer are driven by an increase in mean qc and Nc, whereas near cloud top the variance of qc and Nc becomes the predominant factor. Our findings also indicate that a strong, vertically increasing correlation of qc and Nc plays a large role in the reduction of autoconversion E in the stratocumulus cloud layer and is likely governed by strong inhomogeneous mixing at cloud top.