Sensitivity of Cloud Feedbacks to Turbulence Closure, Microphysics, and Grid Size in Cloud-Resolving Radiative-Convective Equilibrium Simulations

 

Authors

Andrew Thomas Lesage — University of Utah
Steven K. Krueger — University of Utah
Marat Khairoutdinov — Stony Brook University

Category

Deep convective clouds, including aerosol interactions

Description

Radiative-Convective Equilibrium (RCE) simulations were performed using the System for Atmospheric Modeling (SAM). A large set of simulations was used to determine the sensitivity of cloud feedbacks to the turbulence closure, microphysics (single-moment and double-moment schemes), grid spacing (0.5 km to 16 km), and SST (301 K and 305 K). The turbulence closures used included one that predicts the subgrid-scale turbulent kinetic energy (SGS-TKE) but does not allow SGS cloudiness, and the Simplified Higher-Order Closure (SHOC), which does. Precipitable water and precipitation were most sensitive to SST. Cloud water path and ice water path were primarily sensitive to grid size and turbulence closure. Shortwave cloud radiative effects (CREs) were larger in magnitude with SGS-TKE, while simulations with SHOC had larger magnitude longwave CREs. The simulated clouds were primarily upper-level cirrus. Net CREs were negative for single-moment SGS-TKE and positive for double-moment SHOC. Net CREs were positive for the abundant high-altitude cirrus and negative for the other cloud types. Cloud feedbacks were positive with single-moment SGS-TKE but negative with single-moment SHOC. On average, warmer, single-moment and/or SHOC simulations had more frequent strong updrafts. Heavier precipitation events were slightly more frequent in the double-moment simulations. The higher SST SGS-TKE simulations had more precipitation events at all intensities; however, higher SST SHOC simulations had more light precipitation events and fewer heavy precipitation events.