Sensitivity of Radiative-Convection Equilibrium to Divergence Damping in GFDL-FV3-Based Cloud-Resolving Model Simulations

Using a nonhydrostatic model based on a version of Geophysical Fluid Dynamics Laboratory's FV3 dynamical core at a cloud-resolving resolution in radiative-convective equilibrium (RCE) configuration, the sensitivity of the mean RCE climate to the magnitude and scale-selectivity of the divergence damping is explored. Divergence damping is used to reduce small-scale noise in more realistic configurations of this model. This sensitivity is tied to the strength (and width) of the convective updrafts, which decreases (increases) with increased damping and acts to organize the convection, dramatically drying out the troposphere and increasing the outgoing longwave radiation. Increased damping also results in a much-broadened precipitation probability distribution and larger extreme values, as well as reduction in cloud fraction, which correspondingly decreases the magnitude of shortwave and longwave cloud radiative effects. Solutions exhibit a monotonic dependence on the strength of the damping and asymptotically converge to the inviscid limit. While the potential dependence of RCE simulations on resolution and microphysical assumptions are generally appreciated, these results highlight the potential significance of the choice of subgrid numerical diffusion in the dynamical core.