Facets of the Tropical High-Cloud Feedback in a Global Storm-Resolving Model

Tropical high clouds are the cloud type that contributes most to the uncertainty in climate sensitivity, as their feedback on global warming remains poorly constrained. One reason for this is that global circulation models (GCMs) do not resolve the full spectrum of high clouds, ranging from thin cirrus to thick deep convective clouds. We use a set of storm-resolving aquaplanet simulations that resolve this spectrum to investigate how high clouds with different thicknesses contribute to the total high-cloud feedback. We find that the total feedback is positive, arising from all high clouds remaining at fixed temperatures and from an intensification of the diurnal cycle of deep convection. Fixed high-cloud temperatures require an interactive representation of ozone. Prescribing ozone, as often done in GCMs, results in an unrealistic warming of high clouds and hence a less positive feedback. Thick clouds produced by deep convection partly shift from daytime to nighttime in response to surface warming. This reduces sunlight reflection and results in a previously unrecognised positive feedback. The reduction of high clouds in response to warming, assumed to produce a negative feedback, affects thin (warming) and thick (cooling) clouds equally in our simulations, resulting in a near-neutral feedback. By resolving cloud thickness, our storm-resolving simulations show that heterogeneous high-cloud responses to warming combine to yield a positive total feedback.