It has long been observed that the top-of-atmosphere net radiation is about the same in convective and adjacent non-convective regions over the tropical warm pools, despite the fact that the albedo and greenhouse effect of deep-convective clouds are both large. Why this occurs is unknown, but one hypothesis suggests that deep-convective clouds shade the ocean from insolation and thereby mutually interact with sea surface temperature gradients and atmospheric circulation. 

It is posited that this interaction, which I call the “cloud shading feedback,” could constrain the top-of-atmosphere radiation balance and maintain similar net radiation in convective and adjacent non-convective regions. I investigate this hypothesis using satellite data and climate model experiments. 

Evidence of the cloud shading feedback from a combination of geosynchronous and polar-orbiting satellite data is presented, and then idealized climate model simulations are used to investigate if the cloud shading feedback constrains the top-of-atmosphere radiation balance. A control experiment is performed in which convective clouds interact naturally with the ocean and atmosphere by forming over the warmest water and shading it, and a mechanism-denial experiment is performed in which the cloud shading feedback is removed by randomizing cloud shading of the ocean. 

Removing the cloud shading feedback brightens convective clouds and enhances the contrast in top-of-atmosphere net radiation between convective and non-convective regions in the tropics. The results support the hypothesis that feedbacks between sea surface temperature and convection are critical to obtaining similar top-of-atmosphere net radiation in convective and adjacent non-convective regions over the tropical warm pools.