On the difficulties of simulating anvil cirrus

Tropical anvil clouds are an important player in the climate system, but reliably simulating them has proven difficult. Here we pinpoint two sources of uncertainty in predicting anvil cloud area, using cloud-resolving simulations of radiative-convective equilibrium. The first uncertainty is that anvil cloud area is actually surprisingly sensitive to horizontal resolution, due to resolution-dependent mixing between clear and cloudy air. The second is that even though detrainment at the anvil level is constrained by clear-sky physics (so-called "clear-sky convergence"), the lifetime of this detrained condensate depends on uncertain microphysics, leading again to a lack of predictability of anvil cloud area. We provide an analytical understanding of these sensitivities, which may point the way towards future progress.