Dynamical imprints on precipitation cluster statistics across a hierarchy of high-resolution simulations

Abstract. Tropical precipitation cluster area and intensity distributions follow power laws, but the physical processes responsible for this macroscopic behavior remain unknown. We analyze global simulations at 10 km horizontal resolution that are configured to have drastically varying degrees of realism, ranging from global radiative–convective equilibrium to fully realistic atmospheric simulations, to investigate how dynamics influence precipitation statistics. We find the presence of stirring and large-scale vertical overturning, as associated with substantial planetary- and synoptic-scale variability, to be key to having cluster statistics approach power laws. The presence of such large-scale dynamics is reflected in steep vertical velocity spectra. Large-scale rising and sinking modulate the column water vapor and temperature field, leading to a heterogeneous distribution of moist and dry patches and regions of strong mass flux, in which large precipitation clusters form. Our findings suggest that power laws in Earth's precipitation cluster statistics stem from the robust power laws of atmospheric motions.