Predicting deep convective self-aggregation using water vapor and buoyancy length-scales

This study investigates the occurrence of convective self-aggregation (CSA) in cloud-resolving model simulations and introduces a dimensionless parameter, the Vapor-Buoyancy Scale ratio ($\eta_{vbs}$), as a predictor for the development of CSA. $\eta_{vbs}$ is a combination of characteristic spatial autocorrelation length scales of column relative humidity and near-surface virtual temperature fields at the beginning of the simulation, before any signs of CSA appear. As CSA occurrence is viewed as the result of a competition between processes that moisten and processes that dry the domain, the four characteristic lengths are intended to capture the spatial scales over which the relevant mechanisms act, at both local and large scales. Specifically they represent: the size of the moist halo around deep convective cores; the size of the virtual temperature depression at cold pool centers; the size of large-scale clear-sky patches between convective systems; the scale over which the large-scale moisture export by cold pools takes place. When convective moist halos are wider than cold pool centers, convective updrafts are likely to erupt in the vicinities of preexisting ones, thereby encouraging the development of self-aggregation. Likewise, when the large-scale distance between convective systems exceeds the distance over which cold pools act, the formation and expansion of dry patches can promote self-aggregation. The parameter’s predictive skill is tested across a range of cloud-resolving model simulations with varying domain sizes, geometries, resolutions, and physical parameterizations. Results demonstrate $\eta_{vbs}$ robustness and its alignment with known CSA sensitivities, such as enhanced aggregation in larger domains and at coarser resolutions.