Corroborating Observational Evidence of Convective Self-Aggregation with High Resolution Global Storm Resolving Model Simulations

Convective self-aggregation is an atmospheric phenomenon that randomly scattered convection spontaneously reorganizes into a single convection under the absence of external drivers. So far, only a handful of studies have provided observational evidence of the convective self-aggregation of which some used cluster analysis method on observational data and idealized simulations with small domain. Our work applies cluster analysis methods from previous studies on to very high resolution (≤5 km) global storm resolving models of the DYAMOND model intercomparison project and observational datasets of same time period to confirm some of the characteristics of convective self-aggregation. We show that convective self-aggregation increases outgoing longwave radiation at the top of the atmosphere, and it dries the atmosphere. In addition, we find that low-levels cloud increases, and high-level clouds decrease with stronger aggregation. Our study results provide indirect observational evidence of convective self-aggregation, which supports the view that convective self-aggregation is not only discovered in an idealized simulations but also in real world atmosphere.