The study of rainfall arises from the necessity for knowing large and short-term climatic dynamics, as well as their affectations in the context of engineering practices. This research focus on the study of tropical rainfall and it was guided toward the conceptual exploration of the physical mechanism that explains how the multifractal scaling properties emerges in the rainfall field. On the basis of space-time rainfall records and model outputs analysis, it was possible to collect evidence that confirm rainfall multifractality exists and such a statistical property can be also identified in physically-based model outputs. The conceptual exploration that was developed in this research based on either classic--physics conservation principles or modern theories related to the study of the well-known critical phenomena. Among the findings, multifractality is understood as an essential reflection of the atmospheric instability by convection processes. Either instabilities or their resulting multifractality are sub-products of a diffusive mechanism which takes effect in the atmosphere. Under particular conditions of the dynamical system representing the convection processes, diffusion-driven instabilities give rise to the concentration of spatial structures in the rainfall field, and the organization of such structures is described by multifractality. Although open questions remain about the physics of rainfall multifractality, this work sets up a path for building a general theory and to promote innovative engineering design tools.