Simulating the emptying of a closed magma chamber by a self-organized fracture mechanism: a conceptual framework for linking different eruptive regimes

We study the statistical features of eruptive events causing the emptying of a closed magma chamber by using a discrete dynamical system where magma spreads through self- organized fracture networks. Specifically, the complex interactions between magma and rock are modelled by means of three dynamical variables that, respectively, control the occurrence of self-organized crack networks, the magma diffusion and the formation of high-density dikes, which magma uses to eventually reach the surface. In particular, the numerical analysis has been carried out for investigating the main features of erupted volumes and repose times statistical distributions. Interestingly, the model predicts that erupted magma is, generally, a mixture of magma that has continuously stopped during the whole ascent path from the chamber to the surface, except for eruptions above a given size threshold, for which it is possible to distinguish two dominant components deriving from specific depth ranges. Such a finding can provide a theoretical framework for the general feature of many volcanic eruptions whose deposits are characterized by two different magmas. Furthermore, in the distribution of repose times a timescale separation between short and long more probable repose times is found, which increases by deepening of the magma chamber. The identification of two different types of repose times suggests the presence of different patterns, which could help the understanding of the processes responsible for different eruptive regimes.