The magma storage capacity of Mt. Etna plumbing system tracked from 3 decades of compositional lava evolution and excess SO2 discharge
Patrick Allard 1,2, Rosanna Corsaro2, Nicole Métrich1,3
Affiliations: 1Université Paris Cité, IPGP, Paris, France; 2INGV, Catania section, Italy; 3INGV, Pisa section, Italy
Presentation type: Talk
Presentation time: Friday 15:45 - 16:00, Room R280
Programme No: 1.5.5
Abstract
Deciphering the magma plumbing system of volcanoes is crucial to improved understanding and forecasting of their eruptive behavior. Here we provide a quantitative estimate of the magma storage capacity beneath Mt. Etna by combining the remarkable compositional evolution of erupted lavas over three decades since the early 1970s and the excess SO2discharge during this period. Focusing on K2O/Th and Rb/Th ratios (unaffected by the degree of magma differentiation) in pre- and post-70s Etna lavas, we document in detail the progressive replenishment of Etna's mid-crustal plumbing system by a new, more alkaline and more radiogenic trachybasaltic magma that gradually mixed with and replaced the former resident magma. On a few occasions (1974, 1998, 2001-2002) this new magma could bypass the central volcano conduits and reach the surface with no or limited interaction with the resident magma. Instead, on other occasions the imprint of resident magma turned back to increase, indicating a complex storage system geometry. Termination of the compositional lava change after the 2001-2002 eruptions, and broadly steady K2O/Th and Rb/Th ratios in products erupted since then, demonstrate a complete renewal of Etna's magma storage cell. Combining this compositional trend over 3 decades with the cumulated amounts of co-erupted magma and of degassed magma - inferred from SO2 emission rates - constrains an overall magma storage capacity of order 4 km3 beneath Etna, a main proportion of which consists of unerupted degassed magma. Useful comparison can be made with recent results from seismic tomography and ground deformation modeling.