Mayotte 2018 eruption likely sourced from a magmatic mush

Despite significant advancements in understanding crustal melt transport, determining the shallow magmatic architecture at any given volcanic system remains a significant challenge with geophysical methods alone. In this study, we present a new conceptual model combining previously studied models for geomechanical magma reservoirs and dyke-shaped conduits. Supplementary to previous work, we include a model for porous mush adjacent to a visco-elastic magma reservoir during the eruption. This addition enables physically consistent magma recharge into the elastic magma reservoir by porous flow instead of employing an arbitrary mantle magma recharge parameterization. We compare our conceptual model with an illustrative test case - the submarine 2018 Mayotte eruption in the Comoros Islands (∼6.55 km³ Bulk Rock Volume, ∼3.5 km water depth, ∼35 km estimated deep magma reservoir). We estimate a magma effusion rate history using a new continuous-time inversion of surface deformation data. A single magma reservoir cannot match the observed eruptive history for any reservoir geometry or crustal and magmatic material properties. However, the presence of a porous mush adjacent to the magma reservoir helps to reconcile the model with the observations. The additional magma flux from the mush region sustains the long eruption at a low effusion rate, as is observed off Mayotte, with a decreasing effusion rate after few months of eruptions and a low persistent effusion rate since then. The mush flow dynamics are a direct consequence of reduced pressure in the magma reservoir during the eruption. Thus, our model provides a parsimonious and physically motivated explanation of the Mayotte eruptive history. Since many historical basaltic eruptions worldwide, such as sometime observed at Piton de la Fournaise, Volcan Arenal, and Laki, have a similar effusion rate history shape as Mayotte, magmatic mushes may be a key component for explaining long-lived basaltic eruptions.