Petrological monitoring of the volcanoes

Soon or later, the French community will have to face a volcanic crisis on the Antilles islands. Yet, despite on-going real scientific advances, we are still unable to predict the date and the style (i.e. magnitude of explosivity) of the eruption to come. This is partly due to our general incomplete understanding of the relationships between the deep magmatic processes, the precursory signals, and the eruptive dynamics. The deep processes of degassing, crystallization, and fragmentation of the magma can be realistically simulated through ever-burgeoning experimental apparatus. Confronting this knowledge (‘petrological monitoring’) to the geophysical and geochemical monitoring promises valuable advances in the prediction of volcanic eruptions.