Vesuvius is perhaps the most dangerous of the world's Decade Volcanoes. 600 000 people live within 10 km of the crater and it has experienced 9 large (VEI >3) and numerous minor eruptions in historical times. This important volcano has been the subject of extensive earlier field and petrological studies by many groups, which supply an underpinning framework for this program. The sustained explosive eruptions, which characterize Vesuvius's history, are highly destructive events in which stable eruption plumes may give way gradually or abruptly to column collapse and/or phreatomagmatism. The study focuses on the cause of the abrupt transition from stable Plinian eruption to phreatomagmatic explosions generating highly destructive pyroclastic density currents during the 79 AD eruption. The AD 79 eruption, which destroyed Pompeii and Herculaneum (Lirer et al 1973, Sigurdsson et al. 1982, 1985, Cioni et al. 1992), is generally regarded as the 'type' example of Plinian volcanism. The program links macroscopic parameters describing the changes in eruption intensity and style with microscopic observations of pumice texture and chemistry. The study will (i) characterize the mixture of volcanic rock fragments (pyroclasts) emerging from single eruptive vents at a single moment and (ii) contrast patterns of vesiculation and microlite-crystallization for wide spectra of magma composition, and eruptive style and intensity. Large and small crystals (phenocrysts and microlites respectively) and gas bubbles (vesicles) in pyroclasts 'freeze in' aspects of the history of the parent batch of magma in the chamber and conduit, and represent 'microscopic windows' into the subsurface evolution of the volcano. We will interpret the changes in mean and extreme values of vesicularity and microlite-crystallinity before and after the key change in eruptive style. By using such well-constrained historical eruptions, we expect that the results of this work will have broad application to less well-documented historic or prehistoric eruptions and will greatly improve our understanding of abrupt changes in eruptive style/intensity during explosive volcanism, and aid in forecasting not when Vesuvius (and other cone volcanoes) will next erupt but how it will erupt, a question that is equally important from a perspective of hazard management and mitigation.
