Experimental constraints on pre-eruption conditions of the 1631 Vesuvius eruption

Highlights

• We established phase equilibria of 1631 magma at 100 MPa, 950–1050 °C, water content 1.3 - 3.2 wt% and fO₂ NNO+1 to +3

• Natural vs. experimental compositions suggest crystallization at 100 MPa, T 950 ± 30 °C, H₂O_melt = 2.3 ± 0.3 wt%, fO₂ around NNO+1

• A A CO₂-rich fluid was present in the reservoir, related to deeper crystallization and interaction with the carbonate basement.

• Evidence for storage at 3–4 km before 1631 confirms the shallowing of magma reservoirs in the recent evolution of the volcano

Abstract

We established the phase equilibria of a representative tephriphonolitic sample of the 1631 eruption of Vesuvius (Italy). Experiments were conducted at 100 MPa, in the temperature range 950–1050 °C for melt water content ranging from 1.3 to 3.2 wt%, and at an oxygen fugacity (fO₂) of NNO+1 to NNO+3 (one to three log unit above the fO₂ of the Ni-NiO solid redox buffer). Results show that clinopyroxene, biotite and leucite dominate the crystallizing phase assemblage, with minor proportions of plagioclase and amphibole, in agreement with the petrological attributes of the tephra. Comparison between the phase proportions and compositions obtained in experiments and those observed in the rock indicates a pre-eruptive temperature of 950 ± 30° and a melt water content H₂O_melt = 2.3 ± 0.3 wt%, for an oxygen fugacity around NNO+1. These T-H₂O estimates are confirmed by empirical geothermometers based on experimental clinopyroxene and melt compositional trends. As for other Vesuvius eruptions, the most felsic part of the 1631 reservoir appears to have reached pre-eruptive leucite saturation, although a large amount of leucite microcrystals in the studied samples likely grew syn-eruptively. Our results confirm that magma storage conditions beneath Vesuvius became hotter, shallower, and more CO₂-rich after the AD 79 Pompeii Plinian event.

Introduction

The knowledge of the range of possible styles of magmatic activity to be expected from renewed activity at active volcanoes like Somma-Vesuvius is a challenging task, yet such information is of first-order importance for volcanic hazard assessment and emergency planning. Magma composition and pre-eruptive physical conditions directly influence magma rheology, volatile content and style of degassing. In turn, these strongly control the style of volcanic activity and the pre-eruptive state of the volcano, giving clues to interpret the precursory signals of an impending eruption (Blundy and Cashman, 2008; Rutherford, 2008; Cashman, 2004; Oppenheimer et al., 2014).

The study of the products of past eruptions, and the changes of magma composition with time, represent a viable way to infer the range of possible magma conditions for a reactivation of the volcanic activity. Experimental petrology conducted on these products may constrain possible physico-chemical conditions of pre-eruptive magma crystallization, as indicated by the general stability fields of the different mineralogical phases under variable conditions of pressure P, temperature T and volatile fugacity in the coexisting fluid phase.

The last 20 ka of activity at Somma-Vesuvius (Italy; SV) have been punctuated by several explosive eruptions, that dispersed widespread sheets of fallout deposits and often ravaged the volcano slopes and the plain nearby with pyroclastic density currents (Cioni et al., 2008). This fact, with the dreadfully high population density of the many villages encircling the volcano, only few kilometers far from the summit vent, make SV one of the higher risk volcanoes in the world.

SV products are characterized by a variable composition, from slightly evolved tephritic to phonolitic tephritic melts, which especially characterized the most recent activity, up to phonolites and trachytes, generally associated to the largest eruptions in terms of both Intensity and Magnitude (Pyle, 2015). A general perspective about the SV past activity clearly reveals that SV is a highly dynamic system, where the main compositional features of the magma and the frequency and style of eruptions show significant changes with time, at scales from millennia (for the oldest activity) to centuries (for the most recent activity; Cioni et al., 2008, Santacroce et al., 2008; Sbrana et al., 2020). In addition, the large set of available experimental petrology data (Scaillet et al., 2008; Pichavant et al., 2014) clearly suggests a progressive shallowing of the main magma reservoir with time, with two main levels of magma stalling at pressures of about 200 MPa and 100 MPa (about 8 km and 4 km depth, respectively). This evidence is also confirmed by other different, independent datasets, such as the pressure inferences from chlorine in glass (Balcone-Boissard et al., 2016) or from water concentration in melt inclusions in the products of different eruptions of Vesuvius (Cioni, 2000; Fulignati and Marianelli, 2007), or the available seismic tomography data (Auger et al., 2001; Scarpa et al., 2002). The deepest reservoir was active at least from 9 to 2 ka ago, while the shallower was surely active during the last 2 ka of activity, up to the last eruption in 1944. In this context, the definition of the main fields of physico-chemical conditions related to all the main eruptions of SV, and especially to those which occurred during the last centuries of activity, can be of extreme importance for the definition of the possible range of expected magma composition and physical properties in the case of a next reactivation of the volcano. The present strategy adopted by the Dept. of Italian Civil Defense for SV to face a possible next renewal of activity is based on a so-called “Maximum Expected Event” scenario (Barberi et al., 1990; Cioni et al., 2003), considered to be a subplinian eruption (Subplinian I type in the classification of Cioni et al., 2008). Following the Plinian AD 79 event, the two largest events at SV, are of this type (AD 472, AD 1631). While the products of the AD 472 eruption have already been the object of detailed studies aimed at defining PTX magma conditions (Scaillet et al., 2008; Fulignati and Marianelli, 2007), the products of the 1631 event are still not well studied in this respect. In the paper, we present experimental data on the mineral phase stability for this very important eruption, with the aim of constraining pre-eruptive storage conditions and physical properties expected to control similar eruptive scenarios.