Abstract : This study is aimed at better understanding magma degassing processes and associated eruptive dynamics through the study of the bulk rocks and mainly of olivine-hosted melt inclusions from explosive products of Mount Etna. We particularly focussed on the 2001 and 2002 flank eruptions that resulted from dyke intrusions, independent of the central conduits, and on three lava fountains at South-East summit crater, in 2000.
We acquired a complete data set on major elements and volatile contents (H2O, CO2, S, Cl and F) of Etna magmas. We report new data on the amount of volatiles (4 wt%, among which ~3.4 wt% H2O) dissolved in the most primitive K-rich basaltic melt erupted at Mount Etna for the last 140 years. This basalt is actually the parental melt of the trachybasalts erupted during the recent volcanic activity. We also argue that the geochemical evolution of etnean lavas for the last 30 years, results from the mixing between the new K-rich melt and K-poorer trachybasalts, emplaced in the feeding system prior to 1970. Isotopic analysis of water and sulphur in the most primitive melt inclusions (delta(D) between -120 and -90; ‰ ; delta(34S) = +2,4±0,4 ‰) reinforces the idea of an OIB-type mantle source and contradicts the hypothesis of its contamination by fluids derived from the Ionian subduction, as previously suggested.
The systematic determination of CO2 and H2O contents of melt inclusions allow us (i) to provide constraints on melt entrapment and S, Cl and F initial exsolution pressures, (ii) to assess the depths of magma transfer and ponding, (iii) to propose a pressure-related model of the evolution of the dissolved and exsolved gas phase, and (iv) to evaluate the effect of the sulphide immiscible liquid on the sulphur degassing path.
We propose that the 2001 and 2002 flank eruptions were sustained by closed system ascent and degassing of basaltic to trachybasaltic magmas. The most primitive magmas rose from at least ~10 km b.s.l. and sustained the powerful lava fountains that occurred in July 2001 and October 2002. Most of the lavas emitted in 2001 and 2002 derived from trachybasaltic magmas, ponding at 5±1 km b.s.l., that were partially dehydrated because of their flushing by a CO2-rich gas phase of deep derivation. Such a process could be frequent at Etna and most likely at basaltic volcanoes.
The S/Cl molar ratio in the associated gas phase is computed to be of 5.4 and 3.7, depend on the degassing dynamics upon eruption. These calculated values and those of the Cl/F (2) and S/CF (9) ratios confirm a dominant closed system magma ascent and degassing mechanism. They are in good agreement with those directly measured by teledetection during the 2001 and 2002 flank eruptions. Furthermore, any gas/melt separation occurring at greater depth would result in higher S/Cl ratios in the gas phase. This also applies to magmas, already saturated with sulphide globule, that are stored at shallow depths in the central conduits and degas under closed system conditions. Therefore, the modelled evolution of S/Cl and Cl/F ratios, both in the melt and the gas phase, gives access to the depths of gas segregation. The S/Cl and Cl/F ratios in the gas emission at the surface, along with the chemistry of associated solid products, bring strong constraints on the different mechanisms at the origin of lava fountains
Finally, steady state summit degassing, involving convective overturn in conduits, would imply ascent of volatile-rich magma until shallow depth (≤ 1 km under the craters), and thus the drain-back of the shallow degassed magma, in order to supply the excess gas flux at Mt Etna.
As a whole, our modelling offers a general background to interpret gas emissions composition and so to better understand the different degassing processes that occur during the ascent of etnean volatile-rich basaltic melts.