University ‘La Sapienza’, Rome, Italy

Abstract We present new results on the velocity structure of the Somma^Vesuvius volcano,obtained by joint inversion of Pand S-wave arrival times from both local earthquakes and shot data collected during the TOMOVES 1994 and 1996 experiments. The use of a large set of earthquakes,recorded over a period of ten years by both temporary and permanent seismic stations,allowed us to enhance the resolution of the structure beneath the Somma^Vesuvius down to 5 km depth. The results obtained show the presence of a high Vp and Vp/Vs anomaly located around the crater axis,between 0 and 5 km depth,involving the volcano edifice and the carbonate basement westward deepening from the adjacent Apenninic belt. The whole available seismic catalogue between 1987 to 2000 (1003 events) has been relocated in the obtained 3-D velocity model. Seismicity appears to be clustered around the anomalous high rigidity body. Laboratory experiments at high temperatures and pressures on 1944 eruption lava samples,taken representative in composition of the magma masses erupted through the cycle 1631^1944,support the interpretation of this anomaly in terms of magma quenching along the main conduit,because of the exsolution of magmatic volatiles. The effect of volatiles from the melt at the eruption onset and through its explosive phases is to increase the solidus temperature well above its eruptive temperature,causing the immediate quenching of the system. This paper shows a good example of how seismic tomography and experimental petrology constrain magmatic models. Results have important implications for the hazard assessment at Somma^Vesuvius,and at other volcanoes worldwide where similar seismological evidence has been recently observed. A 2004 Elsevier B.V. All rights reserved.

Ash and lava samples from 2002 Mt.Etna (Italy) eruption have been treated experimentally in the MAQUA equipment (magma-water interaction internally heated pressure vessel) installed at the HP-HT Lab of Università di Roma "La Sapienza". This equipment allows the control of experimental T and P, the water injection P, the water/melt ratio, and the measurement of the sample V variations. Within the vessel, it is also installed an ultrasonic probe acting both as passive and active transducer located just below the sample holder, via a silica buffer road. In the passive mode this probe is able to analyze the MWI shock waves or the pressure waves due to the gas exolution, in the active one it monitors Vp variations due to changes of the waves propagation velocity through the sample. The experiments have been performed at an average eruptive T (1050^circC) and a confining P=10MPa. The Pinj (water injection pressure) was set at 100MPa with a water/melt ranging between 0.1 and 0.5. These experiments show that variations in the multiphase system viscosity due to variable amounts of the mixture components: melt+crystals+bubbles at comparable water injection modes, can produce strongly different distributions of the produced mechanical energy in terms of volume expansion, fragmentation and mass transport. The powdered lava samples or the ash grains used as starting material give an interaction behaviour and experimental products significatively different. In particular, the homogeneous melt (powdered starting material) compared to an inhomogeneous melt mixture (ash grains) give a lower amount of fragmentation with a modal distribution around the finest fractions and a minimal transported mass. On the contrary, inhomogenous melts mixtures give a very well developed fragmentation with a unimodal distribution around coarser grain sizes and transported masses reaching more than 20% of the initial mass.

Surface tension plays an important role in the nucleation of H2O gas bubbles in magmatic melts and in the time-dependent rheology of bubble-bearing magmas. Despite several experimental studies, a physics based model of the surface tension of magmatic melts in contact with H2O is lacking. This paper employs gradient theory to develop a thermodynamical model of equilibrium surface tension of silicate melts in contact with H2O gas at low to moderate pressures. In the last decades, this approach has been successfully applied in studies of industrial mixtures but never to magmatic systems. We calibrate and verify the model against literature experimental data, obtained by the pendant drop method, and by inverting bubble nucleation experiments using the Classical Nucleation Theory (CNT). Our model reproduces the systematic decrease in surface tension with increased H2O pressure observed in the experiments. On the other hand, the effect of temperature is confirmed by the experiments only at high pressure. At atmospheric pressure, the model shows a decrease of surface tension with temperature. This is in contrast with a number of experimental observations and could be related to microstructural effects that cannot be reproduced by our model. Finally, our analysis indicates that the surface tension measured inverting the CNT may be lower than the value measured by the pendant drop method, most likely because of changes in surface tension controlled by the supersaturation.