GNGTS 2015 - Atti del 34° Convegno Nazionale

GNGTS 2015 S essione 1.3 173 to a fluid transfer process (D’Auria et al. , 2015). Analyzing the entire data set of Solfatara fumarolic compositions, we infer that fourteen episodes of magmatic fluid injections affected the CFc from 1983 to 2014 enough to produce measurable geochemical anomalies. Their effects are investigated by physical numerical modelling. Modelling magmatic fluid injections into the CFc hydrothermal system. Chiodini et al. , (2012) applied a physical numerical model [TOUGH2 by Pruess (1991), with an axisymmetric computational domain] tomimic the injection of batches ofmagmatic fluids into the hydrothermal system, feeding the fumarolic field of Solfatara. The results highlight the occurrence of the new unrest of CFc which apparently culminated in 2012-2013 with the above cited magma intrusion at relatively shallow levels. Repeated injections of hot fluids at the base of the hydrothermal system, i.e. beneath Solfatara crater, are imposed to the model, keeping a fixed H 2 O-CO 2 ratio and adjusting the flux through a trial-and-error approach in order to reproduce the H 2 O-CO 2 composition measured at the main Solfatara fumaroles. Twelve injections of variable intensity, each involving an amount of deep fluids of the order of the quantities involved in low-medium sized eruptions, well reproduce the compositional changes of the fumaroles in the 1983-2011 period (Chiodini et al. , 2012). The cumulative curve of injected fluids (for a total of ~ 25 Mt) clearly shows a change in the slope at the beginning of the 2000’s which can be interpreted as the beginning of the new unrest phase at CFc, independently suggested by the inversion in the deformation pattern which, roughly at the same time, passed from a subsidence trend to the new uplift regime. In the last years, new researches based on the fumarolic inert gas species suggested that the period studied in Chiodini et al. (2012) was likely affected by depressurization of the gas-magma separation process (Caliro et al. , 2014). This depressurization, which occurred from 1980’s to 2011-2015, should have caused an important increase of the H 2 O/CO 2 ratio of magmatic fluids because H 2 O is more soluble in magma than CO 2 . This implies that the hypothesis of a fixed H 2 O-CO 2 composition of Chiodini et al. (2012) cannot be taken as plausible. We present here the results of new modelling, which accounts for a progressive increase in the water content of the injected fluids. Recently, several studies were aimed to improve the modeling of the hydrothermal system of CFc. They include, for example, the first definition of a 3-D domain with heterogeneous properties of the rocks derived from the density tomography of the caldera (Petrillo et al. , 2013), and the first application to CFc of MUFITS (Afanasyev et al. , 2015), a code which deals with high, magmatic temperatures of the fluids. Here, however, we discuss the results of new modelling performed with TOUGH2 code (Pruess, 1991) and an axisymmetric computational domain, i.e. the same tools adopted inChiodini et al. (2012), in order to compare these new results with the previous ones. TOUGH2 accounts for the coupled transport of heat and a multi-phase (gas and liquid) and multicomponent (water and carbon dioxide) fluid. The used computational domain, discretized in 850 cells of different volume, represents a 5 km diameter and 2 km height cylinder. Bottom and lateral boundaries are impermeable and adiabatic, while the top boundary has fixed atmospheric temperature and pressure. Values of the rock properties (porosity Φ 0.2; permeability k 10 -14 m 2 ; density ρ 2000 kg/m 3 ; thermal capacity C = 1000 J/kg C; thermal conductivity K 2.8 W/m C) are equal to those adopted in previous modelling (Chiodini et al. , 2012 and references therein). The initial state is a steady state reached after 2000 years of injection of 3400 td -1 of a gas mixture at 350 C with a relatively low CO 2 /H 2 O molar ratio, and ideally represents the pure hydrothermal component discharged at Solfatara before the 1982-84 crisis. The transient solution is obtained with the pulsed injection into the hydrothermal system of large amounts of a gas mixture, with H 2 O-CO 2 composition representing the magmatic fluid. The CO 2 /CH 4 anomalies measured at Solfatara fumaroles provide the hint for the number and timing of each injection episode of magmatic fluids (14 in the 1983-2014 period), while the fumarolic CO 2 /H 2 O