GNGTS 2015 - Atti del 34° Convegno Nazionale

174 GNGTS 2015 S essione 1.3 ratio constrains the total gas amount of each injection episode whit a trial-and- error procedure similar to that adopted in Chiodini et al. (2012). While in Chiodini et al. (2012) approach the composition of the injected fluid was constant with time, in this study the H 2 O/CO 2 ratio (by weight) of the injected magmatic fluids increases form the value of 0.67, in 1983, to 1.2, in 2012. This increase of the H 2 O/CO 2 ratio agrees with the hypothesis of an open magmatic system, which depressurizes in time because of degassing. Practically, we depict one possible, but not unique, scenario previously proposed to explain the evolution of the fumarolic inert gas species compositions (see Fig. 8b in Chiodini et al. , 2015). The results of the new model confirm the beginning of the new unrest phase in earlier 2000’s, when the cumulative curve of injected fluids shows an inflection point as already noted by Chiodini et al. (2012). There are, however, two major differences with the previous simulations: 1) in order to reproduce the observed fumarolic compositions, the injected amounts of fluids have to be ~30% higher than in previous model; 2) the system is significantly heated during the process, a feature not observed in Chiodini et al. (2012). The increase of the H 2 O/CO 2 ratio of the injected fluids with time causes, in fact, a remarkable increase of the total amount of steam injected into the system, and in turn of condensation and heating of the whole system. Fig. 3 shows the evolution of the simulated average temperature in the deep central part of the domain, above the injection point (a cylinder of 1 km diameter and 1 km height). The resulting average temperature remains nearly constant from 1983 to 2005 (240-245 C), while from 2006 to 2014 it increases from 245 C to 270 C. The absolute temperature field is in some way controlled by the quite arbitrary choice of the function used to describe the H 2 O/CO 2 increase of the magmatic component, which in turn constraints the total amount of injected steam. On the contrary, the time evolution of the temperature increase is much less affected by this choice, being mainly controlled by other constraints, such as the measured fumarolic CO 2 /CH 4 ratio (frequency of the injections) and the H 2 O/CO 2 ratio (intensity of the injections). It is worth to note that the reliability of the modelled evolution of the temperature finds confirmation on independent observations. The fumarolic content of carbon monoxide, which is the gas specie most sensitive to temperature variations (Chiodini and Marini, 1998), shows the same behavior (Fig. 3). In 2005-2006, concurrently with the beginning of the increase of the simulated temperature, the fumarole of Pisciarelli starts to increase its activity [increase of flow rate and discharge temperature: Chiodini et al. (2015)]. Finally, in 2005-2006 CFc starts to expand and uplift with an accelerating trend very similar to the temperature increase. Conclusion. The almost unique, long time set of fumarolic compositions data at Solfatara highlight important changes in the hydrothermal system feeding the manifestation observed from 1983 to 2015 periods. In particular, during the ongoing unrest of CFc started in 2005, the occurrence of numerous episodes of injection of magmatic fluids into the hydrothermal Fig. 3 – Evolution of the average temperature simulated for the deep central part of the computational domain compared with the measured carbon monoxide (CO) content of the fumaroles. The timing of the simulated magmatic fluid injections (dashed lines) were derived by the analysis of the CO 2 /CH 4 and He/CH 4 fumarolic ratios (see the text for further explanations).