It is generally accepted, but not experimentally proven, that a quantitative prediction of volcanic eruptions is possible from the evaluation of volcanic gas data. By discussing the results of two years of real-time observation of H 2 O, CO 2 , and SO 2 in volcanic gases from Mount Etna volcano, we unambiguously demonstrate that increasing CO 2 /SO 2 ratios can allow detection of the pre-eruptive degassing of rising magmas. Quantitative modeling by the use of a saturation model allows us to relate the pre-eruptive increases of the CO 2 /SO 2 ratio to the refi lling of Etna's shallow conduits with CO 2 -rich deep-reservoir magmas, leading to pressurization and triggering of eruption. The advent of real-time observations of H 2 O, CO 2 , and SO 2 , combined with well-constrained models of degassing, represents a step forward in eruption forecasting.

This work addresses the study of fluid circulation of the Stromboli island using a dense coverage of self-potential (SP) and soil CO 2 data. A marked difference exists between the northern flank and the other flanks of the island. The northern flank exhibits (1) a typical negative SP/altitude gradient not observed on the other flanks, and (2) higher levels of CO 2. The general SP pattern suggests that the northern flank is composed of porous layers through which vadose water flows down to a basal water table, in contrast to the other flanks where impermeable layers impede the vertical flow of vadose water. In the Sciara del Fuoco and Rina Grande^Le Schicciole landslide complexes, breccias of shallow gliding planes may constitute such impermeable layers whereas elsewhere, poorly permeable, fine-grained pyroclastites or altered lava flows may be present. This general model of the flanks also explains the main CO 2 patterns: concentration of CO 2 at the surface is high on the porous north flank and lower on the other flanks where impermeable layers can block the upward CO 2 flux. The active upper part of the island is underlain by a well-defined hydrothermal system bounded by short-wavelength negative SP anomalies and high peaks of CO 2. These boundaries coincide with faults limiting ancient collapses of calderas, craters and flank landslides. The hydrothermal system is not homogeneous but composed of three main subsystems and of a fourth minor one and is not centered on the active craters. The latter are located near its border. This divergence between the location of the active craters and the extent of the hydrothermal system suggests that the internal heat sources may not be limited to sources below the active craters. If the heat source strictly corresponds to intrusions at depth around the active conduits, the geometry of the hydrothermal subsystems must be strongly controlled by heterogeneities within the edifice such as craters, caldera walls or gliding planes of flank collapse, as suggested by the correspondence between SP^CO 2 anomalies and structural limits. The inner zone of the hydrothermal subsystems is characterized by positive SP anomalies, indicating upward movements of fluids, and by very low values of CO 2 emanation. This pattern suggests that the hydrothermal zone becomes self-sealed at depth, thus creating a barrier to the CO 2 flux. In this hypothesis, the observed hydrothermal system is a shallow one and it involves mostly convection of infiltrated meteoric water above the sealed zone. Finally, on the base of CO 2 degassing measurements, we present evidence for the presence of two regional faults, oriented N41 ‡ and N64 ‡, and decoupled from the volcanic structures.

[1] Constraining fluxes of volcanic bromine and iodine to the atmosphere is important given the significant role these species play in ozone depletion. However, very few such measurements have been made hitherto, such that global volcanic fluxes are poorly constrained. Here we extend the data set of volcanic Br and I degassing by reporting the first measurements of bromine and iodine emissions from Mount Etna. These data were obtained using filter packs and contemporaneous ultraviolet spectroscopic SO 2 flux measurements, resulting in time-averaged emission rates of 0.7 kt yr À1 and 0.01 kt yr À1 for Br and I, respectively, from April to October 2004, from which we estimate global Br and I fluxes of order 13 (range, 3-40) and 0.11 (range, 0.04-6.6) kt yr À1 . Observed changes in plume composition highlight the coherent geochemical behavior of HCl, HF, HBr, and HI during magmatic degassing, and strong fractionation of these species with respect to SO 2 .

Accurate and precisely located self-potential (SP), temperature (T) and CO 2 measurements were carried out in the summit area of Stromboli along 72 straight profiles. SP data were acquired every metre and T data every 2.5 m. CO 2 concentrations were acquired with the same density as T, but only along seven profiles. The high density of data and the diversity of the measured parameters allows us to study structures and phenomena at a scale rarely investigated. The shallow summit hydrothermal activity (Pizzo-Fossa area) is indicated by large positive SP, T and CO 2 anomalies. These anomalies are focused on crater faults, suggesting that the fracture zones are more permeable than surrounding rocks at Stromboli. The analysis of the distribution of these linear anomalies, coupled with the examination of the geologic, photographic and topographic data, has led us to propose a new structural interpretation of the summit of Stromboli. This newly defined structural framework comprises (1) a large Pizzo circular crater, about 350 m in diameter;(2) a complex of two concealed craters nested within the Pizzo crater (the Large and the Small Fossa craters), thought to have formed during the eruption of the Pizzo pyroclastites unit; the Small Fossa crater is filled with highly impermeable material that totally impedes the upward flow of hydrothermal fluids; and (3) The present complex of active craters. On the floor of the Fossa, short wavelength SP lows are organized in drainage-like networks diverging from the main thermal anomalies and converging toward the topographic low in the Fossa area, inside the Small Fossa crater. They are interpreted as the subsurface downhill flow of water condensed above the thermal anomalies. We suspect that water accumulates below the Small Fossa crater as a perched water body, representing a high threat of strong phreatic and phreatomagmatic paroxysms. T and CO 2 anomalies are highly correlated. The two types of anomalies have very similar shapes, but the sensitivity of CO 2 measurements seems higher for lowest hydrothermal flux. Above T anomalies, a pronounced high frequency SP signal is observed. Isotopic analyses of the fluids show similar compositions between the gases rising through the faults of the Pizzo and Large Fossa craters. This suggests a common origin for gases emerging along different structural paths within the summit of Stromboli. A site was found along the Large Fossa crater fault where high gas flux and low air contamination made gas monitoring possible near the active vents using the alkaline bottle sampling technique.