*SO2 Fluxes during the May 2016 Eruptive Sequence: Comparison with 2014–2015 Results*

It is well established that SO2 fluxes are directly linked to the rate of magma ascent and degassing [78]. Thus, temporal variations in SO2 fluxes do reflect changes in magma feeding to the volcano's shallow plumbing system, and, as such, may help track transition in activity style, from quiet passive degassing to eruptive periods [81]. On Etna, we now have 3 years of UV camera observations available ([28,30], this study), during which transition from quiescence to eruption has frequently been observed. We, thus, examined our dataset in the attempt to tentatively identify any possible systematic SO2 flux threshold/trend corresponding to such an activity switch.

For this purpose, in Figure 10, we compared the cumulative SO2 flux trends (time-normalized) for three different periods encompassing three eruptive paroxysmal episodes, which occurred in August 2014, December 2015, and May 2016, respectively. For each of three events, we calculated the averaged SO2 fluxes (corresponding to the slopes of the cumulative curve) in the periods before, during, and after eruption, and we found significant similarities between the three events. In each of the three 2014–2016 events, the pre-eruptive fluxes fell in a relatively narrow range, between 1900 and 2500 t/d. The syn-eruptive (during the paroxysmal sequence) fluxes were typically higher, and spanned between 3000 and 5200 t/d (Figure 10), and were the highest during the December 2015 paroxysmal sequence that was consistently the most energetic in the past few years [67]. Lastly, each of the three post-paroxysmal phases was characterized by reduced SO2 emissions, ranging between 600 and 900 t/d, which impliesa reduced magma supply and degassing of a volatile depleted (residual) magma after

p

each eruptive episode. These ranges were fully in agreement with those indicated by Reference [49] and by the analysis of a 13-year-long dataset.

**Figure 10.** Comparison between cumulative SO2 flux trends associated with three lava fountaining paroxysmal sequences that occurred at NSEC (black line) in August 2014, and at VOR in December 2015 and May 2016 (red line and blue line). Stars indicate onsets of the paroxysmal sequence. Pre-eruptive, syn-eruptive, and post-eruptive phases show similar SO2 fluxes for all these three events. In each of these three events, pre-eruptive fluxes range between 1900 and 2500 t/d, syn-eruptive fluxes are the highest (3000–5200 t/d), while post-eruptive fluxes are systematically the lowest (<900 t/d).

Our preliminary results are suggestive of the existence of a systematic pattern in SO2 emissions that, if confirmed, would imply a recurrent degassing process/mechanism prior to, during, and after the Etna's eruptive periods. Clearly, additional data are required to corroborate this initial hypothesis.

#### **5. Conclusions**

SO2 imaging at Mt Etna during 2016 revealed different styles of gas emissions, which reflected changes in volcanic activity, from quietly passive degassing to eruptive activity (lava fountaining, intense strombolian activity, and lava flowing).

To real-time characterize and monitor this very dynamic activity period, we designed a novel routine to automatically calculate SO2 fluxes including computer-based detection of image quality, and calculation of plume speed time-series using computer vision libraries. This automatic processing routine allowed us to obtain real-time information on volcano degassing dynamics at high spatial and temporal resolution, which results in a further step in instrumental volcanic gas monitoring. We validated the methodology through a comparison with manually processed results and integration with independent thermal and seismic observations. All these independent datasets showed coherent temporal variations that validated the use of UV cameras for detecting subtle changes in volcanic and degassing activity. Our novel method, thus, promises a step ahead in instrumental volcanic gas monitoring.

Our automatically derived SO2 flux time-series were used to constrain degassing regimes on Mt. Etna in 2016. We have shown that our automatically processed SO2 fluxes peaked during heightened

activity, such as during the May 2016 eruptive sequence, and during a phase of elevated degassing in July–August 2016, which culminated with the opening of a new summit incandescent vent. The May 2016 sequence was preceded by a circa one-month-long phase of a mild but detectable SO2 flux increase, and was followed by an abrupt drop in degassing. Comparison between our SO2 flux time-series in May 2016 and those associated with two other eruptive paroxysmal sequences (occurred in 2014 and 2015) highlighted strong similarities in SO2 flux dynamics, which implies the possible existence of thresholds that distinguish between degassing regimes, prior to, during, and after eruptions. Pre-paroxysm SO2 fluxes were found to have consistent values (of ~2000 t/d) during the three episodes. Similarly, the highest SO2 fluxes (from 3000 t/d up to 5200 t/d on a daily average basis) were identified during the three eruptive sequences, while post-eruptive were systematically characterized by reduced degassing (<1000 t/d). This result, if confirmed by future observations, may bring implications for identifying switches in volcanic activity regime. We believe this methodology can be successfully exported on other open-vent active volcanoes, after a relatively brief period of calibration.

We also tested the ability of UV camera records to characterizing SO2 emissions during a lava fountain episode. This is, to the best of our knowledge, one of the first SO2 camera record at high spatial and temporal resolution of an ongoing lava fountaining [28,30]. We have reported on high levels of degassing during the lava fountain (of up to 260 kg/s), from which we assessed a lower limit (due to the ash presence within the plume) for the cumulative SO2 mass of ~1700 tons emitted during a lava fountain episode.

#### **Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-4292/11/10/1201/s1.

**Author Contributions:** D.D.D., A.A., and M.R. conceived the manuscript. D.D.D. developed the automatic UV camera processing algorithm. M.B. designed the UV camera system hardware. R.D. and G.T. validated the automatic algorithm. D.C. processed thermal remote sensing data. E.P. collected and processed thermal camera data. M.C. supervised the work and contributed to write the manuscript.

**Funding:** The European Community's Seventh Framework Program under grant agreement 305377 (BRIDGE Project) funded this research.

**Acknowledgments:** We wish to thank Salvo Caffo (Ente Parco dell'Etna) for administrative support. Nino Giuffrida (Funivia dell'Etna) and Filippo Greco (INGV-OE) are acknowledged for support in the field. Three anonymous reviewers are acknowledged for their constructive comments, which improved the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
