S23D-2779
Mount Etna: 3-D and 4-D structure using seismic tomography

Tuesday, 15 December 2015
Poster Hall (Moscone South)
Ceri Nunn1, Bruce R Julian1, Gillian R Foulger1, Domenico Patanè2, Jesús M Ibáñez3, Pierre Briole4 and The MED-SUV Team, (1)Durham University, Department of Earth Sciences, Durham, United Kingdom, (2)INGV National Institute of Geophysics and Volcanology, Catania, Italy, (3)University of Granada, Instituto Andaluz de Geofísica, Granada, Spain, (4)Ecole Normale Supérieure Paris, Laboratoire de Géologie, CNRS, Paris, France
Abstract:
We investigate the time-varying structure of Etna, an active stratovolcano in eastern Sicily, using seismic tomography. In volcanic systems, it is thought that the presence of fluids, cracks and pressurized gases can rapidly and drastically change the elastic properties of the host rocks. Recent work suggests that changes beneath Etna are detectable with seismic methods, and that these changes can be linked to volcanic activity. Temporal changes to Earth structure are commonly investigated by carrying out separate tomographic inversions for different epochs. However, repeated inversions of the same area are expected to vary, even if the structure itself does not change. This is due to variations in the seismic ray distribution and to observational errors. Potentially, changes between epochs which are due to experimental limitations can be misinterpreted as changes to the structure of the volcano. 
Consequently, we use a new tomographic program, TOMO4D, that inverts multiple data sets simultaneously [Julian & Foulger, Time-dependent seismic tomography, GJI, 2010]. This code imposes constraints which minimise the differences calculated between two epochs. The remaining structural variations are thus truly required to fit the data, and reflect changes which almost certainly exist between the two epochs. We have selected and relocated ~400 local earthquakes with at least 5 P and 5 S observations. They cover a period which includes several eruptions, from 1st November 2000 to 31st December 2006. We divide our data into different epochs and invert two epochs simultaneously. The models show a seismically fast central region, surrounded by a slower outer region. This suggests a central system of dykes or sills surrounded by volcanic sediments and country rock. At depths of 0-4 km below sea level the seismically fast region is not below the summit crater but is offset to the southwest. 

By monitoring the changes to the elastic parameters of the host rocks we observe temporal changes within the volcano. The technique has potential for long-term volcano monitoring and hazard assessment since it could be applied to monitoring changes from month to month.