GNGTS 2019 - Atti del 38° Convegno Nazionale

202 GNGTS 2019 S essione 1.3 performed some resolution tests by producing a set of synthetic surface displacements at our real GPS stations, adopting different patch slip patterns. We compute the Green’s functions G relating the slip s on each patch to the 3-D displacement u at GPS stations assuming an isotropic half-space elastic with a Poisson ratio of 0.25 (Okada 1985). We regularize the inversion by applying smoothing via a finite difference approximation of the Laplacian operator ( ∇ 2 ) and an associated weighting factor κ . We apply positivity constraints in order to avoid implausible and overly rough slip distributions. The inversion minimizes the weighted residual sum of squares (WRSS) and the roughness of the slip model. Results achieved from the resolution tests fail in recovering the slip related to deeper patches, highlighting a poor resolution of the inversion on this sector of the décollement surface. Such a poor resolution would be related to the GPS stations distribution on the eastern flank of the volcano with respect to the geometry and position of the décollement surface. Based on these results, in the following inversions we reduced the number of patches along the surface downdip to 9, by excluding the deeper patches (those located offshore the eastern flank of the volcano). Fig. 3 shows results achieved for SSE-001. Results. The surface displacement patterns suggest that two contiguous sectors of the unstable flank, delimited by the Timpe Fault system, have been involved during the 2006- 2016 SSEs sequence. The former, including i) the Timpe Fault system, ii) the area westward of Timpe Fault system and iii) the SE edge of the unstable flank, has been affected by large displacements during SSE-001, SSE-002, SSE-004, SSE-005, SSE-006, SSE-008 and SSE-011. The latter includes the coastal area, close to the NE edge of the unstable sector and was characterized by large displacements during SSE- 007 and SSE-009. The estimated slip distribution on the décollement surface mimics the patterns, with the largest slip values concentrated on its SE edge during SSE-001, SSE- 002, SSE-004, SSE-005, SSE-006, SSE-008 and SSE-011 and close to its NE edge during SSE-007 and SSE-009. At first glance, SSEs involving the NE edge appear to occur after the events affecting the SE edge. Indeed, such a temporal relationship is purely speculative due to the limited number of SSEs. Moreover, SSE-003 is characterized by large displacements on the entire eastern flank, while SSE-010 shows a complex deformative pattern. SSEs’ moment magnitudes (estimated from the static seismic moment released during the SSE) range between 5.3 (SSE-011) and 5.8 (SSE-001); such a magnitude range is generally larger than the maximum magnitude (~5.2) estimated for earthquakes occurring at Mt. Etna in the last 200 years (Azzaro et al. , Fig. 3 - Displacements and slip distributions for SSE occurred in April 2006. Blue and red arrows indicate observed (with 95% confidence ellipses) and predicted displacements, respectively.