GNGTS 2022 - Atti del 40° Convegno Nazionale

GNGTS 2022 Sessione 1.1 35 The amount of scattering accumulated along the seismic ray path is given by (1) Which is the difference between the measured peak delay time for the i-th waveform ( t i PD ( f ) ) and the theoretical peak delay obtained from: (2) with A(f) and B(f) as the coefficients of the resulting linear fit and R Hypo is the hypocentral distance (between 20 and 100 km). High scattering corresponds to positive values of Δlog 10 t ( f ), representing heterogeneities (e.g., fractures and faults) in the crust, while low scattering-negative values characterize compact and rigid crust. Qc measures the decay rate of the coda envelope with increasing lapse time from the origin time of the seismic event (Aki and Chouet, 1975). At first, to obtain Q c -1 we linearize the equation of power spectral energy density (E(t,f)) as a function of the lapse time from the seismic event’s origin time (t) (Aki and Chouet, 1975): where S(f) includes source and site terms and α is equal to 3/2, when propagation is constrained in a single layer in an anisotropic multiple scattering regime (Paasschens, 1997). Then, to image Q c -1 in 2D, we applied an inversion scheme using sensitivity kernel functions for the multiple-scattering regime (De Siena et al. , 2017; Del Pezzo et al. , 2018). The kernel functions are used to build the rows of the inversion matrix used in the inversion scheme. We then tested our resolution through a checkerboard test. Results and Discussion. Peak delay results show a clear difference between the pre- sequence and sequence images, mainly at low frequencies (1.5 Hz – Fig. 1a), where we can define a spatial increase of scattering with time, attributed to structural discontinuities (faults and fractures) and lithology. High-scattering anomalies mapped by the peak delay delineate highly fractured carbonate formations (such as the Lazio-Abruzzi platform), while the Monti Sibillini thrust marks the strongest contrast in scattering. Peak delay also shows a high scattering area corresponding to the Gran Sasso massif and L’Aquila zone, where an important seismic sequence (Mw 6.3) occurred in 2009. In time and at low frequencies (1.5 Hz), the high-scattering is diffuse before the sequence, while between Amatrice and Norcia, the anomaly expands from the southern sector of the seismogenic zone to the fault network located in the W-SW of Norcia. This change in time and space can be interpreted as micro-fracturing processes in the fault network during the seismic sequence. While the structural elements appear pretty clearly in the scattering distribution, we cannot detect a similar pattern in the coda attenuation tomography (Fig. 1b). Absorption highlights a difference between the pre-sequence and the sequence, with high-absorption anomalies at low frequencies distributed around the Apennine Mountain chain. During the 2016-2017 sequence, the high absorption is then focused across the seismogenic zone. Absorption mapping indicates a concentration of fluids along the seismogenic area during the 2016-2017 sequence. The time periods of Amatrice, Visso, and Norcia identify a progression related to a fluid migration from Amatrice. This spatial variation can be attributed to the deep migration of CO 2 -bearing fluids across the fault network.