Separation of intrinsic and scattering seismic attenuation in the Southern Apennine zone, Italy

Abstract

Summary Scattered waves observed at the seismographs of the National Italy's seismic network have been used to investigate the intrinsic dissipation and scattering properties of the lithosphere under the Southern Apennines, Italy. First, we investigate the coda‐Q properties, then we apply the MLTW analysis in the hypothesis of velocity and scattering coefficient constant with depth, and finally we interpret these results with the aid of numerical simulations in a medium with depth dependent velocity and scattering coefficient. Results obtained in the hypothesis of a uniform model show that a low scattering‐Q−1 and a relatively higher intrinsic‐Q−1 characterize the lithosphere of the Southern Apennines. Numerical simulations of the seismogram energy envelopes were performed hypothesizing a strongly scattering crust and trasparent upper mantle, both with reasonable intrinsic dissipation coefficients. In these symplifying assumptions the theoretical curves calculated for the homogeneous model fit to the synthetic envelopes with scattering attenuation coefficients always greater than the synthetic values. This results lead to the consideration that scattering‐Q−1 obtained using MLTW analysis under the assumption of uniform medium are overestimated. The values of the scattering‐Q−1 estimated for Apennines at low frequency (1–2 Hz) in the hypothesis of uniform medium are of the same order of those obtained in several areas around the world. The estimates obtained for frequencies ranging from 2 to 12 Hz are very low if compared with those obtained in the same hypothesis for other areas around the world. Coda Q−1 closely resembles intrinsic Q−1.

Attenuation, source parameters and site effects in the Irpinia–Basilicata region (southern Apennines, Italy)

“…The quality factor Q based on the S-wave part of the recordings shows a pronounced frequency-dependence (close to linear), given by Q( f ) = 28.3 f 0.87 . This S-wave attenuation model differs from the results of other authors reported for neighboring regions (Rovelli et al, 1988;Malagnini et al, 2000;Bianco et al, 2002), and in particular shows a less severe frequency-dependence than the results derived by Castro et al (2004bCastro et al ( , 2008 for southern Italy. However, as stated above, it is important to keep in mind that the comparison of different models is complicated by the fact that the data used to estimate them sample different frequency bands and different crustal volumes.…”

“…The studies by Malagnini et al (2000) and Rovelli et al (1988) sample for 12 instance large regions such as the entire Apennines or the entirety of southern Italy. In contrast, our results are related to a small area limited by the extent of the ISNet network, and also Bianco et al (2002) analyzed data from a much larger area surrounding the ISNet network. This limitation of our study to the small region covered by the ISNet network can however also be seen as an advantage in the sense that with the derived attenuation functions, we provide a means for reliable and consistent attenuation correction for events occurring in close proximity or within the network, avoiding a bias in source parameter or site response estimation due to inappropriate attenuation correction.…”

“…Considering however the typical uncertainty ranges in determining such Q(f) models, the results of both studies are highly consistent with one another, and we only list the result of Castro et al (2004b) in Table 1 and Figure 6. Another study covering the region of interest investigated coda-Q in the frequency range 1-12 Hz (Bianco et al, 2002) and the authors found from a multi lapse time window analysis that intrinsic Q seems to dominate with respect to scattering Q in the area. Other Q-models have been reported for different regions of Italy by Rovelli et al (1988), Malagnini et al (2000) and Castro et al (2004a,b), and the one that matches our results most closely is the model for Umbria given by Castro et al (2004a), with a frequency-dependence N=1.2.…”

Direct simulation Monte Carlo method with a focal mechanism algorithm

“…F 60 denotes the Fisher distribution function with a confidence level of 60%, which was also used by previous studies (Bianco et al, 2002(Bianco et al, , 2005Giampiccolo et al, 2006). The ratios M f ðZ s ; Z i Þ=M f ðẐ s ;Ẑ i Þ were depicted by the shading zones for the confidence areas ( Figure 5).…”

Precursory pattern of tidal triggering of earthquakes in six regions of China: the possible relation to the crustal heterogeneity

“…Thirdly, we compare the integrated energy densities derived from observational data with that derived from theoretical formulation to obtain the best-fitting values of B 0 and L −1 e . The fitting errors are calculated by the following equation (Bianco et al, 2002):…”