[SCG71-01] Hypocenters and focal mechanisms of deep low-frequency earthquakes beneath Zao volcano
Keywords:Zao, deep low-frequency earthquake, focal mechanism
Deep Low-Frequency earthquakes (DLFs) beneath volcanoes are possible evidence for deep-seated magmatic activities between the lower crust and uppermost mantle. At Zao volcano, the number of DLFs started increasing after the 2011 Tohoku Earthquake (Mw 9.0). The hypocenters of these DLFs form two clusters at shallow (20-28 km) and deep (28-38 km) depths. These two clusters are located at central and lower part of a high Vp/Vs zone, respectively (Okada et al., 2014), and the fact suggests different fluid involvement and source processes of DLFs at these two clusters. Using waveform correlation, we performed waveform classification of 1202 DLFs (Jan. 2012-May 2018), which are detected by the matched filter method, and revealed that 6 groups (A, B, D, E, F, G) out of 7 ones are located in the shallow cluster and the other (C) is located in the deep cluster (Ikegaya and Yamamoto, 2019, VSJ). In this presentation, we report the results of focal mechanism estimation using the S/P spectral ratios, and discuss the relationship with the contribution of deep-seated magma.
In the analysis, we used waveform data at 8 stations, operated by Tohoku Univ., NIED, and JMA, surrounding the epicenters of DLFs. We estimated the optimal focal mechanisms that minimize the residual between the theoretical and observed value of the S/P spectral ratio at each station. We assumed the following 6 basis focal mechanisms: (1) Double Couple (DC), (2) Compensated Linear Vector Dipole (CLVD), (3) Tensile Crack (TC), (4) DC + isotropic spherical source (ISO), (5) DC + CLVD, and (6) DC + TC, and estimated the optimal model parameters of these focal mechanisms by grid search. The solution with the smallest AIC among 6 basis focal mechanisms was selected as the optimal one.
As the result, we obtained DC + TC mechanism for group A, D and E, TC for group B and F, and DC + ISO for group C and G as the optimal model, respectively. These optimal models well explain the observed azimuthal dependence of S/P spectral ratios which show maximum peaks in the northeast and southwest directions for group C, and low values in all directions for the other groups. We further decomposed the moment tensors of the optimal solutions into DC, ISO, and CLVD, and revealed that the focal mechanism of group C has the dominant (60%) double-couple component, whereas those of the other groups have the dominant (60-100%) non-double-couple components. This result indicates that the source processes of DLFs distinctly differ between deep and shallow clusters. This difference may be caused by the increase in differential stress near the Moho discontinuity (e.g., Bürgmann and Dresen, 2008), or the difference in the amount and shape of the melt at the central and lower parts of the high Vp/Vs zone. The combination of DC and TC mechanisms is also often obtained as the source mechanism of the shallow (< 10 km depth) volcanic earthquakes, and attributed to such as the dike intrusion (e.g., Hill, 1977). Although it is rather difficult to directly apply this physical model, the focal mechanisms of DLFs in the shallow cluster may be similar to that of the shallow volcanic earthquakes.
The systematic detection of DLFs and focal mechanism estimation in this study contribute to understanding fluid behavior in the lower crust.
In the analysis, we used waveform data at 8 stations, operated by Tohoku Univ., NIED, and JMA, surrounding the epicenters of DLFs. We estimated the optimal focal mechanisms that minimize the residual between the theoretical and observed value of the S/P spectral ratio at each station. We assumed the following 6 basis focal mechanisms: (1) Double Couple (DC), (2) Compensated Linear Vector Dipole (CLVD), (3) Tensile Crack (TC), (4) DC + isotropic spherical source (ISO), (5) DC + CLVD, and (6) DC + TC, and estimated the optimal model parameters of these focal mechanisms by grid search. The solution with the smallest AIC among 6 basis focal mechanisms was selected as the optimal one.
As the result, we obtained DC + TC mechanism for group A, D and E, TC for group B and F, and DC + ISO for group C and G as the optimal model, respectively. These optimal models well explain the observed azimuthal dependence of S/P spectral ratios which show maximum peaks in the northeast and southwest directions for group C, and low values in all directions for the other groups. We further decomposed the moment tensors of the optimal solutions into DC, ISO, and CLVD, and revealed that the focal mechanism of group C has the dominant (60%) double-couple component, whereas those of the other groups have the dominant (60-100%) non-double-couple components. This result indicates that the source processes of DLFs distinctly differ between deep and shallow clusters. This difference may be caused by the increase in differential stress near the Moho discontinuity (e.g., Bürgmann and Dresen, 2008), or the difference in the amount and shape of the melt at the central and lower parts of the high Vp/Vs zone. The combination of DC and TC mechanisms is also often obtained as the source mechanism of the shallow (< 10 km depth) volcanic earthquakes, and attributed to such as the dike intrusion (e.g., Hill, 1977). Although it is rather difficult to directly apply this physical model, the focal mechanisms of DLFs in the shallow cluster may be similar to that of the shallow volcanic earthquakes.
The systematic detection of DLFs and focal mechanism estimation in this study contribute to understanding fluid behavior in the lower crust.