10:45 AM - 11:00 AM
[SCG39-25] Spatio-temporal variation of fluid migration associated with slow earthquakes in the shallow Nankai subduction zone
Keywords:Fluid migration, Spatial mapping, Slow earthquake, Scattering coefficient
Slow earthquakes in the subduction zones are potentially linked to fluids, and such observations have been recently reported in the shallow Nankai subduction zone through heterogeneous variations of seismic structure (Tonegawa et al. 2020, JpGU, AGU). However, the region of fluid migration has not been precisely identified yet. Because those studies have reported that fluid migration may occur before shallow slow earthquakes, the identification of the region of fluid migration is important to investigate the characteristics of shallow slow earthquakes. In this study, we estimated two-dimensional maps of the scattering coefficient by seismic heterogeneous variations through a non-linear inversion, i.e., simulated annealing, and investigated the relationship between the scattering coefficient changes, which reflects fluid migration, and slow earthquakes. Here, this technique was confirmed by synthetic waveforms calculated with a three-dimensional velocity structure incorporating heterogeneous structure and random sources at the sea surface.
In the numerical simulations, we set a region with a velocity reduction, and estimated scattering coefficient by using synthetic waveforms with and without the velocity reduction. We performed this test twice by changing the region of velocity reduction, and these tests showed that the anomalies of the scattering coefficient were suitably estimated at the regions of the velocity reductions.
We applied this technique to the following observed data. We calculated cross-correlation functions (CCFs) from ambient noise records in the vertical component observed at the DONET stations, and the stacked CCFs show ocean-acoustically coupled Rayleigh (ACR) waves at frequency band of 0.5–2.0 Hz. We further calculated cross correlation coefficients (CCs) between the reference CCF and individual CCF, which was corrected by temporal change in seismic velocity. Here, dCC is defined by the difference of the CCs (CCs) between the stable period and the period with a reduction. The dCCs for all of the station pairs were the observed data, and such dCCs were observed multiple times.
As a result, the anomaly of the scattering coefficient was estimated to be surrounding and partly overlapped with the region of slow earthquakes including slow slip event (SSE) observed by borehole pore pressure measurements and very low frequency earthquakes (VLFEs). Figure 1 shows an example of dCC occurred in 2012, and the anomaly region was estimated at the southern part of and around the SSE region estimated by Araki et al. (2017). This observation may indicate that a leakage of fluids around the edges of the source region of slow earthquakes is observed.
In the numerical simulations, we set a region with a velocity reduction, and estimated scattering coefficient by using synthetic waveforms with and without the velocity reduction. We performed this test twice by changing the region of velocity reduction, and these tests showed that the anomalies of the scattering coefficient were suitably estimated at the regions of the velocity reductions.
We applied this technique to the following observed data. We calculated cross-correlation functions (CCFs) from ambient noise records in the vertical component observed at the DONET stations, and the stacked CCFs show ocean-acoustically coupled Rayleigh (ACR) waves at frequency band of 0.5–2.0 Hz. We further calculated cross correlation coefficients (CCs) between the reference CCF and individual CCF, which was corrected by temporal change in seismic velocity. Here, dCC is defined by the difference of the CCs (CCs) between the stable period and the period with a reduction. The dCCs for all of the station pairs were the observed data, and such dCCs were observed multiple times.
As a result, the anomaly of the scattering coefficient was estimated to be surrounding and partly overlapped with the region of slow earthquakes including slow slip event (SSE) observed by borehole pore pressure measurements and very low frequency earthquakes (VLFEs). Figure 1 shows an example of dCC occurred in 2012, and the anomaly region was estimated at the southern part of and around the SSE region estimated by Araki et al. (2017). This observation may indicate that a leakage of fluids around the edges of the source region of slow earthquakes is observed.