9:00 AM - 9:15 AM
[SSS06-01] Aseismic process driving the earthquake swarm, revealed by the dense seismic observation
Keywords:Earthquake swarm, Fluid, Aseismic slip
The models for earthquake swarms are proposed as a migration of highly pressurized fluid (e.g., Parotidis et al., 2005) or an aseismic slip on a fault plane (e.g., Lohman and McGuire, 2007). However, the detailed process of fluid migration and aseismic deformation for the occurrence of earthquake swarms has not been fully understood. In Hakone volcano, central Japan, shallow intense swarm activities have been observed. Based on the dense seismic observation data, a previous study found a diffusion-like migration of swarm hypocenters that follows a diffusion equation and gives hydraulic diffusivity between 0.5 and 1.0 m2/s (Yukutake et al., 2011). The observation implies that the occurrence of the earthquake swarms is related to the fluid migration. On the other hand, during the swarm activity from May 11 to 19, 2019, beneath Lake-Ashi, the western part of Hakone caldera, we observed a precursive seismic activity, repeating earthquakes, and tilt change. These observations may indicate the involvement of the aseismic process in addition to highly pressurized fluid in the occurrence of swarms.
We used 34 seismic stations in and around the caldera of Hakone volcano to determine the hypocenter locations and focal mechanisms. The double-difference method (Waldhauser and Ellsworth, 2000) was applied, using the differential arrival time data obtained by the manually picking and waveform cross-correlation process. The focal mechanisms of earthquakes were determined by using the P-wave polarities and P/SH wave amplitudes. To detect the events with smaller magnitude, we applied the Matched Filter method, using the template events based on the HSRI earthquake catalog. For the detection of repeating earthquakes, we applied a criterion with a correlation coefficient of >= 0.95 at more than four stations for the waveform records of 1 sec before and 5 sec after the theoretical P-wave arrival time. The borehole tiltmeters installed in Hakone volcano were also used as crustal deformation data, after removing the trend and tidal response.
The hypocenter distribution represents a plane-like distribution oriented to the azimuth of the E-W direction and nearly vertical dipping angle. One of two nodal planes of the focal mechanism is generally consistent with the orientation of planar hypocenter distribution. The detailed temporal-spatial distribution of seismicity estimated by the Matched filter method shows that the precursive seismic activity started from May 11 on the plane-like distribution, following the remarkable earthquake swarms from 05:18 (JST), May 18. We observe a downward migration of precursive seismicity to the starting point of the remarkable earthquake swarms at the speed of 70m/day, and the remarkable earthquake swarms show a diffusion-like migration of hypocenter that gives a hydraulic diffusivity of 3.0 m2/sec. Fifty-six groups of repeating earthquakes were identified during the seismic activity. Averaged aseismic slip histories inferred from the repeating earthquakes indicate that a subtle aseismic slip of 0.3 mm occurred during the precursive seismic activity, while the aseismic slip accelerated from May 18 up to the cumulative slip of 3 mm. The tilt changes at the borehole stations can be explained most appropriately by assuming the tensile-shear on the planar hypocenter zone. The shear slip dislocation in this model is estimated as 3mm that is consistent with the result of repeating earthquakes. These observations may imply that the aseismic process due to the intrusion of highly pressurized fluid into the fault plane was also related to the triggering the earthquake swarms.
We used 34 seismic stations in and around the caldera of Hakone volcano to determine the hypocenter locations and focal mechanisms. The double-difference method (Waldhauser and Ellsworth, 2000) was applied, using the differential arrival time data obtained by the manually picking and waveform cross-correlation process. The focal mechanisms of earthquakes were determined by using the P-wave polarities and P/SH wave amplitudes. To detect the events with smaller magnitude, we applied the Matched Filter method, using the template events based on the HSRI earthquake catalog. For the detection of repeating earthquakes, we applied a criterion with a correlation coefficient of >= 0.95 at more than four stations for the waveform records of 1 sec before and 5 sec after the theoretical P-wave arrival time. The borehole tiltmeters installed in Hakone volcano were also used as crustal deformation data, after removing the trend and tidal response.
The hypocenter distribution represents a plane-like distribution oriented to the azimuth of the E-W direction and nearly vertical dipping angle. One of two nodal planes of the focal mechanism is generally consistent with the orientation of planar hypocenter distribution. The detailed temporal-spatial distribution of seismicity estimated by the Matched filter method shows that the precursive seismic activity started from May 11 on the plane-like distribution, following the remarkable earthquake swarms from 05:18 (JST), May 18. We observe a downward migration of precursive seismicity to the starting point of the remarkable earthquake swarms at the speed of 70m/day, and the remarkable earthquake swarms show a diffusion-like migration of hypocenter that gives a hydraulic diffusivity of 3.0 m2/sec. Fifty-six groups of repeating earthquakes were identified during the seismic activity. Averaged aseismic slip histories inferred from the repeating earthquakes indicate that a subtle aseismic slip of 0.3 mm occurred during the precursive seismic activity, while the aseismic slip accelerated from May 18 up to the cumulative slip of 3 mm. The tilt changes at the borehole stations can be explained most appropriately by assuming the tensile-shear on the planar hypocenter zone. The shear slip dislocation in this model is estimated as 3mm that is consistent with the result of repeating earthquakes. These observations may imply that the aseismic process due to the intrusion of highly pressurized fluid into the fault plane was also related to the triggering the earthquake swarms.