10:45 AM - 12:15 PM
[SSS10-P05] Aseismic slip during the swarm earthquakes in the Hakone volcano
Keywords:Hakone volcano, Earthquake swarm, Repeating earthquake, Tilt change, Hypocenter migration, Aseismic slip
The generation mechanisms for earthquake swarms in volcanic regions have not been fully understood. Several physical models were proposed, such as stress changes associated with magma movement (Toda et al. 2002), a reduction in fault strength accompanied by migration of pressurized fluid (Shelly et al. 2014), and the expansion of aseismic slip on a fault plane (Barros et al. 2020). Recent studies indicated that high-pressure fluids may induce aseismic slip during the earthquake swarms at the intra-plate or volcanic region (e.g., Yukutake et al 2022; Nakajima and Hasegawa 2023). In order to understand the aseismic slip during swarm activity, we investigated the earthquake swarms that occurred around Kojiri in the Hakone volcano, central Japan, from May 15 to 31, 2015, including the spatial-temporal distribution of hypocenters, crustal deformation based on the tilt meter and detection of repeating earthquake.
To determine a highly resolved hypocenter distribution, we used the Double-Difference method. To evaluate the occurrence of aseismic slip, we estimated the geodetic fault model using the tilt record at Kojiri station, using the formula of Okada (1992) to calculate a theoretical tilt change. To detect repeating earthquakes, we trimmed waveforms using a time window from 1 second before to 5 seconds after the theoretical arrival time of the S wave. We defined similar events with at least four stations with a high cross-correlation coefficient (>=0.95). For similar events, we select the events whose circular cracks assuming a constant stress drop of 3 MPa, overlapped more than 50% of each other as repeaters.
The hypocenters of earthquake swarms are distributed on a nearly vertical plane with a strike in the WNW-ESE direction. The distribution of elapsed time and distance from the first event of this swarm activity shows a diffusion-like migration of the hypocenter. The diffusion coefficient based on the equation of Shapiro et al. (1997) was estimated to be about 10 m2/s. Compared to the swarm activity in 2019 (Yukutake et al. 2019), the diffusion coefficient was about 10 times higher, indicating the faster migration of hypocenters.
A remarkable tilt change of 2.5 micro radians tilt down in the SW direction was observed at the Kojiri station after May 15, corresponding to the start time of the swarm activity. The observed tilt change can be explained by a right lateral fault with a displacement of 4 cm, which is consistent with the tectonic stress field in the Hakone volcano. If we assume that the moment derived from the best geodetic model can be expressed as the sum of the cumulative seismic moment of all earthquake swarms and the moment due to aseismic slip, 92% of the moment may be released by aseismic slip. Using the repeating earthquakes, on the other hand, we estimated a cumulative slip of 0.4 cm during the same period, assuming the scaling of the seismic moment and slip by Somerville et al. (1999). The total amount of aseismic slip obtained from repeating earthquakes is not consistent with the estimation based on the geodetic model. We will further investigate the detection criteria for repeating earthquakes and the relationship with aseismic slip in future work.
To determine a highly resolved hypocenter distribution, we used the Double-Difference method. To evaluate the occurrence of aseismic slip, we estimated the geodetic fault model using the tilt record at Kojiri station, using the formula of Okada (1992) to calculate a theoretical tilt change. To detect repeating earthquakes, we trimmed waveforms using a time window from 1 second before to 5 seconds after the theoretical arrival time of the S wave. We defined similar events with at least four stations with a high cross-correlation coefficient (>=0.95). For similar events, we select the events whose circular cracks assuming a constant stress drop of 3 MPa, overlapped more than 50% of each other as repeaters.
The hypocenters of earthquake swarms are distributed on a nearly vertical plane with a strike in the WNW-ESE direction. The distribution of elapsed time and distance from the first event of this swarm activity shows a diffusion-like migration of the hypocenter. The diffusion coefficient based on the equation of Shapiro et al. (1997) was estimated to be about 10 m2/s. Compared to the swarm activity in 2019 (Yukutake et al. 2019), the diffusion coefficient was about 10 times higher, indicating the faster migration of hypocenters.
A remarkable tilt change of 2.5 micro radians tilt down in the SW direction was observed at the Kojiri station after May 15, corresponding to the start time of the swarm activity. The observed tilt change can be explained by a right lateral fault with a displacement of 4 cm, which is consistent with the tectonic stress field in the Hakone volcano. If we assume that the moment derived from the best geodetic model can be expressed as the sum of the cumulative seismic moment of all earthquake swarms and the moment due to aseismic slip, 92% of the moment may be released by aseismic slip. Using the repeating earthquakes, on the other hand, we estimated a cumulative slip of 0.4 cm during the same period, assuming the scaling of the seismic moment and slip by Somerville et al. (1999). The total amount of aseismic slip obtained from repeating earthquakes is not consistent with the estimation based on the geodetic model. We will further investigate the detection criteria for repeating earthquakes and the relationship with aseismic slip in future work.