10:15 AM - 10:30 AM
[SCG45-18] Characteristics of slow slip events and induced seismicity changes off the Boso Peninsula
Keywords:slow slip event, afterslip, triggered seismicity, fault constitutive law
The start and the end of a slow slip event (SSE) are not always gradual accelerations or decelerations. Remarkable instances can be found off Boso, Japan, where five SSEs occurred from 1996 to 2014. The crustal deformations of those SSEs let us notice that even repeating SSEs sliding in the same region may start with abrupt acceleration and seemingly terminate sliding abruptly (Fukuda, 2018). We consider those activities are clues to understand the rupture initiation and termination of the SSEs.
Off Boso, there are also seismic activities triggered by SSEs. When SSEs accelerate abruptly, seismicity becomes active within following 1-2 days, whereas when they accelerate smoothly, the seismicity is activated after about 6-20 days (Fukuda, 2018). Furthermore, if we look at the spatial distributions of seismic activities, they appear to consist of two clusters: onshore and offshore. Beneath the land, the seismicity has been low since before the SSE while almost no activity in the offshore area. When an SSE occurs, the seismic activity level of the offshore area increases first, at a time close to the SSE. Subsequently, seismic activity beneath the land tends to increase, as does the aftershock activity. Then, the offshore activity virtually disappears after the SSE, while seismic activity beneath the land appears to return to the original low level realized before the SSE. SSEs accompanied with apparently induced seismic activities also occurred in 2018 and 2024, both of which are similar to the aforementioned case of SSE acceleration with offshore seismic activity increases in close proximity to the SSE, as well as the characteristics of activity evolutions of the distinct two clusters.
How can we understand these non-ordinary time evolutions in SSEs and seismic activity levels? We have been simulating various fault slips using a one-degree-of-freedom spring-blocking system in order to investigate their behaviors in as simple a model as possible. Assuming an effective spring constant associated with the slip of a circular crack in an elastic body, radiation damping is applied while direct effects are neglected. In the simulation results, when the repetition of locking and slow sliding corresponding to an SSE repeater was observed, the start of sliding was accompanied by smooth acceleration, whereas after the maximum sliding speed was exceeded, the behavior was observed to slow down more quickly, followed by locking. This behavior is analogous to the characteristics of offshore seismic activity evolutions. On the other hand, in the case of a slow slip caused by applying a stress disturbance to a steady-state slip (in short afterslip), the behavior was that of rapid acceleration followed by gradual deceleration and return to steady-state slip. This behavior resembles the characteristics of seismic activity evolutions under land. However, in the simulation, the stress disturbance is caused in a short period of time, like an earthquake, whereas what is occurring in the case of Boso is an SSE. The fact that after-effect slip-like behavior is observed during the occurrence of SSEs has been shown by Sato et al. in SSEs in the Bungo channel. Elucidating the mechanism of this phenomenon requires a future scrutiny, but it may be understood as the difference between the spontaneous and triggered SSEs, as our spring-slider model allude.
Off Boso, there are also seismic activities triggered by SSEs. When SSEs accelerate abruptly, seismicity becomes active within following 1-2 days, whereas when they accelerate smoothly, the seismicity is activated after about 6-20 days (Fukuda, 2018). Furthermore, if we look at the spatial distributions of seismic activities, they appear to consist of two clusters: onshore and offshore. Beneath the land, the seismicity has been low since before the SSE while almost no activity in the offshore area. When an SSE occurs, the seismic activity level of the offshore area increases first, at a time close to the SSE. Subsequently, seismic activity beneath the land tends to increase, as does the aftershock activity. Then, the offshore activity virtually disappears after the SSE, while seismic activity beneath the land appears to return to the original low level realized before the SSE. SSEs accompanied with apparently induced seismic activities also occurred in 2018 and 2024, both of which are similar to the aforementioned case of SSE acceleration with offshore seismic activity increases in close proximity to the SSE, as well as the characteristics of activity evolutions of the distinct two clusters.
How can we understand these non-ordinary time evolutions in SSEs and seismic activity levels? We have been simulating various fault slips using a one-degree-of-freedom spring-blocking system in order to investigate their behaviors in as simple a model as possible. Assuming an effective spring constant associated with the slip of a circular crack in an elastic body, radiation damping is applied while direct effects are neglected. In the simulation results, when the repetition of locking and slow sliding corresponding to an SSE repeater was observed, the start of sliding was accompanied by smooth acceleration, whereas after the maximum sliding speed was exceeded, the behavior was observed to slow down more quickly, followed by locking. This behavior is analogous to the characteristics of offshore seismic activity evolutions. On the other hand, in the case of a slow slip caused by applying a stress disturbance to a steady-state slip (in short afterslip), the behavior was that of rapid acceleration followed by gradual deceleration and return to steady-state slip. This behavior resembles the characteristics of seismic activity evolutions under land. However, in the simulation, the stress disturbance is caused in a short period of time, like an earthquake, whereas what is occurring in the case of Boso is an SSE. The fact that after-effect slip-like behavior is observed during the occurrence of SSEs has been shown by Sato et al. in SSEs in the Bungo channel. Elucidating the mechanism of this phenomenon requires a future scrutiny, but it may be understood as the difference between the spontaneous and triggered SSEs, as our spring-slider model allude.