10:45 AM - 12:15 PM
[SSS10-P11] Extraction of epicentral migration by a space-time Hough transform: Application to earthquake swarm beneath the NE Noto Peninsula, Japan
Keywords:Earthquake swarm, Epicentral migration, Space-time Hough transform, Noto Peninsula
Spatiotemporal migration of earthquakes has been observed not only for earthquake swarms but also for slow earthquakes. In both cases, their front exhibits diffusive behavior, which is explained by the diffusion of pore fluid pressure and stress (e.g., Shapiro et al., 1997; Ando et al., 2012). Therefore, the study of seismic source migration is key to understanding the physical process of seismic activity. Sagae et al. (under review) developed a space-time Hough transform to objectively extract the source migration at a constant speed on a fault plane. They found the spatiotemporal characteristics of source migrations with different speeds (10–1000 km/day) inside the diffusion front of tectonic tremors. In contrast, few previous studies have investigated the source migrations inside the diffusion front of earthquake swarm (e.g., Dublanchet, & De Barros, 2021). The space-time Hough transform has the potential to extract the source migration projected onto the ground surface (epicentral migration), even when the fault plane is unclear. In this study, we applied the space-time Hough transform to earthquake swarms beneath the northeastern Noto Peninsula, Japan to extract the epicentral migrations and investigated their spatiotemporal characteristics in detail.
Earthquake swarms beneath the northeastern Noto Peninsula have continued since 2018. The source distribution is spatially grouped into four clusters, which are defined as S, W, N, and NE clusters. The earthquake swarm started in the S-cluster in 2018, became active in the deep part of the S-cluster in December 2020, and then earthquake swarms started in the other clusters. In this study, we used the source distribution (Amezawa et al., under review) for 20,542 earthquakes of magnitude 0.0 or greater between May 2018 and June 2022, which were relocated using the double-difference method (Waldhauser & Ellsworth, 2000).
In the space-time Hough transform, a straight line (epicentral migration) in (2+1)-D space-time (longitude, latitude, and time) is represented by four parameters: the distance from the origin to the straight line ρ, the zenith angle θ, the azimuth φ, and the rotation angle ψ. In particular, tanθ and φ represent the migration speed and migration direction, respectively. By considering a straight line with spatial width (cylinder), we take the uncertainty information of the epicenter into account. We extract the epicentral migration that contains the largest number of earthquakes inside the cylinder. In addition, even if multiple epicentral migrations occur, we can separate each epicentral migration based on the parameters (ρ, θ, φ, and ψ). In this study, we set bins for ρ in the range of 0-20 km at 0.2 km intervals, φ and ψ in the range of 0-360° at 10° intervals, and tanθ in the range of 0.1-10 km/day at 0.1 km/day intervals. The radius of the cylinder was set to 0.25 km. The threshold for the number of earthquakes inside the cylinder was set to 8 or more, and the epicentral migration was objectively extracted.
As a result, we extracted 485 epicentral migrations in the S, W, and N clusters. In most of the clusters, the median duration and the migration speed were longer than six days and slower than 0.5 km/day, respectively. On the other hand, in the deep part of the S-cluster, the median duration and the median migration speed were shorter than one day and higher than 1.5 km/day, respectively.
In the deep part of the S-cluster, the sources are distributed in a ring shape, and we observed rapid epicentral migrations propagating northward along the epicentral distribution. These epicentral migrations occur intermittently, suggesting that rapid and intermittent spatiotemporal changes in pore fluid pressure may occur in the deep part of the S-cluster. In this presentation, we discuss depth variations of earthquakes associated with the extracted epicentral migrations and the possibility of the space-time Hough transform incorporating the depth information.
References:
[1] Kodai Sagae et al. Fine structure of tremor migrations beneath the Kii Peninsula, Southwest Japan, extracted with a space-time Hough transform. ESS Open Archive.
[2] Yuta Amezawa et al. Long-living Earthquake Swarm and Intermittent Seismicity in the Northeastern Tip of the Noto Peninsula, Japan. ESS Open Archive.
Acknowledgement: This research was supported by JSPS KAKENHI Grant Number JP21H05205 in Grant-in-Aid for Transformative Research Areas (A) “Science of Slow-to-Fast Earthquakes”.
Earthquake swarms beneath the northeastern Noto Peninsula have continued since 2018. The source distribution is spatially grouped into four clusters, which are defined as S, W, N, and NE clusters. The earthquake swarm started in the S-cluster in 2018, became active in the deep part of the S-cluster in December 2020, and then earthquake swarms started in the other clusters. In this study, we used the source distribution (Amezawa et al., under review) for 20,542 earthquakes of magnitude 0.0 or greater between May 2018 and June 2022, which were relocated using the double-difference method (Waldhauser & Ellsworth, 2000).
In the space-time Hough transform, a straight line (epicentral migration) in (2+1)-D space-time (longitude, latitude, and time) is represented by four parameters: the distance from the origin to the straight line ρ, the zenith angle θ, the azimuth φ, and the rotation angle ψ. In particular, tanθ and φ represent the migration speed and migration direction, respectively. By considering a straight line with spatial width (cylinder), we take the uncertainty information of the epicenter into account. We extract the epicentral migration that contains the largest number of earthquakes inside the cylinder. In addition, even if multiple epicentral migrations occur, we can separate each epicentral migration based on the parameters (ρ, θ, φ, and ψ). In this study, we set bins for ρ in the range of 0-20 km at 0.2 km intervals, φ and ψ in the range of 0-360° at 10° intervals, and tanθ in the range of 0.1-10 km/day at 0.1 km/day intervals. The radius of the cylinder was set to 0.25 km. The threshold for the number of earthquakes inside the cylinder was set to 8 or more, and the epicentral migration was objectively extracted.
As a result, we extracted 485 epicentral migrations in the S, W, and N clusters. In most of the clusters, the median duration and the migration speed were longer than six days and slower than 0.5 km/day, respectively. On the other hand, in the deep part of the S-cluster, the median duration and the median migration speed were shorter than one day and higher than 1.5 km/day, respectively.
In the deep part of the S-cluster, the sources are distributed in a ring shape, and we observed rapid epicentral migrations propagating northward along the epicentral distribution. These epicentral migrations occur intermittently, suggesting that rapid and intermittent spatiotemporal changes in pore fluid pressure may occur in the deep part of the S-cluster. In this presentation, we discuss depth variations of earthquakes associated with the extracted epicentral migrations and the possibility of the space-time Hough transform incorporating the depth information.
References:
[1] Kodai Sagae et al. Fine structure of tremor migrations beneath the Kii Peninsula, Southwest Japan, extracted with a space-time Hough transform. ESS Open Archive.
[2] Yuta Amezawa et al. Long-living Earthquake Swarm and Intermittent Seismicity in the Northeastern Tip of the Noto Peninsula, Japan. ESS Open Archive.
Acknowledgement: This research was supported by JSPS KAKENHI Grant Number JP21H05205 in Grant-in-Aid for Transformative Research Areas (A) “Science of Slow-to-Fast Earthquakes”.