5:15 PM - 7:15 PM
[SCG50-P01] Analysis of Spatiotemporal variation of the earthquake swarms in the Okinawa Trough by using Matched filter method
Keywords:Okinawa Trough, Earthquake swarm, Dike intrusion, ETAS model, b_value
The Okinawa Trough is located to the northwest of the Ryukyu Arc and is a back-arc basin where rift expansion is still ongoing. Earthquake swarms occur frequently and are confirmed in the vicinity of the rift axis, volcanoes, and hydrothermal vents. The activity over the past 20 years has been concentrated in specific areas, and although the duration is usually a few days, it can sometimes last for more than a week. In order to perform a detailed analysis of earthquake swarms, it is necessary to capture more earthquakes, but the detection capability is reduced when they occur in the Okinawa Trough (Nakamura, 2022). Therefore, we used the Matched Filter (MF) method to detect the main earthquake swarms from 2002 to 2022 and improved the accuracy by re-determining the epicenters using the HypoDD method. We also applied Rivalta's (2010) model to analyze the epicenter migration and conducted the statistical analysis using b-value analysis, the ETAS model, and seismic moment.
As a result of the analysis, we were able to detect about 2 to 3 times as many earthquakes as the JMA catalog. The characteristics of earthquake swarms differ between the Central Okinawa Trough (COT) and the Southern Okinawa Trough (SOT), with the COT being dominated by local-type earthquakes (within 10 km of the active region) and the SOT being dominated by migration-type earthquakes (movement of 10 km or more). The 2007 activity was classified as aftershock-type only. In the ETAS model, if the parameter α is 0.7 or less, it is considered to be a earthquake swarm type, and if it is less than 0.5, it is considered to be the effect of dyke intrusion (Ishigaki, 20 10). In this study, the value of α was less than 0.7 for all of the activities, suggesting that they were triggered by the dyke intrusion. However, the reliability of the ETAS model may be low for the 2007 activity because the Mc value increased rapidly, reducing the detection capability. In the case of the swarm earthquakes in the SOT, the diffusion coefficient of Shapiro et al. (1999) was larger than that of previous studies (D = 10 km²/s), and the possibility of activity due to dyke intrusion is high. The area beneath Futagoyama, which is near the start of the earthquake swarms, is a possible magma reservoir. In the SOT, it is estimated that, as in Iceland and Afar, the progression of dyke intrusion is related to seismic activity. The ETAS model also suggested that the earthquake swarms off Amami Oshima Island were triggered by dyke intrusion, but since no crustal deformation was observed, it is thought that the scale was small. In addition, in the Okinawa Trough, regions with high and low b-values were identified, and the high b-values near the hydrothermal vents suggested the influence of hydrothermal activity (Minami, 2021). In the area off the northwest coast of Okinawa Island, earthquake swarms have been occurring in an area where traces of dyke intrusion have been observed near the axis of the rift, so there is a possibility that dyke intrusion is involved in the earthquakes swarm.
The Okinawa Trough earthquake swarms were more active near the volcanic front (80-90 km depth on the slab) and in the hydrothermal activity area than along the rift axis. It is more likely that the active magma formation at the volcanic front is responsible for the occurrence of earthquake swarms than the rifting effect of the Okinawa Trough axis.
As a result of the analysis, we were able to detect about 2 to 3 times as many earthquakes as the JMA catalog. The characteristics of earthquake swarms differ between the Central Okinawa Trough (COT) and the Southern Okinawa Trough (SOT), with the COT being dominated by local-type earthquakes (within 10 km of the active region) and the SOT being dominated by migration-type earthquakes (movement of 10 km or more). The 2007 activity was classified as aftershock-type only. In the ETAS model, if the parameter α is 0.7 or less, it is considered to be a earthquake swarm type, and if it is less than 0.5, it is considered to be the effect of dyke intrusion (Ishigaki, 20 10). In this study, the value of α was less than 0.7 for all of the activities, suggesting that they were triggered by the dyke intrusion. However, the reliability of the ETAS model may be low for the 2007 activity because the Mc value increased rapidly, reducing the detection capability. In the case of the swarm earthquakes in the SOT, the diffusion coefficient of Shapiro et al. (1999) was larger than that of previous studies (D = 10 km²/s), and the possibility of activity due to dyke intrusion is high. The area beneath Futagoyama, which is near the start of the earthquake swarms, is a possible magma reservoir. In the SOT, it is estimated that, as in Iceland and Afar, the progression of dyke intrusion is related to seismic activity. The ETAS model also suggested that the earthquake swarms off Amami Oshima Island were triggered by dyke intrusion, but since no crustal deformation was observed, it is thought that the scale was small. In addition, in the Okinawa Trough, regions with high and low b-values were identified, and the high b-values near the hydrothermal vents suggested the influence of hydrothermal activity (Minami, 2021). In the area off the northwest coast of Okinawa Island, earthquake swarms have been occurring in an area where traces of dyke intrusion have been observed near the axis of the rift, so there is a possibility that dyke intrusion is involved in the earthquakes swarm.
The Okinawa Trough earthquake swarms were more active near the volcanic front (80-90 km depth on the slab) and in the hydrothermal activity area than along the rift axis. It is more likely that the active magma formation at the volcanic front is responsible for the occurrence of earthquake swarms than the rifting effect of the Okinawa Trough axis.