17:15 〜 18:30
[SCG39-P28] A comparison of the source areas in Hyuganada among afterslip of two interplate earthquakes in 1996 and successive slow slip events
キーワード：GNSS、スロー地震、沈み込み帯、非地震性すべり、九州パラオ海嶺
Around the Hyuganada in southwestern Japan, the Philippine Sea Plate (PHS) is subducting beneath the Amurian Plate. It is known that afterslip and long-term slow slip events (SSEs) occur on the plate interface around this region. Afterslip occurs following two interplate earthquakes in Hyuganada (both are Mw 6.7) in October and December, 1996 (Yagi et al., 2001) and SSEs repeatedly occur near the afterslip area (e.g., Yarai and Ozawa et al., 2013). However, little is known about the spatiotemporal relationship between the afterslip and the SSEs. In this study, we identify the time periods of aseismic slip episodes including the afterslip and SSEs that occurred from 1996 to 2011 and estimate detailed spatiotemporal evolutions from an inversion analysis based on GNSS data.
We use daily coordinates of GEONET GNSS stations operated by Geospatial Information Authority of Japan (GSI) (F3 solutions: Nakagawa et al., 2009). From the coordinate time-series data, we identify the slow transients and their time periods possibly of the afterslip and SSEs by reference to the transients of SSEs reported in previous studies (Yarai and Ozawa, 2013; Takagi et al., 2019).
Five transients are identified. One is afterslip from Oct .1996 to Nov. 1997 and the others are four SSEs: (1) Jun. 1998 - Mar. 1999; (2) Jan. 2005 - Mar. 2006; (3) Jan. 2007 - Nov. 2007; and (4) Apr. 2009 - Apr. 2010. For each transient, we apply an inversion method based on network inversion filter (Segall and Matthews, 1997; Hirose et al., 2014). In the inversion analysis, a spatial distribution of slip is represented by 8 × 7 small subfaults with a size of 20 km x 20 km. These subfaults are assumed to represent the geometry of the plate interface of the subducting PHS estimated by Shiomi et al. (2008) and Nakanishi et al. (2018).
Our result shows clear time variation of the afterslip area. In the time period between the first and the second earthquakes in 1996, slip area extends on the plate interface within the depth range from about 15 km to 50 km. But after the second earthquake, slip at shallower part near the epicenter of the first earthquake (< ~25 km depth) gradually decreases and ends around May 1997, whereas deeper slip continues until Nov 1997. In addition, source areas of the SSEs are located in and around the afterslip area. We also identified additional slip area in the offshore area of Cape Toi, south of the afterslip area, during the transient periods (1), (2), and (4). Among them, the southern slip area is spatially separated from the northern slip during (1) and (2). Comparing the obtained slip distribution of the afterslip with the cumulative slip distribution of the four SSEs, the SSE slip is located in the deeper part of the afterslip area and south of the afterslip.
From the above characteristics of the afterslip and the SSE slip, it is suggested that the deeper area of the afterslip may host longer (or slower) slip and SSE slip while the shallower area may host shorter (or faster) slip. Yamamoto et al. (2013) show the subducted Kyushu-Palau Ridge (KPR) in the PHS slab from a seismic exploration. Comparing the cumulative slip distribution of the SSEs with the location of subducted KPR, the main slip area of the SSEs is located deeper and outside of the KPR. This suggests that the subducted KPR controls the slip behavior on the plate interface.
Acknowledgments:
GNSS data were provided by Geospatial Authority of Japan. This research was supported by JSPS KAKENHI Grant Number JP16H06474 in Scientific Research on Innovative Areas "Science of Slow Earthquakes".
We use daily coordinates of GEONET GNSS stations operated by Geospatial Information Authority of Japan (GSI) (F3 solutions: Nakagawa et al., 2009). From the coordinate time-series data, we identify the slow transients and their time periods possibly of the afterslip and SSEs by reference to the transients of SSEs reported in previous studies (Yarai and Ozawa, 2013; Takagi et al., 2019).
Five transients are identified. One is afterslip from Oct .1996 to Nov. 1997 and the others are four SSEs: (1) Jun. 1998 - Mar. 1999; (2) Jan. 2005 - Mar. 2006; (3) Jan. 2007 - Nov. 2007; and (4) Apr. 2009 - Apr. 2010. For each transient, we apply an inversion method based on network inversion filter (Segall and Matthews, 1997; Hirose et al., 2014). In the inversion analysis, a spatial distribution of slip is represented by 8 × 7 small subfaults with a size of 20 km x 20 km. These subfaults are assumed to represent the geometry of the plate interface of the subducting PHS estimated by Shiomi et al. (2008) and Nakanishi et al. (2018).
Our result shows clear time variation of the afterslip area. In the time period between the first and the second earthquakes in 1996, slip area extends on the plate interface within the depth range from about 15 km to 50 km. But after the second earthquake, slip at shallower part near the epicenter of the first earthquake (< ~25 km depth) gradually decreases and ends around May 1997, whereas deeper slip continues until Nov 1997. In addition, source areas of the SSEs are located in and around the afterslip area. We also identified additional slip area in the offshore area of Cape Toi, south of the afterslip area, during the transient periods (1), (2), and (4). Among them, the southern slip area is spatially separated from the northern slip during (1) and (2). Comparing the obtained slip distribution of the afterslip with the cumulative slip distribution of the four SSEs, the SSE slip is located in the deeper part of the afterslip area and south of the afterslip.
From the above characteristics of the afterslip and the SSE slip, it is suggested that the deeper area of the afterslip may host longer (or slower) slip and SSE slip while the shallower area may host shorter (or faster) slip. Yamamoto et al. (2013) show the subducted Kyushu-Palau Ridge (KPR) in the PHS slab from a seismic exploration. Comparing the cumulative slip distribution of the SSEs with the location of subducted KPR, the main slip area of the SSEs is located deeper and outside of the KPR. This suggests that the subducted KPR controls the slip behavior on the plate interface.
Acknowledgments:
GNSS data were provided by Geospatial Authority of Japan. This research was supported by JSPS KAKENHI Grant Number JP16H06474 in Scientific Research on Innovative Areas "Science of Slow Earthquakes".