14:45 〜 15:00
[SCG39-11] Observation of shallow slow slip event propagating updip to the trough in the Nankai Trough.
キーワード：ゆっくり滑り、南海トラフ、海底観測、スロー地震、海底孔内観測
We report observation of shallow slow slip in Kumano-nada, the Nankai Trough, south of Japan in December 2020 to January 2021 period from nearby seafloor and seafloor borehole observations. We investigated long-term pore-fluid pressure, and tilt data at three seafloor borehole stations (IODP C0002, C0010, C0006), a seafloor tilt station data (BMS1), as well as 13 DONET seafloor pressure station data, to identify slow crustal deformation. Seafloor broadband seismic data were also analyzed to identify low frequency earthquakes (LFT) and very low frequency earthquakes (VLFE) activity.
The observed episodic slow event is hereby reported in several stages (Stage A-D).
Stage A (Dec. 3-13): The slow event was first observed at Dec. 3, 2020 by the seafloor tilt station (BMS1) in DONET1-B Node with no accompanied LFT and VLFE activity, with firstly southward tilting and then turned to northward tilting, through stagnant period in Dec. 6-11 when active burst of LFT and VLFE activity started.
Stage B (Dec. 13-29): After two days in Dec. 13, slow change also started at borehole stations in the south-western side. Observed pore-fluid pressure change were trending continuously negative for C0002 (-8.7 kPa) and C0010 (-9.2kPa) and positive for C0006 (nearest to the trough in Node C, 15.4kPa). These changes were much larger than past episodes reported (Araki et al., 2017 and Ariyoshi et al., 2021). The seafloor and borehole tilt data also responded with large tilt change as large as 1.9 (BMS1) ~9 (C0010) and ~90 microradians (C0006) respectively. Seafloor pressure data between nearby stations in Nodes B, C and D also showed significant relative differences with slow transition, such as +1hPa (KMD06 and KMD07-KMD05), +2hPa (KMC11 and KMC12-KMC09). Pressure KMC12-KMC09 went through notable negative polarity change of ~ -1 hPa in +2 hPa change in overall, suggesting crustal deformation migrated beneath KMC09 site during the period. LFT and VLFE activity during the stage was active.
Stage C (Dec. 29-Jan. 10, 2021): While C0002 and C0010 pore-fluid pressure data continues to trend negatively, C0006 changed its trend from positive to negative after Dec. 29, suggesting the source of observed crustal deformation migrated from north to south beyond C0006 located within 10 km from the trough. Seafloor pressure between KMC11(at the trough axis)-KMC09 exhibited significant change of –1.3 hPa during the period, also suggests crustal deformation migration reached to the trough.
Stage D (Jan. 10-29): Intermittent slow changes in pore-fluid pressure were observed in C0010 and C0006 accompanying local LFT activity.
We performed slip fault modeling to account observed tilt and pore-fluid pressure change in the stages to understand what caused observed crustal deformation, using grid search of best fit Okada models for fault geometry and slip. Modeling was performed in time segments in these stages, first assuming the fault on the subducting plate boundary, and model with higher fault dipping angle to check validity of the assumption. In Dec. 16-21 period in the stage B, the best fit model was Mw 6.1 slip located in south of DONET1-B node, where LFT activity was concentrated. Mechanisms of VLFE events during the period was also analyzed to give pronounced low angle reverse fault solutions. Therefore, we consider the observed event in the period was slow slip in the shallow subducting plate interface, probably initiated downdip in Kumano Basin, propagating updip on the subducting Philippine Sea plate interface toward the trough.
The borehole pore-fluid trend polarity change observed at C0006 (< 10km from the trough axis) in the stage C to D as well as seafloor pressure change at KMC11, strongly suggests crustal deformation took place very close to the trough axis where the plate subduction interface originated. From these direct observations in the seafloor above the plate interface, we argue that the slow slip event reached to the trough.
The observed episodic slow event is hereby reported in several stages (Stage A-D).
Stage A (Dec. 3-13): The slow event was first observed at Dec. 3, 2020 by the seafloor tilt station (BMS1) in DONET1-B Node with no accompanied LFT and VLFE activity, with firstly southward tilting and then turned to northward tilting, through stagnant period in Dec. 6-11 when active burst of LFT and VLFE activity started.
Stage B (Dec. 13-29): After two days in Dec. 13, slow change also started at borehole stations in the south-western side. Observed pore-fluid pressure change were trending continuously negative for C0002 (-8.7 kPa) and C0010 (-9.2kPa) and positive for C0006 (nearest to the trough in Node C, 15.4kPa). These changes were much larger than past episodes reported (Araki et al., 2017 and Ariyoshi et al., 2021). The seafloor and borehole tilt data also responded with large tilt change as large as 1.9 (BMS1) ~9 (C0010) and ~90 microradians (C0006) respectively. Seafloor pressure data between nearby stations in Nodes B, C and D also showed significant relative differences with slow transition, such as +1hPa (KMD06 and KMD07-KMD05), +2hPa (KMC11 and KMC12-KMC09). Pressure KMC12-KMC09 went through notable negative polarity change of ~ -1 hPa in +2 hPa change in overall, suggesting crustal deformation migrated beneath KMC09 site during the period. LFT and VLFE activity during the stage was active.
Stage C (Dec. 29-Jan. 10, 2021): While C0002 and C0010 pore-fluid pressure data continues to trend negatively, C0006 changed its trend from positive to negative after Dec. 29, suggesting the source of observed crustal deformation migrated from north to south beyond C0006 located within 10 km from the trough. Seafloor pressure between KMC11(at the trough axis)-KMC09 exhibited significant change of –1.3 hPa during the period, also suggests crustal deformation migration reached to the trough.
Stage D (Jan. 10-29): Intermittent slow changes in pore-fluid pressure were observed in C0010 and C0006 accompanying local LFT activity.
We performed slip fault modeling to account observed tilt and pore-fluid pressure change in the stages to understand what caused observed crustal deformation, using grid search of best fit Okada models for fault geometry and slip. Modeling was performed in time segments in these stages, first assuming the fault on the subducting plate boundary, and model with higher fault dipping angle to check validity of the assumption. In Dec. 16-21 period in the stage B, the best fit model was Mw 6.1 slip located in south of DONET1-B node, where LFT activity was concentrated. Mechanisms of VLFE events during the period was also analyzed to give pronounced low angle reverse fault solutions. Therefore, we consider the observed event in the period was slow slip in the shallow subducting plate interface, probably initiated downdip in Kumano Basin, propagating updip on the subducting Philippine Sea plate interface toward the trough.
The borehole pore-fluid trend polarity change observed at C0006 (< 10km from the trough axis) in the stage C to D as well as seafloor pressure change at KMC11, strongly suggests crustal deformation took place very close to the trough axis where the plate subduction interface originated. From these direct observations in the seafloor above the plate interface, we argue that the slow slip event reached to the trough.