Takuya NISHIMURADisaster Prevention Research Institute, Kyoto University

E-mail: nishimura.takuya.4s@kyoto-u.ac.jp

Acknowledgments　We thank the Geospatial Information Authority of Japan and the Japan Coast Guard for providing GNSS RINEX data. The Japan Meteorological Agency provided the unified catalogue of earthquakes.

ReferenceMiyaoka, K. and T. Yokota, Development of stacking method for the detection of crustal deformation -Application to the early detection of slow slip phenomena on the plate boundary in the Tokai region using strain data, J. Seismol. Soc. Jpn. (Zisin), 65, 205-218, 2012.

Nishikawa, T., Matsuzawa, T., Ohta, K., Uchida, N., Nishimura, T., and Ide, S. The slow earthquake spectrum in the Japan Trench illuminated by the S-net seafloor observatories. Science, 365(6455), 808-813, 2019.

Nishimura, T., Ozawa, S., Murakami, M., Sagiya, T., Tada, T., Kaidzu, M., and Ukawa, M. Crustal deformation caused by magma migration in the northern Izu Islands, Japan. Geophysical Research Letters, 28(19), 3745-3748, 2001.

Nishimura, T., T. Sagiya, and R. S. Stein, Crustal block kinematics and seismic potential of the northernmost Philippine Sea plate and Izu Microplate, central Japan, inferred from GPS and leveling data, J. Geophys. Res., 112, B05414, doi:10.1029/2005JB004102, 2007.

Nishimura, T., T. Matsuzawa, and K. Obara, Detection of short-term slow slip events along the Nankai Trough, southwest Japan using GNSS data, J. Geophys. Res. Solid Earth, 118, 3112-3125, doi:10.1002/jgrb.50222, 2013.

Nishimura, T., Short-term slow slip events along the Ryukyu trench, southwestern Japan, observed by continuous GNSS, Progress in Earth and Planetary Science, 1:22, doi:10.1186/s40645-014-0022-5, 2014.

Ozawa, S., Shortening of recurrence interval of Boso slow slip events in Japan. Geophysical Research Letters, 41(8), 2762-2768, 2014.

Uchida, N., Nakajima, J., Hasegawa, A., and Matsuzawa, T. What controls interplate coupling?: Evidence for abrupt change in coupling across a border between two overlying plates in the NE Japan subduction zone. Earth and Planetary Science Letters, 283(1), 111-121, 2009.

Key points・ More than 179 short- and imtermidiate- term SSEs (SSEs) with a duration of up to 80 day are detected using 26-years-long GNSS data in the Kanto region, central Japan. ・Two distinctive SSE clusters where relative plate motion is fully released by s-SSEs are found along the Sagami Trough. ・ SSEs rarely overlap shallow tremors and regular M>7 earthquakes along the Japan Trench. However, they overlap M5-6 interplate earthquakes considerably.

Detection of small short-term SSEs using GNSS Data

Short-term SSEs in the Kanto region, central Japan using GNSSdata for a quarter century

Flowchart of SSE analysisPre-process of GNSS data

Remove offsets and postseismic transientRemove noisy data

Start

Event detection using AIC

Finish

Fault model estimation

Rectangular fault on plate interface

Yes

NoShifting a time window

Modified from Nishimura et al.(2013), Nishimura(2014)

Fault model estimation

Rectangular fault on plate interface

Spatial filtering

Estimation of daily offsets With Step function

Comp. of relative plate motion

Duration estimation Stacking method of crustal deformation (Miyaoka and

Yokota, 2012)

Estimation of offsets With Ramp function NS, EW, and UD Comp.

Iteration

Fig. 2 Example of daily GNSS coordinates parallel to a relative plate motion and ΔAIC at 950443 station (Nishimura et al., 2013). Downward steps can be recognized, which is possibly caused by S-SSEs. Large -ΔAIC means significant step in time-series.

−0.05

−0.04

−0.03

−0.02

−0.01

0.00

0.01

0.02

0.03

0.04

0.05

0

20

40

60

80

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

N50ºW Component at 950443 (Sagawa, Kochi Pref.)

Dis

plac

emen

t (m

)-Δ

AIC

Year

Linear function w/wo a step

Kanto Double Subduction Zone and Boso SSEs

138 E 140 E 142 E

N 43

N 63Tokyo

Niijima

Inubousaki

Omaezaki

B

Ito

Odawara

100 km

ZTKNZTKN

Mt. Fuji

LTSI

LTSI

Izu

Boso Peninsula

Kanto

Tokai napaJTr

hcne

iaknaN

Trhguo

100 km

ZenisuMikurajima

KouzushimaimagaSTr hguo

NiijimaIzu Peninsula

Oshima

aguruSTr

hguo

Tokyo

Odawara

Omaezaki

Inubousaki

Ito

Fig. 4 Topographic map of the Kanto Region. Modified from Nishimura et al.(2007)

Fig. 7 Time-series of daily GNSS displacements for 26 years. Six Boso SSEs (black arrows) have been identified by previous studies. A linear trend is removed. See Fig. 6 for station location.

Fig. 5 Slip distribution of past 6 Boso SSEs (a-f) and their recurrence intervals (g). (Ozawa et al., 2019)

PAC

PHS

Data and Preprocessing

Fig. 6 Used GNSS stations and assumed slab geometry. The geometry is based on the model complied by Dr. F. Hirose. Time-series of A-E stations are shown in Fig.8.

SSEs detected on the PHS SSEs detected on the PAC

PAC SSEs controlled by subducting sea-mounts and overriding plates

SSECleaned time-series

’ SNR 1.48

’ SNR 2.69

’ SNR 2.16

’ SNR 2.69

’ SNR 2.04

’ SNR 2.19

Nor

mal

ized

am

plitu

de

100

0

Duration estimation by stacking of GNSS Timeseries

２ (SNR 3.86)

15(SNR 5.15)

40 (SNR 6.06)

72 (SNR 6.65)

200 (SNR 4.63)

Fitting of Ramp function

11 days

・Removal a linear trend・Normalized by noise level・Reverse polarity of data if predicted signal is negative.

・Stacking in order of high SNR.・Stop stacking at the highest SNR of stacking timeseries

②②

Stacking time-series ~ Source time function

Estimate duration by fitting a rump function

33˚

34˚

35˚

20 km

2 mmCal.Obs.

Horizontal displacement (Duration 1 days) 2012/01/0233.509N 132.722E D 30.5km L105.4km W 46.1km Strike 227 Dip 8 Rake 104 Slip 0.0144m Mw6.23 RD 2244.0

Category A

131˚ 132˚ 133˚ 134˚ 135˚

33˚

34˚

35˚

20 km4 mm

Obs. Cal.

Vertical displacement

Fig. 3 An example of an analysis procedure for stacking of GNSS Timeseries. Stacked timeseries are used to es-timate a source time function of small SSEs.

The Kanto region, central Japan is situated under complex tectonics where the Philippine Sea and the Pacific plates subduct from the Sagami Trough and the Japan Trench, respectively (Fig. 4).Several SSEs were reported by the previous studies [e.g., Ozawa et al., 2014]. Recently, Nishikawa et al. found tremors along the Japan trench However, spatiotemporal distribution of SSEs on both plates still remains unclear.

Fig. 9 Fault models for detected S-SSEs on the Pacific plate. (a) Rectangular fault models (blue rectangles) and their slip vectors (yellow vectors). (b) Cumulative slip of SSEs. Focal mechanism is plotted for Mw>4.6 interplate earthquakes (160º ≤ Strike ≤ 220º, 0º ≤ Dip ≤ 30º, and 45º ≤ Rake ≤ 135º) during 1997-2019. Green regions represent source regions of the past megathrust earthquakes.

Many shallow SSEs on the PAC locate in a region where the overriding plate is the North American plate, not the Philippine Sea plate (Fig. 11). This may reflect on a difference of interplate coupling controlled by geology of the overriding plate [Uchida et al., 2009]. It is also suggested that the SSE cluster at ~35.5ºN corresponds to a subducted sea-mount induced from a bathymetry and gravity anomaly.

Fig. 11 Distribution of the PAC SSEs and bathymetry. SSE clusters at ~35.5ºN corresponds to a subducted sea-mounts chain corresponds.

The used daily coordinates at 294 stations from April 25, 1994 to February 21, 2020 are estimated using GIPSY 6.4 software. Offsets related M≥6 earthquakes and maintenance are removed. Although co- and post- seismic displacement of the 2011 Tohoku-oki earthquake is removed by fitting Heaviside, logarithmic, and exponential functions, we excluded a period from December 11, 2010 to September 11, 2011 for the SSE detection because of large scattered data. The transient displacement related the 2000 Miyake-Kouzu volcanic event is also removed by using the model prediction (Nishimura et al., 2001). N80W, N25W, and N40W components are used to detect steps for SSEs along the Japan Trench, Sagami Trough, and Suruga Trough, respectively.

We apply the method of Nishimura et al. (2013) and Nishimura (2014) to detect a jump associated with short-term SSEs in GNSS time-series and estimate their fault models from observed displacements. A rectangular fault on the Philippine Sea or the Pacific plates is assumed for each SSE. The stacking of GNSS time-series based on the displacement predicted by the fault model [Miyaoka and Yokota, 2012] enable us to estimate duration of SSEs by fitting a ramp function.

Fig. 1 Flowchart of SSE analysis

(Fig. 2)

(Fig. 3)

(Miyaoka and Yokota, 2012)

Boso SSEs

Fig. 8 Fault models for detected S-SSEs on the Philippine Sea plate. (a) Rectangular fault models (blue rectangles) and their slip vectors (yellow vectors). (b) Cumulative slip of SSEs. Focal mechanism is plotted for Mw>4.6 interplate earthquakes (240º ≤ Strike ≤ 300º, 0º ≤ Dip ≤ 30º, and 45º ≤ Rake ≤ 135º) during 1997-2019. Green regions represent source regions of the past megathrust earthquakes.

NA

−0.20

−0.15

−0.10

−0.05

0.00

0.05

0.10

0.15

0.20

Dis

plac

emen

t (m

)

94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19

F 93047 N25W

E 93041 N25W

D 950226 N25W

C 93022 N80W

A 950216 N80W

B 93009 N80W

Fig. 10 Moment evolution of SSEs in the three analyzed region. A gray period is excluded in the analysis because of postseismic deformation of the 2011 Tohoku-oki earthquake.

Findings and Implications・Two distinctive SSE patches (Regions A and B)・Total slip on these patches is equal to the relative plate motion.・Medium-size interplate earthquakes occur at a downdip edge

of SSEs patches.・Small cumulative slip in the region of past megathrust

Temporal evolution of SSE moment release

Findings and Implications・Many SSEs occur in two depth ranges, i.e., 10-20 km and 40-60 km.・Little overlapping of S-SSEs and tremor regions in the shallow(10-20 km) depth range.・Both SSEs and fast earthquakes coexist in the deep depth range (40-60 km). It is partly attributed

to detected SSEs include co- and post-seismic slip events of M~6 earthquakes.・No significant change of SSE slip behavior at the 2011 Tohoku-oki earthquake (Fig. 10).・Small cumulative slip in the region of past megathrust earthquakes.

137˚ 138˚ 139˚ 140˚ 141˚ 142˚ 143˚34˚

35˚

36˚

37˚100 km

80

60

4020

100 80

60

40

20

Japa

n-Iz

u-Bo

nin

Tren

chSagami Trough

Suru

ga T

roug

h

1923 Kanto

2011 Tohoku-oki

2008 Ibaraki-oki

2011 Tohoku-oki AS

Boso SSE

Tokai SSE

Kanto

Tokai

IzuIslands

F

DC

A

GNSS stationTremorAIC point

B

E

137˚ 138˚ 139˚ 140˚ 141˚ 142˚34˚

35˚

36˚

37˚

50 km

8060

−40

20

139˚ 140˚ 141˚ 142˚ 143˚34˚

35˚

36˚

37˚

50 km

100

80 60

40 20

0.00.10.20.30.40.50.60.70.80.91.01.11.2

Tota

l slip

(m)

137˚ 138˚ 139˚ 140˚ 141˚ 142˚34˚

35˚

36˚

37˚

80

−60 −40

−20

50 km

0.5 0.5

1

1923

Tokai

Tonankai

1923

Tokai

Tonankai0.00.10.20.30.40.50.60.70.80.91.01.11.21.31.41.5

Tota

l slip

(m)

139˚ 140˚ 141˚ 142˚ 143˚34˚

35˚

36˚

37˚

0.50.

5

0.5

11.5

50 km

100

80 60

−40

−20

2011

2008

2011AS2011AS

2008

2011(A) (B) (A) (B)

139˚ 140˚ 141˚ 142˚ 143˚34˚

35˚

36˚

37˚SSE on PAC plate

0

0.5×1020

1×1020

1.5×1020

2×1020

1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 20180

1×1020

2×1020

3×1020

4×1020

Calendar year

Mom

ent f

or J

apan

Tre

nch

(Nm

)

Japan TrenchSagami Trough

Suruga Trough

Region ARegion B

Mom

ent f

or S

agam

i and

Sur

ga T

roug

hs (N

m)

G3.6 D1727

©Takuya Nishimura. All rights reserved