site amplification and strong ground motion of the 2007
TRANSCRIPT
Earth Planets Space, 60, 161–167, 2008
Site amplification and strong ground motion of the 2007 Noto Hanto, Japan,earthquake estimated from aftershock observation
Masayuki Yoshimi and Kunikazu Yoshida
Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki 305-8567, Japan
(Received July 10, 2007; Revised September 18, 2007; Accepted October 4, 2007; Online published March 3, 2008)
Site amplifications in the lowlands most affected by the 2007 Noto Hanto earthquake, Monzen, Anamizu,and Wajima, are examined using aftershock records observed at eight temporary seismic stations installed justafter the mainshock and at two K-NET stations. The predominant frequencies of spectral ratios at alluviumsites in Anamizu and Wajima are approximately 1 Hz. Site amplifications at the alluvium sites are successfullyreproduced from 1-D response analysis, except for that at ISK005 where 2-D or higher amplification effects areinferred to play a significant role. A source model composed of two asperities reproduces the ground motionsof the mainshock using the empirical Green’s function method. The seismic moments of the asperities are3.76×1018 N m and 2.21×1018 N m, respectively. Peak ground velocity (PGV) at alluvium sites during themainshock are estimated to be 70–110 cm/s for Monzen, 50–110 cm/s for Anamizu, and 60–70 cm/s for Wajima.Key words: Spectral ratio, aftershock observation, local site effects, source model, empirical Green’s functionmethod.
1. IntroductionThe Noto Hanto earthquake occurred on 25 March 2007
on the northwestern coast of Noto Peninsula in Ishikawaprefecture, central Japan. This was a shallow, moderate-sized earthquake (Mj 6.9) whose fault plane extended in theNE–SW direction across the coast line and dipped towardsthe southeast. Seismic intensities of 6+ on the JMA (JapanMeteorological Agency) scale—the second-largest class—were produced in the alluvium plains within 30 km from theepicenter, Monzen, Anamizu, and Wajima. Many woodenhouses in these areas were damaged, but severe destruction,such as total collapse, was limited to small sub-regions inapparent response to local site effects.Waveforms of strong ground motion during the main-
shock were recorded at seismographs in the intenselyshaken areas in the Anamizu and Wajima lowlands. At oneof the K-NET (strong-motion seismograph network) sta-tions, ISK005, operated by the National Research Institutefor Earth Science and Disaster Prevention in the vicinity ofa severely damaged subregion of Amamizu, a peak groundvelocity (PGV) of about 100 cm/s was observed. At twostations within 1.5 km of each other in Wajima, ISK003,and JMA Wajima (an alluvium site), a clear contrast wasrecorded in shaking intensity. PGV at JMA Wajima, in themiddle of a severely damaged subregion, was 100 cm/s,more than twice as large as that at ISK003 (44 cm/s) andlikely due to a nonlinear soil response.This paper summarizes some results of our research into
ground motions in the damaged areas. We focus mainlyon aftershock observations in the Anamizu, Wajima, and
Copyright c© The Society of Geomagnetism and Earth, Planetary and Space Sci-ences (SGEPSS); The Seismological Society of Japan; The Volcanological Societyof Japan; The Geodetic Society of Japan; The Japanese Society for Planetary Sci-ences; TERRAPUB.
Monzen lowlands in order to evaluate site amplificationsand on source modeling in order to estimate ground motionsin the lowlands during the 2007 mainshock.
2. Aftershock ObservationsAftershock observations were conducted in some of the
most intensely shaken lowlands. Eight temporary obser-vation stations had been installed in and around the low-lands by the morning of 27 March 2007, 2 days afterthe mainshock. Two of these temporary stations were inAnamizu, three were in Wajima, and the remaining threewere in Monzen. Each station was equipped with a three-component accelerometer, a 24-bit digital recorder withGPS time calibration, and a battery. These instrumentsrecorded ground motions continuously for 3 weeks. Thelocations of our temporary stations and the permanent K-NET stations are shown in Fig. 1 and listed in Table 1. InAnamizu, stations AYK, APO, and ISK005 define a lineararray with instrument spacing of approximately 400 m. Sta-tion AYK is a rock site on the ridge bounding the easternAnamizu lowland; APO is an alluvium site between AYKand ISK005; ISK005 is an alluvium site in the vicinity ofa severely damaged subregion. In Wajima, stations WFG,WKW, WSB, and ISK003 form a 1.5-km-long linear arraywith instrument spacings of 200–600 m spanning the low-land. WFG is a rock site with thin alluvium deposits onthe western ridge bounding the lowland; ISK003 is also arock site on the eastern bounding ridge; WKW and WSBare alluvium sites. In Monzen, all three stations are sitedon alluvium sites in severely damaged subregions. MHSand MJD are within 200 m of the city office where a shak-ing intensity of 6+ was observed. MTG is in the west ofthe lowland where the total building collapse rate was thehighest of the affected areas.
161
162 M. YOSHIMI AND K. YOSHIDA: SITE AMPLIFICATION AND GROUND MOTIONS IN DAMAGED AREAS, NOTO, JAPAN
136.5 137
37
37.5 20 km07/03/25/09:42 Mj6.9
Anamizu
Wajima
Monzen
1 km
AYKAPO
ISK005
1 km
WSBWKW
WFG
ISK003
1 km MHS
MJDMTG
130 14030
40 Noto
J A P A N
Fig. 1. Locations of temporary seismic observation stations in Monzen,Anamizu, and Wajima. Black triangles are our observation stations.Gray triangles are K-NET stations operated by the National ResearchInstitute for Earth Science and Disaster Prevention (NIED). Gray dotsare epicenters of the mainshock and aftershocks (M ≥ 3) prior to 6April 2007, as determined by the Japan Meteorological Agency.
Table 1. List of observation stations in Monzen, Anamizu, and Wajima.
Station Latitude Longitude Start time Geology*
MHS 37.28644 136.76894 03/26 17:50 alluvium
MJD 37.28875 136.76589 03/27 10:25 alluvium
MTG 37.28872 136.74017 03/27 09:10 alluvium
AYK 37.22975 136.91278 03/26 12:00 rock
APO 37.22930 136.90822 03/26 12:30 alluvium
WSB 37.39267 136.90556 03/26 15:25 alluvium
WKW 37.39436 136.89986 03/26 16:00 alluvium
WFG 37.39628 136.89467 03/26 16:30 rock
*Ishikawa Pref. (1991, 1993).
3. Analysis of Aftershock RecordsA large number of aftershocks were successfully
recorded. In Fig. 2, we show examples of observed groundaccelerations at our temporary stations and the nearbyK-NET stations, ISK003 and ISK005, during the eventsat 10:51 (JST) on 28 March 2007 and 15:18 (JST) on 6April 2007. The waveforms have been bandpass-filteredto between 0.05 and 24 Hz. The horizontal peak values ofground acceleration at the alluvium sites span similar rangesto those at the rock sites. However, the waveforms at thealluvium sites are distinguishable by their pronounced low-frequency components, which indicate site amplification.The spectral amplitude ratios of the alluvium sites APO,
ISK005, WSB, and WKW with respect to the rock sites inthe frequency range 0.2–10 Hz are illustrated in Fig. 3. Sta-tions AYK and ISK003 have been selected as the referencesites in Anamizu and Wajima, respectively. The horizon-tal components of 15 aftershocks in Anamizu and 12 af-tershocks in Wajima, with JMA magnitudes of 3.5–4.9, areshown. Smoothing with a 0.2-Hz bandwidth Parzen Win-dow has been applied in calculating the spectral ratios. Thetime window is 10.25 s long and starts at the S-wave onset.Predominant frequencies are approximately 1 Hz for all thestation pairs. Large amplitude ratios of 5–20 are obtained atthe spectral peaks.
T0=2007/03/28 10:51:5
2007/03/28/10:51 M4.6
-150
15
[gal
]
EWMax=7.3
Max=13.7Max=9.4Max=5.3Max=4.3
Max=10.9Max=4.2Max=3.5Max=4.8Max=3.6
NSMHS
Max=10.8
MJDMax=10
MTGMax=8.5
AYKMax=5.2
APOMax=7.6
ISK005Max=13
WFGMax=2.3
WKWMax=2.6
WSBMax=6.6
ISK003Max=5.2
UDMax=5.4Max=6.3Max=5.1Max=4.3Max=4.8Max=5.4Max=2.2Max=2.3Max=2.5Max=2.1
0 5 10 15Time[s]
T0=2007/04/06 15:18:18
2007/04/06/15:18 M4.3
-750
75
[gal
]
EWMax=27.4Max=35.2Max=19.6Max=50.3Max=37.3Max=94.5
Max=8.9Max=8.1Max=12
Max=17.2
NSMHS
Max=39.8
MJDMax=50.8
MTGMax=23.2
AYKMax=55.6
APOMax=53.3
ISK005Max=33.6
WFGMax=7.5
WKWMax=7.5
WSBMax=20
ISK003Max=11.4
UDMax=28.6Max=18.9Max=14.1Max=32.2Max=34.8Max=32.7
Max=8.5Max=7
Max=11Max=5.8
0 5 10 15Time[s]
Monzen
Anamizu
Wajima
Monzen
Anamizu
Wajima
Monzen
Anamizu
Wajima
Monzen
Anamizu
Wajima
Monzen
Anamizu
Wajima
Monzen
Anamizu
Wajima
Fig. 2. Acceleration waveforms observed at our temporary arrays andK-NET stations. Upper: the event at 10:51 (JST) on 28 March 2007.Lower: the event at 15:18 (JST) on 6 April 2007. From top to bottom:the NS, EW, and UD components of each station. AYK, WFG, andISK003 are rock sites.
Theoretical spectral ratios calculated from 1-D velocitystructures (Fig. 4) are illustrated as solid curves in Fig. 3.The 1-D structure for ISK005 is inferred from P-S loggingdata at the site. Structures for the other stations are modeledmanually based on logging parameters from nearby bore-
M. YOSHIMI AND K. YOSHIDA: SITE AMPLIFICATION AND GROUND MOTIONS IN DAMAGED AREAS, NOTO, JAPAN 163
ISK005/AYK
0.1
11
10
100
Spe
ctra
l_ra
tio
11 10
Frequency[Hz]
Average-EWAverage-NS
Theoretical
APO/AYK
0.1
11
10
100
Spe
ctra
l_ra
tio11 10
Frequency[Hz]
Average-EWAverage-NS
Theoretical
WKW/ISK003
0.1
11
10
100
Spe
ctra
l_ra
tio
11 10
Frequency[Hz]
Average-EWAverage-NS
Theoretical
WSB/ISK003
0.1
11
10
100
Spe
ctra
l_ra
tio
11 10
Frequency[Hz]
Average-EWAverage-NS
Theoretical
Fig. 3. Amplitude spectral ratios of the horizontal components of alluviumsites with respect to the reference sites (AYK for stations in Anamizuand ISK003 for stations in Wajima) in Anamizu (upper) and Wajima(lower). Broken lines show the averages of the ratios. Solid lines aretheoretical spectral ratios from 1-D velocity structures (Fig. 4) with a1-D theoretical response analysis.
0
10
20
30
40
0 200 400 600
Vs
Rho
1.0 1.5 2.0 2.5
ISK005
Vs (m/s)
Density (g/cm )3
Dep
th (
m)
0
10
20
30
40
0 200 400 600
Vs
Rho
1.0 1.5 2.0 2.5
APO
0
10
20
30
40
0 200 400 600
Vs
Rho
1.0 1.5 2.0 2.5
WKW
0
10
20
30
40
0 200 400 600
Vs
Rho
1.0 1.5 2.0 2.5
WSB
ref.
ref.
ref.ref.
Fig. 4. 1-D subsurface velocity structure models, S-wave velocity anddensity, for alluvium sites in Anamizu and Wajima. Inverted trianglesdenote the depths of the reference layers assumed in the 1-D responseanalysis.
holes, such as N -values of the SPT (standard penetrationtest) and lithology (Sato, 1994), in order to fit the frequen-cies of peaks and troughs in the theoretical spectral ratiosto the observations. The S-wave velocities of the referencelayers are assumed to be 600 m/s and 500 m/s at AYK andISK003, respectively. The attenuation factor for each layeris given as a function of S-wave velocity depending on thetype of soil after Kobayashi et al. (1999) and Kyuke et al.(1999). The depths to the basement of the 1-D structures inWajima are consistent with those estimated by Sato (1994).The frequencies and amplitudes of the theoretical spec-
tral peaks agree well with the observed ones at frequencies
136.5 137 137.5
37
37.5
20 km
Mainshock 07/03/25/09:42 Mj6.9
EGF107/03/28/10:51 Mj4.6 10.2km
EGF207/04/06/15:18 Mj4.3 11.7km ISK001
ISK002
ISK003
ISK004
ISK005
ISK006
ISK007
ISK008
ISK009
TYM002
TYM005
TYM007
Fig. 5. Location map of the K-NET observation stations and epicentersof the aftershocks used as empirical Green’s functions. The largest rect-angle on the focal area indicates the fault plane assumed in the forwardmodeling of the mainshock, and the other rectangles are estimated as-perities. The fault plane is 22 km (along-strike) by 20 km (down-dip),the strike is 58◦, and the dip is 60◦. The star shows the location of therupture nucleation point.
of up to approximately 5 Hz. We have confirmed that withthese 1-D structures the ground motions during aftershocksat the alluvium sites, except at ISK005, are well reproducedfrom the ground motions at the reference sites using a 1-Dresponse analysis. The amplitude ratios at ISK005 are sev-eral times greater than those at APO, though the frequenciesof the spectral peaks are similar, indicating similar velocitystructures in the linear 1-D regime. The observed wave-forms at ISK005 are characterized by a large S-wave com-ponent and later phases several times larger than those seenat APO (see Fig. 2). Large amplification of the S-wavetraincan be accounted for if the acoustic impedance at the rock-sediments boundary is made larger. However, the laterphases cannot be reproduced by a 1-D analysis. Moreover,the amplification of the EW component of ISK005 is abouttwice that of the NS component (see Fig. 3). We conclude,therefore, that ISK005 is one location where 2-D or higher-dimensional amplification effects predominate.
4. Source Modeling of the Mainshock Using theEmpirical Green’s Function Method
4.1 Data and methodWe have determined a mainshock source model that
yields broadband ground motions consistent with observa-tions using a forward modeling approach incorporating theempirical Green’s function method (Irikura, 1986; Kamaeand Irikura, 1998). We seek a source model composed oftwo asperities, as two distinct pulses can be recognized insome of the observed waveforms. Two aftershocks, referredto hereafter as aftershock-1 and aftershock-2, are selected assources for the empirical Green’s functions (EGFs). Bothhave focal mechanisms similar to that of the mainshock,and their waveforms are recorded at our temporary stationsas well as at K-NET stations within 100 km of the epicen-ter. The locations of the K-NET stations that recorded both
164 M. YOSHIMI AND K. YOSHIDA: SITE AMPLIFICATION AND GROUND MOTIONS IN DAMAGED AREAS, NOTO, JAPAN
Table 2. Source parameters of the mainshock.
Mainshock
Origin time* (JST) 2007/03/25 09:41:57.9
Hypocenter* (136.686E, 37.221N, 10.7 km)
MJMA* 6.9
Seismic moment (N m)** 1.36×1019
Focal mechanism** (58, 66, 132)
(Strike, dip, rake) (173, 48, 34)
*JMA, **F-net.
Table 3. Source parameters of the aftershocks used as empirical Green’sfunctions.
Aftershock-1
Origin time* (JST) 2007/03/28 10:51
Hypocenter* (136.612E, 37.176N, 10.2 km)
MJMA 4.6
Mo (N m) ** 1.29×1015
Area (km2) 1.21
�σ (MPa)*** 2.36
Aftershock-2
Origin time* (JST) 2007/04/06 15:18
Hypocenter* (136.790E, 37.267N, 11.7 km)
MJMA 4.3
Mo (N m) ** 8.18×1014
Area (km2) 0.64
�σ (MPa)*** 3.89
*JMA, **F-net, ***Brune (1970, 1971).
aftershocks are shown in Fig. 5 together with the epicentersand the focal mechanisms of the mainshock and the EGFaftershocks. The locations of the assumed fault plane andthe computed asperities are also shown. The fault plane’slocation is based on fault models proposed from coastaldeformation (Awata et al., 2008), GPS and InSAR (Ge-ographical Survey Institute, 2007), and strong motion in-version (Horikawa, 2008). Tables 2 and 3 list the sourceparameters of the mainshock and aftershocks, respectively.The source parameters of the aftershocks, rupture area, andstress drop are estimated from displacement source spec-tra computed from acceleration data recorded at KiK-netborehole stations ISKH02, ISKH03, ISKH04, and GIFH10(Brune, 1970, 1971; Madariaga, 1976).4.2 Source model and syntheticsThe source model obtained from the forward modeling is
shown in Fig. 6 and annotated with the source parametersof each asperity. Rupture starts at the nucleation point andpropagates radially at a velocity of 2.8 km/s. An S-wavevelocity of 3.5 km/s is assumed along the wave propa-gation path. Aftershock records bandpass-filtered to be-tween 0.2 and 10 Hz are used as EGFs. The waveformsfrom aftershock-1 and aftershock-2 are used for synthesiz-ing waveforms from the eastern asperity (Asp1) and thewestern asperity (Asp2), respectively. Asp1 accounts forthe overall features of waveforms recorded in the north andthe southwest of the Noto Peninsula, but Asp2 is required toreproduce the sharp velocity pulses seen in the waveformsfrom ISK005 and ISK007. The seismic moments releasedfrom the asperities are estimated to be 3.76 × 1018 N m
0
5
10
15
20
Dis
tanc
e al
ong
dip[
km]
0 5 10 15 20
Distance along strike[km]
N58EN122W
Asp1
Asp2
60 d
eg.
Vr = 2.8 km/s
hypocenter
9.9 x 9.9
Asp1
Asp2
3.76 x 1018
[Nm][km ]2
0.69.44[MPa]
[s]
MoL x W
Rise time
4.0 x 122.21 x 10
18[Nm]
[km ]2
0.515.6[MPa]
[s]
Mo
L x W
Rise time
Source parameters
NL = 9, NW = 9, ND = 9Stress parameter 4.0
NL = 5, NW = 15, ND = 9Stress parameter 4.0
Fig. 6. Source model composed of two asperities, Asp1 and Asp2,obtained from forward modeling. Rupture propagates at the velocity2.8 km/s from the rupture nucleation point shown as a gray star. Sourceparameters for the asperities, seismic moment: Mo (N m), asperity size:L × W (km2), stress drop: �σ (MPa) (Brune, 1970, 1971), rise time,scaling parameters (NL, NW, ND), and stress parameter are listed.
for Asp1 and 2.21 × 1018 N m for Asp2, the sum of whichamounts to about 44% of the whole event.Comparisons of the synthetic and observed waveforms at
six stations are presented in Fig. 8 and pseudo-velocity re-sponse spectra (PVRS) for the waveforms at all the stationsin Fig. 5 are shown in Fig. 7. The velocity and displace-ment synthetic waveforms and their peak values show goodagreement with the observed ones for all the stations andcomponents—with the exception of the NS component atISK006. The PGVs of the synthetic waveforms are 70–150% of those observed. The envelopes of the syntheticacceleration waveforms are also consistent with the obser-vations for all stations—except for ISK006 and ISK005.The poor fit of the displacement waveforms at ISK006, es-pecially for the NS component, is attributable to a differ-ence in the radiation coefficient between aftershock-1 andthe mainshock produced by the large azimuth discrepancy.The PVRS values of the synthetics agree well with thoseof the observations over a broad period range. However, atthe shorter-period end of the PVRS spectrum (0.1–0.2 s),the horizontal components at stations ISK005 and ISK007,both of which are alluvium sites, are overestimated by sev-eral fold, whereas the vertical components are reasonablywell matched. This indicates some degree of nonlinear soilbehavior at those stations.We have also estimated strong ground motions during
the mainshock at our temporary stations. Synthetic groundvelocities from the source model are shown in Fig. 9. Thewaveforms at every alluvium site show significant pulse-like waves followed by monotonic later phases, and thePGVs are two- or three-fold larger than those at nearby rocksites. The calculated PGVs for the alluvium sites are 70–110 cm/s in Monzen, 60–110 cm/s in Anamizu, and 40–50 cm/s in Wajima. Since the synthetic PGV at ISK003in Wajima is 70% of the observed PGV, we simply correctPGVs at the alluvium sites in Wajima to 60–70 cm/s.
5. ConclusionLinear array observations of aftershocks following the
Noto Hanto earthquake in three of the most affected low-
M. YOSHIMI AND K. YOSHIDA: SITE AMPLIFICATION AND GROUND MOTIONS IN DAMAGED AREAS, NOTO, JAPAN 165
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK001-EW
obs.
syn.
0.1 11Period[s]
ISK001-NS
obs.
syn.
0.1 11Period[s]
ISK001-UD
obs.
syn.
ISK001
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK002-EW
obs.
syn.
0.1 11Period[s]
ISK002-NS
obs.
syn.
0.1 11Period[s]
ISK002-UD
obs.
syn.
ISK002
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK003-EW
obs.
syn.
0.1 11Period[s]
ISK003-NS
obs.
syn.
0.1 11Period[s]
ISK003-UD
obs.
syn.
ISK003
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK004-EW
obs.
syn.
0.1 11Period[s]
ISK004-NS
obs.
syn.
0.1 11Period[s]
ISK004-UD
obs.
syn.
ISK004
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK005-EW
obs.
syn.
0.1 11Period[s]
ISK005-NS
obs.
syn.
0.1 11Period[s]
ISK005-UD
obs.
syn.
ISK005
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK006-EW
obs.
syn.
0.1 11Period[s]
ISK006-NS
obs.
syn.
0.1 11Period[s]
ISK006-UD
obs.
syn.
ISK006
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK007-EW
obs.
syn.
0.1 11Period[s]
ISK007-NS
obs.
syn.
0.1 11Period[s]
ISK007-UD
obs.
syn.
ISK007
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK008-EW
obs.
syn.
0.1 11Period[s]
ISK008-NS
obs.
syn.
0.1 11Period[s]
ISK008-UD
obs.
syn.
ISK008
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
ISK009-EW
obs.
syn.
0.1 11Period[s]
ISK009-NS
obs.
syn.
0.1 11Period[s]
ISK009-UD
obs.
syn.
ISK009
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
TYM002-EW
obs.
syn.
0.1 11Period[s]
TYM002-NS
obs.
syn.
0.1 11Period[s]
TYM002-UD
obs.
syn.
TYM002
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
TYM005-EW
obs.
syn.
0.1 11Period[s]
TYM005-NS
obs.
syn.
0.1 11Period[s]
TYM005-UD
obs.
syn.
TYM005
0.1
11
10
100
1000
pSv(
cm/s
)
0.1 11Period[s]
TYM007-EW
obs.
syn.
0.1 11Period[s]
TYM007-NS
obs.
syn.
0.1 11Period[s]
TYM007-UD
obs.
syn.
TYM007
Fig. 7. Comparison of the pseudo-velocity response spectra with a damping factor 0.05 of the synthetic (bold broken) and the observed (thin solid)ground motions at the K-NET stations shown in Fig. 5.
lands have revealed site amplification at the alluvium siteswhich resulted in severe damage. Site amplifications at al-luvium sites in Anamizu and Wajima have predominant fre-quencies of approximately 1 Hz and spectral ratios withrespect to rock sites that are five- to twenty-fold larger.These amplifications are reproduced well with a linear 1-D
model—except at ISK005 in Anamizu.A proposed source model comprising two asperities suc-
cessfully reproduces ground motions during the mainshockusing the empirical Green’s function method. The seis-mic moment released at the asperities—3.76 × 1018 N mat Asp1 and 2.21 × 1018 N m at Asp2—is about 44% that
166 M. YOSHIMI AND K. YOSHIDA: SITE AMPLIFICATION AND GROUND MOTIONS IN DAMAGED AREAS, NOTO, JAPAN
syn.18.24 cm
-40
0
40
Dis
p.(c
m)
0 10 20 30 40Time(s)
obs.21.5 cm
syn.8.18 cm
0 10 20 30 40
obs.6.67 cm
syn.3.1 cm
0 10 20 30 40
obs.6.48 cm
syn.107.76 cm/s
-200
0
200
Vel
.(cm
/s)
0 10 20 30 40
obs.98.22 cm/s
syn.49.28 cm/s
0 10 20 30 40
obs.34.55 cm/s
syn.21.48 cm/s
0 10 20 30 40
obs.18.45 cm/s
syn.2041.06 gal
-2000
0
2000
Acc
.(cm
/s/s
)
0 10 20 30 40
ISK005-EWobs.
780.26 gal
syn.1202.9 gal
0 10 20 30 40
ISK005-NSobs.
472.52 gal
syn.489.06 gal
0 10 20 30 40
ISK005-UDobs.
555.45 gal
syn.6.39 cm
-15
0
15
Dis
p.(c
m)
0 10 20 30 40Time(s)
obs.5.67 cm
syn.6.39 cm
0 10 20 30 40
obs.12.87 cm
syn.3.98 cm
0 10 20 30 40
obs.3.89 cm
syn.19.3 cm/s
-80
0
80
Vel
.(cm
/s)
0 10 20 30 40
obs.21.83 cm/s
syn.28.82 cm/s
0 10 20 30 40
obs.39.4 cm/s
syn.18.07 cm/s
0 10 20 30 40
obs.14.03 cm/s
syn.199.19 gal
-1000
0
1000A
cc.(
cm/s
/s)
0 10 20 30 40
ISK003-EWobs.
395.54 gal
syn.312.92 gal
0 10 20 30 40
ISK003-NSobs.
518.06 gal
syn.130.87 gal
0 10 20 30 40
ISK003-UDobs.
141.12 gal
syn.2.23 cm
-15
0
15
Dis
p.(c
m)
0 10 20 30 40Time(s)
obs.3.79 cm
syn.4.06 cm
0 10 20 30 40
obs.6.63 cm
syn.1.32 cm
0 10 20 30 40
obs.2.15 cm
syn.14.22 cm/s
-80
0
80
Vel
.(cm
/s)
0 10 20 30 40
obs.18.8 cm/s
syn.18.94 cm/s
0 10 20 30 40
obs.23.02 cm/s
syn.5.77 cm/s
0 10 20 30 40
obs.6.83 cm/s
syn.415.87 gal
-1000
0
1000
Acc
.(cm
/s/s
)
0 10 20 30 40
ISK004-EWobs.
588.47 gal
syn.333.33 gal
0 10 20 30 40
ISK004-NSobs.
621.78 gal
syn.200.14 gal
0 10 20 30 40
ISK004-UDobs.
146.46 gal
syn.5.88 cm
-20
0
20
Dis
p.(c
m)
0 10 20 30 40Time(s)
obs.12 cm
syn.2.8 cm
0 10 20 30 40
obs.20.21 cm
syn.6.34 cm
0 10 20 30 40
obs.8.85 cm
syn.103.55 cm/s
-80
0
80
Vel
.(cm
/s)
0 10 20 30 40
obs.50.29 cm/s
syn.38.71 cm/s
0 10 20 30 40
obs.37.86 cm/s
syn.32.99 cm/s
0 10 20 30 40
obs.21.02 cm/s
syn.3502.69 gal
-2000
0
2000A
cc.(
cm/s
/s)
0 10 20 30 40
ISK006-EWobs.
847.31 gal
syn.1383.47 gal
0 10 20 30 40
ISK006-NSobs.
716.03 gal
syn.1293.17 gal
0 10 20 30 40
ISK006-UDobs.
461.51 gal
syn.5.15 cm
-15
0
15
Dis
p.(c
m)
0 10 20 30 40Time(s)
obs.7.59 cm
syn.3.51 cm
0 10 20 30 40
obs.9.15 cm
syn.1.9 cm
0 10 20 30 40
obs.3.13 cm
syn.24.52 cm/s
-40
0
40
Vel
.(cm
/s)
0 10 20 30 40
obs.33.35 cm/s
syn.12.98 cm/s
0 10 20 30 40
obs.26.87 cm/s
syn.7.9 cm/s
0 10 20 30 40
obs.7.78 cm/s
syn.350.17 gal
-600
0
600
Acc
.(cm
/s/s
)
0 10 20 30 40
ISK007-EWobs.
182.38 gal
syn.338.66 gal
0 10 20 30 40
ISK007-NSobs.
202.15 gal
syn.148.01 gal
0 10 20 30 40
ISK007-UDobs.
167.14 gal
syn.7.82 cm
-15
0
15
Dis
p.(c
m)
10 20 30 40Time(s)
obs.13.68 cm
syn.5.92 cm
10 20 30 40
obs.8.12 cm
syn.2.45 cm
10 20 30 40
obs.2.75 cm
syn.23.13 cm/s
-80
0
80
Vel
.(cm
/s)
10 20 30 40
obs.23.45 cm/s
syn.16.18 cm/s
10 20 30 40
obs.15.98 cm/s
syn.10.07 cm/s
10 20 30 40
obs.8.83 cm/s
syn.328.84 gal
-1000
0
1000
Acc
.(cm
/s/s
)
10 20 30 40
ISK008-EWobs.
382.25 gal
syn.372.36 gal
10 20 30 40
ISK008-NSobs.
228.5 gal
syn.314.71 gal
10 20 30 40
ISK008-UDobs.
297.42 gal
Fig. 8. Comparison of the waveforms of the observed (top) and the synthetic (bottom) ground motions at ISK003, ISK004, ISK005, ISK006, ISK007,and ISK008. Numerics are maximum amplitudes. Time is from the origin time. Top: accelerations, Middle: velocities, Bottom: displacements foreach station.
M. YOSHIMI AND K. YOSHIDA: SITE AMPLIFICATION AND GROUND MOTIONS IN DAMAGED AREAS, NOTO, JAPAN 167
Synthetic velocities of the Noto Hanto earthquake 2007/03/25
-1000
100
[cm
/s]
EWMax=49
Max=39.8Max=88.4Max=23.9Max=59.6
Max=107.8Max=21.3Max=41.6Max=43.7Max=19.3
NSMHS
Max=68.7
MJDMax=87.2
MTGMax=108
AYKMax=23.3
APOMax=54.7
ISK005Max=49.3
WFGMax=20.4
WKWMax=38.6
WSBMax=45.7
ISK003Max=28.9
UD Max=29.9Max=12.5Max=25.1Max=14.1Max=17.4Max=21.5Max=12.3Max=15.4Max=16.8Max=18.1
0 5 10 15 20 25 30Time[s]
Monzen
Anamizu
Wajima
Monzen
Anamizu
Wajima
Monzen
Anamizu
Wajima
Fig. 9. Synthetic ground velocities at all the temporary array stationscomputed from our source model using the empirical Green’s functionmethod.
of the whole event. Peak ground velocities (PGV) at thealluvium sites of our array stations during the mainshockare estimated to have been approximately 70–110 cm/s forMonzen, 50–110 cm/s for Anamizu, and 60–70 cm/s forWajima.
Acknowledgments. We thank Dr. Hidetaka Saomoto and Dr.Ikuo Cho for assisting our aftershock observation. Fruitful dis-cussions with Dr. Haruko Sekiguchi at GSJ, AIST, are greatlyappreciated. The manuscript was improved by reviews from twoanonymous reviewers. This study has been partially supported by
a Grant-in-Aid for Special Purposes, “Urgent research on the after-shocks of the 2007 Noto Hanto earthquake” Special CoordinationFunds, from the Ministry of Education, Culture, Sports, Scienceand Technology (MEXT). A portion of the ground motion data setused in this study was recorded by K-NET and KiK-net operatedby the National Research Institute for Earth Science and DisasterPrevention. Some of the figures were drawn using Generic Map-ping Tool (Wessel and Smith, 1998).
ReferencesAwata, Y., S. Toda, T. Azuma, H. Kaneda, and H. Horikawa, Coastal
deformation associated with the 2007 Noto Hanto earthquake, centralJapan, determined by uplifted and subsided intertidal organisms, EarthPlanets Space, 2008 (in press).
Brune, J. N., Tectonic stress and the spectra of seismic shear waves fromearthquake, J. Geophys. Res., 75, 4997–5009, 1970.
Brune, J. N., Correction, J. Geophys. Res., 76, 5002, 1971.Geographical Survey Institute, Leveling survey in the focal area of
the Noto Hanto earthquake, 2007, http://www.gsi.go.jp/WNEW/PRESS-RELEASE/2007/0711.htm, 2007 (in Japanese).
Horikawa, H., Characterization of the 2007 Noto Hanto, Japan, earth-quake, Earth Planets Space, 2008 (in press).
Irikura, K., Prediction of strong acceleration motion using empiricalGreen’s function, Proc. 7th Japan Earthquake Symp., 151–156, 1986.
Ishikawa Prefecture, Subsurface geological map Anamizu, Togi, Tsurugiji,1991.
Ishikawa Prefecture, Subsurface geological map Wajima, 1993.Kamae, K. and K. Irikura, Source model of the 1995 Hyogo-ken Nanbu
earthquake and simulation of near-source ground motion, Bull. Seis.Soc. Am., 88, 400–412, 1998.
Kobayashi, K., H. Kyuke, T. Uetake, M. Mashimo, and H. Kobayashi, Es-timation of Q value for soil layers by simultaneous inversion of transferfunctions at many sites: Part 3, Annual Meeting, Arch. Inst. Jpn., B-2,253–254, 1999 (in Japanese).
Kyuke, H., K. Kobayashi, T. Uetake, M. Mashimo, and H. Kobayashi, Es-timation of Q value for soil layers by simultaneous inversion of transferfunctions at many sites: Part 4, Annual Meeting, Arch. Inst. Jpn., B-2,255–256, 1999 (in Japanese).
Madariaga, R., Dynamics of an expanding circular fault, Bull. Seis. Soc.Am., 66, 639–666, 1976.
Sato, Y., Subsurface structure of Wajima urban area, Hokuriku Jiban Joho(Report of Hokuriku Geothech. Association), 4, 1994 (in Japanese).
Wessel, P. and W. H. F. Smith, New, improved version of Generic MappingTools released, EOS, 79, 579, 1998.
M. Yoshimi (e-mail: [email protected]) and K. Yoshida