ultra-broadband seismic and tsunami wave observation of high …kubota/uploads/draft/kubota... ·...

manuscript submitted to Earth and Space Science

Ultra-broadband seismic and tsunami wave observation of high-sampling 1

ocean-bottom pressure gauge covering periods from seconds to hours 2

3

T. Kubota1, T. Saito1, N. Y. Chikasada1, and W. Suzuki1 4

1 National Research Institute for Earth Science and Disaster Resilience, Tsukuba, Ibaraki, Japan. 5

6

Corresponding author: Tatsuya Kubota ([email protected]) 7

8

Key points: 9

• High-sampling rate is important for pressure gauges to record seismic body waves, 10

Rayleigh waves and tsunamis and their dispersive feature 11

• Theoretical relation between pressure and vertical acceleration (p = ρwΗ0az) is valid for a 12

long time (~3h for the 2010 Chile earthquake) 13

• A relationship between pressure and vertical velocity (p = ρwc0vz) holds only at the first P 14

wave arrival, but not for later phases 15

16


Abstract 17

Recent developments of ocean-bottom pressure gauges (PG) have enabled us to observe 18

various waves including seismic and tsunami waves covering periods of T ~100–103 s. To 19

investigate the quality for broadband observation, this study examined the broadband PG records 20

(sampling rate of 1 Hz) around Japan associated with the 2010 Chile earthquake. We identified 21

three distinct wave trains, attributed to seismic body waves, Rayleigh waves and tsunamis. Clear 22

dispersive features in the Rayleigh waves and tsunamis were explained by theories of elastic 23

waves and gravity waves. Quantitative comparison between pressure change and nearby 24

seismograms demonstrated the validity of the theoretical relation between pressure p and vertical 25

acceleration az for ~3h from the origin time. We also found a relationship between p and vertical 26

velocity vz holds only at the first P wave arrival, but not for later arrivals. Similar results were 27

confirmed for various earthquakes with different source-station distances and magnitudes, 28

suggesting the robustness of these relations. The results demonstrate that the high-sampling rate 29

(≥ 1 Hz) is necessary to observe seismic-wave dispersion and PG can record both seismic waves 30

and tsunamis with reasonable quality for waveform analyses, whereas conventional onshore and 31

offshore seismometers or tide gauges can observe either of seismic waves and tsunamis. 32

Utilizing the high-sampling PG in combination with the seismic and tsunami propagation theory 33

for estimating earthquake source process or analyzing wave propagation processes in the ocean, 34

will deepen our geophysical understanding of the solid-fluid coupled system in the earth and 35

contribute towards disaster mitigation. 36

37

Plain Language Summary 38

Recent developments of offshore ocean-bottom observation networks have enabled us to 39

use high-sampling (1 or more samples per second) seafloor pressure gauge (PG) data. This study 40

investigated PG records with broadband period range (seconds to hours) around Japan during the 41

2010 Chile earthquake. We identified a seismic P wave train arriving ~20–30 min after the focal 42

time. Another seismic wave train due to the surface Rayleigh wave during ~70–110 min and 43

tsunamis during ~24–72 h were also confirmed, which showed a dispersive feature; long-period 44

waves arrive earlier than short-period waves. The dispersion theories obtained from the elastic 45

and fluid dynamics thoroughly explained these features. We also compared wave amplitudes 46

between the PG and nearby ocean-bottom seismometer, and confirmed the validity of the 47


relationship between pressure and vertical acceleration, and between pressure and vertical 48

velocity. This study demonstrates PG can record both seismic and tsunami signals clearly with 49

reasonable quality, while conventional seismometers or tide gauges can either. Ultra-broadband 50

observation of PG plays an important role in deepening our understanding of geophysical wave 51

propagation processes in the solid-fluid coupled system in the ocean, and enables delivery of 52

essential information for earthquake early warning and disaster mitigation. 53

54


1 Introduction 55

Ocean-bottom pressure gauges (PGs) have recently been developed worldwide for 56

tsunami observations (e.g., Titov et al. 2005; Mungov et al. 2013; Tsushima & Ohta, 2014; 57

Rabinovich & Eblé 2015). Compared with tsunami observation using the conventional tide 58

gauges, PGs are advantageous as they are less affected by complex coastal site effects (e.g., 59

Kubota et al., 2018). Owing to this advantage, PGs have significantly contributed to our 60

understanding of tsunami generation and propagation processes derived from solid and fluid 61

dynamics (e.g., Satake et al., 2005; Saito et al., 2010; Inazu & Saito, 2013; Maeda et al., 2013; 62

Rabinovich et al., 2013; Allgeyer & Cummins, 2014; Watada et al., 2014; Lay et al., 2016; 63

Poupardin et al., 2018; Sandanbata et al., 2018; Kubota et al., 2020; and references therein). PGs 64

have further been utilized for observing much longer time-scale phenomena, such as infragravity 65

waves (e.g., Tonegawa et al., 2018) and oceanographic and geodetic phenomena (e.g., Baba et 66

al., 2006; Inazu et al., 2012; Wallace et al., 2016; Fukao et al., 2019). 67

In addition recent studies have reported that the PG observations can record other 68

geophysical waves, including ocean-acoustic and seismic waves (e.g., Nosov & Kolesov, 2007; 69

Levin & Nosov 2009; Bolshakova et al., 2011; Matsumoto et al., 2012; Webb & Nooner, 2016). 70

Dynamic pressure changes associated with the coseismic seafloor vertical accelerations, 71

recognized as a reaction force to the seafloor lifting up the water column, can also be observed 72

(e.g., Filoux, 1982; Nosov & Kolesov, 2007; An et al., 2017; Saito, 2019; Ito et al., 2020). This 73

indicates the ability for PGs to be utilized as vertical accelerometers for the sea-bottom motions 74

(An et al., 2017; Kubota et al., 2017). One strong advantage for using PGs in seismic wave 75

observation is that the signal never saturates, unlike high-sensitivity ocean-bottom seismometers. 76

The recent development of seafloor pressure observations with a sampling rate exceeding 1 Hz 77

(hereafter, high-sampling rate) in comparison with the low sampling rate of the conventional PG 78

(e.g., the DART system,1/15 Hz, e.g., see Rabinobich & Eble, 2015) has contributed extensively 79

to our understanding of ocean-acoustic waves and seismic waves. Given the PG observations of 80

various geophysical waves, PG could be utilized for ultra-broadband geophysical observation, 81

including both seismic and tsunami waves, which will be useful to extract earthquake source 82

information and the ocean’s ultra-broadband wave propagation process. However, PG records 83

have primarily been utilized for analyzing tsunamis, and not extensively for seismic wave 84

signals. Recent studies have discussed PG quality and performance in tsunami observation, 85


which featured a period range of ~102–103 s (e.g., Saito et al., 2010; Tsushima & Ohta, 2014; 86

Rabinovich & Eblé 2015), whereas those in seismic waves have not been examined in detail. 87

In the present study, we examine the high-sampling PG record featuring a period range 88

of ~100–103 s, with particular focus on the seismic wave signals. Section 2 describes the PG and 89

additional datasets used and the method conducted for the analyses. Section 3 outlines the 90

results. Section 4 interprets the results by examining the origins of the identified PG signals 91

based on both the propagation theories of seismic waves and tsunamis and the comparison with 92

other nearby instruments. Section 5 compares the PG waveform with nearby ocean-bottom 93

seismometers (OBS) to discuss the quantitative relationship between pressure change and 94

seismic waves. Section 6 examines the quality of the PG observation by comparing the spectral 95

amplitudes and the background noises to discuss the performance and limitations of the high-96

sampling PG observation. Finally, Section 7 discusses future potential of the high-sampling PG 97

for furthering our geophysical understanding and for contribution to practical disaster mitigation. 98

The conclusion is summarized in Section 8. 99

100

2 Data and method 101

2.1 Data 102

We examined the records of onshore and offshore instruments in northern Japan 103

(Figure 1) during some moderate local to regional earthquakes and major regional to global 104

earthquakes. We here primarily focused on the 2010 Chile earthquake (Mw 8.8, Duptel et al., 105

2012) because of its large magnitude and ideal source-station distance (Figure 1, the other 106

examples are discussed in Section 5). The approximate distance along the great circle path from 107

Chile to Japan is 17,000 km (angular distance of 150°). Station information is listed in Table S1. 108

We use PGs at KPG1 and KPG2 (dark blue inverted triangles in Figure 1b) and three-component 109

acceleration records from a OBS at KOBS1 (red circle) of the Off-Kushiro cabled-observatory, 110

operated by the Japan Agency for Marine-Earth Science and Technology (JAMSTEC, 111

Kawaguchi et al., 2000; Hirata et al., 2002). A coastal tide gauge at Hanasaki Port operated by 112

the Japan Meteorological Agency (JMA, blue diamond), and a Streckensen STS-2 onshore 113

broadband seismometer at Kushiro (KSRF, green triangle) and a nearshore tiltmeter at Samani 114

(SAMH, orange square) operated by National Research Institute for Earth Science and Disaster 115

Resilience [NIED] (Okada et al., 2004; NIED, 2019) were also utilized. The detailed 116


specification of the instruments is described in Text S1. The velocity records from the onshore 117

seismometer are converted to accelerograms. The horizontal components of the original 118

seismograms and tilt records are along the north-south and east-west directions, which are 119

rotated to the radial (R) and transverse (T) directions along the great circle path (Figure 1). We 120

re-sampled all of these datasets with 1 Hz. Note that the instrument response at the onshore and 121

offshore seismometers were not removed. We also examine the PGs of the Deep-ocean 122

Assessment and Reporting of Tsunamis (DART) system deployed by the National Oceanic and 123

Atmospheric Administration (NOAA) in the Pacific Ocean (Titov et al., 2005; Mungov et al., 124

2013, black inverted triangles in Figure 1). Refer to Table S2 for information on the DART 125

stations. 126

127

128

Figure 1. Location map of this study. (a) The 2010 Chile earthquake (Duptel et al., 2012) 129

and great circle paths to the stations (yellow lines). DART tsunami stations are shown by 130

inverted triangles. (b) Enlarged view around northern Japan. Dark blue inverted triangles are 131

the PGs, the blue diamond is the coastal tide gauge, the green triangle is the onshore 132

broadband seismometer, the red circle is the OBS, and the orange rectangle is the nearshore 133

tiltmeter. 134

135

Figure 2a shows the time series after the Chile earthquake from the PGs (dark blue), 136

tide gauge (blue), onshore seismometer (green), OBS (red), and tiltmeter (orange). The 137

waveforms from DART records are shown in Figure S1. Tidal variations are observed by the 138


PGs and tide gauge. We also confirm small tidal fluctuations in the tiltmeter. Figure 2b shows 139

the highpass filtered records with a cutoff period of 10800 s (3 h, ~0.1 mHz). Seismic waves are 140

clearly observed, except with the tide gauge. Figure 2c depicts the bandpass filtered records 141

(passbands of 30–10800 s, ~0.1–33 mHz). Tsunamis are clearly detected by the PGs and the 142

coastal tide gauge. A pressure change of Δp = 1 hPa corresponds to a sea-surface height change 143

of Δη = 1 cm (Δp = ρwg0Δη, ρw ~ 1.03 g/cm3: seawater density, g0 = 9.8 m/s2: gravitational 144

acceleration), therefore the tsunami amplitudes in PGs and tide gauges are comparable. Tsunami-145

related tilt changes are also recorded with the coastal tiltmeter (e.g., Kimura et al., 2013; Nishida 146

et al. 2019). The amplitudes of seismic waves and tsunamis in the PGs are almost comparable, 147

whereas tsunamis recorded in the tiltmeter (orange trace in Figure 2c) have much smaller 148

amplitudes than the seismic waves, by magnitudes of 102. 149

150

Figure 2. Time series at onshore and offshore sensors associated with the 2010 Chile 151

earthquake. (a) Raw data. (b) Highpass filtered data (cutoff period of 10800 s). The 152

waveform around the first P wave arrival is enlarged in Figure 2b’. (c) Bandpass filtered 153

data (30–10800 s). Note that the scales are different in each panel and U, R, and T denote 154

the vertical, radial and transverse component, respectively. 155

156

2.2 Spectrogram analysis 157

We examine the features of the high-sampling PG data based on the spectrogram 158

analysis which calculates the Fourier transform with a moving time window. In this study, we 159

applied Aki and Richards (1980)’s definition of the Fourier transform, as: 160


161

𝑋(𝜔) = ∫ 𝑥(𝑡)𝑒!"#𝑑𝑡$!/&'$!/&

, (1) 162

163

where TL is the time window length of each bin and ω is the angular frequency (=2πf, f is 164

frequency). Here, the spectrogram 𝐴(𝑡, 𝜔) is defined as: 165

166

𝐴(𝑡, 𝜔) = -∫ 𝑥(𝑡 + 𝜏)𝑒!"(𝑑𝜏$!/&'$!/&

-. (2) 167

168

For the calculation of the spectrogram we adopt the Fourier amplitude spectra, but not power 169

spectra. In order to examine the spectral features of the seismic waves, we set the time window 170

length of each bin (TL) at 512 s and the time shift (ΔT) at 60 s (i.e., t = nΔT, n is an integer 171

number). We also investigate tsunamis based on the spectrogram with TL = 8192 s and ΔT = 172

1800 s. We further examined the spectral features with broadband period ranges including both 173

seismic waves and tsunamis, using TL = 1024 s with ΔT = 60 s. Prior to calculating the Fourier 174

transform of each bin, we removed the mean value and applied a Hanning taper (Blackman & 175

Turkey, 1958) with a length of 0.05TL to both edges of the window. 176

177

3 Results 178

The spectrograms for the PGs at KPG1 and KPG2 are shown in Figure 3. The 179

spectrograms during the first three hours from the origin time are illustrated in Figures 3a and 180

3b (TL = 512 s). The spectrogram confirms two distinct seismic wave trains during the seismic 181

wave arrival at ~20 min (periods of T ~3–20 s, ~50–300 mHz) and ~70 min (T ~10–200 s, ~5–182

100 mHz) (black and red arrows in Figures 3a and 3b, respectively). The second wave train at 183

~70 min displays a clear dispersive feature (i.e., long-wavelength waves arrive earlier than 184

short-wavelength waves). Figures 3c and 3d depict the spectrogram for 120 hours after the 185

origin time (TL = 8192 s). A tsunami-attributed wave train, showing a clear dispersion (T ~60–186

1000 s, ~1–20 mHz) can be confirmed in the spectrograms at ~24 h (blue arrow in Figures 3c 187

and 3d). The arrival timing and duration of seismic waves and tsunamis are significantly 188

different; hence, it is difficult to display both in one figure. In order to display both wave 189


signals, we show the spectrograms using the logarithmic scale for the elapsed time in Figures 3e 190

and 3f (TL = 1024 s). 191

192

193


Figure 3. Spectrograms for PGs at (a, c, e) KPG1 and (b, d, f) KPG2. (a, b) Spectrograms 194

for seismic waves. (c, d) Spectrograms for tsunamis. (e, f) Spectrograms for both seismic 195

and tsunami waves with a logarithmic scale on the horizontal axes. The highpass (Figures 196

3a, 3b, 3e, and 3f) and bandpass (Figures 3c and 3d) filtered waveforms are also shown. 197

Distinct wave train arrivals are denoted by the colored arrows (black: seismic P waves, red 198

Rayleigh waves, pink: Rayleigh waves from the opposite direction to the first Rayleigh 199

wave, blue: tsunamis). 200

201

We also calculate the spectrograms for other onshore and offshore instruments to 202

compare with the PG spectrograms (Figure 4, see Figures S2 and S3 for the spectrograms with 203

linear scale for the elapsed time). The seismic wave trains at ~20 min and ~70 min are detected 204

by the tiltmeter, onshore seismometer, and OBS, but not by the coastal tide gauge, which is 205

similar to the PG spectrogram. Another distinct wave train can also be confirmed at ~40–45 min 206

in the horizontal components of the seismometers (e.g., black arrow in Figure 4e). We also 207

identify an additional seismic wave train in the transverse component at the station KSRF at 208

~70 min (e.g., green arrow in Figure 4f), but this is not observed in the PG spectrogram. We 209

recognize small amplitude increases associated with tsunamis in the tiltmeter spectrogram at 210

~24 h (blue arrows in Figures 4a, 4b, and 4c), but the dispersive feature cannot be observed. 211

212


213

Figure 4. Spectrograms for (a) tide gauge at Hanasaki, (b, c) tiltmeter at SAMH, (d, e, f) 214

onshore seismometer at KSRF, and (g, h, i) ocean-bottom seismometer at KOBS1. The 215

horizontal axes are displayed on a logarithmic scale. The Love wave train arrival is denoted 216

by the green arrow. See the caption in Figure 3 for the other description. 217

218

4 Interpretation on results of spectrogram analysis 219

4.1 Seismic wave trains 220

We interpret the wave trains in the PG and the other instruments identified by the 221

spectrogram analysis based on the theory of seismic waves and tsunamis. The onset timing of the 222

first waves in the pressure records (~20 min) is identical to the onshore seismometer (Figure 2). 223

We calculate the theoretical travel time of the core phases such as the PKP phase, the P wave 224

penetrating into the outer core, at KSRF using the global earth structure model AK135 (Kennett 225

et al., 1995, Figure S4). We obtain a travel time of PKP phase of 1193.74 s (~19.8 min), which 226


accurately reproduced the observed arrival time (Figure S4). Thus, we conclude that this wave 227

train is the seismic body P-waves including PKP, and PKIKP, PKiKP, PKPab, and PKPbc 228

phases (e.g., Blom et al., 2015). Furthermore, the theoretical travel time of the SS phases, free-229

surface reflected S waves leaving a source downward, at KSRF, is 2570.55 s (~42.7 min, Figure 230

S4). This also corresponds with the onset of the waves confirmed in the horizontal seismograms 231

at ~40–45 min (Figure 4e), and also indicates the body wave. The direct S waves do not arrive at 232

the stations off Hokkaido because they do not propagate into the fluid outer core (e.g., Shearer, 233

2009). 234

The dispersion is detected in the second seismic wave train at ~70 min. To identify the 235

origin of this wave, we examine the particle motion of the seismometers at KSRF and KOBS1 236

(Figure 5). The pressure waveform at KPG1 is depicted in Figure 5 (red trace). The particle 237

motions in both OBS and onshore seismometer records demonstrate retrograde motion during the 238

arrival of this wave at 4200–4800 s (black arrow in Figure 5). This indicates a seismic surface 239

wave (Rayleigh wave). For the time window during 100–110 min (6000–6600 s) from the origin 240

time, we confirm another retrograde particle motion with the rotation direction opposite to the 241

first Rayleigh wave (gray arrow in Figure 5). A corresponding spectral amplitude increase is also 242

confirmed (pink arrow in Figure 3). These comparisons suggest another Rayleigh wave train 243

from the opposite direction to the first Rayleigh wave arrival, and delayed by ~30 min. It is also 244

worth pointing out that the Rayleigh wave arrivals at 5-10 mHz is slightly delayed compared to 245

those at 20-50 mHz (Figures 3a and 3b). This observation has an opposite sense to typical 246

understanding of surface dispersion that the longer-period waves arrive earlier (e.g., Shearer, 247

2009). A similar dispersion pattern also appears in the KOBS1 record, in particular the vertical 248

component (Figure 4g). 249


250

Figure 5. Particle motions and displacement waveforms of (a) onshore seismometer at KSRF 251

and (b) OBS at KOBS1. Particle motions during 3600–4200 s, 4200–4800 s, 4800–6000 s, 6000–252

6600 s, and 6600–7200 s from the origin time are shown at the top panels. White and black stars 253

indicate the start and end locations of the particle motion in each period, respectively. Black and 254

gray arrows denote the rotation direction of the particle motion. Displacement waveforms are 255

shown in the bottom. Vertical displacement expected at KPG1, obtained by the time-double-256

integration of pressure (based on p = ρH0az) is shown by pink trace. Pressure waveform at KPG1 257

(dark red trace) and pressure expected from the vertical acceleration and the dynamic pressure 258

relation (black) are also shown. 259

260

Examining the particle motion of the seismometers suggests that the wave train 261

arriving to the PG at ~70 min is the Rayleigh wave. We further examine the dispersive feature of 262

the Rayleigh wave train recorded by the PGs based on the elastic wave propagation theory. We 263

calculate the theoretical group velocity dispersion of the Rayleigh wave based on the AK135 264

global structure model (Kennett et al., 1995) incorporating a seawater layer with a thickness of 265

4.5 km. The velocity structure and the dispersion relationship is depicted in Figure S5. Figure 6 266


shows the comparison between the theoretical arrival time of the Rayleigh wave and the 267

spectrograms. The theoretical Rayleigh wave arrivals at each period are consistent with the 268

spectral amplitude peak in the spectrograms (red lines in Figure 6). In addition, the onset of the 269

delayed Rayleigh wave train is also explained (pink lines) when considering a path along the 270

opposite direction of the great circle path on the earth (~23,000 km). This suggests that the 271

delayed wave train is also due to the Rayleigh wave, but radiated to the opposite direction of the 272

first Rayleigh wave train. 273

274

275

Figure 6. Comparison of theoretical arrival times of the seismic and tsunami waves and the 276

spectrograms for (a) KPG1, (b) Tide gauge at Hanasaki, (c–e) onshore seismometer at KSRF, 277

and (f–h) OBS at KOBS1. Theoretical arrival times of P and SS waves (black), the surface waves 278

(red and pink: Rayleigh, green: Love waves), and tsunamis (blue). The pink line is the arrival of 279

the Rayleigh wave along the opposite direction of the great circle path on the earth. 280


281

We also calculate the theoretical dispersion relationship using different structure 282

models to compare with the observed seismograms. We reveal that a model excluding the 283

seawater layer from the AK135 structure model (Figure S6) cannot explain the short-period 284

Rayleigh waves (>~50 mHz, Figure S7). We also compare the theoretical arrival time based on 285

the PREM global structure model (Dziewonski & Anderson, 1981), which includes a seawater 286

layer (Figure S8). The observed dispersive features in the Rayleigh waves are reasonably 287

reproduced by the PREM model, although the theoretical arrival time is slightly delayed by ~10–288

20 min in periods at ~10–50 s (~10–100 mHz, Figures S9 and S10). This result demonstrates that 289

the earth structure model is important for reproducing the dispersive feature, which suggests 290

Rayleigh wave dispersion recorded by the PG record can be useful to constrain the subseafloor 291

structure, as well as the ocean bottom seismometers. 292

In the particle motion of the onshore seismometer, a particle motion along the 293

transverse direction is also confirmed during 3600–4200 s (Figure 5). This indicates the Love 294

wave is also observed in the horizontal components of the seismometers. In order to examine the 295

Love wave train in the PGs, we calculate the group velocity and theoretical arrival times of the 296

Love wave (green line in Figure 6). The spectral amplitude increase due to the Love wave cannot 297

be confirmed in the KPG1 spectrograms, while the corresponding signal is recognized in the 298

transverse component of the seismometer at KSRF (green arrow in Figure 4f). This result is 299

expected given the transverse motion of Love waves (e.g., Shearer, 2009) while the pressure 300

changes due to the seafloor motion are primarily caused by the seafloor vertical accelerations 301

(e.g., Saito, 2019). 302

303

4.2 Tsunami wave trains 304

The wave train arriving at ~24 h which was attributed to a tsunami, also demonstrated 305

the dispersion. Based on the gravity wave theory (e.g., Pedlosky, 2013; Saito, 2019), the tsunami 306

phase velocity c and group velocity U are given as: 307

308

𝑐 = &)*$= "

*, (3) 309

310


𝑈 = +&21 + &*,"

-./0(&*,")4, (4) 311

312

where k is wavenumber (=2π/λ; λ is wavelength), T is period, and ω is angular frequency. The 313

angular frequency ω follows the dispersion relation: 314

315

𝜔& = 𝑔3𝑘 tanh(𝑘𝐻3), (5) 316

317

where g0 is the gravitational acceleration and H0 is the water depth. We calculate the theoretical 318

arrival time of the tsunami energy using the group velocity at each period (blue line in Figure 319

S5b and S5c). The theoretical arrival time explains the spectral peak in the PG (Figure 6a). The 320

tide gauge spectrogram, meanwhile, did not show clear tsunami dispersion (Figure 6b). The main 321

cause for this is its low sampling rate (1/60 Hz). However, the complex coastal site effect 322

associated with local bathymetry might be another possible cause (e.g., Geist, 2018; Tanioka et 323

al., 2019). The dispersive feature was also not clearly recognized in the tiltmeter spectrogram 324

(Figures 4b and 4c). 325

We investigated the spectral feature of the PGs of the DART stations (Titov et al., 326

2005; Mungov et al. 2013, black inverted triangles in Figure 1). The information of the DART 327

stations is listed in Table S2, the waveforms are depicted in Figure S1 and the spectrogram of the 328

DART data is in Figure 7. The dispersive feature can also be confirmed, even in the spectrogram 329

of the DART station 32412, closest to the epicenter (~2,400 km from the source, Figure 7a), 330

although the temporal delay of the shorter period components is too large for other distant DART 331

stations (Figures 7b to 7e) and KPG stations (Figure 4). We can also recognize the dispersive 332

Rayleigh wave signals in these DART records. However, because the sampling rate of these 333

DART systems is low (≤ 1/15 Hz), it is difficult to recognize the dispersive feature from the 334

spectrograms. 335

336


337

Figure 7. Spectrograms for the DART records in the Pacific Ocean. The horizontal axes are 338

displayed on a logarithmic scale. The time window length of each bin (TL) of 1024 s and the 339

time shift (ΔT) of 60 s were used in this calculation. Red and blue lines are the theoretical 340

arrival times of the Rayleigh waves and tsunamis. See the caption in Figures 3 for the other 341

description. 342

343

In the KPG1 and KPG2 spectrograms (Figures 3c and 3d), after the tsunami arrival 344

at 24 h, the amplitudes at periods of 50–500 s (20–200 mHz) increase between 24–36 h. 345

This feature is also confirmed in the DART spectrograms (Figure 7). We also verify 346

amplitude variations above 20 mHz following tsunami arrival in the onshore seismometer at 347

KSRF (Figures S3f and S3g) and nearshore tiltmeter at SAMH (Figure S3k and S3l). No 348

local earthquakes or aftershocks occurred during the tsunami arrivals, so this may suggest 349

that these shorter-period tsunamis are generated by long-period ones with the effects of 350

reflections and scattering by local bathymetry. 351

In addition, the delayed shorter-period tsunamis (T < ~100 s) at > ~30–40 h are not 352

recognized in the DART spectrograms. The static pressure change at seafloor 𝑝 due to sea-353

surface height change (i.e., tsunami) can be expressed as follows (e.g., Saito, 2019): 354

355

𝑝 = 4","5"67-0(*,")

, (6) 356


357

where𝑘 is the wavenumber, 𝜌3 is the seawater density, 𝐻3 is the water depth, and 𝜂3 is the 358

sea-surface height change. Tsunami amplitude decay in the seafloor pressure gauge is 359

expressed by the factor 1/cosh(kH0). Figure S11a shows this decay factor as a function of 360

kH0 (also shown in Saito, 2019) and Figure S11b is the decay factor as a function of 361

frequency assuming a constant seawater depth, which is obtained from the tsunami 362

dispersion relation (equation (5)). At a depth of > 3,000 m, the ocean bottom pressure 363

change due to shorter-period tsunami has an amplitude less than ~0.5 times of the sea-364

surface tsunami, suggesting that the shorter period tsunamis do not reach the seafloor, and 365

thus are not recorded by the DART stations. 366

367

5 Quantitative relation between pressure and vertical seismogram 368

In section 4.1, we verify that the PG clearly observed the Rayleigh wave signal. The 369

current section compares the quantitative relationship between the pressure change and the 370

seafloor seismic vertical motion in detail. When the period is longer than the resonant period of 371

the acoustic fundamental mode, T0 = 4H0/c0 ~6 s (c0 ~ 1.5 km/s: the velocity of the ocean 372

acoustic waves, H0 ~2200 m: seawater depth in this region), the relation between the pressure 373

change p(t) and vertical acceleration az(t) can be expressed as (e.g., Filloux, 1982; Nosov & 374

Kolesov, 2007; An et al., 2017; Saito, 2019): 375

376

𝑝(𝑡) = 𝜌8𝐻3𝑎9(𝑡), (7) 377

378

where ρw is the seawater density. Hereafter, we refer to this relationship as dynamic relation. 379

Using the dynamic relation, we examine the quantitative relationship between the pressure 380

change at KPG1 and vertical acceleration at KOBS1. We focus on the period bands of 10–200 s 381

(5–100 mHz), in which Rayleigh waves are the most dominant signal in the records. The 382

amplitudes and phases of the pressure change (p) and pressure-converted acceleration (ρwH0az) 383

are very similar (Figure 8a). The vertical displacement expected at KPG1 obtained from equation 384

(7) is similar to the vertical displacement at KOBS1 as well (pink trace in Figure 5). We assess 385

the similarity of the pressure change and vertical acceleration based on the correlation 386

coefficients (CC), which is defined as: 387


388

𝐶𝐶(𝑡) =∫ ;(#<()=#(#<()>($!/&'$!/&

?∫ ;(#<()&>($!/&'$!/&

∫ =#(#<()&>($!/&'$!/&

, (8) 389

390

where TL is the length of the time window. We use the two time window lengths for the CC 391

calculation; TL = 60 s (gray line in the middle panel in Figure 8a) and TL = 300 s (black line). We 392

obtain CC ~1 during the dominant Rayleigh wave arrival (~70 min) and also other seismic waves 393

(~1–3 h). Recent studies have reported the validity of equation (6) in the time domain in the 394

coseismic dataset of a few minutes from the origin time (An et al., 2017; Kubota et al., 2017). 395

Since the signal-to-noise (SN) ratios in the KPGs are so high, they had good agreement with 396

seismograms for a much longer duration than the results from previous studies. We further 397

compare the pressure change at KPG1 and the vertical acceleration at KOBS1 for other local to 398

regional moderate earthquakes and regional or global major earthquakes; the 2006 Kuril 399

earthquake (Mw 8.3, epicentral distance ~890 km, Figure 8b), the 2007 Kuril earthquake 400

(Mw8.1, 950 km, Figure 8c), the largest foreshock of the 2011 Tohoku earthquake (Mw 7.3, 401

~390 km, Figure 8d), the 2011 Tohoku earthquake (Mw 9.0, 420 km, Figure 8e), the 2012 402

Sumatra earthquake (Mw 8.6, ~6700 km, Figure 8f), and the 2018 Hokkaido Iburi Eastern 403

earthquake (Mw 6.6, 230 km, Figure 8g). We obtain similar results regardless of their very 404

different source-station distances and magnitude, which suggests the robustness of the observed 405

relation. 406

407


408

Figure 8. Comparison of the pressure and vertical acceleration for (a) the 2010 Chile earthquake, 409

(b) the Kuril earthquake, (c) the 2007 Kuril earthquake, (d) the 2011 Off-Miyagi earthquake (e) 410

the 2011 Tohoku earthquake, (f) the 2012 Sumatra earthquake, and (g) the 2018 Hokkaido 411

earthquake. Magnitudes and epicentral distances are shown in each panel. Blue and red traces are 412

pressure change at KPG1 and vertical acceleration at KOBS1 converted to apparent pressure 413

change using the relation p=ρwH0az, respectively. Bandpass filter with passband of 10–200 s is 414

applied. Correlation coefficient between pressure and acceleration is also shown (gray: TL = 60 s, 415

black: TL = 300 s). Coherency and phase difference between two waveforms are shown in the 416

bottom two figures. In the phase difference, the region where the coherency is less than 0.9 is 417

masked. 418

419

We calculate the temporal variation of the coherency and phase difference between 420

two waveforms in the frequency domain. The procedure to calculate the coherency and phase 421

difference is summarized in Text S2. Our results confirm the high coherency (> ~0.9) and zero-422

phase-difference at periods of ~10–50 s (~20–100 mHz, Figures 8d and 8e) during the Rayleigh 423

wave arrival. We further assess the correlation between the pressure and vertical acceleration 424


with different dominant periods (Figure 9). The correlation between the pressure and 425

acceleration is low in the shortest period bands (T < 10 s). This is due to the dynamic relation 426

being derived under the assumption that the wave period is longer than the resonant period of the 427

acoustic fundamental mode T0 = 4H0/c0 ~6 s. We also find that the correlation between the 428

pressure and acceleration is very high in periods of T ~10–50 s (Figures 9c and 9d), which infers 429

that the dynamic relation holds well at the period bands of ~10–50 s. Conversely, at longer 430

periods (T > 50 s) the correlation is not as high and unstable (Figures 9e to 9h). The low 431

correlation at the longer period is caused by the noises attributed to the seafloor vertical 432

displacement due to the long-period ambient oceanographic water waves. This is referred to as 433

the compliance noise (e.g., Crawford et al., 1991; 1998; Webb, 1998; Webb & Crawford, 1999; 434

An et al., 2020). However, we detected relatively high correlation during the Rayleigh wave 435

arrival (~70–100 min) even at the longer periods up to T ~400 s (~2.5 mHz, bottom panels in 436

Figures 9e to 9g). This indicates that the PGs can observe the Rayleigh wave signals up to the 437

periods of ~ 400s where the Rayleigh wave signals are larger than these compliance noises. 438

439


440

Figure 9. Comparison of the pressure and vertical acceleration at period bands of (a) 4–6 s, (b) 441

6–10s, (c) 10–20 s, (d) 20–50 s, (e) 50–100 s, (f) 100–200 s, (g) 200–400 s, and (h) 400–800 s. 442

See Figure 8 for the other description. 443

444

In contrast to the dynamic pressure relation (equation (7)), a relationship between the 445

pressure change and vertical velocity vz(t) is proposed in several previous studies (e.g., Nosov et 446

al., 2007; Bolshakova et al., 2011; Matsumoto et al., 2012): 447

448

𝑝(𝑡) = 𝜌@𝑐3𝑣9(𝑡). (9) 449

450


We also compare the pressure change at KPG1 and the vertical velocity at KOBS1, with periods 451

shorter than the acoustic resonant period T0 ~ 6 s. As a result, we can recognize high correlation 452

(~1) during the theoretical P wave arrival in this period band (Figure 10a). The comparisons 453

between pressure change at KPG1 and the vertical velocity at KOBS1 for other moderate to 454

major earthquakes confirm that this tendency is robust (Table S3, Figures 10b–10g). 455

If we carefully review the P-wave arrivals, we find that the amplitude of the pressure 456

change p(t) does not agree with ρwc0vz(t), although the phases agree with each other. The 457

amplitude ratio between p and ρwc0vz in the first arrival of the seismic waves differs for each 458

event (see Figure 10). This tendency is most evident in the 2011 Tohoku-Oki earthquake (Figure 459

10e). The local site effect due to the station location difference might be the possible cause. 460

However this seems implausible because the station difference between KPG1 and KOBS is only 461

~ 4 km, and any systematic relation in the amplitude ratio among these events cannot be 462

identified. 463

We also point out that the high correlation is confirmed only in the P-wave arrivals, but 464

neither in the entire wave trace nor in the later arrivals such as S waves (Figures 10b–10g).The 465

pressure p tends to be larger than ρwc0vz around the P arrivals. We should note that a relation 466

p=ρwc0vz can be derived assuming a P-wave propagation through fluid medium (e.g., Saito, 467

2019). However, the wavefields are more complicated than a P-wave propagation through fluid 468

medium, in particular, in the later wave arrivals where the waves are trapped by the reflections 469

between the sea surface and the seafloor. The previous results of numerical simulations suggest 470

this feature; the pressure changes at the seafloor due to trapped P-waves, or ocean-acoustic 471

waves exceed ρwc0vz (Bolshakova et al., 2011; Saito, 2017). Although Matsumoto et al. (2012) 472

compared the Fourier amplitudes of the seismograms and pressure records and found the 473

consistency with the relation p=ρwc0vz in the frequency domain, they discussed only the Fourier 474

amplitudes and not the phases. Our results show that the validity of the theoretical relationship 475

between the pressure and velocity is very limited, and holds only at the first arrival of the seismic 476

waves, not at the latter arrivals. 477

Notably, the time window at ~6300 s demonstrates relatively high correlation (Figure 478

10a, Figure S12a for more detail), which might be due to the P wave associated with the Mw 7.4 479

aftershock (GCMT) which occurred ~1.4 h after the Chile earthquake. We also examined the 480


correlation between the pressure and vertical velocity with several different period bands (Figure 481

S12), although it seems the correlation is low in longer period ranges (T > T0 ~ 6 s). 482

483

484

Figure 10. Comparisons of the pressure and vertical velocity for the multiple events. Blue and 485

red traces are pressure change at KPG1 and vertical velocity at KOBS1 converted to apparent 486

pressure change using the relation p=ρwc0vz, respectively. Correlation coefficient between 487

pressure and vertical velocity is also shown (gray: TL = 60 s, black: TL = 20 s). Bandpass filter 488

with passband of 4–6 s is applied. Comparisons of the pressure change and vertical velocity at 489

the time window around the P wave arrival are also shown at the bottom. Theoretical arrival 490

times of the direct P and S phase are also shown. 491

492

6 Quantitative comparison in amplitudes of tsunami and seismic waves 493

We finally investigate the quantitative relationship between the spectral amplitude and 494

the background noise, using the 10-Hz sampled original data. The time window of 58.25 h (=0.1 495

× 222 s) is used for the Fourier transform. Figure 11 shows the spectral amplitudes of the onshore 496


and offshore records for background (gray) and coseismic (black) signals. In order to 497

quantitatively compare the amplitudes between each record, the units are converted to apparent 498

pressure change ([Pa·s] = [(m-1· kg1·s-2)·s]). We assume that a sea height change of 1 cm equals 499

a pressure change of 1 hPa (102 Pa) to convert to apparent pressure change for the tide gauge at 500

Hanasaki. The vertical accelerations in the OBS at KOBS1 and onshore seismometer at KSRF 501

are converted to the apparent pressure change by using equation (6), with ρw = 1.03 g/cm3 and H0 502

= 2218 m (water depth at KOBS1). 503

PG spectra (Figure 11a) demonstrates an amplitude increase for the period range of 504

~60–5000 s, which are associated with tsunamis. Conversely, in the tide gauge spectra, the 505

significant increase in the tsunami amplitude is limited within the period range of ~2000–4000 s 506

(~0.25–0.5 mHz, Figure 11b), which is due to the coastal site effect (e.g., Geist, 2018; Tanioka et 507

al., 2019). The broadband amplitude increase in the PG spectra demonstrates that PGs are 508

capable of detecting the wider period range of tsunamis and are much less affected by the coastal 509

site effect than coastal tide gauges. 510

Our results further confirm that amplitude increases are related to the seismic waves in 511

the spectra of the PG and seismometers (Figures 11a, 11c, and 11d). Small spectral amplitude 512

increases at periods of ~5 s correspond to the body waves (marked by green arrow). In the 513

periods of ~10–50 s (~20–100 mHz), where the Rayleigh wave is dominant, we verify the signals 514

amplitude increases in the PG (orange text). This Rayleigh wave-related amplitude increase can 515

also be observed in the spectra of the OBS and onshore seismometer. The background noise level 516

in the frequency band of Rayleigh waves in the KPG1 is small compared to the KOBS1. This 517

suggests that the KPG1 can detect seismic Rayleigh wave signals with qualities similar to or 518

better than the KOBS1 during this event. One challenge in the seafloor seismic observation is 519

that the installation environment (i.e., we cannot fix the sensor to the ground) cannot be 520

controlled. This can easily be achieved in the onshore seismic observation. The lower noise level 521

in the PG in this frequency band may suggest that the PG will provide good supplementary 522

information for seafloor seismic observation. 523

524


525

Figure 11. Fourier amplitude for (a) KPG1, (b) HANA, and vertical components for (c) 526

KOBS1 and (d) KSRF. The gray and black lines depict the power spectra with a time 527

window of ~2.5 days before and after the 2010 Chile earthquake, respectively. 528

Characteristic spectral bands of tsunami and seismic waves are shown by blue and red 529

arrows, respectively. Note that the instrumental responses are not removed. 530

531

7 Future applicability of the high-sampling PG 532

We summarize the key results of the spectral analyses in Figure 12. Five key findings 533

can be drawn from our analyses: 534

(1) High-sampling-rate (≥ 1 Hz) is necessary for PGs to detect the dispersive Rayleigh waves in 535

broadband periods (~10–400 s, ~2.5–100 mHz) and dispersive tsunamis (~50–1000 s, ~1–20 536

mHz) very clearly. 537

(2) Sampling rate of the typical DART system (~1/15 Hz) is not sufficient to fully analyze the 538

wave propagation processes of the seismic waves and tsunamis. 539

(3) Seismometers (≥ 100 Hz sampling) can record seismic wave signals clearly but do not record 540


tsunamis, even when installed at the ocean floor. 541

(4) Coastal tiltmeters observe both seismic wave and tsunami signals but do not clearly record 542

the dispersion. 543

(5) Tide gauges cannot clearly detect tsunami dispersion or seismic waves. 544

545

546

Figure 12. Schematic illustration of geophysical phenomena in the ocean. The geophysical 547

phenomena which can be observed by the high-sampling-rate ocean bottom pressure gauges 548

and other instruments are shown by colored arrows. Note that the application of PG to 549

geodetic phenomena is shown by the dashed arrow, because this time-scale range is not 550

covered in this study. 551

552

In particular, finding (1) cannot be explicitly indicated if the sampling rate of the PG is 553

low (e.g., the third-generation DART system, 1/15 Hz, e.g., see Rabinobich & Eble, 2015). Our 554

result suggests that the high-sampling PG observation would be a good candidate for the 555

alternative and backup tools of seismic wave observations. This corresponds with An et al. 556

(2017) and Kubota et al. (2017) who also demonstrated that the seismic waves retrieved from the 557

PG records greatly contribute to accurately determining the centroid moment tensor solution of 558

moderate-sized offshore earthquakes. New, wide, and dense offshore observation networks 559

incorporating OBSs and PGs were recently installed in the deep-sea region (e.g., Kaneda et al., 560

2015; Kawaguchi et al., 2015; Rabinovich & Eblé 2015; Kanazawa et al., 2016; Mochizuki et al., 561

2016; Uehira et al., 2016; Howe et al., 2019). These new PG networks feature a higher sampling 562


rate (≥ 1 Hz). Furthermore, the sampling rate of the next generation DART system (DART4G) 563

will possibly be higher than the third-generation system (e.g., Rabinobich & Eblé, 2015; An et 564

al., 2017; Angove et al., 2019). Utilizing the array of high-sampling PGs will enable us to 565

analyze the ultra-broadband geophysical wave propagation, including tsunamis and seismic 566

waves. Developments of the offshore PG networks will facilitate the analysis of ultra-broadband 567

pressure signals covering periods of 100–103s. PGs can further be employed for observing much 568

longer time-scale phenomena (e.g., Baba et al., 2006; Inazu et al., 2012; Wallace et al., 2016; 569

Tonegawa et al., 2018; Fukao et al., 2019, Figure 12). Using these new PG networks will create 570

new possibilities for understanding the geophysical wave propagation processes in the solid-fluid 571

coupled system and the generation processes of earthquakes and tsunamis. Several studies have 572

already begun to use an array of the high-sampling PGs for the broadband geophysical wave 573

propagation analyses (Fukao et al., 2018; 2019; Sandanbata et al., 2018; Mizutani & Yomogida, 574

2019). 575

Real-time earthquake early warnings are especially important for developing nations 576

exposed to a high tsunami risk (e.g., Mulia et al., 2019). For more reliable and early earthquake 577

warnings, it is desirable to install OBSs in these networks, despite the higher economic costs 578

associated with construction. Nevertheless, this present study demonstrates that seismic wave 579

signals can still be observed by high-sampling-rate PGs, even if they do not incorporate OBSs. 580

Recently, Nakamura et al. (2019) estimated earthquake body-wave magnitudes based on high-581

frequency seismic wave signals in PGs for real-time data analyses. The present study 582

demonstrates the significant potential of longer and wider period seismic signals in near-field 583

high-sampling PG records for real-time estimation of the moment magnitude and centroid 584

moment tensor (e.g., Kubota et al., 2017). The broadband observations of high-sampling PG are 585

significantly valuable for real-time earthquake warnings and disaster mitigation. 586

587

8 Conclusions 588

We examined the performance of PGs in a wide frequency range by investigating the 589

spectral features in the PG records in northern Japan associated with the 2010 Chile earthquake, 590

and comparing these with nearby onshore and offshore instruments. We calculated the 591

spectrograms and revealed that the PGs clearly detected the wave trains due to body waves, 592

Rayleigh waves, and tsunamis. The dispersive features were visibly recognized in the Rayleigh 593


wave (periods covering ~10–400 s) and tsunami (~60–5000 s) wave trains, which were explained 594

by the propagation theories of seismic waves and tsunamis, respectively. The quantitative 595

comparison between the pressure change and the vertical acceleration demonstrated that the 596

dynamic relation holds for ~ 3h from the origin time, whereas, the relationship between the 597

pressure change and the vertical velocity at higher period range (< ~6 s) holds only at the first P 598

wave arrival. This validity was confirmed in time domain from the real observation. Similar 599

results seen for multiple earthquakes, regardless of their very different source-station distance 600

and earthquake magnitude, suggest the robustness of the observed relation. This study 601

demonstrated that high-sampling PGs can observe the broadband seismic wave and tsunami 602

signals. In particular, seismic wave signals can be detected with similar quality as the OBS. The 603

broadband geophysical wave observations of the high-sampling PG are important for furthering 604

our understanding of the geophysical analyses and developing practical disaster mitigations. 605

606

Acknowledgments, Samples, and Data 607

We thank the Editor Benoît Pirenne and two anonymous reviewers for their comments. We thank 608

Katsuhiko Shiomi for the information of the coastal tilt observation of NIED. This work was 609

financially supported by Sasakawa Scientific Research Grant (Grant Number 2019-2037) from 610

the Japan Science Society and JSPS KAKENHI Grant Number JP19K14818 from the Japan 611

Society for the Promotion of Science. Coastal tide gauge data is available at the 612

Intergovernmental Oceanographic Commission (IOC)’s website (http://www.ioc-613

sealevelmonitoring.org). PG and OBS data are acquired by JAMSTEC 614

(http://www.jamstec.go.jp/scdc/top_e.html). DART tsunami data is available at the NOAA’s 615

website (https://www.ngdc.noaa.gov/hazard/dart/2010chile_dart.html). Onshore broadband 616

seismometer is available at the NIED F-net (NIED, 2019, https://doi.org/10.17598/NIED.0005). 617

Tiltmeter record used in this study, which is operated by NIED, is available in 618

https://doi.org/10.17598/NIED.0018. We used TauP Toolkit (Crotwell et al., 1999) version 2.4.5 619

(https://www.seis.sc.edu/taup/) to calculate the theoretical arrival times. Figures in this paper was 620

prepared using Generic Mapping Tools (GMT) software version 6.0.0 (Wessel et al., 2019). We 621

used Editage (www.editage.com) for English language editing. 622


References 623

Aki, K. & Richards, P. G. (1980). Quantitative seismology. Mill Valley, CA: University Science 624

Books. 625

Allgeyer, S., & Cummins, P. (2014). Numerical tsunami simulation including elastic loading and 626

seawater density stratification. Geophysical Research Letters, 41, 2368–2375. 627

https://doi.org/10.1002/2014GL059348 628

An, C., Cai, C., Zheng, Y., Meng, L., & Liu, P. (2017). Theoretical solution and applications of 629

ocean bottom pressure induced by seismic seafloor motion. Geophysical Research Letters, 630

44, 10272–10281. https://doi.org/10.1002/2017GL075137 631

An, C., Shawn Wei, S., Cai, C., & Yue, H. (2020). Frequency Limit for the Pressure Compliance 632

Correction of Ocean-Bottom Seismic Data. Seismological Research Letters, 91(2A), 967–633

976. https://doi.org/10.1785/0220190259 634

Angove, M., Arcas, D., Bailey, R., Carrasco, P., Coetzee, D., Fry, B., … Schindelé, F. (2019). 635

Ocean observations required to minimize uncertainty in global tsunami forecasts, 636

warnings, and emergency response. Frontiers in Marine Science, 6, 350. 637

https://doi.org/10.3389/fmars.2019.00350 638

Baba, T., Hirata, K., Hori, T., & Sakaguchi, H. (2006). Offshore geodetic data conducive to the 639

estimation of the afterslip distribution following the 2003 Tokachi-oki earthquake. Earth 640

and Planetary Science Letters, 241, 281–292. https://doi.org/10.1016/j.epsl.2005.10.019 641

Blackman, R. B., & Turkey, J. W. (1958). The measurement of power spectra from the point of 642

view of communications engineering. New York: Dover Publications. 643

Blom, N. A., Deuss, A., Paulssen, H., & Waszek, L. (2015). Inner core structure behind the PKP 644

core phase triplication. Geophysical Journal International, 201, 1657–1665. 645

https://doi.org/10.1093/gji/ggv103 646

Bolshakova, A., Inoue, S., Kolesov, S., Matsumoto, H., Nosov, M., & Ohmachi, T. (2011). 647

Hydroacoustic effects in the 2003 Tokachi-oki tsunami source. Russian Journal of Earth 648

Sciences, 12, ES2005. https://doi.org/10.2205/2011ES000509. 649

Crawford, W. C., Webb, P. C., & Hildebrand, J. A. (1991). Seafloor compliance observed by 650

long-period pressure and displacement measurements. Journal of Geophysical Research, 651

96(B10), 16151–16160. https://doi.org/10.1029/91JB01577 652


Crawford, W. C., Webb, S. C., & Hildebrand, J. A. (1998). Estimating shear velocities in the 653

oceanic crust from compliance measurements by two-dimensional finite difference 654

modeling. Journal of Geophysical Research, 103(B5), 9895–9916. 655

https://doi.org/10.1029/97JB03532 656

Crotwell, H. P., Owens, T. J., & Ritsema, J. (1999). The TauP Toolkit: flexible seismic travel-657

time and ray-path utilities. Seismological Research Letters, 70, 154–160. 658

https://doi.org/10.1785/gssrl.70.2.154 659

Duputel, Z., Rivera, L., Kanamori, H., & Hayes, G. (2012). W phase source inversion for 660

moderate to large earthquakes (1990-2010). Geophysical Journal International, 189, 661

1125–1147. https://doi.org/10.1111/j.1365-246X.2012.05419.x 662

Dziewonski, A. M., & Anderson, D. L. (1981). Preliminary reference Earth model. Physics of the 663

Earth and Planetary Interiors, 25, 297–356. https://doi.org/10.1016/0031-9201(81)90046-664

7 665

Filloux, J. H. (1982). Tsunami recorded on the open ocean floor. Geophysical Research Letters, 666

9, 25–28. https://doi.org/10.1029/GL009I001P00025 667

Fukao, Y., Sandanbata, O., Sugioka, H., Ito, A., Shiobara, H., Watada, S., & Satake, K. (2018). 668

Mechanism of the 2015 volcanic tsunami earthquake near Torishima, Japan. Science 669

Advances, 4, eaao0219. https://doi.org/10.1126/sciadv.aao0219 670

Fukao, Y., Miyama, T., Tono, Y., Sugioka, H., Ito, A., Shiobara, H., … Miyazawa, Y. (2019). 671

Detection of Ocean Internal Tide Source Oscillations on the Slope of Aogashima Island, 672

Japan. Journal of Geophysical Research: Oceans, 124, 4918–4933. 673

https://doi.org/10.1029/2019jc014997 674

Geist, E. L. (2018). Effect of Dynamical Phase on the Resonant Interaction Among Tsunami 675

Edge Wave Modes. Pure and Applied Geophysics, 175, 1341–1354. 676

https://doi.org/10.1007/s00024-018-1796-y 677

Hirata, K., Aoyagi, M., Mikada, H., Kawaguchi, K., Kaiho, Y., Iwase, R., … Fujiwara, N. 678

(2002). Real-time geophysical measurements on the deep seafloor using submarine cable 679

in the Southern Kurile subduction zone. IEEE Journal of Oceanic Engineering, 27(2), 680

170–181. https://doi.org/10.1109/JOE.2002.1002471 681


Howe, B. M., Arbic, B. K., Aucan, J., Barnes, C. R., Bayliff, N., Becker, N., … Weinstein, S. 682

(2019). SMART Cables for Observing the Global Ocean: Science and Implementation. 683

Frontiers in Marine Science, 6, 424. https://doi.org/10.3389/fmars.2019.00424 684

Inazu, D., Hino, R., & Fujimoto, H. (2012). A global barotropic ocean model driven by synoptic 685

atmospheric disturbances for detecting seafloor vertical displacements from in situ ocean 686

bottom pressure measurements. Marine Geophysical Research, 33, 127–148. 687

https://doi.org/10.1007/s11001-012-9151-7 688

Inazu, D., & Saito, T. (2013). Simulation of distant tsunami propagation with a radial loading 689

deformation effect. Earth, Planets and Space, 65, 835–842. 690

https://doi.org/10.5047/eps.2013.03.010 691

Ito, Y., Webb, S. C., Kaneko, Y., Wallace, L. M., & Hino, R. (2020). Sea surface gravity waves 692

excited by dynamic ground motions from large regional earthquakes. Seismological 693

Research Letters. https://doi.org/10.1785/0220190267 694

Kanazawa, T., Uehira, K., Mochizuki, M., Shinbo, T., Fujimoto, H., Noguchi, S., … Okada, Y. 695

(2016). S-net project, cabled observation network for earthquake and tsunamis.Paper 696

WE2B-3 presented at Suboptic2016, Dubai, 18–21 April 2018. 697

Kaneda, Y., Kawaguchi, K., Araki, E., Matsumoto, H., Nakamura, T., Kamiya, S., … Takahashi, 698

N. (2015). Development and application of an advanced ocean floor network system for 699

megathrust earthquakes and tsunamis. In P. Favali, L. Beranzoli, & A. De Santis (Eds.), 700

Seafloor Observatories: A New Vision of the Earth from the Abyss (pp. 643–662). 701

https://doi.org/10.1007/978-3-642-11374-1_25 702

Kawaguchi, K., Hirata, K., Mikada, H., Kaiho, Y., & Iwase, Y. (2000). An expandable deep 703

seafloor monitoring system for earthquake and tsunami observation network. Proceedings 704

of OCEANS 2000 MTS/IEEE Conference and Exhibition (pp. 1719–1722). IEEE. 705

https://doi.org/10.1109/OCEANS.2000.882188 706

Kawaguchi, K., Kaneko, S., Nishida, T., & Komine, T. (2015). Construction of the DONET real-707

time seafloor observatory for earthquakes and tsunami monitoring. In P. Favali, L. 708

Beranzoli, & A. De Santis (Eds.), Seafloor Observatories: A New Vision of the Earth from 709

the Abyss (pp. 211–228). https://doi.org/10.1007/978-3-642-11374-1_10 710


Kennett, B. L. N., Engdahl, E. R., & Buland, R. (1995). Constraints on seismic velocities in the 711

Earth from traveltimes. Geophysical Journal International, 122, 108–124. 712

https://doi.org/10.1111/j.1365-246X.1995.tb03540.x 713

Kimura, T., Tanaka, S., & Saito, T. (2013). Ground tilt changes in Japan caused by the 2010 714

Maule, Chile, earthquake tsunami. Journal of Geophysical Research: Solid Earth, 118, 715

406–415. https://doi.org/10.1029/2012JB009657 716

Kubota, T., Saito, T., Suzuki, W., & Hino, R. (2017). Estimation of seismic centroid moment 717

tensor using ocean bottom pressure gauges as seismometers. Geophysical Research 718

Letters, 44(21), 10,907-10,915. https://doi.org/10.1002/2017GL075386 719

Kubota, T., Saito, T., Ito, Y., Kaneko, Y., Wallace, L. M., Suzuki, S., … Henrys, S. (2018). 720

Using tsunami waves reflected at the coast to improve offshore earthquake source 721

parameters: application to the 2016 Mw 7.1 Te Araroa earthquake, New Zealand. Journal 722

of Geophysical Research: Solid Earth, 123, 8767–8779. 723

https://doi.org/10.1029/2018JB015832 724

Kubota, T., Saito, T., Suzuki, W. (2020) Millimeter-scale tsunami detected by a wide and dense 725

observation array in the deep ocean: fault modeling of an Mw 6.0 interplate earthquake off 726

Sanriku, NE Japan. Geophysical Research Letters, 47, e2019GL085842. 727

https://doi.org/10.1029/2019GL085842 728

Lay, T., Li, L., & Cheung, K. F. (2016). Modeling tsunami observations to evaluate a proposed 729

late tsunami earthquake stage for the 16 September 2015 Illapel, Chile, Mw 8.3 730

earthquake. Geophysical Research Letters, 43, 7902–7912. 731

https://doi.org/10.1002/2016GL070002 732

Levin, B. W., & Nosov, M. A. (2009). Physics of tsunamis. Netherlands: Springer Netherlands. 733

https://doi.org/10.1007/978-1-4020-8856-8 734

Maeda, T., Furumura, T., Noguchi, S., Takemura, S., Sakai, S., Shinohara, M., … Lee, S.-J. 735

(2013). Seismic- and Tsunami-wave propagation of the 2011 Off the Pacific Coast of 736

Tohoku earthquake as inferred from the Tsunami-coupled finite-difference simulation. 737

Bulletin of the Seismological Society of America, 103(2B), 1456–1472. 738

https://doi.org/10.1785/0120120118 739


Matsumoto, H., Inoue, S., & Ohmachi, T. (2012). Dynamic response of bottom water pressure 740

due to the 2011 Tohoku earthquake. Journal of Disaster Research, 7, 468–475. 741

https://doi.org/10.20965/jdr.2012.p0468 742

Mizutani, A., & Yomogida, K. (2019). Near-fault ocean-bottom observation of an offshore 743

earthquake by DONET: Array-based measurement of tsunami phase speed and early 744

estimation of tsunami height. Paper S33D-0604 presented at AGU Fall Meeting 2019, San 745

Francisco, 9–13 December 2019. 746

Mochizuki, M., Kanazawa, T., Uehira, K., Shimbo, T., Shiomi, K., Kunugi, T., …, & Yamada, 747

T. (2016). S-net project: construction of large scale seafloor observatory network for 748

tsunamis and earthquakes in Japan. Paper NH43B-1840 presented at AGU Fall Meeting 749

2016, San Francisco, 12–16 December 2016. 750

Mulia, I. E., Gusman, A. R., Williamson, A. L., & Satake, K. (2019). An optimized array 751

configuration of tsunami observation network off southern Java, Indonesia. Journal of 752

Geophysical Research: Solid Earth, 9622-9637. https://doi.org/10.1029/2019JB017600 753

Mungov, G., Eblé, M., & Bouchard, R. (2013). DART® Tsunameter Retrospective and Real-754

Time Data: A Reflection on 10 Years of Processing in Support of Tsunami Research and 755

Operations. Pure and Applied Geophysics, 170, 1369–1384. 756

https://doi.org/10.1007/s00024-012-0477-5 757

Nakamura, T., Araki, E., Takahashi, N., Suzuki, K., Yamamoto, S., Korenaga, M., & Noda, S. 758

(2019). Magnitude estimation by using ocean-bottom pressure data. Paper SCG59-P03 759

presented at JpGU Meeting 2019, Makuhari, Japan, 26–30 May 2019. 760

National Research Institute for Earth Science and Disaster Resilience [NIED] (2019), NIED F-761

net, National Research Institute for Earth Science and Disaster Resilience. 762

https://doi.org/10.17598/NIED.0005. 763

Nishida, K., Maeda, T., & Fukao, Y. (2019). Seismic observation of tsunami at island broadband 764

stations. Journal of Geophysical Research: Solid Earth, 124, 1910–1928. 765

https://doi.org/10.1029/2018jb016833 766

Nosov, M. A., & Kolesov, S. V. (2007). Elastic oscillations of water column in the 2003 767

Tokachi-oki tsunami source: in-situ measurements and 3-D numerical modelling. Natural 768

Hazards and Earth System Science, 7, 243–249. https://doi.org/10.5194/nhess-7-243-2007 769


Nosov, M. A., Kolesov, S. V., Denisova, A. V., Alekseev, A. B., & Levin, B. V. (2007). On the 770

near-bottom pressure variations in the region of the 2003 Tokachi-Oki tsunami source. 771

Oceanology, 47, 26–32. https://doi.org/10.1134/S0001437007010055 772

Okada, Y., Kasahara, K., Hori, S., Obara, K., Sekiguchi, S., Fujiwara, H., & Yamamoto, A. 773

(2004). Recent progress of seismic observation networks in Japan - Hi-net, F-net, K-NET 774

and KiK-net. Earth, Planets and Space, 56, xv–xxviii. https://doi.org/10.1186/BF03353076 775

Pedlosky, J. (2013). Waves in the ocean and atmosphere: introduction to wave dynamics. Berlin: 776

Springer Berlin. https://doi.org/10.1007/978-3-662-05131-3 777

Poupardin, A., Heinrich, P., Hébert, H., Schindelé, F., Jamelot, A., Reymond, D., & Sugioka, H. 778

(2018). Traveltime delay relative to the maximum energy of the wave train for dispersive 779

tsunamis propagating across the Pacific Ocean: The case of 2010 and 2015 Chilean 780

Tsunamis. Geophysical Journal International, 214, 1538–1555. 781

https://doi.org/10.1093/gji/ggy200 782

Rabinovich, A. B., Thomson, R. E., & Fine, I. V. (2013). The 2010 Chilean tsunami off the west 783

coast of Canada and the northwest coast of the United States. Pure and Applied 784

Geophysics, 170, 1529–1565. https://doi.org/10.1007/s00024-012-0541-1 785

Rabinovich, A. B., & Eblé, M. C. (2015). Deep-Ocean Measurements of Tsunami Waves. Pure 786

and Applied Geophysics, 172, 3281–3312. https://doi.org/10.1007/s00024-015-1058-1 787

Saito, T. (2017). Tsunami generation: validity and limitations of conventional theories. 788

Geophysical Journal International, 210, 1888–1900. https://doi.org/10.1093/gji/ggx275 789

Saito, T. (2019). Tsunami generation and propagation. Tokyo: Springer Japan. 790

https://doi.org/10.1007/978-4-431-56850-6 791

Saito, T., Matsuzawa, T., Obara, K., & Baba, T. (2010). Dispersive tsunami of the 2010 Chile 792

earthquake recorded by the high-sampling-rate ocean-bottom pressure gauges. Geophysical 793

Research Letters, 37, L23303. https://doi.org/10.1029/2010GL045290 794

Sandanbata, O., Watada, S., Satake, K., Fukao, Y., Sugioka, H., Ito, A., & Shiobara, H. (2018). 795

Ray tracing for dispersive tsunamis and source amplitude estimation based on Green’s law: 796

application to the 2015 volcanic tsunami earthquake near Torishima, south of Japan. Pure 797

and Applied Geophysics, 175, 1371–1385. https://doi.org/10.1007/s00024-017-1746-0 798

Satake, K., Baba, T., Hirata, K., Iwasaki, S. I., Kato, T., Koshimura, S., … Terada, Y. (2005). 799

Tsunami source of the 2004 off the Kii Peninsula earthquakes inferred from offshore 800


tsunami and coastal tide gauges. Earth, Planets and Space, 57, 173–178. 801

https://doi.org/10.1186/BF03351811 802

Shearer, P. M. (2009). Introduction to seismology (2nd ed.). New York: Cambridge University 803

Press. https://doi.org/10.1017/CBO9780511841552 804

Tanioka, Y., Shibata, M., Yamanaka, Y., Gusman, A. R., & Ioki, K. (2019). Generation 805

mechanism of large later phases of the 2011 Tohoku-oki tsunami causing damages in 806

Hakodate, Hokkaido, Japan. Progress in Earth and Planetary Science, 6, 30. 807

https://doi.org/10.1186/s40645-019-0278-x 808

Titov, V., Rabinovich, A. B., Mofjeld, H. O., Thomson, R. E., & Gonzalez, F. I. (2005). The 809

Global Reach of the 26 December 2004 Sumatra Tsunami. Science, 309, 2045–2048. 810

https://doi.org/10.1126/science.1114576 811

Tonegawa, T., Fukao, Y., Shiobara, H., Sugioka, H., Ito, A., & Yamashita, M. (2018). Excitation 812

location and seasonal variation of transoceanic infragravity waves observed at an absolute 813

pressure gauge array. Journal of Geophysical Research: Oceans, 123, 40–52. 814

https://doi.org/10.1002/2017JC013488 815

Tsushima, H., & Ohta, Y. (2014). Review on near-field tsunami forecasting from offshore 816

tsunami data and onshore GNSS data for tsunami early warning. Journal of Disaster 817

Research, 9(3), 339–357. https://doi.org/10.20965/jdr.2014.p0339 818

Uehira, K., Kanazawa, T., Mochizuki, M., Fujimoto, H., Noguchi, S., Shinbo, T., …, & Yamada, 819

T. (2016) Outline of seafloor observation network for earthquakes and tsunamis along the 820

Japan Trench (S-net), Paper EGU2016–13832 presented at EGU General Assembly 2016, 821

Vienna, Austria, 17–22 April 2016. 822

Wallace, L. M., Webb, S. C., Ito, Y., Mochizuki, K., Hino, R., Henrys, S., … Sheehan, A. F. 823

(2016). Slow slip near the trench at the Hikurangi subduction zone, New Zealand. Science, 824

352, 701–704. https://doi.org/10.1126/science.aaf2349 825

Watada, S., Kusumoto, S., & Satake, K. (2014). Travel time delay and initial phase reversal of 826

distant tsunamis coupled with the self-gravitating elastic Earth. Journal of Geophysical 827

Research: Solid Earth, 119, 4287–4310. https://doi.org/10.1002/2013JB010841 828

Webb, S. C. (1998). Broadband seismology and noise under the ocean. Reviews of Geophysics, 829

36, 105–142. https://doi.org/10.1029/97RG02287 830


Webb, S. C., & Crawford, W. C. (1999). Long-period seafloor seismology and deformation 831

under ocean waves. Bulletin of the Seismological Society of America, 89(6), 1535–1542. 832

Webb, S. C., & Nooner, S. L. (2016). High-resolution seafloor absolute pressure gauge 833

measurements using a better counting method. Journal of Atmospheric and Oceanic 834

Technology, 33(9), 1859–1874. https://doi.org/10.1175/JTECH-D-15-0114.1 835

Wessel, P., Luis, J. F., Uieda, L., Scharroo, R., Wobbe, F., Smith, W. H. F., & Tian, D. (2019). 836

The Generic Mapping Tools Version 6. Geochemistry, Geophysics, Geosystems, 20, 5556–837

5564. https://doi.org/10.1029/2019GC008515 838

1

Earth and Space Science

Supporting Information for

Ultra-broadband seismic and tsunami wave observation of high-sampling ocean-bottom pressure gauge covering periods from seconds to hours

T. Kubota1, T. Saito1, N. Y. Chikasada1, and W. Suzuki1

1 National Research Institute for Earth Science and Disaster Resilience, Tsukuba, Ibaraki, Japan.

Contents of this file

Texts S1 to S2 Figures S1 to S12 Tables S1 to S3

Additional Supporting Information (Files uploaded separately)

Dataset S1 is available at https://doi.org/10.17598/NIED.0018. The Dataset file includes tilt record at SAMH during the 2010 Chile earthquake. The .sac files are in SAC format and .csv ascii file is also available. Time zone of the record is UTC.

Introduction

Text S1 describes the specification of the instruments used in this study. Text S2 explains the procedure employed to calculate the coherency and phase spectra. Figure S1 shows the time series for the DART stations installed in the Pacific Ocean during the 2010 Chile earthquake. Figures S2 and S3 show the spectrograms displayed with the linear scale on the elapsed time. Figure S4 shows the comparison between the theoretical and observed arrival times of the seismic phases at KSRF. Figure S5 shows the AK135 structure model and theoretical dispersion relation. Figures S6 and S7 show the AK135 structure model without a seawater layer and its dispersion relation, and the comparison between the spectrograms and theoretical arrival times. Figures S8 and S9 display the PREM structure model and its dispersion relation, and the comparison between the spectrograms and theoretical arrival times. Figure S10 compares theoretical arrival times of the Rayleigh waves between different structure models. Figure S11 shows tsunami amplitude decay in the seafloor pressure gauge. Comparison between the pressure change and vertical velocity at varying period bands are shown in Figure S12. Tables S1 and S2 are the station lists for the different instruments around northern Japan and the DART stations in the Pacific Ocean, respectively. Table S3 is the event list investigated in this study.

2

Text S1. This text summarizes specifications of the observation instruments used in this study. The specification of the Off-Kushiro PG and OBS, which are maintained by the Japan

Agency for Marine-Earth Science and Technology (JAMSTEC), is described in Hirata et al. (2002) and the website of the JAMSTEC Submarine Cable Data Center (SCDC, http://www.jamstec. go.jp/scdc/top_e.html). The OBS uses JA-5III-A servo accelerometer developed by the Japan Aviation Electronics Industry. The details summarized in http://www.jamstec.go.jp/scdc/ html_sysindex_e/obsdetail.html. The direction of the original X, Y, and Z components are converted to the EW, NS, and UD components based on the procedure provided by the JAMSTEC SCDC. The response of the OBS is flat at frequency bands of 0 to 40 Hz. The PG equips the quartz-oscillation type pressure sensor HP2813E developed by Hewlett-Packard, Inc. (http:// www.jamstec.go.jp/scdc/html_sysindex_e/hpdetail.html).

The onshore seismometer at KSRF, which is one of the NIED (National Research Institute for Earth Science and Disaster Resilience)’s F-net observation network, uses Streckeisen STS-2 broadband seismometer (Okada et al., 2004; NIED, 2019). The nearshore tiltmeter at SAMH uses a high-sensitivity accelerometer JTS-3B manufactured by Mitutoyo Corporation (Okada et al., 2014). Text S2.

This text explains the procedure to calculate the coherency and phase difference between the pressure and vertical acceleration waveforms in Figure 8. We first calculate the Fourier spectra X(ω) and Y(ω) from the time series x(t) and y(t) in every 1 s step, with a tapered 512 s time window. The cross spectra SXY(ω) is defined as: !!"(#) = ⟨'(#)(∗(#)⟩, (S1) where ⟨⟩ denotes the ensemble average, and * is the complex conjugate. Considering the average over time as the ensemble average, we calculate SXY(ω) using the time window of ±256 s, in every 30 s step. Using the cross spectra SXY(ω) and the power spectra SXX(ω) and SXY(ω), the coherency CohXY(ω) is calculated as: ,-ℎ!"(#) = |%!"(')|

)%!!('))%""('). (S2)

In addition we also calculate the phase shift between the spectra X(ω) and Y(ω), θXY(ω), using the phases θX(ω) and θX(ω), as: /!"(#) = /!(#) − /"(#) = tan*+ 5,-./0[⟨!(')⟩]50.6[⟨!(')⟩] 6 − tan

*+ 5,-./0[⟨"(')⟩]50.6[⟨"(')⟩] 6. (S3)

1

Figure S1. DART records installed in the Pacific Ocean associated with the 2010 Chile earthquake. The sampling rate is 1/15 Hz except for the station 32412. The sampling rate for the station 32412 is shown in Figures S1b and S1c (red: 1/15 Hz, green: 1/60 Hz, blue: 1/900 Hz).

0 12 24 36 48 60 72 84 96 108 120 132 144

Elapsed time from the origin time [h]

(a) No filter

21413 100hPa

51407 100hPa

51406 100hPa

32411 100hPa

32412 100hPa

0 3600 7200 10800 14400 18000 21600

Elapsed time from the origin time [s]

(b) Highpass 10800s

21413 5hPa

51407 5hPa

51406 5hPa

32411 5hPa1min15s

32412 5hPa

0 12 24 36 48 60 72 84 96 108 120

Elapsed time from the origin time [h]

(c) Bandpass 30s-10800s

21413 10hPa

51407 10hPa

51406 10hPa

32411 10hPa15min

32412 10hPa

1

Figure S2. Spectrograms for seismic waves. (a) Theoretical arrival time of seismic waves. (b–l) Spectrograms for onshore and offshore instruments. See the caption in Figure 3 for a description of each panel.

2

51020

50100200

500F

requency

[m

Hz]

0 1 2 3

Elapsed time [h]

(a)

P SS R

ayl

eig

hL

ove

0 1 2 3

Elapsed time [h]

(b) PG KPG1

101 102

[hPa!s]

0 1 2 3

Elapsed time [h]

(c) PG KPG2

101 102

[hPa!s]

0 1 2 3

Elapsed time [h]

(d) Tidegauge Hanasaki

102 103

[cm!s]

2

51020

50100200

500

Period [s]

2

51020

50100200

500

Fre

quency

[m

Hz]

0 1 2 3

Elapsed time [h]

(e) Onshore seismometer KSRF.U

102

[µm/s2!s]

0 1 2 3

Elapsed time [h]

(f) Onshore seismometer KSRF.R

102 103

[µm/s2!s]

0 1 2 3

Elapsed time [h]

(g) Onshore seismometer KSRF.T

102 103

[µm/s2!s]

0 1 2 3

Elapsed time [h]

(k) Tiltmeter SAMH.T

101 102

[µrad!s]

2

51020

50100200

500

Period [s]

2

51020

50100200

500

Fre

quency

[m

Hz]

0 1 2 3

Elapsed time [h]

(h) OBS KOBS1.U

103

[µm/s2!s]

0 1 2 3

Elapsed time [h]

(i) OBS KOBS1.R

103

[µm/s2!s]

0 1 2 3

Elapsed time [h]

(j) OBS KOBS1.T

103

[µm/s2!s]

0 1 2 3

Elapsed time [h]

(l) Tiltmeter SAMH.T

101 102

[µrad!s]

2

51020

50100200

500

Period [s]

2

Figure S3. Spectrograms for tsunamis. (a) Theoretical arrival time of tsunamis. (b–l) Spectrograms for onshore and offshore instruments. See the caption in Figure 3 for a description of each panel.

0.20.5

125

102050

100200500

Fre

quency

[m

Hz]

0 24 48 72 96 120 144

Elapsed time [h]

(a)

Tsu

na

mi

0 24 48 72 96 120 144

Elapsed time [h]

(b) PG KPG1

101 102

[hPa!s]

0 24 48 72 96 120 144

Elapsed time [h]

(c) PG KPG2

101 102

[hPa!s]

0 24 48 72 96 120 144

Elapsed time [h]

(d) Tidegauge Hanasaki

102 103

[cm!s]

25102050100200500100020005000

Period [s]

0.20.5

125

102050

100200500

Fre

quency

[m

Hz]

0 24 48 72 96 120 144

Elapsed time [h]

(e) Onshore seismometer KSRF.U

101 102

[µm/s2!s]

0 24 48 72 96 120 144

Elapsed time [h]

(f) Onshore seismometer KSRF.R

100 101 102

[µm/s2!s]

0 24 48 72 96 120 144

Elapsed time [h]

(g) Onshore seismometer KSRF.T

100 101 102

[µm/s2!s]

0 24 48 72 96 120 144

Elapsed time [h]

(k) Tiltmeter SAMH.R

100 101 102

[µrad!s]

25102050100200500100020005000

Period [s]

0.20.5

125

102050

100200500

Fre

quency

[m

Hz]

0 24 48 72 96 120 144

Elapsed time [h]

(h) OBS KOBS1.U

102 103 104

[µm/s2!s]

0 24 48 72 96 120 144

Elapsed time [h]

(i) OBS KOBS1.R

102 103 104

[µm/s2!s]

0 24 48 72 96 120 144

Elapsed time [h]

(j) OBS KOBS1.T

102 103 104

[µm/s2!s]

0 24 48 72 96 120 144

Elapsed time [h]

(l) Tiltmeter SAMH.T

100 101 102

[µrad!s]

25102050100200500100020005000

Period [s]

3

Figure S4. Theoretical arrival times of the seismic phases at KSRF calculated from the AK135 structure model, for (left) the P waves and (right) the S-waves.

!100

0

100

200

300

[cm

/s]

PK

IKP

PK

PP

KiK

P

PK

P

U

R

T

KSRF (Raw)

!100

0

100

200

300

[cm

/s]

1150 1200 1250 1300

Elapsed time [s]

PK

IKP

PK

PP

KiK

P

PK

P

U

R

T

KSRF (BPF 4-6s)

!200

0

200

400

600

SS

U

R

T

KSRF (Raw)

!200

0

200

400

600

2400 2500 2600 2700 2800 2900 3000 3100 3200

Elapsed time [s]

SS

U

R

T

KSRF (BPF 20-200s)

4

Figure S5. (a) Velocity structure model based on AK135. (b) The phase velocity (dashed lines) and group velocity (solid lines) of the fundamental mode of the Rayleigh (red) and Love (green) waves and tsunami (blue). (c) Theoretical arrival times of P and SS waves (black), the surface waves (red and pink: Rayleigh, green: Love waves), and tsunamis (blue). The pink line is the arrival of the Rayleigh wave along the opposite direction of the great circle path on the earth.

0

20

40

60

80

100

120

140

160

180

200

Depth

[km

]

0 2 4 6 8 10

Vel [km/s], ! [g/cm3]

PS!

(a) AK135

10!3

10!2

10!1

100

101V

elo

city

[km

/s]

0.2 0.5 1 2 5 10 20 50 100200 500

Frequency [mHz]

25102050100200500100020005000

Love

Rayleigh

TsunamiPhase velocityGroup velocity

(b)

1

2

5

10

20

50

100

200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

2

5

10

20

50

100

200

500

1000

Period [s]

(c)

P SS

Ra

yle

igh

Lo

ve

Tsu

na

mi

5

Figure S6. (a) Velocity structure model based on AK135 without seawater layer. (b) The phase velocity (dashed lines) and group velocity (solid lines) of the fundamental mode of the Rayleigh (red) and Love (green) waves and tsunami (blue). (c) Theoretical arrival times of P and SS waves (black), the surface waves (red and pink: Rayleigh, green: Love waves), and tsunamis (blue). The pink line is the arrival of the Rayleigh wave along the opposite direction of the great circle path on the earth.

0

20

40

60

80

100

120

140

160

180

200

Depth

[km

]

0 2 4 6 8 10


PS!

(a) AK135 without seawater

10!3

10!2

10!1

100

101V

elo

city

[km

/s]

0.2 0.5 1 2 5 10 20 50 100200 500

Frequency [mHz]

25102050100200500100020005000

Love

Rayleigh

Phase velocityGroup velocity

(b)

1

2

5

10

20

50

100

200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

2

5

10

20

50

100

200

500

1000

Period [s]

(c)

P SS

Ra

yle

igh

Lo

ve

6

Figure S7. Comparison of theoretical arrival times of the seismic and tsunami waves based on the AK135 without seawater layer and the spectrograms.

12

51020

50100200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

P

Rayl

eig

h

Tsu

nam

i

(a) PG KPG1

101 102

[hPa!s]

0.1 1 10 100

Elapsed time [h]

Tsu

nam

i

(b) Tidegauge Hanasaki

102 103

[cm!s]

2

51020

50100200

5001000

Period [s]

12

51020

50100200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

P

Rayl

eig

h(c) Onshore seismometer KSRF.U

101 102 103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Rayl

eig

h(d) Onshore seismometer KSRF.R

102 103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Love

(e) Onshore seismometer KSRF.T

102 103

[µm/s2!s]

2

51020

50100200

5001000

Period [s]

12

51020

50100200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

P

Rayl

eig

h(f) OBS KOBS1.U

103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Rayl

eig

h(g) OBS KOBS1.R

103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Love

(h) OBS KOBS1.T

103

[µm/s2!s]

2

51020

50100200

5001000

Period [s]

7

Figure S8. (a) Velocity structure model based on the PREM model. (b) The phase velocity (dashed lines) and group velocity (solid lines) of the fundamental mode of the Rayleigh (red) and Love (green) waves and tsunami (blue). (c) Theoretical arrival times of P and SS waves (black), the surface waves (red and pink: Rayleigh, green: Love waves), and tsunamis (blue). The pink line depicts the arrival of the Rayleigh wave along the opposite direction of the great circle path on the earth.

0

20

40

60

80

100

120

140

160

180

200

Depth

[km

]

0 2 4 6 8 10


PS!

(a) PREM

10!3

10!2

10!1

100

101V

elo

city

[km

/s]

0.2 0.5 1 2 5 10 20 50 100200 500

Frequency [mHz]

25102050100200500100020005000

Love

Rayleigh

Phase velocityGroup velocity

(b)

1

2

5

10

20

50

100

200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

2

5

10

20

50

100

200

500

1000

Period [s]

(c)

P SS

Ra

yle

igh

Lo

ve

1

Figure S9. Comparison of theoretical arrival times of the seismic and tsunami waves based on PREM and the spectrograms.

12

51020

50100200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

P

Rayl

eig

h

Tsu

nam

i

(a) PG KPG1

101 102

[hPa!s]

0.1 1 10 100

Elapsed time [h]

Tsu

nam

i

(b) Tidegauge Hanasaki

102 103

[cm!s]

2

51020

50100200

5001000

Period [s]

12

51020

50100200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

P

Rayl

eig

h(c) Onshore seismometer KSRF.U

101 102 103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Rayl

eig

h(d) Onshore seismometer KSRF.R

102 103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Love

(e) Onshore seismometer KSRF.T

102 103

[µm/s2!s]

2

51020

50100200

5001000

Period [s]

12

51020

50100200

500

Fre

quency

[m

Hz]

0.1 1 10 100

Elapsed time [h]

P

Rayl

eig

h(f) OBS KOBS1.U

103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Rayl

eig

h(g) OBS KOBS1.R

103

[µm/s2!s]

0.1 1 10 100

Elapsed time [h]

SS

Love

(h) OBS KOBS1.T

103

[µm/s2!s]

2

51020

50100200

5001000

Period [s]

2

Figure S10. Comparison of theoretical arrival times of the Rayleigh wave from different velocity structure models at KPG1. Red, green, and blue lines denote the theoretical arrival times of the Rayleigh waves based on the AK135 model, AK135 model without seawater layer, and the PREM model, respectively.

2

5

10

20

50

100

200

500

Fre

quency

[m

Hz]

0 1 2 3 4

Elapsed time [h]

2

5

10

20

50

100

200

500

Period [s]

AK

135

w/o

wate

r

PR

EM

(b) PG KPG1

101 102

[hPa!s]

3

Figure S11. The factor of 1/cosh(kH0) as a function of (left) the wave number, k, normalized by the water depth H0 and (right) the frequency.

10!3

10!2

10!1

100

1/c

osh

(kH

0)

0.01 0.1 1 10 100Normalized Wavenumber, kH0

(a)10!3

10!2

10!1

100

1/c

osh

(kH

0)

1 2 5 10 20 50 100Frequency [mHz]

1020501002005001000Period [s]

H0 =

1000m

2000m

6000m

(b)

1

Figure S12. Comparison of the pressure and vertical velocity at period bands of (a) 4–6 s, (b) 6–10s, (c) 10–20 s, (d) 20–50 s, (e) 50–100 s, (f) 100–200 s, (g) 200–400 s, and (h) 400–800 s. See Figure 10 for the other description.

(a)

!240

!120

0

120

240

[Pa

]

BPF 4-6sKPG1 p

!240

!120

0

120

240

[Pa

]

BPF 4-6sKOBS1 !wc0vz

!1

0

1

Co

rre

latio

n

0 3600 7200 10800 14400 18000 21600

Elapsed time [s]

CC(TL=60s)Correlation

(b)

!240

!120

0

120

240

[Pa

]

BPF 6-10sKPG1 p

!240

!120

0

120

240

[Pa

]


!1

0

1

Co

rre

latio

n

0 3600 7200 10800 14400 18000 21600

Elapsed time [s]


(c)

!240

!120

0

120

240

[Pa

]

BPF 10-20sKPG1 p

!240

!120

0

120

240

[Pa

]


!1

0

1

Co

rre

latio

n

0 3600 7200 10800 14400 18000 21600

Elapsed time [s]


(d)

!240

!120

0

120

240

[Pa

]

BPF 20-50sKPG1 p

!240

!120

0

120

240

[Pa

]


!1

0

1

Co

rre

latio

n

0 3600 7200 10800 14400 18000 21600

Elapsed time [s]


(e)

!120

!60

0

60

120

[Pa

]

BPF 50-100sKPG1 p

!120

!60

0

60

120

[Pa

]


!1

0

1

Co

rre

latio

n

0 3600 7200 10800 14400 18000 21600

Elapsed time [s]


(f)

!60

!30

0

30

60

[Pa

]

BPF 100-200sKPG1 p

!60

!30

0

30

60

[Pa

]


!1

0

1

Co

rre

latio

n0 3600 7200 10800 14400 18000 21600

Elapsed time [s]


(g)

!30

!15

0

15

30

[Pa

]

BPF 200-400sKPG1 p

!30

!15

0

15

30

[Pa

]


!1

0

1

Co

rre

latio

n

0 3600 7200 10800 14400 18000 21600

Elapsed time [s]


(h)

!30

!15

0

15

30

[Pa

]

BPF 400-800sKPG1 p

!30

!15

0

15

30

[Pa

]


!1

0

1

Co

rre

latio

n

0 3600 7200 10800 14400 18000 21600

Elapsed time [s]


1

Table S1. Summary of the records used in this study.

Station Latitude [°N]

Longitude [°E]

Altitude [m] Instrument Sampling

ratea observation # of component

KPG1 41.7040 144.4375 -2218 Pressure gauge 10 Hz Pressure 1

KPG2 42.2365 144.8454 -2210 Pressure gauge 10 Hz Pressure 1

Hanasaki 43.2833 145.5667 N/A Tide gauge 1/30 Hz (30 s)

Sea-surface height 1

KSRF 42.9820 144.4851 18 Onshore seismometer 100 Hz Velocityb 3

KOBS1 41.6870 144.3945 -2329 Ocean-bottom seismometer 100 Hz Acceleration 3

SAMH 42.1330 142.9164 -63 Tiltmeter 10 Hz Tilt 2

aAll records are re-sampled to 1 Hz. bVelocity records are converted to accelerograms. Table S2. Summary of the DART records used in this study.

Station Latitude [°N]

Longitude [°E]

Altitude [m]

Approx. distance from the source [km] Sampling ratea

32412 -17.9750 -86.3920 -4325 2400 1/15 Hz, 1/60 Hz (1 min), or 1/900 Hz (15 min)a

32411 4.9242 -90.6858 -3232 4900 1/15 Hz

51406 -8.4925 -125.0214 -4480 6100 1/15 Hz

51407 19.6368 -156.5192 -4682 10300 1/15 Hz

21413 30.5460 152.1140 -5827 12300 1/15 Hz

aSampling rate depends on the time (see Figure S1 for details).

2

Table S3. Summary of the earthquakes analyzed in this study.a

Event name Date (yyyy/mm/dd)

Time (hh:mm:ss UTC) Mw

Approx. Epicentral distance

[km]

Latitude [°N]

Longitude [°E]

Depth [km]

2010 Chile 2010/2/27 06:34:11.530 8.8 17000 -36.122 -72.898 22.9

2006 Kuril 2006/11/15 11:14:13.570 8.3 890 46.592 153.266 10.0

2007 Kuril 2007/01/13 04:23:21.160 8.1 950 46.243 154.524 10.0

2011 Off Miyagi 2011/03/09 02:45:20.330 7.3 390 38.435 142.842 32.0

2011 Tohoku 2011/03/11 05:46:24.120 9.0 420 38.297 142.373 29.0

2012 Sumatra 2012/04/11 08:38:36.720 8.6 6720 2.327 93.063 20.0

2018 Hokkaido

Japan

2018/09/05 18:07:59.150 6.6 230 42.686 141.929 35.0

aSource information derived from the U.S. Geological Survey (USGS).

ultra-broadband seismic and tsunami wave observation of high …kubota/uploads/draft/kubota... ·...

Documents