the 25 october 2010 mentawai tsunami earthquake (mw 7.8...

1

AUXILIARY MATERIALS

The 25 October 2010 Mentawai tsunami earthquake (Mw 7.8) and the tsunami hazard presented by shallow megathrust ruptures

T. Lay Department of Earth and Planetary Sciences, University of California, Santa Cruz, California, USA C. J. Ammon Department of Geosciences, The Pennsylvania State University, University Park, Pennsylvania, USA H. Kanamori Seismological Laboratory, California Institute of Technology, Pasadena, California, USA Y. Yamazaki, K. F. Cheung Department of Ocean and Resource Engineering, University of Hawaii at Manoa, Honolulu, Hawaii USA A. R. Hutko Incorporated Research Institutions for Seismology, Data Management Center, Seattle, Washington, USA

2

Animation S1. QuickTime Movie of F-Net (Japan) network P wave data back-projected to the Mentawai

source region. The left panel indicates the results of linear stacking of the P waves filtered in the 0.5-2.0

Hz band with a 3 s averaging window. The peak amplitude of the stack as a function of time is shown

below and tracked as the animation runs. The right panel is similar, but shows the cube-root stack

amplitudes of the back-projected data.

http://es.ucsc.edu/~thorne/GRL_MENTAWAI/LAKYCH_Movie.S1.mov

[H.264 Encoded QuickTime Movie, ~0.8 MB]

3

Animation S2. Animation of the aftershock sequence from 25 October 2010 to 25 November 2010. Red

dots identify events shallower than 33 km, orange dots indicate events with depths from 33 to 100 km.

Events fade smoothly from the origin time and vanish after two weeks. Data are from the U.S. Geological

Survey, National Earthquake Information Center.



4

Animation S3. Animation of the rupture model for the inversion of P, SH and R1 STFs. The cumulative

seismic moment at each subfault is shown on the top, while the moment rate release at a given instant of

time is shown below. The outer expanding circle corresponds to the rupture front defined by Vr = 1.5

km/s. The inner expanding circle corresponds to the end of subfault slip front that lags behind by 32 s.

Darker colors indicate larger subfault seismic moment (left) or moment rate (right).



5

Animation S4. Animation of the rupture model for the inversion of only P and SH body waves. The

cumulative seismic moment at each subfault is shown on the left, while the moment rate release at a given

instant of time is shown on the right. The outer expanding circle corresponds to the rupture front defined

by Vr = 1.5 km/s. The inner expanding circle corresponds to the end of subfault slip front that lags

behind by 16 s. Darker colors indicate larger subfault seismic moment (left) or moment rate (right).



6

Figure S1. The fractional signal misfit for simultaneous inversions of P, SH, and R1 STFs for the 25

October 2010 Mentawai earthquake for different assumed rupture velocities. There is a slight preference

for a rupture velocity of 1.5 km/s, and this was used for the final inversions for both seismic data sets.

7

Figure S2. Comparison of observed (black lines) and predicted (red lines) waveforms for the

simultaneous inversion of P, SH and R1 STFs for the 25 October 2010 Mentawai earthquake. The R1

signals have been corrected for propagation effects and provide direct constraints on the long-period level

of the source function, along with being more sensitive to source finiteness than the body waves. The

corresponding slip distribution is shown in Figure 1a. The labeling indicates the phase, epicentral

distance, station code and azimuth for each arrival.

8

Figure S3. Subfault seismic moment (top) and rupture duration (bottom) for the simultaneous inversion

of P, SH, and R1 STFs. The subfault moments are in units of 1019 Nm.

9

Figure S4. Observed (black bold lines) and predicted (red lines) P and SH waveforms for the 25 October

2010 Mentawai earthquake for the body wave only inversion. Data and synthetics are on true relative

amplitudes for P and SH separately, with all SH signal amplitudes being reduced by a factor of 5 relative

to the P signals. The station code and azimuth from the source are indicated. These data have limited

sensitivity to the rupture directivity, but do constrain the depth extent of the rupture. The corresponding

slip distribution is shown in Figure 1b.

10

Figure S5. Subfault seismic moment (top) and rupture source time function (bottom) for the inversion of

P and SH waves with a rupture velocity of 1.5 km/s. The relative moment is indicated by the length of

the arrows, with amplitudes being contoured. The subfault rakes are indicated by arrows at the top. All of

the subfault source durations are parameterized as 7 2-s rise time, 2-s fall time triangles shifted by 2 s,

giving total subfault durations of 16 s.

11

Figure S6. (a) Plot of the 25 October 2010 aftershock distribution versus distance along the trench (assumed to

strike 324°N) from the mainshock hypocenter and days relative to the mainshock origin time. Symbols are scaled

proportional to the cube of the maximum of mb, Ms, or Mw. Data are from the U.S. Geological Survey, National

Earthquake Information Center. (b) Similar projection of the aftershock locations in the dip direction (234N,

orthogonal to that shown in (a)). The data suggest the updip aftershocks shut off before those in the downdip regime.

Aftershocks updip appear more plentiful and smaller in magnitude, the downdip region seems to have fewer but

larger aftershocks. Caution is needed, however, since these events are all near the threshold of global detection, and

a pattern in one earthquake can at best suggest such a characteristic.

12

Figure S7. P-wave ground velocity based estimates of the source spectra for large Indonesian earthquakes

compared with reference ω-squared source spectra. The three tsunami earthquakes (2 June 1994; 17 July

2006; 25 October 2010) are shown on the left, with all of the observations plotting well below the

reference curves, in contrast to the behavior for deeper conventional megathrust events of 2 November

2002, 20 February 2008, and 12 September 2007 shown on the right.

13

Figure S8. Moment-scaled seismic energy for large recent interplate earthquakes (black dots), tsunami

earthquakes (red circles), and intraplate earthquakes (blue circles). The tsunami earthquakes tend to have

relatively low ER/Mo ratios.

14

Figure S9. The tsunami bathymetry computational grid from the GEBCO database, decimated to ~1800

m resolution. The location of DART 56001 is indicated by the white dot, and the epicenters of large

tsunami earthquakes along Sumatra and Java are indicated by the black stars with red shading indicating

the approximate rupture areas.

15

Figure S10. (a) The slip distribution from the finite fault inversion using P, SH and R1 STFs. (b)

Computed total sea bottom vertical deformation for the finite-fault model; the dynamic sea-bottom

motion is used as the initial condition for the tsunami modeling. The red dot indicates the mainshock

epicenter.

16

Figure S11. (a) The slip distribution from the finite fault inversion using P and SH with a layered source

velocity structure. (b) Computed total sea bottom vertical deformation for the finite-fault model; the

dynamic sea-bottom motion is used as the initial condition for the tsunami modeling. The red dot

indicates the mainshock epicenter.

17

Table S1. Local 1D Velocity Model Used for P and SH Finite Fault Inversion

Vp (km/s) Vs (km/s) Density(kg/m3 Thickness (km)

1.5 0.0 1000 3.0

2.5 1.7 1500 1.0

3.1 1.9 1700 2.0

5.0 2.9 2300 4.0

6.0 3.46 2600 10.0

6.7 3.87 2800 20.0

7.7 4.5 3300 Half-Space

the 25 october 2010 mentawai tsunami earthquake (mw 7.8...

Documents