Extreme Slip in the Tohoku Earthquake: What Happened and A Role for Geodesy in Assessing Hazard from Extreme Events
Greg Beroza Stanford Geophysics Department School of Earth Sciences
PNT Challenges and Opportunities Symposium
Tectonic Setting
Japan Trench ~83 mm/yr convergence Old, cold Pacific Plate
Most of plate boundary is underwater so it’s hard to tell what’s going on out there.
Figure of GPS velocity vectors from GSI website Historical Earthquakes
Large (up to M 8+)
Not gigan c (M 9)
1896 Sanriku event (pre-instrumental era in seismology) weakly felt but big tsunami
Image: Chuck Ammon; Rupture outlines: Hiroo Kanamori; Earthquake Loca ons: USGS ERC Segmenta on Model
Off Miyagi, 6 earthquakes – 1793(M8.2), 1835(M7.3), 1861 (M7.4), 1897(M7.4), 1936(M7.4), and 1978(M7.4) – occurred in the past 200 years with an average interval of 37.1 years.
30-year probability of 99% (highest in Japan).
M7.2 earthquake in 2005. Since less than 7.5, the “expected” earthquake was not considered to have occurred.
Adapted from Preliminary report of the 2011 off the Pacific coast of Tohoku Earthquake Mar. 25, 2011 Yoshimitsu Okada (President, NIED) Geologic Hints of Larger Earthquakes
Last earthquake of this size may have been 869 AD Jogan earthquake. From tsunami
modeled as Mw 8.3 (Minoura et al., 2001), but could have been larger.
Unclear how extensive tsunami from 1611 Keicho earthquake was.
Also noted deposits from two earlier tsunamis indica ng, “…gigan c tsunamis occurred three mes during the last 3000 years.”
They wrote, “More than 1100 years have passed since the Jogan tsunami and, given the recurrence interval, the possibility of a large tsunami striking the Sendai Plain is high.” March 11, 2011 Mw 9.0 Mainshock Real-Time Seismology
0 minutes Earthquake began at 11 Mar 2011 05:46:23 UTC. (ruptured for ~160 seconds) 3 minutes JMA M 6.8
5 minutes PTWC M 7.3
10-20 minutes PTWC, ATWC, & JMA M 7.9 tsunami warnings (tsunami didn’t reach shore for ~30 minutes).
20 minutes USGS M 9.0 – clearly an extraordinary event.
Now M 9.0, but es mates range from 8.8 to 9.1.
Global Wavefield
M 7.9 a ershock
Everywhere on Earth moved by several cm. Dip Ide and Takeo (1997) 15 deg Ide (2002) 440 440 x 240 km fault plane discretized using km spline basis with 10-km interval in space and 5-s interval in time for 160 s Linear Inversion (NNLS) 240 km Space & Time Smoothing (Bayesian) ABIC Minimum Solution
Slip Evolu on in the Tohoku-Oki Earthquake
Ide et al. (2011) ANIMATION OF SLIP EVOLUTION
Slip-rate distribution
Time in seconds relative to the mainshock origin time is shown in the legend in the lower right.
Slip-rate (in meters/second) is shown with colors as the earthquake evolves.
Ide et al. (2011) FINAL SLIP
Moment-rate function indicates rapid growth about 50 s after initiation.
Moment rate is equivalent to one
Mw 8.0 earthquake every 2 seconds for an extended period.
Exceeds 30 m near the trench.
Slip extends along strike near the trench.
Ide et al. (2011) Other Slip Models
Broadband P P + R1 + cGPS Other Slip Models Ini al Updated Other Slip Models 1 sps GPS Measurements
Courtesy of Ronni Grapenthin and Jeff Freymueller, University of Alaska Preliminary GPS me series provided by the ARIA team at JPL and Caltech. Original GEONET RINEX data provided by the Geospa al Informa on Authority (GSI) of Japan. Permanent Displacements from GPS (Geospa al Informa on Authority of Japan) Seafloor Geodesy
Sato et al. (2011) Seafloor Geodesy Confirms Extreme Slip
Several cm accuracy achievable. Error for these measurements is about 10-20 cm for MYGW, MYGI, FUKU and 50-60 cm at KAMN and KAMS. Sato et al. (2011) Free Surface
As rupture approaches surface on a reverse fault, dynamic normal stress behind tip drops, causing higher stress drop. The effect propagates back down the fault, extends duration, and increases slip (Shi et al, 1998; Oglesby et al., 1998).
ShearStress
Excess stress drop
Ma and Beroza (2008) Oglesby et al. (1998) Slip on Fault with Shallow (15°) Dip
Tamura and Ide Free Surface
Ide et al. (2011) Extreme Shallow Slip Exceptional Tsunami
Excep onal tsunami excita on because largest displacement occurs near the trench where:
(1) it translates most directly into seafloor deforma on
(2) the displaced water column is deepest. Japan Coast Guard Ship Sailing Over Tsunami Extreme Shallow Slip Exceptional Tsunami
Excep onal tsunami excita on because largest displacement occurs near the trench where:
(1) it translates most directly into seafloor deforma on
(2) the displaced water column is deepest. May also have happened in the 1896 Sanriku earthquake. Another Possible Example: 1907 Sumatra
Near-trench slip, excep onal tsunami Kanamori et al. (2010) Another Possible Example: 2010 Mentawai Earthquake
Lay et al. (2011) Accurate Hazard Characterization is Required to Mitigate Risk Effectively
Extreme Events: Size of the earthquake was unan cipated. The last earthquake of this size there may have been the 869 AD Jogan earthquake. Record in Japan is among the longest anywhere, yet earthquakes over the last 1000 years have been considerably smaller. Other parts of the world where unexpectedly large earthquakes might be an cipated?
40,000 km of subduc on zones worldwide. Most are thought not to be capable of great earthquakes due to “segmenta on.” Examples include: Java, Central America, Mexico (1787?), Tonga, southern Europe, Makran.
In most places reliable historical records only for last few hundred years.
Most plate boundaries, par cularly in subduc on zones are largely underwater. A global seafloor geodesy effort could drama cally increase our understanding of hazard from great earthquakes.