The 2011 Tohoku Earthquake and Tsunami Hazard … 1896 Meiji Sanriku tsunami 1896 Meiji tsunami: 22,000 casualties (more than 2011 tsunami) M 7.2, Max tsunami height 38 m Weak shaking

The 2011 Tohoku Earthquake

and Tsunami Hazard Assessment

Kenji Satake

Earthquake Res. Inst. Univ. [email protected]‐tokyo.ac.jp

Outline

1. The 2011 Earthquake Source

2. The Tsunami

3. Past Tsunamis and the 2011 Source Model

4. Long‐term Forecast of Large Earthquakes

5. Tsunami Hazard Assessments at NPS

Outline


2. The Tsunami




The largest earthquake in Japan’s historyMarch 11 Earthquake38°06.2’ N, 142° 51.6’ Edepth 24 km14 h 46 min 18.1 s (JST)M 9.0

The earthquake“Off the Pacific Coast of Tohoku Earthquake” by JMA

The disaster “The Great East Japan Earthquake Disaster”by Cabinet Office

The 2011 Tohoku Earthquake

11 March Tohoku Eq.

• Interplate earthquake due to subduction of Pacific plate at 8 cm /year or 8 m / century

• Largest size (M=9.0) in Japan’s history

GPS data and slip distribution

GSI (2010, 2011)

March 11, 20111998‐2000

Sato et al. (Science 2011)

Max observed slip: 24 m horizontal3 m vertical

Max slip on fault (estimated): > 50 m

7

Large slip revealed by seafloor observations

Repeated bathymetry sounding: ~ 50 m offsetFujiwara et al. (Science 2011) GSI (2011)

The 2011 Earthquake Source

Aftershocks: 1day (JMA)Slip distribution:

strong motion data Yokota et al. (20112)

Slip distribution: GPS and GPS/AGSI (2011)

8

Outline


2. The Tsunami




Seafloor displacement(cross‐section)

March 11, 2011 tsunami

11

Iwate

Miyagi

Fukushima

Sendai plain

Large inundationabout 1 hour after eq.

Sanriku coast

High tsunami about 30 minutes after the earthquake

The Observed Tsunami

Outline


2. The Tsunami




The 2011 and past Tsunami sources

The 1896 Meiji Sanriku tsunami

1896 Meiji tsunami: 22,000 casualties (more than 2011 tsunami)

M 7.2, Max tsunami height 38 mWeak shaking but large tsunami“Tsunami earthquake” Width: 50 km, slip: 6 m

Near trench axis15

Tanioka and Satake (1996)

The 1896 Meiji Sanriku tsunami

1896 Meiji tsunami: 22,000 casualties (more than 2011 tsunami)

16

Tsunami heights along Sanriku coast

1933~1896~2011

1933< 1896~2011

1933~1896 < 2011

The 869 Jogan earthquakeNihon Sandai Jitsuroku (Chronicle of Japan)A large earthquake in MutsuPanic stricken by violent tremblingsFallen houses, wide‐opened ground fissuresRoaring like thunder heard from the sea Sea rushed into castle, a few hundred milesAbout 1,000 people were killed

Tsunami deposit studiesSand layer brought by tsunamibelow volcanic ash (AD915)

distributed ~ 5 kmfrom the coast

Tsunami sand

Peat

To‐aAD 915

18Dr. Sawai, AIST

The 869 depositsPossible 869 depositsNo deposits

869 Simulation 2011 inundation

The 869 Jogan EarthquakeA fault model

Satake et al. (2008)Namegaya et al. (2010)

19

Past tsunamis

Sanriku coast• Attacked by tsunamis in 1896 (22,000 casualties) , 1933 (3,000 casualties), and 1960 (Chilean eq., 142 casualties)

• The 2011 tsunami was similar heights with 1896

Sendai Plain• Historical and geological data recorded AD 869 tsunami• The AD 869 tsunami was very similar to 2011• Tsunami hazard maps or break waters were prepared for a more frequent (99%, M~8) tsunami

20

The 2011 and past Tsunami sources

Temporal and spatial distribution of slips

22Satake, Fujii, Harada and Namegaya (2013)Bull. Seism. Soc. Am. (2011 special issue)

Deep and Shallow Subfaults

Ver 8.0 (55 subfaults)

23

TotalAverage slip 9.5 m Moment 4.21 x 1022 Nm (Mw = 9.0)

Deep Subfaults (Jogan model type)

24

“Jogan” typeAverage slip 6.4 m Moment 2.1 x 1022 Nm (Mw = 8.8)

Shallow Subfaults (Tsunami Earthquake type)

25

Tsunami earthquake typeAverage slip 18.6 m Moment 2.1 x 1022 Nm (Mw = 8.8)

Comparison with Coastal Tsunami Heights

Delayed slip Instantaneousslip

simulation

55 subfaults 26

Ver 8.0

1896 type

869 type

Deep and Shallow Subfaults

Long wavelength

Large inundation in Sendai plain

Short wavelength and large peak

High tsunami on Sanriku coast

Outline


2. The Tsunami




Why wasn’t it forecasted?

Eastward rebound on March 11Westward motion in 1998‐2000

Coseismic slip(released strainduring earthquake)

Slip deficit(stored strain)

29GSI (2010, 2011)

Why wasn’t it forecasted?

Slip deficit(stored strain)

Yamanaka and Kikuchi (2004)

Large (M>7) earthquakes in 20th century

30

Westward motion in 1998‐2000

GSI (2010, 2011)

Earthquake Recurrence and Hazard Rate

Interval between eq.Earthquake Occurrence

periodic

Random (Poisson process)

Future (e.g., in next 30 years) probability can be computedfrom past recurrence data

Sanriku long‐term forecast

+3m, 24 m

+0.9 m, 5 m

‐0.7m, 15 m

Long term forecast by ERC

Northern Sanriku

Miyagi‐oki

Southern Sanriku

Ibaraki‐oki

Sanriku long‐term forecast

+3m, 24 m

+0.9 m, 5 m

‐0.7m, 15 m


Northern Sanriku

Miyagi‐oki

Southern Sanriku

Ibaraki‐oki

+3m, 24 m

+0.9 m, 5 m

‐0.7m, 15 m


Long‐term forecast of earthquakes

Iwate

Miyagi

Fuku‐shima

The 2011 type earthquake (Mw 8.4 – 9.0)4 events in last 2500 yrsonce in 600 yrs

Along Japan trenchTsunami earthquakes (Mt 8.6 – 9.0) 4 events in last 400 yrs(1611, 1677, 1896 and 2011)once in 103 yrs somewhere along Trench

Outer‐rise normal fault eq.1 event in 400 yrs (1933)once in 400 – 750 yrs somewhere

Revision (update) on November 2011

Long‐term forecast by Japanese government

Region TypeAs of January 1, 2011 As of January 1, 2012M Probability M Probability

Northern Sanriku

Interplate 8 0.5‐10 % 8 (Mt 8.2) 0.7‐10 %

Others 7.1‐7.6 90% 7.1‐7.6 90%Central Sanriku ‐ ‐ ‐ ‐

Miyagi‐okiInterplate 7.5 99% 7.4 unknownOthers 7.0‐7.3 60%

Southern Sanriku

Interplate 7.7 80‐90 % 7.9 0%

Others 7.2‐7.6 50%2011 Tohoku

typenot estimated Mw 8.4‐9.0 0%

Fukushima‐oki Interplate 7.4 (swarm) < 7 % 7.4 (swarm) 10%

Ibaraki‐okiInterplate 6.7‐7.2 > 90 % 6.7‐7.2 > 90%Others 6.9‐7.6 70%

Japan Trench "Tsunami eq." Mt 8.2 20% Mt 8.6‐9.0 30%Japan Trench Normal‐fault 8.2 4‐7 % 8.2 (Mt 8.3) 4‐7 %

Earthquake Research Committee (2009, 2011)

Supercycle of earthquakes

Seismologists assumed earthquake cycle (~35 years) from past records of two centuries and made forecast (99% in 30 years), but there seems to be a supercycle (~700 years) on top of it.

36

Long‐term forecast of earthquakes

Long‐term forecast • based on earthquake history in last few centuries• 99 % probability in next 30 years but smaller size (M~8)

March 11 earthquake was much larger (M=9.0)• GPS data suggested such slip deficit• Earthquake supercycle may exist

37

Outline


2. The Tsunami




Tsunami Assessment Method for Nuclear Power Plants in Japan (2002)

• Japan Society of Civil Engineers (2002)• Used as a standard guide for industry• English version published in 2006• Annex II in IAEA Safety Standards SSG‐18 (2011)

• Deterministic Method– The maximum and minimum water levels – Parametric study of fault parameters

Deterministic MethodParameter study by changing parameters

of the Standard Fault Modelin rational ranges

Fault Models for Scenario Earthquakes

Fault Model for the Design Tsunami

Standard Fault Model

LandTarget site

Sea

The design tsunami is the highest among scenario tsunamis at

the target site.

Direction of coastline

Target Site

The Design Tsunami

Scenario Tsunamis

Scenario Tsunamis

Tsun

ami height

Records of Historical tsunamis

JSCE (2002)

Standard fault model

Design tsunami for NPSDesign water levels

Fault model(s) based on historical tsunami

Parametric study andScenario tsunamis

Fukushima‐Daiichi NPS (TEPCO)

Fukushima‐1

Design height 6.1 m based on 1938 Shioya‐oki earthquakes (Mw 7.9)

Tsunami earthquakeSimilar to 1869 off Fukushimaestimated tsunami heights: 8.4 ‐15.7 m

The 869 Jogan‐type earthquakeestimated tsunami heights: 7.8 ‐ 9.2 m

However, probabilities of tsunami earthquakes and Jogan‐type earthquakes were considered to be low

Data source: TEPCO (2012)

Tsunami Damage at NPS

Onagawadesign tsunami height 13.6 msite level 14.8 m

Fukushima‐1design tsunami height 6.1 msite level 10‐13 m

Fukushima‐2design height 5.2 msite level 12 m

Tokaidesign height 4.9 msite level 8.0 m

13 m

16 m

15 m

5 m

2011 tsunami

Data source: IAEA (2011))

Tsunami Source ModelsTsunami sourcesFault models

Magnitude rangeProbability of earthquakesBranches for logic trees

Numerical Models for tsunami generation and propagationSeafloor deformation modelTsunami propagation modelBranches for logic trees

Logic trees and numerical simulation

Tsunami hazard curves

Tsunami heights for fragility curve

Probabilistic Tsunami Hazard Assessment

Atomic Energy Society of Japan 2011Implementation Standard Concerning the Tsunami Probabilistic Risk Assessment of

Nuclear Power Plants

Japan Society of Civil Engineers 2011

Annnaka et al. (2007, Pagepoh)Sakai et al. (2006, ICONE14)

Tsunami Source ModelsTsunami sourcesFault models

Magnitude rangeProbability of earthquakesBranches for logic trees

Numerical Models for tsunami generation and propagationSeafloor deformation modelTsunami propagation modelBranches for logic trees

Logic trees and numerical simulation

Tsunami hazard curves

Tsunami heights for fragility curve

Probabilistic Tsunami Hazard Assessment

Annaka et al. (2007)

Probabilistic Method

Aleatory (random) variabilityHazard curves

1896, 1611Tsunami Eq.

1677Tsunami Eq.

No Tsunami eq.

Epistemic (modeling) uncertainty Logic tree

Only at specific (past) locations

Anywhere along trench axis

年間超過確率

Tsunami Earthquakes

Annaka et al. (2007)

Conclusions

1. 2011 Tohoku earthquake was the largest (M~9) in Japan’s history2. Recently‐installed marine geophysical instruments revealed huge slip

near trench axis and recorded tsunami propagation toward coast3. Sanriku coast and Sendai plain experienced similar tsunamis in AD

869 (Jogan eq.) and in AD1869 (Sanriku tsunami eq.)4. The 2011 tsunami source was a combination of the Jogan‐type and

tsunami earthquake type 5. Long‐term forecast based on historical data estimated 99 %

probability but M~8 in Miygai‐oki6. The tsunami hazard assessments at NPS were made by a

deterministic method based on historical tsunamis. The probabilistic tsunami hazard assessment including the epistemic uncertainty was proposed but not implemented

The 2011 Tohoku Earthquake and Tsunami Hazard … 1896 Meiji Sanriku tsunami 1896 Meiji tsunami: 22,000 casualties (more than 2011 tsunami) M 7.2, Max tsunami height 38 m Weak shaking

Documents