multi-scale simulation of tsunami from tohoku earthquake and diffracted tsunami waves in hokkaido...

Post on 20-Dec-2015

215 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Multi-scale Simulation of tsunami from Tohoku

Earthquake and Diffracted

Tsunami Waves in Hokkaido

ZHU AIYU, Zhang DongningInstitute of Geophysics, China Earthquake AdministrationYuen Dave A.University of MinnisotaSong ShenyiComputer Network Information Center, China Academy of Science

Outlines Governing equations and numerical

methods

Numerical simulation

Tsunami diffraction

Generation Propagation Runup Inundation

Governing equations

2D nonlinear shallow water equations

Numerical methods Finite volume method (clawpack)

fy

fx

Sy

bghghhv

yhuv

xhv

t

Sx

bghhuv

yghhu

xhu

t

hvy

huxt

h

)2

1()()(

)()2

1()(

0)()(

22

22

3/4222

3/4222

hvuvnS

hvuunS

fx

fx

bottom frictional terms

Computing Domain longitude from 135° to 160°E latitude from 30 ° to 45°N DART 21401, DART 21413, DART 21418 and

DART 21419

http://www.ndbc.noaa.gov/dart.shtml

Generation

length L=510km width W=150km the fault dips δ=14˚ rakes λ=81˚ strikes θ=193˚ average slip D=20m

Figure 1 The tsunami source model based on USGS

The first source model

Figure 5 Offshore DART buoys wave height records versus computed results of the F1 model

Figure 3 (a) Surface projection of the slip distribution superimposed on ETOPO2 (http://www.geol.ucsb.edu/faculty/ji/big_earthquakes/2011/03/0311_v3/Honshu.html); (b) The tsunami source model based on UCSB

The second source model

consists in a 190 sub-faults system

(a) (b)

Figure 6 Offshore DART buoys wave height records versus computed results of the F2 model

finite element method,

uses the GPS data for constrain condition

Figure 3 The tsunami source model of GPS

The third source model

Figure 7 Offshore DART buoys wave height records versus computed results of the F3 model

Which is the best source model ? The arrival times of the first wave, the results of F2

and F3 gives have a generally better agreement with the observed data, but not for case of F1 that is much slower than the observation.

The amplitude of the first wave, all of the model have there bad side and good side

It is indeed difficult to distinguish which one is the best

correlation coefficient

Table 1 the correlation relation of the different models

ModelsBuoy gauge

Correlation coefficient

Average correlation strength

F1 21401 -0.7806 0.9226

21413 -0.9334

21418 0.9975

21419 0.9786

F2 21401 -0.8497 0.95602

21413 -0.9404

21418 0.9962

21419 0.9979

F3 21401 -0.4821 0.7902

21413 -0.6827

21418 1.0

21419 1.0

DD

EEE

DD

Covxy

)(),(

The second source model is the best!

Grid is 1800×3000PropagationVisualized by Avizo

Characters: (1) The tsunami wave is split into two waves, one is towards to the northeast coast of Japan, Another wave faces to the Pacific Ocean (2) tsunami wave is that first arriving wave is negative and then the positive wave which agrees with the observed data.

Table 2 Maximum run-up values for calculated and observed values

LocationCoordinates(Lat,. Long.)

Computational height (m)

Observed height (m)

Shimakita (41.3636, 141.2306) 2.2838 2-3

Hachinohe* (40.5317,141.5278) 7.4699 2.7 or more

Miyako* (39.6436,141.9753) 15.6069 8.5 or more

Ofunato* (39.0194,141.7536) 20.4789 8.0 or more

Onahama (36.9369,140.8919) 3.9363 3.5- 4

Ohoshigyoko (35.7444,140.8583) 3.3574 3-4

Mera (34.9189,139.8247) 1.4855 1-2

Chiba (35.5681,140.0456) 0.5038 0-1

Okada (34.7894,139.3914) 0.5864 0-1

runup

http://www1.kaiho.mlit.go.jp/KANKYO/TIDE/real_time_tide/sel/index_e.htm the notation * means the maximum height of tsunami cannot be retrieved so far due to the troubles, and actual maximum height might be higher according to the observed data

Adaptive mesh refinement method

Inundation

Tohoku tsunami diffraction

movie

Add the Gaussian noise to the topography

Topography

Topography add Gauss noise

movies

No flooding phenomenon at Hokkaido Island ?

Possible reason The earthquake source is not exactly correct The shallow water equations are not a perfe

ct model for the wave motion, The 1-mintute resolution etopo ocean bathy

metry is maybe not fine enough The Gaussian noise maybe not appropriate

Thank you

Diffraction

top related