Uncertainty of anticipation of seismic intensities Uncertainty of - - PowerPoint PPT Presentation

uncertainty of anticipation of seismic intensities
SMART_READER_LITE
LIVE PREVIEW

Uncertainty of anticipation of seismic intensities Uncertainty of - - PowerPoint PPT Presentation

Uncertainty of anticipation of seismic intensities Uncertainty of anticipation of seismic intensities -A study of fluctuation of anticipated seismic intensities by the method of current JMA Earthquake Early Warning JMA intensity Mitsuyuki


slide-1
SLIDE 1

Uncertainty of anticipation of seismic intensities Uncertainty of anticipation of seismic intensities

  • A study of fluctuation of anticipated seismic intensities

by the method of current JMA Earthquake Early Warning –

Mitsuyuki HOSHIBA Kazuo OHTAKE Kazuhiro IWAKIRI

JMA intensity

Mitsuyuki HOSHIBA, Kazuo OHTAKE, Kazuhiro IWAKIRI (Meteorological Research Institute, Japan) Tamotsu AKETAGAWA (Japan Meteorological Agency) Hiromitsu NAKAMURA (National Research Institute for Earth Science and

  • tsu

U ( at o a esea c st tute o a t Sc e ce a d Disaster Prevention, Japan) Shunroku YAMAMOTO (Railway Technical Research Institute, Japan)

Apr 22 2009 EEW Workshop

  • Apr. 22, 2009. EEW Workshop

Anticipated Seismic Intensity=f (Source Factor, Path Factor, Site Factor) Source Factor : 1 scalar (for example; Magnitude) Path Factor : Attenuation relation At Present Path Factor : Attenuation relation Site Factor : 1 scalar Without consideration of spectrum contents

slide-2
SLIDE 2

JMA seismic Intensity

Accelerogram, Accelerogram, Ans (t)、 Aew (t), Aud (t) Filter Filter V(t)=

JMA intensity (T t l d ti f V(t)

( ) (Ans

2(t)+Aew 2(t)+Aud 2(t))1/2

JMA intensity is evaluated from 3 comp. accelerograms (Total duration of V(t) > Va) is 0.3s g

I=2 log10(Va)+0.94

Approximate Relation between JMA Intensity scale and Modified Mercalli scale JMA pp y 1 2 3 4 5 L 5 U 6 L 6 U 7 Modified Mercalli 1 2 3 4 5 6 7 8 9 10, 11,12

slide-3
SLIDE 3

Method for anticipation of Seismic Intensity used in JMA Earthquake Early Warning used in JMA Earthquake Early Warning

1. Estimation of PGV from attenuation relation (Si and Midorikawa, 1999) using Magnitude, Hypo. Dist. and Depth 2. (PGV at surface)=(PGV at basement)×site factor

  • 3. JMA intensity=2.68+1.72log(PGV at surface) ・・・

Mid ik t l 1999 Midorikawa et al, 1999

3

slide-4
SLIDE 4

According to this idea

Two Earthquakes occurred at the same place with the same magnitude

According to this idea,

3.1 3.4 4.2 3.1 3.4 4.2

At each site, Intensity

4.0 4.0

・paths are the same ・Site factors are the same

Earthquake

M4.8 M4.8

At each site, Intensities of the two earthquakes are expected to be the same!

R ll ? N ? Really? or Not ?

slide-5
SLIDE 5

In actual observation, even with the same magnitude … From HP of JMA Even if the same Intensities are not Even if the same magnitude, Intensities are not always the same

slide-6
SLIDE 6

3 4 3 2 Point B Point C Point A 3.1 3.4 4.2 4.0 3.3 3.2 4.2 4.3 Intensity Point D Point A Earthquake M4.8 M4.8

How large fluctuation?

Point A 3.1 - 3.3 = -0.2 Point B 3.4 - 3.2 = 0.2 Point C 4.2 - 4.2 = 0.0 Point D 4.0 - 4.3 = -0.3 ・・・

Investigated the distribution of fluctuation

slide-7
SLIDE 7

Hypocenter and magnitude are from JMA unified catalogue

Data

y g g Intensity observation by JMA, municipalities and NIED May, 1996 – July, 2007 M3.5~5.5

Select of the earthquake pair

Distance between those epicenters is less than 5km Difference of focal depth is less than 5km Having the same magnitude

Select of Intensity data

Exclude the range of hypocentral distance at which seismic intensity is Exclude the range of (distance between hypocenters×10)

Select of Intensity data

  • bserved to be less than

0.6 hypocenters×10)

Intensity Intensity

Hypocentral Distance (km)

slide-8
SLIDE 8

M4 M5 00-20km 20-40

・Same magnitude ・distance is less than 5km Focal depth difference is less than 5km

M5 M6 20-40 40-60 60-80 80-

N=100

Focal depth difference is less than 5km ・Intensities are observed commonly more than 10 stations

slide-9
SLIDE 9

① ② ①-② (Intensity difference ) RMS: 0.50 Distribution

  • f Intensities
  • f Intensities

I t it f At this station, Inteisity of

Extent of uncertainty even when the earthquake occurred adjacently with the same magnitude

Intensity of earthquake ① is larger than ② by 0.6 Inteisity of earthquake ② is larger than ① by 1.7

slide-10
SLIDE 10

① ② ①-② Distribution of Intensity differnce RMS: 0.80

slide-11
SLIDE 11

・Same magnitude ・distance is less than 5km

Histogram of RMS of

Focal depth difference is less than 5km ・Intensities are observed commonly more than 10 stations

N=100 Intensity Difference Average of RMS: 0.41

Extent of uncertainty of Extent of uncertainty of intensity even when the magnitude, path and site are the same are the same

slide-12
SLIDE 12

Two Earthquakes occurred at the same place with the same magnitude

i i 3.1 3.4 4.2 4 0 3.6 3.9 4.7 4 5 Seismic Intensity 4.0 4.5

At each site,

地震

・paths are the same ・Site factors are the same At each site Intensities of the two earthquakes are expected to Difference of At each site, Intensities of the two earthquakes are expected to be the same!

Really? or Not ?

slide-13
SLIDE 13

AVE.=0.49 RMS=0.80 No._=230 Removing Inter‐event Shift by 0.49 Inter event residual AVE.=0.00 RMS=0.64

  • No. =230

Estimate the RMS for each pair of _ p earthquake

slide-14
SLIDE 14

Magnitude is used Magnitude is used as source factor

Average of RMS:0.41 → 0.33 g

Average of Intensity difference is used as difference is used as source factor

slide-15
SLIDE 15

Two Seismic Intensity meters located adjacently

Earthquake 2 Earthquake 3 5.1 2.2 5.3 2.4 Earthuake 1 Earthquake 2 Earthquake 3 Earthquake 4 3.1 3.2 3.3 3.4 Intensity When Hypocentral distance is much larger than the distance I t it

For each eathquake

between the 2 intensity meters Intensity meter A Intensity meter B

For each eathquake

・Source factor is same ・path is same For each earthquake, difference of intensity is expected to be h the same

Really? or Not ?

slide-16
SLIDE 16

Hypocenter and magnitude are from JMA unified catalogue

Data

y g g Intensity observation by JMA, municipalities and NIED May, 1996 – July, 2007 M3.5~5.5

Select of the pair of seismic intesity meter

Distance between those inteisity meters is less than 5km Distance between those inteisity meters is less than 5km

Select of Intensity data

Exclude the range of hypocentral distance at which seismic intensity is Exclude the range of (distance between hypocenters×10)

Select of Intensity data

  • bserved to be less than

0.6 hypocenters×10)

Intensity Intensity

Hypocentral Distance (km)

slide-17
SLIDE 17

Two Intensity meters meters Distance between intensity meters : intensity meters : 0.9km Average:‐0.15 Diff f it Difference of site factor Standard d i i deviation :0.52 Fluctuation of seismic intensity Intensity meter A is larger than B by 0.9 Intensity meter B is larger than A by 1.0

slide-18
SLIDE 18

Distance Distance between Intensity meters :1.0km

Average:‐0.63 Difference of site factor factor Standard deviation :0.40 Fluctuation of Fluctuation of seismic intensity

slide-19
SLIDE 19

・distance is less than 5km ・Intensities are observed commonly more than 10 stations Histogram of standard N=791 Histogram of standard deviation of intensity difference Average of standard deviation:0.29

E t t f t i t Extent of uncertainty even when the sites located adjacently Standard Deviation

slide-20
SLIDE 20

Extent of uncertainty in anticipation of seismic intensity seismic intensity

Uncertainty in anticipation of seismic intensity when Source factor is represented by 1 scalar Uncertainty (RMS) in case of same magnitude, same path, and same site

0.41/√2 = 0.29

For source factor, Seismic intensity is

0.33/√2 = 0.23

used instead of magnitude Uncertainty in case of same source

0 29/√2 0 21

Uncertainty in anticipation of seismic intensity when Site factor is represented by 1 scalar Uncertainty in case of same source, same path, and same site

0.29/√2 = 0.21

In addition to these uncertainty due to the In addition to these, uncertainty due to the attenuation relation should be considered