1

Overview of Active Fault Researchat Geological Survey of Japan

Yasuo. AWATA and Toshikazu. YOSHIOKA

Active Fault Research Center,Geological Survey of Japan, AIST

UJNR, November 10, 2006, Tokushima

2

Main Subjects on Active Fault Research

Paleoseismological investigation

Segmentation and scaling relations

Prediction of earthquake magnitude

3

Paleoseismological Investigation

4

Active Fault Research Project in Japan, 1st Stage

■1995 to 2004 FY

5

Active Fault Research Project in Japan, 1st Stage

■1995 to 2004 FY■98 Active fault regions■ Investigated mainly by

local govern. and GSJ■Evaluated by ERC, HERP

by local governmentsby GSJ

6

Probabilistic Shaking Map for Coming 30 Years( ERC, 2005 )

Active faultsSubduction boundary

Other sources

7

Active Fault Research Project in Japan, 2nd Stage

■From 2005 FY

■18 faults re-investigated by GSJ, in 2005 - 2006

40 fault regions will bere-investigated

Improved segmentation and scaling relations will be used ?

by Local Govern.by GSJby GSJ’s budget

(from 2003)contract with MEXT

(2005 and 2006)

Nagamachi-Rifu F

Itoigawa-Shizuoka F

Tachikawa F

Kego F

8

Paleoseismicty of an Active Fault, Which Triggered during the 2004 Mid-Niigata Earthq.

Triggered surface rupture, with a slip of 0.2 m

Characteristic two faulting eventseach with a slip > 1.5 m

9

Prediction of “Most-Likely Earthquake-Segment”by ERC

5-km-thresould (Matsuda, 1990)Scaling relation between total rupture length and slip145 :Most likely earthquake segments12 :further segmented based on paloseismicity alone

10

Fault Segmentationand

New Scaling Relations

11

Behavioral Segment & Earthquake Segment

Time

Estimate the extent of behavior segment ?Probabilistic prediction of earthquake segment ?

12

Segmentation of 15 Surface Ruptures in Japan

Paleoseismicity, Rupture process, GeometrySegment length <= 35 kmJog and Gap >= 2 km

13

Scaling Relation : for Behavioral Segments

13 segments / in 6 ruptures

y = 0.1965xR2 = 0.905

0

5

10

15

0 20 40 60 80

Behavioral Segment Length (km)

strike-slip

Dmax =

2.0 x

10e-4

L

Japan (strike-slip)

14

Scaling Relation : for Behavioral Segments

y = 0.169x

y = 0.2047xy = 0.2028x

0

5

10

15

0 20 40 60 80

Behavioral Segment Length (km)

strike-slipreversenormal

World (all types)

15

Scaling Relation : for Behavioral Segments

y = 0.1965xR2 = 0.905

0

5

10

15

0 20 40 60 80

Behavioral Segment Length (km)

strike-slip

y = 0.169x

y = 0.2047xy = 0.2028x

0

5

10

15

0 20 40 60 80

Behavioral Segment Length (km)

strike-slipreversenormal

Dmax =

2.0 x

10e-4

L

Japan (strike-slip) World (all types)

16

Criteria for Behavioral Segment

Geometry :Jog & Gap >= 2 km :Bend >=20 deg.

Paleoseismicity :Timing of faulting event:Recurrence interval

17

Behavioral Segments - Fault Length

km431 behavioral segments

( Length >= 10 km ) ( Slip rate >= 0.1 mm/y )

18

Behavioral Segments - Fault Length

431 behavioral segments ( Length >= 10 km ) ( Slip rate >= 0.1 mm/y )

145 prob. earthq. segments.(by ERC, 2005)

290 behavioral segments

19

Behavioral Segments - Fault Length

Mostly <= 40 km

20

Fault Length v.s. Slip per Event

Dave. = 1.2 x 10e-4 L = ~ 0.6 Dmax

21

Prediction ofFuture Earthquake Segments

22

Prediction of Most Likely Earthquake Segments

Application of 5-km-thresould ( Matsuda, 1990 )

431 b-segments → 256 earthquake-segments

23

Scaling Relation : Between B- & E-Segments

Predicted

24

Scaling Relation : Between B- & E-Segments

Predicted

25

Scaling Relation : Between B- & E-Segments

Predicted

26

Scaling Relation : for Historical RupturesTotal Length v.s. Maximum Slip

0

2

4

6

8

10

12

14

0 100 200 300 400 500

Total Length of Surface Rupture (km)

Data from Wells & Coppersmith(1994) and others after 1994

27

Summary

We have been contributingto the Paleoseismological investigation.

We are workingon fault segmentation and new scaling relations

We have to attack to the problemson the probabilistic prediction of future earthquakes

( geology, mechanics of earthquake & faulting )

28

“Active Fault Database” at AFRC Website( http://www.aist.go.jp/RIODB/activefault/ )

500 segmentswith fault strand mapParameters of faultingSite information on long-term slip rate and paleoseismicity

Tokushima●

29

“Active Fault Database” at AFRC Website( http://www.aist.go.jp/RIODB/activefault/ )

Segmentation of MTLStrike N 80 EDip 45 NLength 52 kmSense of Faulting Right-lateral Upthrown Side North-sideSlip Rate 4 m/kySlip per Event 6.9 mRecurrence Interval 1.7 ky

AD 1500 to 1868 Age of Last FaultingAD 1596 (documented)

Elapsed-time Rate 0.18ca. 0 % by BPTRupture Probability

in Next 30 years ca. 2 % by Poisson

Chichio Segment

Tokushima●

302005 Kashmir Earthquake, PAK

Thank You for Your Attention