comparisons of darfieldand christchurch ground...

Comparisons of Darfield and pChristchurch Ground Motions with

NGA W1 GMPEsNGA‐W1 GMPEs

David M BooreDavid M. BooreWorkshop on Update of Pacific Northwest

Portion of the U.S. National Seismic Hazard MapsPortion of the U.S. National Seismic Hazard Maps (NSHMs)

March 21‐22, University of Washington, Seattle

(http://www.geonet.org.nz/canterbury-quakes/)

Stations for which Rjb<= 200 km

-42o

-41o40'

ate:

2012

-03-

17

Cl C it t

43o

-42o40'

-42o20'

42

e ap_r

_le_

200.

draw

;D

a

Most class C sites to the north

Class C sites at greater distance may have smaller depth of sediments than

-43o40'

-43o20'

-43o

Latit

ude

chris

tchu

rch_

stat

ion_

ma northsediments than

those near Christchurch

-44o40'

-44o20'

-44o

Christchurch (M 6.1)NEHRP class DNEHRP class Chypocenter

orks

hop_

21-2

2mar

12\c

170o 171o 172o 173o 174o-45o

Longitude

e:C

:\pac

nw_w

o

1000 DarfieldChristchurchNEHRP B

NEHRP C 12-0

3-17

;Tim

e:16

:40

• Many sites with Vs30 = 275 and

500

m/s

)

NEHRP C

vs_r

jb.d

raw

;D

ate:

20660 m/s because values were estimated, not measured

• Distributions of Vs30 similar for both events

300

400

VS

30(m

NEHRP D

chur

ch_d

arfie

ld_v

s30_

v

• Most records with R<100 km are from class D sites

200

NEHRP E op_2

1-22

mar

12\c

hris

tc

1 10 100 1000

RJB (km)

:\pac

nw_w

orks

ho

Note: Some sites classified as “E” by Bradley & Cuprinovski (2011)

Comparison of Darfield and Christchurch Ground Motions

• Motions from both events at close distances are comparable

1 12-0

3-17

;Tim

e:16

:

• Apparently no or small site effect (but note difference in spatial locations for different class sites, so source effects could

1

(T=0

.2s)

(g)

2_ob

s.dr

aw;

Dat

e:20

1

compensate for site effects)

• Nonlinear soil response could have reduced motions for the larger event

0.1

ampe

dP

SA

urch

_dar

field

_psa

_t0.

2

Observed RotD50 reduced to

g

0.01

5%-d

NEHRP Class D (Darfield, M 7.0)NEHRP Class D (Christchurch, M 6.1)NEHRP Class C (Darfield, M 7.0)NEHRP Class C (Christchurch, M 6.1)

p_21

-22m

ar12

\chr

istc

huGMRotI50 using factors from Boore (2010); maximum effect is 4% at T=5 s.

In this and subsequent1 2 10 20 100 200

RJB (km)

\pac

nw_w

orks

hopIn this and subsequent

comparisons, the Y-axis spans three orders of magnitude.

Concentrate on H components, as the 2008 NGA GMPEs were only for H.

1

12-0

3-17

;Tim

e:16

:

0.1

(T=1

.0s)

(g)

0_ob

s.dr

aw;

Dat

e:20

1

• Christchurch (M 6.1) class D motions comparable or even larger than Darfield (M 7.0) at close distances

0 01am

ped

PS

A

urch

_dar

field

_psa

_t1.

0• Darfield class C motions greater than Christchurch motions (R>50 km)

0.015%

-dNEHRP Class D (Darfield, M 7.0)NEHRP Class D (Christchurch, M 6.1)NEHRP Class C (Darfield, M 7.0)NEHRP Class C (Christchurch, M 6.1)

p_21

-22m

ar12

\chr

istc

hu

1 2 10 20 100 200

RJB (km)

\pac

nw_w

orks

hop

0.1 12-0

3-17

;Tim

e:16

:

(T=5

.0s)

(g)

0_ob

s.dr

aw;

Dat

e:20

1

• Christchurch (M 6.1) class D motions now smaller than Darfield (M 7.0) at all distances

0.01

ampe

dP

SA

urch

_dar

field

_psa

_t5.

0• Apparent site effect small or not existent

0.0015%-d


p_21

-22m

ar12

\chr

istc

hu

1 2 10 20 100 200

RJB (km)

\pac

nw_w

orks

hop

Comparison of Observed Motions and Motions from NGA‐Motions and Motions from NGA

W1 GMPEs: Darfield

Darfield (M 7.0)

Dat

e:20

12-0

3-21

;

• Reddish GMPEs for class D

• Bluish GMPEs for class C

Don’t try to follow curves for individual1

(T=0

.2s)

(g)

mpe

s_bl

ue_r

ed.d

raw

;

• Overall comparison good

• Don’t try to follow curves for individual GMPEs

0.1

ampe

dPS

A

NEHRP Class DNEHRP Class CAS08, Vs30=275 m/sBA08 Vs30=275 m/s _o

bs_n

ga_r

_le_

200.

gm

Overall comparison good

• Apparent lack of site effect in data is consistent with GMPEs

0.01

5%-d BA08, Vs30=275 m/s

CB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s

ar12

\dar

field

_psa

_t0.

2_

1 2 10 20 100 200

RJB (km)

orks

hop_

21-2

2ma

1Darfield (M 7.0)

Dat

e:20

12-0

3-21

;

0.1(T=1

.0s)

(g)

mpe

s_bl

ue_r

ed.d

raw

;

• Note separation in observed class D values; GMPEs predictions between the two groups

0.1

ampe

dPS

A


bs_n

ga_r

_le_

200.

gm• Site effect in GMPEs, but not apparent in observations (but little overlap in distance range for class C and D)

0.015%

-d BA08, Vs30=275 m/sCB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s

ar12

\dar

field

_psa

_t1.

0_

1 2 10 20 100 200

RJB (km)

orks

hop_

21-2

2ma

Darfield (M 7.0)

Dat

e:20

12-0

3-21

;

0.1

(T=5

.0s)

(g)

mpe

s_bl

ue_r

ed.d

raw

;

• GMPEs underpredict observations (except class C at greater distances)

• Site effect in GMPEs, but not

0.01

ampe

dPS

A


bs_n

ga_r

_le_

200.

gmapparent in observations (but little overlap in distance range for class C and D)

0.001

5%-d BA08, Vs30=275 m/s


ar12

\dar

field

_psa

_t5.

0_

1 2 10 20 100 200

RJB (km)

orks

hop_

21-2

2ma

Comparison of Observed Motions and Motions from NGA‐Motions and Motions from NGA

W1 GMPEs: Christchurch

1

Christchurch (M 6.1)

;D

ate:

2012

-03-

2

1

A(T

=0.2

)(g)

gmpe

s_bl

ue_r

ed.d

raw

• Overall comparison good

• Apparent lack of site effect in data is consistent with GMPEs

0.1

dam

ped

PS

A

NEHRP Class DNEHRP Class CAS08, Vs30=275 m/sBA08 Vs30=275 m/s .2

_obs

_nga

_r_l

e_20

0.

0.01

5%-d BA08, Vs30=275 m/s


12\c

hris

tchu

rch_

psa_

t0

1 2 10 20 100 200

RJB (km)

ksho

p_21

-22m

ar1

1 Christchurch (M 6.1)

;D

ate:

2012

-03-

2

0.1

A(T

=1.0

)(g)

gmpe

s_bl

ue_r

ed.d

raw

• GMPEs tend to underpredict class D motions at close distances

• Site effect in GMPEs, but not

0 01dam

ped

PS

A


_obs

_nga

_r_l

e_20

0.apparent in observations (but little overlap in distance range for class C and D)

0.015%

-d BA08, Vs30=275 m/sCB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s

12\c

hris

tchu

rch_

psa_

t1

1 2 10 20 100 200

RJB (km)

ksho

p_21

-22m

ar1

0.1Christchurch (M 6.1)

;D

ate:

2012

-03-

2

0.01A(T

=5.0

)(g)

gmpe

s_bl

ue_r

ed.d

raw

• GMPEs severely underpredict observations at shorter distances

• Agreement better for greater 0.01

dam

ped

PS

A


_obs

_nga

_r_l

e_20

0.distances

• Site effect in GMPEs, but not apparent in observations (but little overlap in distance range for class C

0.0015%-d BA08, Vs30=275 m/s


12\c

hris

tchu

rch_

psa_

t5and D)

1 2 10 20 100 200

RJB (km)

ksho

p_21

-22m

ar1

Sensitivity of Predicted T=5 s PSA to sediment depth and toto sediment depth and to

magnitude

• Show AS08 and class D onlyShow AS08 and class D only

• Z1.0=1000 m from Bradley & Cuprinovski (2011)

• M 6 33 from Holden (2011) 0.1Christchurch (M 6.1)

2_6.

33.d

raw

;D

M 6.33 from Holden (2011)

• Use of these data in NGA-W2 without Z1.0 and larger M could result in biased results. T=

5.0)

(g)

ompa

re_z

sed.

add_

m6.

2

• Does Z1.0 vary spatially, with it being smaller for stations at greater distances? If so, this would help explain the

0.01

ampe

dPS

A(T

00.c

lass

_d_a

s_on

ly.c

o

p pdiscrepancy at greater distance. This could also be due to a difference in geometrical spreading due to lateral changes in crustal structure

0.0015%-d

aNEHRP Class DAS08, Vs30=275 m/s, Z1.0=480 mAS08, Vs30=275 m/s, Z1.0=1000 mAS08, as above, M 6.2AS08, as above, M 6.33

_t5.

0_ob

s_ng

a_r_

le_2

in crustal structure.

1 2 10 20 100 200

RJB (km)

r12\

chris

tchu

rch_

psa_

Vertical MotionVertical Motion

(Fry et al., 2011)

(Bradley & Cuprinovski, 2011)

Effects Producing Spatial Variability inEffects Producing Spatial Variability in Ground Motions

•Source: Radiation Pattern & Directivity•Path: volcanic vs sedimentsPath: volcanic vs sediments•Basin Waves•Sediment Depth•Shallow Site Response

•LinearN li•Nonlinear

Fault Normal and Fault Parallel Velocity Time Seri

00

HVSC,Fault Normal,158

00

CMHS,Fault Normal,158

00

PRPC,Fault Normal,158SHLC,Fault Normal,158

0 20 40 60 80

-100

050

10

Time (Sec)

Vel

ocity

(cm

/s)

HVSC F lt P ll l 68

0 20 40 60 80

-100

050

10

Time (Sec)

Vel

ocity

(cm

/s)

CMHS F lt P ll l 68

0 10 20 30 40 50

-100

050

10Time (Sec)

Vel

ocity

(cm

/s)

PRPC F lt P ll l 68

10 20 30 40

Time (Sec)

SHLC F lt P ll l 68 CMHSCMHSPRPCPRPCSHLCSHLC

HVSCHVSCSHLC, PRPC,CMHSSHLC, PRPC,CMHS0 20 40 60 80

-100

050

100

Time (Sec)

Velo

city

(cm

/s)

HVSC,Fault Parallel,68

0 20 40 60 80

-100

050

100

Time (Sec)

Velo

city

(cm

/s)

CMHS,Fault Parallel,68

0 10 20 30 40 50

-100

050

100

Time (Sec)

Velo

city

(cm

/s)

PRPC,Fault Parallel,68

10 20 30 40

Time (Sec)

SHLC,Fault Parallel,68 CMHSCMHS

LPCCLPCC

S C, C,C SS C, C,C S

5010

0

cm/s

)

LPCC,Fault Normal,158

65° Dip0 20 40 60 80

-100

0

Time (Sec)

Vel

ocity

(c

100

LPCC,Fault Parallel,68

0 20 40 60 80

-100

050

1

Time (Sec)

Velo

city

(cm

/s)• Drastic change in frequency

content going from sites on rock (HVSC & LPCC) to sites oc ( SC & CC) to s teson Quaternary sediments.

(from B. Chiou)

Evidence for Nonlinear Soil Response

0.5

1

g

NNBS

Tim

e:21

:09:

00

-0.5

0

as7

7wg

w;

Dat

e:20

12-0

3-17

;T

10 15 20 25-1

0

0.5g 2_

NN

BS

_V2A

.xls

.dra

w

-0.5

0

as1

3eg

011\

2011

0221

_235

142

10 15 20 25-1

0.5

1

g eala

nd_c

hris

tchu

rch_

2

-0.5

0

aup

g

File

:C:\n

ew_z

e

10 15 20 25-1

Time (s)

200

400

/s)

2001 M 6.8 Nisqually (SDS)

NS

-200

0

200

Acc

el(c

m/s

/

me:

08:0

9:29

20 30 40 50-400

200

400/s

) EW

Dat

e:20

12-0

3-18

;Tim

-200

0

Acc

el(c

m/s

/

_200

1\sd

s_un

filt.d

raw

;

(see Frankel et al., 2002, for details)

20 30 40 50-400

100

200

s/s)

UD File

:C:\n

isqu

ally

_

-100

0

Acc

el(c

m/s

20 30 40 50-200

Time (s)

Negative vertical accelerations are “clipped”. This may be due to a different nonlinear process than that producing the cusps shown in the previous figures.

(Fry et al., 2011)

p g

Conclusions• M 7.0 Darfield and M 6.1 Christchurch motions similar

for close distances, short periods• M 7.0 Darfield motions higher than M 6.1 Christchurch

motions for longer periods (as expected from the difference in magnitudes)

• Site response not too obvious, but this may be because of the different spatial distributions of the site classes (most close sites are class D)

• Observed motions influenced by many effects, including• Lateral changes in geology• Local linear and nonlinear site response

B i (?)• Basin waves (?)

• GMPEs are in reasonable agreement with observations for close distances, short periods

d di l i d i i• GMPEs underpredict longer period motions, using metadata in current NGA‐W2 flatfile

ENDEND

From B. Chiou, Source: GNS Science

Directivity Effect and Velocity Pulse

HVSCHVSC

SHLC, PRPC,CMHSSHLC, PRPC,CMHSLPCCLPCC

S C, C,C SS C, C,C S

65° Dip

Source: George Walker

(from B. Chiou)

0282 Christchurch, New Zealand

3.4

-43

)SHLC

PRPC

CMSH

-43.

6La

titud

e (° CMSH

HVSC

LPCC

-43.

8

172 4 172 6 172 8 173172.4 172.6 172.8 173Longitude (°)

(from B. Chiou)

Fault RuptureFault Rupture

• Reverse faulting on a buried faultReverse faulting on a buried fault• Assumed fault plane

St ik 68° (f USGS CMT)– Strike = 68° (from USGS CMT)– 65 ° dip, to the south

( )– Top of rupture is at 2 km depth (assumed)– Bottom of rupture is at 12 km (assumed)– Rupture length ~ 15 km (length of the aftershock zone).

(from B. Chiou)

300

2001 Nisqually, Washington, earthquake (M 6.8)

SDS (NS) filt d SDS filt d f 10 20 H

0

100

200

on(c

m/s

ec2 )

0

50

11:0

4:41

SDS (NS): unfiltered SDS: filtered from 10--20 Hz

• SDS within 200 m of SDN

-300

-200

-100

Acc

eler

ati

-50

filt_

filt.d

raw

;D

ate:

2003

-09-

17;

Tim

e:• liquefaction at SDS, not at SDN• Note cusps at SDS and increased

20 30 40 50 20 30 40 50

100

200

300

m/s

ec2 )

50

\rec_

proc

_stro

ng_m

otio

n\sd

s_sd

n_un

f

SDN: filtered from 10--20 HzSDN (NS): unfiltered

amplitude at high frequencies

200

-100

0

100

ccel

erat

ion

(cm

50

0

File

:C:\m

etu_

03\

20 30 40 50-300

-200

Time (sec)

Ac

20 30 40 50

-50

Time (sec)

(after Frankel et al., 2002)


d.dr

aw;

Dat

e:20

0 01(T=5

.0)(

g)

blue

_red

.com

pare

_zse

d

0.01

dam

ped

PS

A

NEHRP Class DNEHRP Class C ga

_r_l

e_20

0.gm

pes_

b

0.0015%-d

NEHRP Class CAS08, Vs30=275 m/s, Z1.0=480 mAS08, Vs30=275 m/s, Z1.0=1000 mCB08, Vs30=275 m/s, Z1.0=326 mCB08, Vs30=275 m/s, Z1.0=1000 mCY08, Vs30=275 m/s, Z2.5=2.5 kmCY08, Vs30=275 m/s, Z2.5=4.1 km hu

rch_

psa_

t5.0

_obs

_n

1 2 10 20 100 200

RJB (km)

2.5

1-22

mar

12\c

hris

tch


_m6.

2.dr

aw;

Dat

0 01(T=5

.0)(

g)

ed.c

ompa

re_z

sed.

add_

0.01

dam

ped

PS

A

NEHRP Class DNEHRP Class C _l

e_20

0.gm

pes_

blue

_re

0.0015%-d

NEHRP Class CAS08, Vs30=275 m/s, Z1.0=480 mAS08, Vs30=275 m/s, Z1.0=1000 mAS08, as above, M 6.2CB08, Vs30=275 m/s, Z1.0=326 mCB08, Vs30=275 m/s, Z1.0=1000 mCY08, Vs30=275 m/s, Z2.5=2.5 km _p

sa_t

5.0_

obs_

nga_

r_

1 2 10 20 100 200

RJB (km)

CY08, Vs30=275 m/s, Z2.5=4.1 km

mar

12\c

hris

tchu

rch_

-41o40'Darfield (M 7.0)

NEHRP class DNEHRP l C

2012

-03-

17;T

-42o40'

-42o20'

-42o NEHRP class Chypocenter

_r_l

e_20

0.dr

aw;

Dat

e:

Stations for which Rjb<= 200 km

-43o40'

-43o20'

-43o

Latit

ude

\dar

field

_sta

tion_

map

_

-44o40'

-44o20'

-44o

wor

ksho

p_21

-22m

ar12

\

169o 170o 171o 172o 173o 174o-45o

Longitude

le:C

:\pac

nw_w

1

03-1

7;Ti

me:

16:1

5

0.1) bs.d

raw

;D

ate:

2012

-0

PG

A(g

hurc

h_da

rfiel

d_pg

a_ob

0.01


op_2

1-22

mar

12\c

hris

tch

1 2 10 20 100 200

RJB (km)

C:\p

acnw

_wor

ksho

1

2-03

-17;

Tim

e:16

:

0.1(T

=2.0

s)(g

)

_obs

.dra

w;

Dat

e:20

12

0.01ampe

dP

SA

(

rch_

darfi

eld_

psa_

t2.0

_

5%-d

a


_21-

22m

ar12

\chr

istc

hur

1 2 10 20 100 2000.001

RJB (km)

\pac

nw_w

orks

hop_

100

03-1

7;Ti

me:

16:1

5

10

/s)

bs.d

raw

;D

ate:

2012

-0

PG

V(c

m/

hurc

h_da

rfiel

d_pg

v_ob

1


op_2

1-22

mar

12\c

hris

tch

1 2 10 20 100 200

RJB (km)

C:\p

acnw

_wor

ksho

Observed RotD50

1 Darfield (M 7.0)

te:2

012-

03-2

0;Ti

m

Observed RotD50 reduced to GMRotI50 using factors from Boore (2010); maximum effect is 4% at T=5 s. 0.1) s_

blue

_red

.dra

w;

Dat

at 5 s

In this and subsequent comparisons, the Y-axis spans three orders of magnitude.

PG

A(g

NEHRP Class DNEHRP Class CAS08, Vs30=275 m/s s_

nga_

r_le

_200

.gm

pes

Concentrate oh H components, as the 2008 NGA GMPEs were only for H.

0.01 BA08, Vs30=275 m/sCB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s

mar

12\d

arfie

ld_p

ga_o

bs

1 2 10 20 100 200

RJB (km)

,

_wor

ksho

p_21

-22m

1Darfield (M 7.0)

Dat

e:20

12-0

3-20

;

0 1(T=2

.0s)

(g)

pes_

blue

_red

.dra

w;

D

0.1

ampe

dP

SA

(

NEHRP Class DNEHRP Class CAS08, Vs30=275 m/s ob

s_ng

a_r_

le_2

00.g

mp

0.01

5%-d

a BA08, Vs30=275 m/sCB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s r1

2\da

rfiel

d_ps

a_t2

.0_o

1 2 10 20 100 200

RJB (km)

,

orks

hop_

21-2

2mar

100

Darfield (M 7.0)

te:2

012-

03-2

0;Ti

m

100

/s)

s_bl

ue_r

ed.d

raw

;D

at

10

PG

V(c

m/

NEHRP Class DNEHRP Class CAS08, Vs30=275 m/s s_

nga_

r_le

_200

.gm

pes

1

BA08, Vs30=275 m/sCB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s m

ar12

\dar

field

_pgv

_obs

1 2 10 20 100 200

RJB (km)

,

_wor

ksho

p_21

-22m


Dat

e:20

12-0

3-21

;

0.1) pes_

blue

_red

.dra

w;

D

PG

A(g


s_ng

a_r_

le_2

00.g

mp

0.01 BA08, Vs30=275 m/sCB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s r1

2\ch

ristc

hurc

h_pg

a_o

1 2 10 20 100 200

RJB (km)

,

orks

hop_

21-2

2mar

1Christchurch (M 6.1)

Dat

e:20

12-0

3-2

0.1(T

=2.0

)(g)

mpe

s_bl

ue_r

ed.d

raw

;

0.01

dam

ped

PS

A

NEHRP Class DNEHRP Class CAS08, Vs30=275 m/s 0_

obs_

nga_

r_le

_200

.g

5%-d BA08, Vs30=275 m/s

CB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s 2\

chris

tchu

rch_

psa_

t2.0

1 2 10 20 100 2000.001

RJB (km)

,

ksho

p_21

-22m

ar12


Dat

e:20

12-0

3-21

;

10

/s)

pes_

blue

_red

.dra

w;

D

1

PG

V(c

m/


s_ng

a_r_

le_2

00.g

mp

1 BA08, Vs30=275 m/sCB08, Vs30=275 m/sCY08, Vs30=275 m/sAS08, Vs30=660 m/sBA08, Vs30=660 m/sCB08, Vs30=660 m/sCY08, Vs30=660 m/s r1

2\ch

ristc

hurc

h_pg

v_o

1 2 10 20 100 200

RJB (km)

,

orks

hop_

21-2

2mar

(Fry et al., 2011)

comparisons of darfieldand christchurch ground...

Documents