You are on page 1of 32

BC Hydro (2011) Subduction GMPE and Comparison

with Data from Recent Subduction Zone Earthquakes

Kofi O Addo, Ph.D., P. Eng.


Nov 4, 2011

COSMOS
‘PROXIMITY’ to the CSZ

2
‘AGE’ of the Facilities

16
Rockfill or Timber
14
(Concrete/Earthfill)
PLETED

Earthfill
12
Deterioration
F DAMS COMP

C
Concrete
10 General Ageing
8

6
NUMBER OF

4
2
0
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
YEAR
The PSHA Model is an ‘asset’
Dam safety reviews and
Continual management system
surveillance audits

Does the dam


meet all Dam Safety Investigations
No or ?
performance •Assessments against “good
and safety practice”, level of risk
requirements?
Yes

Is the dam
Yes safe enough?
g
Operate normally
and maintain No
emergency
preparedness Improve dam through
Capital Projects or
Issue is modify operations or
completed
in
decommission
database
Approach
SSHAC Level 3 Study for Ground Motion

Evaluated available models


ƒ Limited to M8.5, but needed M9.2 for Cascadia
ƒ Computed
p g
ground motions have a very
y large
g range
g
ƒ Models worked well for some cases, but not for others
Decided to develop a new model with epistemic uncertainty – several
alternative versions

Strong Motion Data


ƒ Amalgamated available strong motion data from existing models
ƒ Youngs et al (1997), Atkinson & Boore (2003), Zhao et al (2006)
ƒ Lin and Lee (2008), Macias & Atkinson (2009)
ƒ Added more recent data
Datasets

292 Earthquakes
((9946 record p
pairs

163 Interface
(3557 record pairs

129 Intraslab
(6389 record pairs
Slab Dataset
Issue: truncation of
small amplitude data
Final Dataset
Magnitude Scaling (R=100 km)
Hypocentral Depth Scaling
Forearc or Backarc?
Forearc or Backarc?

12
Model Parameters

Source
ƒ Moment Magnitude (Mw)
ƒ Event type (intraslab or interface)
ƒ Hypocentral depth
Path
ƒ Rupture Distance (interface)
ƒ Hypocentral Distance (intraslab)
ƒ Forearc
o ea c / Backarc
ac a c
Site
ƒ vs30
Model Applicability
Median
ƒ Ave horizontal spectral acceleration: 0.01 – 10 sec

Range of applicability
ƒ Intraslab:
ƒ 5.0
5 0 < Mw < 8
8.0
0
ƒ 40 < Rhypo < 300 km
ƒ 35 < Depth < 150 km
ƒ Interface:
ƒ 6.5 < Mw < 9.0
ƒ 20 < Rrup < 300 km
ƒ 10 < Depth < 50 km
Sigma
ƒ ‘Traditional’ sigma
ƒ Single-station sigma

Correlations of epsilons (see Al-Atik’s presentation)


ƒ Within-event, Between-event, total
Regional Differences
Regional Differences
Regional Differences
Regional Differences
Recent Earthquakes
Since developing the model, two large interface subduction zone earthquakes
have occurred:

2010 Maule (Chile) Mw8.8 Earthquake


2011 Tohoku (Japan) Mw9.0 Earthquake

How do model predictions compare with data from these earthquakes?


Mw8.8 Maule (Chile) Earthquake: PGA
Mw8.8 Maule ((Chile)) Earthquake
q ((T=1.0s))
2011 Tohoku (Japan) Mw9.0 Earthquake
PGA Attenuation and BC Hydro Model

Forearc Backarc
T=1.0s
0s Attenuation
tte uat o and
a d BC
C Hydro
yd o Model
ode
Forearc
Backarc
BC Hydro Model, Event Terms: PGA
BC Hydro Model Event Terms: T=1 sec
Event
Terms

T=3
T 3.0
0 sec

PGA
Magnitude Scaling (R=100 km)
Break in Magnitude Scaling

Period (sec) Original Model Revised for


Tohoku & Maule
PGA 7.8 8.0
0.3 7.8 8.0
0.5 7.8 7.9
1.0 7.8 7.8
20
2.0 78
7.8 77
7.7
3.0 7.8 7.6
Summary

BC Hydro Subduction GMPE


ƒ Global model
model, updated for recent megathrust earthquake
ƒ Regional variations in distance attenuation not considered
ƒ Different slope in Japanese data
ƒ Regional variations in median GM is about ± 0.2 LN units
ƒ Key features
ƒ Applicable to M9 earthquakes
ƒ Depth dependence strong for intraslab, but not for interface
ƒ Difference between forearc and backarc
Availability of Model?
ƒ Journal Paper by Dec 15, 2011
ƒ Model available after paper is submitted
Recognition

Ground Motion Team


Norm Abrahamson
Bob Youngs
Nick Gregor
g
Walt Silva
Gail Atkinson
Kofi Addo

Peer Review Panel


Carl JJ. Stepp
Ken Campbell
Kevin Coppersmith

You might also like