Transcript Campbell – Hybrid-empirical Ground Motion Model for

Ground Motion Prediction
Equations for Eastern North
America
Gail M. Atkinson, UWO
David M. Boore, USGS
(BSSA, 2006)
Atkinson and Boore 2006 ENA
relations



Based on stochastic finite-fault model (of
Motazedian and Atkinson, 2005)
Key source parameter is stress drop –
determined from compilation of instrumental
and historical data
Attenuation model based on empirical
trilinear model of Atkinson, 2004 for ENA
Stochastic finite fault
model (Silva;
Beresnev and
Atkinson; Motazedian
and Atkinson, 2005)
Key features:
- Subsources are
Brune point sources
- long-period
controlled by
moment, short-period
controlled by stress
- Results
independent of
subfault size
Parameters needed to apply stochastic finitefault model




All parameters needed for stochastic point source model:
stress drop, attenuation with distance, site amplifications,
physical properties
Geometry of source (can assume fault plane based on
empirical relations such as Wells and Coppersmith on fault
length and width vs. M, in this case assumed to be 1/3 of
western fault area for a given M)
Direction of rupture propagation (assume random)
Slip distribution on fault (assume random)
High-frequency spectral level depends on
stress drop (as in point-source model)
Illustration of typical decay of spectral amplitudes for
m1=3.75. Line is trilinear shape fitted to the data. Transition
distances are r01=70 km, ro2 = 140 km. Slopes of geometric
attenuation: -1.3, +0.2, -0.5
Amplitudes decay faster than 1/R at R<70 km. This has
important implications for ENA ground motion relations.
ENA stress drops,
based on highfrequency spectral
level. Mean = 140
bars.
Comparison
of AB06
eqns to
M4.9 2005
Riviere du
Loup data
Comparison
of AB06
eqns to
several M4.5
events
Comparison of equations to data for M5.8 and M7.6
Saguenay and Bhuj events – note high PGA for
Saguenay, well above AB06 predictions for M5.8
Comparison of groundmotion equations of
this study (solid red
lines) for M 5.5 and 7.5,
with previous
predictions (Atkinson
and Boore, 1995, black),
and mean and standard
deviation of alternative
EPRI (2004) predictions
(blue), all for hard-rock
site conditions in ENA.
AB06 ground-motion equations
are given for 2 site conditions



Base condition for development and
comparison to ENA data is hard-rock
(vs30>2000 m/s)
We also perform the same set of simulations
for B/C boundary site conditions, for an
assumed B/C profile (from Frankel et al.,
1996)
Separate sets of equation coefficients for
each set of simulations
Significant Issues





Limited M > 4 ground motion database
Vertical vs. horizontal ground motion
Stress drop of ENA source spectrum
Near-source geometrical attenuation
Kappa of NEHRP B-C site profile