Next Generation Attenuation

(NGA) Projects

Yousef Bozorgnia, Ph.D., P.E.

Executive Director,
Pacific Earthquake Engineering Research Center (PEER),
University of California, Berkeley
NGA= Next Generation “Attenuation” Relation
 Three separate NGA projects:
 For active crustal EQs (like California, Middle
East, Japan, Taiwan,…): NGA-West
 For subduction EQs (like Pacific Northwest
and northern California, Japan, Chile, Peru,…):
NGA-Sub
 Stable continental regions (like central &
eastern US, a good portion of Europe, South
Africa,…): NGA-East
NGA-West1 (Original NGA Project)

 Initiated October 2003

 PEER compiled a very comprehensive
database of ground motions recorded in
shallow crustal earthquakes in active
tectonic regions
 Numerous supporting research studies
were also carried out
NGA-West1
 In 2008, NGA ground motion prediction
equations (GMPEs) were finalized
 USGS adopted the NGA-West1 GMPEs for
the US National Seismic Hazard Maps
 NGA-West2 is a follow-up of NGA-
West1
Sponsors of NGA-West2

 Supports of the
sponsors are
appreciated
NGA-West2 Sub-Projects
Update worldwide database
Update worldwide database
Update worldwide database
Update worldwide database

NGA-West2 database includes over

21,000 three-component recordings…
(over 63,000 records)
From NGA-West1 to NGA-West2 the size of
database was increased by a factor of 6
Examples of data added to NGA-West2
database

Earthquake Name* Year M N Rec Rrup Range (km)

Tottori, Japan 2000 6.61 414 1-333
Niigata, Japan 2004 6.63 530 8-300
Chuetsu-oki, Japan 2007 6.8 616 10-300
Iwate, Japan 2008 6.9 367 5-280
El Mayor-Cucapah, CA 2010 7.2 238 11-240
Darfield, New Zealand 2010 7 114 1-540
Christchurch, New Zealand 2011 6.1 104 2-440
Wenchuan, China 2008 7.9 263 1-1500
L'Aquila, Italy 2009 6.3 48 5-230
*subset of added events
 Many NGA-West1 records were re-processed to
widen the useable bandwidth

 Number of
reliable long-
period
records was
increased by
more than a
factor of 2.5

Courtesy: Tim Ancheta

We added small magnitude data
 Database has 10,706 recordings associated with
magnitude ≤ 4.5
 Motivation:
 NGA-West1 models over-predicted motions
for small magnitude
 In the future, we can analyze multiple events
recorded at same site to characterize the site
variability (single-station Sigma)
 In the regions that have mainly small
magnitude data, they can compare NGA with
their data
 Magnitude scaling at small
magnitude

PGA SA(1.0s)
0 0
10 10

10-1
Acceleration (g)

10-1

10-2

10-2
Rrup = 0 km 10-3
Rrup = 10 km
Rrup = 50 km
Rrup = 200 km

10-34.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0

-4
8.5 10 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5
Moment Magnitude Moment Magnitude
 VS30 distribution in database

 Measurements versus inferred values

(estimated by various methods such as
slop, geology,…)

Courtesy:
J. Stewart, E. Seyhan
 VS30 distribution in database

 Measurements versus inferred values

(estimated by various methods such as
slop, geology,…)

Courtesy:
J. Stewart, E. Seyhan 16
Database “flatfiles”

 It has “data” (PGA, PGV, RotD50) and

 “Metadata”= Data about data
Database “flatfiles”
 Flatfiles are publically available at:
http://peer.berkeley.edu/ngawest2/databases/
 These are the files we used to develop
GMPEs
 The flatfiles have response spectra
 There are 11 flatfiles available for 11
damping values
Public availability of NGA-West2
time series
 Actual time series (“time-histories”) will
be downloaded from PEER web site
 Beta version of the online tool to select
and download time series is being
tested by a small group of users to test
and find errors, typos
 Widespread version of the online
database will be available in few weeks
NGA-West2 GMPEs for horizontal motion

 Using the latest database, and

 Using supporting research on:
 Directivity of ground motion
 HW/FW model using simulations data
 Update of nonlinear soil response
 New classification of “main shock” vs
“aftershocks”
 …
NGA-West2 GMPEs for horizontal motion

 Five teams of GMPE developers worked

interactively and developed five GMPEs
for horizontal components
 Abrahamson-Silva-Kamai (ASK team)
 Boore-Stewart-Seyhan-Atkinson (BSSA)
 Campbell-Bozorgnia (CB)
 Chiou-Youngs (CY)
 Idriss (I)
NGA-West2 GMPEs for horizontal motion
 Scope of horizontal GMPEs:
 RotD50 as the main intensity measure
 PGA, PGV, and spectral periods at least at 21
periods: 0.01, 0.02, 0.03, 0.05, 0.075, 0.1,
0.15, 0.2, 0.25, 0.3, 0.4, 0.5, 0.75, 1, 1.5, 2,
3, 4, 5, 7.5 and 10 sec
 No GMPE for PGD
NGA-West2 GMPEs for horizontal motion
 Scope of horizontal GMPEs:
 Applicable magnitude range: M ≤ 8.5 for strike-
slip; M ≤ 8 for Reverse; M ≤ 7.5 for Normal
faults
 Applicable distance range: 0-300 km
 If a GMPE developer team chooses different
ranges, they have to provide reasoning
 GMPE developer teams can select a sub-set of
the database, and provide reasoning for the
selection
Example of comparison of horizontal
GMPEs
Example of comparison of horizontal
GMPEs
RJB=30, Strike-Slip, VS30=760
1

0.1

0.01
ASK
BSSA
0.001 CB
CY
I
0.0001
3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
M
Example of changes since 2008

Reduced
conservatism at
small-moderate
magnitude
Magnitude “saturation” at short periods is now
a common feature

Magnitude Saturation
Status of NGA-West2 GMPEs
for Horizontal Motion
Status of Horizontal GMPEs
 PEER reports have been published and
available free of charge to download
 Horizontal GMPEs were reviewed and
adopted by the USGS for the 2014
National Seismic Hazard Maps
 Major impact on seismic design and
evaluation of all facilities in the western US

29
Status of Horizontal GMPEs
 A collection of papers have been
submitted and accepted by journal of
Earthquake Spectra
 Accepted papers have been posted and can
be downloaded from EERI web site
 Hard copy papers will be printed in
August 2014 issue of EQS

30
Excel file of GMPEs
 There is an Excel file of the five GMPEs
that you enter input parameters and it
gives you weighted average of the
prediction. Download it free of charge
from: http://peer.berkeley.edu/ngawest2/databases/

31
GMPEs for Vertical
Ground Motion
NGA-West2 vertical GMPEs

 Four teams have completed development

of GMPEs for vertical component:
 Gulerce-Kamai-Abrahamson-Silva (GKAS)
 Stewart-Seyhan-Boore-Atkinson (SSBA)
 Bozorgnia-Campbell (BC)
 Chiou-Youngs (CY)
Example of V/H spectra ratio

34
Status of NGA-West2 Vertical GMPEs
 A single PEER report, with different
chapters documenting different vertical
GMPEs, was published

 Journal papers to be submitted to

Earthquake Spectra
Damping scaling of
response spectra
Damping scaling of response spectra

 Scale GMPEs
for damping
other than 5%:
0.5% to 30%

Ref: Rezaeian et al. (2012)

Damping scaling of response spectra
Damping Scaling Factors
 A model was
developed to
scale spectral
ordinates as a
function of
 Damping ratio
 Magnitude
 Distance
Ref: Rezaeian et al. (2012)

Duration was important, but took care of the

effects approximately through magnitude and
distance
Different models for horizontal and vertical
components
Directivity
 Most NGA-West1&2 35

models have not 1992 Landers, CA, EQ

explicitly included
directivity of ground
Forward directivity
region

motion Lucerne Valley

136 cm/sec
34.5
 Five directivity models

Latitude
have been developed in
NGA-West2
Rupture
propagation Epicenter

 Wide-band and narrow- Joshua Tree

band models 34 43 cm/sec

This effort will continue

Backward directivity
 region

in 2014 -117 -116.5 -116

Longitude
 Directionality (Polarization)
 NGA models are for
“average” horizontal
components
 Developed max and
min rotated spectra,
as a function of mag,
distance,…
 Examine relationship
of max/min spectra
with RotD50 (50
percentile) spectra
Ref: Boore (2010)
 Directionality key findings
 Only for the cases when the site is
within 5 km of the fault and spectral
period is longer than 0.5 sec, is the
orientation of maximum rotated spectral
ordinate is more likely to be closer to the
strike-normal than the strike-parallel
direction
 Beyond these ranges, the angle of the
maximum rotated motion is random
Ref: Shahi and Baker, PEER report 2013/10
Site Response

 NGA-West1 site amplification factors

were inconsistent with NEHRP site
amplification factors
 Goal: To make NEHRP and NGA site
amplifications more consistent
 Proposed, and got major approval,
to revise NEHRP factors
 NGA-West2 Status

 All tasks and sub-projects completed on

the proposed schedule
 13 PEER reports have been published and
can be downloaded free of charge from:
http://peer.berkeley.edu/news/2013/05/nga-
west2-flatfile-gmpes/

 Database flatfiles have been made

available to the public:
http://peer.berkeley.edu/ngawest2/databases/
32 people have been involved in NGA-
West2

 Technical Coordination Committee:

 Abrahamson, Bozorgnia, Campbell

 External reviewers and oversight

committee:
 Chris Wills, Mark Petersen, John Anderson, Roger
Borcherdt, Silvia Mazzoni, Farzad Naeim

 Funding agencies representatives:

 Badie Rowshandel & Tom Shantz
 32 people have been involved in NGA-
West2
 Database: Ancheta, Darragh, Chiou, Silva,
Stewart, Seyhan, Graves, Wooddell, Katke, Boore,
Kishida, Al Atik, NGA developers
 GMPE Developers:
 Abrahamson-Silva-Kamai
 Boore-Stewart-Seyhan-Atkinson
 Campbell-Bozorgnia
 Chiou-Youngs
 Idriss

 Damping: Rezaeian, Bozorgnia, Idriss,

Abrahamson, Campbell, Silva & GMPE developers
32 people have been involved in NGA-
West2
 Vertical: GMPE developers
 Directivity: Spudich, Chiou, Baker, Shahi,
Rowshandel, Somerville, Bayless, Watson-
Lamprey & GMPE developers
 Directionality: Baker, Shahi, & directivity group
 Site Response: Stewart, Seyhan, Anderson,
Borcherdt, Crouse, Graves, Idriss, Power,
Silva, Shantz
NGA is about solving large complicated
puzzle through collaboration and
team-work

NGA
NGA-Subduction

48
NGA-Subduction

 PEER has initiated a multidisciplinary

multi-year research program to develop
Next Generation Attenuation models for
Subduction earthquakes: NGA-Sub

 We are in the middle of NGA-Sub

 So far the focus has been on development of
the database
On-going efforts for subduction
events data collection
Japan, Taiwan, Solomon Islands,
Cascadia, Mexico, Chile, Peru
 Subduction events and recordings
processed by PEER: Japan

M # of # of
Events Records
> 7.5 11 3,800
6.5-7.5 41 20,000
< 6.5 32 13,000
Total 84 36,800
Subduction events and recordings
processed by PEER: Chile

M # of # of
Events Records
> 7.5 7 70
6.5-7.5 6 24
< 6.5 14 52
Total 27 146

For Chile, PEER also

coordinated with USGS
and surface-wave site
measurements at 31
recording sites have
been completed
NGA-Sub Plan
 Following completion of the database,

 Develop at least one GMPE for

subduction

 If we can secure enough funding, the

plan is to develop more than one GMPE

 Schedule: To complete by May 2016

NGA-East

54
NGA- East
 Goal: To develop next
generation Ground
Motion Prediction
Equations (GMPEs)
for Central & Eastern
US
 Stable Continental
Regions (SCRs)
 Sponsors: NRC, EPRI,
USGS, DOE
NGA-East sub-projects
 Database:
Compiled a database of recorded
motions from Central Eastern North
America (CENA)
 About 30,000 components have been
processed
 Magnitude range: 2.7 to 6.9
 Largest events: Nahanni 6.9, Saguenay 5.9,
Mineral 5.7

 Distance range: 0-1500 km

NGA-East sub-projects

 Geoetchnical Engineering:
Linear and nonlinear site responses; reference
rock condition; reference “Kappa”; …

 Source-Site Path: Investigate

geometrical spreading (1/R, 1/Rx); Q;
Regionalization of CENA,…

 Ground Motion Simulation: Generate

validated simulated GMs for CENA to be
used by GMPE developers
NGA-East sub-projects

 GMPE Development scope

 Next generation attenuation models for
CEUS; applicable to magnitude 4-8,
distances of 0-1000 km

 5% damped response spectra at periods

0.01 to 10 sec
NGA-East status
 Database has already been developed
 It is being QA’d internally
 Regionalization efforts are completed
 Final report is being drafted
 Simulation validation has been
completed
 Running simulations is started

 Complete the project by mid-2015

59
 Thanks!

Yousef@Berkeley.edu