Embed Size (px)
Citation preview
11/30/10
1
The 100-Runs Project
Ruth A. Harris (U.S. Geological Survey)
Harris November 2010
Presentation for the NGA-East SSHAC Workshop 1Critical Issues and Data Needs
Berkeley, CANovember 18, 2010
SOURCE(S) PRODUCED
REGULAR GROUND MOTION EXTREME GROUND MOTION
West Face of Yucca Mountainfrom the Floor of Solitario Canyon
Rickey Ridge, Pahute Mesa, NTS
Harris November 2010
Photos courtesy of Tom Hanks
‘100 RUNS’ = 100 SOURCE REALIZATIONS
Harris November 2010
PROCEDURE FOR THE ‘100 RUNS’ PROJECT
1) This is a computational project so we need toMake sure that the computer codes work:determine if they can produce repeatable regular and extreme ground motion
2) Determine how to produce a range of M6.5 source realizations:Hold workshop and learn about methods for heterogeneous stresses
3) Choose a method and request review-committee comments and approval
4) Conduct ‘100 runs’ spontaneous-rupture exercise
5) Summarize results and provide recommendations for future activities
Harris November 2010
PROCEDURE
1) Make sure that the computer codes work:determine if they can produce repeatable regular and extreme ground motion
11/30/10
2
Harris November 2010http://scecdata.usc.edu/cvws/participants.html
Project CoordinatorRuth Harris, USGSSoftware Engineer
Michael Barall, Invisible SoftwareCurrent Modelers
Brad Aagaard, USGSJean Paul Ampuero, Caltech
Joe Andrews, USGSRalph Archuleta, UCSB
Victor Cruz-Atienza, UNAM, MexicoLuis Dalguer, ETH, Switzerland
Steve Day, SDSUBen Duan, TAM
Eric Dunham, StanfordGeoff Ely, USC
Alice Gabriel, ETHYoshi Kaneko, UCSD
Yuko Kase, GSJ, JapanJeremy Kozdon, StanfordNadia Lapusta, Caltech
Shuo Ma, SDSUHiro Noda, Caltech
David Oglesby, UCRKim Olsen, SDSU
Daniel Roten, SDSUSurendra Nadh Somala, Caltech
Seok Goo Song, URS
Harris November 2010
For more information about our group, please read our SRL article
available on our websitehttp://scecdata.usc.edu/cvws
Harris, R.A., M. Barall, R. Archuleta, B. Aagaard, J.-P. Ampuero, H. Bhat, V. Cruz-Atienza, L. Dalguer, P. Dawson, S. Day,
B. Duan, E. Dunham, G. Ely, Y. Kaneko, Y. Kase, N. Lapusta, Y. Liu, S. Ma, D. Oglesby, K. Olsen, A. Pitarka, S. Song, and E. Templeton,
The SCEC/USGS Dynamic Earthquake-Rupture Code Verification Exercise, Seismological Research Letters, vol. 80, no. 1, 2009.
Harris November 2010
Overall Goal of our Code Verification Group
Compare the computational methods currently being used by SCEC and USGS scientists to simulate (spontaneous) earthquake rupture dynamics
Some Specific Objectives
Understand if our methods are producing the same results when using the same assumptions about friction, crustal structure, fault geometry, etc.
FundingThis project has been funded by the Southern California Earthquake Center,
the U.S. Geological Survey, and the U.S. Dept. of Energy/PG&E-USGS Extreme Ground Motion Project
Harris November 2010
Failure
11/30/10
3
Harris November 2010
Code Comparison StrategyStart simply
Spontaneousrupture on a vertical strike-slipfault set in ahomogeneous(materials)elastic Fullspace
SomeResults
homogeneousinitial stresses
slip-weakeningfriction
REGULAR GROUND MOTION
Code Comparison StrategyIncrementally add complexity
Rupture on a Vertical Strike-Slip fault set in aHomogeneous (materials) elastic Halfspace,Homogeneousinitial stresses,Slip-weakening friction
Rupture on aVertical Strike-Slip fault set in aHomogeneous (materials) elastic halfspace,HeterogeneousInitial stresses,Slip-weakening friction
Rupture on aVertical Strike-Slip fault set in aHeterogeneous(Materials) elastic halfspace,homogeneous initial stresses,Slip-weakening friction
TPV4 TPV5 TPV6-7
Harris November 2010REGULAR GROUND MOTION
Code Comparison Strategy
Rupture on a Vertical Strike-Slip fault set in aHomogeneous(materials) elastic halfspace,Depth-dependentInitial Stresses,Slip-weakening friction
Rupture on a Dipping Dip-slip fault set in a Homogeneous(materials) elastic halfspace,Depth-dependentinitial stresses,Slip-weakening friction
Rupture on a Vertical Dip-Slip fault set in aHomogeneous(materials) elastic halfspace,Depth-dependentinitial stresses,Slip-weakening friction
TPV8 TPV9 TPV10-11
Harris November 2010
Incrementally add complexity
REGULAR GROUND MOTION
Code Comparison StrategyIncrementally add complexity
Rupture on a Dipping Dip-slip fault set in a Homogeneous(materials) Plastic yielding halfspace,Depth-dependent‘strong supershear’ initial stresses,Slip-weakening friction
TPV12 TPV13
Rupture on a Dipping Dip-slip fault set in a Homogeneous(materials) elastic halfspace,Depth-dependent‘strong supershear’ initial stresses,Slip-weakening friction
Harris November 2010EXTREME GROUND MOTION
11/30/10
4
Harris November 2010
Andrews et al., BSSA, 2007Figure 7
Extreme Ground Motion -
Harris Nov. 2010
3D
2D
elastic plastic
Harris November 2010
3D elastic
2D elastic 2D plastic
3D plastic
0 seconds 8
-6
0
m/s
VerticalVelocityatStation1-kmfrom fault,0.3 km-depth,footwall sideof the fault
(3 Hz filter)
Harris November 2010
0 seconds 8
-6
0
m/s
Horizontal(Fault-tracePerpendicular)VelocityatStation1-kmfrom fault,0.3 km-depth,footwall sideof the fault
(3 Hz filter)
3D elastic 3D plastic
2D elastic 2D plastic
11/30/10
5
Harris November 2010
PROCEDURE
1) Make sure that the computer codes work:determine if they can produce repeatable regular and extreme ground motion
2) Determine how to produce a range of M6.5 source realizations:Hold workshop and learn about methods for heterogeneous stresses
o.k.
Harris November 2010
assign 100xfixed
assign 100xFailureCriterion
Computer Codesthat Simulate Earthquakes as
Spontaneous Ruptures
already tested
‘100 runs’ Project – Produce 100 Regular M6.5’s
‘100 runs’ goal- generate this 100x,for 100 M6.5’s
SCEC Heterogeneous Initial Conditions Dynamic Rupture WorkshopFriday May 21, 2010
Davidson Conference Center, USC, Los Angeles, CA
09:00 Introduction Ruth Harris and Norm Abrahamson09:30 The Archuleta/Lavallee Method Ralph Archuleta and Daniel Lavallee10:30 Break10:50 The Dalguer/Mai Method Luis Dalguer and Martin Mai11:50 Lunch12:40 The Olsen Method Kim Olsen and Daniel Roten13:40 The Song Method Seok Goo Song14:40 Break15:00 The Andrews Method Joe Andrews and Michael Barall16:00 More Discussion All17:00 Adjourn
Harris November 2010 Harris November 2010
PROCEDURE
1) Make sure that the codes work:determine if they can produce repeatable regular and extreme ground motion
2) Determine how to produce a range of M6.5 source realizations:Hold workshop and learn about methods for heterogeneous stresses
3) Choose a method and request review-committee comments and approval
o.k.
o.k.
11/30/10
6
Develop Hybrid Initial-Conditions Method and Receive Committee Approval
Harris November 2010
Features:*Gradual Forced Nucleation*Slip-weakening failure criterion*Rigid barriers at fault edges*24 km x 13.6 km fault plane with constant dip*Initial normal stress distribution: depth dependent*Initial shear stress distribution: 1-point statistics = Levy distribution 2-point statistics = von Karman power spectrum*M6.5*1-D 3-layer shear-modulus (and velocity) model*Elastic behavior
Harris November 2010
PROCEDURE
1) Make sure that the computer codes work:determine if they can produce repeatable regular and extreme ground motion
2) Determine how to produce a range of M6.5 source realizations:Hold workshop and learn about methods for heterogeneous stresses
3) Choose a method and request review-committee comments and approval
4) Conduct ‘100 runs’ spontaneous-rupture exercise
5) Summarize results and provide recommendations for future activities
o.k.
o.k.
o.k.
Harris November 2010
What we are finishing (Fall 2010)
With SCEC DOE ExGM and PGE funding:
*Three people have used their tested codes to run 100 M6.5 spontaneous rupture simulations.
The results are being summarized and archived.
This will be followed by a transfer of the results to the kinematic-rupture ground-motion project.
Harris November 2010
PROCEDURE
1) Make sure that the computer codes work:determine if they can produce repeatable regular and extreme ground motion
2) Determine how to produce a range of M6.5 source realizations:Hold workshop and learn about methods for heterogeneous stresses
3) Choose a method and request review-committee comments and approval
4) Conduct ‘100 runs’ spontaneous-rupture exercise
5) Summarize results and provide recommendations for future activities
o.k.
o.k.
o.k.
o.k.
11/30/10
7
Harris November 2010
One Initial Stress-Conditions Realization from the 100 Runs Exercise
Distance along strike (km)
Dis
tanc
e do
wn
dip
(km
)
0.40 0.42 0.44 0.46
initial shear stress / initial normal stress
0
10
-10 0 +10
Harris November 2010
Rupture front contour plot and ground motion from this one realization
Distance along strike (km)
Dis
tanc
e do
wn
dip
(km
)
-10 0 10
10
0
Rupture-front contours (0.5 sec intervals) on the fault-plane
Harris November 2010
Ground motion from this one realizationUnfiltered velocity (+0.1 to -0.2 m/s) vs. time (0-15 secs) at YM station
Vertical velocity
Horizontal velocity(fault parallel)
Horizontal velocity(fault-traceperpendicular)
Harris November 2010
The 100-Runs Project and its Heirs: A To Do List
1) For this exercise: Archive and summarize the results
2) Determine if the heterogeneous conditions assumptionsare consistent with the hypotheses behind them and, develop a new hybrid method
3) Figure out the reasons for mismatches among codes
Strategy: A 2011 heterogeneous stresses group benchmark
11/30/10
8
The End
Harris September 2010
Final Slip (m) for Simulation #0000
0 0.5 1.0 m
-10 0 10 km along strike
0
5
10kmdepth
