Southern California Earthquake Center
Sciente Director
K Ak, Eecut,s e Director
T I-lenses Southern Cilit,’rn,a Earthquake Center
90089 Southern California Earthquake Center Annual Meeting
Institutional Represenians es October 29-31, 1991 Labc,rators California Institute ii Technologs Pasaijena CA 9112P
D Jackson
Department f Earth and Spacebsiences NSF/USGS Los Angeles CA Site Review 90024 SCEC Advisory Council Meeting R Arcfnstieta
Department it Geological S tn_c’
9310t, November 1, 1991
K Mc\aIl, Earth octence.
Board f rud,es UC if Santa ‘run Q50b4
B Stinger
torn
UCSD University Hilton Hotel Los Angeles, California
L Seeber l.amont-[toherrc Geological Obs Columbia Unisersir, Palisades Vt
T Hearin LSGS- OEVF 0- 25S f%,I-,nAse 0 Pasadena CA 9U0 . :
I
Telephone (213) 740-5843 FAX (213) 740-0011 complete.
Note: SCEPP, Individual The SCEC Members Members SCEC 1991 SCEC SCEC SCEC Master Group Group Group Group Group Group Group Group Friday, Thursday, Wednesday, Tuesday, Visitors Evaluation Organization Overview, Annual Academic
This and of of Investigator SCEC NSF/USGS C Model F D E G B H A Activities November Meeting
report SCEC (Fault-Zone (Regional Program (Crustal (Strong (Subsurface (Physics (Master October (Engineering October 1991 Criteria Co-Investigators Seismic October
Advisory infrastructure (Aki) is Agenda Ground Deformation)
Site Model Table 29 of Reports 1 Seismicity) 31 with preliminary Hazard Geology) Imaging Earthquake 30 Council Review Applications) Construction
Motion of regard Analysis) (1991) of
Team
Contents has Sources) Seismogenic Prediction) and
to not and not
education/outreach, been intended
Zones)
included. to be 27 20 E-1 F- H-i D-i G-1 C-i B-i A 1 1 1 7 4 2 6 5 3 1 6 8 9 - 1 1
Oregon Harvard
Davis San
Crouch, Massachusetts
University
Princeton Stanford
University California
University University
Columbia
University
University University
(Coordinating
Diego
and
State
Bachman
University:
University:
University:
University:
Institute
Namson,
State
of
of of of
of of
of
University:
California,
California, California, Southern
California, California,
Nevada,
Institute
University:
Academic
Institution)
Associates,
of
Consulting
Technology;
Reno:
PARTICIPATING
California:
of
Santa
San
Riverside:
Santa
Los
Technology:
INDUSTRY
CORE
Angeles:
Diego:
Inc.:
Barbara:
Cruz:
Geologists:
Co-Investigators
INSTITUTIONS
PARTICIPANTS
INSTITUTIONS
Thomas
John
Geoffrey Peter
Egill Robert
Paul Keiiti
Ralph Leonardo Donald
David
Thorne Ruth Yehuda Duncan
Stephen
Edward
James John
John
AiIm Stephen
Karen
Bradford
Steven
James Robert
Thom
Davis
Beavan
Hauksson
Anderson Harris
Leary
Suppe Cornell
Aid
Archuleta
Jackson
McNally
Rice
Davis
Crouch
Clayton
Day
Yeats
Helmberger
Lay
Agnew
Bock
Miller Keller
Henyey
Park
Martin
Hager
Seeber
(1991)
Ta-hang
William
Charles
Hiroo
Christopher
John
Kenneth
Yan
Kerry
Lynn
Craig Leon
Bernard
Bruce Stephen
Sandra
Steven
Thomas
Steven
Kagan
Orcutt
Knopoff
Sykes
Nicholson
Kanamori
Sieh
Shaw
Ward
Seale
Wesnousky
Sammis
Petak
Minster
Teng
Richard
Rockwell
Hudnut SchoLz SOUTHERN CALIFORNIA EARTHQUAKE CENTER (SCEC) ORGANIZATION
/ /.
/
6 / / V / Steering Committee (BOD, Assistant Directors, Group Leaders)
Group Leaders
Projects
Lunch
Afternoon each Morning
TUESDAY,
Coffee Coffee
group
Education
Group Group Group Group SCEC/USGS
Terrascope Group Working USGS SCEC
Data Group Group Group NSF
Pinon SCEC
GPS Refteks
CUBE
provided
SCEC
Break
break.
Comments
Center!Catalog
session:
Overview
Comments
C
E
A
Administrative
B
Flat
F
0
D
report):
H
session:
OCTOBER
(Fault
(Demonstration
(Geodesy)
(Master
(Strong
(Regional
(Earthquake
Group
(Subsurface
and
(Engineering
at
Pasadena
ANNUAL
hotel
Zone
Outreach/SCEPP
Model)
Motion)
Reports
Seismicity)
Geology)
Update
Office
(noon
Report
Imaging)
Physics)
29
Applications)
9:00
of
1:00
of
to
MEETING
probability
Year
1:00
a.m.
p.m.
1
p.m.).
to
Progress
to
maps
noon
5:30
Tom
Tom
Kei
Elaine Jim
Karen
Kei
Kerry
Steve
Rob
Egill
Leon
Dave
Geoff
Egill
Rob
Hiroo
Dave Aaron
Bernard
AGENDA
by
(discussion
p.m.
Aki
Hays
Aki/Luci
Clayton
Heaton
Clayton
Henyey
Day
Hauksson
Hauksson Jackson
Knopoff
Jackson
Steve
Luci
Martin
Sieh
McNally
Kanamori
Padovani
Martin
Minster
Jones)
Park
Jones!
follows 3
Evening
Coffee
Afternoon
Lunch
Morning
WEDNESDAY,
McNally
Group Group
Group Group
Group
Group Group
Individual
provided
break.
Session
E Session:
G
C
B
D
F
A
session:
and
at
on
Working
OCTOBER
hotel
Tom
Education
(11:30
Henyey.
Group
7:30
3:30
9:00
12:30
and
30
a.m.
Meetings:
Outreach:
p.m.
a.m
pm.
p.m.
to
12:30
to
to
to
noon.
6:30
Discussion
3:30
p.m.).
Egill
Rob
Leon
Ralph Kerry
Kei Dave
p.m.
Aki
Clayton
p.m.
Hauksson
Jackson
Knopoff
Sieh
Archuleta
led
by
Karen 4 5
THURSDAY, OCTOBER 31
Morning session: 9:00 a.m. to 12:30 p.m.
Reports on future plans from working group leaders:
Group E Dave Jackson Group B Ralph Archuleta Group C Kerry Sieh Group D Rob Clayton
Coffee break.
Group F Egill Hauksson Group G Leon Knopoff Group A Kei Aki
End of SCEC Meeting.
Lunch Meeting (12:30 to 2:00) SCEC/IRIS Issues Forum: Tim Ahern and Tom Henyey will lead discussion.
Issues: PASSCAL Instruments Post-Earthquake Response Data Management
Afternoon session: 2:00 p.m. to 5:00 p.m.
SCEC Steering Committee* meets to summarize workshop, and prepare report to Advisory Council and preliminary plan for NSF.
*SCEC Board of Directors, plus Kerry Sieh, Egill Hauksson, Leon Knopoff, and Geoff Martin.
Get-acquainted meeting of SCEC Advisory Council.
Steering
Afternoon
Committee.
team.
2.
and
Morning
FRIDAY,
Discussion
NSF/USGS
Executive
Executive
Lunch
SCEC
Committee.
Session:
NOVEMBER
Steering
for
Session:
should
Session:
Session:
Site-Review
SCEC
Committee
consider
Advisory
SCEC
NSF/USGS
1
Team.
Noon 8:30
2:30
1:00
SCEC
Advisory
meeting
Council
p.m.
p.m. a.m.
to
Progress
Site-Review
Evaluation
1:00
and with
Council
to
to
to
NSF/USGS
noon
2:30
4:00
SCEC
report
p.m.
Criteria.
Team
and
p.m.
p.m.
Advisory
and
SCEC
and
Site-Review
plans
SCEC
Steering
Council
for
year 6
earthquake
the
seismologists.
such
groups,
integration model
had
various
southern
and various
probabilistic
the
earthquake
cooperative
has
NSF
an
enormous
been
developed
many goal
future
workshop
by
The
In
center
My
Tom
regular
meetings
disciplines
joining
California.
the
doing
of
of
part
meetings
present
hazards
source
and
programs
seismic
working
specific
consortium
the
past
program,
Another
into
is
monthly
I
a
was
all
and
share
called
center,
superb
eight
and
3-day
forms
the in
in
and
held
hazard
disciplines
workshops
one
We
southern
the
with
relationship
SCEC
earthquake
meetings
was
the
exploring
months
of
workshops,
job
which
meeting
applicable
SCEC
was
last
responsibility
propose
universities,
“science
model,
in
attended
various
members.
held
by
OVERVIEW,
June
attended
where
his
California.
since
of
is
is
Keiiti
related
new
part,
all
to
to
including to
at
director”.
what
between
at
our
government
the
by
it
Pasadena,
earthquake
of
construct
members, integrate
meeting
also
Aki
Santa
managing
funding
will
the
first
we
center
physicists,
1991
sciences
by
meetings
directors
be
call
geotechnical
I
attempt
Barbara,
deadlines
invited
about
research
am
was
and
the
debated,
hazard
sources,
agencies
the
the
in
responsible
workshops
rock
born, master
a
of office
outside
order
for
complex
“master
week
on
mitigation.
for
findings
improved,
individual
mechanicists
we
constructing
and
concerned
the
at
engineers
model
the
to
ago
experts.
for
SCEC.
have
designed
model”,
operation
physics
developing
develop
demanding
to
from
achieving
updated,
working
through
already
explore
Tom
with
One
this
and
and
the
for
for
of
of
a 7
discretization characteristics, conclusion
power complexity normal
frequency law
can
friction
a model.
the meetings
describe master
SCEC place.
fault longer
for
goal. seismologists.
homogeneous
integration
master
start
with
zone
This
I
activities
law stress model.
times
I
law
This
believe
some
and
realized,
constructing
relation.
presented
geology
parameters
of
model,
morning,
or
distribution
my
is
will
of
than
of
fault
I
of
else
fault
tend
that
true
and
attended
personal
the
the
research
produce
however,
and
and
others
at
slip
to
to
we
He
plane
with
I
important
continuum,
varying
will
our
several
imaging.
computational
emphasize
then
would
now
such
suggested
for
most
view
most
before
a which
master
and
some
never
those
simple
have
a
SCEC
as
like
with
of
of
results
interaction
of
uniform
Thus,
obeys
the
theoretical
these
or
of
such
model.
the
a
that
relevant
the
to
produce
meetings
cyclic
depth
difficulties
solid
the
Gutenberg-Richter
to
direction
directly
tell
my
observational
meetings
the
the
interaction
the
elements
fault
of
physical
occurrence
overview
as
you
This
rate-
to
and
reasons
by
the
absence different
relevant
functions
the
we
with
may
computational
what
and
Jim
foundation
and
of
observed
foundation
and
are
application
why
of
elements,
oversized
began
Rice.
state-
be
of
the
I
of
disciplines
to
power
moving.
the
integration
found
of
earthquakes
some
due
the
the
master
According
to
first
dependent
is
temperature
elements
upon
spatio-temporal
law
including
construction
problems.
see
of
slip-weakening
to
based
through
people
8
I
model
toward
the
the
magnitude
months
inadequate
shall
which
can
of
to
on
friction
master of
efforts
obtain
these
GPS,
Rice,
same
take
take
first
and our
the
the
we
of
of 8 9
Since Rice’s conclusion is fundamental to the construction of our master model, I would like to propose to call it Rice’s law. An important implication of
Rice’s law is that heterogeneities or nonuniformities in material strength or fault plane geometry with a wide range of scale lengths are essential ingredients of fault physics. I observe that this is now a consensus of the working group G on fault physics under the leadership of Leon Knopoff, and I feel that this consensus was made possible by the center-mode research in which a broad-minded interaction has taken place among physicists, geophysicists and geologists. These fundamental propositions on fault physics are directly relevant to the application of our master model in the form of probabilities of earthquake occurrence. Two basic parameters of the earthquake recurrence statistics are the mean recurrence time (T) and its spread (a). The 1988 working group has adopted the log normal distribution proposed by Nishenko and Buland (1987) who found that a=0.21 from global data on recurrence intervals.
Seismicity simulation can give us some insight into how these parameters should depend on the physical and geological conditions of a particular fault segment. For example, Steven Ward estimated the value of a using a model of segmented faults for the earthquakes in the Middle America trench, and found that a is about 0.7, much larger than the Nishenko-Buland value. He found also that, in general, the stronger segment shows smaller a and tends to generate more characteristic earthquakes. This is intuitively acceptable, because weaker segments would be affected more easily by the interactions of other segments than stronger ones. In fact, Savage’s chi square testing of the Nishenko-Buland hypothesis, revealed that, although the statistical significance was marginal, the recurrence intervals were more 10
constant for two Chilean segments, and more variable for the Parkfield segment than the hypothesis predicted. These differences may be attributed to the differences in the physical and geometrical characteristics of the respective fault segments.
Ideally, we would like to develop a master model which can account for all the interactions between segments of a given fault zone, and between segments of different fault zones such as the main San Andreas fault, subparallel strike slip faults, and thrust faults in southern California, as well as coupling of seismic fault slips with aseismic slips and deformation in the ductile part of the lithosphere.
These interactions require a full 3-D model for their understanding. On the other hand, the construction of a realistic 3-D master model for southern
California is probably impractical at this stage. A viable approach may require two parallel complementary research directions. In one of them, we shall study each of the interaction process separately using a full-3-D model but under idealized geological and geophysical conditions. In the other, we shall construct a 2-D master model of earthquakes for the realistic conditions of southern California. The 3-D studies will be used to check the validity of approximations of 2-D studies locally. The global aspects of the problem will be studied using the 2-D model. We shall need to assign parameters of these models through geological and geophysical characterization of earthquake faults.
Here, more intensive integration of efforts of different working groups is needed. We need more productive interaction among groups working on fault geology, seismic imaging, fault physics, seismicity and crustal deformation.
How we accomplish this integration of efforts in the immediate future is the most crucial issue of this meeting. The success or failure of our center 11
depends on how we can integrate these various disciplines to obtain adequate estimates of parameters to be used in our seismicity model for Southern California.
How do we promote necessary interactions among different working groups? For example, during the Santa Barbara workshop Jim Byerlee proposed the existence of strong pore-pressure gradient in the fault zone in the direction perpendicular to the fault plane. Such a zone shall affect the characteristics of trapped modes, head waves and other seismic signatures of the fault zone and should be an excellent target for the fault zone seismology group. Effective points of interaction among different groups may be brought out by asking questions to other groups.
Such a question which may serve as a focal point of interaction among different groups was asked by Dave Jackson at the recent UCLA meeting. His question is about the recent increase in the activity of moderate thrust earthquakes in the LA Basin, from the Whittier Narrows to the Sierra Madre with an apparent northward migration. He asks if this activity has anything to do with the San Andreas fault. Ken Hudnut has a model in which these thrust faults may be connected through aseismic slip across a detachment fault. Can we estimate any change in probability of earthquake occurrence on the San Andreas fault as a result of these moderate thrust earthquakes?
We need to ask questions like these to promote interactions among different working groups, and I encourage you attend as many working group meetings scheduled tomorrow as you can.
Let us now shift our attention closer to the application element of master model. As I mentioned earlier, our center sponsored a one-day
overestimates
The
recorded
the
frequencies strong
for
motion.
California non-linear
is effect
Zonation
leaders
with of
state-of-the-art
interaction
seismologists.
workshop
future
greater
my
frequencies
other
weak
my
The
The
motion
between
of
personal
held
working
at
about
presentation
at
hand,
revealed
working
site
motion
The
between
second
evidence
on
soil
for
distances
at
amplification
amplification
lower
papers
a
the
the
the
sites
data
Stanford.
This
stations
involvement,
week
amplification
groups
evidence
that,
relationship
average.
Liam
weak
accumulating
is
than
than
of
from
Gamble
for
ago
longer
a
on
at
recent
later
rock
Finn
about
(low
I
systematic
factor
soil
at
the
at
the
is
the
shall
as
today.
USC’s
sites
soil
factor
the
House
than
sites and
average,
seismological
strain)
recent
5
regional
data
you’ll
is
go
between
Hz,
sites.
at
results
myself
Gamble
greater
within
about
into
applied
difference
The
all
throughout
workshop
and
International
motion
hear
the
frequencies
some
network
Gamble
from
the
50
who
at
50
House
weak
geotechnical
evidence
very
details
km
soil
and
relation
details
in
km,
the
the
was
were
House
motion
well
in
from
in
at
frequency
site
the
Loma
Conference
on
world
but
Pasadena
least
on
supporting
a
central
asked
is
to
the
than
workshop
strong
other
natural
this
amplification
reversed
gave
the
Prieta up
indicates
hypocenter.
engineers
rock
to
subject
dependent
to
to
subjects
and
strong
(high
considerable
on
a present
2
explore
was
growth
earthquake.
for
sites
pervasive
Hz.
southern
that
Seismic
because
motion
started
strain)
higher
factor
from
only
and
site
the the
the
On
of 12 13
These observed non-linear effects are consistent with the prediction by geotechnical engineers based on the laboratory results on the behavior of soil under a strong shaking.
My presentation was followed by heated discussions which lasted for the whole day. By the end of the day, however, when we went around the table asking what we should do in the future, I felt a consensus emerging from the participants, which may be summarized in the following three recommendations. (I). There is an urgent need for characterizing the geologic conditions under which the site effects are non-linear, linear or transitional. For this purpose, we need to reoccupy the seismograph sites where strong motion data already exist to collect weak-motion data. (2). We need dense 3-D arrays of high-quality seismographs in an area where strong shaking is anticipated, so that the type of research done by C. Y. Chang using the data from SMART-I array in Taiwan, which also demonstrated the non linear effect, can be repeated for other sites. (3). We need to broaden our views encompassing both geotechnical engineering and seismology in order to interface the one-dimensional non-linear approach of the former with the
2-D, 3-D linear approach of the latter including the source and propagation path effect. This is somewhat similar to the dual approach I recommended earlier for seismicity simulation.
The workshop was very refreshing to me because everyone spoke out what they really think, believe and want to do. One of the reasons for this free direct communication may be because nobody from the funding agency was there. The successful Gamble House workshop may offer another justification for the center mode of doing research.
by
of map
manner condition.
source-receiver customer,
with earthquake knowledge
futuristic
receiver maps The
future
on
acceleration the
are
identified
aspect,
the
Charlie
Seismic
also
relation
user
such
view
meeting
When
Since
I
seismic
subgroup
I
of
would
pair
diverse.
dream.
like
in
community
a
and
of
to
Sammis,
probabilistic
on
We
Zonation
time
task.
with
I
any
my
using
all
this
the
to
zoning.
pairs,
now
the
prepared
attach
then
members
series,
of
bring
state-of-the-art
the
The
parameter
ground
parameter.
In
It
I
earthquake
like
the
their
mentioned
can
is
we
was
is
a
community
the
GIS
seismic
Out
we
state-of-the-art
recent
diverse,
to
the
impractical
to
extract
can
needs.
very
motion
probability
address
can
one
technology,
proceed
for
state
synthesize
Repeating
meeting
earlier.
source,
much
hazard
compute
paper
a
of
the
and
of
I
of
given
parameters
the
important
envisioned
to
beyond
ground
users
impressed
the
the
of
presented
of analysis.
method
propagation
application
prepare
however,
There
the
the
the
exceedance
ground
SCEC
the
art
of
a
motion
time
same
results
areas
printed
is
the
the
can
paper,
on
occurrence
by
common
subgroup
at
apparently
a
motion
of
series
following
earth
procedure
the
the
parameter
path
of
need
be
the
probability
into
master
standard
I
unanimous urgent
basis
extracted
omni-purpose
thought
Stanford
for
science
parameters for
and
on
site-specific,
of
is
procedure
model.
a
for
mapping
improvement
of
requested
research
recording
hazard
the
ready
given
in
information.
this
our
all
enthusiasms
Conference
from
the
particular
required
relevant
to
In
map
source-
current
chaired
zoning
was
as
needs
by
or
usual
deal
this
site
the
the
in
to
a
a
a 14
term
earthquake
of
lasting.
earthquakes. both
and
the
information
observational
GIS
their
construction
goal
In
What
technology
summary,
impact
hazard
and
without
The
we
be
elements
and
need
mitigation
tailored
progresses
on
I
sacrificing
based
am
application
earthquake
now
of
very
to
maps
the
is
the
in
we
much
to
master
the
southern
specific
of
are
which
proceed
hazard
excited
the
importance
making
model,
can
master
need
California.
more
mitigation
by
accommodate
appear
of
and
the
vigorously
model
of
a
momentum
try
customer.
to
the
research
to
of
be
long-term
focus
with
southern
constantly
very
being
will
the
on
fundamental,
integration
the
California
gained
be
goal
updated
short
long
of
in 15
Dr.
Mr.
Dr.
Mr.
Dr.
I.
James
Paul
Barbara
James
M.
University
Davis,
Civil Pasadena,
Southern
United
Menlo
345
California
1110
University
Sacramento, Berkeley,
Department
1416
FLORES
IDRISS
Southern
(Jim)
Middlefield
(Jim)
East
916-752-8924 415-329-5163
916-752-5403 415-329-4867
818-795-9055
818-795-2030
9th
ROMANOWICZ
916-445-5718 510-643-5690
916-445-1923
510-643-5811
Engineering [email protected]
CA
Park,
States
DIETERICH
California
CA Street,
Division
DAVIS
of
CA
Green
of
95616
of
CA
CA
California,
Geological
California,
Geology
91106
94720
California
Road
Room
94025
Street
95814
ADVISORY
Department
of
Earthquake
(fax)
(fax) (fax)
(fax)
(fax)
(phone) (phone) (phone)
(phone) (phone)
Mines
(chair)
and
1341
#300
Davis
Berkeley
Survey
and
Geophysics
COUNCIL
Earthquake
Geology
Preparedness
(e-mail)
Project
Center 16
Dr.
Dr.
Dr.
Dr.
Miss
Dr.
John
William
Robert Thomas
Dennis
Shirley
University
Livermore,
Department Earth
Lawrence
University University
Salt Los
Fort
Safety
Los
Department
Colorado Department
Room Cambridge,
City Massachusetts
RUNDLE
Angeles,
Lake
Angeles,
Collins,
(Bob)
MILETI
Administration
(Tom)
(Bill)
415-422-4918
415-294-5236
213-740-5943 213-687-8213
213-740-2411
801-581-5560 303-491-5951 213-485-6400
303-491-2191
801-581-6553
617-253-6208
617-253-3381
300,
Sciences
MATTINGLY
and
State
City,
Livermore
of
SMITH
of
City
PETAK
CA
Park
MA
of
JORDAN
of
of
Systems
CO
CA
CA
Utah
Southern
Institute
Geology
Sociology
Earth,
UT
University
Hall
Department
94550
80523
02139
90012
90089-0021
Office
East
84112-1183
Management (fax) (fax) (fax)
(fax) (fax)
(fax)
(phone)
(phone) (phone)
(phone) (phone)
(phone)
National
Atmospheric
California
and
of
Technology
Geophysics
Laboratory
and
Planetary
Sciences 17
utsJde
c!SGS
NSF
Visitors
John
Susan Wayne
Visitors
Charles Robert
Jases
Elaine
James
Shih-Chi
Visitors
Anchorage,
University
H2M Tucson,
2550 General University
Gould
Geoscience
L.
Department
Pennsylvania
Deike
L.
K.
H.
F.
L.
Padavani
Abo
A.
Thatcher
Beck
Whitcomb
Hays
Liu
Hill
Denali
Wesson
Simpson
Building Langston
AZ Engineering -
-
Program
AK
of Park,
85721 of Department
-
Director,
NSF/USGS St.,
—
USGS State
— for
-
Arizona
Volcanos
99503
Geosciences
Building
Office
Geophysicist,
NSI
Director,
8th External
PA
Coordinator Director,
University
Coordinator
November
Manager
Tel
Tel Tel
Tel Tel
Tel
16802
Bead,
Floor SITE
Eartb
and
(202)357—7356,
(202)357—7958,
(415)329—4810, (703)648—6722,
(703)648—6714,
(202)357—9780,
Geophysics
Research VISIT
USGS
Sciences
Engineering
Earthquake
1,
Menlo
for
for
1991
OF
Of
Tel
Tel the
Tel
FAX FAX
FAX
fic.
THE
the
Park
Division
SCEC,
(907)278—2551
(814)863—7823
(907)277—9736 (602)621-2672 (814)865—0083
(602)621—8628
Hazard SCEC
SCEC,
FAX
ofEarthquakes, FAX FAX
FAX
FAX
FAX
Office
(202)357—0364
(703)648—6717 (202)357—0364
(202)357—9803
(703)648—6717
(415)329—5163
Deputy
Program
Mitigation
chief 18 THE MASTER MODEL
( Instrumentation, Deployment Strategies, Special Studier)
Observations Data Base Other National Maintenance & International Data Bases (Geology, Gravity, Heat Flow, I etc.) } Modeling & University & Model Government Construction Research } } Models & Model Maintenance
Visitors
Interpretation & J Use of Models
hazard
public
developed new
other. and integrated A.
apply geophysical, terms
II. earthquake-related
represented through model the
Office
institutions, I.
seismic
scientists
SOUTHERN
Center
acquire
research
Master
The
Research
Research
The
Goal
(i.e.
of
and
Its
will of
analysis
hazard
various
a
Earthquakes,
substance
from
for
Southern
transfer
private
master
is
into
represent
consensus
in
and
of
and
in
Model
to
results
the
terms
model
Directions
the
partnership
meetings
forms
integrate
integrate
of
CALIFORNIA
seven
seismological
and
purpose
sectors.
model
to
science
southern
Center
Construction California
will
of
become
EVALUATION
a
the
(master
Research
applicable
on
Volcanoes,
two
distillation core
be
and
user
research
the
and
is
some
of
in
debated
or
with
institutions
California.
a
pertinent
workshops,
developing
available.
model)
Objectives
community)
more
order
information Earthquake
framework
issues,
Performance
EARTHQUAKE
the
to
findings
of
and
in
differing
to
earthquake
Center
for
United
regularly
data
Engineering
CRITERIA
develop
This
while
and
a
southern
and
SCEC
Center
pertinent
the
prototype from
for
in
thinking
updated
distillation
opinions.
States
a
master
on
which
model
scheduled
a
number
the
hazard
will
(SCEC)
other
California.
prototype
(USGS).
to
CENTER
various
Geological
probabilistic
on
develop,
improvement
model
geologic,
which
earthquakes
issues
mitigation
may
a
of
is
regular
workshops,
disciplines
participating
composed
on
is
probabilistic
be
The
The
refine
it
developed
(SCEC)
one
geodetic,
stated
Survey’s
may
basis
seismic
goal
will master
on
in
hand,
and
and
the
the
be
be
of
as
of
in
in 20 21
Master Model Improvement and Maintenance
The requirements of the master model will guide data acquisition and interpretation through the processes of interaction and feedback. As such, the master model will be constantly improved and updated as new research results become available. To facilitate master model improvement and define its research directions and objectives, the Center has been configured into eight disciplinary working groups as follows:
A) SeismicHazard Analysis and E) Crustal Deformation Master Model Construction F) Regional Seismicily B) Strong Ground Motion Prediction G) Physics of Earthquake C) Fault Zone Geology Sources D) Subsurface Seismic Imaging H) Engineering Applications
The research results of working groups B to F will provide input to the master model, and when fully integrated by working group A, will be the best representation of the earthquake process. Group A will maintain the master model in its most current form. The master model will be maintained as a data base, a set of model parameters, and a set of products (principally digital maps) at the SCEC data center in Pasadena. Working group H will provide for engineering applications of the master model.
Master Model Output
The products derived from SCEC research will consist largely of maps and data bases related to probabilistic estimates of earthquake occurrence and strong ground motion. Estimates of strong ground motion depend on a knowledge of fault failure as well as propagation path and local site conditions, particularly since the population distribution in southern California is concentrated away from the main San Andreas fault. The Center will undertake the hazard analysis in two steps. The first step, which will be completed in two years, will involve updating the geologic data on faults (mainly from trenching), and determining the propagation and site effects to construct maps of exceedance probabilities for strong ground motion parameters in southern California. The second step, which will be completed in five years, will involve adding to our hazard analysis the new geodetic data, particularly those data relevant to blind thrusts. In its product, the Center will combine all pertinent information on earthquake hazards using a Bayesian approach. 22
B. Measures of Research Performance • Publication • What is the scientific and/or technical impact of research results in guiding other research and in improving probabilistic hazard analyses? • Are papers published in peer-reviewed scientific journals? • Is there timely dissemination of information regarding center workshops to the scientific community and publication of technical reports detailing various elements of the master model? • Are information circulars published from time to time, for example, after major earthquakes or to assist in classroom instruction? • Interaction • Is the Center, through its working group structure and its focus on the master model, and through workshops, monthly meetings, special symposia, and joint publications effective in facilitating scientific interaction? • Does the Center facilitate interaction between the various earth science disciplines involved in earthquake studies and seismic hazard analysis? • Does the Center facilitate interactions between scientists at the participating institutions and the USGS? • Does the Center facilitate interactions between earth scientists and engineers? • Master Model Impact • What is the impact of the master model on the earth sciences and on earthquake hazard reduction methodologies? • Do the scientific and earthquake hazard mitigation communities accept the master model as a useful way of both structuring scientific research programs on earthquakes and developing probabilistic earthquake forecasting and strong ground motion prediction strategies? • Is the master model significantly improved from year to year beyond relatively elementary models? • Data and Ideas • Is the Center assuming the role of a clearinghouse for data and ideas pertaining to earthquake research and hazard mitigation?
•
•
A. III.
will
• Post-Earthquake
Other
• • •
• • •
•
assume
The Education
Objectives
post-earthquake
What
Is earthquake? maintained the
Is What
Does Earthquake of Does for and who its
government? Does officials, dissemination ground IRIS, Is
sources Earthquake southern compatibility
(California
(Southern
Funding
the
the
function
the
Center
own
general
timeliness
accessibility
are
the the
future
is
the
the
is
Center
Center
southern
motion,
such
the
participants,
funded
the
California?
USGS,
Center
Center
Center
and
and
has
California
and
use?
Division
quality
Prediction
Prediction
with level
Information
leadership
effective
effective
as
with
with
of
a
updated?
the
Outreach
scientific
from
fault
FEMA,
develop
and
of
primary
assume
assume
California
other
data
of
which
the
and
other
press
UC
use
of
other
zone
but
Earthquake
in
in and
SCEC
Evaluation
Evaluation
quantity
organizations
Mines
roles
Berkeley?
foundations,
effective
by
assimilating
investigations?
role
its those
a also
data
an
following
geology,
sources?
other
responsibility
both
ability
effective
earthquake
data
in
in
from
centers
data
and
of
the
Center
earthquake
information
communication
center
Council,
Council)
Preparedness
new
to
Geology),
are
etc.)
other
education
business,
data
a
such
leverage
role
such
archived
data
damaging
in
and
generated
catalog
for
and
researchers
Pasadena,
in
and as
following
as
(seismic,
non-Center
research
timely
to
ideas
the
CEPEC
the
funding
and
of
those
NEPEC
by
Project),
scientists,
and
being
individuals
coordination
USGS,
earthquake
by
the
state
not
and
as
coordination
operated
GPS,
a
world-wide
SCEC,
and
(California
data
from
well
only
damaging
(National
scientists
and
properly
accurate
CDMG
SCEPP
hazard
public
strong
center
as
from
other
local
who
and
by
its
in
of 23
hazard
business
projected minorities
Finally, elsewhere
reduction.
undergraduates
B. in a doctoral
• increase
groups,
•
•
means
general.
Students,
Special •
Women
•
• •
• •
• •
•
•
Measures
reduction
What
visitors
Where
What
Has
to Has Center?
scientists
What Does the school
preparedness
in
newscasters, Is Are
coordinators,
relationship
Do in
of
community
given
fellows,
earthquake
make
the
population into
understandable
there
in
encouraging
various
Interest
and
It
SCEC,
such
To
the
the
Post-docs,
the
the
is
efforts
activities
must
is
students
do the
and
attracted
the
Center
the
Center
this
Minorities
of
meaningful
effective
Are
persons Center
the
U.S.
under
in
students
and
scientific
Performance
core
Groups
also
SCEPP,
high
quality
societal
southern
in
and
etc.
end
they
are
quality
awareness
and
insurance
and
demographics
reaching
to
developed
provided and
of
the
and
careers
be
and
include
to
the
being
who
school
involved
it
in
invite
format,
the
communication
response
and
the
the
of
visitors sensitive
teachers,
contact
desires
Visitors
public
mainstream,
participating demand?
and
importance
California,
of
Center?
are
graduate
world.
made Center?
post-docs
in
out
participation
students
an
interaction
the
among
the
underwriters,
concerned
the
an
earth
in
at
program?
to
with
opportunity
to
to
perhaps
officials,
to
decision
effective
experience
USGS
large,
Center’s
of
the
Do
The
the
students,
train
recruit
sciences,
Center
the
in
particularly
public
the
end
institutions?
of
above
they
with
need
the
and
Center
with
develop
Center
with
in
earthquakes
by
graduate
up
southern
making?
program
research
corporate
and/or
assume
conjunction southern
groups
scientists
to
for
women
realtors,
to
groups?
post-doctoral
officials,
after
earthquake
for
specifically,
visiting
communicate
bring
must
college
has
given
an
visitors
interest
their
leadership
students
findings.
of
such California
and
a
California
effective
more
also
disaster
and/or
new
responsibility
and
outreach
the
scientists
the
tenure
with
undergraduates
minorities?
awareness?
as
such
reach
swriters
and
and
fellows,
women
present
media,
earthquake
emergency
with
projects
SCEPP?
and
roles?
planning
with
science,
working
hosting
persons
region.
area
to
out
post
these from
high
and
and
and
the
and
the
as
to
to
at 24 25
IV. Technology Transfer
A. Objectives
The master model is the product of the Center. Its essence must be transferred to other scientists, earthquake engineers, emergency preparedness and response personnel, government officials and policy makers, the business community, the media, and the general public. SCEC will make use of workshops, technical reports, special publications, maps, and data bases to transfer the master model to the users. Forms of information transfer will necessarily be different for the different user groups. Newsletters and master model updates will issued on a regular basis. SCEC will team up with SCEPP in the transfer of information to the less technical user community. It will cooperate with the USGS and CDMG (California Division of Mines and Geology) in the transfer of more technical information.
B. Measure of Performance • Are workshops held for the user groups? How effective are they? • Are the publications which are required for technology transfer being produced by the Center?
• Are the master model products accepted and/or used by other scientists, engineers, private consultants, public officials, and the public at large?
• How universally applicable is the master model concept? • Is the Center effective in increasing earthquake awareness in southern California and the rest of the nation? • Does the Center have an impact on seismic policy in California and the nation as a whole? • Is it being used as a consultative body on matters pertaining to the study of earthquakes and how to use the results of such studies for earthquake hazard mitigation? • Does it help facilitate the implementation of federal and state programs in earthquake research? 26
VI. Institutional Support and Management
• Institutional Support • What is the level of institutional support at the core and participating institutions? • Management • How effective is the management in keeping the Center focused, interactive and generally productive? • Are the scientific participants and officials at the participating institutions satisfied with the managers and management structure? • How is the Center management perceived by NSF, the USGS, and other outside organizations with which SCEC has relationships? • Governing Boards • How effective are the Center Board of Directors, the Steering Committee, Group Leaders, and Advisory Council in providing guidance and oversight? Do these groups show a genuine interest in making the Center viable and productive? • Research Groups • Are the research groups and group leaders effective in planning, organizing, carrying out, summarizing, and integrating the research activities and results? • Funding • Have objective procedures been devised for distributing research funds to the various principal investigators, including the development of an overall scientific plan, and accounting for their expenditure vis-a-vis research productivity and scientific quality? Southern Call forn I a Earthquake Center
1991 Visitors and Post-doctoral Fellows
Visitor Host Institution Research Project
V. Keilis-Borok (Academy of UCLA Precursor Phenomenology and Modeling Sciences of USSR) (Ph.D.-1947) and Tanya Levshina
Peter Molnar (MIT) (Ph.D-1970) UC-Santa Barbara Rotation Tectonics and Crustal Shortening across the Transverse Ranges (Possibility of M8 Thrust Earthquake)
Geoffrey King (IPG, Paris) USGS/USC Blind Thrust and Mechanics of Fault Zone (Ph.D.-1970) Deformation
Minoru Takeo (ERI, Tokyo) Caltech Strong Motion Simulation for Southern (Ph.D.- 1982) Call fo rn ía
Gianluca Valensise (ING, Rome) UC-Santa Cruz Terrace Topography and Palos Verdes Fault (Ph.D.- 1987)
Jian Lin (Woods Hole) USGS Mechanical Analysis of Blind Thrust Fault (Ph.D.-1988) Network in the Los Angeles Basin
Sergio Barrientos (Univ. of Chile) UC-Santa Cruz Fault Segmentation Study Using Dislocation (Ph.D.- 1987) Model
David Scott (Oxford) USC Role of Fluid in Mechanical Models of (Ph.D.-1987) Earthquakes
Rachel Abercrombie (Reading) USC Fault Mapping and Seismic Scaling Law in the (Ph.D.-1991) Scale Range I m to 10 km
T’J Groun A: Master Model Construction and Seismic Hazard Analvsk
Grrnin leader: Keilti Aki
Probabilistic Seismic Hazard Assessment of the Peterson-Sykes-Jacob (Columbia) A-I Greater Los Angeles Region
Groundwork for a Master Model Wesnousky (Nevada-Reno) A-4
Development of a short-term prediction program Minster and Agnew (UC-San Diego) A-5 and Master Model Definition
Seismic Hazard Due to Interacting Fault Segment Cornell (Stanford) A-6
Comparing Historic and Geologically Estimated Fault Slip Jackson-Davis (UCLA) A-7
Determination of Site-Specific Weak-Motion Amplification Aki and (USC) A-9 Factor for California
GIS Working Group Report Sammis (USC) A-14
Digital Geologic Map Database for Southern California Park (UC-Riverside) A-16
Mapping participation in SCEC Miller (UC-Santa Barbara) A-19
Design and implementation of the Master Model GIS Richard (UC-Santa Barbara) A-22
Living with the San Andreas fault McNally (UC-Santa Cruz) A-23
soft
alluvium
*observed hard
Geology
Table
SanJacintoValley(M7)
Anza(M7)
Coachella
San
San
Mojave Table
Segment
with
computed These
period
time recurrence
near
Ontario, and
wave coastal
Bernardino probabilities
from
0.3 by
average segments the
of and segments velocity California
magnitude
earthquakes
compares
Mark
Larnont-Doherty
from
Using
Probabilistic
A
exceeding
the
rock
s,
Poissonian
Bernardino
Bernardino
hard-rock
San
probabilistic
rock-igneous
the
San
dependent
strong
2
1
motion.
a
D.
hazard
Characteristic
for
Working
factor
(M7.5) areas.
Time-dependent
hard
attenuation
Jacinto
-Quatemary
response
and
-Tertiary
Petersen,
Bernardino
of
Valley
hazard
of
amplification
hazard
hard
Earthquake
models
7.5
motion
the
area,
along
San
the
rock
is
curves
sites.
of
20
The
(average
Valley
Hazard
rock
recurrence
Mtns.(M7.5)
along
greater
San
faults
Seismic
Group
San
2.0
(M7.5)
Bernardino
seismic
Geological
calculated
curves
spectra
sites
cm/s
exceeding
the
Lynn
in
function
data
30-year
Our Andreas
higher
sites
Jacinto
are
indicate regions
0.3
(standard
amplification
and
San
the
(M7)
Probabilities
(Table
curves
(1988):
than
factor
Earthquakes
at
for
long-term)
R.
data
particularly
hazard
at
(PSV)
s
Hazard
range
because
southern
Andreas
calculations
and
period
1
those
Sykes,
using
at
1.2
1.1
1.3
the
Observatory
and
probability
using
fault
20%
fault
indicate
30
s
40%
is
2).
that
1
recurrence
time-dependent
are
deviation)
period
(0.7) (0.2)
(0.7)
the
1971
s
assessment
for
2.4,
the
from
km Poissonian
three
time-varying
and
The
and
(Table
for
both
and
the computed
the
factor
Mojave, and
for
large
recurrence
rate.
(1988) Assessment
San
from
adjusted
high
(M 5 =6.5)
that
a
20%
most
the
area calculations
30-year
Klaus
sites
along
time-dependent
for
(Figure factor
time-dependent
San
of
Andreas of
earthquakes
probabilities,
1).
We
mean
San
the
because
0.13
0.29
0.17 0.22
0.36
0.32
for
is
exceeding
near
is
Columbia
both
of
Jacinto
amplification
(standard
and
I
San
Jacob
to
for
located
conducted
We
San
Recurrence
calculate
of
s
Bernardino
the
the
probability and
the
probabilities
San
period
1).
PSV 2.0
-
Los
are
2.7
Bernardino
time-dependent
both
incorporate
2.0
2.0
1.2
Andreas
for
fault
next
the
greater
indicate
Poisson
San
Fernando
for
faults.
Uncertainties
caused
higher
Angeles
amplitudes
in
(M University,
deviation)
(0.8)
(1.4)
(0.8)
area 20
rock
hazard
the
and
30
the
estimated
an
for
recurrence
Andreas
>
factor
cm/s
Valley,
(Figure
years
Los
alluvial
recurrence
The
or
is
Poissonian
that probabilities
7)
at
*
the
primarily
and
Models
Greater
recurrence
Poissonian
of
Mountains
earthquake
located
to
San
3
along
site
calculation.
Angeles
is
greater
exceeding
the
60%
s
are
alluvial
is
65%
Palisades,
on
2).
San
particularly
are
and response
Jacinto
amplification by
(standard
3
about
valley
and
median
about
six
s
and
alluvium
probability
close
near
high
by
Jacinto
period
models
the
recurrence
Los
Los
Poissonian
probabilities,
fault
0.19
0.15 0.11 0.14
0.14
0.17
region,
2.7 0.6
2.0
and
of
San
sites
for
the
two
30%
that
a
faults.
Working
20
San
(up Angeles
to
30-year
exceeding
NY
factor
that
Angeles
deviation)
(0.4)
(0.9)
(1.0)
alluvial,
Coachella
scatter
segments
times
Valley
cm/s
for
the
may
Jacinto
in
than
high
to
for
Bernardino
10964
factor
excluding
we
of
Los
characteristic
probabilities.
San
40%)
Probabilities
1.3
amplify
Poissonian they
an
PSV recurrence
probability
larger
region
in
calculated
associated
Group
and
soft-rock
Angeles,
higher
Andreas
event
pseudo-
Region
the
defined
Faults
Valley
for
are
at
Anza
than
that
San
the
the the
for
1
on
on
of
at s .D gO.2 p0.8 g02
Figure 0.6 1.0 1.0 0.0 1.0 0.0 1.0
(a)
(b)
recurrence recurrence (lower alluvium - ......
*-5-f
1. 20 20 20
Hazard
dashed 0.s-aNuvium 0.3s.aIuvum PSV 40 PSV 40 0.3s-rock PSV PSV
TIME-DEPENDENT 0.3s-rock 40
sites
probabilities. (cm/i) (cm/I)
probabilities
(cm/I) (cm/i)
POISSONIAN 60 60
curves 60
curve)
at -- :------
0.3 - 80 80 - 80
s,
showing
confidence .— 100 100
1 100 - .
s,
as . go. 0.8 g02 ‘0.8
and 0.6 1.0 OC_
- 1.0 0.0 1.0 1.0 0.0
input.
the -__ -
3 ...
levels -
median s 20 20
periods. -.
PROBABILITIES
(b) --.. PSV 1s-aIIuvum PSV
for 40 PSV 1 Is-rock PSV 40 s-aik3vIum Is-rock
(solid
hazard
(cm/i)
(cm/i) (cm/i)
a (cm/I)
PROBABILITIES 60 60
(a)
site 60
curve),
hazard
in 80 80
curves 80
the 100 100 100 .
city
85%
curves .D p0.8
gO.2
computed p0.8 j:
gO.2 0.6 i.o ° 1.0 0.0 1.0 1
of 0.0 .0-
(upper 0lX
San
computed - - 20 20
Bernardino
dashed
using
3s-alluvium PSV PSV 40 3i-alhjvium PSV 3s-rock PSV 40
3s-cock
(cm/i) (cm/i) (cm/i)
using (cm/i)
time-dependent
curve) 60 60
for
poissonian 80 80
rock and 100 100
15%
and Figure
-
parameters
decreasing
sites
1.0
1.0
00__6080
1.0
1.0
1.0
2.
Los
at
Hazard
1
soil
s
the
Angeles
rupture
constant.
PSV
PSV
PSV
allenuatbn PSV
period
PSV
probability
magn4tude
axnplltlcatton
curves
given
(cml)
(cm/c)
(cm/a)
(cm/a)
(cm/a)
length
6080
6080
parameter.
computed
for
80
Median
the
1X
0C_40
cities
___
curves
by
i
1.0
1.0
1.0
1.0
.o
of
varying
20
Los
are
Ontario
Angeles,
soIl
rupture
solid,
PSV
PSV
PSV
40
PSV
aflercjatcn
PSV
probability
magnitude
one
aniplificatlon/
(cm/a)
(cm/a)
(cm/c)
(cm/s)
(crrVs)
length
60
60
input
dashed
Ontario,
80
80
parameter
curves
IX
1
and
1.0
1.0
1.0
1.0
i.o
San
obtained
San while
___
Bernardino
Bernardino
soil
holding
rupture
- by
PSV
PSV
PSV
PSV
attenua1n
PSV
40
probabi?y
magnitude
amplification
(cm/i)
(cm/a)
(cm/s)
(cm/s(
(cm/c)
increasing
length
60
for
all
alluvial
80
other
or
1X 14-3 SCEC Progress Report: Groundwork for a Master Model Period: April 1, 1991 to October 18, 1991 P.1.: Steven G. Wesnousky
The Southern California Earthquake Center has embarked upon an effort to develop a master model of seismic hazard for southern California. The underpinnings of the effort will be data describing the locations, slip rates, and paleoearthquake histories of active faults in the region. Hence, we have begun to synthesize the slip rates, paleoearthquake histories, and historical record of earthquakes along mapped faults and related seismogenic structures in southern California. To date, we have finished our synthesis of data bearing on the major strike-slip faults of southern California: The San Andreas, San Jacinto, Elsinore, Rose Canyon, Newport Inglewood, and Palos Verdes fault systems. For each fault, we have prepared a written summary describing the results of paleoseismological studies for each fault zone, maps showing the sites of the respective studies, and a complete bibliography of the published literature describing the details of the particular studies (Appendix). Currently, we are placing the information in a tabular format that may (1) readily be input into a GIS data base and (2) used as the basis for constructing long-term seismic hazard maps. Subsequently, work will go toward completing a similar update for remaining faults and other seismogenic structures (e.g. blind thrusts) in the region. On a separate front, we have also began looking at characteristics of the earthquake frequency distribution for particular faults by combining paleoseismological data with historical and instrumental earthquake statistics. Understanding the shape of the b-value curve for individual faults is needed to assess the hazard due to moderate to large earthquakes which are not primary surface rupturing events. Our observations to date suggest that the San Andreas and Garlock faults, the faults with greatest cumulative offset are characterized by a ‘characteristic earthquake’ frequency distribution whereas the San Jacinto and Newport Inglewood faults show a classical Gutenberg-Richter distribution. Finally, an initial digitized data set of faults and slip rates has been provided to Steve Park at UCR to incorporate into a GIS data base and I have provided the output of initial hazard map calculations to Lucy. Jones so that she may incorporate aspects of probability gain due to the occurrence of possible foreshocks into hazard map calculations. A-c
Southern California Earthquake Center Annual Activity Report Jean-Bernard Minster Duncan Agnew Institute of Geophysics and Planetary Physics, Scripps Institution of Oceanography, University of California, San Diego, La Jolla, CA 92093-0225
Dr. Minster, as member of the Board of Directors representing the Scripps Institution of Oceanography (UCSD) attended all Board meetings but one, and all monthly seminars but one (in each instance of an absence, UCSD was represented by Dr. Agnew). He also organized and hosted the July, 1991 monthly seminar series at the Institute of Geophysics and Planetary Physics, in San Diego. Drs. Minster and Agnew both participated in the activities of the Working Group on the Master Model, and contributed a presentation to the working group meeting of July 29, 1991, discussing how the master model definition should be refined in terms of its practical applications. Dr. Minster initiated, in collaboration with Dr. Archuleta of UCSB, the drafting of a Memorandum of Understanding between SCEC member institutions, to coordinate the use of portable (PASSCAL) instrumentation for field experiments and for rapid response to earthquakes. This Memorandum of Understanding should be followed by another one between SCEC and the Incorporated Research Institutions for Seismology (IRIS), outlining modes of cooperation between SCEC and the IRIS PASSCAL program in the use of portable instruments in southern California. After the Sierra Madre earthquake, Dr. Minster and Dr. F. Vernon of UCSD deployed four portable broadband (STS-2) stations in the epicentral area. The data collected in this deployment was rather sparse, in view of the paucity of aftershocks, but the experience pointed to the need for a more formal earthquake response plan. This observation was transmitted to SCEC in the form of a memo suggesting the formation of an ad hoc working group to develop such a plan. Dr. Agnew has addressed the question of the completeness of the pre-instrurnental earthquake record for southern California. This involved preparing a new version of the Evernden algorithm for predicting seismic intensities (available to the working group for using in predicting future earthquake effects), computing the intensity distribution for specific events, and comparing these with the recorded intensities--or, for hypothetical events, those that might have been recorded (which required a re-examination of the sources for the historical catalogs). The result shows that where constraints are available from paleoseismology these are almost always better than the historical record can provide. The intensities reported for the event of 8 December 1812 are consistent with a rupture on the San Andreas fault near Wrightwood, extending some distance along the fault from this point. Discussions have also been held with Dr. Lucy Jones on the application of the foreshock probabilities to a larger suite of faults in southern California than those already studied. TO : PHONE NO. : 1213740it OCT.22.1991 1:48PM P 2 FROM : Dr. C. nilin Com11 Prtc1a LJa11ey,C 94028 PHONE NO. : 415 854 8053
Southern California arthquake Center Annual Report: Feb. 1, 1991 to Jan, 31, 1992 Prof. C. Aiim Cornell, P.1. Dr. Steven R. Winterstein, P.1. Department of Civil Engineering Stanford University Subject: Se,smic Hazard Due to Interacting Fault Segments Funded Period: Oct. 1, 1991 — Jan 31, 1992 The Center is committed to making increasingly more accurate probabilistic seismic hazard estimates for sites in Southern California. The current seismic hazard methodology includes at most renewal recurrence models of independent faults and fault segments, coupled with ground motion prediction equations. Our research will extend the models to recognize that there is mechanical coupling between fault segments on the same and neighboring faults, and that slip on one fault segment may accelerate or retard the events on other segments thereby increasing or decreasing the estimate hazard contributed by those segments. The work extends past analytical recurrence work by the investigators (e.g., Cornell and Winterstein; BSSA, 1988) and work which has been in progress through a Ph.D. student, S. C. Wu, in cooperation with USGS Drs. James Dieterich and Robert Simpson. Indeed, primitive notions of this work (in a fault—recurrence rather than a site—hazard form) appeared in an Appendix of the 1990 ay Area NEPEC Working Group report, a group in which Cornell and Dieterich both served.
We approach the problem (1) from the perspective of segments as the atomic unit, because this creates the possibility of verification and application, and (2) from the phenomenological perspective (e.g., relating segment interactions through increments to stochastically modeled inter—event times rather than explicitly as stress state changes.) The focus will be on development and exploration of recurrence and site-oriented hazard models, analytically and via simulation as necessary. Further the work should be conitent with other SCEC workers, e.g., the hazard mapping of Wesnousky and Jones, on one hand, and the non-linear, multi—mass slider dynamics of Jackson, Kagan and others, on th. other. Some preliminary results have been presented orally at the International Conference on Stochastic Mechanics in Corfu, Greece, September, 1991, We are preparing a paper based on that presentation for submission to BSSA.
Student: Ph.D. student funded for four months: S. C. Wu Cooperating Investigators: James Diatarich and Robert Simpson, USGS, Menlo Park
—-1—
episodic
as.i.mic
of
future,
record,
Northern
energy
instrumental up
nor
ha.
Th.
a
The
8
N
All
AU.
All
Region
Table
magnitude
moment
wheuld
segments, based
prsclicited
We
earthquakes.
regions
agrees purpose seismicity
slip, earthquake we
certain
Calif
Calif
occurr.no•
with
L:4
southern
been
result.
COMPARING
cumulative
used
Calif
Calif
Calif
compared
release
1
en
or
especially
and
the
released
occur
M
slip
if
with
Calforriia
in
the
or
is
Seismic
threshold
their
1810—1990
1810—1990
1933—1990
.ignif
1810—1933
1810—1990
Dates
6.
which
it
with
geological
the
show
moment
FE
faults
is to
recording
record
California
method
geological
is
of
th.
Th.
HIS’IORIC
there.
occurs.
has
distribution
stationary,
see
south
geologic
not
cant
large that
that
in
influences
moment
fault
results
that
also
is
rates
are
whther
with
historic
is
of
observed
Historical
We
predicted a.eismic
further Most
(1857,
earthquakes
became
rats
AND
shown
Wesnousky,
presently
since been
slip
October
has
compared
DAVID
predictions,
the
release,
slip
on
10
18
14
are
4
6
the
of G!OLOGICALLY
and
had or
the
each of
much geologically
implied
earthquakes
1872,
in
q.od.tically
routine.
1933;,
behind.
summarized
th.
rate
D
release
slip
the
moment
to
by
24,
enough
the size
JACKSON
these
of
loaded,
slower
1018
historic
1986j
determine
must plate
plate
1952)t
and
must 15
15
15 Geological
1991
about
figure.
8
6
by
implie,
of
to
Th.
for
of
Nm/yr Thus
earthquakes
numbers
the
the
increase
ESTIMATED
earthquakes
we
determine
occur.
tectonics,
since
tectonic
above in
ready
eeiaxnic
estimated
rate Neither
earthquakes
100
now,
moment
length
calculated
historic
the
Table
whether
that
fault
of
1933,
with
about
to
it Sinc,
historic
in
moment
the
seismic
rate,
FAULT
1:
northern
release
of
20
whether 13 suffer
34 must
34
34
Plate
to
Our
the rate
which
the
segments
certain
when
the
this
rat.
keep
above
be
but
SLIP
Teat
of
,q 7 C
C) 0
D
0
0 CD 0
I —. U)
0 C (J) C.”
1<
—
r’
vi
0.
-q c
—nil
•
r—
XO!
•
cm—.
x
P § xe
un..
ox
ox
Ox
OX
rn ,,i..
0
i
r OX OX
OX OX
x
—r
u—- ox OX X
X ox
ox 0
•
Ox X r OX
OX Ox
OX ,
o 0 ‘ 0
8
OX
— 0
OX
OX ox ox oc
ox ox
-IT,
o o
I
XO Xe
I• ,— tr 0
XO
*0 Xc
—
P B e
x
Year
.lin
i
I
•
• ,— —T. ______
per
I.
r—r
P R
—
—
,‘,..,..
5
r
rrvn
i ,,
Events
nit
—r , -,
I
X0
Ti]
P
lit xe
.,,
0* ox
OX Ox
OX ox Ox
ox oX
OX Ox
X OX ox
ox
•,1r,
OX Ti ox
OX
ox OX
ES—JCLA
O Ox OX $
OX
OX .... OX
ox
• OX
OX ox 6 0 O
X u-,-
FROM
——rv
,
x
XO •
•_t
*0 O --
.
X
Q
Al
.....
•
..
i
I
v
-i--i
______
u - :24
o
o 3 CD 5.-’ z-. CD o
CD; CD -‘ U)
pervasive
used
weak Imperial
understanding simple frequencies. 2 Formation
zonation, Southern greater greater although results Hz
since that
site of
median in Determination
sediments, Factor
K.
shows
the
Central
have
amplification
Aki
the
seven
the
motion
station.
show
Figures
way We
The
Our
The
amplification
at
geologic
and
Valley
been
general
the
station
the
California
non-linear
as
rock
for
and
California
symbols
(3)
then
results
relation
weak
of
results
well
strong
standard
site
F.
This
determined
the
Fig.
Mesozoic
estimating
Miocene
sites
region
3
California
surficial
extended Su,
amplification
trends
age
factors
as
motion
and
nonlinear
is
from
correlation
1
to
for
between
site
for for
than
in
Dept.
is
at
for
error
from
4
indicate
has
of
contrast the
Southern
all
both
the
effect.
for
sediments
1.5
give
geology
at
granitic
five
the site
using
soil
site
very
frequencies,
results
for
158
coda Geol.
behavior
comparison
and
each
Site-specific
weak Central
same
the
amplification
station
effect
sites
between
results
Southern
seven
high
stations
132
to
12
rocks,
is
waves
California.
spatial
Sciences,
frequency
which
the
motion
work
through
more
for
Hz,
CA,
well-calibrated
site
at
and
different
of
site
are
strong
frequencies site
a
and
although
of
sediment
respectively.
of
California
amplification,
of
90089-0740
to
complicated
shows
Southern
specific
distribution
important
groups:
the
site
amplification
the
the
185
Cretaceous
(5)
factors
Univ.
in
motion
By
USGS-CALTECH
site
amplification
Pre-Cretaceous
Southern
strong
local
the
these
Weak-motion
the
sites.
comparing
site
(1)
amplification
of
California
is
for
at
mean
stations
higher
age
larger
in Southern
amplification
earthquakes
Quaternary
of
with
the
motion
As two
especially
earthquake
Marine
Southern
California
dependence
factors
the
site
frequencies
shown
than
regions
known
than
of
Figures
with
site
show
amplification
site
the
metamorphic
California,
sediments,
ranges.
about and
that
for
seismic
in
California.
amplification
sediments,
geologic
USGS
hazard
amplification
(Su
surficial
that
region
factor
are
these
surficial
low
of
is
1
Amplification
of
5
et
younger
and
Central
stronger
almost
We
Hz,
frequencies.
1.5,
network.
seismic
aL,
analysis,
two
which
and
(4)
age
Los
factors
2,
geology.
found
rocks.
geologic
suggesting
(2)
3,
1991).
determined
we
figures,
Franciscan
provides
California
the
factors
Angeles,
sites 6,
for
results
Pliocene
becomes
network
can
that
seismic
and
Figure
versus
same,
lower
have
The
The
age
see
the
we
12
of
in
a a 0-
It)
7 o tL It)
______
1 1
AGE
10 10 — ______I
—
F—1.5
(million
F=6.O
standard
standard
standard
standard
Hz
Hz years)
•1
error
error
error
error
100 100
I
of
of
mean
mean
1000 1000 I’) - It)
0 It) iq I’, 0
0 It) 0•
I I —
______
1 1 ______,.,,,,..I
•1
10
10 F=12.O
I
F3.O
712i/
standard
standard
standard
standard
Hz
Hz
error
error
error
error
100 100
I
of
of
mean
mean
A,o
1000 1000 7 o o o U,. — 1’ — ., 1 I ______AGE 10 . 10 I I F=1.5 (million F=6.O standard standard standard standard Hz Hz years) error error error error 100 100 . I I I of of mean mean 1000 1000 . . a U, a U, 0- I , U, I I ‘—V., ______‘“1 1 1 10 10 . F=12.O I F=3.O 5tandard standard standard standord Hz Hz error error error error 100 . 100 I I I of of mean
mean FH1 1000 1000 SOUTHERN CALIFORNIA SITE EFFECT (F= 1 .5 Hz)
30’ 20’
10’ ln(A/A) 36° 0 -2.5 -- -1.0 50’ 0 -1.0-- -0.4 40’ 0 30’ -0.4 -- 0.2 20’ + 0.2--0.8 10’ 0.8 -- 1.4 35° . 1.4 -- 2.1 50’ 2.1 40’ 0 --3.2 30’ 20’ 10’ 34° 50’ 40’ 30’ 20’ 10’ 33° 50’ 40’ 30 SOUTHERN CALIFORNIA SITE EFFECT (F=12.O Hz)
ln(A/A) o o -1.0-- -0.4 o -0.4 -- 0.2 + 0.2--0.8 • 0.8--1.4
• 1.4--2.1 •
1
map
sets.
developed
version
the
which ARCINFO
sheets California Many chose
geologic
and pipelines etc.
agencies Barbara, data
additional
government center.
sets
GIS
Charles
compilation
even
It
such
Working
Our
information
an
2)
2)
We
ARCINFO
1)
1)
this
universities
supplies
completed
also
This
which
G.
is
maps and
In
It
It
at It
plan
etc.
offer
have
as
a
already
overlay
does
thus
database
is
Sammis,
is
c) a)
b)
is
its
allows
all
planning program
front-end
Caltech
geology
Group
makes
not
currently
ARCINFO
As
ESRI
should
for
of the
current chosen
courses
levels,
the
well
of
have
is
over
data
ARCINFO WAVEFRONT
when
Southern
a
(with
a
dealing
have
industry
exist.
of
the
currently
geological
program)
stopgap,
for
is Steve
true
Report it
-
suited
eliminate
some
allows
active
infrastructure
topography,
as
ARCINFO
program
agencies.
developing
form
we
and
calculation
being
in
an
it
the
updated
ARCINFO
uses
3-d
required.
with
is
Park,
are
its
ideal
California.
maps
drawbacks:
to
standard
data
up
following
it
we
faults,
the
released
database.
used
is
the
data
use.
installing
the
is
this
this and
called
will
Steve
to
vehicle
files.
currently
geology)
overlap
as
of
ORACLE not
a
of
assessment
It
but
bases.
by
These
produce
in
topography,
running
objection.
true
problem
the
such
the
work
in
spatial
is
user
Doug
Miller,
reasons:
the
ALACARTE
their
It
not
it
the
principal
widely
3-D
for
infrastructure
and
has
as
systems
with
later
USGS
testing
friendly.
by
database
to
at
3-D
Morton
geography
interaction
geography
is
variables
GIS
highways,
and
display
of
the Also,
UC
1995.
visualize
as
STARDENT
this
used
depth
visualizations
seismic
GIS
to
which
capability
a
Steve
follows:
can
Riverside,
ESRI
which
week
plans
which
the
produce
menu
22
by
of
data
based
to
in
use
community.
departments
with
we
true
USGS
bridges,
Richards
government
October,
multiple
hazard.
water
Southern
(the
at
to
is
simplifies base
driven
the
will
and/or
to
USC.
on
UC
true 1:100,000
have
3-D
of
company
has
drape
table,
for
these
use
Santa
The
center
3-D.
data
data
1991
20 We the
fit. Other for
feels
will
file
also coordinates.
sheets.
Current
such
based
produced
be
that
We
We
Collection
In
Implementation
investigators
Steve
compatible
projects
order
labeling
plan
are
it
on
9) 7) 4)
5)
6) 3) 2)
8)
1)
This
is
Park,
currently
an
the
Topography
All
Detailed
Detailed
Physiographic Active
Simple
Water
to
to
very
Infrastructure
of
is
area
attribute
so
store
corner
overlay
faults
in
are
Database
done
with
important
that
of
collaboration
table
faults geology
-
free
mapping
working
geology
the
ARCINFO
Steve
using
the
coordinates maps
-
maps
code
to
database in
-
layers
boundaries
center
we use
-
for
hand
Park
a
produced
near
for
on
they
Alluvium
along
“tic
have
the
with
any
the
database.
labeling
at
in
of
SanBernardino
file”.
must
center
UCSB
the
of
Wesnowski’s
read-only
Doug following
-
the
by
in
the
vs
SanAndreas
Steve
contain
7
the
hand
government
to
bedrock
Morton
data
1/2
establish
geology.
Park
format
layers:
at
minute
layers
reference
as
files
at
Caltech
has an
-
the
at
Kerry
geologists
The
a
as
example
quadrangle
created
Caltech.
standard
USGS
they
group
Sieh
see
this has will unavailable Second, the ARC/INFO The to next next form, ARC/INFO the between quadrangle currently room mapping for digitized an are for done incorporating such attributes and meetings the We Date: Project: the specific Institution: P.1.: have Results 1:750000 SCEC. earlier attributes approximately basin stored geologic purchase be have database 36N year final year. Stephen as negotiated with for and October available Wesnousky’s results from I Digital Quaternary, have he geology is showing to selected on growth result will which the in one and the All First, to University (Redlands) anywhere discuss is stores database a has latitude of K. is Wesnousky’s this stabilize work in resulting three much I remaining supervising. between used for 19, a Geologic will multiple an currently Park tell for subsurface as order been high is $1500, the ARC/INFO project the 1991 data institutional progress. with new of the also by a the smaller slip and Tertiary, else geology as of priority the San compilation the temporarily is database. the to region in 114W user research data, a Steve be and California, copies shown longitude Map three have than fall six map subset two data examine Bernardino USGS geology Primary used scale what manuals quadrangles Some for of into for and ways. Database between Wesnousky as from center is will and contract into in of southern by of eventual a the directions I than SCEC Figure in ARC/INFO. is pre-Tertiary have in three of installed disk the be Riverside effort 122W. the and liquefaction Doug SCEC. emphases the The Maine, geographic our are operation both basin. he done the Working geologic with drive developed maps for because feature. 2. categories: geographic for database. $350. has is inclusion California Morton in San published are At We Southern on spent and Earth the by implementing attached used of in that units Jacinto realized. at We will a year’s maps, Anyone San information Group potential the this we the Desigu borrowed UCR in (USGS-UC previously. a Systems point, in have linked pursue and water Bernardino features draft is are evaluation will coding California organization; the and for end. to and available. A by in now spans also of determine the the Center’s of example. and unpublished obtaining in table two his Research San to SCEC a Sun system An an system geology (lines, borrowed formed revising standardized the the Riverside) hold My seismic the avenues Andreas attribute of basin will example workstation geographic This San can database. highest other region database regular, will points, what In (GIS) equipment be Institute a at (Figure this now hazard addition, working Bernardino map workstation. maps of a incorporated be fault digital system research for scale between of priority research coding for coding and obtain flexibility if distinguishes a We the features design; at zones. and analysis infrequent, 1) (ESRI) SCEC completed of polygons) data group dedicated based a UCR. in are regional 1:24000 system. map for system a can is 32 digital in basin. copy sets, now thus and The and into and this this are for for for on be at N
I %-i4
are
Figure
dotted
1 \) I
-
Location where _\ /
—
inferred. map
\‘\
\ S
‘
B
\
for
,\ DO
..
Symbols
c. study
\ r •1
-
area. used
—
Basement
are:
RDL
SBDO,
rocks
San
are
Bernardino;
shown
(I,-..
with and
___,_\
pattern, RDL,
i
Redlands.
and faults
Figure been
[
lumped
2
Tertiary
Quaternary
Quaternary Quaternary
-
Digital
together
(landslide) (pleistocene)
(holocene)
map
for
for
presentation;
Redlands
N
Fault Pre-Cretaceous Cretaceous
Fault
7.5’
(inierred)
the
quadrangle.
stored
map
Quaternary
is
0
more
complicated.
units
2km have Year 1 Annual Report, Oct 22, 1991
Mapping Participation in the SCEC
Stephen P. Miller
Dept. Geological Sciences, UCSB, Santa Barbara, Calif.93106 805-893-2853, [email protected]
During the first partial year of operation (May-October, 1991) we have established a SCEC GIS mapping and 3-D visualization facilityat UCSB. Color output is available on a filmrecorder and on a 36 inch electrostatic plotter. The following goals have been accomplished:
1. ARC/INFO GIS software has been installed on a Sun Sparcstation 2 GX computer, including all available software options, and has been upgraded to a multiple user license. The machine is accessible on the Internet as rapa.geol.ucsb.edu (128.111.108.22). See Stephen Millerfor the SCEC username and password. Before sending a plot, SCEC users need to telephone 805-893-4704 to make sure the plotter is actually turned on.
2. A 670 MB disk has been added to the Sun workstation for SCEC usage.
3. A parallel interface has been installed to enhance throughput between the Sun and the CalComp 5835XP color electrostatic plotter.
4. ARC/INFO coverages have been exported from UCR, and preliminary maps have been rendered on the color electrostatic plotter, depicting Southern California fault and basement types.
5. Software has been developed to read USGS 1:250,000 DEMtopography grids. Four panels have been combined, along with hand-digitized seafloor bathymetry, to create a composite LAbasin terrain model. The sample spacing is approximately 90 m.
6. 3-D color shaded relief images of the LAbasin are being rendered on the
UCSB Raster Technologies graphics system, with output to a filmrecorder. Eikonix
Digitizing
(realistic
Scanner
3-D
shadIng)
Tables
Adobe
and
light
9-Tracks
4-D
UCSB
UCSB
Illustrator
Exabytes
sources,
viewing
Macintoshes
Color-fill
Micro
Mapping
Suns
Mapping
VAXs
Fiber
CERFnet
ARC/INFO
contouring
Optics
Hardware
Software
CalComp
Raster
Laserwriter
electrostatic
Tech
DI-300
ERDAS
color
L-DGO
3-D
Plotter
Graphics
GMT 3-0 imaging of the LAbasin, based on composite USGS 1:250000 OEM topographygrids(work in progress). Seafiooi data are being digitized to provide a complete gridded repre5entjon. C.t £_.I_.JJ. AC.)’ CI ,.)_.i I S./ U t.L I..I_t..d &C ._-_._I_..
_1Ah4S7TUTEFORCRUSTALS11JDfES
SCEC Progress Report
Design and Implementationof MasterModel GIS
Stephen M. Richard, P1
Oct 17, 1991
During year one, I have developed an method of implementing the seismichazard assessment techique of Wesnousky (1986) using ArcfInfo. A summary of the algorithmto be implemented and a discussion of the nature of the data stsuctujewere presented to the groupat the Iuly SCC meeting, with a request fo coi3mlents.As none were received, I am proceeding with the implementation at present. I attendedmeetings of the MasterModel group at U. C. Riverside and at USC to discuss the nae and goalsof the database structure. Between now and the end of 3anuy the implementationof theWesnouskymethod will be completed,and his data base for activefaults will be integratedinto the MasterModel Arc/Thfodatabasebeingbuiltjointlyat UCSE, USC and UCR.
UNIVERSITYOF CALIFORNIA,SANTABARBARA,CA 93106-1100 (805) 893-8231 FAX(805) 8938649
TOTAL P.01 Fl INST TECTONICS—UCSC 18.25. 1991 18:28
Living with the San Andreas - K. McNally
Cooperative Arrangement Between SCBPP and SCEC:
SCEP? is to form a cooperative agreement with the Southern California Earthquake Center to assist the Center with its outreach program. In the broader sense, outreach consists of the translation and transfer of scientific information generated by Center investigators to practitioners. These practitioners include design professionals, hazard managers, educators, journalists, risk managers, policy makers/elected officials and other research scientists. More specifically, outreach would be implemented through the assignment of staff and development of a work program mutually agreed upon by SCEPP and the Center,
A basic element of the proposed arrangement is the assignment or hiring of a professional staff employee whose position would be funded by the Center. This person would provide liaison between SCEPP and SCEC and would work with a multi-disciplinary advisory committee to formulate a general plan for SCEC outreach.
The outreach program will be built around a number of anticipated Center outputs or products. These outputs Include new scientific information which will contribute to a better definition of seismic risk in southern California. This risk Information may take the form of revised earthquake probabilities or new information on faults, liquefaction potential and soils amplification. The Center will be heavily involved in the analysis of all significant earthquakes which occur in the region. They will also follow developments and work toward improved earthquake prediction techniques.
The outreach program may include:
Workshops for scientists regarding the needs of practitioners and the public which can be met by Center research;
Earth science education for practitioners and the public (primary, secondary, undergraduate and continuing education). This program element also includes curriculum development and mechanisms to update curriculum;
-ur1
desirable.
Work
provide government, Development
Public
user on
the
SCEPP California. earthquake SCEPP
SCBC/SCEPP information
IN5T
a
outside
groups.
earthquakes
Center communications
for
Coordinate preparedness
information designed
and what Emergency Conduct
Outline Develop a
assessing work
Development
publication
out schools
TECTONICS—UC5C
education
and
updates,
possibility);
the
an
role
which
on
with
SCEC
world.
private
of
brochure,
Workshops:
public.
information
a
an
(K special
the
to
SCEC
short-term
GIS.
dissemination
and
other
of
and
Preparedness
addressing
for
earthquake
contains
and coordinate have
-
June and
of
sector,
SCEC
Would
12).
extent
publish
a
with
various
workshops
could
a
Application public/private
response
a
also
28th
Geographic
non-technical
quarterly
dissemination
USGS,
Assess
scientists
seismological,
constituents,
earthquake
include
jointly
of
play
the
earthquake.
the
response
periodic
organized
strategy
Commission,
scientific
(e.g.
development
earthquake
CDMG,
in
and
curriculum
and
produced
tiewsietter,
Information
as
participation
the
sector
BICEPP,
seminars
plan
to
newsletters for
10.25.
Integration,
report
risk;
strategy.
aftermath
groups
how
etc.
involvement
These
SCEPP
geotechnical
translated
use
for
a
etc.); potential
1991
development,
and
on
their
bulletin
Schools
speciaL
System
to
the
of
publications
concerned
of
and
current of
leizO
E.g.
use
translate
compatibility and
GIS
research
Center
representatives
an
SCEC scientific
in
which
on
reports
and Task
of
to
workshops,
other
(a
earthquake
southern
the
seismic
these
merge
mechanisms
workshop
and
sociological
with
can
scientific
will Force,
may
might
Sierra
on
knowledge;
identify
be
systems
with
coflaborate
SCEC
earthquake
recent
include
potential
seminars. the
used
in
from
be
Madre
local to
F.
in
is a 3
ruI,
quarterly Plan Center
iriI
to
products.
develop
newsletter,
written
E.g.,
next
earthquake
communications
big
earthquake bulletins,
18.25.
and
what document.
1951
periodic is
18:21
SCEC?,
dissemination
general of P. 4 Group B: STRONG GROUND MOTION PREDICTION
Group Leader: Ralph Archuleta
Determination of Site-specific Weak Motion Amplification Factor Aki and Su (USC) B-I for California
Characteristics of Non-linear Soil Response: Predictions of Anderson (Nevada-Reno) B-6 a Time Domain Code
Strong Motion Data and Linear Attenuation Archuleta (UC-Santa Barbara) B-9
Strong Motion Prediction for Soil Sites Day (San Diego State) B-13
Strong Ground Motions in Los Angeles Heimberger (Caltech) B-15
Investigation of Site Response of the Los Angeles Basin using Kanamori-Hauksson (Caltech) B-18 Portable Broadband Seismographs
zonation,
understanding
weak Imperial used pervasive
simple
frequencies.
Southern
greater
greater
since
that
although
2 Formation
site
median
sediments, results
of
in
Hz
Factor
Determination
K.
shows
the
Central
have
amplification
Aki
the
seven
motion
the
station.
way
The Figures
show
Our
amplification
We
at
The
geologic
Valley
general
and
the
station
been
the
California
as
non-linear
rock
for
and
symbols
California
(3)
then
relation
results
of
weak
well
results
standard
strong
site
This
F.
the
determined
estimating
Fig.
Mesozoic
Miocene
sites
region
3
surficial
California
extended
Su,
amplification
trends
as
age
nonlinear
factors
and
motion
is
from
correlation
1
to
between
for
site
for
for
than
in
at
is
Dept.
error
for
from
indicate
4
has
contrast
of
all the
Southern
the
both
effect.
1.5
for
give
sediments
geology
at
grarntic
five
soil
the
site
using
site
very
frequencies,
for
results
158
behavior
comparison
coda
Geol.
and
each
Site-specific
weak
Central
same
the
effect
station
amplification
sites
results
seven
between
Southern
high
stations
to
132
12
is
rocks,
waves
California.
spatial
frequency
Sciences,
which
the
motion
work
more
through
for
Hz,
site
well-calibrated
CA,
at
of
different
and
are
site
strong
a frequencies
site
sediment
although
and
of
respectively.
amplification,
of
Caiifornia
of
complicated
to
specific 90089-0
important
Southern
distribution
shows
groups:
the
site
the
the
amplification
185
Cretaceous
(5)
factors
Univ.
in
motion
By
USGS-CALTECH
site
amplification
strong
Southern Pre-Cretaceous
local
the
these
Weak-motion
the
sites.
740
comparing
site
amplification
(1)
California
of
is
for
higher
mean
at
stations
age
larger
in
amplification
Southern
with Quaternary earthquakes
of
motion
the
As
especially
two
earthquake
Marine
Southern
California
dependence
the
factors
site
shown
frequencies
known
than
regions
than
of
with
site
Figures
show
amplification
site
metamorphic
the
ranges.
sediments,
California,
about
and
that
for
in
seismic
amplification
California.
geologic
sediments,
hazard
amplification
USGS
surficial
(Su
that
region
factor
are
these
low
surficial
of
is
1
of
Amplification
5
et
younger
and
stronger
Central
We
almost
Hz,
frequencies.
network.
analysis,
1.5,
seismic
al,
two
which
and
age
(4)
geology.
2,
Los
factors
found
rocks.
suggesting
geologic
(2)
3,
1991).
figures,
we
determined
Franciscan
provides
California
the
factors
sites
6,
for
results Angeles,
becomes
Pliocene
network
can
that
seismic
and
Figure
versus
same,
lower
have
The
The
age
see
the
we
12
of
in
a a h C,’ I C -S b
C 0 p UI
L
y
OL
OL —
09.d
QL—J
uoiiiiw)
ZH
ZH
pJDpUDS
popuo.
(IJD(
QOL JO.U JDJJ aOL
bLO
uo•w
000L 000L
• -S
0 bi 0 b 0 0 UI
L
OL
oc—.i
OZL
ZH
ZH
OOL
000L
000L
...... I
U,
I’,
0
0•
U,
U)
U,
0
1
1
AGE
10
10
(million
—
—
F-’8.O
—
F—1.5
standard
standard
standard
standard
Hz
year,)
Hz
error
error
error
error
100
100
of
of
mean
mean
H
1000
1000
In
I
iq
0
I 0
.
•
....I
.,..I
1 1
._
10
10
F—12.O I
—
F—3.O
standard
standard
standard
standard
Hz
Hz
error
error
error
error
100
100
I
I
of
of
mean
mean
6-3
1000
10 )0 SOUTHERN CALIFORNIA SITE EFFECT (F=1.5 Hz)
ln(A/A) o -2.5---1.O o -1.0-- -0.4 o -0.4 -- 0.2 + O.2--O.8 • O.8--l.4 • 1.4--2.J •
121°40’ 20 1200401 20’ 119°40’ 20’ 118°40’ 20’ 117°40’ 20’ 116°40’ 20’ 115°40’ 20’ 73 SOUTHERN CALIFORNIA SITE EFFECT (F=12.o Hz)
20’ 10’ 36° ln(A/A) 50’ o 40’ o -1.0-- -0.4 30’ o -0.4 -- 0.2 20’ + 0.2--0.8 10’ 35° • O.8--1.4 50’ • 1.4--2.1 40’ • 30’ 20’ 10’
340
50’ 40’ 30’ 20’ 10’
330 50’ 40’ 30’
1 U- SEHII-91
Status by Characteristics The
part the time-domain to will
Our
impulse a(t) Our
shear model
DESRA, perfect resondtea Figure with acceleration the of initial at acceleration oscillate Figure
non-linear at time much series, This non-linear
stress-strain Studies
soil series, regard obtained We little
..utly
examne
John
the low O.5
viewpoint
vismo1ogica1
affect
of surface study ED
most
—
ar.
response
a
Report: series 2. 4. energy 5. 3. I. stress
C0iexp(—).t),
as
smaller represents
surface
the pulse
range
frequencies,
where 1 2 decreases agreement. compared Anderson,
to Spectral Impulse in SensitIvity Effect Effect
2 the now
as
one executing show,
ihows with extensive
9:53
is
to
strong
through
complex question
acceleration, the
cases the casep shown
parameter is
peak separated of of
is code,
this is
relationship of month. completing
of
will of the about
In
decreases
a characteristics
of broadened of
introduced fundamental the the
the
the input
response
fundamental ratio seismologlats
non-linear
this 20 community motion. point. are actuaUy the
acceleration
niffn.ss to non-linearity non-linearity
cause Yu
DESKA. in
5
of
In input input.
studies
4, site acceleration
Fourier below
with m
1.5
the are
reduced
standard CKS
part Figure
tests
the the peak preliminary
increase Guang, into
of
response
inideot
tImes
due
We sail response
our
7 increases,
that
hive
unconsolidated linear above
by
of
to
Thus, about in
hive
Is several
selected
accelerations.
soil amplitude
anticipate
frequency We
to
1. has
calculations Incomplew. the decrease, of the SCHOOL suddenly
this
frequency who rilative
Ii at th.
arid
deviation
smoothly
more
the
on on
time
The thu are mode, been ped1ctions response
assumed
from
the of
2.5 seismogram become
for non-linearity,
of
model peat
results parts: Raj want
prethotability
non-linearity,
DESRA
strong to
not
surfac. non-1.inearity
a spectra
Hz. linear energy
series
OF
on
to
below.
have
linear
spectra that
as Increased Slddharthan against
predicted
acceleration
judging
to
of in
th. with
at
the SES
saud inoreasingly
and
for
have might
between
predictions
With
The
motion
understand
peak end about
the predicted
the
a
to
predicts
in
linear
show is
reipons.
sytcm.
the
preparing corner depth
The overlying of
predictions model
reduced figure
demonstrated
the first
manuscript
rajd
while small
be
the
in
of
calculations.
by 1(t)/Oax
3
until
very e.ccelerographs.
as
model. about
normalized input
Inferred
Hz.
the at strong
sign due
cade,
the
that frequency
predicted
as parameters ampLitude of interested shows RX
the of
the more to
for
little by soil
a the
Uear We a
to rock.
a
change pulse
25
the questlee
coda
manuscript
but of base O. tiiiie
sediment motion a
might At a
overburden
corresponding
deposit
surface,
from have :(t)/a11, thoroughly
factor
Hz effect
pulse of that
amplification a
frequencies only
spectra
analysis has
The of by O241166 Haskeli
of
of domain
an n motion,
and
of
be
the tested
tho
the
the a 3, about
The the non-linear with
ensembl,
of from
the according examining
shear
layer
widely
s(
ready 20
ii
non-linear
spectra to seismogram
modeL than 2.
shape:
where since program, pressure. appears strain
object
cod.
now
Hz,
Input the
describe
the
20
the
velocity At for
to
to
above
for
of
the
used Hz.
finite-element
the peak
a
non-linear
of non-linear she the the
in
a
and
a.,,.
to peak peak soil
of
As horizontal th.
submission
linear
several
that
ranaom
secti-a
Figure
response
the
strains and acceleration The same 20 with
the this
The
acceleration
the
continues velocity and
response
output
lithe
acceleration
acceleration
Hz,
nnn-linear
hysteretic
results
accepted
velocity
study peak prethctions.
elmost layer
max.inium time
pulses
2.
of
effects
are
response time
hi peak
from
sigasfi
In
tit,na the
Wi!h
code
.8 B-4
is
the
to
as
at SP—i1—91
I 0
4D
—0.5
—1.0
0.5 1.5 oo 1.0
9:4
0.0
SCDDL
0.5
OF
MENES
1.0
Trne
Figure
NO
(sec)
7C2’841
4
1.5
‘ob
2,0
13-7
2.5
vipalg
amjtdy
LawwJ
pazt7rnLLLOf(
0
0
0
Q
Q
Q
I
I
0
I
I
I
(‘1
I
I W
:2
-
- aD
UCf—.3i91
auto-regressive
nearly
Yerba Although
Prieta
Arias The
accelerogram
(BRN)
regressive
covariance slope
determine concept series
the
history. possible
method previously
deployments; Howard
southern
earthquakes.
recorded
Valley
have
related
earthquake
Golden,
connect
existing wheel,
earthquakes,
(STRONGMO), portable
site.
auto regressive
intensity
by recorded
at
Buena
earthquake.
identical
This
InstUution
with
Title Name
Aftershock,
we
to
that
is
strong-motion
nonlinear Co.
The
the
removing
to
the
Bundock),
three
California.
i23
recorders
method
We
a
the
which
and
recorded
decided
project
of
of
the
nonlinear
simple
the
of
Arias
focus
Using
Island
onset
time
basic
however,
because
1952),
Project
method
have
assignment
for
auto-regressive.
or
method
P1
Arias
four
We
FUM
existing
is
effects
provides
histories
M
more
involves
the
on
intensity
high
idea
that
of
In
method
a
use
have
large
(Franciscan)
magnitudes
best
been
we
large
with
6.7).
simple
intensity
the
of
Figure behavior
two
digitized
nonlinear
and
The
Crusti
there
of
to
the
frequencies.
Is
earthquakes,
SCEC
the
we
also
have University
—
Strong
Ralph
effect
earthquakes.
examining
However, NCEER
sites.
of
Arias
that
rock
earthquakes
the
by accelerographs
three
well-determined
can
differs
portable
most
ftp
Fourier
have
efforts
are
1
names
connected
random
auto-regressive
ANNv
Stuthes
for
a
of
we
the
will accelerograms.
also
command
Although
Archu.leta
and
Motion
(Parkfield,
intensity
We
and
response.
obvious
tasks:
efficient
complete
been
local
by
an
compare
nonlinearity
strong-motion
of
lead
be
of
we
are of
amplitude
alluvium
Thus
recorders
Treasure
a
and
various
aUuviu.m
the
Uu.b
California,
vibration
related
site
looking initiate
factor
In
earthquakes
REPoRT
have
to
Data
in
now
to
sites
plotted
means
over
each
order USGS
76
the
the
M
conditions.
a
southern
that
listing
peaks
The
Arias
looking
shift
been
and
of
stations
3.9)
Arias National
method
time
looking
sites.
to a
There
Island
Internet,
at
method
have
site
spectra, 2/90
100,
should
to
strong
techniques.
in
of
second
the
have
Santa
Tu
along
Linear
two
and
in
in
intensity able
examine
Menlo
database
establishing
of
signal
(ASH,
intensity
the -
(e.g.
the
current
are
more
a
California;
not
The
1)91
have
(fill).
different
combinations
to
at
strong-motion
smoothed
Rather
to
motion
Geophysical
has
Barbara
strongly
recorded
we
shape
with
MUSIC,
Attenuation
fundamental
problems
method
data
the
The
determine
been
auto-regressive
Park
Agnew
closely
analysis
recorded
are
its
strong-motion
for
We
method
We list
may
the
at
from
of
than
Southern
strengths,
data
methods.
able
(Paul
a
released
a
dampen
have
the
of
22
Lamont-Doherty
have acceleration
power Pronys
is
nearby
at
database
show
in
previous
State
reinventing
the
sites
Data
base,
(1940
to
techniques
that
stations
Arias
based
the
of
the
two
Spudich
frequency
applied
applied
retrieve
stations
1989
simulating
a
Hospital). California
data
spectrum.
8
databases
that
the
Center
rock for
co-locate
the The
method, Imperial
break
or
data
curve
method
stations
on
was
Loma
auto
more
large
from
have
have
field
time
time
first
and
site
the
the
the
the
the
for
in
in
to
at
to
is
at B-9
method
genetally alluvium
defirdtely
with
i.-.’4
sites
dominated shows
weak
have
rrLM
differences
motion
strong
by
rLSt
a
recorded
single
peaks
between
DtuC1
peak
around
at
rock
the
near
same
sites
1.0
2.6
Hz
sites.
and
Hz.
and
alluvium
We
2.6
Hz
will
sites
while
now
(Figure
rock
compare
sites
2).
The
this
are B—to
ui—a—1W1
t
14:24
‘C C
‘C C
-b
‘C C
+
FRC$i
Crustai
Studies
1W
ZL’Wa-
r. 8.-li —. ‘t. I
-l
-H C
g
.4-
9 C (1
• •
- . - -•-- ••-• .rt_r3r.RM1mt
continuing.
relative
strong earthquake
Source
event
analysis effect
component
use
this
Thus,
substantial motion
successfully
measuring
13 (“Hotel around
Valley, digital
Stardust
Field
source amplification
epresentatlons
(1) empirical
Objective:
source which application
P1.:
Institution:
km
in carried
site,
obtained
experiment.
is
this
ground
estimating
and data
may
models
the
to
models.
Circle”).
and
[tithe
simulations.
of
which golf
evident
Mesozoic
estimation
deployment
site
the
as
golf
problems
150
for
logger
lead
of
out
recorded
wave
course,
Among
an
effect
motion
city
S
responses, for
methodologies
remains
on
Steven
San
the
km.
of
course
to
waves
an
assemblage
and
in
Previous
site
Mesozoic use
with
soil
propagation
of a
15
Diego
i:
the
respectively.
crysalline
at
experiment
During
located
practical
there
of
with
2
the
San
on
in
M.
amplification
km urban
During
and
is
magnitude
indicates
unidentified, a
provides
time
site-specific
the
rock
short-period,
quite
Day
primary
with
Strong
State
is
Diego.
event,
noise.
alluvial
work
a Green’s
within
a
crystalline
Sites,
of
domain
predictive
sites)
two
major
for
sites.
portable,
the
developed,
University
subevents
problem,
Motion
a
a to
and
has
Fortunately,
the
Both
using
2.6
goals
spectral month
The
two-month reasonable
about
exceeding
sites
test
function
amplification
was
regional
in
Our
amplitudes
compares
prediction
October.
earthquakes,
shown
Status
3-component
data
the
of
rock
portable
event-triggered
method
Prediction
of
intermittent.
the
600
is
extension
experiment,
can
via
the
with
amplification
2
in the
logger
Report
in
summation
shopping
feasibility
meters
to
reproduce that
2
10
Greens
feasibility
the
however,
1991
experiment
the
an
SCEC
recordings
this
40
is
of
data
over and
factors.
code-development
adjacent
to
a
at
for
eastern
Is
strong
Hz.
of
sensor,
of
recording
source
epicentral combine
loggers,
durations,
cvent-thggered.
During
much
this
Soil
we
center
are
Interstate
function
instnnrientation,
band.
method.
many
of
factor
uCt
the
test
deployed
ground
the
work In
part
Sites
had
in
concentration
at
directly
of
model
most
less
the
a
for
and
the
theoretical
this
of
£g,i
development,
the
to
Figure
relatively
(Mission
distance
of
sufficient
and
to
8.
summation,
current
the
recordings the than
western
(ii)
motion.
2
the
severe
Incorporate
site,
The
which
to
a
preliminary
features
measuring
strategy
begun
stage, city.
ICincinetrics
2
1
10Hz
The
of
area
kilometers
we
shows
for
project,
Valley
quite
treatment
noise of
signal
part
approximately
represents
A
site
An
use
investigating
encountered
major
immediately
and
of
bond.
testing,
of
with
of
of
period, strong
improved
the
source
the
approach
alluvium
recorded
in
is
level
we
Mission
directly
the
work
Fourier
on
strong
hotels
away.
SSR1
direct
of
north
same
have
site
site
and
the
for we
the
an
at
is
1.-j.. B-Is 1
+
‘C 4 If r
+
‘C
+ = C ‘C C-,
U I—
p -I
C C C I
Fig 1. Rrded ground motion (north-south component) for site on Quaternary alluvium in Mission Valley, San Diego, Caiifonua (top ucc). axnpared with ground motion I on Mesozo crystalline sites in eastern San Diego (lower 3 traces).
however,
parameter
layers, The
synthetics deeper of
is and
SR of record receivers.
is
seismic amplitude
such Basin,
the
with velocity
structure variations velocities
This
provided
significant used
of more the
Results:
also region
and Objective:
also
reflected
the
the
the
phase
October
area.
dip
fault
timing
be
very
data
short
to
accurate
direct
dipping
L.A.
displayed
just both
component,
which
in
used
create
velocity of
gradient
of
Large
defmition.
by
the the
Better
set
and low
of
in
for
(FD)
of
the
range A
because
above
Basin.
of once
A
the
16
4,
S
one
to
the
the
the
distinctive
is
records
crust,
have
focussing
considerable
the
the
layers
a
shear
layered
and
2D
1987
km
constraint
not
basin
variations
with
clearer
structures
solid
seismic
and
of
first
off
with
at
from
pulse
first
the
and
from
direct
been
The
essentially
only
the
the
see
the
velocities
Whittier
the
thickness
were
Our
observations
could
between
multiple
earthquake
bedrock,
Strong
a
model.
multiple.
3D
sediment
bottom
well-recorded
reflected
the
picture
low
flat
relative
free
feature
the
valuable
the
recorded
California
structure.
P
major
the
velocity
in
varied
allow
lower
number
arrivals
only
layered
multiple
Pasadena,
epicenter,
seismic
surface
seismogram
Ground
aftershock
lower
the
the
in
of
By
of
of
of
and
of
efforts
to
to
be
seismic
more
The
the
the
Don
hypocentral
to
panel
in
the
by
figure the
phase
earthquake
forward
where
the the
models
basement
Institute
These
was
of
strong
model,
fit
complex
fit
velocities assessing
before
top
conventional
horizontal
Summary
into
dip
rupture
shallowest Motions
Whittier
events
Heimberger
records
direct
California
see
by
detailed
the
to
of
included
from
SS.
1.
few
energy
many was
waveforms
a
date
increasing
effects
figure
figure
reflections
timing
will
modeling,
distinct
so
Figure
reaching
of
SH
basin
layers
have
rock.
It process
found
two the
depth,
source
it
from
of
has
in
aftershocks.
Technology
Green’s
increase
is
of
shear
was
may
arrival
layers.
Los
the
1.
1.
as 91125
have
nature
between
difficult
the
stations
SMA
centered
occurred
2
and
pulse,
The
was
the
an
The
A
to
displays
taken
to
the
sediments
the
be
(outside
are Angeles
(source
the
of
wave
details
structure
been
be
additional
ridge
functions
values
basin
was
the
shape
focussed
adopted
seismic
velocity of
strong
When
best
receiver)
generated
seismic
a
as
to
located
the
as
seismic
on
very
(a
precision
in
modeled
controlled
seen
model
structures
are
fitting
physics).
receivers
the a
the
body
expected
of
arrival
the
structural
the
primary
motion
contrast
of
energy
to
sensitive
explained.
the
of
comparison
which
basin)
velocity,
by
in
constraint,
on
Los
in
dipping
absolute
by
approximate
sources
this wave
the
model
the
the
basin
the
the
for
of
of
the
Angeles
by
instruments.
is
common
deeper
Secondly,
constraint
occur
with
5 records.
can
multiple
and
the
locations
lateral
L.A.
constraints
records
tangential
the
ray
to
structure
the
is
depth
layers
to
but
timing
direct
be
the
10
the
large also
date
of which
at
can
km
the
the
to
of at Whittier Narrows Aftershock (Oct. 4, 1987)
34.2
-J
I.. z0
33.8
-118.4 -118.2 -118 -117.8 We5t Longitude
0
N S
7
* Vp. 6.760 V*. 3.O0 d•rgV. 2.750 —. VP. 6.410 Ve 3.700 •r1ilV. 2.720 Vp. 5.370 Vi 3.100 OInhItV 2.700 Vp. 6.020 Vi. e.Qoo O.r%IIIV. 2.600 4.600 Vi. I Iv. e.soo dñii. 2.500 zi —I 3.980 Vs 2.300 O•ni!V 2.400 Vp. 3.290 Vs 1.900 derStV. 2.200 2.670 Vi. 1.360 •ni’w. 2.OSO 2.050 Vi. 0.950 i.eoo
Figure 1. Upper panel displays station locationswhich recorded the magnitude ML =5.3, Whittier event. Most of the observations are simple except those in the basin. Lower panel displays a 2D model showing the position of the event relative to stations. (after Scrivner & Helmberger, 1991). Realand SyntheticWaveformsfor DWN,Velocity
Vertical Radial Tangential
E ‘3 44r - 4Dcto I 2iliIIliliIIIil I ‘I I I’ ‘I’ H*MH
Synthetic ,c — lI3I1I3I1II 1111I31313131 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 Seconds Seconds Seconds
Realand SyntheticWaveformsfor BEC,Velocfty
Vertical Radial Tangential
T —I I 1
E ‘3
Doto — Il13131113131
1’
. -! Synthetic
—it’ 1 I 1 I 1 I I iL 0 2 4 6 8 10 12 14 Seconds Seconds Seconds
Figure 2. Comparison of observations at DWN and BEC with theoretical predictions from the model given in figure 1. OCT—22—91TUE14:15 CALTECHSE1SMOLOG F Nu, 1bo4Uiib ru Hg
PIs Hiroo Kariamori and EgiH Hauksson
Institution California Institute of Technology
Title of Project Investigation of Site Response of the Los Angeles Basin Using Portable Broadband Seismographs
A SCEC Project_Goals and Progress R.eport. 16 Oct. 1991
Instrument testing. We deployed two Ref-Tek instruments at the Pasadena Station (PAS) from March to July 1991 to compare and calibrate several different sensors to the Streckeisen STS-l. The sensors being tested were: 1) two 3-component GuraLp CMG-4 broad band seismometcrs; 2) a 3-component broad-band Ranger; and 3) a single component FBA with veLocity output; and 4) 3component short-period Ranger. The results of the test showed that the Guraip CMG-4 were not suited for monitoring teleseismic events to obtain the response of the Los Angeles basin. Subsequently, the Guraip CMG-4 seismometers have been returned to the factory and they will be replaced by Guraip CMG-3ESP seismometers. The CMG-3ESP is a much more Sensitive instrument, although it will not record strong ground motions on scale. This problem with the sensors has caused some delays in the actual field work initially proposed. On 27 July 1991 we inspected several potential field sites in the Los Angeles basin with Mr. Rudy Lee of the Los Angeles County, Dept. of Public Works. On 5 August we deployed one of our Ref-Tek instruments with the broad-band Ranger in the the Imperial Yard, near the intersection of the Rio Hondo Flood Control Channel arid the Imperial Highway. The broad-band Ranger had performed satisfactorily in the tests done at PAS. So far we have recorded several teleseisms at the Imperial Yard site, although it has proven to be noisy. In the near future we plan to move the instrument to another site, which is the LA County Sheriff station on Firestone Blvd. Enclosed are two Figures showing seismograms, spectra and spectral ratios for both the CMG-4 and the broad-band Ranger. The spectral ratios show that the ranger performs much better a longer periods than the Guraip. 0
C C
r’ R Cl
C n
4D
C
C r
C
Cf c-I ci C r C ‘Mc C
tfl
‘Mc -I
C
0 C-) C-) C C
C, 110
C
io_2 10_i 100 10_i io0 10_i 100 ‘l Frequency (Hz) Frequency (Hz) Frequency in (Hz) -o ‘7 0 w
C C
ft
C rr
4:
ci
C r rr C
U rr ‘7 U 0 C r C ‘C C
I 0
‘C-
C) -c
10
-t C S 1o1 100 io2 io 100 la_I io Frequency (Hz) Frequency (Hz) Frequency (Hz) 0 Group C: Fault-Zone Geology
Group Leader: Kerry Sieh
Seismic Hazard Potential of the Santa Monica-Hollywood Doland and Sieh (Caltech) C-i Fault System
The Role of Active Crustal Shortenmg and Left-Lateral Rubin and Sieh (Caltech) C-2 Faulting along the Central Transverse Ranges, CA
ARC/INFO as an Analytical Tool for Problems in Paleoseismicity Sieh and Lilje (Caltech) C-3 and Neotectonics
Structure versus Earthquakes along the Southern San Jacinto Fault Zone Seeber and Anders (Columbia) C-4
Tectonic Geomorphology of the Los Angeles Basin Keller and Pinter (UCSB) C-7
Paleoseismic Studies on the Palos Verdes and Whitter faults Rockwell (San Diego State) C-8
Marine Terraces and Earthquake Occurrence Ward (UCSC) C-9 ct-i
SEISMIC HAZARD POTENTIAL OF THE SANTA MONICA-HOLLYWOOD FAULT SYSTEM: A PROGRESS REPORT
J. F. Dolan and K. E. Sieh (Seismological Laboratory, Caltech)
Over the past three months we have focused our efforts on a geomorphic study of the WNW-trending Santa Monica-Hollywood Fault system (SM-HFS) in order to elucidate the kinematics, subsurface structural style, and seismic hazard of this portion of the northwestern Los Angeles Basin. Our studies reveal that the NNW-trending Newport-lnglewood Fault (N-IF) separates the SM-HFS into two geomophologically and structurally distinct segments. To the east, linear scarps indicate that recent activity on the Hollywood Fault is concentrated along the topographic mountain front. Active deposition of numerous small alluvial fans at the mountain front, coupled with the absence of any significant fan incision or segmentation, suggests recent, rapid uplift of the Santa Monica mountains in this area. Similarly, the absence of geomorphic or structural evidence for active uplift/tilting south of the mountains in Beverly Hills and West Hollywood indicates that the main active trace of the Hollywood Fault daylights at the mountain front. In contrast, the active trace of the Santa Monica Fault west of the N-IF forms a series of left-stepping, en echelon scarps 3-4 km south of the mountain front. The presence of several large, deeply dissected and highly segmented old alluvial fans between the active fault and the mountain front suggests relatively little, if any, recent mountain-front uplift. The oldest and most steepiy dipping segments of the old fans lie at the mountain front. These surfaces dip only 2-2.5°, indicating that no appreciable tilting of the mountain front has occurred since fan deposition. This absence of active tilting implies that the southern Santa Monica Mountains do not overlie the forelimb of an active, south-verging fault-propagation fold associated with a blind thrust at depth. Rather, most recent contractional deformation is apparently accommodated along the en echelon scarps that cut across downtown Santa Monica. Consistent up-to-the-north scams along the entire SMHFS, coupled with growth of the Santa Monica Mountains (evinced by geomorphic features such as the Cahuenga Pass wind gap), clearly indicate that the Santa Monica and Hollywood Faults are, at least in part, south-vergent contractional structures. However, the trend of the Hollywood Fault (about N7OE-N85E) is essentially parallel to that of the Raymond Fault along strike to the east, which Jones and others (1990) have, shown to be a predominantly left-lateral fault. This observation suggests that the similarly oriented Hollywood Fault may also exhibit a significant left-lateral component of displacement. Our geomorphic analysis failed to reveal any evidence of left-lateral displacement along the Hollywood Fault, however, possibly because active deposition of alluvial fans over most of the fault trace at the mountain front has buried many of these features. Although obvious left-lateral geomorphic indicators (e.g., laterally offset drainages) are similarly absent from the Santa Monica Fault, the left-stepping en echelon pattern of faulting in Santa Monica does suggest an oblique left-lateral component to the Santa Monica Fault. Resolution of the obliquity of displacement on the SMHFS awaits three-dimensional trenching studies. The active traces of the Santa Monica and Hollywood Faults trend into the N-IF at high angles, and are separated from one another by 1.5-2 km. The N-IF thus appears to locally be acting as a tear fault between the two segments of the SMHFS. Whether or not the N-IF acts as a segment boundary for the SM-HFS during earthquakes remains an important, unanswered question. The role of active crustal shortening and left-lateral faulting along the central Transverse Ranges, California Charlie Rubin and Kerry Sieh, Seismological Laboratory, Caltech, Pasadena, Ca.
Current geologic studies on the Frontal Fault System of the central Transverse Ranges suggest that the Sierra Madre segment is an active tectonic feature and represents a significant seismic hazard along the mountain front. Geologic mapping has concentrated on Quaternary geomorphic features using the 1928 Fairchild aerial photograph collection from Whittier College and USGS topographic maps (6 minute series, 1941) which were published without any cultural information to obscure topographic detail. Although Crook et al. [1987] produced the most detailed map of the the surface trace of the Sierra Madre segment and concluded, that the fault has not been active during Holocene time, our preliminary studies show that the fault has cut Holocene sediments the Holocene, our recent work suggest that the fault has been active during late Quaternary time and represents a significant seismic hazard along the mountain front. Paleoseismic data is sparse for the frontal fault system and the Sierra Madre segment is not designated as active under the Alquist-Priolo Special Studies Zone Act. Between Duane and La Verne, fresh fault scarps have been identified using the 1928 photographs and verified in the field. Numerous geomorphic criteria, including the presence of hanging valleys, thangular faceted alluvial surfaces and deformed late Holocene alluvial fans, are consistent with an active fault. In addition, three potential excavation sites have been identified along the frontal fault system. It appears that the Sierra Madre segment is propagating southward of the mountain front between San Antonio and Monrovia Canyons, whereas, to the west, active strands may still be present in crystalline rocks. The Progress and implementation * * between Systems version ARC/INFO * * The Pilot Caltech suspected along utility observations
ARC/INTO Handle Ablity System System image initial include After After maps. 3) 2) final One 1) Project the of is campus to Institute, SCEC to large
of bringing must must stages Garlock a extensive considering traces paleoseismic task handle date very the
the on - - - - and SPARCstation MByte Additional CALCOMP Hand Existing Existing and System be have was a and
pilot following of of as network powerful and SUN photogrammetric able data Redlands, in Fault. the small system to ESRI.
an testing, GIS Dr. drawn projects data trench internal build system. WILDsoft many anticipated. SPARCstation must to from published
maps analytical and Kerry Fujitsu or Her data produce and through formats: 9500 vector installation maps GIS-like a localities. the neotectonic be a platforms II California). project hard will is software/hardware and sets. wide They Sieh with able accessible following series 1.2 format, based data both maps include several disk. techniques) spectrum and to 32 include GByte
capabilities tool involves II 36”x Although were at from and publication MBytes GIS Anne (UNIX) data least Our AutoCAD unrelated synthesizing hardware
for via software external 48” made the computer manufactured sets. of purchase Lilje, the deal internet.
problems following: digitizer. and memory, sources. this as system mapping There the quality with the California combining channels formats, was SCSI packages, database week files core was and “massaging” that are ordered: These 19” drive 8.5” by including of databasing of made in of several could TIFF, (WILDsoft color local geologic is Oct this Institute it
paleoseismology x them sources quite was 11” through 1st. be system. GIF, company graphics considerations into concluded applications the output used small, The a contacts, of may EPSJ’S, data format files, a Technology a ARC/INFO to machine finalized sitewide and accelerated the (Environment set be AUTOCAD, that of results produce used observed collected DEM, programs binding files
and is we database. licensed to located in will not would DLG, monitor, manage, larger it’s neotectonics fault by the originally Research such point visual Sally agreement present on and opt traces, traditional as analyze and the out to others. McGill use the a 400
c-3
OCT
The
be product
On pervasive
expected Hills,
earthquakes. resolution
southeast confirmed We
bedrock magnetization. complicating
suggest Badlands of
structure Borrego including
southern We
ruptures.
ruptures
(Quaternary) zone
and seismogenesis offers
our
L.
historic The
their smicrural
Structure
carried
this
Seeber
22
the
made
have
attention
field
for
definition
behavior
and
far
a
sequence
‘91
basis
similar
a
of
unique
and
surface
formations
for
in confirming
beyond
California.
work
made
The
to
southeastern
and
folding
to
prowess
out
field
end
of
transeurient
and
a
the
resolve
Versus
the
psieoseismic
the
oroas
9:S9M
of
Structure
tectonic
on
factors, and
have
behavior
by
Mark
over
from conceptual
of
planned mapping,
fault
Our progress
these
Salton
opportunity
same
Saii
buried this
and
both
its
earthquake
short
of
fault
in
a
many
fault
understanding
in
termination
Feipe the
issue H
zones
the A
constrain
that
Earthquakes
large
and
rotations
portion
such
the
LDO—SEIS1OLOG’
the
Trough.
extension
the
and
paper
fauittng
of
(historic)term
hypothesized
below
Anders,
for
Ama
In
sediments,
interactions
cross
gravity
other
analysis
ruptures.
principal
data
the
and
models
Borrego
elsewhere.
normal
the
as
gravity
basin.
spring
to
sources
concerning
of
fault
SJFZ
Gap
demonstrated
the
compare
analysis
the
by
on
faults
Folding
results
in
Lamont
a
at
of
paleornagnetic
sediments.
This
concerning surface
fault.
of
paleomagnetic,
rate
weak
component
1991
modeling of
and
drag,
the
The
Badlands,
Investigators.
but
the
at
along
gravity
reach leading
for
on
these
behavior
The
the
we
of
surface,
may
of
study
is Clark
in
Lorna
long
The
has
sediments
the
-Doherty
non-cylindrical
hazard
these
rotation.
Borrego ruptures.
consistent paicomagnetic
have
the
goal
issues
into
provide
the
been
basis
the
San
data
arid
indicate
to
Bedrock
has
fault
the
sediments
The
Prieta
This
results
of
and
developed
of
distinct
the
possibility
assessment
Southern
Iscinto
short-term
techniques
resolved
segmentation
postponed
along
in
the
of
and
this
Geological
reults
The
Badlands
as
with
Equally
may
seismogenic
result
a
with
the
event
a
that
a
SJFZ
model
wiil
project
earthquake
morphology
set
results
basement
supetirnposcd
the
Borrego
sub-horizontal
is
be
fault
folding,
data
the
establish
many
of
may
diagnostic
a
in
interpreted
soon
of
SJFZ
a by
in
important,
deformation
detachment
San
in
folds
to is
for
central
hidden
extension
multiple
from
zone
Observatory
is
also
California
using
Plio-Quaternary directly
of
this
be
of
to
be
the
basement
Badlands
compaction,
from
feature
Jacinto
faults
data.
in
surprising
compare
the
identify
the
ready
the
at
reflects
late
fault
California
a
the
source
features
the
characteristic
primarily
cross
variety
characteristic
ma.gnetostratigraphy however,
the
Ocotillo
pertinent
decoupling
and
fall
and
model
sediments
into
along
relays
zone
Borrego
for
Borrego
Fault
improves
Gravity
a
northeast
crustal
of
long
faults
This the
in number
submission.
the
of
of
and
large
further
this
of focussed
and
view heavily
basins
stability
as
data
to
the
are
San
work
Badlands
the
Zone
dip
current
a
thinning fault
layers.
and
data
upper
by
the
of
of
Feipe
SXFZ.
north
in
will
on
the
has of ______
CT 22 ‘91 U9:59RM LDG0—9EISMQLQG’ P. 1i’ J.
>
40 I
I o Q =1 cI 1’ j c -cii \ extefltof basin I identifiedfrom I I • seismicdata I I I I
0
4
o 5 10 distance O) observedBouguer ••••, gravity Brawley and Ocotillo Fms. — — basement depththat gives BorregoFin. sediment beat-fit to gravity Palm Springs P. density basement geomtery —2,3g./cm caiculatedfrom JrnperiaiFm. 3 suagrapbic data 2.7 g/cm • MezozoicjPaleorofc basementrocks 3 a se&ment-basemem - interface fromefsic data W Sub.basement (after FuIs et at) — 3.1 3gfcm I.
Figure 1. An attempt at modeling avity along the San Iacinto fault zone, fram the Borrego Badlands to the Superstition Hills (Coweyand Seeber, preprint).The results are inconsistent with available stiuctural data.We suspect thc problem is with these data and our mapping effort Is addressingthis inconsistency and otherissuesbrought out by the gravity modeling. 1969 JPF 1968 AFTtRitouc.5 Al n
Al I-. C I-
10 20 30 40 50 60
9’ SicoiiD Eij N OF t1f*5t4OLJC. 13+J CoYoTli.ceK FAULT
SW
4. c1l
Rgure 2. Sketch a tectonic modd ftz the southern San Jadn)o fault zne: an cxtenoqfrotacion allochthon within two parallel mas fmuh and a (letachinent (fioni Cbwey and Steber, ptqñ). flvklenoe fr’xn gravity aal sinictaral data include left-lateral aid (kwn4o-the-soothcagcxwEçoocnts on all the cioss faiks; 20°-3(r clockwise n*ations ii the Botrego Badlands fiom lciwmgne& data; a ibsitialpmi of the 1968 main sheck m released in a aub-nt with a nearly hocmta1 nodal to plane and a !lip vector parallel to the fault zone (Pdeaen et aL, JGR in press). fli 1968 the allocbthon moved southeast relative the west wall and the flo& of the failt zone. A irstable prediction of this model is thai displacement on the Coyote Ork and Oark faulss thuld increa and decrease to the southeast, rcspcciivcJy TECTONIC GEOMORPHOLOGY OF THE LOS ANGELES BASIN: A PROGRESS REPORT
E.A. Keller and N. Pinter September, 1991
The objective of this project is to produce tectonic geomorphic maps of the Los Angeles region at a scale of 1:100,000. The maps are intended to show landforms such as fault-bounded mountain blocks, structural hills, mountain fronts, fault-line valleys, prominent marine and river terraces, and fault scarps. The purpose of the mapping is to provide a comprehensive overview of the tectonic geomorphology of the Los Angeles Basin as a whole and to delineate areas where more detailed work is necessary to better understand the tectonic framework and related earthquake hazard. The philosophical framework we are building our research upon is that surface morphology, given the principles of landscape evolution and an understanding of tectonic and geomorphic processes, reflects the tectonic and seismic history of a region. The study of tectonic geomorphology is based upon the premise that the form of the landscape is a function of process (geomorphic and tectonic); earth materials (broadly defined to include composition, texture, and geologic structure); and time. In other words, tectonic landforms are a function of process and materials integrated through time. This simple relationship provides a way of summarizing levels of tectonic geomorphic inquiry (Gregory, 1978): • Level 1: Identify of landform elements (basins, ranges, alluvial fans, terraces, fault scarps, etc.) and relating these to tectonic geomorphic processes and earth materials, independent of each other and of time. • Level 2: Obtain relationships between the geomorphic processes, the materials, and the landforms, independent of time. • Level 3: Evaluate relationships between the tectonic landforms and earth materials, integrated through time. Our project is nearing completion of Level 1 at this time, which is the identification of tectonic geomorphic features of the L.A. Basin. We expect that the project in its final form at the end of this year will comprise Level 1and aspects of Level 2. In order to take into account the variable scale inherent in tectonic geomorphology,we are mapping and classifying landforms according to a scale-based heirarchy (Crofts, 1981).
Table 1. Landform Heirarchy for Tectonic Geomorphic Mappin2 (after Crofts, 1981). Scale Examples Imagery Land region Los Angeles Basin 1:145ksatellite Land system small basins, ranges, drainage networks 1:100k topo maps Feature alluvial fans, prominent terraces 1:63k air photos Site scarps, sag ponds, etc. 1:18k air photos
To date we have completed mapping at the ‘region,’ ‘system,’ and ‘feature’ scales. This mapping has delineated interesting topography related to tectonic processes as interpreted from localized zones of incision and of aggradation, from drainage networks, and fromotherfeatures. Our next step will be to incorporate all information into a uniform base-map at a scale of 1:100,000. For local sites of particular interest, more detailed maps will be produced at a larger scale. The 1:100,000 scale base-map and accompanying explanatory text and local maps will comprise the final product of this work for the first year.
References Cited Gregory, KJ., 1978,A physical geography equation: National Geographer, v. 12, p. 137-141. Crofts, R.S., 1981, Mapping techniques in geomorphology: In A. Gondi (ed.), Geomorphic Techniques, London, George Allen and Unwin, Ch. 2.9, p. 66-75. SCEC SUMMARY REPORT - YEAR ONE
P.!.: Dr. Thomas Rockwell Department of Geological Sciences San Diego State University San Diego, CA 92182
Project Title: Paleoseismic studies on the Palos Verdes and Whittier faults
Results: Trenching of the Palos Verdes fault at Harbor Park is requiring unanticipated permit aquisitions. Consequently, actual trenching of the fault does not appear possible before the end of the rainy season in 1992. Therefore, I focused on a related fault, the Rose Canyon-Newport-Inglewood fault zone, that also has a poorly constrained slip rate and that poses a substantial seismic hazard to most of coastal southern California. Most of the 1991 summer was spent trenching the Rose Canyon strand of the zone. The northern Elsinore-Whittier fault is presently under study and trenching will continue through the end of the first project year on this fault. Both studies complement previous or ongoing studies supported by the NEHRP. Trenching of the Rose Canyon fault (complementing NEHRP support to Scott Lindvall) demonstrated a minimum of 8.7 m of dextral slip in less than 8100 years B.P. (dendrochronologically corrected), yielding a minimum slip rate of 1.1 mm/yr for most of the Holocene. Analysis of the geomorphology of the fault zone suggests a best estimate of 1.6O.5 mm/yr for the entire fault zone in San Diego. The fault parallels the coast to the northwest and is continuous (as a zone) with the Newport-Inglewood fault that bisects the greater western Los Angeles metropolitan area. Consequently, this study also provides a minimum dextral slip rate for the fault that generated the 1933 M6.3 Long Beach earthquake. Paleoseismic studies along the northern Elsinore-Whittier fault are underway as of September, 1991. 3-D trenching of the fault this fall will further constrain the slip rate for this important fault zone and, with continued study, may provide information on the timing and size of previous slip events. FROM LHST TECTONICS—UCSC 10. 17. 1991 10:04 P. 2
Progress Report, October 1991: Marine Terraces and Earthquake Occurn. SCEC Working Groups: C (Fault Zone Geology) and E (Geodesy). Steven Ward, Institute of Tectonics, University of California, Santa Cruz
SynIIctit Seisniiti,. Most of my SCEC rclatcd rcl3caivl1iii the past seven months has been directed toward demonstrating bow synthetic soismicity calculations, based on the concept of fault segmentation and incorporating the physics of faulting through static. dislocation theory, can improve earthquake recurrence statistics and hone the probabilities of hazard. I have made substantial progress in this project as documented in two Jouria1 of GcophysicalResearch papers, one accepted and one submitted. Compared to forecasts constructed fwin a hand full of earth quake rccurrencxi intervals, forecasts constructed from yntlieLk scismicity (see Figure 1) are more robust in that they embody regional seismicity information over seveLal units of magnitude, they can extrapolate seismicity to higher magnitudes than have actually been observed, and they are formulated from a catalog which can be extended as long as needed to be statistically siguilicant. Synthetic seismicity models can also be used to judge the stability of common rate eslima.tesand the appropriateness of idealizations to the earthquake cycle. I find that estimates of fault slip rate are unbiased regardless of sampling duration while, on the other hand, estimates of earthquake recurrence time are strongly biased. Recurrence iutmvai estimated from wiemicity samples less than about ten times the actual recurrcnce interval will almost certainly be too short. For the Middle America Trench, it would Lakeabout 200 and 400 years of monitoring to constrain slip and recurrence rates to ±10%, Increasing gap time generally imicreasesconditional probability of earthquake occurrence, but the effect is weak. For the MAT, the spread parameters of the best fitting lognorrnal or Weibull distributions ( 0.75) are much larger than the 0.21 ijutrhisic spread proposed in the Nishenko-l3uland hypothesis (See Figure 2). Stres8 interaction between fault segments disrupts time or slip predictability and causes earthquake recurrence to be far more aperiodic than has been suggestcd. With the arrival of Dr. Barrientos, a SCEC visiting researcher, in January 1992, and continued funding, we intend to apply this technique to the San Andreas fault with an eye toward reviewing the conditional probabilities proposed by the Working Group on California Earthquake Prediction.
Ward, S. N., A Synthetic Seismidty Model for the Middle America.Trench, Jour. Geophys.Res. (in press), 1991a.
Ward, S. N., An Application of Synthetic Seisrnicity Calculations in Earthquake Statistics: The Middle American Trench, Jour. Geophys. Res., (submitted), 1991b.
Marine Terracss. Space gwdosy tells us that 8 mm/y of roughly North-South mnnl.inni being accorn.modatcd between Palos Verdes and the base of the Transverse Ranges. The goal of this work is to employ geological and gcornorphol.ogicaldata to isolate the fraction of this ntotion on some of the faults closest to the coast. I have rcccntly met with Dr. Ken Lajoie of the USGS Menlo Park and have reviewed his substantial data set on the heights and ages of marine terraces around Palos Verdea and Huntington Mesa and other geological information on the dome structures furthcr north and inland. I don’t foresee too much difficulty in translating the terrace data, through dislocation modeling, into slip rates for the Palos Verdee, and possibly the Newport Ingl.ewood, faults. Figure 3 presents preliminary results which suggest repeat times on the Pales Verdes fault of 150 and 900 years for magnitude 6 and 6.5 earthquakes respectively. Dr. Lajoie and I have discussed other data. sets (i.e. Ventura anticline) a.ndenvision continuing the project into 1992. The arrival of Dr. Valensise, a SCEC visiting researcher, in January 1992, should increase the pace on this aspect of the project. 900
C,,
flE20’
U)
0 FDM
U)
a.Iong comprising within (1991a).
Each statistics this -
Fig.
work. - tHST
horizontal ¶
the
60
the
3
1.
of
1300
Top. TECTON;35—UcsD I
Generic I I I I. p .
! initial r —‘ this . a—
IL
recurrence -
kin I
realization - L The
line Z
900
9! I I I • ; :
synthetic I —
segment
200
sawtooth 21
fault —-—
years ‘
V’i of I •
: - — — II
various
of
strike,
of — — — —
the 41
seismicity
ma.ps
T a
1
curve ‘—: —• —
plate
4000 I
400 size
Each
a I
I i ‘- .- —
M shows
boundary.
earthquakes year 5
pillow
calculation
- —I.--—. -i-- 1 I • • e
7
ruu,
the 71
earthquake.
reprents I I I e I— -
800 - -L— —
a.
distribution
Bottom. Middle.
-a----- 0. • • : i
within
for — i? t 1991
the ItO
if I
: — -
the
—1
Accumulation Bold
each
Space-time
Middle III
i- _____
800 a
r I.
—
slip
of
l 10:05
Z
segments
segment i
t - — •
segment •
I -
from 15 — -
America - — - I h .
distribution — 13
an — —T 1000
—- ——-;—;— —-I
of
h
are —
lengths I
individual I
tectonic
I
the —.— . 14 — — =_-
M
Trench I
L ,
a 16)16’
prizt’e —
-T—-f -
TL I
and
8 ______
—I
disp1a.cznent
of
events.
1200(KM)
from
earthquake 1 ,— I
I
seibmicity. concern
strengLh 1 P. — -, I
Ward I
c—ID 3
The 181 - —J
of - •
2 FRDF INST TECTONICS—UCSC 18. 17. 1991 1e:5 . 4 3 C-U
I
0
1
s—I
00-I 0 TIME (years)
Fig. 2. Best fits of lognormal (sop) and Weibull (bottom) distributions to the model MAT statistics. Each distribution was constrained to have the appropriate mean recurreue timc. The remaining free parameter was searched for best fit at each 0M The Weibull distribution better , paraniet.ers represents the sta.tfsics of recurrence thafl does a Iognorma.1distribution. The spread of the distributionw are not constant, but generally decrease with magnitude. lu al] cases they are substantially larger than the intrinsic spread value of 0.21 proposed by Nishenko and Buknd (1987). The larger spread reflects a. higher degree of aperiodicity in the recurrence of ma.jOr quakes. FROM INST TECTONICS—USC 10.17.1991 18:06 P. 5 4
Fig. 3. Map of the Palos Verdesregibn of southern California zhQwingmajor drainages (wggIy dashed lines) and domestructures associated with the Palos Verdesand Newport.Inglewoodfault systems. Marine terrace data at San Pedro indicate that the Pa.losVerdes peninsula has been uplifting at a rate of about 0.21 rn/ky over the last 2 rniilion years. If this uplift is the result of repeated thrust earthquakes on the Pa.los Verdes fault, it is possibleto estimate its grnetry and long-term slip ate. The cootours over the peninsul.arepresent the uplift rate in 0.1 rn/ky intervals which would be generated from repeated slip on a thrust striking 132°, dipping 25, and comingto within five kilometersof the surface, The fault (dashed rectangle) is 25 km long, 6.0 and 6.5 10 km wide, and slips at 1 nm/y. For a.fault of this size and slip rate, magnitude earthquakes could recur every 150 and 900 years rcspetively. Group D: Subsurface Imaging of Seismogenic Zones
Group Leader: Rob Clayton
Offshore Newport-Inglewood Fault Zone and Adjacent Crouch (Private) D-1 Compressional Structures: Dana Point to Oceanside
Regional Map-View and Cross-Sectional Determination Suppe (Princeton) D-4 of Fault Geometry for Slip for Blind Thrusts in Populated Areas of S. California
Subsurface Geology of Northern Los Angeles Fold-Thrust Belt Yeats (Oregon State) D-7
Seismic Potential of Thrust Faults of Southern California T. Davis (Private) D-1O
Analysis and Inversion of Teleseisms P. Davis (UCLA) D-13
Scattering and Instrinsic Q for Hawaii, Long Valley, Central Aki (USC) D-16 and Southern California Between 1.5 and 15 Hz
Field Study of San Andreas Fault Zone Structure from Aki-Ferrazzini (USC) D-18 Parkfield to Salton Sea
Trapped Wave Studies Leary (USC) D-20
Experiments of Fault Zone Trapped Waves Li-Teng D-22
Crustal Framework for the Master Model: Collection and 3-D Okaya-Henyey (USC) D-24 Visualization of Geophysical and Tectonic Data for So. California
Block and Structural Velocity Models for the L. A. Region Clayton-Hauksson (Caltech) D-25 OCT—21—91
nvestigaticr
southern belt to Inglewood horizons illustrate map rat. publication. structure normal relationships
to seismic-reflection being the Results
include
beds
in
underlain these 10
convergence the length distance of
(not years overlying been affect
uplifted deforming
probably is rotation
MON
ADJACENT
depth
the
deformation
to
the
concluded along
inner
Multichannel
order
seaward normal
Across
or
folded).
devoid
12:20 analyzed,
15
same THE
the
compression
initial
method
NIFZ
strata
riftinq,
middle
California
and on
degrees
Fault
with of contour
detachment
and formed southern
by
OFFSHORE
of structural
area folded
aseismically
strata
COMPRESSIONAL a
to
portions
rate
7 and
is an
of
thrust
number 0.6-0.7
Offshore that were
extension
for 1cm. the
at
selected
(early Miocen. As is
J.K. Zone evaluation estimated
earthquakes
in
in
parallel
during
detachment,
but
seismic-reflection
normal
the depth Santa
strata
recognized. much
are
retrodeforming
NIFZ
which 414
these used profile, to
terms California
margin
NEWPORT-INGLEWOOD
The
or
of Crouch
(NIFZ)
surface.
of
are details James
mm/yr
seafloor.
Miocene) about
compressed
E. wells,
and
as reverse
has these map Barbara, reverse depth
to
because
within a
the
to
west
these
faults
of
otherwise
to STRUCTURES:
to Cota
period
between
7
of
constrain
younger
and the been K.
and has &
of
also 3.6
the
offshore
3cm
the
be
and
of
profile, Assoc’s,
shortening
bottom of sections
margin.
Crouch
2nd
magnitude structures development The
the
East west
reveal uIFz faults. the
about been and m.y.
and they
early
the NIFZ
assessed. CA the However,
provide
of uplift
these
into
Dana
Floor
detachment
region
strata
relatively
overall
thrust 93101
folded early-late
profiles
of
of
offshore
constructed
NIPZ is samples, th. to
ar.
NIFZ
DANA FAULT
0.4 (Figures
history
tight
a have Inc.
are
Features
the
Pt the
estimated strata,
Recent.
along
Using
fold-and-thrust
and
timing
2.0 riding important
km
appear occur over
west
a
dip
PT.
being
faults
strata of
NIFZ,
zone, and ZONE
been Also,
shortening
mar3ed
asymmetric
convergence
over
end
across Newport
-OCEANSIDE
and
the
of surface the
undaforined seaward.
the
1&2).
the
Miocene
of
Oceanside
the
has of
that
on
AND
converted
and
drafted
to The greater.
the
continuous
however,
a to
geologic a
and
may NIFZ
margin.
the
line
past clues fault
an
asymmetry
maximum
cut
generally
the
be
this
NIFZ, are
Key
be
N!?Z. itself
about
west
and
folds on
or
10
for
as
Dr-I
It
data.
Figure
Location 1 — W41’ ——— —
—
Tracklifle
South
of
Coast
figure
Survey
map
2
of
indicated
offshore
SOUTH
o
O
OFFSHORE
4
by
CALIFORNIA
COAST
seismic—reflection
4
8
A
Oceanside 12
—
8
CALIFORNiA
A’.
&
2OKmi
OCEANSIDE
Survey
i2M*
6 h
2 0 - z i-’. ).ij (A rtCDZ Oifr (t1 CD HctI CDL) QICD I ,- I CDQ o i-’. o rtQ w )CD CD
rt I- 0 OCD N Z i-i. o w CDrt . Ct’ Cjt1 ti 1 H-0 CD
c
CD
I-brtO o rt-t) 0 I-’. 0• CD C) ri CtZO HO o Tj
II
/ / z. p-f •1 N \
\
\
11 c—c-i p U
LU
LU
U
Principal
Institution: Data
major obtained Angeles axial normal, interpretive key fault kinematic, projects fault major Santa tear measured
unsegmented magnitude offsets history kink
area; later. I rates rates Basin been fault
years. recurrence Expanded log
F
areas
r\
data faults
AcquIsition bend slip surfaces
bands
Regional beneath seismogenic recognized
in of
Barbara and
can blind Our
These (Fig.
I
may
and
We
of
Basin, U Investigator:
this
fault of
permission
and
through L from
in
from>
the Slip
be and thceton
which focus
sedimentary
I
quantitatively the oblique-slip
intervils
are way
2). act thrust
of the
U
case kink
subsurface
faults IN
or
the slip combined
underlying
Channel,
dynamic from
the potential for
earthquake also Map-View
to
direction
If
of kink Li
to
in
16km, and
may
r Santa
Ms fault bands
can segment which
the faults.
Blind
determining Los
i.
date
seismic using
in
which
V to University hii
of
bands
• Research:
ages
be terminate
the
segments use
faults.
potential Angeles
sections I
produces
7.1 modeling. Barbara
which California, fault has
Thrust
of to
with
calculated. .
past
record Suppe empirical fault
L Oblique-slip
process
we and
faults
of center, of
motion construct
and is
generated
been IN
geometry
selected
an will U
predicted. This
estimates
and narrow
record
Cross-Sectional
fault •
well
reported
Faults
Channel and Basin, rates or
earthquakes; the 0
extensive
and
data
of
Li
construct
(S) Axial
including
method
future offset
relationships
bav
kink
for collecting — section
slip
limIt
Using
balanced
upward of
wider
by is
growth
acquisition
on
In
will
DC
including determining
the
fault surface
in
band
of slips and bends
constrained Populated
Syntectonic, the
one
the
ruptures proprietary
of
these
provide and
I
this fault
the
in
kink .
kink geodesy,
major map
In into
horizons deep rupture
slip. (A
significant
I” at
cross-sections. the
in
distribute
Determination
method, the magnitude
between maps
slip.
wells
a -
band
underlying logs younger in view
band
A’), Santa
rate
seated
a
long-term
case
fault
may Areas
Sediments
key
basis
area
slip that
The seismic
or
from cross-section widths (Fig. are
are
analysis
of
and
of we Barbara areas
amounts earthquake
maps segmentation
growth
be
thrust,
we and
axial of 17 section,
used
known of for
Fig,
underlying
of have the
1).
predicted
a
U faults
Southern
have
slip
reflection
mm/year.
integrating
than
the
L for
single
of
In of
fault
deposited will
to surface
t
The
Newport-Inglewood I
which
Channel sedimentation
axial
Fault the determined of rates. addition,
by
long-term the
subsurface
or
constructed
do
highlight is
magnitude width subsurface
construction,
rupture Independent event. growth
interval
the
purpose
sediments
surfaces. map California
fault
ramps for
Geometry survey from
We
Further,
fastest
earlier diverse
and
of
bidden
constraints
also
slip Moreover, have
triangles,
ramp.
and
the
to
that is geometry
upward
the
Los of
through
electric
and in
about
highlights
determine atop
In kink
rates.
and
Mapping constrain
research deposited imaging
these
recently
methods, the
the
area
Angeles thrust,
I
crustal
the
rupture
These
L
least
trend.
these band
Los
have . deep
at
1200
sup well the
on the slip
of
of
a • — V IL_j_. I’IJ . L)Z L_)c) -. UL1 IlL
A Area184Okm 2
— :.:.:..:
• K Active axial iurfao. o = Inactive axial surface
IIslands
FIgure 1: A map of axial suri.o.s in he ea.t•rn Santa 5arbara Channd, CA. Th. axial surface. bound a lc)nkband formed during PlIoc.n. and Ouat.rnary mallen on w und.rlVIng oblique, iight lateral blind thrust (S s total slip) Continuity of the ectlv. axial surface (A) suggests that th. underlying fault ramp, which approaches the .abcttom south of Santa Cruz Island, Is continuou, overan area of laa.t 4O square km. ______
C r C C C
Cn CII PcIfic Electdc
I
C
— ____,_._.___.E_a — r I_ L - — — __. JT —— — .,. C fl_ —— —
— -- -,
- ——- —---—f - — . r..-‘---• C —--— ø- — - .—--. _.___-c — - —E2z:_$ __) I — jfr. -øc——.’ -_‘---—;=__ —‘,‘- —— -,=— — =-—;;— ..g:._ ---— C .---- — ____J_—. .=- -. —. re— .- !‘ — —•“---=- w .•‘ r — ‘—- - I._’ -..______4 is C
C — C —T
- 1 0
F—
=-.‘- .—.—--.I.- — — =- —. I --a- ‘—‘j—--- —‘.--— 2
70 ia.oa
=;=N__ —— __:-___...... —.— I - — — - —- — b‘-W - - -‘,— ,-—— — _•1 c ---.--- 1D2121W 1
I-, Los Angeles Basin (1
I
Fig. 2A. Example of active growth kink-bend smicture above blind Fig. 2B. Well section through active growth structure above blind thrust on seismic section from central LA Basin seismic survey. thnist fault in central LA Basin. Preliminary estimates of the Slip on die fault began at the time of deposition of the layer at 07 s earthquake magnitude for this hidden fault based on axial-surface on the left, the layer for which the kink-band begins to narrow mapping similar to Fig. 1 is in excess of Ms = 7. The seismic upward. survey will allow accurate assessment of this magnitude. ) C SUBSURFACE GEOLOGY OF NORTHERN LOS ANGELES FOLD-THRUST BELT The base of Pleistocene gravels (base of San Pedro?) locally is conformable with the underlying Pico Formation and locally overlies Pico and older rocks with angular unconformity (Figure 1). Structure contours (Figure 2) define a broad east- west arch (Wilshire arch) along Wilshire and Olympic Boulevards east of La Cienega Boulevard and, farther west, crossing 1—405 at Wilshire Boulevard. The arch has a broad left step at the northern projection of the Newport—Inglewood fault west of La Cienega Boulevard. The Los Angeles basin syncline ends westward at this projection of the fault. Farther north, the base of the Pleistocene dips gently northward toward the Hollywood fault of Dibblee (1991) near La Cienega Boulevard. These data suggest that the Hollywood fault is not a near—surface feature of a blind thrust, although the Wilshire arch may be related to a blind thrust. A structural cross section shows another major unconformity at the Pliocene-Miocene contact. Our impression is that the active anticlinorium postulated based on the Whittier Narrows earthquake is a more complex structure than previously assumed. RESIDUM. GRAVITY MODELING OF BALANCED CROSS SECTIONS We1 have “gravity—truthed” a cross section constructed by Namson and Davis that extends from Santa Monica Bay NNW across the Santa Monica Mountains at 1—405, across the San Fernando basin and Merrick syncline, to the San Gabriel fault. This exercise is not meant to be a criticism of Namson and Davis, but rather an example of checking a balanced cross section with an independent data set and how that might affect a structural interpretation. The isostatic residual gravity is from A. Griscom (written communication, 1991; see Jachens and Griscorn, 1985 and Griscom and Sauer, 1991). The isostatic residual removes the effects of topography and the Moho so that the residual, high-frequency gravity anomaly is the gravitational expression of near-surface structure. The section is digitized and modeled using the GM-SYS modeling program. Densities for the various rock units are from McCulloh (1960) and Corbato (1963). The greatest discrepancy between the gravity and structure is over the Merrick syncline, which is resolved by decreasing sediment thickness and by making the Lopez fault low dipping to the north. Furthermore, there is less sediment in the San Fernando Valley than modeled. Both these corrections affect total bed length of the section. Gravity modeling of the published Davis et al. (1989) section requires such minor changes that shortening estimates are not affected. REFS. Corbato, C., 1963, U. Cal. Pubs. Geol. Scil 46:1—32 Dibblee, T., 1991, Dibblee Foundation Maps DF 30 and 31 Davis, T., et. al., 1989, JGR 94:9644—9664 Griscom, A. and Sauer, P., 1991, AAPG Bull. 75:366 Jachens, R. and Griscom, A., 1985, in Hinze, W., ed., The utility of regional gravity and magnetic anomaly maps: Soc. Expl. Geoph. McCulloh, T., 1960, USGS Prof. Paper 400B, B320—326 tyfr S Laurel 0H2 N Dorothy Hay CH #3 PE #1 Mamson CH #1 Packard CH #1 TowerCH#1 Nonmarine Pleistocene
within Marine Pleistocene
o% M. Pico m. Pliocene
c U. Pico
Marine Pleistocene % M. Pico
U. Mohnian M. Pico
Figure 1: Stratigraphic correlation of the base of Pleistocene gravels. The logs shown are spontaneous potential on the left and resistivity 100 ft. on the right. Numbers indicate 100’ drill depth 0 —J intervals. Biostratigraphic symbols represent 0 2000 4000 ft. benthic foraminifera age picks. Wells located by Vertical Exaggeration = X8 stars on Figure 2. DQ F 0 2 4 6 6 * wells used in cross section, Fig. 2 thousands o4 loot 9 miles
Baldwin Hills
Figure 2. Structure contour map of the base of the Pleistocene gravels, northern Los Angeles basin (Beverly Hills and Hollywood 7.5’ Quadrangles). Contour interval is 200’ (solid lines), with supplemental contours at -100’ and +100’ (broken lines). Faults from Dibblee (1991 a,b). thrust intermediate between Mountains major the folds. anticline. the nine will our entirely most from geometries Namson age seismicity cross-strike of and cross existing a California,
are southern segments, 1:250,000. hypothesized ramps region. maps map focusing deformation 2). also The In regional these capable sections factor We in order These buried arid surface cross analysis be objective and California the as believe and Other and In cross and anticlinorium responsible
Seismic Our a discontinuites derived in on Davis, Ventura detachments to of balanced focused our extend three-dimensional sections interpretations fault geodetic regional such existing estimating SCEC generating develop discontinuities study including and sections thrust initial these we of Davis Potential 1990 that slip as our from have from subsurface on for basin Annual and cross for the ramp data ramps detachment geologic, a work and work only shows have Jay andl99l) the or and three-dimensional the tied the thrust forming of the may have large Elysian Point and sections convergence with SCEC 6-8 of areas Namson we those Namson is have cross Valencia, Summary been Pacific extend size together in the Thrust reveal plotted to geological the base earthquakes. have km deep faults, the slip San that develop the of Park of focused three ramps was plotted sections. at Morales depth. and Coast the future regional for to plotted well, along and Luis fault 13-16 late Consulting Faults have ramp total initiated CA Report, thrust the thrust rates. dimensional plotting restorations major that Thom a maps on on Cenozoic thrust to 91355 zone and In picture strike surface earthquakes. regional been km fault thrust Over the tops a resolving of
the thrust Most most are ramps, below thrusts regional October The geophysical depths Southern in Davis and positions segmentations and constructed San slip and below Geologists and the responsible September in of such of cases input folds subsurface fault geophysical the (Davis geometry comparing vectors the Andreas bottoms the remaining the distribution compare to the although 22, base the as Santa be and fault California the of data Cuyama ramp The geometry. thrust 1991 the data. Point the the et map across and ramps 1991. convergence for along thrust al., fault geometries is of Monica San bottom map these bottoms which fault some regional time studies. a late the fault slip at San Most 1989; of basin. series Cayetano southern (Figs. a are strike of fault zones regional ramp with rates Cenozoic period are scale are Luis and of are of at In data the 1 The a of zone tops on map of at we +; E a ‘.4
a
a t C
U
.0
‘U C ‘U
HF I1IIi1flI:i j 11 if jI U IIiIi11I1if 11111 ii di il 11 11 11 iI 11130
Ventura Anficline
W.st OH
+WeC 111’00 H AP EXPLANATIO4 ar.nney __ WnoQe d.po&o Ond.e and Cwança p.. Fonando UO,s4 Pane kru IIL an Saq.m Fonoao. Santa UeP.P,.q L Ir h4.ccen. depoith ,dan me b-eq Monica Bay E:: onFenando Foi. “.‘v Vadey. P.aoi Vele. Pan. Panoa and and “the anan. a.po.oi kan,dan a. Cain,.y.. Unt Canyon. dOde4c Idonleney Poiy Puante, Tan Canyon and Thpinga Fonneioni . Id.ooin. .eVuv. and vaeo. bn& qnan.e ed,d.5 the Con.u Mono lay May CCn OPO. W’C El Tie dI.n“ Fenoo..
Le U.o.n WM OIoinn. eooi e,deo a. Gaon4i 10001 0 10 20km E S.oi’e. Ianp S,,nnde’ Vaq,ro, end Vaan.e, Fon,nal.ono I I • I dleneniai.d Eooine ie Uppe’ .kuesa i..n. depoona Foinancan Aaaantag. ., Ccaoi P.ange Opthokiq E:1 — - .nd.a.n.nnd!an iendn ‘iaano, and ,oiIlnanpina nocku — (Moiano.c M
Figure 2
_—4—:1
for proposed.
a.
sites
purckased
stand our
SW
scatterer.
degraded
semblance cross-correlation
on effect,
been
1990).
associated First
correlation used
at
in
Herbert
site
As
Our
the
.outheru The
aftershock
arrival,
firit
arid
in
to
seen
arrival
preparation
at
out,
more
using
search
The objeethre
the
etack
E
Rendon
Frazier
attempt
:4:
using
correlation
times by
We a
th.
with
of
beam
Sari
between
than
Reftek
next
California..
travel
the
other,
the
waveforms
studies,
for
-“
will Mojave
the
the
Gabriel
of
Park.
ha.
model
ii
between
200
NORESS
stack.
step
coherent
to
for
then Pg
to
time
method
Salton
recorder
F?:M
(Hadley
nearby
acquired
search
sites
will’
use
a
should
and
involve,
and
We
of
With
Analysis
have
l..ige
Mountams
residual.
to
E—C
When
teleieiimic
of
occur individual
Sung
trough
Pu
The
form
scatterera
propose
of
array.
for
ettion.
the
(UCLA
and
th.
the
students
data
Hedlin
scale
waves;
systematically
Sierras.
(199),
spatially
a
southern
high
where
Kanatnori,
need
assistance
and
PauIM.
have
beam.
We
and
from,
in
imaging
to
matching
et
phase
he.i
stations
In
Then
noise
transverse
propose
preparation
purchase
arrive.
fs.r’,1Ha.r
been
al.,
scattered
Inver.ion
Sung
The
the
arid
A
been
coherent
California
Davis,
(19l)
tune,
least
inversion
1
stations
of
tel..ei.mic
calculated
1977;
next
experhaeut
begun
and
and
stripping
Lund.)
to
used
Jim
with
a
square.
ranges
and
search
radiation UCLA
step
who
Jackson
further
Raike.,
the
for
of’
variation
Mon
analysis
network.
to
the
are
for
waveforms
and
Teleaeiem.
detected
an
will
average
correct
(figure
coda
can
of
for
technique
upper
(USGS)
instrinuents
in
excluded,
Imaging
2
have
1980;
(1991,
crusts]
involve
be
is
of,
body-body
southern
instruments
in
has
Semblance
greatest.
seen
2a).
for
13
mantle
waveform
body
installed
correlation.
Humphrey.
to
begun
UCLA
in
well-recorded
which
local
..nd
transect a
image
in
Travel
it
press)
systematic
-
California.,
an
the
is
Rayleigh
anomalies.
uppermost
scattering
by
statics
analysis graduate
Figure
it,
and
maxi.mi.e.
anticipated
residuals,
lateral
found
sites
time
which
examining
that
on
and
Two
to
and
permitted
(figure
scattering
anomalies
search telaseisms
2b
install
has
from
variation
Clayton,
has
we
is
also.
student
regions
mantle
off,
u
shows
based
cross..
been
have
been
that
has
the
the
for
1).
at at
I
4
-08
-0.6’
-0.2
0.2
0.4
0.6
0.8
—1
0•
4.:
0 1—
—
:L
2
x670467
z670467
iM
4
Seismic
Stack.d
Section
Saismogrn
6
dm.
-
time
L
8 S
(sic)
(sec)
-
Weighted
L.S.
Weight
10
ccnect.d
cc.mc.
12
Athval
Art.
lim.
Tim.
14
16 16 Lab Lab Lab La) Lab Li) II Lx La) t 3 Lx Lx I-
t.la
a rr
± ‘S C, •1 • ‘1 U 0 2 1 a V PS a P. p C, P.. C 4- p I p •0 p ‘0
I-. I—
.1
11
regions
Q-l,
to
for general
from
freedom. range
Q-l.
distributed isotropic normalized
time
to normalized successive Unlike
fundamental energy, separate
wave
A
f-I
a
frequencies
Comparative
Valley,
common
whereas
.4
windows
the
To
in
and
The
A
to
the
the
conclusions
the
show
Q-l
the
other
new
f
body
Within
its’
-2.0,
figures
time
scattering
method
scatterers
new
energy
the
effects
at
separability
Central
above
site
method
coda
the
whereas
regions.
wave
less
were
the
energy
windows
method
by
and
the
show
accuracy
90%
than
as
in
of
Study
of
6.0
Kevin
on
scattering then
Q-I
and
error
Hawaii, recently
a
scattering
source.
and
Wu
in
Hz the
the
function
confidence
or
the
Despite
relies
of
as
is
each
uniform
simultaneously
Mayeda,
of
the (1985),
equal
nature
Southern
bars,
anelastic
of
source,
strongly
frequency-dependent
a
Scattering
the
function
Long
developed
We
opposite
on
window
in
Q 1
of
results
the
to
model
a
which
the
of
plotted
intrinsic
level
hypocentral
site
Stuart
6.0
medium
Valley,
and
Q-l
rather
frequency
attenuation
integration
California
of
Hz,
and
was
for
from
using
fitting,
intrinsic
is
involves
by
hypocentral
Koyanagi and
fit
the
considerably
large
scattering
many
Q 4 .
path
central
found.
Hoshiba
of
to
southern
geometric
the
Intrinsic
dependent,
randomly
distance.
Monte-Carlo
the
error
effects
in
Q- 4
of
events
scattering,
integration
F-distribution
Between
and
vertical
all
and
We
the
et
Q 4
from
distance.
bars
four
California
al.
southern
for
S-wave
found
Keiiti
less
spreading-corrected
recorded
Q 4
is
The
and
(1991)
an
decreasing
we coda
error
regions.
greater
frequency
anelastic
1.5
simulations
of
analysis
for
data
uniformly
Aki
that
can
energy
Using
waves,
for
the
was
and
California.
bars
agree
Hawaii,
at
make
for
in
than
40
entire
many
used
In
all 15.0
and
give
proportional
of
all
the
degrees
dependent.
for
we
with
general,
intrinsic
some
four
the
assuming
three
total
to
the
Hz.
three
stations S-wave
Long
5-
those
of 0-I’ u-fl
0.01
1Q_
10
10 1 10
p Southern CA —*-- Central CA Long Valley --0—— Hawaii 0.01
1Q-
Intrinsic 1Q 1 1 10
p Southern CA Central CA Long Valley -—a-— Hawaii 0.01
1Q-
Total 1Q
1 1 10 Frequency (Hz)
in
instruments
softwares other
time
and rate
the
20Mbyte filtered
the
converter
modify
Corporation. perpendicular
size
relative
At Figure
second
variation
Input
At
To
To
We
early
these
box
response
consist
of
of
the
and
from
operate
carry
programs
proposed
from
100
a
1,
instrument
according
our
November.
of
to
present
lunch
sites
duration
hard
in
for
with
along
the
the
out
Parkfield
samp/sec.
in
old
will
the
the
of
the
It
the
loading
recorder
to
fault
disk,
we
this
box.
a
is
MIT-built
the
fault to
are
time
3-component,
the
be
conversion
instrument
the
Onset
equiped
of
will
to
collect
experiment,
geophone
easily will
strike
used
which
The
northern
Field
fault.
the the
The
zone
the
The
different
set to
which
be
to
recording
recording
sampling
instruments
to
microseisms
Salton
in
up
adapted
spectral digital
power with
structure
corresponds
operate
set
Study
obtain
one
the
time
one
part
to
up
should
1Hz
8
programs
range
we
must
10
may
tape
analog
instrument
of
of
Sea
Ak.
at
rate
for
analysis
a windows.
of
system
in
sec.
geophone,
needed
by
time
the
varying
2.5
Progress
are
triggering
first
San
last
recorders,
the
be
of
before
data
using
studying
to
It
San
input
microseconds
1
being
supply
cJ..
correction
purpose
is
two
about
to
is
program
to
Andreas
of
to
along
then
Andreas
outside
a
distance
set
8sec.
being
The
3-component
improve
the
pins
tattletale
weeks
is
tested
mode.
report
up
externally
10
we
the
first
amplified,
the
data
instrument
Vr’e
of
connected
hours
the
the
digitized.
have
long-period
and
the
We
other
in
passed
fault, San fault
from
at
will
our
and
time
Onset.
data
on
the
fault
to
USC
have
developed
of
Andreas
recording
experiments
or
unload the
be
an
zone
field.
from
at
from
continous
logger
through
zone
is
to
The long-period
done
internal
by
selected
and
This
gain
first
then
a
anisotropic
San
lithium
structure
an
12-bit
data
the
the
fault
as
built
one,
of
as
is
a
external
left
an system
reference
Juan
new
data
the 10,
field
clock.
done
recording
13
are
by
equalizer
to
by
analog
in
in
batteries
data
100,
sites,
developing
recorded instrument
Bautista
determine
files
the
a
experiment
Onset
using
microseisms.
amplification
and
direction
clock
are or
field,
site
recorded.
to
shown
rather
at
1000,
which
a
Computer
collected.
digital
placed
and
and
to
Macintosh
a
on
lkSC.
and
the
sampling
with
Cayuma.
new
than
a
will
in
the
a
and
extends
two
These
inside
a
start to SOTQ$UVfl
ff:
awejotrj
oigiaq
mspnt uunf
UHci
971,
M.
Malin Science
al.
(1991)
Geophys.
core
trapped
[1] such possible
with
means
de
Andreas
geometrically
distributions
geometrically
trapped
precise Gutenberg-Richter
velocity dimension
clustering
resolved
located
rock
advanced borehole phenomena
simulated rise
and
starting trapped
study scale small fracture
are
(Gutenberg-Richter
1)
Physique
sensitivity
&
(1989)
recognized
Kagan
civic
event
numerical
low
to
is
P. range
volumes.
sequence
(1991)
that
Proc.
evidently
Interpreting
investigation
To
There
Observation
wave
waves
power
256,
within
E.
density
waves;
point
to models,
fault
and
velocity
J. velocity station
use
clustering
groupings
illustrate
Science
(Fig.
an
MaIm
match
Y.
is
and
occur
were
R. du
3-component
Conf
to
Science 226.
excitation
complex
there
are
irregular
observed
at
Y.
for
in
irregular
to
law
methods
a with
Globe
of
astr.
fluctuations
small
3)
The
ib).
a
Parkfield,
complex
head
records
planar
related
conducted
&
(1990)
be
observed
model
[3].
process
failures.
such
(VS
recording
is
[4]
‘Earthquake spatial
b-values
distributions
244,
seismic
of
L.
as
of
8
to relation),
‘critical
that
Soc.
events;
a
251,
The
wave
[7].
station
provided Physical
Parkfield
much
Geological
fault
lie
Knopoff
the
need
investigations:
Bak
(Fig.
in
based
volume
inversions
four
J.
low
1.8-2.0
most
direct
in
waveguide
on
clustering
557.
107.
equivalent
data
deep
fault
The
1592.
crust
To
Geophys.
Parkfield
in
this
data,
zone
which
P.
trapped
systems’
to
spatial
of
P
2)
a
2b)
3-component
aspects
SCEC
velocity
crystalline
on
turn
narrow
nearly earthquake
&
and
allows observe
systems
simulations
fixing
the
fault
about
event
summer
zone
Prediction:
(1980)
[3]
borehole
[6]
kmls)
has
near
by
consisting
and
are,
a
C.
[9]
Sciences,
large
these
[9].
low
uniform
resolution
S
Li
Research
the
of
Tang
King
wave
structures.
shown
Res.
to
Michelini
fault zone
and
trapped reproduces
fractal
constraints
head
clustering
event
the
fault
within
the
1.1
events
Parkfleld
Y.-G.
fault
layer.
velocity
in
hazardous
three
Geophys. achieving
scale
on
rock
seismic
can
possibility
fault
which
95,
log
to
G.
(1989)
zone
spatial
Parkfield
observations trapped
will
waves;
the
demonstrated
University
spatial
of
Peter
planar
power
State-of-the-Art,
dimension
structures or
events
Statement,
1.15
(Hurst
be
limits,
waves as
G.
21793. &
are
fracture
a
A
massively
provide
(Fig.
A.
catastrophic
material
as
trapped
microearthquakes,
is
uniform
Simulations
elements
data
highly
P.
from
P.
the
Leary
150-250
recorded
J.
J.
spatial
fault
separations
feasible.
a
occurrence
[2].
&
4)
parallel
waves
law
Leary,
(1983)
function
rescaled
with
R.
borehole
propagate
Geophys. of
visual
independent
on
T.
1).
of
[101
trapped
achieving
density
recognizing
better
astr.
seismic
unpredictable
distributions
Relocated for
October
V.
wave
properties
with
the
at
Southern
fault
parallel
resolution
locations
that
A interact
m
[6,7];
Ben-Zion
waveguide
BSSA
McEvilly
Parkfield
Pageoph
down
clustering
a
failure
nature
fractal
of
range
of
Soc.
wide
velocity
seismic
cubic
Strasbourg.
at
in
gouge possibly
observations
wave
(‘Kagan-Knopoff’
Res.
easily
seismic
time
fluctuations
activity
wavelengths
relation
Connection 1991
fault
good
over
62,
to
80,
California
consider
do
of
can
of
[2]).
shown
Parkfield
fault
S of
dimension
volume
94;
and
observations
magnitude
can
Y.
characterize
crustal
(1991) requires
head
network
not
303. models
in
121,
[4].
the
fault
zone
(Fig.
1245.
[10].
be a
velocity
trapped
over
non-uniform
Physical
15635.
&
space.
velocity
large
to
zone
be
change
precipitated
in
size
Brittle
[81
waves
K.
the
761;
the
with
zone
[2]
approximately
P
la)
is
Machine
Fig.
and
Figs.
mechanics
BSSA
as
[7]
events
appropriate
and
range
Aki
Ben-Zion is
head
provides
fault
the
distribution
D
scale
D
wave
large
models
and
Leary
-1
Leary
incompatible
Meakin
a
microearthquakes:
significantly.
is
phenomena
2a
relation
structures
[5]
=
event
fracture
will
[8];
=
2c.
earthquake
(1990)
ability
150-250
[5];
81;
consistent using
2.3;
waves
of
the
shows
1.3
using
Maim
propagation
on
an
by
at
there
low
allow
Constraining
scale
P.
524;
locations.
an
P.
at
to
of
the
statistics
creates
earthquake
this
Y.
a
which
to
P.
C.
fault
BSSA
in
relatively
to
least
[1]),
scientific
excellent
of
C.
modeled
the
velocity are
superior
that
are
Institute
&
P.
m
that
the
lengths
Lees
a
(1991)
(1991) crustal
fractal
giving
model
events
O—ao
et
more
P.
sizes
wide
E.
fault
zone
with
with
This
well
well
it
San
and
one
80,
the
the
are
al.
E.
of
in
et
J. is 10 km 1....I 20km . . . 0
.. .c.
•1’ , . : ‘ - e. t ,. 10 km
(a)
jQ3_ -
to1 I I U., S tol’ “U, %,
tO 4. S
“I 4” •‘ to 4 too to’ t02 tO
nrrER-Fr DLSTANE
(b)
Figure 1. (a) Comparison between the distribution of Parkfield microearthquakes (above) and a synthetic fractal distribution of points (below). The synthetic distribution has a fractal dimension of D = 1.3. (b) The frequency distribution of inter-event distances of Parkfield earthquakes (solid dots) and inter-point distances from the synthetic distribution (x’s). The distribution of Parkfield events is given by Maim et al. [5].
C-)
0.5 2.5
0.5
3.5
3.5
2.5-
00
0
3
1
0
xlO- 5
x10 5
MAXIMUM
MAXIMUM
10
10
fROSS
TRACE
TRACE
20
20
/
AMPLITUDE
AMPLITUDE
TRACE
/
TRACE
fl
30
30
\
NO
FAULT
NUMBER
NUMBER CROSS --
--
40
40
xXfault00y
xXnofaulty
FAULT
50
50
vs
vs
xXfaultOly
FAULT xXfault0Oy
;
60
60
70
H 70 have past have keep on of basement catalogue during of that the trapped appropriate appropriated between of appropriate the gouge that 5-15 indoor fault—zone recording amplitude interest. fault-zone layer on is deployment surface for schedule fault
observation Yong-Gang meters. the Mussall, preferable the instrument distance instrument surface three in The 5. 3.
4. 2. 3. Report 5. 4. 2. We 1. km we 1. been been is (thicker sandwiched November or at following Smith’s touch trace distance waves can 0 characteristic The We have in Four The above is The The The with in system. The Six The of Anza, and selected granted months, trapped The of have November
width on trapped Gerble,
of Li/Ta-Liang obtained the seismic from overlain for low location event with fault—zone location appropriate location of instruments units to deployment. using the units 500 location numerical prepared than 6
deployment Experiment fault to criteria farm, criteria: container. being southern fault range velocity both record PASSCAL the of by May, help we waves. Smith. m. by after 50 of locations Mader events. for surface fault waves of have basement fault enough zone by The should
the Teng, all training m) should amplitude zone. them should of of shipped from should modeling fault—zone six trapped PASSCAL are for . a locations geological instruments numerical Frank achieved has The should a and
former private of California” trapped layer zone The
increases. USC, thick will even month wave-guide, waveform the be The recording
crucial Fault the not be been numerical O’leary outcrops will numerical from to be is in
Vernon. Group earthquake appropriate waves be will shows instruments be in and four use if sedimentary data be properties. between the the field
waves Zone trapped within for carefully IRIS placed modeling be placed the survey data to in they
PASSCAL D will phase properties, destroy following segment owners, that decreases logging. funded stations.
instrument (Subsurface obtained on i.e.
modeling Trapped center modeling the observe at seismicity experiment, during tens or inside waves be both source are this in in the offset information area shows a selected instruments. are or close by of completed the and the or safe narrow and The things: San The excited
width Waves sides. rapidly using During fault weathering shows six shows fault deployment structures from
SCEC. where Imaging) SCEC. available two region that we vertical areas range hundreds location area places, Jacinto permits to months based range fault units still “The 1 that UCSD that zone zone the The The the the the by on as Hz is of so
of is 1
Figure
Figure
appropriate
Anza,
observation
2
1
An
southern
The
abandoned
location
place
of
California.
fault—zone
for
dry
map
instrument
well
of
trapped
six
within
PASSCAL
deployment.
waves
the
trace
instruments
at
the
of
San
fault
Jacinto
(circles)
zone
fault,
is
for an visualization software formats, capable so These able sets platforms of Vicar, used located discuss the cost. hardware if 7.5 30 inhouse; obtain obtain capabilities fully bathymetry, geological which stated Objectives (C) (b) (a) Summary P.l.’s: Program: necessary, planar that m framework data determine develop Geophysical make mm, USGS to (i.e., examine for x Gridded WAVEFRONT, can Visualization The these from software above. Efforts David exchange 30 GIS spatially with of this (academic (e) 15 application discussion and/or database in a configuration m, Report Crustal at be and rendering of initial 3-D mm, of data the be a use and system. UCSB data application. Okaya however some 2 this thus graphical used database the the for digitial We competitive geophysical arcsec USGS block provide volumetric at at and 30 regional can data data the - at research best visualization Framework far are depth). SCEC October cost. (a) and or and mm, software with and either is Tectonic planar be master topography for Stardent SUN of sets commercial we x with currently to selected. available regional which visualization JPL). of visualization to 2 land-use Tom incorporated and the this gravity define Such will institutions, arcsec, Consultations geological no various which data. Arcinfo. The program: packages regarding 1991 datasets and USGS model. program which 1 try or Henyey will for Data A degree AVS platform. data software (gridded) assembling the little volumetric and set methodology to are the The draped (Digital allow 2 computer degrees obtain for for arcsec have crustal is of It are systems. cost. capability magnetics. will Master most and such site (d) into ultimate have is the increments. offshore Southern a for desired available with application occur more be been Bouguer over Elevation and geophysical them three-dimensional leading easily x Currently, as the framework data focussed of a able hardware/software). 3 Model: UCSB preliminary to SunPHIGS, sophistication SCEC-wide topography goal arcsec complete considered Full simultaneous available once California via to sets, display California to obtainable Initial at candidate gravity be the accumulation is Model These accept Collection and must various we on we to and (c) able of information community attempts the determine and have set JPL to Southern have work set bathymetry (a) of 3 or SunGKS, DEM roughness group include to very all first arcsec is spatial are to Southern oblique-perspective displays visually DEM) display of will be share determined an SCEC and on also be 2-D diverse two data digital easy are licensing, be of arrangement different to the California if Startdent a 3-D x is underway. sampling data different and objectives and made possible software- define of juxtapose being 3 will institutions data with to available using software California arcsec topography Visualization multiple data use, point with SunVISION. be the hypocenters is before computer made and a this purchased AVS, regional already and types (b) Arcinfo (will intervals: crustal data and as to from such to (f) data be at views r)%Lf to (b) a be JPL in be and and no to of and time is seismograms.
concentrated obtained profiles can mass-storage operational.
the
structural
be
inversions
The
This
picked
used
Velocity
across
a
object
study
complete
on
on
component events
features
the
and
ALTECI
Unix assembling of
is
Structure
basin.
this at
waveform
produced
R. phase
computers,
of
a
SEISMOLOGY
Clayton
study
very
of
the
of
The
catalogue
and the
the
greater preliminary by
is
data.
Caltech
latter
and
data
reformatting
to
and the
Greater
determine E.
for
The center Los
step SCSN ____FX
are
Hauksson
1984
Los
stage now
primary Angeles
is
and is
now
NO the
the
Angeles
to
starting
now
and
the
81e5640715
data,
1990 possible large
source
Basin
to
(almost)
archived
Basin
to date
in
scale
We
from
of construct
a
because
has
form
have the
velocity
travel
data
that
now
the
P
I,’
02 Group E: Crustal Deformation
Group Leader: Dave Jackson
Extremal Fault Slip Models for Southern California from Johnson and Agnew (UCSD) E-1 15 Years of USGS Trilateration Data
Surface Deformation and Seismicity Compared in a Structural Context Beavan and Seeber (Columbia) E-3
Velocity field in Southern California from GPS and VLBI Hager et al. (MIT) E-4
Geodetic Approach to Quantification of Earthquake Hazards in the Hudnut (Caltech) E-6 Los Angeles Basin Within the Next 3 to 5 years
Tectonic Deformation in Southern California Jackson (UCLA) E-7
Evaluation of the Characteristic Earthquake Concept Rice (Harvard) E-IO in Southern California
Marine Terrace and Earthquake Occurrence Ward (UC-Sanla Cruz) E-12
Permanent GPS Geodetic Array Operation at SlO Bock (UC-San Diego) E-15
Mechanics of Blind Thrust Faults in the Los Angeles Basin Stein et al. (USGS) E-18 E—( Extremal Fault Slip Models for Southern California from 15 Years of USGS Trilateration Data R4ulIY JoIsoN Institute of Geophysics and Planetary Physics, Scripps Institution of Oceanography. 0225, University of California, San Diego, La Jolla CA 92093
The USGS geodolite measurements made over the past 15 years throughout southern California provide an unequaled measurement of crustal deformation in the area. We have used several different techniques to investigate what this dataset can resolve about the distribution and rate of interseismic slip on the faults of the San Andreas system. Most recently we have developed a procedure to invert for extremal models to try to better pin down the rates of slip. Instead of trying to determine the com plete slip distribution on each fault (which requires hundreds of model elements to be found), we invert only for the minimum value of the maximum slip rate on each fault: the “minimax” slip rate. In this procedure we seek the slip model s that minimizes the misfit 11E’(d— A )IIE subject to the constraint 0 s u. Where d is the vector of line-length rates of change, Z is the covariance matrix of the measured data, A is a matrix of partials which relates the slip on an individual fault patch to a change in line-length, and u is a vector of upper bounds on the model elements. çl’he model elements s are also constrained to be greater than or equal to zero which means that only right-lateral movements are allowed on the fault patches.) In this way, each model element is con strained to lie between both a lower bound and an upper bound—thus the technique name “bounded value least squares.” The general strategy employed is to impose a different upper bound on each fault and then let the algorithm calculate the least squares solution to the model. If the imposed bounds are very “loose” (large range of slip values allowed), the routine will have no difficulty fitting the data and will result in a small model misfit—since this is still an underdetermined problem, it is quite possi ble that the misfit will be 0. On the other hand, with tight bounds on all of the model elements (or perhaps only a select subset—and this is the virtue of the approach), the routine will have difficulty fitting the data adequately and may result in a model misfit (distributed as 2XN’ where N is the number of measured data) which is deemed statistically unacceptible. In this latter case, we can then assume that at least one of our assumptions at the outset was incorrect. If we choose our assumptions (i.e. bounds on the model) so that all but one is non-controversial, we are left with the conclusion that this bound was incorrect. By imposing a relatively loose upper bound on the slip patches on, say, the San Jacinto, Elsinore, and Newport-Inglewood faults and then progressively lowering the upper bound on the patches of the San Andreas Fault, we eventually come to a point where the data can no longer be satisfied (at some level of statistical confidence). At this point we say that we have determined the minimax slip rate for the San Andreas Fault. That is, there must be at least one fault patch on the San Andreas which must be slipping at a yearly rate greater than or equal to this value. We have calculated this minimax rate for each of the four southern California faults and have found values which axe geologically reasonable for both the San Andreas and San Jacinto, and have found that the data axe fit acceptably by a model in which either the entire Jacinto or Newport-Inglewood is held to no slip at all (though both cannot be simultaneously held to zero slip). The results for the San Jacinto are shown in the accompanying plot. The minimax solutions also reflect where the algorithm wants to place the subsurface slip. As it turns out, this placement shows similarities to both our previous inversion results (though with different slip rates) and to the patterns of seismicity along the faults. Acceptability Region for San Jacinto Minimax Slip Rate 60
50>%
Co C) -C C)
D
-cC) 4-0 300 C 0 aa) D
10 2 4 6 8 10 12 14 16 18 20 Minimax slip rate on San Jacinto Fault (mm/yr)
Contour levels ore one minus the probability that the best—fitting Inversion model would have had the calculated misfit level based solely on the error estimates for the geodolite data. To use this plot: select some prejudicial upper bound slip—rat. to be imposed on all fault patches on the SAF, EF, and NIF on the vertical axis (e.g. 35 mm/yr), then trace horizontally to the intersection with an acc.ptible confinderice level (e.g. 99Z), and finally down to the horizontal axis to get th. SMALLEST minimax slip—rote not violating these assumptions (about 12 mm/yr); hence, somewhere there is a patch on the SJF which Is slipping at least this fast. All acceptible minimax slip—rates then exist in the pie shaped region not excluded by the haichures. L.I
Investigations
interseismic modeling VLBI,
inferred changes sain2 in increase about progress the sets observations release from upon primarily on Results
Park, Johnson number yet
geodetic
The
C
the
lack
We
begun; The
Surface
to
the
USGS
segmentation,
and
GPS
co-seismic
make
Can broad-scale
from propose
i
in from
in
of of
tri)ateration
(October, of on
backdrop and
concerned
a
sess
such
data?
scheduled we
suitable
and
IGPP,
“cleaned”
geodetic
d.eforrnation J-1DHLI the
trilateration by
the
the
Lamont-Doherty microseismicity
a Deformation
anticipate
start
other
data
questions SCEC. to
orientation
USGS
actual
Can
and
U.C.
quantitatively
deformation of
1991)
geodetic
with block
in
on measurements
data
field geodetic
version the post-seismic combined
the
behavior
San
network
these
working
lnterseismic
partitioned
long-term
networks,
asked rotation,
experinnts. set
years
(inferred Diego.
and
data,
of
for John and/or
problems.
Geological
data.
Palisades,
elements
a to studies
of
in
above,
compare
on southern
Seismicity
substsntlal
and
come. deformation
faults, Beavan,
etc.?
southern
of
Analysis
deformation
But but
the
from between
that
deformation.
(914)
deformation
it
of
and
data
for also
of
is
The
inferred Observatory
detailed We New
We
seismicity seisrnicity
surface
California
359-2900
anticipated the Leonardo
making
Caiifornia. to
subset after
of
using
propose eatest
elastic
Compared
anticipate
master York
use
associated
these
reflected
from
near-fault
Christmas, deformation
fields
our
of
recent
the
and
and
10964
and
Secber data
of
limitation model
to
has the
that
results geological
connection
Columbia
use This
near-fault stress that
be
recoverable
data
In with
in been
In GPS
SCEC
reconciled studies
currently
can smacture.
the
once
we
comparison
a
has
to
from measwements)
major
obtained
on
data)
Structural way
be
will better
been
the structure? University will
geodesy
between
such will inferred
geodetic
described earthquakes;
available suain,
first
be
with
provide with
The
given
plan be
studies
from
able
will
P1 required.
give
focus
seismic from
the this
and
future
Context
has
data to Can
be
both be
USGS
geodetic
above
a
deformation
at us
broad-scale
information
completed permanent analysis
recognised
substantial
carried will
near-fauic
(primarily present we
by
geodetic
to
moment
How
will Hadley has
not Menlo
make
data
out
of
not
is
be is be
a
E—3 -—z’i. JJ•kJ rr..LI FU I UUL)t1 .. LLJL)I’,HII U
Progress Report to SCEC Velocity Field in Southern California from GPS and VLBI
B. H. Hager, P1; T. A. Herring and R. E. Reilinger, co-rs Massachusetts Institute of Technology, Cambridge, MA 02139 October21, 1991
We proposed two projects directly related to SCEC goals: 1) to produce a map of the velocity field in southernCalifornia using all available GPS, VLBI and conventional data in the region, and 2) to interpret this velocity fleid using both “traditional”dislo cation models of creep at depth on faults and more “realistic” models that include vis. cous relaxation in the intracrustal asthenosphere. Most of our effort to date has been directed at task 1).
The velocity field from space geodesy is being estimated using Tom Herring’snewly developed “Global” software, which uses a Kalman filter to combine station-position covariance matrices and estimates to form a single rigorous solution for positions and velocities. Up to now, we have concentrated on reanalyzing OPS data from two subsets of data - the “Trex” experiment (primarily funded by NSF) analyzed by Mark Murray at MiT for his Ph. D. thesis, and the Ventura Basin experiments, (primarily funded by NASA) analyzed by Andrea Donnellan of Caltech for her Ph. 1). thesis. To strengthen the fiducial tracking network for this analysis we also used data from the 1988 Global Tracking Experiment (GOTEX); from the 1991 IERS GIG campaign;and recent data from the PGGA. In all, almost 200 OPS expexirnents were analyzed. Since the bulk of the data analysis for all these experimentswas carried out at MIT,, it made sense to start work on these data sets. It surprised us how much work such a seemingly straightforwardproject turned out to be, requiring discovery of such pitfalls as non-standardized nomenclature for fiducial sites.
Danan Dong, the student supported by this grant, has made substantial progress, both theoretical and in the area of software development, in combining conventional and space geodetic survey data. His approach is to remedy the rank deficiencies of terrestrial geodetic data by using the coordinate and velocity information from the space geodetic solutions as the “model coordinates” in the inversion, or, alternatively, to use a Gauss-Markov approach. An example, using VLBI and trilaterafion in southernmost CA, is shown in the attached figure.
In terms of code development, our next step is to combine GPS and VLBI solutions. This is expected to be straightforward. We are also attempting to acquire all useful thiateration data, primarily from USGS, and triangulation data from NGS. We also need to acquire “clean” GPS data from SCEC.
Figure: Gauss-Markov solution for velocity field of southernmost CA obtained using the VLBI solutions at monu (Monument Peak), piaf (Pinyon Flats, assumed — asbe), blkb (Black Butte, assumed oroc), and pear (Pearbiossom, assumed • brin) combined with denser trilateration data. 11
-1 U-)-u
0 U m LO
Qo C] m U U 7 D
0
Th 73 U 0- i i i flu I • LJflL.Z.Lfl
Report to SCECfor the October 1991 Workshop: Group E, #4 Geodetic Approach to Quantification of Earthquake Hazards in the Los Angeles Basin Within the Next 3 to 5 Years Prirsc4oilnvestigaor: KerryE. Sieh Institution: Seismologicallaboratory Co-Investigator: KennethW. Hudnut CaliforniaInstituteof Technology
Using historical geodetic data and our GPS data from 1990 and 1991, we are working towards providing displacement results and to estimate earthquake hazards using SCEC funding to help support our data analysis. interpretation and modelling. Initiai results from work in 1991 are still preliminary, but we plan to submit a paper on these results for publication during early 1992. Much of our effort in 1991 has necessarily been re-directed at studying displacements associated with the Sierra Madre earthquake in detail - these results of our SCEC work are to be presented at the Fall AGU meeting,and apreprint of that study will be ready at that time. The purposes of this project, specifically related to earthquake hazards, are as follows: 1) assess overall slip distribution pattern and relate this to individual structures in the Los Angeles basin, 2) analyze for locked versus aseismically slipping portions of the detachment/ramp fault system (e.g., is the Santa Monica Mtns. fault-propagation fold accumulating stress interseismically?), 3) model and determine the interaction of the San Gabriel Mtns. thrust-sheet wedge with the San Andreas fault and determinehow this interactionmay affectrecurrence interval modulationon the San Andreasfault. A first attempt at determininglocked versus slippingpatches will be done with simpledislocation modelling. In addition to these earthquake hazards goals, the following more general aims are being addressed by our study: I) is fold development consistent with the kink-band model of Suppe, or another model?, 2) what role do left-lateral faults play in the LA Basin, and rigid-body block rotations occur in the San Gabriels or in the Fontana-Chino area? This project is a major component of the geodesy group efforts toward the SCEC goal of assessing earthquake hazards in the Los Angeles Basin. Our contribution is presently in determining the spatial disthbution of strain during the intervals 1933-1978, 1978-1990, and proposed work will include the 1990, 1991 and 1992 survey data comparisons and interpretation of those data. Currently, we use dislocation models, and with various possible geometries and slip patternsfor buried and surficialfaults in the study area. We will soon complete our comparison of 1978 trilateration data to our 1990GPS data (from a 23-station network across the LA Basin), and of the 1933 triangulation with these later surveys to specifically study strain across the San Fernando Valley. Our work may shed light on possible mechanisms for the substantially increased M>4.5 seismicity in this region since 1987, as discussed by Jones et al. (SCEC Bull. No. 1, 1991). Caltech GPS Stations in Southern California ::—4—:
We
the KOJAVE12, data, construct velocities southern
used avaisbie. 4•termined BLKBUTTE, to MOJAVE12;
that Angeles north
to We the thrusts velocitie, consistent
newly Figure as good data, Th. the limits
slip relaxation agreening significant obviates rheology),
(5-10
used
determine PVERDES
used
data
map,
USGS
relevant model
C4
VNDNBERQ, GP8
agreement
rate
which
of
discovered
mm/yr
implying
DAVID
on 2.
a
basin
beneath
the
the
we California.
combined
latitude
trilateration
the
combination
a NON
These
PVERDEB,
that
We
th.
shows well TECTONIC on
of
from
with
and
at although
map
are, started
and
data
geologically
between
displacement
D.
the
M
allowed
time
the
need
PEAK, SAFE,
we
convergence
depth.
CENT,
is,
rotation
with
that
with
JACKSON,
of
data VNDNBERG
Global
geological
the
significant
on
remaining
find described
Mojave
San
with
35
scales
horizontal
for
SANPRUL,
they
DEFORMATION
with
average,
and CENT,
basin.
plate
Pales
geological
the
The it then
TWIN,
are
for
kndreas
of’
N-S
time
historic Positioning
networks,
doss
VNDNBERG.
are
OCTOBER
rates
Shear
deformation
VIBI, DEPT
map
Vary
using
all
creep
en
naasusred
(from determined
TWIN,
convergenc,
tectonic or
Verdes
velocities.
and
above,
findings
dependent
PINFIATS,
on
deformation
the
not
the
is
moving
dirergence
EARTH
at velocity
Long
of
Zone
GPS,
measurements
meet
the
on
a
IN
estimates.
shown
LA.CU, triangulation
24,
Wallace Ro.gri
uric
most
to
12
and few
as
each
SOUTHERN
We Baseline
System
Pacific
as
theory
slip
&
arid
of
at of
1991 determine
crustal
rates
the
wall
years
the SPACE
strain our
accurate in Psi2ossor’i, then GAVI
of
a
of
the The
about
the
fault,
triangulation
fault
rates of fault.
offshor,
Creak
Figure velocity
about
channel
tectonic an.lastic
(GPS)
as
and do
The
measured
and velocity
plate. site
possibility
block..
CALIFORNIA
SCIENCES,
to
Interf.rometry
of
blocks,
rates
40
available
not
(0-5
and
the
and
on
all
a
good
geodetic WILP is
relative
7
We
1.
data. mm/yr
along
few
JPL,
mm/yr,
vary
faults,
south Islands),
of 33
trilateration
Within
translation
we
also
of niin/yr).
To deformation
(anel,astic
Results
agreement
stress
the
vectors
as
thousand).
+
with points
the included
UCLA
data
construct
DEADMANL,
Finally
the
relative
much
3
of
data
shown
included
of
data
position relative
mm/yr,
the
and
input
repsect
coast
blind
Lat The
to
(VLB)
show
over
but
show
from
Los
in
we
the
34
in
to
no
E—7 1
LL OLL— O9LL— 0111— craii— O’611— OLCI— a’ 0’
-
(44 I Cs 0 Cd 0 cc
(‘a CM 01
(A
0 0
(A CM a’ 0’
(A a’ Ga 0’ 0 0
UI C, Ci DI C, DI
LA CM 0 0 3AVOI 01 SJWOO1JA NOIIVIS I&flA ‘5) SdO
-_:_‘:1
Q: —1,If
_
?.L
CAL.
IFORNI
fRM
E—(
A
20mm/yr
B
1
ock
S
L
‘p
xxx
N
FIGURE
t
-
E-9 3
response.
interval,
found
geometric earthquake remarkably
of differences
discrete
interesting distances
show laws, means direction
strike
simulations
the
strength). a
state-dependent properties
long
analysis
that can
models
earthquakes.
we the response
slip earthquake
a
way
related
the
cool/shallow
seek
ends
be
and
each
vertical
the which
has
to array
Annual
specifically
large
tripped
Only
The
Physics
Studies
can
models
rather
stress
is to
be
of
Evaluation
USGS
possibility
or
point
and
for
yet
obstacles
At
or
done Department
contrast
understand
recurrence
in
fault
sources
resistant
of
general
with
lead
work
are
have
and
limited
strike-slip
state
geometry
been
each
on
However,
constitutive
into
segments
distributions Summary
than
on
are
perturbed,
of
(such
program,
friction
by
rather
segments
to
examining
crust
sufficiently
no
accomplished
the
transitions
earthquake
constrained
depth,
done
a
underway
spatio-temporally
boundary
to
in
and
to
to
that
characteristic
to
study
well
of
fault
as
the
further
and
quasi-independent time
slip
whether
slow
of
breakup
along
fault
on
applies
the
there
into
and
cellular
division
Report.
constitutive
conclusions
Earth
defined
using
can
properties,
positive
of
sometimes
transmission
crustal
moves,
Harvard
(if
and
limitations
characteristic
computations,
between
conclusions
coordinated
refined
which
the
strike.
integral
on
is
study,
sources.
the
lead
by
uniform
on
models
and
into
which
much
so
automata
identifying
response
Southern
of
data
continuum
local
earthquake
deformation.
in
the
in far
to
Planetary
University,
a
can
spatially
there
grids
long-term complex
elastically
hot/deeper
that properties
characteristic
equation
long
current
which
only
on
on
fault
use
to
properties
that
Strong
fault
at
be
on
earthquake
characteristic
this
granite
the
would
would
James
to
in or
segment
lab California
October
planar
and
studied),
slightly,
earthquake
surface,
properly
possible
the
other
approach
the
models
spring-block
limit
concept.
complex
mode
topic
interest
Sciences
based
slip
methods
geometrical
which
average,
topic Cambridge,
deformable
in
regions
presence
under
R.
be
1
seem
fault
applied
investigators
upon
response.
the
has
ends
1991
earthquake
by
Rice
consistent
from and
suggest
rate-
response,
incorporate
physical
which
Earthquake
are
in
constrains
models
the
slip
to
zone modest
These
and
been
hydrothermal
behavior
based
the
reduction
A—B
concept
(V=slip
as
at
very
be
uniformity.
everywhere), and
of
continuum
arrays
patterns
Division
an
a
variations
show
concept
crustal
into
MA
a on
that
significant
basis
We
involve
[=Vdt 55 (V)IdVl
origins
imposed
state-dependent
strong
tentative.
on
are
variations
with,
with
behavior,
computational
have
of
segments the speed;
3D
with
uniform
02138
in
are
Center,
the
spatio-temporally
either
of
for
blocks.
along
the
an
Southern
that
of
range
say,
elastic
cell
case studying
drawn
conditions, Chinnery
of
limit projects
simplified such
characteristic
Results
Applied
intermediate
segments.
“plate”
may
property t 5 =steady
characteristic
uniform
essentially
Nevertheless,
at
of
a
or
USC
strike.
fault
for
of
along
two
(which,
as
least
interactions
These
frictional
grid
based
lead
characteristic
the
jogs
varies
on
thus
such
friction
velocity. analysis
California
models
P.O.#569928
or
solution.
Sciences
for
classical
size).
variations
This
being
strike
in
to
necessity
the
more
are
If
on
or
state
unfortunately,
far,
possibilities
the
uniform
phase
complex
recurrence
earthquake
a
this
physics
with
constitutive
stepovers
inherently
driven
sequence
earthquake
is
laws.
are
negative
the
with
from of
along-strike
variation
between
frictional
in
The
feature
Rate-
friction
slip
depth,
locking
results
of
along
slip
fault
Here
such
of
slip
on
and
of
at
in
in
is
in
m a E—IQ
fault ±
per periodicity if
straining found recurrence
thickness, event hypothesis
earthquake,
the cycle
motion earthquakes,
creeping
events. crust,
zones plane
effects
extensively Parkfield
prior
by consequence
Prieta
segment the
(and
by
crust
greater uniform
(Plausible somewhat underlying the deformation,
support mode.
loading
Parkfleld
11
Parkfield
our
occurrence
km
fault
support
yr
perhaps
segments
earthquakes,
is
they
would
extending
and
and
in
earthquake).
Crustal
ongoing
of
of
Future
The
at
for
3D
Palmdale
coupled
Free-slip
from
of
along
fitting
zone
is
fault
predicted
patch
prior
35
thickness
a
and
everywhere
ranges
are
constrained
a
interval
of
in
lower to
finite
shear
earthquakes
of
slip stiffer
focus
studied
fault
be
mm/yr
ruptures.
over
of
in
characteristic
1906)
time.
some an
is
segment
lower.
extensions neither
a
strike
deformation.
with
time
great
other
must
predicted
the
USGS
the
rates
to divided
to
postdoc,
NSF/EAR
element
flow
crust
boundary
for
segment
network,
in
the
and
and
the the
by
would
interseismic
model
extent
so
These earthquake(s),
assumed
their
of In
and
the
over
earthquakes.
inevitably
relaxation
geodetic
in
If
top
independently,
models
in
top mantle
in
more
far
freely
the
the
by
relaxation
supported
that
the
Parkfield
are
first
into
periodic
the
the
neighbors.
calculations
Y.
are
for
ous a
12.5
tend consists
of
affected
is
relaxation
next simple
which,
in
show
periodic
length
creeping
grant.
responses
conditions
viscous
Ben-Zion,
mantle,
that
creeping locked
The
the
year
for
slipping
envisioned
about
previous
flow
of
work
data
km
to
event.
have
properties
period,
mantle.
the
the
the
non-uniformity
aim
1906-
analysis
lengthen
modeling
time
earthquakes
is
at
scale
in
Crustal
of
Likewise, and
by
through
along
1992
by
portions,
on
earthquakes least
relaxing
type
viscoelastic
consistent
loading
zone
some
times
zones
a
is
parts;
is
are
region
accumulated
its
For
are
7.5
Li
the
modeling
who the
17.5
near
and
of
long
to
to
±
neighbors;
That
examined).
the
employed
sometimes,
and
are
in
of
imposed deformation
significant
creeping apply
yr, time
example,
with
Parkfield
9
are
analyze
a
along
here,
locations
1857-type
km
is
a
rate,
the
postseismic compared
San yr
layer
between
viscoelastic
the
on
higher
rather
locked
Rice
plane
are
studying
consistent
with
dependence,
if
time
elastic
lower
which
at a
upper
NW
2
of
in
Andreas
in
and
the
strike
a
physical
(only
Parlcfield
slip
on
interactions
(JGR,’87)
time.
response
to
SAF
contrast
zone
Pacific/N.
for
is
then using
than
segment.
since
responds
fault
lower are
influence
end
hence
them.
model
locked
earthquakes
crust
the
mantle.
upper
is
slip to
deficits
Parkfield,
similar
corresponding
the
geodetic
constrained
to
deformation
lower
the
with
the assumed
the
of
Shortly
fault
system,
zone
the
the
model
is
the
crust
ratio
may
seismic
to
dictated
the
An
may
such
to
far
Parkfield
1988 crust,
imposed
against
and
statistics
the
1857 America
in
The
that
near
issues
plane NW.
on
crust,
The
of to
this
imposed
1857
field
be
a
of
relaxation
not
after
data,
range, of
these
Fares
examine
interactions.
other
the
the
±7
to characteristic-like
generate
discussed
a
SCEC
event.
relaxation
time
somewhat
Middle
crust
behavior,
crustal
by
by
7.5
be
Thus
provide
slip,
below
velocity,
of
segment
range
following under
time
kinematically
yr
to
an
segment.
including
of
relative
are
timing
possibly
driven
which
seismicity. and
albeit
dependent
lan
estimate
constant
the
has
past
adjacent
except
funds
Thus,
the
time
dependent
the
Mountain,
time
stressing
slip
being those
complementary
Rice
viscoelastic
1857
on
a
the
of
time
events
further
loading
while should
poorly
regions
properties
and
by
single
The
motion
dependence
rates
the
are
the
a
characteristic
is
The
in
best
that
an
locked
(EQS
based
about far-field
particular a
great
and
postseismic
to
15
loading,
great
model
An
steady
previous used
underlying
1989
at
later
earthquake
and
along
we
elastic-brittle
from
vertical
estimate
layer
constrained,
constrained
be
surface
yr,
rate
the
of
the
1906
and
,88).
important
on
1
find lower
affected
rupture
to
are
the Loma
and
in
to
next
most
times
shear the
of
the
share
in
their
fault
the
2
page,
1857
that
fault the
1995
of
yr of
FROM
o Pa1is
data structures
Dr. difficulty
te.rmces
USGS motion
of
being
Ward,
Working
fault
Ward,
San beet
researcher,
moee spread
of
Middle are
and recurrence
forrnulaL,d
than
Synthetic
the
robust
can
one
quake substantial
fault
can
directed
Middle
?Jarine
(in
this
this
earthquake
Syntheik
unbiased
appropriateness
Vakn.Ise,
extrapolate
Andreas
recurrence wepted
improve
sets
fitting
ZHST
Verdes,
press),
segments
about
aperiodic
accommodated
segmentation
Menlo
S. S.
recurruuc,
work
upect on
propved
in
America
around
Group
N.,
In N.,
(i.e.
Progress
toward
American
that
further
sdsrnicity
some
TECTOHICS-UCSC
Ternices.
time
in
from
lognorma]
progrta
Steven
translating
ten
A
199 Sesmicit
Is
An
Park
earthquake
and regardless
fault
and
of
January
Ventura
a
occurrence,
thcy
SCEC
Synthetic
to
Palo.
disrupts
than
rates
of
on
times
are
seismicity
is..
SCEC
Application
the
a
Trench,
in
demonstrating
intervals,
possibly
north
ore
the employ
catalog
and Report,
with
Ward,
California
embody
the
strongly
Trench,
and
of
in
models
pwjct.
Space
has
Verde.
to
Working
between
faults
the
idealizations
or
submitted.
this
have
anticline)
192,
visiting
Nishenko-Buland
of
the
and
time
an
±10%.
iuwrporating
Most
recurrence
it
SeIsmicity
been
Welbu]]
Institute
actual
the which
geological
sampling
but
to
I
October
forecasts
project
would
eye
geodesy
terra.ce’daLa,
can
regional
reviewed
Jour,
closest
inland. and
biased.
of
higher
and
or
Earthquake
Newport
of
Groups:
Palos
suggested.
the
Synthetic
researcher,
toward
also
Increasing
how
slip
can
Huntington
recurrence
my
and
distributions
continued
take
Compared
as
Geophijs.
‘
effect
to
of
to
1901:
statistics
Model
constructed
duration
magnitudes
tells be
Recurrence
aeisixiicity
be
Although
Verde.
predictability
documented
SCEC
ud
synthetic
his
the
Tectonics,
envision
the
the
about
used
reviewing
Inglewood,
C
extended
iubstantial
is
hypothesis earthquake
us
Sciarnicity
geomorphological
Marine
through
(Fault
Prediction.
coast.
With
for phyIca
gap
In
weak.
related
Interval
Rca.,
to
and
funding,
that
Mesa.
to
while,
and
200
January
the
continuing
I information
judge
(
ieiainidty
time
from
forecasts
intervals
the
have
than
University
15.
I
Zone
the
as
Terraces
the
in
bone
8
and
Middle
(submitted),
have
0.75)
and
of
dislocation
For
faults
and
research
long
on
two
arrival
will
Calculations
datA
(See
generally
0.1991
cycle.
the
synthetic
base
faulting
nimfy
have
not
we
conditional
400
Geology)
other
1992,
the
the
the
recently
causes
are
Journal
as
almost
stability
intend
estimated
constructed
America
yet
calculations,
Figure
set
the
Dr.
of
and
years
actuafly
over
MAT, other
probabilities
needed
of
I
of
data
much
geological
in
of
find
should
the
on
done
14133
through
prQject alifornle.,
Lajele
Dr.
modeling,
increases
aelarnicity
Earthquake
earthquake
1991b.
the
roughly
several
and
met
certainly
to
of of
the
2).
hand,
that
Transverse to
- hi
probabilities of
the
larger
Trench,
to
Barrient.os,
so, Gcophsical
monitoring
past
apply
been
from
common
Earthquake
certainly
E
Isolate
with
heights
Stress
and
be
from
into
estimates
spread
static
I
Information
(Gendesy).
based
units
estimates
..“
North-South
don’t
of
conditional
seven
statistically
than observed,
sciarnicity
(see
into
Santa
I
be
Dr.
this
Jour.
1992.
have
hazard.
recurrence
a
Occurrence.
interaction
the
and
of
too
Range..
dislocation
rate hand
FIgure
parameters
on
increase
Ken
slip
foresee
the
technique
mouths
proposed
to
a
magnitude,
Rejcardz
—
Crui
of
fraction
discussed
Statistics.
ages
SCEC
short.
the
The
Geophy8,
of
estimate.
constrain
fault
0.2].
rates
Lajoie
full
and
on
I
samples
earthquake
probability 1)
significant.
have
concept
of
motion
too
the
arrival
The
to
the
of has
are
intrinsic
between
they
slip
For
visiting p•
for rna.rinc
papers,
of
by
theory,
to
be
of
of
earth
much
other dome
made
more
pace
been
they
81
Rss.
The
rate
this
and
less
slip
the
the
the
the
the
the
arc
far
of
is
of 1 ‘.7 1
• o TIME (YEARS) : • 0 DISP STEEN -n , (METERS) (BARS) 0
— I I I I I I I I I I z U’ tO -4 0 0 -4 m C, -l 0 = •4 C, T C I C, N U’ -o C, 0
—o 0
—o 0 a ‘a ‘a
—o a 0 a
I.’ 0 —o 0
10 —o 0
FROM
of
reprenLs
rPT1m
quakiis. (1987).
are
tatistics.
E-4
Fig.
t.he
suhatantially
I-I
ning
INST
distributions
0
0
1
2.
I
The
Each
free
the
Best
TECTOHICS—UC5C
larger
statistics
parameter
dietribution
fits
larger
are
spread
of
IognorrnaJ
than
of
not
wa
recurrence
refiecta
coustant,
was
the
searched
lutrinsic
contre.ined
(tvp)
a
than
TIME
higher
but
for
and
spread
does
generally
bt.
to
Weibuil
degree
a
have
fit
va’ue
iognormal
ie.
(yeas)
at
deciease
the
of
(bottom)
each
B.I99
of
aperiodicity
appropriate
0.21
14.
distribution.
with
proposed distrlbutiou5
141S5
The
magnitude.
In
mean
Wdbni
the
by
The
NisTienko
recurrence
recurrence
to
distribution
spread
the
In
all
model
and
parameters
cases
time.
of
Bukrid
P.
better
major
MAT
they
The
5 a PERMANENT GPS GEODETIC ARRAY OPERATIONS AT SlO (1991 SCEC ANNuAL REPORT) Yehuda Bock, Principal Investigator Scripps Institution of Oceanography, IGPP 0225, UCSD 9500 Gilman Drive, La Jolla, CA 92093-0225 Office phone: (619) 534-5292, Fax: (619) 534-5332 E-mail: [email protected]
1. Introduction
The Permanent GPS Geodetic Array (PGGA) has been operated in California since the spring of 1990 by SlO and JPL with assistance from Caltech, MIT and UCLA. Funding for the maintenance of the network is provided by NASA and SCEC.. The goals of the PGGA are to monitor crustal deformation related to the earthquake cycle in California, continuously, in near real-time and with millimeter accuracy, using a fully automated and economically feasible system. The roles of the PGGA include providing reference sites and precise GPS orbital information to support detailed GPS geophysical surveys in California. For a more comprehensive overview of the PGGA please refer to the article “Continuous monitoring of crustal deformation” by the principal investigator in the June 1991 issue of GPS World. 2. Accomplishments a) We have developed an automated system to collect, analyze and archive data from the PGGA sites at JPL, Piñon Flat Observatory (PFO), and SlO and from a globally distributed set of GPS tracking stations. We monitor data at a two minute sampling rate to all visible satellites, 24 hours a day, 7 days a week. All raw data in the SlO archive are translated into the Receiver Independent Exchange (RINEX) format for GPS data. These are archived on an optical storage device, the Epoch-i Infinite Storage Server. All data collected to data are on-line and accessible via anonymous ftp. We have also archived the Cahrans California High Precision Network data collected from April to August 1991, the San Diego County High Precision Network collected in April 1991, and the GIG ‘91 data collected in January to February 1991. b) Since September 1991, we have been generating a precise satellite ephemeris in support of GPS surveys in southern California. This orbit is available within one week of data collection. The time series of relative position for the JPL to SlO baseline using these orbits is shown in Figure 1. We are currently evaluating the orbits and estimate their precision to be at the 10-30 parts per billion level. c) We monitored the position of the JPL site during the period of the Sierra Madre earthquake. We did not detect any significant motion at the several mm level in the horizontal and 10 mm level in the vertical. The time series of position of the baseline from Goldstone to JPL is shown in Figure 2. d) We are evaluating P-code receivers for possible purchase and deployment in the array. We have run zero and short baseline tests with the ROGUE SNR-8 and Ashtech P-code receivers and are preparing a report for the Crustal Deformation Committee. e) We have hired Dr. Peng Fang to serve as the primary Network Data AnalysL f) We have upgraded our computing facilities to streamline our automatic operations. (g) We have made significant improvements to the GAM1T GPS software package in coliaboration with MiT, and have instructed UClA SCEC investigators in its operation. (h) A highly stable monument has been built at Vandenberg Air Force Base with funding from MiT. We plan to deploy receivers there and at Parkfleld by December 1991, with assistance from JPL. 1 JPL Lo Scripps (PGGA Solutions; Length 171,195.727 m)
4 North 4 Vertical
2 2-
0- 0-
—2 -
[4 —4 I I I I I
- 220 240 260 280 220 240 260 280 ci U) 04 East 4 - Length ci) > 2- ci) 0 0-
—2 2-
—4 —4 I I I I I I 220 240 260 280 220 240 260 280 Day Number, 1991
1. Time seiies of position between the JPL and SlO PGGA sites. Each point represents a 24 hour solution. T ‘-1 GoldsLone 1,o J1L (PGGA Solutions; Length 179,254.461 ru)
4 North 4 Vertical
2- 2
0- 0-
—2
—4 I I I I I I I I I - I I I I 172 176 180 184 188 172 176 180 184 188
. East 4 Length
2 a) 0 0
—2 —2
—4 —4 I I I I I I I I I I I I I I I I I 172 176 180 184 188 172 176 180 184 188 Day Number, 1991
2. Time series of position between Goldstone and JPL during the period of the Sierra Madre Earthquake. The vertical arrow indicates the first point after the earthquake (day 179). We detect no significant motion of the JPL site in Pasadena. Mechanics of Blind Thrust Faults in the Los Angeles Basin 1991 Annual Report to the Southern California Earthquake Center*
Ross S. STErN U.S. Geological Survey,Menlo Park
JJANLINt WoodsHole OceanographicInstitution
GEoFFREY C.P. KINGt Institut de Physique du Globede Strasbourg, France
Goals and Objectives. Geodetic and geological evidence has revealed that the crust beneath the Los Angeles basin is contracting at a rate of 3-15 mm/yr in a north-south direction, and that a suite of blind reverse and thrust faults underlies the basin to accommodate this contraction. We seek to understand the deep fault geometry in the Los Angeles basin, to deduce or constrain the fault slip rates and, where possible, earthquake repeat times. In addition, we seek to understand the earthquake cycle on the blind faults, and to learn how adjacent blind faults interact as they are stressed and relaxed during the cycle.
Our first step was a systematic review of blind earthquake rupture and blind fault structures in southern California. The review was aimed at honing our modeling efforts to fault geometries, crustal rheologies, and earthquake sources appropriate for southern California. Seismic, geologic and geodetic data from the Los Angeles basin played a key role, as well as observations from other fold and thrust belts where the earthquake or seismic reflection record is superior to southern California.
Following the review, we began a set of numerical experiments on simple, isolated blind faults. Thus far we have pursued four broad lines of inquiry, all designed to probe the strain and displacement field associated with slip on blind faults, which we describe and illustratebriefly here:
Boundary Conditions. Here we considered the remote tractions and fault slip distribution best suited to reproduce patterns of observed geodetic contraction across the basin, coseismic surface displacement, and inferred earthquake stress drop. This led to a study of the strain field produced by uniform slip, in which slip is constant along the fault, and by tapered slip of the same mean value, but with an ellipticaldistributionthat minimizes the stress intensity factors at the fault tips. Although thegroundsurfacedisplacementfor these two end members are similar (they are indistinguishable for blind faults), the strain fields at depth are quite different. We illustrate this result in Figure 1, for a reverse fault that cuts the earth’ssurface. Because remote basin contraction will produce tapered slip on freely slipping faults, and because tapered slip minimizesthe artificially high strains at the fault ends, weelected to use tapered slip fortheremainingstudies.
tCcnter Visiting Fellow, 1991-92 *J)foflpjfl Group 22 October 1991 BLIND FAULT MECHANICSIN ThE L.A. BASIN Page 2 E-19
_LLLL 111111111111 Tapered Slip ‘UniformS1ipj: I,1IIIIIIIIIII (semi-elliptical) ____1__1__l1111111 I
I I I I I I 8 6 4 2 0 2 6 4 2 0 2 4 Distance From Fault Tip (km) Depth (km) Distance From Fault Tip (km)
Fig. 1. Boundary-element visualizations of the strain-change produced by faults in an elastic halfspace; fault length along strike is infinite; fault width, W, is 8 km. The mean fault slip is 1 m, and thus the strain drop, u/W, is 125 ppm. The deviatoric shear strain is plotted, or the shear-strain maxima independent of orientation, (elI-E where the unit vector i is horizontal and 2 is vertical; nodes are 0.51”22+’/km apart. Maximum strain plotted (black) is >50 ppm or 40% of the strain drop. For uniform[e slip, shear strain is concentrated at the base of the fault and diminishes toward the124ground surface. For tapered slip, shear strain diminishes little toward the surface. In both) cases, high shear strain extends farthest from the fault trace on the downthrown block (right side of fault).
Coseismic Strain Field at Depth. Here we investigated the shear and dilatation fields associated with tapered slip on blind faults, since patterns of aftershocks furnish a means to test whether the predicted strain changes are observed in the earth. We examined a suite of fault geometries that are found or are believed to exist in the Los Angeles basin, studying the effects of fault dip and the proximity of the fault to the free surface. We found that when D/W (the depth of the fault tip, D, divided by the fault’s down-dip width, W) <1, ground-surface interaction significantly modifies the shear and dilatant strain fields. When D/W>>1, a high shear strain halo encircles the fault, with slightly concentrated strain at the fault ends. As the fault approaches the free surface, however, the halo breaks up into several isolated zones of high shear strain (Figure 2, left panel). These zones correspond to sites of high-angle reverse faults and aftershocks found in the cores of many active anticlines, which lie above the fault tip. The antisymmetric dipole pattern of dilatant strain that prevails when D/W>>1 is also disrupted as the fault nears the free surface, with the contraction lobe (-)above the fault tip becoming dilatant (+) (Figure 2, right panel). This site also corresponds to the region of secondary faulting and oil storage in anticlines lying above reverse faults. If the rocks at this shallow depth are strong enough to store elastic strain, the combined effect of increased shear and dilatation promote secondary failure. 22Octoberl99l BLIND FAULTMECHANICSIN ThE LA. BASIN Page 3
Distance From Fault Tip Distance From Fault Tip (km)
Fig. 2. Strain Field produced by 1 m of tapered slip on a 4-km-wide thrust fault dipping 25°. Coseismic strain drop is 250 ppm. Maximum shear strain shown is >50 ppm or 20% of the strain drop. Peak dilatant strains plotted (black and white fields) are >±30 ppm; neutral grey is ctlO ppm. Note that the nearly symmetricalpattern of strain below the fault is disrupted by the presence of the free surface above the fault.
Long-term Deformation and Strain. We are exploring several ways to reproduce the long-term evolution of structures associated with blind faults, with the goal of linking their coseismic deformation to the geological deformation, with kilometers of cumulative slip through repeated earthquakes. We have tested a two-modulus-region model in which both regions are elastic but one has a lower stiffness. The boundary between the regions can take any form, and thus can reproduce the Los Angeles basin configuration in 2 dimensions, with a basin of compliant sediments overlying stiffer basement rocks, in which faulting can take place in one or both regions. We have tested and successfullyused this to simulate the modulus structure at the sites of the Coalinga and Loma Prieta earthquakes, where the seismic velocity structure is better known than in the Los Angeles basin.
The second mechanical model we are evaluating is one in which an elastic layer overlies an inviscid fluid, in which gravitational forces are balanced and the effects of erosion and deposition are included. Thus far we have tested surface-cutting reverse faults only. Finally, we are examining the deformation field for blind faults in an elastic incompressible solid (eg., Poisson’s ratio, v=O.5). Implicit in the methodology of balanced, retro deformable cross-sections is the rule that cross-sectional area is preserved. In an incompressible solid, the dilatant strains are everywhere zero, so this rule is satisfied exactly (Figure 3). In practice, such a criteria means that mineral transport (dissolution, migration and precipitation) must occur on a more local scale for an incompressible solid than for a Poisson solid (where v=O.25), since there are no large-scale pressure gradients to drive the migration. The shear strain field for an incompressibleelastic solid is similar to that for the Poisson solid, except that the deviatoric shear strains are uniformly higher. 22
Fig. equivalent in
above fault
Next mean drop,
aftershocks associated For strain experiments
pattern study network
October
an
the 3.
faces
Steps.
incompressible
rosettes ufW,
zones slip,
the
Displacement
of
long-term
199 in
tip
surface
with do
to
u.
the S
and
and
of
with
of
For
a
not
to
We
Los
plastic
regional ramp-decollement
the constant focal
investigate
overlap,
an
coseismic
deformation
models,
plan
Angeles
fault,
elastic
field
mechanisms
elastic yielding,
Distance
to
BLIND
horizontal
seismic
in
as
explore
for
we
layer
preferred
Basin
432
the
models,
they
halfspace.
FAULT
a
will
produced
such
6-km-wide
same
over
From
would moment,
will and
with
alternative
continue
contraction
MECHANICS
as
we
fault
position
adjacent an
be
2
the
these
by
Fault
for
have
inviscid
optimized
Tapered
orientations
fault
reverse
uW.
slip
a
studying
results.
Poisson
slip
restricted as
parallel
IN
Tip
of
on
tip
anticines
We ThE
substrate.
4.24
normalizations,
slip the
fault
or
to
(kin)
We 0 LA.
will incompressible
fault detect solid.
blind and
faults,
of
km.
dipping
ourselves
also BASIN
also 3
interactions.
bends,
are
Finally,
faults
km
slip
An
and
plan
observed
compare
45°,
is
anticline on
to
so
and
to such
to
imposed
these
examine
we
buried
that
study
solids.
models
carry
intend
as
blind
in
the
faults.
is
constant
4 the
2 the
at
1
8 out
0
seen on
9
Center’s 10 slip
We
with
3
to
earthquake
strain field.
Depth km additional the
(km)
lines study
will
to
Page
1
depth
strain fault,
form
m
field
The GPS
also
and
4
the
of Group F: Regional Seismicity Group Leader: Egill Hauksson
Towards, real-time, routine broad-band Seismology Hauksson-Kanamori-Helmberger (Caitech) F-i
Shear Wave Splimng in the Los Angeles Basin Li-Teng (USC) F-4
Seismicity Studies in Southern California Associated Nicholson (UC-Santa Barbara) F-6 with the San Andreas Fault System
Fault Kinematics from Earthquakes in Southern California Seeber-Armbruster (Columbia) F- 10
Research on Regional earthquake Source Imaging Lay (UC-Santa Cruz) F- 13 and Regional Propagation
meeting
Using
Kanarnori,
Hauksson,
Publications
California,
Geologists
Earthquake As
step earthquake
magnitude,
use.
quickly
software,
dial-up
many
The
software
TERRAscope The
Results
E.
Investigations
Project
Cahech Towards,
techniques
The
Report
SUMMARY
Hauksson,
a
a
Sierra
objectives
goal
part
paper
TERRAscope
as
assembled
1
1991
system
as
six
development.
the
of
of
E.,K. in
H
1978
possible,
for
data
Madre
presenting Arias TERRAscope
real-time,
seismic
software
southern this
our
,
H.
broad-band
analysis
E.
of
at
in
1990, Hutton,K..
Kanamori,
REPORT:
for
project
contribution
Hauksson,
by
group
Pasadena.
earthquake
various
--1991
we
moment,
Egill
Southern
used
California.
submitted
the
of
routine
determined,
This
F
onscale
is
seismic
stations.
last
Hauksson
formats
are
Douglass,
for
Sierra
and
to
and
is
To
of
to
the
13
and
establish
California.
to
REGIONAL
D.
the
to:
synthesize
28
broad-band
years
network.
broad-band
T.
obtain
Sierra
the
Macire
to
The
Helmberger
focal
Engineering
first
June
from
peak
Heaton,
and
make
master
of
data
Madre
rapid
major
quantitative
Earthquake--,
mechanism.
L.
seismicity
contributions
1991
1
ground
The
them
existing
were
Jones,
data
Experiment
model,
data
earthquake
event
provided
seismology
goal
Geology
SEISMICITY
quickly
more
from acceleration,
-
Earthquake
analysis
has
earthquake
of
that
we
information
southern
useful
by
this
In
been
(abstract)
of
retrieved
a
have
Toward
was PIs.;
is
realistic
addition
project
southern
being
methods
written
to
recorded
Atlas
started
California.
local
data
11
the
via
submitted
on
Realtime
automated
September
is
test
(GROUP
end
to
California,
and
for
for
magnitude,
to
the
the
for
a
developing
for
present
onscale
the
user.
Southern
to
project
Caltech
earthquake
data
develop
our
Engineering
Seismology
to
for
1991
As
fall
from
ongoing
1991.
existing
F)
with
routine
called:
energy
gopher
a
AGU
new
first
new
the
as as F-I PIs Title: Institution: A INVESTIGATIONS realistic methods recorded RESULTS event Caltech (34°15.7’N; quickly first parameters and maximum and Although instrumentation energy-moment good timely Madre the stress-drop” automatically and results of SCEC significant 1989 the the The The The To energy The that two Presently, enough use should earthquake as gopher test for Project: on radiated seismic six acceleration obtain was Montebello goal short Sierra of some possible, hours later, scale data 118°O.1’W). for TERRAscope using earthquakes. TERRAscope magnitude, be to recorded earthquakes. dial-up of Goals and ratio duration Egill Heimberger California Towards, our we from energy; the available Madre moment quantitative minor the be after is this the we ongoing software are ML=5.8 Hauksson, at earthquake, used and similar for results the system with data project performed Pasadena the of earthquake stations, implementing changes Progress Real-time, The seismic TERRAscope Institute this using within data 3) for the as collected software occurrence: to of Sierra implementation determination at information many data Hiroo earthquake for is emergency SH 1 the PAS, the these Pasadena. (7% and 30 Report. the of were to moment, the pulse Routine were Madre of to as Kanamori, 1988 records by development. Technology belongs of GSC, quick establish following Sierra broad-band software 60 six 28 the g); 11 made 1) at quickly Pasadena minutes indicate TERRAscope services. PFO, earthquake of on June Sept. Broad-band 2) This Caltech Pasadena and Madre determination from greatly determinations to and the determination rapid the analyses for SBC, to a 1991 is 1991 retrieved focal seismic Don of focal PAS group that earthquake. determine Gopher. the earthquake earthquake some facilitated ISA, data the and of The provided Seismology stations. first mechanism mechanism within the and 6/28/1991 occurrence network. of and the analysis via uniform of of of proved “high major Sierra SVD. These SVD large local ML the and the the the the as a
Earthquake----,
Kanamori, Engineering
Realtime Hauksson,
Publications
Southern presenting
southern
to project
this
As
make
a
project
part
called:
California, them
California.
E.,K.
Seismology
of
the
H
Geologists
is our
- Earthquake
,
more
last
to
(abstract) Hutton,K.
- E.
•
contribution
present
I
1978
13
useful
Hauksson,
1991.
in
years
Using
submitted
Douglass,
1990,
southern
existing
Atlas
to
of
to
the
TERRAscope
submitted
and
for
2
the
seismicity
end
earthquake
California.
and
to
Southern
T.
master
fall
user.
L.
Heaton,
to:
Jones,
AGU
has
model,
Engineering
As
data
California.
---1991
been
Earthquake
meeting
a
Experiment
in
we
first
written
various
have
Sierra
step
Geology
1991
The
Atlas
started
a
for
formats
Toward
goal
Madre
paper
for
the
of
of
a F—S
obtained
major
the
observed
sequence
initial
ts
gradually
SCS. 42 earthquake
shocks and 28,
during
1.77 wave reverse-fault
splitting events
from
having
the
0.06
but
near
medium
average
under
Angeles
basement increases
crystalline
leading
split
last
basin
active
Angeles
to
km
other
1990
with
4.1
Santa
tp
after
three
of
the
shear
various
The
value
earthquake.
before
(on
the
last
tectonic
NNE
with
van and
(6)
(5)
near-vertical
crack
(4)
(3)
Montebello
(2)
We
(1)
shear
crack
evidence
the
observed north
basin.
three
containing
62
basin
in
temporal
suggest
as
present
the
the
delay
Monica
that
three
years.
waves
basement
(P)
Report
abruptly
3-component
of The
3-component
Shear-wave A
are
We
large
km
the
leading
of
controlled
density
directions
Yong-Gang
event.
USC
and
wave
density
multi-station
1.74
M>5.0
end
region.
on
contains
Shear-wave
seismic
systematically
SCS.
travel
ENE
observed
years.
examined
time
oreand
regional
for
has
epicenters
3-component
that
Dec.
Because
after
at
campus)
of
t
Mountains
to
polarized
variation
double
shear
on
The
aligned
decrease
beneath
the raypath,
at
station
But,
been
of
of
the
1.66
the distance
major
increased
We
dilatancy
We
3, network
region
the
splitting
Shear
have
three
0.04
SCS
temporal
vertical
ratib
data
Li,
data
after
initial
Newport-Inglewood
shear-wave
wave
N-S
have
we 1988,
when
did
observed
whether
northern
there
array
earthquakes
splitting
has
earthquakes
cracks
SCS
show
the
USC,
in
in
in
and
a recorded
found
examining
by
there
obtained
then
may
in
M>5
not
compressional
the
show
obtained
scattered
Wave
N-S±15°
recorded
polarized
the
value
data
22
both
Los
the by (LABSNET)
might
with
the
were
(SCS)
microcracks
a
M5.8 related
variation
find
Montebello
that
Group
change
factor
returned
basement
km
is,
is
a
part
may
double
there
remote basement
from
that
Angeles
show
complicate
splitting
time
10-130
after factor
used
no
significant by
in
Splitting
the
occur
NE from
installed
Sierra
(MD)
the
of
—60
be
the
pattern,
3-component
independent
(F,
parallel
to
three-component
of
general,
events
station
F
is
the
polarizations
in
several
the
delay
of
main
shear-wave
to
following
due
the
(Regional
to
earthquakes.
U
3
of
time probable
the
rock
ms
events
basin.
stress
shear-wave
in
for
crack
SCS, increases.
is
fit
in
occurring Madre
fault
after
main
NW.
the
and
—1.5
to
major
at
interpreted
the
borehole
difference
occurring
shock
GFP
the
suggesting
to
shear-wave
shear-wave
evidence
in
and variation
beneath
delay
months. no a
the
initial
regime.
SM)
The
M5.5
(INF)
the
normal
the
within
preferential
rock
shock.
the
before
seismogenic
results:
earthquake of
shear-wave
located
tsltp
northern
seismograms
earthquakes
temporal
Seismology)
occurred
NW
main
splitting
time
varied
the
in of
on
splitting
value
Los
around
Upland
station
the
also
They
We at
the
in
this in
of
station
leading
from
Ray
ratio
ts/tp
in
June
location
fault
and
that
delay
the
time
shock.
traveltimes
Los
splitting
on
splitting.
shear-wave
terms
Angeles
shear-wave
show
also
before
margin
time
after
tracing
alignment
and
variation
ratio
horizontal
are
during
basement
depths
from
DHB
(i.e.
trace.
(SM)
the 12,
related
earthquake
Angeles
microcracks
splitting
delay
close
layer
of
in
shear
of
found
recorded
of
of
period,
abruptly
The
several
the shear-wave
1989,
focal
the
decreased
Figure
the
SCEC
and
data
of
events
events
at
anisoiropic
on
for
Ray
For
of
Basin
the
to
of
beneath
to
M5.0
slow between
waves
delay
depth the
area
that
splitting
window
June
of
basin.
of
after
the 6-18
outcropping
focal
the observed.
to
zone
them.
10
and
tracing
we
double
those
months.
Los
by
microcracks
increased
(U)
vertical.
shear-wave
(Figure
1)
of
the
anisotropic
shear
km
slow
of
within
of
Pasadena
time
could 28,
km
the
yields
USC
traveling
from
zones with
before
splitting
the
Angeles
the
on
interest
ratio
the
east
1500m
crustal
events
during
in
yields
in
1991,
M4.4
main shear
wave
Feb.
then
MD
The
Los
Los
2).
For two
this
not
the
the
the
the
an
of
of of
to
in a
Figure
Figure
2
1
diagrams
used
events
in
by
filled
two horizontal
shear USC
extended
beneath
equal-area
a
horizontal
in
bar
campus.
waves
triangle
within
the
Left
having
the
An
of
seismograms
polarization
Left
example
the
are
Los
projections
the
and
Horizontal
panel.
components
Top:
the
aligned
distributions
Angeles
shear-wave
by
maximum
The
of
3-component
an
TWMLTU
diagrams.
of
shear-wave
by
open
of
centre
projections
basin
horizontal
two
the
from
window
of
(S)
length triangle,
lower
vertical
with
of
polarization
90
The
stations
each
seismograms
925
splitting
indicates
components
epicenter
hemisphere
onsets
of
of
respectively.
3
short
panel
TRAV€1TIUE
the
the
MSH
5998
observed
five-station
of
polarizations
directions
lines,
corresponds
the
118
the
8.9
and
recorded
of
(3)
direction
from
1147
leading
respectively.
km
EXG.
directions
Bottom
for
1823
east
stations
of
array
an
The
to
of
leading
and
by
of
event
of
Right
the
the
the
SCS
station
out
maximum
slow
SCS
MSH
station
Bottom
leading
maximum
occurring
of
shear
at
Aspect
station
shear
C—
55°.
the
MSH.
and
location.
Mi&
waves
USC
aspect
shear
waves
EXG.
Right
ratio
array
principal
at
Bottom
The
the
campus
waves
from
ratio
diagrams
Arrivals
are
Equal-area
installed
depth
corresponding
pointed
Left:
stress.
represented
fifty
are
from
of
of
for
shown
events
Time-
18
in
split
by
fifty
rose
km
the
the
a 2 for of regional REFERENCES Network 3) completed on Funding Nicholson, Hull, active hypocenters which mainshock other Nicholson, Hartzell, have velocity tomographic mechanisms 2) Ma), California, ANNUAL Microearthquakes exhibits 1986 1) this systematic Revision the Res., Databases A finite Coachella southern sequence Springs 1986 begun has geophysical North preliminary A.G. subsurface suggests earthquake project stress for oblique-slip S. North anomalies (SCSN) 94, fault: accommodated PROGRESS was C., to (1989). C. and [Hartzell, Earthquake, this and images of Palm and (Figure California, (abstract), 7515—7534. be field of Valley until mislocation R.L. and that Palm a Application completed project submitted C. manuscript organized to earthquake faults. evidence, along (Figures high-resolution Springs both is the Comparison 15 J.M. determine of Wesson, Nicholson along Associated 1) Springs along oriented from Institute September 3-D distribution 1989] was REPORT and Seismicity on EQS Bull. segments Geophys. Lees only velocity [Hull 2 the inversion for discontinuous, (NPS) and suggest received to the that for on arrival and Santa earthquake D. and Trans. about Seismol. for the the re-analysis. (1991). fault a northern the off (1991). of publication Given, and may 3). small 1991; with most Crustal tomographic correct 1986 of of earthquake Barbara, Res. seismic perturbations seismotectonics 45°—55° that the times NSF-USC Studies AGU, 8 that Nicholson, high-velocity of the be May High Soc. this of the amount Lett., active Craig North J. the aftershock and Seismotectonics Elsinore southern correlate present Studies, from orientation, Boatwright the non-vertical waveform A 67, report 1991. axis Cal{fornia Am., from San relocation resolution major [Nicholson in 4 Nicholson Palm aftershock fault along PO# inversion the 1089—1090. pp., of within of 19911. Southern the Andreas fault 19 documents Due University in anomalies San cumulative maximum Southern arrival of purpose with 572726 segments Springs, submitted. pp., general the the inversion and geometry, the 93106—1100 zone of to the Andreas and travel-time Based most of submitted. prior the southern current hypocenters Northern and of C.R. times northern possibly Fault P-arrival of appears strike California of California progress California, may 1948 the horizontal exhibit this commitments, on of results Lees, horizontal Cal Allen and of and northern outline SCSN geological, seismic of Desert study jfornia System San Elsinore Coacheila the tomography to northern curvilinear the style times through (1986). for 1991]. similar be [Nicholson 1986 Regional Andreas earthquake, active is compression the catalog Hot relatively the offset of to slip Elsinore in fault asperity ML rupture work evaluate Valley deformation Springs Aftershocks oblique-slip seismological the San 15 Elsinore The during fault 5.9 in October (—15 fault Seismographic zone, of vicinity et Jacinto did young the preliminary reveal J. North earthquake responsible history fault at., earthquake (a 1 ) the segments. km), not has Geophys. the fault. southern northern degree 1991. 1986], of of begin faults of
along (—2.5 F-b high- focal zone, been Palm NPS and and of the the the a Figure
from diagram been
a
the moved
1:
indicates
Elsinore-Temecula
Lower-hemisphere,
slightly
the
orientation
from
their
trough.
equal-area,
a
(100
original
Compressional
groupings)
single-event
locations
of
the
so
quadrants
focal
that
35
P-axes;
focal
mechanisms
are
a
spheres
average
shaded.
for
do
azimuth
Some
35
not
selected
overlap.
mechanisms
is
353°±52°.
earthquakes
Inset
have
rose
a
U-) 0
F7
permanent Figure
for
orientations
layers
2.
Greyshade
1
C.
A.
stations
through
shown
Layer
Layer
3:
1:
in
used 6,
IIb4/J
3.2-5.0
0.0-1.6
plots
figure
as
to
derived
km
of
km
monitor
4. P-wave
6lknvi 1 .40
.45 from
H6:34) the .5Ionin1
%Sk,,.iii
velocity
Springs
the
seismicity. 3-D - 427 - 10
10 perturbations
kin/i tomographic 8
8
F.
D.
B.
Layer (a)
1
Layer
Layer
Line
6:
4:
2:
(in
<:
12.0-24.0
L
inversion.
5.0-8.0 profiles
IIb4U :yiI
1.6-3.2
%
slowness)
km km
km (A
t2 1!
, Triangles kmkznli
500ki, I 8 — .45 .40
0.0
D) relative IlbiW 8 %3)ownin. ‘SInii
%S)o—n..
reflect
are
to u.u5 4.b.inji
fl92
cross
portable a kin/i 8
8
l-D
section
model and ‘.1
— 0
I I
I g ‘4 w
CA (A (A
0 0 1 01 c1
to Greyshade vertical cross sections of velocity perturbations (in % slowness) relative Figure 3. are shown in individual layer velocities as specified in the 1-D model. Cross section orientations side are shown as small Figure 2a. Earthquake hypocenters projected from within 5 km to either slip distrubution of the open circles. In D-D’, superimposed contour lines are the projected dynamic NPS mainshock lafter Hartzell, 19891. m OCT 22 ‘91 09:55RM LDGO—6LISr1OLOG’ F-o Fault Kinematics from Earthquakes in Southern California
L. Seeber and John G. Armbruster, Lamont-DohertyGeologicalObservatory
1.Bac eauake data. We are generatinga file of relocated hypocenters and quality-selected focal mechanisms -- earthquakedata readyfor structural interpretation -- for the period of good network coveragein southernCalifornia (1981-1990).We are accomplishingthis task in a singleprocedurefor an area coveringmost of southern California (32.5N-35.5N; 114.5W-120.OW).We have tried a variety of schemes to obtain accurate locations; as a.firstiteration,we are satisfiedwith ajoint relocation technique based on location-dependentstation-corrections(1DSC). We have divided southern California in 40x.55equal size boxes;correctionswere assignedto each box far each of about 175stations on the basis a subset of the best constrainedhypocenters in each box. Other, less frequently used stations were assied fixed corrections.Hypocenters were obtained by assigningdifferent 1-dimentionalvelocitystructureto each ray path. These path-dependent structuresare derivedby combiningstructuresobtainedfrom the letterature for ten sub-regions of southern California.Smoothingwas applied to the LDSC’s according to the density of data, Finally,locationswere obtainedon the basis of LDSC with smoothing again applied to avoid stcp in locationbias at theboundaries between boxes. About 20,000quality-selectedhypoéenterswere obtained.These locations are relatively accurate, but they may have substantialSyStematiCerrors. The task in the next stage is to define these errors and to accountfor them by appropriatemodificationsof the LDSC’s. Results were compared with known locations (e.g., blasts) and with results from other studies involving subsets of the data we are considering. The entire process was repeated with differentparameters.Satisfiedwith the preliminaryresults, we have proceeded to obtain focal mechanismsby a grid-search schemewe have developed. After quality selection desiiedto oprlrni2eresolutionand based on variousparameters, we have obtained about 10,000well-constrainedfocal mechanismsfor southernCalifornia,The structuralinterpretationof the mechanismsandthe correlationbetweenseisuica11ydefined and mapped faults will provide constraintson systematiclocationerror. 2. A StructuralModel of Seismicity, The main contributionof this project will be to provide a set of focal mechanj.smsfor which the nodal plane correspondingto the seismogenicfault has been chosen -- a structuralinterpretationof the earthquake data..This model, which is now beginning to take sha?e, includes a.l1mechanismsthat can be associatedwith faults postulated on the basis of the earthquakedata. The main criterionfor our Interpretationwill be correlationbetweenfocal planesand planes defined by hypocenter distribution.The main test for the model is provided by the correlationwith mapped surface faults. We have developed a set of programsfor manipulatingand visua]iaing locations and focal mechanisms in 3-D and for assigningfault planes. We intend to expand its capability to include mapped faults and other pertinentdata. With these tools, the massive task of interpretingall the availabledata for Californiais a realisticone. From our experience in central California, we expect that at leasthalf the focal mechanisms,or about 5,000events, will be included In our model. We expectour effort to yield a valid and useful interpretation of the data, but certainly not the only possible one. The final product of this effort will ineludt th tools to derive alternativesuctural models from the mechardsms. 3. EarthquakePrediction. In centralCaliforniawe have used the strucnrai model to monitor changes in s’ess during 20 years prior to the Loma Prieta earthquake, We found that a sudden change in the patternof selamicitystarting6 years prior to the main shock could be modeled as a perturbationon the strainbuild-up caused by hardening of the fault patch about to rupture (Seeber and Armbruster,submittedfor publication). A 6 year precursor for a M=7 earthquakeis consistentwith worldwide data on precursor-time versus magnitude. A structural modelof seismicityis a prerequisite for this type of analysis.We intend to search for precursor on past and on future earthquakesin southern California as soon as our first structural modelbecomes availablefor this area. “-I 3i 37 34.13G4 116.143 --116.590 I —I’6.9 + I + + C -; C S + 4 -I •1 + -f
+ cli •1 -I- 4 + 4 * 4 cD 4 (.)1 -t :4 -I 0] $ D + 4 + + -I. F Ip I— - + + + + C .4 -I. -I. 4 + U] + •1 m + (1) + 4 -1• 1 0 + + F 4 + 4 + 0 4 •1 C + + + + + I * I. a I 4 -e t +. 4 + +
.4 :U’. 3-I. I’1 n. qe
Figuie 1. bxamples of hypocen obtained by the location-dependent station cxneciion technk (U)SC). Above: two ilentical sections acioss the 1986 North Palm Springs aftershock zor (looking northwest): left by U)SC on data 1mm the fixed netwodç right by Jones et aL, 1988 incorporating data from temporary stations. Compare overall shape of zonc; the two sets inco.porate different eanhquakcs. Below: Section along San Jamb fault (looking northeast). Depth distribution ofU)SC hypocen is similar to depth distribution T1 obtained fitxn detailed studies ofdata from the Anza Netwc& 0
•0 C
C
0
0
0
on,
C
OCT
Figure
represented Mountain, fault; San
structures complex included in the
0
the
22
Borrego
0
Jacinto
0
C
lPFInspiraon 00
0
1968 00
0
‘91 2.
0
pattern in
I
0
0 Relocated
in
In 0
a and
as fault
0
Badland 9:57M
0
terms
vety C the
0
a
1954 nodal
of
structural zone.
large
of
faults
a
epicenters
tectonic
earthquakes,
burri.eci LDGO—SEI51’iQLQG
Point
plane
SHF—San
rupture.
and
0
model
Fault;
with
folds
knot seismogenic
a
and
1
i.ar*
•
its
O
of
Cl.Clark
along Felipe or
will
surface
kineriarics.
seisnilcity,
whether
help
0 the 0
Hill
a
faults.
stuctures
Lake
us San
It
fault;
0
Qo0C to
0
many
can
The
Jacinto
•
A
understand
;S—Borrego
0
queson
CCF—
aiso
comparison
hi
of
the
fault
these
operate
0
C
0
Coyote
Borrego
0
•
we
the
operates
0
earthquake
C
C
0
Sink;
wish
as
significance of
Creek
an
this
Badlands
C BM=Borrego to
asperixy only
N
model
address
fault
will
as
of
a area be -I
3KM
with
and
CF—Clark
I
ban.er,
these
0
is
C
whether
of
be
the
g
the
as
*
*0
000
0
a
-4
a
0
0
a
4
Ammon,
Publications
inversions
The
wave
Kuge
inaccurate
We
California models
1991
surface
for collaboration We
Results
regional
U.S.
Developing
assess
Thorne
Investigations
Propagation
Project
Research
many
projects
axe
have
Sierra
inversion
installed
Conducting
the
wave
5
Lay,
that
testing
California
C.
implemented
surface
is
applicability
Green’s
events.
Madre
and
regionalized
required.
can
that
UCSC
dispersion
with
on
the
methods,
3.
the
be
we
wave
synthetic
Vidale,
earthquake.
Howard
CMT
earthquakes.
Regional
function
Dr.
used
sensitivity
have
to
between
a
Zhang,
paths
code
models
for
and
new
Tomography
near-real
undertaken
Patton,
simulations
and
stable
in
here
extensive
generalized
Dr.
Earthquake
of
PAS
spatial
the for
this
the
at
time
S.
inversion
short
western
and
UCSC.
involve
method
Schwartz
CMT
with
without
distribution
testing
inversion
SCZ
3
period
inverse
the
U.S.
inversion
Our of
broadband
long-term
is
rays,
CMT
of
Source
Rayleigh
regional
and
being
and
of
first
the
theory
of
B.S.S.A.,
the
myself
inversion
California,
sensitivity
the
data applied
method
long
development
stations.
surface
wave
approach
Imaging
broadband
application
are
period
submitted
to
code
propagation
to
completing
regionalized
and
to
waves
source
to
of
develop
waves
array.
robustness
of
and
tomography.
H.
wifl
June
from
rapid
Kawakatsu
mislocation,
recorded
involve
the
a
Dr.
dispersion
Regional
1991
inversion
southern
study
regional
western
Keiko
of
the
the
for
of
In to /
FROI’
UNIVERSiTY PAX: bwlm (405) —-p.
Dr.
Qmirman:
DearEglll Imaging Southern Re:
myself, graduate for Howard
generalized paths wlLhout
Inversion and
stations, period inversion method array. an for
earthquake, LLNL source
squired.
have wave
equipment datasaizatlen. 45316.4
(405)
IHST
of
Egill
excellent
siort
this
Research myself
Tatic.
459-2127
in
also
Inversion
My
An
waves
of and
Graduate
postdoctoral
The
ilys’ TECTOHICS—UCSC
to
the project.
and
Haubson
which California
Patton, student
period
of
code knowledge
Group
We
source
AGU
closely
OF
group’s Nevada
are
for
western
projects
regional
Inverse
Regional
for
PhD.
progress by
recorded
are
CALFR)RNJA,
will
of completing
the
methods,
Jeroen which
this
absuact C. Rayleigh
Dr. mislocailon,
student
H.
about
Earthquake
Woildn coordinated two
progress student
recordings.
augment
Aninn
U.S.
researchers
theory
ICawakatsu method
stuface
that about
Kelko
Propagation”,
report
Ritsema.
we
for new
on
is
and
and
we
3eroen
wave
regional who have
a
being theprograrn.
schedule
approach
Group UCSC Kuge and
the
on
study waves
for
Is
California,
cxtcnsWe
have
inaccurate
Ccnt&
SANTA
being with this
came data Dr.
U.C.SC.
collected 3.
to
propagation
Dr.
written
Ritsema
installed
VIdale
of
undertaken
assess
VBB
from
(3aarles
California
P.1. project
relative
et
Ammon
Caltech
surface
here
applied
to
testing
used
Thcrne CRUZ
Green’s
to
Our
this southern
tomography,
stations
project:
the
(B.S,S.A.
the
from
is
develop
Ammon, the Is
to
by
‘week
first wave Is
applicability in In
conducting
Terrascope
of to
zunitnarized
earthquakes.
involve
our
CMT concentrating
Lay
Ammon
Uecht,
futdon the
the
reglonalized
the 9.
which
data
“Research
California
for expected
dispersion
dispersion
submitted
choice ($30,000,
western
5.1991
sensitivity
Dr.
code
application presentation
described
long-term
will
and
Kelko S.ga
and and
and synthetic
to
here
below.
of
rate on We
Patton
events.
be U.S.. near-real
spatial
he
Cnzr,
on
Sincerely,:
4bccncLay Director, 12s00
Earth
UCsC between
inversion
models June
April
UCU
insbument
and
Reional
Knge,
deployed,
has at
developing
of are
In
development
will
CaIif
UCSC,
robustness at
progress
1991).
The
Science
He
distribution
simulations
testing good data,
1,
the
Dr.
the
that and
involve
time
PAS 1991-January
tngtimte
has
personnel
for
preprint
Earthquake
Thang,
Fail
theoretical
and
Dr.
sensors to
can
and 95064
In
inversion
the Dept. regionalized
Implimenied
and many
on
IATA
September
abet
meeting
collaboration
Jiajun
of the
a
be
has
these
of scnsitiviyt
of
of
request Dr.
SC
the
with
IAUADA
used
“Tomography
insminxnt
Tectonics 1991
rapid
the
surface
and
involved
been
8.
Inversions
Zhang,
Source
31,
preparation efforts.
of
boadband
broadband
the
for
Schwartz
recording
- Pasadena
Jeroen
5,
the
•
regional Is
models
1992).
a
P.
a
I.43TA stable
1991 of
major
wave CMT
In
with
new
long
and
are
2
the
We
and
for
is
is Group G: Physics of Earthquake Sources Group Leader: Leon Knopoff
Laboratory Studies of the Effects of Liquid Invasion on Frequency Gao and Aid (USC) G- 1 Dependence of Intrinsic Q and Scattering Q Dynamics of Fault Interactions Davis (UCLA) 0-3
Spontaneous Rupture Propagation across Fault Steps Harris (UC-Santa Barbara) 0-6
Simulation: Random Stress and Earthquakes Kagan (UCLA) 0-7
Pattern Recognition Methods for Estimation of Earthquake Probabilities Katz and Aid (USC) 0-8
Barner Model of Seismicity with Geometrical Inhomogeneity and Creep Rheology Knopoff (UCLA) 0-13 and Simulated seismicity on a Hybrid Barrier/Asperity Model
Evaluation of the Characteristic Earthquake Concept in Southern California Rice (Harvard) 0-15
The Earthquake Cycle in a Simple Model: Comparing the Behavior of the Shaw (UC-Santa Barbara) 0-17 Model with Seismological Observations
Application of Models of dynamical Systems to Seismicity, Earthquake Patterns, Scholz (Columbia) 0-18 and Predictions
Precursor Phenomenlogy and Modeling Levshina (UCLA) 0-21 from the other which where that attenuation attenuation boundary attenuation higher dependent;(Fig.1) we the 1 of suggest related Coulomb attenuation, attenuation. mechanisms researched mechanisms in liquid Frequency have liquid it where in the is at nonlinear recognize the the data
Coulomb Qr’ 5, 4, Thus 2, 3, predominant effected 1, least invasion a whole effect is than Q- 1 a mechanism We that We range We attenuation To strong The friction, higher is or analyses is Gao as frequency two may we The dependency a notice find range estimate suggest coefficient,a, this we 6MHz of is the which molecular part constant among preliminary of function and by that friction.(Fig.2) ,L.S. can parts the Q-factor; analyses correlation exist than Q- 1 that which 2-6 should is remaining liquid of that mechanisms is from there partial and caused of nonlinear there use = = on that function MHz. frequency liquid range;(Fig.1) likely the in linear Coulomb our of of cx€.. is K.Aki, one the levels. all digitized the suggest of is invasion. emphasize material, the frequency, expressed melting observable result Also are and contents due + by another with observed linear records of attenuation intrinsic whole USC first whole attenuation in the under the depending several we then to friction This the reminds in that the e peak(amplitude part. records fact the peaks may (Progress measurement, high raise but on is measured which as the use research source internal Q-factor 0 the internal sample attenuation that the is frequency find different We to is earth, attenuation and due of the the on frequency strain leads us often calibrate from one liquid attenuations that then the attenuation coefficient spectrum. (O.2MHz-7.5MHz). to scattering attenuation friction friction that Report) we’ve change of of when circumstance. to masked amplitude. process mechanisms experiments dependence dependent if may obtain the Al-Ga a the tendency. independent. is we the implemented constant at frequency is attenuation related amplitude affect to recognize of We nonlinear the nonlinear observed by 0, such different consists a in deduct waves linear some grain part) how may and The the for Q 1 - on all as to is If I )
U)
C U
U 0
is -0.5
0.5
3
2
—1
1•
0
1.5-
-0.5 0.5 0
2
—
1-
1000 0
xl
0-
ioo
F
2000
:
2X)
:
Corre.Lanon
3000 l/Q
3000
versus
Avetge FrequencyinlGlz
between
frequencyforsolidAl-Ga
Frequency
4000
4000
of
S.ui
l/Q(2MHz7MNz):7651e-006
Sp.
in
5000
ai 5000
KHZ
L
fr44A4 aun
6000
0-f
6000
-
•tij-
cI4
c
7000
ô4c-
4j’
7000
tI.
I
8000
8000
::—24—’H
rather
bed
modelling
defects
stress
with to ex.mhie
figure
and greater
axe
crack
problem.
analytic
the crack
law
1959)
crack faster
to to
intersecting, of
to
This
examine
For
The
the
We
a
each
be
orders
for
number
distribution.
reduction
shear
values
propagation
corrosion
surfaces.
(Davis inverted
2),
and
hive
than
connection
(factor
the
end
autiplane
:8E;
exanilnkg
work
than
node.
model
solutIon.
to
that
crack
field.
4)
of
1000
determination
calculated,
conditions used
antiplane
a.
of
strongly
for
and
magnitude
a
set.
A
of
large
of
A
have
nodes
to
crack
.ei.mlcity
power
propagation
By
the crack
a
the
boundary
more
program
Approximately
Knopoff,
find
up machine
connection
simple
out
1)
FR’M
taking
extent
non.’tnteraction
been
non-interacting
with
the
the
brings
interacting
from
function..
for
law
of
nodal
problem
than
ES—C
different
plane
Isolated
framework
physics
foreshock,.
of by
investigated
was
or, a
advantage
we
cancel
integral
Dynamics
1991),
pairwi.e
slip-weakening
fracture
the
through
50)
displacement,
iterating
machine
in
are
written
and
the
9
extreme numerical
than
The
cracks
models
out.
from
Paul
able
3) antiplane
nodes
method
3)
multiple
energy
displacenient
roughness;
cue. the
to
of
a
Future
objective
to
on
multiple
the
to
include
to
M.
heterogeneous
value.
of
the
handle
(figure
per
model
Infinite
of
cases,
solve
the
generate
Fault
Davis,
model
However
and
massively
has
displacement
of
which
the
1
crack
crack
scattering
step.
tBM
if
the
the
1)
1000
the
been
interactions.
a
1).
analytical
effect.
is
exten.lou
they
periodic
Interaction.
In
UCLA
fields
stress
yielded
overlappIng
give
(Knopof,
to
interacting
interaction
involve:
the
problems
3090.
Crack
time
some
interacting
tested
use
were
parallel
stress
of
co-ordinates
a
processes
caused
Intensity
constant
function,
array
fault
this with
a
to
sires
cracks
solutions
A
non-Interacting.
8,800x8,800
against
comparison
field
2D
1958),
micro-processor
oriualJy
processing
type
boundary
crack
crack.
by
crack
interactions.
of
were
antlplane
laplane
arising
(Chetterj.e
have
failure
interaction.
colinear
stress
in
a
of
2)
of
Into
ensemble
slip
chosen
number
2)
all
modelling
1000
displacement,
the
interaction
of
proposed.,
Green’,
from
criterion; (8192
integral
the
agreement.
condition
determined
cues,
cracks
bulk
cracks
bent
enhancement
random,
from
energetic.
of
at
1.
pr.-exl.tlng
(
processors)
ii
properties
Increasing
histogram
canonical,
ii.,
subjected
antipla.ne
functions
ussigned
as
equation
only
(Koiter,
a
namely
matrix
i.e.,
on
1978);
power
a
non
New
y
that
test
the
8%
to
of a 1000 IniIerctin Cr*ck
1200 I I I ‘.D
rnC ) - . . ——. - — - ‘--—I.. 1000 \‘ •_..— . . •.\ __.‘ -. —- .. - ..-.. ,— 4% ; -; 4 • . . . •. •• :.—r d-.7 •. . - - I..-’ %. ..— .. .,• -• -; . . —. - -.. a •/ “- •q • /• — ‘,%: —>-‘ 800 • a- • —
••-‘:- .. ‘“\a —.. -. . •i.. / . - - ;•• • • ,.‘ :• :.‘. \ 600 •_;—.-.: :. - • — .•\ _•%%. . . lb_I4: . ‘• :,,:
s• - •.•_ . i_A. .-. ..
- -.•. — 4. 400 .— .4.. I .-\ / : I 200 Nç : , :-. . -.- \ ..•— I .““V.—/ 1 ,-.- - V •4%
S S —
-J I I I —200 — —200 0 200 400 600 800 1000 1200 c ratio interacting/non interacting displacelRent3 1400
(t
1200 -
1000 -
800
600
400
200
Jfl4 -I 0- —4 —3 —2 —1 0 1 2 3 4 Nodsi Diphcemeut Ratio
year
SCEC continue
8hould able
modelling
The
case,
a
could
stress
stress stress
for
remains rupture stress
determining
the
Jacinto
Jacinto steps
earthquake
upper
material. dimensional if
maximum earthquake how
dimensional
simulate Summary:
Working Proposal
third
it
rupture
a
A
rupture
10/91
to
For Two
would
the
parallel My
second
fields not
field
field drop,
in
meetings,
earthquake
limit, separate
determine
also
fault
fault
after
the Year
rupture discussion
the
be
research
additional
spontaneous
of
-
This distance
on more
as
10192 on
rupture continue
Group:
be
Title:
supershear as
propagated first jumped
rupture, Northern
is
the
since the
rupture.
(non-collinear) calculations,
a
study
I
the
the
evidenced
rupture
the
that
study
result able
of
SCEC
dynamic was
maximum although
than
fault
ruptures
since which
starting first
most the
second,
involved it
process.
which
are
with
results
to
the able
which would
to
assuming
is compresslonal
(formerly
SCEC
Anatolian rupture
The
has fault
segment.
Physics
Spontaneous
determine
I rupture
that
of
also RESEARCH
propagate
down first
waves
am
SCEC
by
Interaction
to I
an implying
the
the
conditions
propagated
non-collinear
of
‘jumpable’
dynamic
will
be
jump I segment,
vertical
5-km the
at
supported I
time using
Primarily,
earthquake
have
this
faults
segments have
propagations the
hard present
of
velocity 1 contribute
great
atUCSB,
be
scientists parallel
fault study
to
the
This Earthquake
appears
just
first
the
on
ineligible
a also
to plane
will sthke-slip
the steps
of
rupture
Rt.mi
M8
finite-difference
first and PROGRESS
for
down distance,
predict the
in
vertical
a
Rupture
started
by
fault
parallel was
the are are
started
second
be
fault rupture
National
vertical
end
1939.
Erzincan
a
strain.
it do,
now fault second
about
field
to to
intention that done
separated
A. second
also
across
the
done
segment, on
for
a segment. the be
leading
If
Sources
will faults
HARRIS
at
to components
4.5
segment
strike-slip observations
fault
1)
and
first an
could The
the
changes
this
in
the
NSF-based
Propagation The
rupture
USGS,
Research look
strike-slip
fault
the
by
earthquake
include
earthquake
REPORT
fault
earthquake
the km
that
first
Interact
fault
of maximum
fault segment. Implication
by
to
object
rupture
using
jump
(Working
as
at
near
computer stepover
this segment.
The
a
ruptured.
less
steps. 2)
fault
Menlo
the
jumping
both
segment,
before,
lower
faults 1
Interaction
3-d
Council/USGS
dilatlonal
future. 2-d
to
results
and
the of
than case
of
SCEC
which
faults
POR
velocity
which
could
the a
segment
finite-difference Another
on this
The
study
displacement. ‘jumpable’
across
apparent
parallel
Park)
thereby static
Group width of
but With. 4
of
program
The
shear
the
I
across
YEAR
as
for
km.
study
show
salary
set
simulations
this
then jump
earthquakes
ruptured
then
steps
was
the
modelling
is
first
the
stress
results
this
show
0) changes
in
simulation,
Fault fault
stress affect very
result
used rupture
initial
was
4-5
the
a
to
case
support
1
unlike linearly
posLd0V.
distance
fault
appear study to
step
determine
significant segment,
tha
field, through
km
to steps,
the of numerically the
and
for
Steps of
numerical conditions
I
were
in
determine
which
segment. the
at velocity.
this I
to hope
the residual
the a dynamic
the
to
will
for normal
the future
where 30 elastic
which
for
be
and
shear
slow
first
fault two-
step
two. the
San San
bar and
re the
be
the
to
an
In I quake risk. geometry defect. tht Is prias. the papers that rock similarity being hypothesis. (a) allow I earthquake confidence artifacts. of hypocentez ‘characteristic’ popular We Published which [7) Invited May 10. believe During (a) (b) 9. 7. 4. 2. 8. 5. S. 3. Papers 1. revealed these propose. Muter have circum-Pacific the material p]aees 39, In Papers Paper. Inpress. Res., predicted ui potential choice Is and are Kagen, Kapa, Kapn, Kagan, Moishari, Kapa, Kagnu, Kagan, Kagun, Kagas, form phenomena. [5] rotation also 1091, to. due the talk that [1), of between : in that model, correct, submitted. occurrence. we define of distribution, our quantitative defects Based or [4], determined concentration, [5], of to period the which [2], recent the (UnIon test earthquake Y. Y, Y. Y. Y. Y. during Y. Y. general afcr by Y. earthquake prepared earlier Kagan, defect. [8], [7], .cal..lnvarlant of CL and and a Y., Y., Y., Y., Y., Y., belt Y., Y., two both result, Y., and an the earthquake which and results R•DM proper from M., a and earthquake and meeting). 1991. 1991. 1991. 1991. 1191. 1991. 1991. the [3] long (b) modes idea, with work validity 1991. we Although (earthquake theoretical [] parameter arid criteria hls the Y. [10) Apr’11 descrIbe should of Phyilci Ess—c:A results D. hypothesis will determine paper.: in quiet mitigation cia.. of Seiimidty: SeIsmic Fractal CorrelatIons tJkellhood On no Statlatics of Y., finally Y. the period: power-law (see P. of earthquake.. hypocenter. 3D use focal rnmulations large of 1, properties geometry Y. condensed for Talk Jackson, of period activity correspond papers (UCLA): these of the below) of 1991, the scale-InvarIant rotation argument, value. Kagan, models dimension demonstrate the moment faults) other mechanisms earthquake earthquakes quantitatively can till.: of isisnuc measuree. Canchy American analysis papers optimisation can ‘Turbulence’ to distribution contribute ‘characterIstIc’ of have are of be of 1991. and of researchers October of for matter This 1991. Q.omstr earthquake be of the both earthquake SCEC to seismicity, explained distribution, earthquake dedicated gap and of in are larger earthquake distribution rejected double-couple the of Cauchy Sank eource. ,Ini11r1ty ii that brittle orientation, Manuscript Geophysical rarthquake known hi properties by model. earthquake not deformation: controlled pattern the 21, to recent of of 1991 of computer than of earthquake, and connected either our with focal solids, 3D earthquake 1991 to trying dlstr1bution earthquake earthquakes fracture, gap fault Earthp.hs and rupture We Geapliya. focal certain might ucue.1 our to annual decade. understanding of stochastic a hypothesis I mechanisms, of of [10] earthquak. by unknown. system, prediction by model large catalogs, real find have Union at NonlInear own simulation focal earthquake.: process. fluid various the .riumc prepqatlon be the directly J. evaluates aspects earthquake are that prediction degree made ha,. I. else, Nønirnser .urnmary used tests, Fmadin. (manuscript, non.planarity thus Cauchy came niechamamp (AGU) turbuLence PEPI, eom,try Is., lvlueaced G.oplip,. the and statistical Therefore, tea hacard, Sci. remained Manuscript and to with w sources, the of of OeopIi.. w time of of 106, the hypothesi, gain ears submitted. are rupture distribution, its the Spring confidence. the following through T.dq, method., chow fractal the faults ScL the of ‘chazact.nstlc’ to report. optimisatlon, able and 123-134. a evaluation whereas J. of earthquake by In earthquake argue .patlal after, SCEC of Geoptq•. better biases, relatively that these 1, earthquakes. I. hid., Meeting preparation). accepted. initiation. to the a [9] rupture temporal elastic 1-16. led., of 3D work simulate and review. that J th. The brittle 106, immediate insight result. Increased this self-similarity the or (hophj.. submitted. random for of J. in proofs medium quiet. by there at faulting data propagation. phyeic., segment, J. distribution 135-148. the earthquake feature. the The 1r4., two fracture Baltimore, statistical the In Into give Giq1i,.. We suggest ieiirnic Is a appro earth of field.: Paper above initial stress goal., R..., show 106, great more with both as us and the of It of of of a
window
earthquakes. two
was trained method the
regression attributes,
that and results uncorrelated
functions attributes
complex on
parameters
combinations
time moving
clustering moving
Danger
distribution the Earthquake
equivalent magnitude were
window, Among this A S.Katz,
threshold, stability University
shows
Danger
calculation
time-window
number
to
layer
it work
formed
functions.
introduced
The
is
predict
the
of
to
K.Aki
failed.(figure
the
time time
with
of
capable
attributes
b)
time-eigenvalue
is
of
function
adaptive
t
generate
Function
parameters
there
c)
sequence
two
and
to
of
of
the he
relating
basis
of
the
defined
of
with
basis
complex
In
the
window,
window
prediction
the
the
of
Southern
Danger
values
of
earthquakes
the
b)
methods
spatial
Analysis.
number
experiments
is
Complex
basis
average
and
to
the
attributes
the
width
used
earthquakes
Adaptive
7
earthquakes
neural
attributes.
is
a
work
and
to
by
of
to
use
positive and
set
calculated
equal
of
3.).
of
earthquake
attributes ,
attributes.
function
the
d)
to
large
9
the
e)
California
15
sets
of
were
of
the
of
spatial
of
and
the
efficiently calculate
nonzero
network
Results
the
attributes
analysis time-space
Adaptive
earthquakes
results
to
time
we
for
eigenvalues
to
Neural
basis
Danger
of
earthquakes
with
differences
maximal
The
used
80
signal parametric
the
during
studied
several
basis
in
was
distribution for
constructed
and
month.
The
and
number distribution
this a
of
maximal
of
eigenvalues
(ANN)
to
show
the
Networks
the
moving
defined
are
Function:
several
with magnitude
and
the
time
respective
process maximal
complex
distribution
neural
magnitude were
Pattern
matrix
magnitude
region
with
defined
of
use
The
that
complex
and
in
representation small
-
of
was
a
eigenvalue
number
percentage
time the for
months
of
as
matrix
of a
nonzero
results prediction
in
network
(ANN)
the
calculated
of
magnitude
of
a)
number
attributes
the
specifically
linear
(depending
each
the
training of
following
as
one
ANN
Recognition
crosscorrelation
window,
set
Linear
Southern
ranges
of
of
time-dependent
the
earthquakes
linear
case
of
of
before
month
of
earthquakes
earthquakes
of
month
modified
combinations
eigenvalues.
the
large
look
crosscorrelation
linearly
analysis.
the of
of
seismic
of
data
of
multi-parameter in
of
calculated
calculated
and
larger
earthquakes
training
mutually
ones:
f)
its
use
from
16 California.
designed
a
promising.
large
time-window.
the
as
earthquakes.
fractal
moving
sets..
quadratic
values
basis
the
independent
of in
defined
depth
attributes
a)
then
It
in
the
in
functions).
g)
such
ANN
attributes
maximum
is
data
the
The
Linear
in of
in
a
width
is
and
based
a
time
a basic
a
or
The
a
given
set
In
as
for
time
way
in
It of
precursor.
e)
of
time-window:15
d)
b)
faults’ c)
conditions
measure means earthquakes by
methods months
attribute a) conclusions the
sequence on ANN the the
attributes are shows amplitude
line
latitudes
two
320
1190
The
The
The
large
The
the
the
Individual
hidden
moving
much
danger
In regions
-
the
may
interactions
the
use
same
that
signal
fact
>
probability
these
most
prior
number
of
in
is
of longitude
output
of
poorer
and
results
of
of
in
be
layer
functions
there
the
that
earthquake
linear
time-window
prediction
follow
the
large
of
indexes.
figures
effective
the
at
an
maltivariate
large
used
the
attributes
same
Southern
linear
the
output
of
of
was
area
ANN
to
is
than
of
interdependence
earthquakes,
Danger
of
may
nonlinear
earthquakes.
the
to
seismic
from
20
output
>
the
the
as
and
used.
In
much
occurrence
prediction
those generate
1150;
training
of
months.
signal
Neural
signal
prediction
in
both
dashed
play
earthquake
taken
California
results
of
danger
with
figure
regression.
function.
of
One
attributes
larger
shown
the
360
cases
interdependence
important
the
is
as
Net
separately the
of
cannot
artificial
line
large
can
output
of
a
Training
>latitude>
scheme
1,
function:
neural of
the
methodology.
rule
then
width
trained
between
the
our
shown
in
prediction
are
shows
see
a Therefore
ANN while
may
One
large
the
two
serve
is
of
experiments
the
shown
role
that
network
earthquake
failed
15
small
are
figure
the
was
to
in
can
serve the
layers
rapture
earthquake
was
in
months.
individual
in as
330;
before
figures
generate
unreliable
Neural
between
the
danger
for
performed
in
see
crosscorrelation,showing
earthquakes’
and
or
indicator
performed
2.
as
that
figures
ANN
absence
zero.
the
Figure
that
zone.
The
ANN
an
large
precursors.
with
Respective
1
Network
uses
function
second
a
and
attributes
the results
is
artificial
This
with
following
positive
predictors.
of
This
ANN
1
was
2
earthquake
defined
continuously
as
of
set
2.340
and
in
:
a
occurrence
means three
range
its
large
value
region
successful
means
a
of
of
are
and
and
2.
Figure
narrow maxima
input
earthquake
signal
and
prediction
seismic
>latitude>
by
(Figure
nods
marked
continues
of
of
linear
that
the
seismic
the
based
that
a
two
3
in
in
set
of 1:
z
by
latitudes:
curve)
Figure
the
5
41
6
0
same
for
1
360
Danger
the
indexes.
>1!jp
10
the
Function
part
‘\1’ (,)2O
>
of
330
1k
2
South
TIME
(dashed
Respective
California
\
(months)
line)
maxima
and
5.
region
the
4
on
output
with
both
the
of
curves
2
A
ANN
range
1
are
0
(continuous
of
80
marked
Hz
1I
II
4
90
(
I —
6
C
by latitudes:
curve)
Figure
the
same
for
2
340
Danger
the
indexes.
>lañtude
the
Function
part
>
of
320.
TIME
South
(dashed
Respective
California
(months)
line)
maxima
and
region
the
on
output
with
both
the
of
curves
ANN
range
are
(continuous
of marked
Z
—
prediction
the
320. Figure
part
0
3
of
Danger
based
South
10
Function
on
California
the
20
mukivariate
(dashed
30
region
TIME
line)
with
40
regression
(months)
the
and
50
range
the
(continuous
result
of
60
latitudes:
of
earthquake
70
curve)
340
80
>latitude>
for
j
the 90
1UiZ1”1
greater
with only
earthquakes model
of scale—size
scale have models
this regions. conditions.
the of
feature
d)
c)
b) a)
radiation distributed.
dynamic
br.ak.ing largest
properties introduce
frictional cracks
performed largest version
realistic
that,
conditions
including
PROGRESS
16:21
the
plane
no
large
seisiidcity
that that
means
simulate
a We
We
without
one
large
view
which
models
p.riodicitIes occur
Earlier
of exclusivity
seismicity
barriers,
Contenporary
size
fracture
because
with
seisinogenic
stress
there
be
have
barrier,
value
fault
events
of
of
have
than piece.
motions
of
small
seismicity
in damping
strengths
REPOPT
end
confined”, prior
213
earthquakes
that is
in
event,
breaking
of
inhomogeneities
the
with
variable
when
we
at
Carison
at
We
i
faulting
importance constructed
analysis
itself
as
gauged
reduction
the
structure
We
homogeneity
instantly
conditions
the
begun
interactions
the
is
those
i.e.
do
the
B—K
only
lengths
find,
have
wel].
scale—seismicity
quasistatic
to
206
approximately
have
in
OF
B—K given
nonstationary
even
of
are
the is not
fault
studies
zone,
edges
the
models
cannot
thickness rupturing
3051
strength
LEON
either
one
nature,
places
friction as
loci
of
and
the
just
model.
as
an
not
of
these is also
at
have
i.e.
strengths
beginning
of only
and
as
upon
recent
greater whichever
equal
in
in seisuricity
transverse
geonretties
principal
KNOPOFE
on
of
the
the
attack
quasistatic
of
largely
Langer,
in
of
are
great
beyond
an
done.
assiuned
zero
by
whert
be
the
a
somewhere
that
the
among
a
of high
regions
if breaking
the
all
at variants
edges.
version
The
in
in
frictional
M(6.5.
suitable
of
additional
to
results
the B—K
we
direction
dynamic-al
and
motion
invoked
lattice
lattice.
elongated
smaU
strength,
a
occurs
past
FOR
of
there
the
We
problem
on are
strengths
that
than the
breaking
the UCLA
were
length governed
is large
dyne.mical
make
events
scale
may
along
model
that
this
are
We
other
find WORK
seismogenic edge
history
can
the of
the
of
of
tEUPthS1CS
Hence
of
great. have
planar
earthquakes
a-
into
15 be
for
have
sites be
in
to
to
earthquake.
a
project
use
here
of
size, some
shown
contacts Bak,
the l—D
Under
the
the
the PERFORMED
the
km
of
smaller. be both
one strengtl
problem can predictth. ingredient parameters.
transverse faults, frictioal of
where
are
no
episodic,
shown
the
the
by
mechatism
the the
introduced
a
stress
and of
San
casq,
edge of
mad*
evidence add
sur 1 -ive
and Burridge—Knopoff
is
seimicity,
(2—D)
et
friction
Ve
an
modeling
crack. lo inoçel
one—and
events
this
sich
sandard dack
the t siowed
4ence
4ndreas
zone.
ii.,
that tress—free
I
with
fracture
three_dimension,J,ity
into
hence
of
will
believe of
have
UNDER
and
intensification
be
end
But
versions
consistent distribution
large
contact,
the in
or
of
and
of
and
real forever
time fractures.
to
we
The
appropriate We
non—overlapping
the two—dimensional
to
f].uctuations
for
that
drops
For
of
order
fractals,
condition,
B—K be
large
energy
retained
fault.
the
the
SCEC
have
support
localization
cannot in faulting,
the
the
earthquakes
earthquakes
we
of
that
vhith
California
events
length
transverse
transverse
the
if
model
models
fractally
thickness
unbroken
and
dynamics
boundary
AUSPICES
the
the
that
system, assumed
we
of
had
to
mutual
events
lass.
model,
model local
with
used
Work
next
this
that
B-K
B—K:
all
end
can
its the
the
the the
of
at
is in in
in to a
1O/21’91
Exploration
observed
observed
quasistatic.
writing,
are
with
considering
understanding
intermediate—term
of dilatant
investigation pumping/valving
corrosion
required
has
remain an
fluids last,
barriers noted,
increases 7 represent arise
fractures
inhoinogeneities georetry-
to imagine, asperities
assumption A
realistic
either
into size
16:22
concerted
otherwise
the
no
left
‘We
different
7e
presumably
the has
results
fluid
unsutuz-ed
becomes
because
have
influence
that in
l—D
any
foreshocks
have
the and
to
in
due
for
are
understanding
other.
213
cracks
strong
on
shown
that
of
models
be
frictional
our
or
first
significant
regim.
investigated
veak.engd
seismicity
fault
friction.
of
the
tiltiriately
to
the
these
reservoirs expressive
to
paramount.
large
begun
2—0
model
sizes.
of
of
breaidng
206
strong,
the for
view
the
in
describe
nature
W’e
attempt
that
prograitming
of
clustering
of
obstacles
zone
are
and
ensues.
3051
geometrical
on
features
lifetime a
the
fluid
breaking
scale
a
are
percolation
of
fracture
time
crustal
study ve.must
not
these
aftershocks
strength
precursory
in
of
earthquake
influences
Here
Simpson
unbroken
triggered.
of
strength
currently
diffusion.
the
at
as weakens
in
ilnportan.ce
Fractures
a
geometrical
asymptotic
clustering.
the
weakened
of
that
introducing
yet
is
of
influence
to
we Thus
of
surface.
the strengths.
features
consider
phase
heterogeneity
fluids.
the
continuing.
fluids
for
nature
of
on
dynamical
frictional
iCLA
is
and
—2—
seismicity have
is
making
the
to
on
on
lover
relative
clustering.
fluids.
arising
In
lowered
the
this
During
state,
earthquakes
the
geometry
to
bturi1L,
long-term
of
asperities
short
these
offsets
of
of
and
our
the
In
relied
interaction
In
of a
elasticitr
problem;
As
are
an
slip
crust.
the
simulation
niodels
In
their this our
possibility
variable
models healing
which
this from contact progressively, rupture.
effort
influences
the
in
compred
fluid
of
simulation.
in
are
the
we
best
e
mc?dej.
significantly
l-LD
e.rthquake
view,
ciffusion and
recursory
on
ime
relationsbip
I
kening
Thus
we
olutionary we
one
faults,
ge
asperities
simply
on
to
B—K
but
of
or
in
thg
remein,
suppose
fusion
transverse
interval, In
case,
represented
the
construct
the
with
have
on
depends
of
the of
these
models.
a
pore
largest
far,
we
models
the
influence
barriers
amending small
We
from vhich
and
fault
interior
faults.
mapped
among
seismicity
we
have
patterns. begun
producing
other,
the
to
we
also
at are
pressure
features
on
further imagine
between
in
on stress
cracks
scale,
be a
scale.
There
scale
have
zone
more
this must
time
not
are
due
and
the
the the
the
one
the
an of w As
of
an by , V V
response.
geometric found
of
interval,
earthquake
differences
show
strike laws,
remarkably
discrete interesting distances
means
direction
simulations
strength). the
a
state-dependent
analysis that
long
properties
can
we
slip
models
earthquakes.
a
the
earthquake response
way
related
the
cool/shallow
seek
ends
be
each
and
vertical
the
which
has
array
to
specifically
large
Only
tripped
Annual
The
models
Physics
rather
can
Studies
be
is
stress
to
possibility
of
or USGS
for
point
Evaluation
yet
obstacles
and
At
contrast
or
recurrence
done
in
understand
of
general
Department
resistant
with
fault
sources
have are
work
lead
and
limited
state
strike-slip
been
geometry
each
constitutive
However,
on
segments
than
into
distributions
on
Summary
perturbed,
(such
are
friction
of
program,
by
rather
crust
sufficiently to
segments
examining
no
accomplished
transitions
the
constrained
done
depth,
earthquake
a
to
in
to
boundary
to spatio-temporally
underway
that
and
to
study
well
characteristic
as
the
further
fault
of
time
quasi-independent
slip and
whether
slow
breakup
along
fault
of
into
applies
and
cellular
division
on
there
the
conclusions
constitutive
defined
Report,
properties, using
Earth
transmission of positive
sometimes can
(if
moves,
crlLsral
Harvard
and
computations,
limitations
conclusions
coordinated
between
characteristic
refined
the
which
study,
integral
strike.
the
is
on
by
sources.
lead
uniform
into
on
models
which
much and
so
automata
response
of
identifying
data
local continuum
Southern
earthquake
the in
deformation.
in
far
to
a
can
there
spatially
Planetary
grids
University,
that
long-term
long hot/deeper
elastically
properties
complex
equation
which
only
characteristic
on
current
fault
on
use
properties
to
at
that
be
fault
Strong
on
earthquake
characteristic
granitô
this
would
would
segment
or
in to
the
planar
James
lab
and
studied),
slightly,
October
California
surface,
the
earthquake
properly
other
possible
the
approach
spring-block
limit
complex
models
concept.
topic
mode
interest
Sciences
based
slip
methods
which
geometrical
topic
average,
deformable
presence in
regions Cambridge,
1
be
seem
under
fault
applied
R.
investigators
upon
the
ends
response.
has
consistent
earthquake
1991
from
by
and
suggest
response,
Rice
rate-
incorporate
which
are physical
constrains
in
Earthquake
models
the
to
slip
zone
behavior
modest
hydrothermal
been
These
and
reduction
based
as
A—B
the
(V=slip
at
very
concept
be
everywhere),
uniformity.
of
arrays
and
continuum
patterns
a
an
Division
variations
into
show
a
crustal concept
significant
that
basis on
MA
We
[=Vdt(V)/dV)
strong
involve
tentative.
imposed
are
origins
with, state-dependent
on
with
variations
behavior,
have
computational
of
segments
the
with
speed;
uniform
3D
the
02138
are
spatio-temporally
either
in
of
for
Center,
the
along
blocks.
an
say,
range
that
case
elastic
of Southern
cell
drawn
conditions,
Chinnery
studying
limit
characteristic
simplified
of
projects
Results
such
intermediate
segments.
may
“plate”
Applied
property
t=steady
uniform
Nevertheless, characteristic
a
essentially
or
of
at
strike.
fault
for
USC
two along
of
(which,
as
These least
interactions
grid
frictional
based
lead
characteristic
the
jogs
thus
varies
friction
velocity. such
models on
analysis
or
solution.
California
P.O.#569928
Sciences
classical
size).
variations
strike
This
for
to
being
necessity
in
more
the
If
on
are
state
unfortunately,
or
far,
phase
the
earthquake recurrence
possibilities
uniform
complex
this
a
with
physics
inherently
constimtive
is stepovers
driven
sequence
earthquake
laws.
are
with
the
negative
from
along-strike
of
variation
between
frictional
in
feature
The
Rate-
friction
locking
depth,
slip
results
of
along
slip
fault
such
Here
slip
of
on
and
of
at
is
in
in
in
in a
fault
per
thickness,
straining ±
found periodicity
event
if
the hypothesis recurrence
cycle
earthquake, creeping
motion
earthquakes,
zones events.
plane
crust, extensively
Parkfield
by effects
consequence prior
(and segment
the Prieta
by
crust
(Plausible greater
uniform
somewhat
underlying
mode.
support the
loading
deformation,
Parkfleld
11
Parkfleld
km
occurrence
our
fault
support
segments yr
perhaps
earthquakes,
they
would
is
extending
and
in
and
earthquake).
Future
of
of
ongoing
Crystal
for
at
3D
The
Palmdale
Free-slip
coupled
of
from
fitting
predicted
fault
along
zone
isto
patch
thickness
35
prior
and
a
everywhere
are
ranges
interval
of
lower
in
constrained
finite
a
earthquakes
shear
slip
stiffer
studied
of
focus
be
mm/yr
fault
ruptures.
over
of
characteristic
time.
1906) in
some
an
lower.
extensions segment
neither
is
with
time
a
strike
other
deformation.
predicted
great
must
rates
the
USGS
to the
divided
to
postdoc,
NSF/EAR
element
crust
boundary flow
for
network,
segment
the
and
in
by
and
the
would
the
These
assumed extent model
their
interseismic
so
of
earthquake(s),
and
In
geodetic the
earthquakes.
inevitably over
relaxation
in
If
independently,
models
top
top
in
more in
the
far
mantle
relaxation
freely
the
that
by
supported
the
Parkfleld
are
periodic
into
first
neighbors.
the
the
are
calculations
for
Y.
tend 12.5
a
consists
our
of
relaxation
next
affected
is
simple
which,
periodic
show
creeping
responses
in
conditions
length
grant.
viscous
creeping Ben-Zion,
mantle,
that
for
the
locked
envisioned
about
The
slipping year
of
previous
data
flow
work
km
to
event.
have
properties
period,
the
the
mantle.
the
non-uniformity
lengthen
1906-
time
earthquakes analysis
modeling
aim
at
is
in
of scale
and
Likewise,
along
Crustal
by
through
1992
type
relaxing
earthquakes by
portions,
viscoelastic on
least
zone
loading
consistent
some
times
is
zones
a
are
parts;
accumulated
region
is
its
For
7.5
are
the
Li
modeling
17.5
of
the
and
who
near
long
to
to
±
examined).
the
That
neighbors;
employed
sometimes,
are
and
significant
imposed
in
yr.
of
example,
with
time creeping
deformation
9
analyze
are
apply
Parkfield
along
a
here,
locations
1857-type
km
is
San
layer rate,
yr a
between
the
postseismic
compared
the viscoelastic
on
higher
rather
plane
are
Rice
locked
consistent
studying
with
if
dependence,
time
elastic
2
at
which
lower
upper
NW
in
Andreas
a
of
in
the
strike
and
a
(only
physical
Parkfleld
slip
interactions
on
time.
response
to
(JGR,’87)
zone contrast
SAF
Pacific/N.
then is
using for
than
since
segment.
influence
lower
responds
are
fault
end
them. hence
model
earthquakes
locked
the
crust
mantle.
upper
deficits
slip
to
is
Parkfield,
corresponding
similar
the
with the
geodetic
constrained
to
deformation
the
lower
Shortly
assumed
of
system,
the
fault
the
the
zone
is
crust
model
the
ratio
may
seismic
dictated to
may An
the
to
such
1988±7 far
crust,
Parkfleld
against imposed
statistics
and
the
1857
America
in
near
that
on
The
of
plane
NW.
this
The
issues
crust,
to
imposed
field
1857
of
relaxation
be
a
not
after
range,
other
interactions. data 1
the
these
the
Fares
of
examine
generate
to
characteristic-like
discussed
a
relaxation
event.
time
SCEC
Middle
crust
somewhat
behavior,
by
7.5 crustal
by
provide below
slip,
Thus
velocity,
be
range
of
time
segment
following
yr under
kinematically
to
an
of
segment.
including
possibly
are relative
timing
seismicity.
and driven
which
albeit
dependent
km
estimate
constant
the
has
past
adjacent
except
funds
Thus,
dependent
the
time
the
time
Mountain,
being
slip
those
stressing
complementary
viscoelastic
Rice
1857
the
a
on
of
time
events
poorly
while
loading
further
should
regions
and
single
properties
The
by
motion
rates
dependence
characteristic
the
is
the
are
a
best in
an
The
that
locked
based
about
(EOS,’88).
far-field
particular
great
a
and
to
postseismic
15
loading,
great
model
underlying
An
previous
steady
1989
used
later
at
we along
earthquake
vertical
layer and
estimate
constrained,
from
surface elastic-brittle
yr,
be
constrained
the
rate
of
1906 the
and
important
on
1
find
lower
affected
rupture
and
in to
the Loma
are
to
next
most
times
shear
of
the
the
their
the
share
2 fault
in
that
page,
1857
fault
the
1995
yr
of of 16’2i’91 14:54 885 893 2431 th5T TH PHYS S8
1991 SCEC Progres fl.eport uf Bruce Shaw
My icarcli tlii asl year has cc11Lred oii Iiy;lig to 111ZlC’(()Jll(CtIOI1 betwecn very simple dyin iiiicaJ models of an earthquake fault, ann scimiiu logical obscrvalious. ‘Ihe foundation of the modeling i a one dujwiiioiiml homogeneous determinist ic partial differrimlial xuation with thc !rmuhiIuaIiii— grcdwiits of an elastw c’uii1ing a loading cchanisiim,and frictional iimsta— hility. The sini )tI, clc rtiatic,n of the equation gives thi li.&sic model in seismology of Biiriidgc and lnopoff. ‘l’Ilisis timemodel we stud numerically. havc been devlupiti wayb of ita.siiring the model in ordcr to coinpali’ it with tismnological obscrvaIioiis. ‘1’he mrmeaiirernentsinclude cxaii mnlimigthi smr)allevent activity preceding large events, studymng Ilic, averagi’ UiOTflCllt spectra for eventa of different sizes, and plotting relations betwceii vai’ioiis source J)arameters. 1 have bet’mi tudynig vamious pamaineler meginmesof thc’ friction function, and exairtimmimigthe behaviors of timemammlelhi lle clifleremit regi ,Tw.
One of the tlibgs missing fromimthe siuipl1 model is aftershocks— thcrc is oniy ruptiii’i tiiiicsualc and a loading tirnescale. ‘l’oadd a new timimescale which produces foresimockamidaftershock clustermug, I have (h(vc’lopeda gemi eralizatiomi of the model which couples the diplaceiiieiit. field to a HCW diffus ing field. I obtain On,ori’s law analytically and numerically. I ]ivc beguti to study the spatial and temporal didrihititiomiof afteisliocks ni real cnmthqimkc catalogu to tr’ lu lct this model. Southern California Earthquake Center Research Summary, 1991 Application of Models of Dynamical Systems to Seismicity, Earthquake Patterns and Predictions
P.1., C.H. Scholz, Lamont-Doherty Geological Observatory, Columbia Univ. 1. Study of a quasi-static block slider model.
We have been studying a 2-D block slider model similar to that of Brown et al (1991), and have reached two main conclusions. The first is that the behavior of this class of models is strongly dependent on model parameters, namely the ratio of Cr, the critical value, vs. the spring stiffnesses. Results for the case with the spring stiffnesses K 1if’cullin =50 are shown in Fig. la. At some values of Cr the model behaves in a way anaf ou to’steacy-state creep (Cr=610); there are no whole array events and the cumulative event size distribution is exponential. At other values (e.g. Cr=590), the behavior is dominated by events that break the entire grid. It is only at the transitions between these two (e.g. Cr=600) that we find power law distributions. The second conclusion is more serious. For earthquakes, stress drop is scale independent so mean slip scales with rupture dimension L. This means that moment, 0M should scale with L in 1-D models and with 3L in 2-D models. However, in our 2-D model, 0M scales linearly with area (Fig. ib), meaning that slip is scale independent (it is controlled by the coupling spring constant). A crossover phenomenon begins at about Area=200 and settles down at about Area=400 to another linear region with a higher slip. This crossover corresponds to a change in dominance of scaling from the coupling to the pulling springs. We believe that this scaling problem is common to all spring slider models, both dynamic and quasi-static, that are now being studied. It is caused by a lack of long-range interactions. Ref. : Brown, S.R., Scholz, C.H., and Rundle, J.B., A simplified spring-block model of earthquakes, Geophys. Res. Lett. 18, 215-218, 1991.
2. Study of Earthquakes as SOC phenomena.
The second part of our work concerns the study of those observations of earthquakes and faulting that are relevent to the argument that this system is one that exhibits self-organized critical (SOC) behavior. Two studies have been completed and reported in papers now in press. In the first (Scholz, 1991) the properties of earthquakes and faults were reviewed with this question in mind. Both were shown to obey self-similar scaling laws, have fractal size distributions, and internal fractal irregularity. However, in contrast to models, the natural system contains a characteristic dimension, the seismogenic thickness, that profoundly effects the behavior. In the case of earthquakes, small earthquakes, those below this dimension, obey different scaling laws than large earthquakes, larger than this size. They consequently belong to different fractal sets and have different size distributions. In the second paper (Pacheco et al, 1991), this difference in the size distribution between small and large earthquakes was specifically tested using a new global earthquake catalogue which had been corrected for problems with the magnitude scale and tested for completeness. Both show = j-B, a cumulative size distribution that is a power law, 0N(M where B=2/3 for small earthquakes and B=1 for large earthquakes. The crossover) is at M=7.5 for subduction zone (L—6Okm) for transform faults (L—lOlcm)(Fig.2).5 The recognition of this events and M=6 distinction has profound implication for earthquake hazard analysis. SCEC supported publications.
Scholz, C.H., Earthquakes and faulting: self-organized critical phenomena with a characteristic dimension., In Proc. NATOA.SJ. , Spontaneous Formation of Space-time Structures and Criticality, (T. Riste and D. Sherrington, eds.), Kluwer, Dortrect, in press, 1991. Pacheco, 3., Scholz, C.H. and Sykes, L.R., Changes in frequency-size relationships form small to large earthquakes, Nature, in press, 1991. I.,
-. I-.
1. Moment Mean Slip (Z slip/area) or Cumulative Number -a 0
1.11.1.1 I I . I I .111.1 I I II 11111 I I
cm L.I. OflO -1 -i -i II II II S o\o’1J LA —co 0 ci 00 Os S -J —I CD
C’, ci, 0 tFQ N 1 S Cfl LA. 1
0 S 0 CD
J.
CID
LA.
sI— &_ao
1 Id•Ce. 0 0 -1 Ge transform faults Mw 19OO-199°
I 5 6 7 8 7.0 7.5 8.0 8.5 9.0 9.5 magnitude (Mw)
• ii 10/21/91
under This various
following spatial retrospective are vtb algorithui the stability reoccurrence seismic this has strong dissipative Grassberger
nonlinear correlates degrees process, might
order
these in. (correlation correlation 1965. considering
predic i’inlyzing attractors” 1988,690—694.
a
algorithm
being
bean I
1.Software,
2.1
software
3.1
the
work
l.v.I.:Kelis—Borok,L.Knopoff,I.M.Rotwain
2.P.Grassberger
higher
supervision
to
systems be.
have earthquakes
Some 09:28
tion.
tasks
distribution
regions.
of
obtain
results continue
started
even
has
dissipative investigated. of
has
the
started
tested. with
freedom,
Va
preliminary the been
Pbysica
and dimensional arises
of
exponent)
nonlinear
of
these been
as alarm
alloys
of behavior
will
though
been.
assume 213
earthquakes
the
a
designed
nuiu.rous sorting,
Procaccia vhich
occurrence working
earthquake
the of
I
done:
to
strong
of
in
1991
under
results
due
correlation tend.
am 90 The developed
it
of
prof.V.I.Keilis—Borok,
for 206
work
the new it dynamical
and
that
1990
of will
(1983),189—208.
going my
of
those
results ANNUAL
to
space,
dynamical
algorithm 3051
The the is merging in
algorithm
works,
Pacific
shock
for
T
work
the
I.Procaccie. correlation
Ve on .
neiartheless even
deterministic
[2J, with
with
LEVSHINA
in
be
unclear
the
prediction
processing
to times
idea.
intermediate—sized
SCEC
local
are the
and
SUMMARY
obtained
under collaboration
thus
system exponent
have
in
International
where though test
magnitude
respect
originally
arid
seismic
problem
trying
CN. is a
has
some
since
systems.
structure
of what
far
been
selection
focal
supervision
the
with
integral
to
REPORT UCLA
the
been
the CN
of
from possess
strong using
of
results
“Keastiring to
belt.
July,8
chaos kind obtained. algorithm
use to
of
N UPH1SiLs fractal catalogues
[1J. a
zone has this
seismic
the
proposed
apply
a successfu]fly
and
large with
Institute
recognitin
of
a
a
of They
of
4
study we
been
or
a
earthqukes” earthquakes
attractr
variat4on single-Jvariable
modification
of
1991.
have on strange
We
data.
of
C-R.Allen.
nonlinear
lov-.dkinensional
hope
I.Vorobieva,
due
or this; behviour
andcompa.re
a
proess
the1
th ha1ve
inta1led
ha1ve
prof.L.K.nopoff.
of
fr
of
even trong
been
to
of
to
During
San
approach
In aftershock
attractors
strangeness earthquakes
of
begun
we
Earthquake
of random
applied
the
determine
an
addition proposed
Andreas
obtained. can
that system
embed the
of earthquake.
initial
of
infinite
“Intermediate-term
,Nature,
on
this investigation
previous
the
a
time
be
earthquake
have
possibility the
noise. to
the
attractor.
study
the
to
sequences
modeled
this
tovard
fault,
it
period
recognition
whether
the
Prediction series. to
methods
a
parameters
of
UCLA
This
triggered
some
number
attractor
executes
Nov
perform
vol.335,
results
measure
of
process
seismic
strange
We
Since
which
VAX.
since
work
as
fIov
the
the
new
the
are
the
of
of
of of
of
In
By
a Group H: Engineering Applications
Group Leader: Geoff Martin
Research Program to be Initiated Early in 1992 A RESEARCH. PROJECT BETWEEN THE
SOUTHERN CALIFORNIA EARTHQUAKE CENTER
AND
• CALIFORNIA DEPARTMENT OF TRANSPORTATION • CITY OF LOS ANGELES DEPARTMENT OF PUBLIC WORKS • LOS ANGELES COUNTY DEPARTMENT OF PUBLIC WORKS
FOR A JOINT PROGRAM ON
THE CHARACTERISTICS OF EARTHQUAKE GROUND MOTIONS FOR SEISMIC DESIGN
I. BACKGROUND
The California Department of Transportation and the City and County of Los Angeles, all presently have major design or seismic retrofit programs related to the development and improvement of their bridge infrastructures. Historically, bridges have proven to be very vulnerable to earthquakes, sustaining damage to both superstructures and foundations, and in some cases have been totally destroyed. The 1971 San Fernando earthquake damaged more than 60 bridges on the Golden State Freeway in California. The cost to the State for repair, including indirect costs due to bridge closures, was approximately $100 million. The 1989 Loma Prieta earthquake damaged more than 80 bridges and caused the deaths of more than 40 people. The costs of repair (not including indirect costs) have been estimated at between $1.8 and $2.0 billion.
The specification of earthquake ground motion characteristics play a critical role in establishing seismic design codes for both bridges and other structures. Both the California Department of Transportation (Caltrans) and the AASHTO Guide specifications for seismic design use smoothed acceleration response spectra keyed to peak ground acceleration maps to characterize ground motions. The shape of spectra also reflects site soil classifications, With increasing costs of bridge construction and the need for seismic retrofit of a large number of older bridges in California, it is clear that research focused on increasing confidence levels and optimizing earthquake ground motion characteristics for design is of paramount importance. With the above objective in mind, design engineers with Caltrans and the city and County of Los Angeles (to be designated )3C have jointly agreed to sponsor a three year research program through the Southern California Earthquake Center (SCEC). A tentative list of research tasks was initially proposed by Caltrans, and was the subject of joint discussions between representatives of 3C and SCEC at a meeting in Los Angeles on March 30, 1991. Further discussions to refine and prioritize research tasks were held during a one day workshop at the University of Southern California on May 22, 1991.
II. SOUTHERN CALIFORNIA EARTHQUAKE CENTER
In February 1991. the National Science Foundation and the U.S. Geological Survey designated the University of Southern California as the focal point for a new
1
Co California
Task
from
procedures. been Table
structure the
Martin).
framework
to years.
appended. seismic
III. and the
regional identified combine USGS.
managed be through Division organizations
in Sciences,
Henyey, Columbia Southern
major Santa
in
Aki,
member
reflect
Southern Pasadena,
PIs:
expected
engineering
history
how
holder
identified
1.
working
participating
Smoothed
RESEARCH
From Participating
The
SCEC Barbara,
H-i:
The
waves
Funds
of
additional
pattern
of
is
the
An
existing
Professor California
the
shown
University
on
as
by
program
Research
Mines
principal
has
Characteristics
the
for
of
the
the
of
California.
Dr.
Southern
eighth
surface important
coordinates
Existing
and
May from
need
generated
the
the
for
group fault
are
received
UC
Executive
estimating
results National
V.
design
of
TASKS
elastic
in
and
for
year
knowledge
W.M.
22,
Santa
research.
Lee for
Assistant group
future is
Earthquake
being
co-investigators
strain
tasks
Table
institutions.
and
of
movements
(Lamont-Doherty
task
headings,
expected
establishing
smoothed
Geology
1991,
coordinating
of
to
the
and
profession,
I. Scientists
more
Keck
response
by
his
Academy
Director.
California
Cruz,
former
sponsored
(Group
the
developing
used
2.
buildup,
seismic
of
earth,
earthquakes
Further
Dr.
earthquakes.
of
agencys
a
than
It
Director,
C 3 /SCEC Foundation
are
of SCEC
to
The
each
UC
Center.
M.
research
spectral
will for
Earthquake
and
H
last
given
from
earthquakes
chairman
also
a
of
spectra
$5
hazards
Trifunac,
San
These
Thomas
C 3 and
the
discussion products
--
by
is
with
from
major
scientific
represent
statistics
2
involvement
a at
Sciences,
million
workshop
Geological
participating
Engineering
the
the
research
several
Diego
its
master
Engineering
physical
includes
least
The
program
shapes
in
a Chair
groups
SCEC
geological
earthquake
NSF
are
in Heaton,
group
development
University
of
Southern
center
of
for
five
(Scripps
of
Response
Southern
of
is
a research
with
model
and
other
institutions
the in
the
scientific
each
and
the
held
program
properties
scientific
Observatory)
civil
years the
past
function
comprising
in
leader
Geological
Applications)
chief
in
USGS
includes
cornerstone
USC
diversity, first
the
new
Applications
site
universities
for
of
Science
at
the
engineering
Institution
data
of
California.
earthquake
and
California.
the
the at
scientist
are
or
center.
research
year
Southern
Spectra
Department
Southern
of
correction
will
consensus SCEC
knowledge
are
within
of
potentially
and
coordinator the
tasks
USC,
structured
University
Sciences
a
9
Director,
shakes
from
and
earthquake
be
research
shown
‘master
research
was
eight
information
of
of
Caltech,
and
is
of
with
(Dr.
faculty California
California coordinated
in
the
occurrences,
the
This the
Research
given
the
factors
as
Oceanography),
recently
seismic
in
at
the
to
of
in
as
and
participating
the
USGS.
under
to
as
organization
NSF
Geoffrey
of
response
effort.
USGS
model,”
USC
model
long
topics
Table
be
Geological
UCLA,
below.
what
California
shown
Southern members
Thomas
needs
sources,
and
used
include
gained
center
and
design
added
seven as
areas
office
Keiiti
will
and A-i
can
the
the
has
UC
11
or
to
in
R.
of
in
a H2 for and Co code
in Task hazard such functions Southern The to attenuation records sites). research Engineering shapes design magnitude, (EQUINFOS). which develop PIs: either the program • • • • • • as use. The The
The H-2: Writing Carrying equations Running Finalizing earthquake Engineering by Digitizing scaling Generating subjected evaluations. the adjustment strong (these guidelines being The studies recurrence cover will California deterministic program, research the results evaluation recent The assessment epicentral
evaluations Southern empirical Technology Dr. relations will Department attenuation be keyed motion relations 8 progress the database of all 1 K. out frequency to and for the Pasadena, of encompass: strong or local (including data. Sieh required Department or 5% relationship the period strong a All the to data will detailed spectral distance, processing codes, design available of systematic scaling
for California smoothed damped site and for will estimates corrections processed reports project will relations take the motion has Strong-Motion dependent motion 1933 spectral use Malibu, Dr. response regression have also the use. also amplitudes been maximum and analysis the under coefficients (fault E. and will data the in elastic include Whittier accelerograph to response of will data been Hauksson, form provide evaluation site acceleration recorded
for Fault regression Upland, amplitudes. a several provide the
peak 3 and base magnitude using the final subsequently parameters in instrument analyses of of determined earthquake response the present, using EQUINFOS cover earthquake direction
regression and a the ground Earthquake spectra on report. reflecting wave and California (coordinated hundred unique the development of site records the analyses. frequency stations
periods Earthquake vs. to Sierra means response and (rock, propagation acceleration. frequency can spectra will of complete be determine from characteristics magnitude number resource analyses records aftershocks, Institute Dr. used are the will Madre be between be for Data stiff with operated a Trifunac). of optimized on available spectra Hence, effects dependent re-assessing limited for for be of frequency soil of earthquake for response Task
which Parameters Information earthquakes file frequency analysis, a of events Task undertaken. spectral newly 0.04 In fault sites research and by 11-5) in for spectral for of data this to have 11-7). for to the the and per is records and C 3 sites earthquake over attenuation and the 15 amplitudes three dependent base, design processed capable dependent data yet USC year) USC engineers on deep obtained baseline seconds. System manner scaling seismic where These topics 1,000 to from with year Civil base Civil and soil are be
of 14-3
of
the characterizing
Hence,
to one
Co Task
research
Loma recordings. conditions fundamental
regularly
seismometers
three
for will
understand
out nature design
understand influence Co Task commentary.
expressed
opinion
Ongoing
critical
added
shaking
parameters
structures.
times
PIs:
this
PIs:
peaks
be
source
dimensional
Prieta
The
The
Site The
it
H-4:
H-3:
selected
codes
research
used
parameters
of
task.
at
among
research
studies
is
and
in
the
of
occurring
are
a
objective objective
duration
empirical
essential
Characteristics
and
site
the
and
earthquake,
Duration
Source-site Effects
the
to
given
California
Dr.
Dr.
response.
reflect
wave
Accelerations
The
parameters,
California
the
The
intensity
at
or
sponsor
those
effects
effects.
often
M.
K.
in strong
form
tasks
geological
can
the
analyses
designed
is
extent
failure
site
Aki
Northern
Trifunac
for
propagation
to
this
of of
of
weak
of
continually
be
of
correlations,
SCEC
differences
stations
where
of
have
of
and
this
and
strong
this
the
regular
wave
and
will
motion
a
These
Local
in
readily
of
to
soil-non-linearity
and
fault
Strong
(depths
Co-PIs
scientists
earthquake levels
Dr.
will effects, the
research
assessment
and
an
research
different
CaLifornia
also
which
frequency
seen
propagation
plays
ground
studies
G.
adequate
empirical
map
Site
to workshops
defined.
stations
research
Dr.
also
modifying
of
be
of
R.
will
Motion to
allow
and
desirable,
the
to
opinion
V.
with
ground
a
Martin,
monitored.
task
Characteristics
involved
4
task
structural
be
provide
sponsQred
complement
major
range
motion
Lee,
one
motions
characteristics of
linear
type
description
an
manner.
Funding
in
conducted.
comparisons
is
studies
studies
is
to
the
on
University
shaking
Shaking
expert
develop.
dimensional
University
associated
from
to
role
to
Southern
focus
of
and
the
is
site
in
response
maximum
components,
develop
systematically
with
by
one
in
soil
Southern
will
direction
continue
and
source-site
response.
Whereas
to
opinion
expressed
the
of
opinions
the
governing
on
in
of
prepare
of
stratigraphy)
commentary.
not
those
be
of
linear
of
Results
California,
between
USGS
a
range
empirical
the
site
Southern
Southern
Southern
Ground
used credible
ground
only
consensus
and
California
on
to
task
depends
of
principal
and
elastic
response
in
and
modelling
the
study
required
has
to be
the
the
of
needed
past
probabilistic
11-1
characteristics
issues,
install
shaking.
NSF
extent
levels
needed
on-going California
largest, where
approach,
values
been
Research
on
are
and
of
will
site
strong
fault
following
analyses
coordination
the
professional
earthquake
to
broad
to
maps
known
of
of
non-linear
refine
effects
exceeded.
but
site
to
of
separate
number
studies,
sponsor
Current
damage
ground
similar
terms.
motion
funds
these
fully
more
of
band
and
soil
the
the
to
to
at
of all In this task, the duration of strong motion will be defined as that portion of the accelerogram which contributes 90% of all energy available for excitation of linear structures. This definition was proposed by Trifunac and Brady in 1975, and over the period 1975-82, used to progressively develop empirical equations between duration, magnitude, epicentral distance and site geology by utilizing the available earthquake data at that time. For this study, the research will utilize the expanded data base described in Task H-l, and will encompass:
• Generation of band-pass filtered data to describe the frequency components making up the strong motion duration.
• Performance of regression analyses on the frequency dependent duration components for scaling in terms of source magnitude, epicentral distance and site conditions.
• A comprehensive final report.
Results from the project will provide the means for assessing the magnitude and number of strong motion pulses in any frequency band for a given site and design earthquake. This in turn will provide the basis for assessing the appropriate means for including duration as an important seismic design parameter in design guidelines or codes.
Task H-5: Geotechnical Site Data Base for Southern California
Co PIs: Dr. Vucetic (UCLA) and Dr. G. R. Martin (USC) -
Site soil conditions, expressed in terms of soil stratigraphy which is defined by soil classification and shear wave velocity profiles, form an essential component of research on the effects of site response from earthquake ground motions. (Tasks H- 1, H-3 and H-4). Such site data expressed in simplified category form (e.g., rock, stiff soil, deep. soil profiles, soft soil site, etc.) also play a role in existing seismic design codes.
A large data base on site shear wave velocity profiles in southern California exists in a variety of open file publications and consultants reports. Some of these data are associated with strong motion accelerograph station sites. Additional synthetic shear wave velocity profiles have been generated from correlations with geotechnical data such as standard penetration blow counts.
The objective of this research task, is to systematically collect and catalog these data, which will ultimately form a data file in the planned SCEC geographic information system. Additional synthetic shear wave velocity profiles will be generated from collected borehole and cone penetrometer geotechnical data at selected locations. Particular attention will be focused on key strong motion instrument locations which lack site profile data. For these sites, shear wave velocity profiles will be generated by either extrapolation from nearby sites, or through direct field measurements using a ‘seismic cone penetrometer.
5
San
from
liquefaction southern Task
showing
California
developed
techniques characteristics
major attenuation chance the
being
uniform bridge Japan
magnitudes Task
evaluate from
moving collected
researchers damage correlated
documented bridges.
acceleration
same San
aid
Co Task
AASHTO
PIs:
in
Fernando,
Bernardino,
damage
the
Earthquake
This
degree.
Design
Whereas
In
H-8: and
damage
11-6:
considered code, H-7:
of
the
and
risk
towards
liquefaction
case
California
this
through
(for
not
results
with
for
research
development
New
deterministically.
relations
and
potential
and
time
are
code
in
Liquefaction
Evaluation
Probabilistic
the
levels.
Dr.
earthquake California
Risk)
being histories
occurring
research
Southern
selected
Loma
the
a
to
bridge
of
Observed
the
often
J.
recurrence
Zealand,
pertinent
associated
final
for
histories).
defining
etc.),
induced
correspondence
of
Hall
bridges
development
by
uniform
have
exceeded
task
levels
Prieta)
and
In
for
example,
potential
more
Tasks
keyed
report.
damage
in
risk
task,
and
the
in
this
and
of
spectral
of
is
been
more
response
in
California
Earthquake
structural
many
agencies.
liquefaction
of
precisely,
and response
levels)
Dr.
Characteristics
relationships
Evaluation
proposed
levels
National performance
has
approach,
recent
to
H-i
Bridge
particular,
risk
in
It
ground
site
are
resulting
(in
R.
prepared
detail,
a
In 50
of
is
other
been
overseas
particular
ordinates
Scott
and
keyed
of
and
spectra anticipated
an and
artificial
case
specific years.
spectra,
addition,
are
However, Damage
to
ground
spectra
damage
6
Earthquake
shaking,
the
may
empirical
in
Sites
11-2.
structures
of compare
(Caltech)
personal
Time
will
be from
histories
of
to
for
by
routinely
addition
criteria
subject
earthquakes
saturated
Revised
be
level
are
initiated
for
maximum
such
time
investigation
various
specific Response
be
and
shaking
studies
The
strong
warranted.
in
to
that
Histories
can
in
personal
required.
shapes
manner)
seismotectonic
assessed
of
telephone
histories of
as
for
terms
older
of
to
during
Hazards
objectives
Recent
used.
building
Potential
provide
risk.
much
earthquake
cohesionless
in
those
extensive
faults earthquakes
requiring
agencies
the
will
ground
(estimated
have
year
and
bridges
discussions
Spectra
of
for
for
Standard
Numerous
seismic
in
of
for
past
Results
specified
be
techniques
Earthquakes
or
seismic
a
calls
Reduction
spectral
2,
not
magnitudes
terms
the
a
accelerators
will
significant
use
Southern
(CDMG
of
seismo-tectonic
made
study,
and
maximum
earthquakes.
number
damage
in
been
zones
soils
or
to
information
in
retrofit
of
in
(Uniform
be
spectral
design
will
and
of
particular,
recorded
by
the
known
research
research
design.
of
ordinates
reports,
lessons
has
studied
to
the
Program,
the to
draw
of
in
to
site
with
U.S.
probabilistic
database
with
earthquake
of
resulted
assess
codes
bridges
regions
causes
shapes
Southern
evaluate
overseas
visits
AASHTO
(Alaska,
existing
learned
will
heavily
can
ground
spectra
studies
City
to
a
zones,
when
Maps
with
now
90%
the
are
the
be
be
to
of
in
to
in
in
of in
with will
Director Geoffrey in
experience components, of the
committees NSF/USGS to Task
administrative selected southern potential sponsored
and designated design spectra quantitative
methods
monitor
IV. more
year
vertical
determine
existing
comprise
reporting
other
To
The The
PROGRAM Seismic
Dr.
11-9:
recent
criteria
3)
progress,
for
for
and
and
suitable
California.
ensure
R.
sponsored
Martin
will
proposed
accelerations by
in
SCEC
will
overall
SCEC
Characteristics
and
southern
Engineering
Martin, to
the
approach
seismic
invited
the
for design
the
the
representatives,
in
information
be
be
establish
activities.
COORDINATION
Principal
that
recommendations
will
for
design
topic
vertical
their
to
make
established
Task
NSF
research
objective
SCEC
Comparisons
loads
study
codes
to
mapping,
SCEC
relevant
be
California
for
areas.
particular
and
may
participate
P1(s),
suggestions
codes.
methods
the
Applications.
research
Investigators
for
acceleration
the primarily
Assistant
assessment
USGS
regarding
be
of
program
principal of
for
structural
results
using
They
and invited
frequency
AND
this
significant.
liquefaction
will
each
to
Vertical
tasks.
are
groups
made
will
in
MANAGEMENT
task
on
evaluate
will
the
7
address
be
Director
of
task
the
currently
interface
SCEC will
research
for
of
site
prepare
Review
components
design.
for
results
related
(commencing
be
made
and
is
from program
post
each
Ground
The
be
integrated
responsible methods
scientists/engineers,
conditions,
post-liquefaction
horizontal
potential
amplitude
for
with
of
coordinated
between
the
committees
quarterly
objective
However,
research
in
liquefaction
directions,
of
ongoing
Engineering
progress
because
SCEC
the
response
to
of
Accelerations
in
into
maps
earthquake
include
year
for
characteristics
research
accelerograph
scientific
SCEC
and
being
of
in
progress
NSF
C 3
and
will
the
of
some
to
and this
ground 2)
ground
to
spectra
and
to
research,
sponsored
vertical
their
improve
conducted
will
Applications.
provide
technical
take
and
meet
task
tasks,
develop
pool.
managed
cases,
C 3
reports.
ground
be
deformations.
expertise
for
technical
the
into
deformation
(commencing
quarterly
of
task
records
acceleration
Committees
re-evaluate
have
the
integration
predictive
horizontal
research.
Assistant
response
direction
in
a
account
motions
by
review
effects
other
more
been
The
and
and
Dr.
in
to H-7