of 2
Eos,
Vol. 79, No. 18, May 5, 1998
VOLUME 79 NUMBER 18
MAY 5,1998
PAGES 217-228
Major
Improvements
in
Progress
for
Southern California
Earthquake
Monitoring
Jim
Mori, Hiroo
Kanamori,
James
Davis, Egill Hauksson, Robert
Clayton,
Thomas Heaton, Lucile
Jones,
Anthony Shakal, and
Ron
Porcella
PAGES
217,
221
Major
improvements in seismic and
strong-motion monitoring networks are being
implemented in southern California to better
meet
the needs of emergency response per­
sonnel,
structural engineers, and the re­
search
community in promoting earthquake
hazard reduction. Known as the TriNet pro­
ject,
the improvements are being coordi­
nated by the California Institute of
Technology
(Caltech),
the
U.S.
Geological
Survey
(USGS),
and the California Division of
Mines
and Geology
(CDMG)
of the state's De­
partment of Conservation. Already the ambi­
tious instrument and system development
project
has started to record and disseminate
ground motions from a spatially dense and ro­
bust network of high quality seismographs.
Deficiencies
in the southern California
seis­
mic
monitoring network and in other networks
became
apparent
with the
1994
Northridge,
California,
earthquake and the
1995
Kobe,
Ja­
pan, earthquake. Both earthquakes occurred
in relatively well-monitored regions but certain
vital
information was lacking in the instrument
data. In the Northridge event, in which there
was about
$30
billion in losses, there was exten­
sive
damage to a large number of structures in­
cluding the well-publicized problem of cracks
in the welds of large steel frame buildings
[Krawinkler
et al.,
1995].
There were, however,
relatively
few strong-motion recordings to
document the actual ground shaking that
caused the damage. In
Kobe,
where over
6,000
people were killed and some
$200
billion in
damage was caused, there was a lack of real­
time information that might have facilitated
For
more information, contact Jim Mori,
U.S.
Geological Survey, 525 S. Wilson Ave.,
Pasadena, CA 91106 USA
emergency
response, as well as a lack of
strong-motion data.
Main
Objectives
The
TriNet project has three main objectives.
The
first is to provide ground shaking data
within just several minutes of a damaging earth­
quake so that the effectiveness of emergency
response can be increased. Point measure­
ments at the critical locations and contour
maps of the affected areas will be distributed
quickly
to local emergency response groups so
that appropriate measures can be taken in
heavily
impacted areas.
The
second objective is to record ground
motion data in order to improve seismic pro­
visions
in building codes and in research.
Regulatory
decisions on building and retrofit
practices
are in need of this better informa­
tion on expected ground motions. The array
of
wide-dynamic range and broadband in­
strumentation will also provide some of the
best
data on
local,
regional, and teleseismic
waveforms forseismological research.
The
third objective is to develop a proto­
type early warning system. Ground motion in­
formation will be rapidly recorded, analyzed,
and communicated to test user
sites.
Seismic
waves travel at the speed of a few kilometers
per second, so rapid interpretation of earth­
quake information is essential for an early
warning to distant locations (over 100 km)
that large seismic waves from an earthquake
are on the way. This is the situation for Los
Angeles
with respect to the San Andreas fault
where a
10-40
second warning could be
given before the intense shaking from a large
San
Andreas earthquake arrives in Los Angeles.
The
Caltech/USGS
segment of the project
involves
upgrading
the regional seismic net­
work into a modern system with broadband
and high dynamic-range recording com­
bined with new digital telemetry technolo­
gies.
The
CDMG
segment is closely tied in
with Caltech and
USGS
to integrate extensive
strong motion recording into the system. This
instrumentation will make possible the produc­
tion of several new products for distribution
through
the Internet and other electronic com­
munication channels. These include
•Quick
maps of potentially damaging
ground shaking (ShakeMap) following signifi­
cant
earthquakes. Up to 670 instruments in
southern California will be reporting informa­
tion.
•An
easily
accessible
database of all para­
metric
and waveform data. This earthquake in­
formation from local and teleseismic events
will
provide one of the best collections of
seis­
mic
data for utilization by engineers and
seis­
mologists.
•Frequency-dependent amplification
maps for the Los Angeles region. These maps
will
quantify the geology-dependent site ef­
fects
that locally amplify or attenuate shaking.
•A
prototype early warning system. A sys­
tem will be developed to test the feasibility of
providing very rapid warnings of expected
seismic
waves. This will include a data utiliza­
tion component which will examine the
problems of implementing a system and
edu­
cating
users about this technology.
Evolution
of the
Project
Following
the Northridge earthquake in
1994,
the
USGS
received
funds
appropriated
by
Congress to the National Earthquake Haz­
ard Reduction Program (NEHRP) to improve
seismic
recording in southern California. The
allocation
was originally split into two parts,
real-time system improvements to the regional
network, the other for the
upgrade
of strong-
motion recording. Early in the planning, the
USGS
decided it was more appropriate to com­
bine
the two and begin to build a single system
that integrated the functions of the regional
seismic
network with the engineering strong-
motion network.
Seismologists
and engineers
of
USGS
(in Pasadena and from the National
Strong
Motion Program) and Caltech worked
together to design a system that took advan­
tage of modern instrumentation which could
record
both weak motions from small events
and the strong shaking from larger earthquakes.
This
page may be freely copied.
Eos, TRANSACTIONS, AMERICAN
GEOPHYSICAL
UNION
Eos, Vol.
79,
No.
18, May
5,
1998
During the 3 years before TriNet formally
began, experimental systems were designed
and hardware was installed. For example, 60
state-of-the-art digital seismic stations were in­
stalled
with continuous communication to
Caltech/USGS.
New technology telemetry on
frame-relay
telephone lines and spread-spec­
trum
radios were tested. Also, the
USGS
Na­
tional Strong-Motion Program installed 30
free-field
strong-motion sites, with many near
critical
structures. Some 40 more sites will be
completed in the next year. In addition, 5
high-rise buildings were outfitted with instru­
ments. Other monitoring installations (in a
nonductile concrete building and in connec­
tion with a building drift experiment) will be
completed in the next year. Software that
combines
and broadcasts information from
the new digital signals and the existing ana­
log
system also has been designed and imple­
mented. Experimental maps of contoured
ground shaking are available on the Web
(http://www-socal.wr.usgs.gov) within a few
minutes after felt and damaging earthquakes
[WaldetaL,
1997].
Leading up to its participation in TriNet,
CDMG's
California Strong Motion Instrumen­
tation Program
(CSMIP)
developed a system
for
near real-time data recovery from strong-
motion stations in its network
[Shakal
et al.,
1996].
The
CSMIP
strong-motion monitoring
system
uses standard digital accelerographs
at field stations throughout California which
automatically transmit data via high-speed
dial-out communication links to Sacramento
computers using conventional phone lines.
While
in a pilot mode, the system automat­
ically
and rapidly transmitted and processed
data from stations after several events. In con­
trast to the classical strong-motion network,
the near real-time system creates a new type
of
strong-motion network with the ability to
quickly
recover and process data. Although
not truly real-time like traditional seismic te­
lemetry, the economy of the approach is a
valuable asset given a large number of strong-
motion stations. The concept of reference sta­
tions,
developed by
CDMG
and its advisory
committees,
also led to the development of
objectives
of the TriNet project. Reference
stations are ground response stations that pro­
vide recordings of the
input
motion in the
built environment. This is key to providing in­
formation for shaking maps in the urban area
and complements the traditional recordings
from
regional seismic stations. In the period
leading up to TriNet,
CDMG
also installed in­
struments at many structures (buildings,
bridges, and downhole arrays) and other sta­
tions in California.
Implementation
of
TriNet
With
major
support
from the Federal
Emergency
Management Agency
(FEMA)
hazard mitigation program through the Cali­
fornia
Office
of Emergency
Services,
the im­
plementation of TriNet began in
1997.
The
project
is divided into two elements with the
real-time information emphasis centered at
Caltech
and
USGS
in Pasadena and the exten­
sive
collection of strong-motion data for engi­
neering purposes centered at
CDMG
in
Sacramento.
By
2002,
there will be
670
sta­
tions in southern California. Caltech and
USGS
will install broadband and strong-mo­
tion sensors with continuous telemetry, ex­
panding the efforts started by the
USGS
to a
total of 200 high dynamic-range sites.
CDMG
will
install new sites and
upgrade
existing in­
strumentation to provide digital strong-mo­
tion recording at 400 stations. Many of the
strong-motion instruments will be
part
of the
CDMG
plan for establishing engineering refer­
ence
sites throughout the region. Also in­
cluded are the 70 digital strong-motion sites
being installed by
USGS
from the NEHRP
Northridge funds.
Connecting
the real-time telemetry from
the
Caltech/USGS
stations to the central proc­
essing
site in Pasadena is a major challenge
for
the project. One of the tasks is to provide
fast
and reliable communications that will be
robust even
during
the strong shaking of an
earthquake. A variety of telemetry paths will
be
utilized to minimize any single points of
failure.
New technologies such as frame-relay
telephone lines, spread-spectrum radios, and
digital microwave links will also be utilized
in the system.The 400
CDMG
sites (plus the 70
USGS
strong-motions sites) will be connected
by
dial-up
telephone telemetry. These instru­
ments will provide significant contributions
to the database of strong motions recorded
in southern California and provide the basis
for
improvements to the seismic provisions of
building codes. The near real-time communi­
cation
from these sites will provide addi­
tional information (on the order of tens of
minutes to an hour) to the initial estimates of
shaking distributions that will be provided by
the real-time system. Extensive software de­
velopment will be needed to handle the
large volume of digital data. Central process­
ing will be designed to handle data very
quickly
for the early warning applications, as
well
as to produce rapid estimates of earth­
quake source parameters and shaking distri­
butions. The system will also provide
organized databases of parametric and wave­
form
data with accessible user interfaces.
Acknowledgments
The
TriNet project is being built using ex­
isting infrastructure and with collaboration
between federal and state government, uni­
versity,
and private sectors. Southern Califor­
nia's first digital network began with the
installation of seismographs known as
TER-
RAscope,
made possible by a grant from the
Whittier
Foundation and the ARCO Founda­
tion.
Pacific
Bell,
through its
CalREN
Pro­
gram, has provided new frame-relay digital
communications technology for telemetry.
The
CUBE
(Caltech-USGS
Broadcast of Earth­
quakes)
project, started in
1991,
has formed a
consortium of government agencies and pri­
vate industry concerned with earthquake haz­
ards
in southern California. This group has
built
support
for the project and provided
valuable user
input
into the design of the
earthquake information systems. In addition
to the initial
support
from
USGS
and the cur­
rent
support
from
FEMA,
numerous govern­
ment and private organizations have made
significant
contributions to the funding.
These
include the National
Science
Founda­
tion, California Trade and Commerce
Agency,
and
Pacific
Bell
through the
CalREN
project.
Cost-sharing of the
FEMA
funding is
also
provided by Caltech and
CDMG.
The
Southern California Earthquake Center Data
Center
continues to provide facilities for data
storage and distribution.
References
Krawinkler,
H., J. Anderson, V.
Bertero,
Steel
buildings,
in
Northridge
Earthquake
of
Janu­
ary
17, 1994
Reconnaissance Report, Vol­
ume
2,
edited by W.T. Holmes and P.
Somers,
Earthquake
Spectra,
11, Suppl. C, 25-
47,
1995.
Shakal,
A., C. Petersen, A. Cramlet, and R.
Darragh,
Near-real-time
CSMIP
strong motion
monitoring
and reporting for guiding event
response,
in
Proc.
11th
World
Conf.
Earth­
quake
Engineering,
Sociedad
Mexicana
de
Ingenieria
Sismicica,
A.C.
(ed.),
Acapulco,
Mexico,
1996.
Wald,
D. J., T. Heaton, H. Kanamori, P.
Maechling,
V. Quitoriano,
Research
and De­
velopment
of TriNet
"Shake"
Maps
(abs.),
Eos,
Trans.,
AGU, 78,
Fall
Meet.
Suppl.,
F45,
1997.
This
page may be freely copied.