P1252428529LfQrh

1
Thailand Earthquake Tsunami Warning Training
Program
Session III.2 Global and Local
Arrays Presented by R.F. Mereu Department of
Earth Sciences University of Western
Ontario London, Ontario, Canada May 17,
2006 Bangkok, Thailand Sponsored by TMD,ATT,
USGS, USAID
2
Topics covered

• Comparison of networks with arrays
• Examples of global networks
• Example of global arrays
• How are arrays and networks used
• Challenges in setting up a local network
• ----- Example the Canadian experience
  with the SOSN/POLARIS networks
• Challenges for Indian Ocean

3
Seismic network

• Each stations clock is independent
• Data recording may be at the station or at
  a common data center
• Waves may not be coherent as they propagate
  across the network Stations are far apart

Seismic array

• Has common time base (one clock)
• Has common recording center
• Waves remain coherent as they propagate
  across the array
• Stations are close to each other.

4
Example Showing Coherent traces
Recorded by the SOSN/POLARIS Seismic Network
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
Map of regional network seismometer stations in
the USA in 1999
From Stein and Wysession, 2003 , An
introduction to seismology, earthquakes and earth
structure
13
Canadian National Seismograph
Network
14
The Polaris Network in
Canada 4 sub-networks ---- Southern Ontario
(SOSN), Northern Ontario (FEDNOR),
Northwest Territories (NWT)
and British Columbia (BC)
15
In a network, the location of an earthquake is
found by triangulation. Information from all
stations is used.
16
From USGS Web Site
17
The 2004 Sumatra-Andaman Isl quake radiated very
long period seismic waves that were recorded by
global networks.
18

Dispersed Surface Waves frum Sumatra-Andeman
Earthquake Recorded by station SOSN ACTO in
Ontario, Canada
19
GSN data are used to study the internal seismic
structure of the Earth
Courtesy of Adam Dziewonski, Harvard University
20
Seismic Arrays
Signals are coherent across stations
21
The Large Aperture Seismic Array
(LASA)
Seismometer geometery of the Large Aperture
Seismic Array (LASA), 1969, J.Geophys. Res.,74,
3182-94
22
An array can be used as an antenna to determine
the direction from which the seismic waves
arrive. This process, called beamforming, tells
where the earthquake is located.
23
EARTHSCOPE The USArray component
of the EarthScope experiment is a
continental-scaleseismic observatory designed to
provide a foundation for integrated studies of
continental lithosphere and deep Earth structure
over a wide range of scales. USArray will provide
new insight and new data to address fundamental
questions in earthquake physics, volcanic
processes, core-mantle interactions, active
deformation and tectonics, continental structure
and evolution, geodynamics, and crustal fluids
(magmatic, hydrothermal, and meteoric). The
USArray facility will consist of three major
seismic components 1. A transportable array of
400 portable, unmanned three-component broadband
seismometers deployed on a uniform grid that will
systematically cover the US 2. A flexible
component of 400 portable, three-component,
short-period andbroadband seismographs and 2000
single-channel high frequency recorders for
active and passive source studies that will
augment the transportable array, permitting a
range of specific targets to be addressed in a
focused manner and 3. A permanent array of
high-quality, three-component seismic stations,
coordinated as part of the US Geological Survey's
Advanced National Seismic System (ANSS), to
provide a reference array spanning the contiguous
United States and Alaska.
24
The NORSAR Array
25
WRA, GBA, YKA Seismic Arrays Constructed in
1960s
WRA
YKA
GBA
Upper mantle (400 and 650 km) discontinuities
detected in the 1960s from analysis of array data
26
Comparison between 2 Upper Mantle Velocity
Models Model 1

Model 2 Smooth increase in velocity
with depth This
model has a 400 km and 650 km transition zone
Upper mantle travel-time curve has no
triplications The travel-time curve
has 2 triplications Seismic
arrays may be used to provide direct measurements
of the slopes of the travel curves
as is shown on the next figure
27
A record section from the Yellowknife array in
Canada Two phases are seen arriving at the array.
The distance to the earthquake was 21.69
degrees. Slowness measurements shown on the right
graph clearly shows the two phases are arriving
with different slownesses This indicates the
presence of 2 travel-time branches Numerous
slowness measurements from many arrays confirmed
that upper mantle has 2 discontinuities or
transition zones.
28
Seismic traces recorded at 4 sub-arrays of the
NORSAR array in Norway The earthquake distance
is approximately 2750 km Two phases are
clearly seen which have take different paths in
the mantle. Note the strange inconsistent
amplitude behaviour of the phases. These
variations are due to differences in the site
responses of the rocks beneath each sub-array
29
Upper Mantle Ray Paths
30
Arrays used in Controlled Source Seismic
Experiments Numerous closely spaced
stations Good coherency between traces
allows velocity filtering
31
Project Early Rise 1966
All stations were located near lake
shores Transportation to all sites was by
Beaver airplane
Instruments Batteries Analog Amplifier
and filter Geotech FM tape recorder
32
The 1986 GLIMPCE Seismic
Experiment Energy Source Ship with air gun
--- Shot spacing 60 100 meters A coincident
seismic refraction/reflection experiments
Receivers --- hydrophones in lake seismometers
at land stations
33
Seismic instruments used by the University of
Western Ontario in the Canadian-Co-Crust seismic
refraction experiments 1979-1985 and the GLIMPCE
experiment 1986 Electronics analog FM
Seismometer 1Hz Mark Products
34
1986 GLIMPCE Experiment ---- Small part of one
record section This figure shows that the PG
wave train is shingled Note trace spacing
was 60 meters
35
1986 GLIMPCE Experiment --- Record section from
stations at N and S ends of Line A
36
Seismic Reflection and Refraction Lines Across
the Canadian Shield in Ontario and Quebec
1982-1992
Red lines 1991-1993 Lithoprobe --- Near Vertical
Reflection Lines Blue Lines --- 1992
Lithoprobe AG ---- Longe range Seismic
Refraction Lines 42 shot points each
recorded by 400 stations Green Lines--- 1982
COCRUST ---- Longe range Seismic Refraction
Lines 8 shot points each recorded by 36
instruments
37
Controlled Source Seismology ---- Near Vertical
Reflection Method -
Truck above is a vibroseis truck. Usually 4 or 5
trucks are used together. Sends a chirp signal
into ground. Normally recorded by arrays of up to
1000 stations to obtain high resolution image of
crust. Chirp
Signal
From Lithoprobe web site
38
1992 Lithoprobe Abitibi-Grenville Reflection
Migrated Image Section
Moho
From Lithoprobe web site
39
The 1992 Lithoprobe seismic Refraction experiment
across the Canadian Shield in Ontario and
Quebec. Participants 5
Canadian Universities
Geological Survey of Canada
United States Geological Survey
Industry --- GPR 44 shots each shot
recorded by 400 stations Analysis --- Tomographic
200 Stanford SGRs
(Recorded on digital tape cassettes) Seismic
Instruments used 200 GSC PRS1 (No tapes
solid state memories)
40
TWO Seismic Refraction Record Sections from the
1992 Lithoprobe Seismic Refraction
Experiment. Note PmP is much clearer
in lower section PmP is the wide-angle
reflection form the Moho From Winnardhi
and Mereu, 1997.
41
The 1992 Lithoprobe Seismic Refraction
Experiment- Tomography Analysis Rays traced
through laterally heterogeneous crust. Seismic
modelling is done with a triangular mesh Each
triangle has a linear velocity gradient
From Winnardhi and Mereu, 1997.
42
Seismic Tomography velocity model and velocity
anomaly results Data is from the 1992
Lithoprobe seismic refraction experiment
From Winnardhi and Mereu, 1997.
43
Challenges in Setting up a Local Network
Example the Canadian experience with the
SOSN/POLARIS networks
( 1991-present time)

44
Seismicity Map for Southern Ontario, Canada
45
(No Transcript)
46
The Western Lake Ontario Region of Southern
Ontario Canada Major
Urban Area ---Toronto Niagara Falls
Toronto
Niagara Falls
47
SOSN Seismic ---gt station
48
Seismic Vault at SOSN- PKRO
49
(No Transcript)
50
Construction of the Satellite Receiving Dish for
POLARIS
51
(No Transcript)
52
(No Transcript)
53
(No Transcript)
54
(No Transcript)
55
(No Transcript)
56
(No Transcript)
57
(No Transcript)
58
(No Transcript)
59
Pg
60
(No Transcript)
61
Shake-Map for Lake Ontario
Earthquake Aug
8, 2004, Magnitude 3.7
From POLARIS
web site ---- www.polonet.ca
62
Challenges for The Indian Ocean
63
(No Transcript)
64
From Geist, Titov, and Synolakis , 2006
Scientific American , January.
65
From Geist, Titov, and Synolakis , 2006
Scientific American , January.
66
Summary of Challenges 1. Much
more detailed information on seismicity of region
is required to map active faults
--- increase precision of location of earthquake
hypocenters --- increase
precision of earthquake source parameters
--- not possible to cover whole region in
detail therefore need regional studies coupled
with more focussed studies 2. More
detailed information required on crustal and
upper mantle structure of region.
The transition from oceanic crusts to continental
crusts around the ocean coupled with
the presence of a major subduction zone poses a
real challenge to unravelling the detailed
geometry of the region ---
recommend a series of off-shore-onshore
coincident seismic reflection/wide-angle
refraction surveys --- to determine
earthquake hypocenters, one needs to know the ray
paths from source to station. ---
This requires good knowledge of both lateral and
vertical velocity variations in the crust and
upper mantle.
3. The location of earthquake hypocenters
--- To improve precision increase the
number of stations-- cost goes up as number goes
up. --- decisions will have to be
made where to put the stations to maximize the
scientific return --- networks must
be optimized for a particular application
e.g. For regional earthquakes ---
station spacing at about 25 km
For teleseismic studies ---- minimum no
of stations 20 with a station spacing at 1-2 km
Mantle studies ---
broad-band instruments to determine surface wave
dispersion curves etc
Urban areas --- strong motion instruments space
to maximize information on soil types
-
67
4. Simulation Studies Before
stations are deployed, a detailed simulation
program should be undertaken. Here the
computer is used to insert earthquakes of various
focal mechanism orientations at various
locations along the active faults.
Theoretical seismograms may then be computed to
various possible station sites.
Standard inversion methods may used to analyse
the simulated data. The answers may then
be checked with the known input data.
The simulation exercises could reveal some
surprises and if carried out properly could
prevent costly deployment mistakes.
5. Seismometer vaults ---
ideally on bedrock --- if rock not
available --- choose quiet areas on clay or
gravel type soil noises --- Avoid
noisy sites such as those close trees , highways,
factories etc --- Vaults should
be leak-proof and preferably buried 6. Power
to vaults in isolated areas --- solar panels,
wind generators --- major problem
--- vandalism 7. Communication with the
stations --- always a problem ----
ideal to have real time communication-- phone
line , radio, internet and satellite
---- satellite is best from logistic point of
view but is also expensive ---- near
real-time communication --- dial-up stations
8. Off-shore stations such as Ocean Bottom
Seismometers (OBS). ---- very
expensive but scientific return may be worth it.
---- may have to be linked to a buoy
and satellite to provide real-time data
information ---- recommend use in
special studies along the Sumatra fault.

68
9. More studies on seismic hazard analysis
Predicting the effects of ground
motion shaking on populated centers.
Simulated studies could be carried out here and
then compared to real observations.
11. Where possible make use of other
geophysical data e.g. Gravity
surveys, deformation measurements from GPS etc
12. Data acquisition, management, storage and
distribution. This is very important
and has to be done right. The
acquisition of data is very expensive. If the
data sits unused on some laboratory DVD ,
all the money in acquiring it is
effectively wasted. 13. Compromises
Every seismic study starts off
with ideal aims. When the work
starts, unexpected problems always arise.
It is easy for costs to get out of hand.
Plans have to be compromised.
69
The
END Thank You!!