6.6 Seismometers and seismological networks 409
System (IMS) network used to monitor nuclear testing (Section
1.2.8).
Although the present global network of broadband seismometers relies on land sites, it is hoped that the global network
will soon include permanent ocean bottom seismometers
(OBS), especially in the Southern Hemisphere, where there
is much less land, and coverage is currently very uneven.
Although OBS instruments are currently used mostly for temporary deployments, the technology is evolving to the point
where permanent sites are practical.
An important aspect of the different networks of high-quality
broadband seismometers is considerable standardization in
data processing and formatting. All 7 terabytes of seismic data
archived by the IRIS DMC 8 as of 2000 are available in a format
called SEED (Standard for the Exchange of Earthquake Data),
which is the standard for the FDSN. SEED data can be converted into whatever format an investigator requires.
It was not until the mid-1990s, more than 30 years after the
start of the WWSSN, that the global number of permanent
digital broadband seismometers surpassed the number of
WWSSN stations at its heyday. However, digital data from all
parts of the FDSN can be retrieved as if it were a single array,
making it more powerful than the WWSSN for seismic analyses. Many stations now report in real time through satellite
telemetry, so seismic signals arrive at data centers a fraction
of a second after they occur, allowing better quality control.
Efforts are being made to eventually have all GSN stations
report in real time, which will be important for applications
like tsunami warning. Software has been developed to take
real-time data from different networks and display it on
the Internet as if it were from a single array. Hence, anyone
with a computer and access to the Internet will soon be able
to examine global seismic data within seconds of them being
recorded.
6.6.8 Arrays
For global networks, the precise configuration of individual
stations is less important than the total coverage. However,
the geometries of seismic arrays are optimized for certain
investigations. Arrays can be linear, two-dimensional, and
even three-dimensional, incorporating borehole seismometers
(Fig. 7.3-8).
There is always a trade-off between the benefits of linear
versus two-dimensional arrays. The same number of stations,
and therefore cost and time for installation, provides greater
resolution if deployed in a linear manner, but the resulting
two-dimensional “slice” into the earth does not image the third
dimension. Linear arrays have long been the mainstay of active
source reflection and refraction experiments. 9 A marine linear
array is easily deployed by towing hydrophones behind a ship,
and similar linear deployments are used for land-based studies.
These data are analyzed using techniques discussed in Sections
3.2 and 3.3.
Linear arrays are most useful if the structure being investigated varies most in one direction, as is often the case at plate
boundaries. For instance, Fig. 5.3-10 (bottom) showed the
seismic structure of the East Pacific rise obtained from an array
of OBSs. Because the structure of the lithosphere changes much
more significantly perpendicular to the ridge than parallel to
it, most of the OBSs were deployed in a line crossing the
ridge. Most of the remaining seismometers were placed in a
second line, parallel to the first. Both lines were aligned along
a great circle path to the seismogenic zones of Tonga and
South America, so as to maximize the chance of obtaining good
signals from distant earthquakes. Similarly, at subduction
zones and transform faults structure varies more significantly
across the plate boundary than along it, so refraction lines are
often placed perpendicular to the boundary. For example, Fig.
3.2-17 showed a cross-section of the western US lithosphere
perpendicular to the San Andreas fault that was derived from
refraction surveys.
Two-dimensional arrays can create a three-dimensional
image of a small region. As a result, two-dimensional arrays
have been deployed around hot spots, rifts, plateaus, transform
faults, and subduction zones to study their structure and tectonics. Reflection data are also now commonly gathered by
two-dimensional surface deployments. An important contributor to this development has been advances in computers and
graphics software that make it possible to analyze and model
such data and display the resulting earth structure in a comprehensible fashion. Such three-dimensional images are of great
importance in exploring for oil and gas and managing existing
oil and gas fields.
Special two-dimensional arrays, often consisting of shortperiod vertical seismometers, have been used to monitor the
locations and magnitudes of underground nuclear tests. The
most ambitious such array was the circular Large Aperture
Seismic Array (LASA), which operated in Montana from the
mid-1960s until 1978. LASA was an array of arrays totaling
525 high-frequency vertical seismometers. Twenty-one clusters of 25 seismometers, each covering 7 km 2 , were deployed
with a total array diameter of 200 km (Fig. 6.6-17). A similar
array is the Norwegian Seismic Array (NORSAR), built in
1971, with 22 sub-arrays spanning an area of 100 km 2 . Part
of NORSAR, the NORESS array, has 24 seismometers distributed within a 3 km-diameter circle. It has counterparts in
northern Norway, Finland, and Germany. As with the WWSSN,
arrays designed for nuclear monitoring have also been important for studies of earth structure. Array data can be stacked
(Section 6.5), allowing small seismic signals to be extracted
from noise. The characteristics of the inner core boundary were
first quantified using stacked array data for PKiKP waves, which
reflect at the boundary but are rarely identified on individual
seismograms due to their small amplitudes.
8 Because all data are duplicated in a sort order, and also stored off site, the computer storage needed is four times greater, or 28 Tbytes.
9 Active experiments include their own seismic sources, as opposed to passive
experiments using earthquake sources.
System (IMS) network used to monitor nuclear testing (Section
1.2.8).
Although the present global network of broadband seismometers relies on land sites, it is hoped that the global network
will soon include permanent ocean bottom seismometers
(OBS), especially in the Southern Hemisphere, where there
is much less land, and coverage is currently very uneven.
Although OBS instruments are currently used mostly for temporary deployments, the technology is evolving to the point
where permanent sites are practical.
An important aspect of the different networks of high-quality
broadband seismometers is considerable standardization in
data processing and formatting. All 7 terabytes of seismic data
archived by the IRIS DMC 8 as of 2000 are available in a format
called SEED (Standard for the Exchange of Earthquake Data),
which is the standard for the FDSN. SEED data can be converted into whatever format an investigator requires.
It was not until the mid-1990s, more than 30 years after the
start of the WWSSN, that the global number of permanent
digital broadband seismometers surpassed the number of
WWSSN stations at its heyday. However, digital data from all
parts of the FDSN can be retrieved as if it were a single array,
making it more powerful than the WWSSN for seismic analyses. Many stations now report in real time through satellite
telemetry, so seismic signals arrive at data centers a fraction
of a second after they occur, allowing better quality control.
Efforts are being made to eventually have all GSN stations
report in real time, which will be important for applications
like tsunami warning. Software has been developed to take
real-time data from different networks and display it on
the Internet as if it were from a single array. Hence, anyone
with a computer and access to the Internet will soon be able
to examine global seismic data within seconds of them being
recorded.
6.6.8 Arrays
For global networks, the precise configuration of individual
stations is less important than the total coverage. However,
the geometries of seismic arrays are optimized for certain
investigations. Arrays can be linear, two-dimensional, and
even three-dimensional, incorporating borehole seismometers
(Fig. 7.3-8).
There is always a trade-off between the benefits of linear
versus two-dimensional arrays. The same number of stations,
and therefore cost and time for installation, provides greater
resolution if deployed in a linear manner, but the resulting
two-dimensional “slice” into the earth does not image the third
dimension. Linear arrays have long been the mainstay of active
source reflection and refraction experiments. 9 A marine linear
array is easily deployed by towing hydrophones behind a ship,
and similar linear deployments are used for land-based studies.
These data are analyzed using techniques discussed in Sections
3.2 and 3.3.
Linear arrays are most useful if the structure being investigated varies most in one direction, as is often the case at plate
boundaries. For instance, Fig. 5.3-10 (bottom) showed the
seismic structure of the East Pacific rise obtained from an array
of OBSs. Because the structure of the lithosphere changes much
more significantly perpendicular to the ridge than parallel to
it, most of the OBSs were deployed in a line crossing the
ridge. Most of the remaining seismometers were placed in a
second line, parallel to the first. Both lines were aligned along
a great circle path to the seismogenic zones of Tonga and
South America, so as to maximize the chance of obtaining good
signals from distant earthquakes. Similarly, at subduction
zones and transform faults structure varies more significantly
across the plate boundary than along it, so refraction lines are
often placed perpendicular to the boundary. For example, Fig.
3.2-17 showed a cross-section of the western US lithosphere
perpendicular to the San Andreas fault that was derived from
refraction surveys.
Two-dimensional arrays can create a three-dimensional
image of a small region. As a result, two-dimensional arrays
have been deployed around hot spots, rifts, plateaus, transform
faults, and subduction zones to study their structure and tectonics. Reflection data are also now commonly gathered by
two-dimensional surface deployments. An important contributor to this development has been advances in computers and
graphics software that make it possible to analyze and model
such data and display the resulting earth structure in a comprehensible fashion. Such three-dimensional images are of great
importance in exploring for oil and gas and managing existing
oil and gas fields.
Special two-dimensional arrays, often consisting of shortperiod vertical seismometers, have been used to monitor the
locations and magnitudes of underground nuclear tests. The
most ambitious such array was the circular Large Aperture
Seismic Array (LASA), which operated in Montana from the
mid-1960s until 1978. LASA was an array of arrays totaling
525 high-frequency vertical seismometers. Twenty-one clusters of 25 seismometers, each covering 7 km 2 , were deployed
with a total array diameter of 200 km (Fig. 6.6-17). A similar
array is the Norwegian Seismic Array (NORSAR), built in
1971, with 22 sub-arrays spanning an area of 100 km 2 . Part
of NORSAR, the NORESS array, has 24 seismometers distributed within a 3 km-diameter circle. It has counterparts in
northern Norway, Finland, and Germany. As with the WWSSN,
arrays designed for nuclear monitoring have also been important for studies of earth structure. Array data can be stacked
(Section 6.5), allowing small seismic signals to be extracted
from noise. The characteristics of the inner core boundary were
first quantified using stacked array data for PKiKP waves, which
reflect at the boundary but are rarely identified on individual
seismograms due to their small amplitudes.
8 Because all data are duplicated in a sort order, and also stored off site, the computer storage needed is four times greater, or 28 Tbytes.
9 Active experiments include their own seismic sources, as opposed to passive
experiments using earthquake sources.