26 Introduction
damaged more than 90% of the houses. The prediction is said
to have been based on precursors, including ground deformation, changes in the electromagnetic field and groundwater
levels, anomalous animal behavior, and significant foreshocks.
However, in the following year, the Tangshan earthquake
occurred not too far away without precursors. In minutes,
250,000 people died, and another 500,000 people were
injured. In the following month, an earthquake warning in the
Kwangtung province caused people to sleep in tents for two
months, but no earthquake occurred. Because foreign scientists have not yet been able to assess the Chinese data and the
record of predictions, including both false positives (predictions without earthquakes) and false negatives (earthquakes
without predictions), it is difficult to evaluate the program.
In summary, despite tantalizing suggestions, at present there
is still an absence of reliable precursors. The frustrations of this
search have led to the wry observation that “it is difficult
to predict earthquakes, especially before they happen.” Most
researchers thus feel that although earthquake prediction
would be seismology’s greatest triumph, it is either far away
or will never happen. However, because success would be of
enormous societal benefit, the search for methods of earthquake prediction will likely continue.
1.2.7 Real-time warnings
Some recent efforts are directed to the tractable goal of realtime warnings, where seismometers trigger an immediate
warning if a set of criteria is met. For tsunamis, the warning
may be several hours in advance, which is enough time for
preparations. This is because tsunamis travel more slowly
than seismic waves. A P wave travels from Alaska to Hawaii
in about 7 minutes, whereas a tsunami traveling at about
800 km/hr across the ocean takes 5.5 hours. After the damage
done to Hilo by the 1946 Alaska earthquake, the Seismic Sea
Wave Warning System was organized for countries that rim the
Pacific Ocean. Information from seismometers and tide gauges
was phoned to the Tsunami Warning Center in Honolulu,
Hawaii, which issued tsunami alerts if necessary. 12 Tsunami
warning systems have since become more automated, using
real-time digital seismic data to locate large earthquakes and
derive information about their magnitudes, depths, and focal
mechanisms. An assessment can be made of the likelihood of
a tsunami, which usually results from vertical motion at the sea
floor.
The situation is much more complicated with seismic waves.
Although local seismic networks can automatically and immediately locate an earthquake and assess if it is hazardous, the
warning time is short. For example, a warning after a major
earthquake on the New Madrid fault system instantly relayed
via Internet or radio to St Louis would arrive about 40 seconds
before the first seismic waves. Seismologists, engineers, and
public authorities are thus discussing what might be done with
such short warning times. Although such times would not permit evacuations, certain steps might be useful. For example,
real-time warnings are used in Japan to stop high-speed trains,
and it may be practical to have gas line shut-off valves or other
automatic responses connected to such a system. The questions
are whether the improved safety justifies the cost and whether
the risk of false alarms is serious.
A related approach is to provide authorities with nearreal-time information, including data on the distribution of
shaking, immediately after major earthquakes. Seismic networks are working to provide emergency management services
with information that can help direct the needed response to
the most affected areas during the chaotic few hours after a
large earthquake, when the location and extent of damage are
often still unclear.
1.2.8 Nuclear monitoring and treaty verification
Another important societal application of seismology is the
monitoring of nuclear testing. Although atomic physics destabilized world politics through the invention of the atomic bomb,
seismology has partially restabilized it. Throughout the cold
war between the USA and the Soviet Union, seismology helped
verify that treaties were being observed.
The role of seismology in nuclear monitoring began in 1957
when the USA detonated RAINIER, the first underground
nuclear explosion. By the early 1960s it became clear that
radioactive elements produced by atmospheric nuclear testing
posed significant health threats. In 1963, 116 nations signed
the Limited Test Ban Treaty, which banned nuclear testing
in the atmosphere, in the oceans, and in space, and required
testing to occur underground. At about this time, the US Air
Force helped fund the deployment of the World Wide Standardized Seismographic Network (WWSSN). WWSSN stations
provided important information for monitoring nuclear testing
and a wealth of data that played a major role in modern geophysical seismology.
In 1976, countries began to abide by the Threshold Test Ban
Treaty, which limited the size of underground nuclear tests to
150 kt (equivalent to 150 kilotons of TNT). Before then, the
largest atmospheric test had been 58 Mt, and the largest underground test had been 4.4 Mt. Figure 1.2-18 shows the yields estimated seismologically for underground nuclear tests carried
out by the Soviet Union. Although it was initially thought that
some of the post-1976 explosions were greater than 150 kt,
this turned out to reflect the different geologies of the western
USA and central Asia. The conversion of seismic body wave
magnitude m b values into TNT yields was calibrated using the
Nevada test site, but the western US crust is more seismically
attenuating than the more stable Soviet sites in Kazakhstan
and Novaya Zemlya (see Section 3.7.10). The yields of
explosions in kilotons, Y, can be related to the observed seismic
magnitudes by
12 Serious or older television viewers may recall the episode of Hawaii 5-0 in which
criminals force the center to issue a spurious tsunami warning to prompt evacuation
of downtown Honolulu and facilitate a robbery.
damaged more than 90% of the houses. The prediction is said
to have been based on precursors, including ground deformation, changes in the electromagnetic field and groundwater
levels, anomalous animal behavior, and significant foreshocks.
However, in the following year, the Tangshan earthquake
occurred not too far away without precursors. In minutes,
250,000 people died, and another 500,000 people were
injured. In the following month, an earthquake warning in the
Kwangtung province caused people to sleep in tents for two
months, but no earthquake occurred. Because foreign scientists have not yet been able to assess the Chinese data and the
record of predictions, including both false positives (predictions without earthquakes) and false negatives (earthquakes
without predictions), it is difficult to evaluate the program.
In summary, despite tantalizing suggestions, at present there
is still an absence of reliable precursors. The frustrations of this
search have led to the wry observation that “it is difficult
to predict earthquakes, especially before they happen.” Most
researchers thus feel that although earthquake prediction
would be seismology’s greatest triumph, it is either far away
or will never happen. However, because success would be of
enormous societal benefit, the search for methods of earthquake prediction will likely continue.
1.2.7 Real-time warnings
Some recent efforts are directed to the tractable goal of realtime warnings, where seismometers trigger an immediate
warning if a set of criteria is met. For tsunamis, the warning
may be several hours in advance, which is enough time for
preparations. This is because tsunamis travel more slowly
than seismic waves. A P wave travels from Alaska to Hawaii
in about 7 minutes, whereas a tsunami traveling at about
800 km/hr across the ocean takes 5.5 hours. After the damage
done to Hilo by the 1946 Alaska earthquake, the Seismic Sea
Wave Warning System was organized for countries that rim the
Pacific Ocean. Information from seismometers and tide gauges
was phoned to the Tsunami Warning Center in Honolulu,
Hawaii, which issued tsunami alerts if necessary. 12 Tsunami
warning systems have since become more automated, using
real-time digital seismic data to locate large earthquakes and
derive information about their magnitudes, depths, and focal
mechanisms. An assessment can be made of the likelihood of
a tsunami, which usually results from vertical motion at the sea
floor.
The situation is much more complicated with seismic waves.
Although local seismic networks can automatically and immediately locate an earthquake and assess if it is hazardous, the
warning time is short. For example, a warning after a major
earthquake on the New Madrid fault system instantly relayed
via Internet or radio to St Louis would arrive about 40 seconds
before the first seismic waves. Seismologists, engineers, and
public authorities are thus discussing what might be done with
such short warning times. Although such times would not permit evacuations, certain steps might be useful. For example,
real-time warnings are used in Japan to stop high-speed trains,
and it may be practical to have gas line shut-off valves or other
automatic responses connected to such a system. The questions
are whether the improved safety justifies the cost and whether
the risk of false alarms is serious.
A related approach is to provide authorities with nearreal-time information, including data on the distribution of
shaking, immediately after major earthquakes. Seismic networks are working to provide emergency management services
with information that can help direct the needed response to
the most affected areas during the chaotic few hours after a
large earthquake, when the location and extent of damage are
often still unclear.
1.2.8 Nuclear monitoring and treaty verification
Another important societal application of seismology is the
monitoring of nuclear testing. Although atomic physics destabilized world politics through the invention of the atomic bomb,
seismology has partially restabilized it. Throughout the cold
war between the USA and the Soviet Union, seismology helped
verify that treaties were being observed.
The role of seismology in nuclear monitoring began in 1957
when the USA detonated RAINIER, the first underground
nuclear explosion. By the early 1960s it became clear that
radioactive elements produced by atmospheric nuclear testing
posed significant health threats. In 1963, 116 nations signed
the Limited Test Ban Treaty, which banned nuclear testing
in the atmosphere, in the oceans, and in space, and required
testing to occur underground. At about this time, the US Air
Force helped fund the deployment of the World Wide Standardized Seismographic Network (WWSSN). WWSSN stations
provided important information for monitoring nuclear testing
and a wealth of data that played a major role in modern geophysical seismology.
In 1976, countries began to abide by the Threshold Test Ban
Treaty, which limited the size of underground nuclear tests to
150 kt (equivalent to 150 kilotons of TNT). Before then, the
largest atmospheric test had been 58 Mt, and the largest underground test had been 4.4 Mt. Figure 1.2-18 shows the yields estimated seismologically for underground nuclear tests carried
out by the Soviet Union. Although it was initially thought that
some of the post-1976 explosions were greater than 150 kt,
this turned out to reflect the different geologies of the western
USA and central Asia. The conversion of seismic body wave
magnitude m b values into TNT yields was calibrated using the
Nevada test site, but the western US crust is more seismically
attenuating than the more stable Soviet sites in Kazakhstan
and Novaya Zemlya (see Section 3.7.10). The yields of
explosions in kilotons, Y, can be related to the observed seismic
magnitudes by
12 Serious or older television viewers may recall the episode of Hawaii 5-0 in which
criminals force the center to issue a spurious tsunami warning to prompt evacuation
of downtown Honolulu and facilitate a robbery.