N
R
L
N
Rayleigh
N
Love
155°W
150°
145°
62°N
60°
58°
56°
Aftershock
area
Epicenter
Anchorage
R u p tu re
T r e n c h
a x i s
G
u i f o f A l a s k a
100 km
N
Point source
Data and finite source
4.4 Moment tensors 239
5 Some of the earthquake damage is shown in Fig. 1.2-11.
Fig. 4.3-15 Focal mechanism for the great
1964 Alaska earthquake, and the surface
wave radiation patterns it predicts if the
source is treated as a point (top left). Love
and Rayleigh waves are shown as solid and
dashed lines, respectively. The observed
patterns (jagged lines) are quite different,
but are reasonably consistent with those
predicted by a finite source propagating
southwestward along the 600 km-long
fault plane, consistent with the large
aftershock area (bottom). (Kanamori, 1970b.
J. Geophys. Res., 75, 5029–40, copyright by
the American Geophysical Union.)
amplitude radiation patterns are quite different, and modeling
shows them to be consistent with rupture propagating southwestward along the 600 km-long fault plane. This dimension is
consistent with the large aftershock area, and together with the
seismic moment (Section 4.6) implies an average fault slip of
about 7 meters, bearing out the gigantic nature of the earthquake. 5 In fact, postseismic deformation is still observed with
geodetic data (Fig. 4.5-15).
4.3.5 Once and future earthquakes
Combining body and surface wave modeling with first motions
is often valuable for studying seismograms from older earthquakes. This application arises often in tectonic studies, because in many cases the largest earthquakes occurred prior to
the development of global seismic networks in the early 1960s
(Section 6.6). Since about 1930, a few stations have reported
first motions to the International Seismological Summary. The
number of points per earthquake is far less than that available for a modern study, and the data from nonstandardized
seismometers are often discordant. However, in some cases
body and/or surface wave modeling is useful, especially if the
first motions constrain at least one nodal plane. One technique
is to use the ratio of Love and Rayleigh wave amplitudes.
This discussion brings out an important difference between
first motion and modeling studies. For first motion studies,
all we need to know about the seismometer is the polarity, so
compressional arrivals are in fact “up” on the seismograms.
However, modeling requires knowing the response of the
instruments. Fortunately, modern instruments are (at least in
theory) standardized, and their calibration can be checked.
This is a problem for studies of older earthquakes, because
calibrations were often quite poor.
In recent years, modeling approaches have become steadily
more powerful. High-quality data from digital broadband
seismometers (Section 6.6) have become standard. In addition,
laterally homogeneous models for seismic velocity and attenuation have been developed and improved. As a result, inversions of body and surface wave data for many earthquakes, as
discussed in the next section, are giving large focal mechanism
datasets for tectonic and earthquake source studies.
4.4 Moment tensors
4.4.1 Equivalent forces
Our approach so far in this chapter has been to view earthquakes as due to slip on a fault and to estimate their source
parameters by forward modeling the radiated seismic waves.
R
L
N
Rayleigh
N
Love
155°W
150°
145°
62°N
60°
58°
56°
Aftershock
area
Epicenter
Anchorage
R u p tu re
T r e n c h
a x i s
G
u i f o f A l a s k a
100 km
N
Point source
Data and finite source
4.4 Moment tensors 239
5 Some of the earthquake damage is shown in Fig. 1.2-11.
Fig. 4.3-15 Focal mechanism for the great
1964 Alaska earthquake, and the surface
wave radiation patterns it predicts if the
source is treated as a point (top left). Love
and Rayleigh waves are shown as solid and
dashed lines, respectively. The observed
patterns (jagged lines) are quite different,
but are reasonably consistent with those
predicted by a finite source propagating
southwestward along the 600 km-long
fault plane, consistent with the large
aftershock area (bottom). (Kanamori, 1970b.
J. Geophys. Res., 75, 5029–40, copyright by
the American Geophysical Union.)
amplitude radiation patterns are quite different, and modeling
shows them to be consistent with rupture propagating southwestward along the 600 km-long fault plane. This dimension is
consistent with the large aftershock area, and together with the
seismic moment (Section 4.6) implies an average fault slip of
about 7 meters, bearing out the gigantic nature of the earthquake. 5 In fact, postseismic deformation is still observed with
geodetic data (Fig. 4.5-15).
4.3.5 Once and future earthquakes
Combining body and surface wave modeling with first motions
is often valuable for studying seismograms from older earthquakes. This application arises often in tectonic studies, because in many cases the largest earthquakes occurred prior to
the development of global seismic networks in the early 1960s
(Section 6.6). Since about 1930, a few stations have reported
first motions to the International Seismological Summary. The
number of points per earthquake is far less than that available for a modern study, and the data from nonstandardized
seismometers are often discordant. However, in some cases
body and/or surface wave modeling is useful, especially if the
first motions constrain at least one nodal plane. One technique
is to use the ratio of Love and Rayleigh wave amplitudes.
This discussion brings out an important difference between
first motion and modeling studies. For first motion studies,
all we need to know about the seismometer is the polarity, so
compressional arrivals are in fact “up” on the seismograms.
However, modeling requires knowing the response of the
instruments. Fortunately, modern instruments are (at least in
theory) standardized, and their calibration can be checked.
This is a problem for studies of older earthquakes, because
calibrations were often quite poor.
In recent years, modeling approaches have become steadily
more powerful. High-quality data from digital broadband
seismometers (Section 6.6) have become standard. In addition,
laterally homogeneous models for seismic velocity and attenuation have been developed and improved. As a result, inversions of body and surface wave data for many earthquakes, as
discussed in the next section, are giving large focal mechanism
datasets for tectonic and earthquake source studies.
4.4 Moment tensors
4.4.1 Equivalent forces
Our approach so far in this chapter has been to view earthquakes as due to slip on a fault and to estimate their source
parameters by forward modeling the radiated seismic waves.