334 Seismology and Plate Tectonics
the Gulf of Aden or the Red Sea. With time, the ocean widens
and deepens due to thermal subsidence of oceanic lithosphere
(Section 5.3.2), and thick sediments accumulate on the continental margins, such as those on either side of the Atlantic.
These margins are not plate boundaries a the oceanic and
continental crust on the two sides are on the same plate a and
are called passive margins, to distinguish them from active continental margins, which are plate boundaries. Subduction often
begins along one of the passive margins, and the ocean basin
closes, such that magmatism and mountain building occur, as
along the west coast of South America today. Continental collision like that currently in the Himalayas occurs eventually, and
the mountain building process reaches its climax. If the continental materials on either side cease to move relative to each
other, this process leaves a mountain belt within the interior
of a single plate. At some future time, however, a new rifting
phase can begin, often near the site of the earlier rifting, and a
new ocean will start to grow. Thus the Appalachian Mountains
record a continental collision that closed an earlier Atlantic
Ocean about 270 million years ago, and remain despite the
opening of the present Atlantic Ocean during the past 200 Myr.
As a result, continental and oceanic crust have very different
life cycles. Because the relatively less dense continental crust is
not subducted, the continents have accreted over a much longer
time than the 200-million-year age of the oldest oceanic crust.
Hence the continents preserve a complex set of geologic structures, many of which can be sites of deformation, including
earthquake faulting. Thus both plate boundary and intraplate
deformation zones within continents are more complex than
their oceanic counterparts.
Earth scientists seek to understand the continental evolution
process for both intellectual and practical reasons. The process
is fundamental to how the planet works, but also provides
information about geologic hazards (earthquakes, volcanism,
uplift, and erosion) and mineral resources. In addition, the
large mountain belts have major impacts on earth’s climate.
Seismology contributes to these studies by providing data
about earthquakes and velocity structure in regions where different parts of the evolutionary cycle occur today or occurred
in the past. These data are combined with other geophysical
and geological data to form an integrated picture of the complicated continental evolution processes. Hence, although the
processes are not fully understood, important progress continues to be made.
5.6.1 Continental plate boundary zones
As for oceanic boundaries, we seek to first describe the motion
(kinematics) within boundary zones, and then to combine
the kinematics with other data to investigate their mechanics
(dynamics). One example is the East African rift (Fig. 5.6-2),
a spreading center between the Nubian (West Africa) and
Somalian (East Africa) plates. The extension rate is so slow,
less than 10 mm/yr, that it is hard to resolve in plate motion
models, and the two plates are often treated as one (Fig. 5.2-4).
However, the rift topography, normal faulting, and seismicity
distribution show the presence of an extensional boundary
zone broader, more diffuse, and more complex than at a
mid-ocean ridge. For example, the seismicity ends in southern
Africa and has no clear connection to the southwest Indian
ridge, where the plate boundary must go. A recent estimate is
that the northern East Africa rift opens at about 6 mm/yr,
whereas the southern part opens at about half that, because the
Euler pole is to the south. Some of the complexity of such
continental extensional zones results from the fact that, unlike
a mid-ocean ridge, the lithosphere starts off with reasonable
thickness and then is stretched and thinned in the extending
zone. The rifting process can eventually progress far enough
that a new oceanic spreading center forms. This has already
occurred in the Gulf of Aden and the Red Sea, which are newly
formed (and hence narrow) oceans separating the Arabian
plate from Somalia and Nubia at rates of about 22 and 16
mm/yr, respectively. Whether the East African rift will evolve
this far is still unclear, because the geologic record shows many
rifts that, although active for some time, failed to develop into
oceanic spreading centers and simply died. As we will see, these
fossil rifts can be loci for intraplate earthquakes.
The earthquakes also indicate that the thermal and mechanical structure of continental rifts is more complicated than
on mid-ocean ridges. Normal-faulting earthquakes extend to
depths of 25–30 km, considerably deeper than at mid-ocean
ridges. Hence the lower crust appears to be surprisingly
stronger and colder than might be expected in an active rift.
Continental transforms are also more complicated than their
oceanic counterparts. As we saw in Section 5.2, the transform
portion of the Pacific–North America plate boundary in western North America is an active seismic zone hundreds of
kilometers wide (Fig. 5.2-3), in contrast to widths of less than
10 km for oceanic transforms. Thus the focal mechanisms
show primarily strike-slip motion on the San Andreas fault
itself and demonstrate complexities including thrust faulting
for events like the 1971 San Fernando and 1994 Northridge
earthquakes and normal faulting due to the regional extension
in the Basin and Range province. The earthquakes and spacegeodetic data show that although most of the motion occurs
along the San Andreas (Fig. 4.5-13) and nearby faults, a reasonable fraction of the motion occurs elsewhere (Figs. 5.6-3 and
5.2-3). The boundary zone is further complicated by volcanism
in areas including the Long Valley caldera in eastern California
and the Yellowstone hot spot, which also have associated
seismicity. Hence, we think of a boundary zone in which the
overall steady motion between the plate interiors is distributed
in both space and time (Fig. 5.6-4). Although much of the
motion occurs in occasional large earthquakes or steady creep
on the main boundary segment, some deformation occurs elsewhere in the zone.
The breadth of continental plate boundary zones has important implications for seismic hazards within them. Because
ground shaking decays rapidly with distance (Fig. 1.2-5),
nearby smaller earthquakes within a boundary zone, but not
the Gulf of Aden or the Red Sea. With time, the ocean widens
and deepens due to thermal subsidence of oceanic lithosphere
(Section 5.3.2), and thick sediments accumulate on the continental margins, such as those on either side of the Atlantic.
These margins are not plate boundaries a the oceanic and
continental crust on the two sides are on the same plate a and
are called passive margins, to distinguish them from active continental margins, which are plate boundaries. Subduction often
begins along one of the passive margins, and the ocean basin
closes, such that magmatism and mountain building occur, as
along the west coast of South America today. Continental collision like that currently in the Himalayas occurs eventually, and
the mountain building process reaches its climax. If the continental materials on either side cease to move relative to each
other, this process leaves a mountain belt within the interior
of a single plate. At some future time, however, a new rifting
phase can begin, often near the site of the earlier rifting, and a
new ocean will start to grow. Thus the Appalachian Mountains
record a continental collision that closed an earlier Atlantic
Ocean about 270 million years ago, and remain despite the
opening of the present Atlantic Ocean during the past 200 Myr.
As a result, continental and oceanic crust have very different
life cycles. Because the relatively less dense continental crust is
not subducted, the continents have accreted over a much longer
time than the 200-million-year age of the oldest oceanic crust.
Hence the continents preserve a complex set of geologic structures, many of which can be sites of deformation, including
earthquake faulting. Thus both plate boundary and intraplate
deformation zones within continents are more complex than
their oceanic counterparts.
Earth scientists seek to understand the continental evolution
process for both intellectual and practical reasons. The process
is fundamental to how the planet works, but also provides
information about geologic hazards (earthquakes, volcanism,
uplift, and erosion) and mineral resources. In addition, the
large mountain belts have major impacts on earth’s climate.
Seismology contributes to these studies by providing data
about earthquakes and velocity structure in regions where different parts of the evolutionary cycle occur today or occurred
in the past. These data are combined with other geophysical
and geological data to form an integrated picture of the complicated continental evolution processes. Hence, although the
processes are not fully understood, important progress continues to be made.
5.6.1 Continental plate boundary zones
As for oceanic boundaries, we seek to first describe the motion
(kinematics) within boundary zones, and then to combine
the kinematics with other data to investigate their mechanics
(dynamics). One example is the East African rift (Fig. 5.6-2),
a spreading center between the Nubian (West Africa) and
Somalian (East Africa) plates. The extension rate is so slow,
less than 10 mm/yr, that it is hard to resolve in plate motion
models, and the two plates are often treated as one (Fig. 5.2-4).
However, the rift topography, normal faulting, and seismicity
distribution show the presence of an extensional boundary
zone broader, more diffuse, and more complex than at a
mid-ocean ridge. For example, the seismicity ends in southern
Africa and has no clear connection to the southwest Indian
ridge, where the plate boundary must go. A recent estimate is
that the northern East Africa rift opens at about 6 mm/yr,
whereas the southern part opens at about half that, because the
Euler pole is to the south. Some of the complexity of such
continental extensional zones results from the fact that, unlike
a mid-ocean ridge, the lithosphere starts off with reasonable
thickness and then is stretched and thinned in the extending
zone. The rifting process can eventually progress far enough
that a new oceanic spreading center forms. This has already
occurred in the Gulf of Aden and the Red Sea, which are newly
formed (and hence narrow) oceans separating the Arabian
plate from Somalia and Nubia at rates of about 22 and 16
mm/yr, respectively. Whether the East African rift will evolve
this far is still unclear, because the geologic record shows many
rifts that, although active for some time, failed to develop into
oceanic spreading centers and simply died. As we will see, these
fossil rifts can be loci for intraplate earthquakes.
The earthquakes also indicate that the thermal and mechanical structure of continental rifts is more complicated than
on mid-ocean ridges. Normal-faulting earthquakes extend to
depths of 25–30 km, considerably deeper than at mid-ocean
ridges. Hence the lower crust appears to be surprisingly
stronger and colder than might be expected in an active rift.
Continental transforms are also more complicated than their
oceanic counterparts. As we saw in Section 5.2, the transform
portion of the Pacific–North America plate boundary in western North America is an active seismic zone hundreds of
kilometers wide (Fig. 5.2-3), in contrast to widths of less than
10 km for oceanic transforms. Thus the focal mechanisms
show primarily strike-slip motion on the San Andreas fault
itself and demonstrate complexities including thrust faulting
for events like the 1971 San Fernando and 1994 Northridge
earthquakes and normal faulting due to the regional extension
in the Basin and Range province. The earthquakes and spacegeodetic data show that although most of the motion occurs
along the San Andreas (Fig. 4.5-13) and nearby faults, a reasonable fraction of the motion occurs elsewhere (Figs. 5.6-3 and
5.2-3). The boundary zone is further complicated by volcanism
in areas including the Long Valley caldera in eastern California
and the Yellowstone hot spot, which also have associated
seismicity. Hence, we think of a boundary zone in which the
overall steady motion between the plate interiors is distributed
in both space and time (Fig. 5.6-4). Although much of the
motion occurs in occasional large earthquakes or steady creep
on the main boundary segment, some deformation occurs elsewhere in the zone.
The breadth of continental plate boundary zones has important implications for seismic hazards within them. Because
ground shaking decays rapidly with distance (Fig. 1.2-5),
nearby smaller earthquakes within a boundary zone, but not