Geoscience Reference
In-Depth Information
(Obara, 2002). Tremor consists of long-duration trains of weak ground motions that
do not have easily identifiable body wave arrivals. Soon after slow slip events were
discovered, it was shown that the slow slip episodes in Cascadia correlate with
periods of enhanced seismic tremor, with the term episodic tremor and slip (ETS)
being used to describe the combined phenomena (Rogers and Dragert, 2003). Efforts
to establish the direct relationship between tremor and slow slip are under way, with
possibilities including heterogeneous frictional conditions on the deep megathrust as
well as activations of multiple faults due to fluid motions and changes in strain
produced by the slow slip. ETS has now been reported in many other subduction
zones but with variable manifestations. In southern Mexico the slip events are as
much as five times larger than in Cascadia and less frequent (Kostoglodov et al.,
2003); tremor has also been observed in the Mexican subduction zone (Payero et al.,
2008). Episodic slip events have been reported in the New Zealand Hikurangi
subduction zone (Douglas et al., 2005; Wallace and Beavan, 2006), while tremor was
elusive (Delahaye et al., 2009) until recently observed (Kim et al., 2011). The
diversity of fault slip processes has been particularly well documented in Japan,
where there is high density of both geodetic and seismic instrumentation. For
example, borehole tiltmeters in the Nankai Trough have been used to detect slip
events that were much too small to be indentified on GPS receivers. The migration of
both tremor and slip on the fault zone interface was subsequently imaged, with
migration speeds of ~10 km per day (Obara et al., 2004), comparable to observations
in Cascadia. Earthquakes depleted in short-period radiation (low-frequency
earthquakes, or LFEs) have been identified near the down-dip edge of the unstable
megathrust zone (
Katsumata and Kamaya, 2003
)
, and it currently appears that tremor
involves superposition of many small LFEs (or even normal earthquakes).
Conventional earthquakes involve large amounts of energy release in small
amounts of time, with rupture spreading over the fault at very high velocities of
several kilometers per second (see Figure 2.11). The ETS and LFE observations make
it clear that fault slip occurs on an immense variety of temporal scales that appear to
scale differently than for fast ruptures. Some of the large slow slip events have the
equivalent strain release of large conventional earthquakes (e.g., Kostoglodov et al.,
2003). In some regions, where the total seismic slip budget falls very short of the total
plate tectonic convergence budget, such as the Marianas and Tonga subduction zones,
the entire megathrust may be failing in slow slip or stable sliding processes, as
appears to be the case along North Island, New Zealand. This indicates the
importance of understanding the full range of frictional processes that appear to play
a huge role in plate motions. The simple earthquake cycle model that has been
invoked for decades needs to be expanded to accommodate these new observations.
Much of the effort thus far has focused on categorizing these events—where and
when they occur and how big they are. And while slow slip appears to be related to a
frictional behavior intermediate between that of steady sliding and stick-slip
earthquakes, theoretical developments are needed in order to make advances in our
understanding of these new observations of fault slip. Laboratory studies of rock
mechanics spanning the full range of fault slip velocities play a key role in
quantifying the observations.
Search WWH ::
Custom Search