S13A-01 INVITED 13:30h
Episodic Tremor and Slip in the Explorer Plate Region Beneath Northern Vancouver Island, Canada
In the Cascadia Subduction zone beneath southern Vancouver Island, distinct non-earthquake tremor-like seismic signals have been found to correlate temporally and spatially with deep slow-slip events on the subduction zone. This slip is characterized by the systematic displacement of continuous GPS sites over a period of one or two weeks in a direction opposite to the longer-term elastic strain accumulation, and a surprisingly regular repeat period of roughly 14 months. Tremors appear as low-frequency (0.4 to 4 Hz) "noise" with bursts that correlate from station to station. We refer to this associated tremor and slip phenomena as "Episodic Tremor and Slip" or ETS activity. Using continuous seismic data spanning the past 7 years, tremors have now been identified beneath northern Vancouver Island in the region of the Explorer Plate. For the data examined, most months contained some tremor activity. Periods of significant tremor activity lasting 5 to 30 days were few but accounted for two-thirds of all tremor activity. This is similar to the tremor activity to the south. Although the accompanying surface displacements have not been clearly mapped for these northernmost significant tremor sequences, their seismic aspects are similar to those of tremors observed to the south. The source region is on or above the subducting plate in the depth range of 20 to 40 km, and the recurrence interval is also about 14 month but clearly about 6 months out of phase with the ETS activity in the region of southern Vancouver Island and the Olympic Peninsula. The northern extent of tremor activity appears to coincide with the northern edge of the Explorer plate and most of the strong tremor sequences seem to initiate here and migrate to the southeast at a rate of about 10 km per day. This migration does not appear to be impeded by the presence of the Nootka Fault Zone (NFZ), the boundary between the Explorer and Juan de Fuca plates. The total amount of tremor activity in northern and southern Vancouver Island appears to be the same although the plate convergence rate of the Explorer Plate with North America is only 2 cm/yr, approximately half of the Juan de Fuca rate. This suggests that tremor activity is independent of plate convergence rate. Preliminary measurements at two newer GPS sites near the NFZ show predominantly eastward displacements of 3 to 4 mm associated with the two most recent north-island tremor sequences. A greater density of continuous GPS measurements is required before meaningful slip-dislocation models can be used to determine slip magnitudes which are expected to be about one half of average slip displacements in the south if the process is related to the relief of stress due to plate convergence.
S13A-02 13:50h
Deep low-frequency tremors and short-term slow slip events in the southwest Japan subduction zone
Slow slip events accompanied with major activities of deep low-frequency tremors have been frequently detected in some places along the belt like distribution of deep tremors in the southwest Japan subduction zone by monitoring of crustal tilt movement. All slip events represented as step-like tilt changes occur periodically coincident temporally and spatially with the tremor activities. The tilt step is not instantaneous, but the duration of each episode is ranging from a few days to a week. The recurrence interval of the episode is different in some areas. In the western Shikoku area, southwest Japan, the short-term slow slip events occur with a recurrence rate of approximately 6 months correlated with migrating major tremor activities for 2 years, 2001 and 2002. The direction of the tilt vector changes according to the migration of the tremor during each episode. This indicates that the source of the slow slip also propagates with tremor. In the eastern Shikoku, the step-like tilt change accompanied with the tremor activity repeats with a recurrence interval of a few months. In some cases, the periodic tremor doesn_ft accompany with the tilt step. Recently, the short-term slow slip event has been recognized in the northern Kii peninsula. The recurrence rate of the episodic slow slip and tremor is approximately 6 months. Around the both edges of tremor source area, long-term slow slip events have been detected. On the other hand, in the Bungo channel between Kyushu and Shikoku Island, a crustal deformation detected by tiltmeters and GPS lasted for a few months accompanied with a series of tremor activity from August to November, 2003. However, in Tokai area, a slow slip event continues for a few years from 2000 and there is no clear correlation temporally with the tremor activity. The relationship between the slow slip events and the tremor activity indicate regional dependences, which might reflect the differences of frictional properties.
S13A-03 14:05h
Repeating Short- and Long-Term Slow Slip Events With Deep Tremor Activity in Southwest Japan Subduction Zone
Non-volcanic deep low-frequency tremors are observed at the Nankai trough subduction zone, southwest Japan (Obara,~2002). This observation is thought to be a manifestation of dehydration process in the subducted slab. This plays an important role in exploring not only water circulation and migration of materials during plate subduction, but also an effect on a shallower great earthquake generation process. Recently, Obara~and Hirose~(2003) reported that crustal tilt change which correlates with the deep tremor activity is observed by high sensitivity accelerometers (tiltmeters) equipped with NIED Hi-net (National Research Institute for Earth Science and Disaster Prevention, Japan; High sensitivity seismograph network). This deformation can be interpreted as a slow slip event (SSE) on the plate interface at depth. In western Shikoku, this activity repeats every six months. Each episode has a duration of about five days and a corresponding moment magnitude of 6.0. The estimated depth is 30--40~kilometers so that this activity may occur at transition zone of interplate coupling. We call this phenomenon `short-term SSE.' We have another evidence of `long-term SSE' which occurred beneath Bungo channel, adjacent to the western Shikoku, where the short-term SSE with tremors repeatedly occurs. In 1997, a SSE was detected by a GSI's (Geographical Survey Institute, Japan) nationwide GPS network (Hirose et al., 1999). This event lasted for about one year. The similar crustal deformation reappeared in 2003 (GSI, 2003). This deformation is observed by both GPS and the tiltmeters. The tilt change begins from late August and lasts for about three months. The maximum tilt change is about 0.6~$\mu$rad. The surface displacement also shows a large and relatively fast change which begins from late August, though very small, relatively slow change in displacement seems to start on March or April, 2003. Both the observed GPS time series and the surface displacement field for both events in 1997 and 2003 are quite similar to each other. This evidence suggests that the same source area slips in these two episodes. We have two smaller short-term SSEs in November, 2003 and February, 2004. Their source areas are adjacent to each other. These two areas cover most of the source area of single event with larger moment release, which occurred in 2001 and 2002. This indicate that several patches, where slip is always slow and transient, may exist on the plate interface. We call this patch `SSE patch.' Each SSE patch has its own slip features, such as duration and magnitude of the slip. A long-term SSE patch may exist in the Bungo channel region, and several short-term SSE patches may exist along a belt-like zone of tremor sources.
S13A-04 INVITED 14:20h
On the Physical Mechanisms of Short- and Long-Interval Silent Slip Events in Deeper Subduction Interfaces
Recent high resolution observations of crustal movements have revealed the occurrences of silent slip events along the deeper part of the subduction zone. There are two types of silent slip events: short- and long-interval silent slip eve,Zts. Silent slip events occur repeatedly with an average interval of 1.2 years on the Cascadia subduction zone (Dragert et al., 2001) and several months in southwest Japan (Obara and Hirose, 2004). These silent slip events are accompanied by activities of low-frequency tremors (Obara, 2002; Dragert et al., 2004). On the other hand, silent slip events occurring in the Tokai region (Ozawa et al., 2002) and beneath the Bungo Channel have much longer durations and intervals. We investigated the mechanisms of these short- and long-interval silent slip events. One possible mechanism of silent slip events is caused by frictional behavior around the unstable-stable transition. Shimamto (1986) investigated the frictional behavior using halite around the transition zone and found that steady state friction decreases with slip velocity at low slip velocity and increases with slip velocity at high slip velocity. Using the rate- and state-dependent friction law which exhibits velocity weakening at low slip velocity and velocity strengthening at high slip velocity, we simulated episodic silent slip events at the deeper part of the seismogenic zone. Intervals and durations of silent slip events increase with the critical weakening displacement. The maximum slip velocity of these events is around 10$^{-9}$ to 10$^{-7}$ m/s. Long-interval silent slip events can be explained by this mechanism. There is a possibility that short-interval silent slip events are caused by high pore-fluid pressure in subduction interfaces since these events are accompanied by low-frequency tremors. We developed a model of short-interval silent slip events considering frictional dilatancy due to slip and compaction at the deeper frictional stable region. In our model, pore-fluid pressure obeys a diffusion type equation with a source tern due to frictional dilatancy and compaction. Porosity decreases at very low slip velocity due to compaction. As a result, pore-fluid pressure increases. When pore-fluid pressure approaches lithostatic pressure, slip starts to accelerate. With an increase of slip velocity, porosity increases due to frictional dilatancy. As a result, pore-fluid pressure decreases. This decrease in pore-fluid pressure results in strengthening the fault and deceleration of slip. We found that short-interval silent slip events can occur only at the condition in which pore fluid pressure is very close to lithostatic pressure. The intervals of these silent slip events are from a few months to a few years. The maximum slip velocity of silent slip events is around 10$^{-9}$ to 10$^{-7}$ m/s. One important parameter is the dilatancy coefficient which determines the amount of frictional dilatancy. With an increase of dilatancy coefficient, changes in pore fluid pressure and changes in the fault strength become larger and the intervals of silent slip events become longer. Approaching the main event, the intervals of silent slip events become shorter. This numerical result suggests that monitoring of silent slip events will be very important for forecasting the main earthquake.
S13A-05 INVITED 14:40h
Deep-low Frequency Earthquakes on the Downward Extension of the Seismogenic Faults
In around the world, deep low-frequency (DLF) earthquakes have been usually observed near active volcanoes or the volcanic front and attributed to magma activity. However, there are several examples of DLF activity beneath active faults in the backarc region without volcanoes in the vicinity. Examples in southwest Japan are in western Tottori prefecture, mid Kyoto prefecture, and some other regions. These events are probably another example of DLF earthquakes that occur beneath active fault zones. In this paper, we will focus on the DLF activity in the Western Tottori district, southwest Japan. On October 6, 2000, a Mw6.7 crustal earthquake occurred in western Tottori prefecture, southwest Japan. Beneath the focal region of the earthquake, DLF earthquakes were observed at depths of around 30 km. Five DLF earthquakes were detected within 3 years before the mainshock and the activity was enhanced after the mainshock. Observed features in the waveforms of these DLF events are as follows; (1) Predominant frequency is 2 Hz - 4 Hz, however they are not monochromatic events, (2) Both P- and S- waves are observed however S-waves have larger amplitudes, and (3) The onset of the P-waves sometimes have a high-frequency component. Duration of these events are one minute or longer, and in some cases they exhibit durations of several minutes. Ohmi and Obara (2002) analyzed the focal mechanism of the DLF earthquake that occurred 9 hours before the mainshock (M0.8), using amplitude ratios of the S-waves to the P-waves and polarization patterns of the S-waves. Both analyses indicated that a single-force source mechanism is more preferable than a double-couple source mechanism, which suggests the transport of fluid such as water or magma. More than 400 DLF earthquakes, whose magnitudes are mostly up to 1.5 or so, were observed during the 3 and half years after the mainshock. Since May 2002, several larger DLF earthquakes, whose amplitude magnitudes are more than 2.0, have been observed. They exhibit clear longer P-wave onset pulses than those of previous events. It indicates they have rather longer source time duration than those expected from ordinary scaling law between source dimension and seismic moment, which shows that source mechanisms of these DLF events are quite different from those of ordinary tectonic earthquakes. In April 2003, an intense activity of low-frequency events was observed in the region. They exhibit continuous tremor-like feature of several tens of minutes. This activity lasted for about three days. However, no associated crustal deformation was detected by the tiltmeters installed in the focal region. Seismic tomography analysis (e.g Zhao et al., 2003) indicates the existence of a low-velocity body at depths from the lower-crust to the upper-mantle in the focal region of the DLF_fs in the western Tottori region. The resistivity structure (e.g. Oshiman et al., 2003) exhibit low resistivity portion that also indicates the fluid activity in the focal region. The DLF_fs in the western Tottori are probably direct evidence of fluid activity in the seismogenic zone. Recent studies (e.g. Iio and Kobayashi, 2002) indicate that the seismogenic faults have downward extension in the lower crust, whose aseismic slip accumulate stress on the seismogenic faults. It is important to understand the nature of DLF earthquakes beneath active faults, in relation to the behavior of fluids in the lower crust that might affect the aseismic slip of the downward extension of the seismogenic faults.
S13A-06 15:00h
Seismicity of deep and intermediate-depth low-frequency earthquakes in the crust beneath the Towada volcano, northeast Japan
We have investigated seismic activities of high- and low-frequency earthquakes beneath the Towada volcano in northeast Japan. The most significant feature of crustal seismicity in the area is the occurrence of both deep (depth ~ 25 km) and intermediate-depth (depth ~ 10 km) low-frequency (LF) microearthquakes. Though similar deep LF earthquakes occur mainly beneath volcanic areas in the northeast Japan, the intermediate LF earthquakes are rare in the region. The waveform characteristics of intermediate LF events are common to the deep LF events; they are characterized by the predominant frequencies of 2-5 Hz, the dominance of S-wave, and long duration compared with the ordinary high-frequency (HF) earthquakes with similar magnitudes. Another notable feature of seismicity is the separation of focal areas among the deep and intermediate LF events and the HF earthquakes; the deep and intermediate LF earthquakes share the same epicentral area but have different focal depths, while the intermediate LF and HF events form separate clusters in almost the same depth range. Peculiar tremor-like signals with duration of more than 10 minutes were also observed at several stations around the volcano. The signals consist of wave packets in the frequency range of 2-4 Hz and duration of about 5 s but with variable amplitude. Judging from the similarity of both locations and particle motions with the intermediate LF events, we interpret these signals are the succession of LF earthquakes in the mid-crust. We also examined velocity structure in the focal area by utilizing receiver functions at several stations around the volcano. A high-velocity anomaly at depths about 35 km appears at all stations and probably corresponds to velocity increase at Moho. On the other hand, low-velocity anomalies appear at depths around 27 and 12 km only at the closest station to the source of LF earthquakes. These depths correspond to the focal depths of deep and intermediate LF earthquakes, which suggest that the LF events occur at the top of local low-velocity anomlies in the lower and middle crust of the area.