U32A-01 INVITED
Triggered Tremor as a Slow Slip Meter?
A growing number of observations show that seismic waves trigger tremor. These observations are providing key constraints on mechanisms of tremor generation. Triggered tremor refers to bursts of seismic energy with the same characteristics as ambient tremor but are temporally modulated by the triggering seismic waves. We hypothesize that triggered tremor serves as an indicator of slow slip. If correct, triggered tremor would be a useful tool for detecting slow slip in the many regions with triggered seismic stations but lacking geodetic and continuous seismic monitoring. This hypothesis is premised on the growing observations of correlated rates of ambient tremor activity and slow slip. Thus, if the probability of triggering tremor depends on the ambient tremor rate, we can infer that it also depends on the amplitude or rate of slow slip. We test the assumption that the probability of triggering tremor depends on the ambient tremor rate by employing ideas used to study earthquake triggering. We measure tremor rates for Cascadia using a new tremor catalog spanning 2007 and 2008, a period containing two episodes of slow slip and increased tremor activity. We also document the characteristics of large teleseismic waves that passed through the region during this period, and which of these triggered tremor. Data used come from the Pacific Northwest Seismic Network and USArray. Preliminary qualitative results suggest the probability of triggering tremor correlates with the ambient tremor rate. To test further the hypothesis that triggered tremor serves as an indicator of slow slip, we perform a blind test, proposing that we should observe triggered tremor where and when slow slip is occurring. We attempt this in the upper Cook Inlet of Alaska, where a Mw7.2 slow slip event occurred between 1998 and 2001, coincident spatially and temporally with the BEAAR broadband seismic deployment. An initial examination reveals no triggered tremor, but the lack of constraint on the evolution of the slow slip warrants a higher resolution temporal sampling of the BEAAR data over the entire time period, which is the focus of ongoing work.
U32A-02
Source Process of the 2007 Boso Slow Slip Event With an Earthquake Swarm From Tiltmeter Data
Around the Boso peninsula (BP), central Japan, the Philippine Sea plate (PHS) is subducting along the Sagami trough. Around the southeast coast of BP, slow slip events (SSE) have been observed with the GEONET (the GPS network operated by GSI) and the NIED tiltmeter network every 6--7 years (Ozawa et al., 2003; NIED 2003). The unique characteristics of the Boso SSEs are that earthquake swarms (ES) have also occurred in association with the SSEs. The latest activity of the SSE and the ES took place in August 2007 (Ozawa et al., 2007). A simple rectangular fault model of the SSE was estimated from static offsets in displacements and tilts (Sekine et al., 2007). The fault model has the shallow northward dipping fault plane consistent with the configuration of the subducted PHS (Mw 6.6). The model illustrates a rough sketch of the overall activity, but the tilt data show sub-day time-resolved deformations associated with the SSE and ES. In this study, we applied a time-dependent inversion procedure for tilt data (Hirose and Obara, 2008) to the Boso SSE tiltmeter data to discuss the detailed source process of the SSE and the relationship between the slow slip and the ES. For a preliminary analysis, an expanded plane fault based on the previous fault model was assumed and divided into 6~× 7 square subfaults with an area 10~× 10~km2 for each. We invert detided and detrended tilt data from 11 to 22 August 2007 with the fixed slip direction to the same as that in the previous model, and obtained hourly slip process of the SSE. The summary of the result is as follows: (1) Slow slip and earthquake activity started on the offshore area east of KT2H (the station Katsuura East located at the southeast of BP); (2) Most earthquakes before 17 occurred around the northern edge of the slip area; (3) The slip area and most epicenters migrated to the onshore area north of KT2H after 17; (4) The total released moment corresponds to Mw 6.6, comparable to the result of Ozawa et al. (2007), which is based on GEONET data.
U32A-03
Which Came First, the Tremor or the Slip, and is There Slip-Free Tremor in Cascadia?
While both tremor and slow slip have been well-documented in a variety of environments worldwide, the underlying physical mechanisms remain poorly understood. Are tremor and slow slip two manifestations of the same process (e.g. fault slip), or do they represent two phenomena (e.g. fault slip and hydrofracture) that are linked by a physical process, such as the redistribution of fluids? We seek to address the relationship between tremor and slow slip by exploiting the new data from collocated borehole tensor strainmeters and three-component seismometers deployed by the Plate Boundary Observatory (PBO) in Cascadia. The PBO strainmeters/seismometers have three distinct advantages for this task. First, strainmeter data are continuously sampled at 20 Hz and can track aseismic deformation from the seismic band to periods of weeks. Thus, they span the unexplored frequency band between tremor and slip, which may hold the key to understanding the physical process linking these phenomena. Second, because they measure strain at a point, the strainmeters provide better spatial resolution of strain than GPS. Third, the strainmeters, combined with the collocated borehole seismometers, provide dramatically improved sensitivity to both slow slip and tremor. We use these instruments to focus on two issues that constrain the relationship between tremor and slip. (1) Which comes first, tremor or slip, or are they simultaneous? Although existing network seismic and GPS data suggest tremor comes first, this could be due to bias from the lower sensitivity of GPS to deep slip. The greater sensitivity of the strainmeters should dramatically reduce this bias. (2) Is tremor always accompanied by slip? There are times when tremor has been observed, but without slip inferred from GPS. Given the increased sensitivity to slow slip, the strainmeter data can provide a more stringent evaluation of this question. Our approach to addressing (1) is to characterize the local tremor activity for each PBO installation using the borehole seismometer data augmented by nearby seismic stations, and to characterize the slip activity from the corrected horizontal shear strains. We focus on three major tremor and slow slip events in northern Cascadia in 2005, 2007 and 2008 where we have the highest strain signal levels. As an example, we have analyzed both tremor and slip for the May, 2008 event as observed at PBO station B018. We find both significant local tremor activity as well as a very clear strain signal (of order 0.1 μstrain) with a sharp onset time on both shear components. The tremor and slow slip initiate simultaneously within the resolution of the 30-minute strainmeter data, suggesting a very close relationship between both phenomena. Our approach to addressing (2) is to examine the strainmeter data for detectable strain when there is clear tremor activity local to a PBO installation, but no clear slip inferred from GPS. Preliminary results from Wang et al. (2008) illustrate the potential value of this approach, where they have detected a slip event near station B012 that had not been previously identified by GPS.
U32A-04
Tremor triggered near Parkfield by teleseismic waves
We perform a systematic survey of triggered tremor around the Parkfield section of the San Andreas fault for the 31 teleseismic earthquakes since 2001 with Mw ≥ 7.5 and depth < 100 km. We identify triggered tremor as bursts of high-frequency (~2-8 Hz), non-impulsive seismic energy that are generated during the passage of teleseismic waves and coherent among many stations. 10 teleseismic events triggered clear tremor around Parkfield. About 35% of the tremor is concentrated south of Parkfield near Cholame, where ambient tremor has been identified previously, and 23% near Bitterwater in the creeping section of the San Andreas fault. Tremor is most commonly initiated by and in phase with the Love wave particle velocity. However, the pattern becomes complicated with the arrival of the Rayleigh waves, and sometimes tremor continues after the cessation of the surface waves. We identify two cases where tremor is triggered during the teleseismic PKP phase. These observations indicate a mixture of driven, instantaneous, perhaps Coulomb-friction response with an added component of self-sustaining activity more suggestive of ongoing slow slip or triggered fluid flow. We also examine the ambient tremor occurrence rate before and after the teleseismic events, and find the aforementioned transient increase of tremor rate during the passage of the teleseismic surface waves, followed by a period of relative tremor quiescence. This suggests that the occurrence time of tremor is temporally advanced by the dynamic stress of the teleseismic waves. Larger amplitude in the teleseismic waves correlates well with the occurrence of triggered tremor, and the inferred tremor-triggering threshold at Parkfield is ~2-3 KPa. The relatively low triggering threshold suggests that the effective stress at the tremor source region is very low, most likely due to near-lithostatic fluid pressure.
U32A-05 INVITED
On the Thermal and Metamorphic Environment of Subduction-Zone Episodic Tremor and Slip
Episodic tremor and slip (ETS) have been detected in the Cascadia and Southwest Japan subduction zones [Dragert et al., 2001; Obara, 2002], where subducting crust is relatively warm because of the young incoming lithosphere (<20 Ma) and modest plate convergence rates (~40 mm/yr). In the Southwest Japan subduction zone, low-frequency earthquakes occur on the plate interface at depths of 35-40 km beneath Shikoku where finite-element thermal models predict temperatures of ~500°C in the subducting oceanic crust and at depths of 40-50 km beneath the Kii Peninsula where predicted temperatures are ~350°C. Warmer temperatures of ~575°C are predicted at ETS depths beneath southern Vancouver Island in the Cascadia subduction zone, but here tremor also occurs within the overlying forearc crust where temperatures are lower. Several studies have proposed a link between ETS and high pore pressures that trigger hydrofracturing [e.g., Obara, 2002] or extend the region of conditionally stable slip [e.g., Kodiara et al., 2004]. In the Southwest Japan and Cascadia subduction zones, subducting oceanic crust passes through the blueschist, greenschist, and amphibolite metamorphic facies where mineral dehydration reactions are complex. The different temperatures predicted for the two subduction zones suggest that ETS does not coincide with a specific temperature or metamorphic reaction. Several lines of evidence indicate that a free H2O-rich fluid is present, at least transiently, in subducting oceanic crust and fluids released by prograde metamorphic dehydration reactions may help trigger or enable ETS within the subducting plate. Less clear is the role H2O may play in tremor observed in the Cascadia forearc crust where retrograde hydration reactions are expected to consume H2O derived from the underlying subducting plate, but free H2O may exist locally in faults and fractures lined by hydrous minerals.
U32A-06
Dilatancy Stabilization vs Thermal Pressurization as a Mechanism for Controlling Slow vs Fast Slip
We explore the possibility that rate-state friction nucleates slip under drained conditions but that as slip accelerates dilatancy induced pore-pressure reductions quench the instability, resulting in slow (non-inertial) slip [Segall and Rubin, 2007 AGU]. Accelerating slip leads to shear heating and consequent thermal pressurization of pore-fluids which destabilize slip. This suggests that competition between thermal weakening and dilatant hardening may control whether slip is ultimately fast or slow. We model friction-dilatancy interactions assuming 2D elasticity, rate-state friction and a simplified dilatancy law [Segall-Rice, 1995, JGR]. As a first step, we consider isothermal membrane diffusion. Linearized stability analysis suggests that E = 1-a/b defines a boundary between slow and fast slip, where E ≡ f0 ε/ β b (σ - p∞) measures the strength of dilatant hardening (ε is the dilatancy parameter and f0 is nominal friction). Numerical simulations with E < 1-a/b accelerate to radiation damping limits for sufficiently large slip zones, whereas for E > 1-a/b the maximum slip-speed remains well below inertial. We next consider one-dimensional (fault normal) homogeneous diffusion, computed by finite difference with automated re-meshing to maintain accuracy. In the limit of a vanishingly thin shear zone (h → 0) it can be shown that h · φ = (1-φ) · h , where · h is the change in thickness of the shearing layer (measured in some lab experiments). Dimensional analysis shows that the dilatant efficiency scales with ε h/ ( β (σ - p∞) √ v∞ / chyd dc ). Slow slip is favored by low effective stress and hydraulic diffusivity chyd, as expected. Numerical results indicate slow slip is possible without extremely low effective stress, conditions that would result in dynamic slip without dilatancy; ε = 10-4, h =1 mm, dc = 100 μm, a/b = 0.8, chyd = 10-6 m2/s, and σ - p∞ = 20 MPa. With thermal coupling, the ratio of dilatancy to shear heating efficiency scales with where cth is thermal diffusivity, ρ is density, and cp heat capacity. This suggests that slow slip is strongly favored by low effective stress, consistent with some seismic observations.
U32A-07
Slow Slip Predictions Based on Gabbro Dehydration and Friction Data Compared to GPS Measurements in Northern Cascadia
For episodic slow slip transients in subduction zones, a large uncertainty in comparing surface deformations predicted by rate and state friction modeling [Liu and Rice, JGR, 2007] to GPS measurements lies in our limited knowledge of the frictional properties and fluid pore pressure along the fault. In this study, we apply petrological data [Peacock et al., USGS, 2002; Hacker et al., JGR 2003; Wada et al., JGR, 2008] and recently reported friction data [He et al., Tectonophys, 2006, 2007] for gabbro, as a reasonable representation of the seafloor, to a Cascadia-like 2D model in order to produce simulations which show spontaneous aseismic transients. We compare the resulting inter-transient and transient surface deformations to GPS observations along the northern Cascadia margin. An inferred region along dip of elevated fluid pressure is constrained by seismological observations where available, and by thermal and petrological models for the Cascadia and SW Japan subduction zones. For the assumed a and a-b profiles, we search the model parameter space, by varying the level of effective normal stress σ̅, characteristic slip distance L in the source areas of transients, and the fault width under that low σ̅, to identify simulation cases which produce transient aseismic slip and recurrence interval similar to the observed 20-30 mm and 14 months, respectively, in northern Cascadia. Using a simple planar fault geometry and extrapolating the 2D fault slip to a 3D distribution, we find that the gabbro gouge friction data allows a much better fit to GPS observations than is possible with the granite data [Blanpied et al., JGR, 1995, 1998] which, for lack of a suitable alternative, has been used as the basis for most previous subduction earthquake modeling, including ours. Nevertheless, the values of L required to reasonably fit the geodetic data during a transient event are somewhat larger than 100 microns, rather than in the range of 10 to a few 10s of microns as might be expected from lab results. We propose elsewhere at this meeting [Liu et al., submitted abstract] that dilatancy of fault gouge, and related frictional stabilization because of its assumed infiltration by dehydration fluids, may be important to resolving that discrepancy. Those dilatancy effects are known from Segall and Rice [JGR, 1995] to be important in stabilizing otherwise unstable friction at conditions of low σ̅ like those assumed, and they have been shown by Segall and Rubin [EOS, 2007] to be capable of producing episodic slow slip transients.
U32A-08 INVITED
Toward a Unified View of Tremor and Slow Slip
Evidence from Japan suggests that deep non-volcanic tremor consists of myriad Low Frequency Earthquakes (LFEs) on the subduction interface, with focal mechanisms consistent with plate convergence [Shelly et al., Nature 2007; Ide et al., GRL 2007]. Thus I adopt the view that tremor and the accompanying slow slip represent two manifestations of slip on the same interface, and that they should be explainable using the same constitutive framework. Here I explore various incarnations of rate-and-state (r-s) friction. Episodic slow slip may result from (1) a transition from velocity-weakening to velocity-strengthening behavior at less than slow-slip speeds [Kato, EPSL 2003; Shibazaki and Iio, GRL 2003]; (2) "standard" r-s friction, meaning with velocity-independent material parameters, on a fault whose length is properly "tuned" and that possesses high pore pressure (low effective stress) [Liu and Rice, JGR 2005; 2007], or (3) fault-zone dilatancy coupled with pore pressure reduction and diffusion [Segall and Rubin, EOS, 2007]. Using the most appropriate laboratory constitutive law it seems that (2) requires too much "tuning" [Rubin, JGR, in press], but many aspects of slow slip events, including their low stress drop (~10 kPa in Cascadia and Japan), are most easily explained by effective stresses as low as 1 MPa. This is consistent with expectations from seismology and petrology, and with the sensitivity of tremor to tides and surface waves. LFE's are most simply interpreted as resulting from material heterogeneity that makes their source region more seismogenic than the surroundings, for which embarrassingly many options exist. The largest LFE's in Japan appear to have magnitudes about 1.5 but durations roughly 10 times longer than "typical" M1.5 earthquakes. A fundamental question is "what makes them slow?". Two answers are (1) elastodynamics and (2) something else. For circular ruptures moment is proportional to stress drop and the radius cubed, so for circular earthquakes rate-limited by elastodynamics a factor of 10 increase in duration can be explained by a factor of 1000 decrease in stress drop. This seems unreasonable, but a factor of 100 decrease (so a factor of 5 increase in duration) is not, if the LFE stress drop is the same 10 kPa as the slow slip stress drop. This is within striking distance of the observations, and if this explains tremor, the question becomes "where are the M2, 3, and 4 elastodynamic events that would be expected of a typical Gutenberg-Richter distribution?". One possibility is that there is a characteristic length scale for slip in the tremor source region [Watanabe et al., GRL 2007], which would be of order 300 m for a 10-kPa M1.5 event. A potential length scale is the "compaction length" that arises during porous flow in a viscously-deforming matrix. A second possibility is that LFE moment is only weakly sensitive to the size of the underlying heterogeneity. Colliding creep waves increase the slip speed locally by multiple orders of magnitude, and for some constitutive laws might give rise to this insensitivity. Such a tremor source is appealing because creep waves arise in all sorts of simulations that include material heterogeneity. It could also give rise to spatially-elongate LFE sources, which might help explain some aspects of their spectra as well as allow larger stress drops for the same duration. If LFE's are rate-limited by "something else", then the question becomes "What process can reasonably increase the speed of >105 sources to the point that they radiate detectable energy, without letting them slide or propagate fast enough to be limited by elastodynamics?". At the moment, "slow but elastodynamic" seems like the most promising line of investigation.