Heidi Houston

Rapid, continuous streaking of tremor in Cascadia

Nonvolcanic tremor is a recently discovered weak seismic signal associated with slow slip on a fault plane and has potential to answer many questions about how faults move. Its spatiotemporal distribution, however, is complex and varies over different time scales, and the causal physical mechanisms remain unclear. Here we use a beam backprojection method to show rapid, continuous, slip-parallel streaking of tremor over time scales of several minutes to an hour during the May 2008 episodic tremor and slip event in the Cascadia subduction zone. The streaks propagate across distances up to 65 km, primarily parallel to the slip direction of the subduction zone, both updip and downdip at velocities ranging from 30 to 200 km/h. We explore mainly two models that may explain such continuous tremor streaking. The first involves interaction of slowly migrating creep front with slip-parallel linear structures on the fault. The second is pressure-driven fluid flow through structurally controlled conduits on the fault. Both can be consistent with the observed propagation velocities and geometries, although the second one requires unlikely condition. In addition, we put this new observation in the context of the overall variability of tremor behavior observed over different time scales.

Influence of depth, focal mechanism, and tectonic setting on the shape and duration of earthquake source time functions

Source time functions of 255 moderate to great earthquakes obtained from inversions of teleseismic body waves by Tanioka and Ruff [1997] and coworkers were compared in a systematic way. They were scaled to remove the effect of moment and to allow the direct comparison and averaging of time function shape as well as duration. Time function durations picked by Tanioka and Ruff [1997] are proportional to the cube root of seismic moment if moments from the Harvard centroid moment tensor catalog are used. The average duration of scaled time functions is shorter and the average shape has a more abrupt termination for deeper events than shallower ones, with a distinct change occurring at ∼40 km depth. The complexity of the time functions, as quantified by the number of subevents, appears to decrease below ∼40 km depth. Furthermore, among events shallower than 40 km, the average duration of scaled time functions is shorter, and their average shape has a more abrupt termination (1) for events with strike-slip focal mechanisms compared to thrust events and (2) for the few thrust events associated with an intraplate setting compared to the majority associated with an interplate (subduction) boundary. In each of these cases, events in more technically and seismically active settings have a longer duration and a more gradual termination. This can be interpreted in terms of lower stress drops and/or slower rupture velocities at active plate boundaries, suggesting that fault rheology depends on slip rate and may evolve as total fault slip accumulates. Furthermore, differences in average time function shape and duration associated with different subduction zones suggest that differences exist in the rheology on the plate boundaries at the various subduction zones. Supporting data table is available via Web browser or via Anonymous FTP from ftp://kosmos.agu.org, directory “append” (Username = “anonymous”, Password = “guest”); subdirectories in the ftp site are arranged by paper number. Information on seraching and submitting electronic supplements is found at http://www.agu.org/pubs/esupp_about. html.

Orientation of the stress field surrounding the creeping section of the San Andreas Fault: Evidence for a narrow mechanically weak fault zone

We determine how the orientations of the principal stresses change with distance from the creeping section of the San Andreas fault (SAF). Our data are 2392 earthquake focal mechanisms, which are grouped according to distance from the SAF into 33 bins parallel to fault strike; the width of the bins are dictated by the spatial organization and density of the seismicity. Only events with M ≥ 1.5, depth ≥ 2 km, and well-constrained focal mechanisms are used. The focal mechanisms are inverted simultaneously for the orientations and relative amplitudes of the principal stresses using two different inversion programs. A bootstrap resampling method provides confidence limits on the stress orientations. Our general result is that the maximum horizontal compression lies mostly at high angles to the SAF except in a narrow zone on the fault (no wider than 1–3 km), in which it appears to lie at a smaller angle (∼45°) from the fault strike. In particular, the maximum horizontal compression lies at a high angle to the SAF trend immediately adjacent to the fault. This stress state is most readily interpreted as due to a narrow mechanically weak SAF in this region. The narrowness of this zone contrasts with the region wider than 20–30 km of rotated stresses found surrounding the SAF in southern California [Hardebeck and Hauksson, 1999], suggestive of different mechanical behaviors of the SAF in these two regions. This approach may provide a way to define the width of the mechanically weak part of faults.

Time functions of deep earthquakes from broadband and short-period stacks

To constrain dynamic source properties of deep earthquakes, we have systematically constructed broadband time functions of deep earthquakes by stacking and scaling teleseismic P waves from U.S. National Seismic Network, TERRAscope, and Berkeley Digital Seismic Network broadband stations. We examined 42 earthquakes with depths from 100 to 660 km that occurred between July 1, 1992 and July 31, 1995. To directly compare time functions, or to group them by size, depth, or region, it is essential to scale them to remove the effect of moment, which varies by more than 3 orders of magnitude for these events. For each event we also computed short-period stacks of P waves recorded by west coast regional arrays. The comparison of broadband with short-period stacks yields a considerable advantage, enabling more reliable measurement of event duration. A more accurate estimate of the duration better constrains the scaling procedure to remove the effect of moment, producing scaled time functions with both correct timing and amplitude. We find only subtle differences in the broadband time-function shape with moment, indicating successful scaling and minimal effects of attenuation at the periods considered here. The average shape of the envelopes of the short-period stacks is very similar to the average broadband time function. The main variations seen with depth are (1) a mild decrease in duration with increasing depth, (2) greater asymmetry in the time functions of intermediate events compared to deep ones, and (3) unexpected complexity and late moment release for events between 350 and 550 km, with seven of the eight events in that depth interval displaying markedly more complicated time functions with more moment release late in the rupture than most events above or below. The first two results are broadly consistent with our previous studies, while the third is reported here for the first time. The greater complexity between 350 and 550 km suggests greater heterogeneity in the failure process in that depth range.

Rapid calculation of seismic amplitudes

The ability to calculate traveltimes and amplitudes of seismic waves is useful for many reflection seismology applications such as migration and tomography. Traditionally, ray tracing (C⁁erveny et al., 1977; Julian, 1977), paraxial methods (Claerbout, 1971), or full‐wave methods (Alterman and Karal, 1968) are used for such calculations. These methods have in common considerable computational expense. Recently, Vidale (1988, 1990a) presented two‐dimensional and three‐dimensional methods to efficiently compute traveltimes of the first arrivals to every point in a regularly spaced grid of points, given an arbitrary velocity field sampled at these points. The computational cost of finding each traveltime is roughly one square root operation.

Cascadia tremor spectra: Low corner frequencies and earthquake!like high!frequency falloff

The discovery of non-volcanic tremor (NVT) has opened a new window to observe major Earth plate boundaries. However, the spectral characteristics of NVT have not been well studied due to poor signal-to-noise ratio (SNR) on individual seismograms. We estimate the spectral content of Cascadia tremor between 2.5 and 20 Hz by suppressing noise using array analysis, and compute empirical path corrections using nearby small earthquakes. We demonstrate that the displacement spectra of the Cascadia tremor have corner frequencies around 3–8 Hz and fall off at f−2 to f−3 at higher frequencies. Our results have the following implications. (1) The high-frequency falloff of tremor agrees with the observations of regular earthquakes, suggesting that tremor can be analyzed using standard spectral models. Prior analyses that have shown a tremor spectral falloff proportional to f−1 may reflect only the spectral behavior over a limited frequency band. (2) Tremor may be no different from a swarm of microearthquakes with abnormally small stress drops on the order of kPa, likely due to the presence of fluids. Alternatively the low corner frequencies of tremor may reflect abnormally slow ruptures. (3) Fitting a standard Brune (1970) spectral model implies a moment release rate of Cascadia tremor of 3.8 × 1010 N·m/s assuming the tremor signals are P waves (or 1.4 × 1010 N·m/s assuming S-waves). This implies that a typical 20-day long tremor episode releases moment equivalent to Mw 5.1 (P-wave) or Mw 4.9 (S-wave), although these may be underestimates if the spectra deviate substantially from the Brune model at very low frequencies.

The non‐double‐couple component of deep earthquakes and the width of the seismogenic zone

It has recently been proposed that all deep earthquakes (> 400 km depth) are caused by a shear instability associated with the transformation of metastable olivine to its high-pressure polymorphs within the cold core of subducting slabs (Green and Burnley, 1989; Kirby et al., 1991). If so, then the seismogenic zone would narrow with depth below 400 km following slab isotherms. Because the ratio of fault slip to length does not increase markedly with depth, and stress orientations indicate that deep fault planes are not parallel to the subducting slab, a fault plane large enough to generate a very large, deep earthquake would not fit within a narrow seismogenic zone and multiple fault planes would be needed. If the fault planes are not parallel due to spatial variations in stress, a deviatoric non-double-couple component (volume-preserving, non-shearing part) of the seismic moment tensor is likely to result. Thus, if the seismogenic zone narrows with depth, one would expect an increasing non-double-couple component (1) for large earthquakes as depth increases from 400 to 700 km, and (2) for very deep earthquakes (below 550 km depth) as seismic moment increases. Data from the Harvard CMT catalog, augmented by moment tensor inversions for the two largest deep earthquakes during the past thirty years, are examined for these features. Although significant scatter is present, the data roughly follow the trends described above, suggesting a relationship between the non-double-couple component of deep earthquakes and the geometry of a seismogenic zone narrowing with depth.

The temporal distribution of seismic radiation during deep earthquake rupture

The time history of energy release during earthquakes illuminates the process of failure, which remains enigmatic for events deeper than about 100 kilometers. Stacks of teleseismic records from regional arrays for 122 intermediate (depths of 100 to 350 kilometers) and deep (depths of 350 to 700 kilometers) earthquakes show that the temporal pattern of short-period seismic radiation has a systematic variation with depth. On average, for intermediate depth events more radiation is released toward the beginning of the rupture than near the end, whereas for deep events radiation is released symmetrically over the duration of the event, with an abrupt beginning and end of rupture. These findings suggest a variation in the style of rupture related to decreasing fault heterogeneity with depth.

Waveform effects of a metastable olivine tongue in subducting slabs

We constructed velocity models of subducting slabs with a kinetically-depressed olivine → β- and γ-spinel transition, and examined the effect that such structures would have on teleseismic P waveforms using a full-wave finite-difference method. These two-dimensional calculations yielded waveforms at a range of distances in the downdip direction. The slab models included a wedge-shaped, low-velocity metastable olivine tongue (MOTO) to a depth of 670 km, as well as a plausible thermal anomaly; one model further included a 10-km-thick fast layer on the surface of the slab. The principal effect of MOTO is to produce grazing reflections at wide angles off the phase boundary, generating a secondary arrival 0 to 4 seconds after the initial arrival depending on the take-off angle. The amplitude and timing of this feature vary with the lateral location of the seismic source within the slab cross-section. Careful analysis of waveforms from earthquakes with depths near 400 km, simple sources, and adequate station coverage in appropriate geometries will be required to resolve whether MOTO is present.

Slow slip: A new kind of earthquake

Sandwiched between the shallow region of sudden, infrequent earthquakes and the deeper home to continuous viscous motion lies an intermediate realm of intermittent sliding and rumbling. Discovered in recent years, it still harbors many secrets.