Earthquake hazards on the cascadia subduction zone

Large subduction earthquakes on the Cascadia subduction zone pose a potential seismic hazard. Very young oceanic lithosphere (10 million years old) is being subducted beneath North America at a rate of approximately 4 centimeters per year. The Cascadia subduction zone shares many characteristics with subduction zones in southern Chile, southwestern Japan, and Colombia, where comparably young oceanic lithosphere is also subducting. Very large subduction earthquakes, ranging in energy magnitude (M(w)) between 8 and 9.5, have occurred along these other subduction zones. If the Cascadia subduction zone is also storing elastic energy, a sequence of several great earthquakes (M(w) 8) or a giant earthquake (M(w) 9) would be necessary to fill this 1200-kilometer gap. The nature of strong ground motions recorded during subduction earthquakes of M(w) less than 8.2 is discussed. Strong ground motions from even larger earthquakes (M(w) up to 9.5) are estimated by simple simulations. If large subduction earthquakes occur in the Pacific Northwest, relatively strong shaking can be expected over a large region. Such earthquakes may also be accompanied by large local tsunamis.     

An inverted double seismic zone in chile: evidence of phase transformation in the subducted slab

Data from two microseismic field experiments in northern Chile revealed an elongated cluster of earthquakes in the subducted Nazca plate at a depth of about 100 kilometers in which down-dip tensional events were consistently shallower than a family of compressional earthquakes. This double seismic zone shows a distribution of stresses of opposite polarity relative to that observed in other double seismic zones in the world. The distribution of stresses in northern Chile supports the notion that at depths of between 90 to 150 kilometers, the basalt to eclogite transformation of the subducting oceanic crust induces tensional deformation in the upper part of the subducted slab and compressional deformation in the underlying mantle.     

Interaction of the san jacinto and san andreas fault zones, southern california: triggered earthquake migration and coupled recurrence intervals

Two lines of evidence suggest that large earthquakes that occur on either the San Jacinto fault zone (SJFZ) or the San Andreas fault zone (SAFZ) may be triggered by large earthquakes that occur on the other. First, the great 1857 Fort Tejon earthquake in the SAFZ seems to have triggered a progressive sequence of earthquakes in the SJFZ. These earthquakes occurred at times and locations that are consistent with triggering by a strain pulse that propagated southeastward at a rate of 1.7 kilometers per year along the SJFZ after the 1857 earthquake. Second, the similarity in average recurrence intervals in the SJFZ (about 150 years) and in the Mojave segment of the SAFZ (132 years) suggests that large earthquakes in the northern SJFZ may stimulate the relatively frequent major earthquakes on the Mojave segment. Analysis of historic earthquake occurrence in the SJFZ suggests little likelihood of extended quiescence between earthquake sequences.     

Weak interplate coupling by seamounts and repeating M approximately 7 earthquakes

Subducting seamounts are thought to increase the normal stress between subducting and overriding plates. However, recent seismic surveys and laboratory experiments suggest that interplate coupling is weak. A seismic survey in the Japan Trench shows that a large seamount is being subducted near a region of repeating earthquakes of magnitude M approximately 7. Both observed seismicity and the pattern of rupture propagation during the 1982 M 7.0 event imply that interplate coupling was weak over the seamount. A large rupture area with small slip occurred in front of the seamount. Its northern bound could be determined by a trace of multiple subducted seamounts. Whereas a subducted seamount itself may not define the rupture area, its width may be influenced by that of the seamount.     

Rapid intraplate strain accumulation in the new madrid seismic zone

Remeasurement of a triangulation network in the southern part of the New Madrid seismic zone with the Global Positioning System has revealed rapid crustal strain accumulation since the 1950s. This area experienced three large (moment magnitudes >8) earthquakes in 1811 to 1812. The orientation and sense of shear is consistent with right-lateral strike slip motion along a northeast-trending fault zone (as indicated by current seismicity). Detection of crustal strain accumulation may be a useful discriminant for identifying areas where potentially damaging intraplate earthquakes may occur despite the absence of large earthquakes during historic time.     

Repeating deep earthquakes: evidence for fault reactivation at great depth

We have identified three groups of deep earthquakes showing nearly identical waveforms in the Tonga slab. Relocation with a cross-correlation method shows that each cluster is composed of 10 to 30 earthquakes along a plane 10 to 30 kilometers in length. Some of the earthquakes are colocated, demonstrating repeated rupture of the same fault, and one pair of events shows identical rupture complexity, suggesting that the temporal and spatial rupture pattern was repeated. Recurrence intervals show an inverse time distribution, indicating a strong temporal control over fault reactivation. Runaway thermal shear instabilities may explain temporally clustered earthquakes with similar waveforms located along slip zones weakened by shear heating. Earthquake doublets that occur within a few hours are consistent with events recurring before the thermal energy of the initial rupture can diffuse away.     

Large trench-parallel gravity variations predict seismogenic behavior in subduction zones

We demonstrate that great earthquakes occur predominantly in regions with a strongly negative trench-parallel gravity anomaly (TPGA), whereas regions with strongly positive TPGA are relatively aseismic. These observations suggest that, over time scales up to at least 1 million years, spatial variations of seismogenic behavior within a given subduction zone are stationary and linked to the geological structure of the fore-arc. The correlations we observe are consistent with a model in which spatial variations in frictional properties on the plate interface control trench-parellel variations in fore-arc topography, gravity, and seismogenic behavior.     

Seismic models of the root of the sierra nevada

Seismic waves generated by earthquakes or explosions show a delay in travel times as they propagate across the Sierra Nevada from all directions except that of the Nevada test site. Early arriving waves from the test site can be explained if they emerge through a rock layer with high seismic velocity from the sharp eastern edge of the Sierran root. Such a layer could be formed by the subducted ophiolite slab that crops out in the western Sierra Nevada foothills. A synthesis of all seismic data indicates that the Sierran root projects downward into the mantle to a depth of about 55 kilometers beneath the high Sierra.     

Subducted seamount imaged in the rupture zone of the 1946 nankaido earthquake

The Nankai Trough is a vigorous subduction zone where large earthquakes have been recorded since the seventh century, with a recurrence time of 100 to 200 years. The 1946 Nankaido earthquake was unusual, with a rupture zone estimated from long-period geodetic data that was more than twice as large as that derived from shorter period seismic data. In the center of this earthquake rupture zone, we used densely deployed ocean bottom seismographs to detect a subducted seamount 13 kilometers thick by 50 kilometers wide at a depth of 10 kilometers. We propose that this seamount might work as a barrier inhibiting brittle seismogenic rupture.     

Seismic precursors to the mount st. Helens eruptions in 1981 and 1982

Six categories of seismic events are recognized on the seismograms from stations in the vicinity of Mount St. Helens. Two types of high-frequency earthquakes occur near the volcano and under the volcano at depths of more than 4 kilometers. Medium- and low-frequency earthquakes occur at shallow depths (less than 3 kilometers) within the volcano and increase in number and size before eruptions. Temporal changes in the energy release of the low-frequency earthquakes have been used in predicting all the eruptions since October 1980. During and after eruptions, two types of low-frequency emergent surface events occur, including rockfalls and steam or gas bursts from the lava dome.     

Mantle phase changes and deep-earthquake faulting in subducting lithosphere

Inclined zones of earthquakes are the primary expression of lithosphere subduction. A distinct deep population of subduction-zone earthquakes occurs at depths of 350 to 690 kilometers. At those depths ordinary brittle fracture and frictional sliding, the faulting processes of shallow earthquakes, are not expected. A fresh understanding of these deep earthquakes comes from developments in several areas of experimental and theoretical geophysics, including the discovery and characterization of transformational faulting, a shear instability connected with localized phase transformations under nonhydrostatic stress. These developments support the hypothesis that deep earthquakes represent transformational faulting in a wedge of olivine-rich peridotite that is likely to persist metastably in coldest plate interiors to depths as great as 690 km. Predictions based on this deep structure of mantle phase changes are consistent with the global depth distribution of deep earthquakes, the maximum depths of earthquakes in individual subductions zones, and key source characteristics of deep events.     

Accelerated plate tectonics

The concept of a stressed elastic lithospheric plate riding on a viscous asthenosphere is used to calculate the recurrence interval of great earthquakes at convergent plate boundaries, the separation of decoupling and lithospheric earthquakes, and the migration pattern of large earthquakes along an arc. It is proposed that plate motions accelerate after great decoupling earthquakes and that most of the observed plate motions occur during short periods of time, separated by periods of relative quiescence.     

Extending earthquakes' reach through cascading

Earthquakes, whatever their size, can trigger other earthquakes. Mainshocks cause aftershocks to occur, which in turn activate their own local aftershock sequences, resulting in a cascade of triggering that extends the reach of the initial mainshock. A long-lasting difficulty is to determine which earthquakes are connected, either directly or indirectly. Here we show that this causal structure can be found probabilistically, with no a priori model nor parameterization. Large regional earthquakes are found to have a short direct influence in comparison to the overall aftershock sequence duration. Relative to these large mainshocks, small earthquakes collectively have a greater effect on triggering. Hence, cascade triggering is a key component in earthquake interactions.     

Temporally unstable recurrence of earthquakes due to breaks in fractal scaling

Observed sequences of large earthquakes are not consistent in either recurrence time or energy release; long-term prediction has been impossible even in areas, such as Parkfield, with well-defined recurrence intervals. The seismic gap hypothesis, which predicts characteristic earthquakes in areas of the circum-Pacific belt that have not produced recent great earthquakes, has also failed to predict the observed clustering of high-energy events. Models in which fractal scaling is broken at high magnitude predict that characteristic events and recurrence behavior will be unstable in time. The central predictions of these models are supported by recent observations at Landers and Big Bear in California.     

Recent earthquake prediction research in Japan

Japan has experienced many major earthquake disasters in the past. Early in this century research began that was aimed at predicting the occurrence of earthquakes, and in 1965 an earthquake prediction program was started as a national project. In 1978 a program for constant monitoring and assessment was formally inaugurated with the goal of forecasting the major earthquake that is expected to occur in the near future in the Tokai district of central Honshu Island. The issue of predicting the anticipated Tokai earthquake is discussed in this article as well as the results of research on major recent earthquakes in Japan-the Izu earthquakes (1978 and 1980) and the Japan Sea earthquake (1983).     

A model for a seismic computerized alert network

In large earthquakes, damaging ground motions may occur at large epicentral distances. Because of the relatively slow speed of seismic waves, it is possible to construct a system to provide short-term warning (as much as several tens of seconds) of imminent strong ground motions from major earthquakes. Automated safety responses could be triggered by users after receiving estimates of the arrival time and strength of shaking expected at an individual site. Although warning times are likely to be short for areas greatly damaged by relatively numerous earthquakes of moderate size, large areas that experience very strong shaking during great earthquakes would receive longer warning times.     

Double seismic zone for deep earthquakes in the izu-bonin subduction zone

A double seismic zone for deep earthquakes was found in the Izu-Bonin region. An analysis of SP-converted phases confirms that the deep seismic zone consists of two layers separated by approximately 20 kilometers. Numerical modeling of the thermal structure implies that the hypocenters are located along isotherms of 500 degrees to 550 degrees C, which is consistent with the hypothesis that deep earthquakes result from the phase transition of metastable olivine to a high-pressure phase in the subducting slab.     

Osmium Recycling in Subduction Zones

Peridotite xenoliths from the Cascade arc in the United States and in the Japan arc have neodymium and osmium isotopic compositions that are consistent with addition of 5 to 15 percent of subducted material to the present-day depleted mantle. These observations suggest that osmium can be partitioned into oxidized and chlorine-rich slab-derived fluids or melts. These results place new constraints on the behavior of osmium (and possibly other platinum group elements) during subduction of oceanic crust by showing that osmium can be transported into the mantle wedge.     

Slow deformation and lower seismic hazard at the new madrid seismic zone

Global Positioning System (GPS) measurements across the New Madrid seismic zone (NMSZ) in the central United States show little, if any, motion. These data are consistent with platewide continuous GPS data away from the NMSZ, which show no motion within uncertainties. Both these data and the frequency-magnitude relation for seismicity imply that had the largest shocks in the series of earthquakes that occurred in 1811 and 1812 been magnitude 8, their recurrence interval should well exceed 2500 years, longer than has been assumed. Alternatively, the largest 1811 and 1812 earthquakes and those in the paleoseismic record may have been much smaller than typically assumed. Hence, the hazard posed by great earthquakes in the NMSZ appears to be overestimated.     

Regular intervals between hawaiian earthquakes: implications for predicting the next event

During the years 1941 through 1983 five earthquake mainshocks of moderate magnitude occurred at regular intervals of 10.5 +/- 1.5 years within a 6-kilometer radius in Hawaii. It is proposed that these Kaoiki earthquakes will continue to occur at regular intervals because the strain accumulation rate and the strained volume remain constant. With appropriate instrumentation, it may be possible to refine predictions of subsequent Kaoiki earthquakes.