Koshun Yamaoka

Variations of fluid pressure within the subducting oceanic crust and slow earthquakes

[1] We show fine-scale variations of seismic velocities and converted teleseismic waves that reveal the presence of zones of high-pressure fluids released by progressive metamorphic dehydration reactions in the subducting Philippine Sea plate in Tokai district, Japan. These zones have a strong correlation with the distribution of slow earthquakes, including long-term slow slip (LTSS) and low-frequency earthquakes (LFEs). Overpressured fluids in the LTSS region appear to be trapped within the oceanic crust by an impermeable cap rock in the fore-arc, and impede intraslab earthquakes therein. In contrast, fluid pressures are reduced in the LFE zone, which is deeper than the centroid of the LTSS, because there fluids are able to infiltrate into the narrow corner of the mantle wedge, leading to mantle serpentinization. The combination of fluids released from the subducting oceanic crust with heterogeneous fluid transport properties in the hanging wall generates variations of fluid pressures along the downgoing plate boundary, which in turn control the occurrence of slow earthquakes.

The excitation and characteristic frequency of the long-period volcanic event: An approach based on an inhomogeneous autoregressive model of a linear dynamic system

We present a method to quantify the source excitation function and characteristic frequencies of long-period volcanic events. The method is based on an inhomogeneous autoregressive (AR) model of a linear dynamic system, in which the excitation is assumed to be a time-localized function applied at the beginning of the event. The tail of an exponentially decaying harmonic waveform is used to determine the characteristic complex frequencies of the event by the Sompi method. The excitation function is then derived by operating an AR filter constructed from the characteristic frequencies to the entire seismogram of the event, including the inhomogeneous part of the signal. We apply this method to three long-period events at Kusatsu-Shirane Volcano, central Japan, whose waveforms display simple decaying monochromatic oscillations except for the beginning of the events. We recover time-localized excitation functions lasting roughly 1 s at the start of each event and find that the estimated functions are very similar to each other at all the stations of the seismic network for each event. The phases of the characteristic oscillations referred to the estimated excitation function fall within a narrow range for almost all the stations. These results strongly suggest that the excitation and mode of oscillation are both dominated by volumetric change components. Each excitation function starts with a pronounced dilatation consistent with a sudden deflation of the volumetric source which may be interpreted in terms of a choked-flow transport mechanism. The frequency and Q of the characteristic oscillation both display a temporal evolution from event to event. Assuming a crack filled with bubbly water as seismic source for these events, we apply the Van Wijngaarden-Papanicolaou model to estimate the acoustic properties of the bubbly liquid and find that the observed changes in the frequencies and Q are consistently explained by a temporal change in the radii of the bubbles characterizing the bubbly water in the crack.

An isotropic source of volcanic tremor-observation with a dense seismic network at Izu-Oshima volcano, Japan

Abstract Volcanic tremor is frequently observed associated with activities of volcanoes. In order to study the nature of the volcanic tremor in detail we installed a dense seismic network covering the area of the summit crater of Izu-Oshima volcano. We succeeded in observing volcanic tremors preceding a small explosive eruption on November 16, 1987. The explosion was followed by a significant draining back of lava which had filled the summit crater of the 1986 eruption. We analyzed the low-frequency components of the tremor (near 0.7 Hz) and found that it was generated by an isotropic (volumetric) source located 350 m above sea level. Assuming a cyclic pressure change at the source the amplitude of the seismic moment is 3.4 × 10 10 Nm. This corresponds to a pressure fluctuation of 3 × 10 3 Pa (0.03 bar) in the system containing the cavity of 10 7 m 3 , which is comparable to the volume of subsidence at the 1987 eruption.

Ground deformation associated with volcanic tremor at Izu-Oshima Volcano

Izu-Oshima volcano had summit and fissure eruptions in November, 1986 after 12 years of dormancy, and three small eruptions followed these events within one year. Episodic and continuous volcanic tremors were observed for the period containing these eruptions. It is a remarkable discovery that the episodic volcanic tremor was accompanied by a small but sharp ground deformation, the polarity of which was variable. The distribution of tilt vectors reveals that the source of ground deformation was always located beneath the northwestern flank of the volcano, where a magma reservoir was predicted by other studies. On the other hand, the seismologically detected tremor source was determined to be at a shallow depth below the central pit crater, a few kilometers away from the predicted magma reservoir. It is thus inferred that the tremor source near the crater generated pressure increases or decreases that were simultaneously transmitted through the vent to the magma reservoir and lead to its expansion or contraction.

Spherical shell tectonics: buckling of subducting lithosphere

Abstract The lithospheric plate is a spherical shell rather than a plane plate. A spherical shell must buckle when it is bent inward. We examined the possibility of lithospheric buckling upon subduction. The lithosphere with its subducted portion is simulated by a hemispherical elastic shell bent inward at its circumferential edge by a uniform radial load. The buoyancy force acting on the lithosphere seaward of the trench is simulated by clamping the shell along a parallel. The deformable portion between the loaded and clamped edges corresponds to the subducting slab of lithosphere. Buckling analyses were made using the techniques of a linear stability analysis and a finite element method. The wavelength of buckling depends on the thickness of the shell and the length of its deformable portion: a longer wavelength is associated with a thicker shell and a longer deformable portion. By scaling the shell thickness and the length of the deformable portion to the elastic thickness of lithosphere and the length of subducting slab, respectively, the wavelength of buckling can be compared favourably to the length of one unit of arcuate trench in a chain-like continuation of island arcs. The load to be applied for initiation of buckling is called the critical load: a lower critical load is associated with a thinner shell with a longer deformable portion. The excess weight of subducting slab provides a load large enough to initiate lithospheric buckling at a relatively early stage of subduction. Radial displacement of the circumferential edge at the critical state is an order of magnitude smaller than the depth of the leading edge of the Wadati-Benioff zone, indicating that most of the present subducting lithospheres are under a postbuckling state. Undulation of the shell in the postbuckling state is not purely sinusoidal but a successive continuation of arcs with cusps in between, invoking the continuation of arcuate deep-sea trenches with cusps at their junctions. The cuspate feature of island arc chains is thus a natural consequence of lithospheric buckling and does not require any such irregularities as seamounts colliding with a trench. Collision of seamounts, however, can aid buckling greatly if they are aligned along a trench at an interval not very different from the inherent buckling wavelength.

Spherical shell tectonics: on the buckling of the lithosphere at subduction zones

Abstract The buckling phenomena of the subducting lithosphere due to the sphericity of the earth are studied by a nonlinear finite element method (FEM), and the effects of geometrical nonlinearity are examined. The subduction of the lithosphere is modelled by a hemispherical shell squeezed at its circumferential edge. In scaling the geometry of the subducting lithosphere, two parameters are employed; one is the thickness of the hemispherical shell (i.e., the thickness of the lithosphere) and the other is the length of its deformable portion (i.e., the length of the descending slab). To simulate the bending of the lithosphere at the trench, uniform inward load is applied to the free circumferential edge of the deformable portion, which corresponds to the leading edge of the subducting lithosphere. At the opposite edge of the deformable portion (i.e., at the trench), a built-in boundary condition is imposed; no displacements and no rotations are allowed. With a load greater than some critical value, the deformable portion of the shell buckles and undulates azimuthally. The investigation of this buckling phenomenon can be summarized as follows: 1. (1) Buckling with a shorter wavelength is associated with a thinner and shorter shell. Mechanical irregularities simulating lithospheric tearing or seamount collision do not alter the buckling wavelength very much. After appropriate scaling, the lengths of subduction zones coincide with the buckling wavelength calculated by F.E.M. 2. (2) Sharp cusps pointing outward from the shell appear during the postbuckling undulation without any irregularities in shell property and in the applied load. The shape of this cuspidal undulation is quite similar to that actually observed at an arc-arc junction. 3. (3) Buckling can take place in the stress environment likely to exist in the earth. These results strongly suggest that the arcuate shapes of subduction zones are in fact generated by buckling of the subducting lithosphere.

A new model for the fault beneath the sedimentary basin in the 1891 Nobi earthquake

We have investigated the geometry and detailed location of the Gifu-Ichinomiya (GI) fault, a buried fault considered to have ruptured during the 1891 Nobi earthquake. Based on an inversion of coseismic vertical displacements obtained by leveling surveys, we obtain an inclined fault plane showing a reverse fault-type mechanism, rather than the vertical fault plane assumed in previous models. The fault dips 60? to the east and its slip during the earthquake is estimated to have been 1.48 m along the fault dip. The fault is located 5 km east of the location assumed in the previous models. Recent earthquakes have been scattered along the new fault location, not the previously estimated one. Focal mechanisms of the earthquakes that have occurred around the fault are dominated by a reverse fault component, which is consistent with the focal mechanism of the GI fault obtained by the inversion of coseismic displacements. The seismic intensity distribution calculated using the new fault geometry and location explains well the distribution of observed damage caused by the 1891 Nobi earthquake.

Continuous observation of seismic wave velocity and apparent velocity using a precise seismic array and ACROSS seismic source

We report the results of continuous monitoring—using a seismometer array—of the travel time of seismic waves generated by an ACROSS artificial seismic source. The seismometer array, which was deployed in a surface vault located 2.4 km from the source, recorded both direct P- and S-waves and refracted P- and S-waves that traveled along a velocity boundary between the granite basement and overlying sedimentary rocks. We analyzed temporal variation in differential travel time and apparent velocity for these phases for a period of 1 month and found significant temporal variation in the differential travel time. Most of the variation can be attributed to changes in environmental conditions, such as atmospheric temperature and rainfall. Variation is even observed in the seismogram that is located 50 m from the vibration source, although much smaller variation is observed in the vibration of the foundation to which the source is attached. The spectral study revealed that the effects of temperature and rainfall depend strongly on the frequency range used by ACROSS and that a large variation occurs in the 15- to 20-Hz range, especially between 17 and 20 Hz. The environmental effect on the temporal variation is comparable to the record of refracted S waves and that of a distance of 50 m, whereas a larger variation was observed in the direct S wave. This result shows that the signal is affected by the environmental change near the vibration source. The environmental effect can be drastically reduced when the signal from the 15- to 20-Hz range is eliminated in the analysis.

Source spectrum and source time function of volcanic tremor determined with a dense seismic network near the summit crater of Izu‐Oshima Volcano, Japan

Digital seismic records of episodic volcanic tremor, obtained with a dense seismic network near the summit crater of Izu-Oshima volcano, were analyzed to determine source spectrum and source time function. Source spectrum and transfer function could be separated because the seismic records showed a systematic change with distance from the source. The source spectrum of velocity amplitude had a different frequency, ƒ, dependence above and below a corner frequency of 8 to 10 Hz. At high ranges, the spectrum was proportional to ƒ−2, while at low ranges, it was proportional to ƒ2. Inversion of this frequency-dependent source spectrum yields a source time function that can be represented by an impulse that attenuates in about 0.1 s. Repeated impulses could explain observed volcanic tremor that persists for many minutes or longer and that have complicated phase spectra. The source spectrum gives an energy release rate of about 5.2×102 J/s, so that the total energy released is about 1.0×105 J during a tremor episode of about 3 min at Izu-Oshima. Such energy release is comparable to the seismic energy released by an earthquake of magnitude 0.1.

Why do island arcs form cusps at their junctions

An interesting feature of an island-arc chain is a cusp of the arcuate trench at the arc-arc junction where the Wadati-Benioff zone and the volcanic front are also sharpened landward. This cusp has been interpreted as a result of the collision of an aseismic ridge with the trench. We propose a different mechanism, intrinsic to the subduction of a spherical shell of lithosphere. When the spherical lithosphere is bent inward at a subduction zone, buckling takes place, and the slab is deformed into a wavy configuration making the trench chainlike. We analyzed this buckling phenomenon of spherical shells by a finite element method, wliich has been developed in the fields of architecture and civil engineering. Our calculations show that the lithospheric shell buckles with a wavelength roughly consistent with the length of one unit of island arcs (i.e., the distance between the two neighboring junctions). The resultant undulation is not purely sinusoidal, but is a succession of arcs with cusps in between, very similar to the shape of the actual arc-trench system in the world. This agreement strongly suggests that the arcuateness of the trenches and the cuspate form of their junctions are the manifestation of buckling of the spherical lithosphere, although the collision of seamounts and aseismic ridges against the trench may promote the initiation of buckling.