Kunihiko Shimazaki

Mapping active magma chambers by b values beneath the off-Ito volcano, Japan

The b value of the frequency-magnitude relation, and thus the mean magnitude, in the off-Ito volcanic area is not uniform based on detailed b value tomography using about 10,000 events in an area of 10 km radius and in the upper 15 km of the crust. A high b value anomaly (b=15) below 7 km depth and with a radius of about 2 km, located below the coast southeast of Ito, contrasts with lower values of typically b=0.7 north of it and at shallower depths. On the basis of surface deformations, a tensile crack dislocation source was located above 7 km depth. Thus we surmise that the magma chamber in the off-Ito area was located below 7 km depth. The correlation of a high b value anomaly with this model supports our hypothesis that active magma chambers may be mapped by high b value anomalies. The magma body may be described as consisting of two parts: a lower, well-established chamber below 11 km depth and a shallower part (11–7 km depth). The b value in the volume surrounding the magma chamber increased from 0.7 during the 1980s to 1.5 during the earthquake swarms of the 1990s. This may reflect a trend of increasing crack density due to the intrusion activity.

Rupture process of the Miyagi-Oki, Japan, earthquake of June 12, 1978

Abstract The faulting mechanism and multiple rupture process of the M = 7.4 Miyagi-Oki earthquake are studied using surface and body wave data from local and worldwide stations. The main results are as follows. (1) P-wave first motion data and radiation patterns of long-period surface waves indicate a predominantly thrust mechanism with strike N10° E, dip 20°W, and slip angle 76°. The seismic moment is 3.1 × 1027 dyne-cm. (2) Farfield SH waveforms and local seismograms suggest that the rupture occurred in two stages, being concordant with the two zones of aftershock activity revealed by the microearthquake network of Tohoku University. The upper and lower zones, located along the westward-dipping plate interface, are separated by a gap at a depth of 35 km and have dimensions of 37 × 34 and 24 × 34 km2, respectively. Rupture initiated at the southern end of the upper aftershock zone and propagated at N20°W subparallel to the trench axis. About 11 s later, the second shock, which was located 30 km landward (westward) of the first, initiated at the upper corner of the lower aftershock zone and propagated down-dip N80°W. Using Haskell modelling for this rupture process, synthetic seismograms were computed for teleseismic SH waves and nearfield body waves. Other parameters determined are: seismic moment M 0 = 1.7 × 10 27 dyne-cm, slip dislocation u = 1.9 m , Δσ = 95 bar, rupture velocity ν = 3.2 km s −1 , rise time τ = 2 s , for the first event; M 0 = 1.4 × 10 27 dyne-cm , u = 2.4 m , Δσ = 145 bar , for the second event; and time separation between the two shocks ΔT = 11 s. The above two-segment model does not explain well the sharp onsets of the body waves at near-source stations. An initial break of a small subsegment on the upper zone, which propagated down-dip, was hypothesized to explain the observed near-source seismograms. (3) The multiple rupture of the event and the absence of aftershocks between the two fault zones suggests that the frictional and/or sliding characteristics along the plate interface are not uniform. The rupture of the first event was arrested, presumably by a region of high fracture strength between the two zones. The fracture energy of the barrier was estimated to be 1010 erg cm−2. (4) The possible occurrence of a large earthquake has been noted for the region adjacent to and seaward of the area that ruptured during the 1978 event. The 1978 event does not appear to reduce the likelihood of occurrence of this expected earthquake.

Pre-seismic crustal deformation caused by an underthrusting oceanic plate, in eastern Hokkaido, Japan

Abstract Eastern Hokkaido is being considered as one of the prime candidates for a future great earthquake in Japan. Geodetic work and seismological data suggest that the continental plate is being compressed and dragged down into the asthenosphere by the underthrusting Pacific plate. A quantitative examination of this idea was carried out by an application of the finite-element method to a static-field problem of crustal deformation. The effect of heterogeneous medium with Poissons ratio differing from 0.25 is significant. An inverse problem was solved by applying a quadratic programming method to the observed vertical displacement. A model, which assumes a nearly uniform tangential displacement along the interface of the oceanic and continental lithospheres down to a depth of 100 km, explains a gravity-change as well as the general features of crustal deformation observed in eastern Hokkaido. The displacement rate is estimated to be 2.7 cm/year. It indicates that a rebound of the continental lithosphere now brings about a thrust-type earthquake whose reverse slip is comparable to that estimated for great shallow earthquakes along the Kuril trench. A possibility of aseismic rebound is suggested at the lower part of the interface.

Focal mechanism of a shock at the northwestern boundary of the pacific plate: Extensional feature of the oceanic lithosphere and compressional feature of the continental lithosphere

Abstract First motions of P-wave and S-wave polarization angles for a shallow shock that occurred on July 25, 1965, at about 70 km oceanward from the Kuril trench indicate a double couple dip-slip source with a horizontal tension axis in the direction perpendicular to the trench. This suggests that the Pacific plate in this region is being extended in this direction. This event cannot be attributed to be triggered by any adjacent great earthquakes. On the contrary, the adjacent area is now being considered a prime candidate for a future great earthquake. Triangulation surveys, tide gauge observations, and levelling surveys performed in the eastern Hokkaido facing the shock across the trench indicate that the continental crust is contracting in the direction normal to the trench axis and tilting towards the Pacific Ocean. These tectonic features are conformable to the idea that the continental plate is being compressed and dragged down into the asthenosphere by the sinking oceanic plate. The extensional feature of the oceanic plate in the region oceanward from the trench suggests that a gravitational pull acting the cold part of the sinking plate dominates the tectonic process now going on in this region.

Mantle structure and seismotectonics of the Sunda and Banda arcs

Abstract We have examined the mantle structure and seismotectonic features of the Sunda and Banda arcs, Indonesia, based on the P-wave tomographic images, focal mechanism solutions, gravity anomaly and heat-flow data. On the basis of slab morphology and seismicity, we can divide the arcs into three parts, the Western Sunda, Eastern Sunda, and Banda arc. The slab-like tomographic image penetrates to a depth of about 500 km below the Western Sunda arc where seismicity does not exceed a depth of 250 km. In the Eastern Sunda arc, where a seismic gap exists between 300 and 500 km depths, the slab appears to be continuous and to penetrate into the lower mantle. Beneath the Banda arc, with seismicity down to a depth of about 650 km, the slab dips gently and does not appear to penetrate into the lower mantle. The positive gravity anomaly shows a systematic pattern, namely, the anomaly along the Eastern Sunda arc is larger than that in the Western Sunda and the Banda arcs. Along the back-arc side of the Sunda and Banda arcs, the heat flow decreases from the west to the east. Seismic strain release from the shallow earthquakes calculated from the CMT solutions show the strain axes to be oblique to the structural trends. The CMT solutions show that the Eastern Sunda arc is characterized by normal earthquakes along the trench and back-arc thrusting earthquakes north of the volcanic line. In the Western Sunda and the Eastern Sunda arcs, earthquakes of the down-dip extension type dominate the slab down to a depth of 200 km while down-dip compression earthquakes occur below 500 km depth. In the Banda arc, deep earthquakes show down-dip extension to a depth of 500 km; below this depth the state of stress is not clearly defined.

Three-dimensional P-wave velocity structure beneath the Indonesian region

Abstract We present the P-wave seismic tomography image of the mantle to a depth of 1200 km beneath the Indonesian region. The artb inversion method is applied to a dataset of 118,203 P-wave travel times of local and teleseismic events taken from ISC bulletins. Although the resolution is sufficient for detailed discussion in only a limited part of the study region, the results clarify the general tectonic framework in this region and indicate a possible remnant seismic slab in the lower mantle. Structures beneath the Philippine Islands and the Molucca Sea region are well resolved and high-velocity zones corresponding to the slabs of the Molucca Sea and Philippine Sea plates are well delineated. Seismic zones beneath the Manila, Negros and Cotabato trenches are characterized by high-velocity anomalies, although shallow structures were not resolved. The Molucca Sea collision zone and volcanic zones of the Sangihe and Philippine arcs are dominated by low-velocity anomalies. The Philippine Sea slab subducts beneath the Philippine Islands at least to a depth of 200 km and may reach depths of 450 km. The southern end of the slab extends at least to about 6°N near southern Mindanao. In the south, the two opposing subducting slabs of the Molucca Sea plate are clearly defined by the two opposing high-velocity zones. The eastward dipping slab can be traced about 400 km beneath the Halmahera arc and may extend as far north as about 5°N. Unfortunately, resolution is not sufficient to reveal detailed structures at the boundary region between the Halmahera and Philippine Sea slabs. The westward dipping slab may subduct to the lower mantle although its extent at depth is not well resolved. This slab trends N-S from about 10°N in the Philippine Islands to northern Sulawesi. A NE-SW-trending high-velocity zone is found in the lower mantle beneath the Molucca Sea region. This high-velocity zone may represent a remnant of the former subduction zone which formed the Sulawesi arc during the Miocene. The blocks along the Sunda and Banda arcs are less well resolved than those in the Philippine Islands and the Molucca Sea region. Nevertheless, overall structures can be inferred. The bowl-shaped distribution of the seismicity of the Banda arc is clearly defined by a horseshoe-shaped high-velocity zone. The tomographic image shows that the Indian oceanic slab subducts to a depth deeper than 300 km i.e., deeper than its seismicity, beneath Andaman Islands and Sumatra and may be discontinuous in northern Sumatra. Along southern Sumatra, Java and the islands to the east, the slab appears to be continuous and can be traced down to at least a depth of the deepest seismicity, where it appears to penetrate into the lower mantle.

Three-dimensional seismic attenuation structure beneath the Aegean region and its tectonic implication

Abstract The three-dimensional attenuation structure beneath the Aegean Sea and the surrounding regions is determined by inversion of seismic intensity data. A large number of seismic intensity data have been accumulated in a uniform scale in the Aegean region, where the seismic activity is much higher than that of the other Mediterranean regions. Nearly 7000 seismic intensity data from 50 earthquakes that have occurred in these regions are used to determine seismic attenuation structure and source acceleration of the earthquakes. The resultant structure reveals a remarkable contrast of attenuation. In the top layer (depth 0–40 km), low Q is dominant in the Aegean Sea, while high Q is dominant in the surrounding land areas, except for the southwestern Turkey. The low- Q regions correspond to areas of Neogene-Quaternary grabens where the high seismicity of shallow earthquakes appears. In the upper mantle, high- Q zones corresponding to the subducting African Plate dominate along the Hellenic arc. The irregular shape of high- Q zones might reflect the split or disintegration of the African slab. Some low- Q spots corresponding to the distribution of volcanoes exist along the volcanic arc. The low- Q spots might correspond to diapirs causing the subduction volcanism.

Dislocation model for strain accumulation in a plate collision zone

Following a scheme developed for a subduction zone by Savage in 1983, which was successfully applied to a transform plate boundary by Matsu’ura and others in 1986, a dislocation model for a plate collision zone is formulated. The solution consists of a rigid plate motion and a tensile dislocation. In addition to this, a strike-slip dislocation is needed when the collision boundary is not perpendicular to the plate converging direction. Theoretically predicted gradual change in horizontal displacements over the collision zone well explains the results of GPS continuous observation in central Japan where the Eurasian and Okhotsk plates are thought to be colliding. The maximum uplift rate is predicted as 1/π times that of converging velocity, however the observed uplift rate is much smaller than that, although vertical movements observed by GPS network is much less accurate than horizontal movements. A comparison of the theoretical results with the observation suggests relatively thin elastic plates, whose thickness is about 30 km. The obtained dislocation model has a close connection with a horizontal detachment fault. The displacement fields above the advancing and retreating edges of a horizontal rectangular detachment fault are mathematically equivalent to those of collision and rift zones, respectively, and lateral edges to a transform fault boundary. The displacement fields at a colliding and a rifting boundaries are the same except for their signs.

Seismic tomography: 3-D image of upper mantle attenuation beneath the Kanto district, Japan

Abstract With the same principle as being used in medical X-ray imaging, the absorption of S-waves is utilized to map attenuation variations in the upper mantle beneath the Kanto district, Japan where the existence of two sinking slabs, the Pacific and the Philippine Sea slabs, is inferred. Nearly 4500 seismic intensity reports on local shallow and intermediate earthquakes are used as measures of the maximum ground acceleration of S-waves which are assumed to be radiated isotropically from earthquake foci. An ordinary damped least-squares inversion is employed to obtain the attenuation structure. The inversion reveals a remarkable contrast of the attenuation coefficient and three prominent features. The first is a low-attenuation zone corresponding to the descending Philippine Sea slab. The attenuation mapping suggests that an aseismic portion of the Philippine Sea slab continues to a depth of 120 km toward the north. The second feature is a hard “root” (low-attenuation zone) of the plate bearing the Japanese islands, attaining at least a depth of 60 km beneath the northeastern part of the Kanto region. This “root” appears to be responsible for the anomalously deep interplate seismicity in this area, where a strong seismic coupling exists at a depth of 40–70 km between the surficial plate and the underlying Philippine Sea slab. The third feature is a zone of extremely high attenuation in the upper mantle, lying on the continental side of the volcanic front, where many Quaternary volcanoes are located. Especially, a spot of very high attenuation is found at a depth of 60–90 km beneath an active volcano, Mt. Fuji.

Characteristic Earthquake Model and Seismicity around Late Quaternary Active Faults in Japan

A total of 172 late Quaternary active fault zones in Japan are examined to determine whether the Gutenberg–Richter relationship or the characteristic earthquake model more adequately describes the magnitude–frequency distribution during one seismic cycle. By combining seismicity data for more than 100 active fault zones at various stages in their seismic cycles, we reduced the short instrumental observation period compared to the average recurrence interval. In only 5% of the active fault zones were the number of observed events equal to or larger than the number of events expected by the Gutenberg–Richter relationship. The average and median frequency ratios of the number of observed events to the number of expected events from the Gutenberg–Richter relationship are only 0.33 and 0.06, respectively, suggesting that the characteristic earthquake model more appropriately describes the magnitude–frequency distribution along the late Quaternary active faults during one seismic cycle. Moreover, the larger an average slip rate is or the shorter an average recurrence interval, the larger the gap in magnitude tends to be between the characteristic earthquake and the largest among other events. A fault zone with a shorter average recurrence interval and/or a larger average slip rate has generally produced more earthquakes in the past or is likely to be at a more mature or developed stage. Thus, these tendencies may reflect a change in the magnitude–frequency distribution related to the maturity or development of fault zones.