All issues
Volume 49, Issue 1
Displaying 1-14 of 14 articles from this issue
- |<
- <
- 1
- >
- >|
-
Part 1. A Case of the 1993 Hokkaido Nansei-oki Earthquake TsunamiKuniaki ABE1996 Volume 49 Issue 1 Pages 1-9
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSLocal amplifications of tsunami were found on the maximum inundation heights at Japanese 10 coasts facing islands in the 1993 Hokkaido nansei-oki earthquake tsunami. For each island-coast geography a peak height (H) at a coastal focus, a background average height (H0), a peak width (Wd), a coastal focus distance from the island (L) and an island size (L0) are defined on the space distribution of maximum heights obtained in the surveys and the relations are discussed.
Main results are as follows: Amplification ratio H/H0, plotted as a function of coastal focus distance from an island, are distributed in the vicinity of 2. The peak width relative to the island size Wd/L0 is proportional to the focus distance relative to the island size L/L0. There is a negative correlation relation between the amplification ratio and the relative peak width. These facts suggest that the amplification was caused by a superposition of incident waves on the sea in the rear of the island slope after being separated into two and refracted on the slope. Thus, the amplification at the coast facing islands is explained from a focusing effect of islands on a tsunami.View full abstractDownload PDF (3254K) -
Part 2. A Case of the 1983 Nihonkai-chubu Earthquake TsunamiKuniaki ABE1996 Volume 49 Issue 1 Pages 11-17
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSFocusing effect of islands on the 1983 Nihonkai-chubu (central part of the Japan Sea) earthquake tsunami was identified at eight coasts facing to islands from peak formations of the maximum inundation heights. Defining parameters of the peak height (H) at a coastal focus, the background average height (H0), the peak width (Wd), the coastal focus distance from the island (L) and the island size (L0), we discussed relations among them. As the result peak width Wd is approximated as
Wd/L0=0.43 (L/L0)1.0
Amplification ratio H/H0 is about 1.5 for islands of epicentral distances smaller than 400km and shows an increase with the epicentral distance for islands of epicentral distances larger than 400km. These facts are explained as a focusing effect of islands on tsunamis, in which incident wave, refracted in a sloped region around the island after divided into two, superposes on each other in the back side. In islands distant from the source the incident wave is coherent and the amplification ratio increase. The amplification ratio and relative peak width are compared with those in the another focusing effect, previously found on the 1993 Hokkaido nansei-oki earthquake tsunami. The similar proportionality of peak width and amplification ratio between two tsunamis suggest that the amplification is caused by the same mechanism.View full abstractDownload PDF (3977K) -
Tokutaro HATORI1996 Volume 49 Issue 1 Pages 19-26
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSAdding the newly data of the 1994 Sanriku-Oki tsunamis, the distribution of the radiating tsunami energy off Sanriku, North Japan, is investigated during the last 100 years, 1896-1995. Judging from the diagram of the attenuation of wave-height with distance, magnitudes of the Sanriku-Oki tsunamis on Apr. 8, and Dec. 28, 1994 are m=-1 and m=1.5 on the Imamura-Iida scale, respectively. These values are normal compared to earthquake magnitudes (M=6.6 and M=7.5). The source area of the tsunami on Dec. 28, 1994 corresponding to an aftershock area lays extending about 150km to the unusual E-W direction. During 1896-1959, the maximum cumulative energy was radiated in the source area of the 1896 Sanriku tsunami (m=3.5) with the order of 1020 ergs per 15′ mesh. And the areas of radiated energy of 1018 ergs extend to Shimokita Peninsula and the Miyagi coasts. On the contrary, the amount of cumulative energy during 1960-1995 drops 1019 ergs and the radiated area moves to the north side. The remarkable seismic gaps still remain near Shimokita Peninsula and the trench side far off Miyagi Prefecture.View full abstractDownload PDF (977K) -
Takayuki HINO, Yoshinobu TSUJI1996 Volume 49 Issue 1 Pages 27-38
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSThe Tonankai earthquake of December 7, 1944 occurred on the upper plane of the subducted slab located at the boundary between the Philippine Sea plate and the Eurasian plate. Recently it was pointed out that in the Kumano-coast region, the southeast part of the Kii Peninsula, the direction of the strike of the subducting Philippine Sea plate runs from southwest to northeast, while in the Ensyu-coast region, which is located between Nagoya and Shizuoka, the direction runs from east to west. We assumed that the dislocation of seismic faulting took place on the determined plate boundary and the dislocation is equal on each fault plane. As the plate boundary discontinues beneath the Ise Bay, two different fault planes were considered in our fault model: the one is located along the Kumano-coast on which the amount of slip is 200cm and its direction is N50°E, and the other is along the Ensyu-coast on which the amount of slip is 50cm and its direction is N50°E. We checked that the crustal movement calculated by our model is similar to the observed pattern. We carried out a numerical simulation of the tsunami based on our model. It was found that the sea level changes calculated by our model agreed well with the tide gauge records at four stations on the coast of the Ise Bay.View full abstractDownload PDF (1312K) -
Daisuke YAMAZAKI, Takumi KATO, Eiji OHTANI1996 Volume 49 Issue 1 Pages 39-53
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSDeformation experiments using a multianvil apparatus have been performed to investigate the preferred orientation of olivine aggregates at confining pressures of 6GPa and 9GPa, temperature of 1200°C and differential stresses of 80-400MPa. In this study, the lattice preferred orientations are produced by recrystallization (grain boundary migration) driven by dislocation energy. In a compressional stress field, c-axis concentrates on the plane perpendicular to the compressional direction, whereas a- and b-axes rotate with time at high temperature, and then a- and b-axes approach to the direction of an angle of about 45° from the compressional direction. In a tensional stress field, a- and c-axes concentrate on the plane perpendicular to the tensional direction and b-axis concentrates toward the tensional direction. The preferred orientation at the confining pressure of 9GPa is weaker than that observed at confining pressure of 6GPa. The magnitude of the anisotropy on P-wave velocity in most samples are 2-4%, which is nearly equivalent to that observed in the upper mantle.View full abstractDownload PDF (4231K) -
Isamu AIDA1996 Volume 49 Issue 1 Pages 55-63
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSEdo (now Tokyo) area suffered serious damage from a great earthquake in 1855, Ansei era. The earthquake is conceived to be essencially an inland earthquake, although the crustal movement might be extended to the bottom of Tokyo Bay. Whether or not a tsunami was generated by the earthquake is investigated basing on historical documents. Although, no tsunami damage was reported, it seems likely that some phenomena associated with a tsunami, which was too small to be observed obviously, took place. Numerical experiments on three fault models, that are supposed on the upper surface of the subducted Philippine sea plate, are made in order to see possible tsunami generation in Tokyo Bay. Maximum double amplitudes of computed tsunamis amount to 10 to 20cm which are in harmony with historical documents. It may be one of the reasonable interpretations for the source model of the earthquake, therefore, that the Ansei Edo earthquake occurred on the plate boundary. Water oscillations induced by the seismic acceleration in a tank or a pool are numerically investigated. It becomes apparent that such water oscillation is largely controlled by the resonance in the reservoir.View full abstractDownload PDF (1339K) -
Takamori ITO, Ryuichi SUGISAKI, Iwao KAWABE, Koichiro NAGAMINE1996 Volume 49 Issue 1 Pages 65-73
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSMineral spring gas compositions at several sites in central Japan have been continuously monitored in our laboratory for the earthquake prediction study since 1979. The mineral spring gas flow rate has been also monitored since 1993. Southern Hyogo prefecture earthquake occurred on January 17, 1995. At this event, we observed preseismic and coseismic anomalous behavior of mineral spring gases at Byakko spa, a monitoring site of ours, 220km away from the epicenter. The gas flow rate decreased remarkably and rapidly 3 hours before the event. This decrease may be attributed to a decline of water flow rate due to a preseismic crustal deformation at the site. Three gas ratios, He/Ar, N2/Ar and CH4/Ar, conspicuously increased after the event. The ratios of N2/Ar and He/Ar began to increase one or two days before the event, while CH4/Ar did at the occurrence time of the event. These changes can be interpreted as the result of changes of a mixing process of the gases dissolved in ground water. The present observation indicates the efficacy of monitoring of subsurface fluids for earthquake prediction.View full abstractDownload PDF (1120K) -
Kenichi KATO, Masayuki TAKEMURA1996 Volume 49 Issue 1 Pages 75-83
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSDirection of rupture propagation φ and fault length L for the 1994 Sanriku-Haruka-Oki earthquake are evaluated from the azimuthal dependence in duration of strong ground motion observed at eleven stations along the Pacific coast of Tohoku and Hokkaido region. After defining the strong motion duration D as the time interval between onset of S wave and 85% of cumulative power curve derived from 5 to 10Hz band-pass filtered accelerogram, the azimuthal dependence of D is examined. We find that D systematically shows directivity: D is the shortest at northern part of Tohoku and the longest at southern part of Tohoku and eastern edge of Hokkaido. By making use of the directivity on observed strong motion durations, we apply the method by IZUTANI and HIRASAWA (1987) to deducing φ and L together with VR/β, where VR and β are rupture velocity and S wave velocity. Assuming that rupture initiated from the epicenter determined by the Japan Meteorological Agency (JMA), we obtain L/VR=53, VR/β=0.62, and φ=W9°N (case 1). This result indicates that the rupture was headed toward northern part of Tohoku, and terminated near the western end of aftershock area. Detailed source-process inversion [for example, SATO et al. (1996)] suggests that main rupture nucleated near the center of aftershock area, corresponding to the epicenter determined by Harvard University. We redo the calculation assuming that the rupture initiated from the Harvard epicenter, then resulting in L/VR=28, VR/β=0.71, and φ=W18°N (case 2). This result also indicates that the rupture propagated toward northern part of Tohoku, and terminated at the western end of aftershock area. SATO et al. (1996) suggest that this earthquake consists of threestage rupture process: the first and the second sub-events corresponding to JMA and Harvard epicenter, respectively, and the third sub-event located near the western end of aftershock area at the distance of about 50km from Hachinohe city. The location where rupture terminated in both cases of this study is consistent with the location of the third sub-event, confirming that the rupture extended to the western end of aftershock area.View full abstractDownload PDF (1144K) -
Shozo MATSUMURA1996 Volume 49 Issue 1 Pages 85-88
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSDownload PDF (2117K) -
Harumi AOKI1996 Volume 49 Issue 1 Pages 89-90
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSDownload PDF (235K) -
Masayuki TAKEMURA, Takashi NOZAWA1996 Volume 49 Issue 1 Pages 91-100
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSDownload PDF (2828K) -
Stick of Fault and Slip on DetachmentYoshihisa IIO1996 Volume 49 Issue 1 Pages 103-112
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSA possible generating process of the 1995 Southern Hyogo Prefecture Earthquake was inferred from the crustal strain data for about one hundred years before the earthquake. The process consists of two different components, stress concentration and strength reduction. It is found from the crustal strain data that the compressional strain in the east-west direction is smaller around the focal region of the earthquake than in the other areas, and that the tensional strain in the north-south direction is dominant especially around the earthqauke fault. The first result suggests that the earthquake fault was stuck and stress was concenrated on the fault, while faults in the other areas were slipping quasi-statically. It is inferred from the second result that the normal stress on the fault was reduced by the increase of tensional stress in the north-south direction, and that the fault became easy to slip. The tensional strain in the north-south direction is thought to result from the slow slip on the detachment which exists in the north of the earthquake fault.View full abstractDownload PDF (1511K) -
Yasuo AWATA, Kiyohide MIZUNO, Yuichi SUGIYAMA, Ryusuke IMURA, Koichi S ...1996 Volume 49 Issue 1 Pages 113-124
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSAssociated with the Hyogo-ken Nanbu earthquake of January 17, 1995 (MJMA 7.2), a 10.5km long surface fault ruptures, the Hokudan Earthquake Fault System, appeared along the Nojima Fault and the Mizukoshi Flexure on the northwest coast of Awaji Island, central Japan. The surface ruptures consist of a NE-SW trending right-lateral fault system with high-angle reverse component. The system has two fault strands, namely the Nojima Earthquake Fault of 8.8km long and the Ogura Earthquake Fault of 3.0km long. These two surface faults show the right-stepping arrangement and form an extensional jog at the southwest of the rupture system. Composite displacement of the right-lateral and vertical shift along the fault system are generally 1.4 to 1.8m (1.2 to 1.6m in right-lateral and 0.5 to 0.8m in vertical). The maximum displacement measures 2.5m (2.1m in right-lateral and 1.3m in vertical) at the central part of the surface faults. At the southwest of the surface ruptures, the fault displacement gradually decreases to non-traceable in a distance of 1.9km. On the other hand, surface ruptures reach as far NE as to the coast of Akashi Strait, about 4km west of the epicenter, with large displacement. No surface rupture appeared in the severely damaged Kobe area, northeast of the epicenter. Geodetic and seismological data suggest that the main faulting event of the earthquake occurred in the Awaji Island area, followed by a smaller event (or events) in the Kobe area. Based on the long-term slip-rate of the Nojima fault in the late Quaternary and the amount of slip of the 1995 event, the recurrence interval of large faulting events is estimated to be 1.3 to 2.5 thousand years.View full abstractDownload PDF (2003K) -
Implications for Generation of the 1995 M7.2 Hyogo-ken-nanbu EarthquakeYuji KANAORI, Shin-ichi KAWAKAMI1996 Volume 49 Issue 1 Pages 125-139
Published: May 24, 1996
Released on J-STAGE: March 11, 2010
JOURNAL FREE ACCESSSeveral boundaries of tectonic blocks can be defined as the lines linking large-scale active faults in Southwest Japan. Using available data concerning geology and geophysics, tectonic features of the block boundaries have been briefly reviewed. These features imply that Southwest Japan is separated into a few microplates. Based on the microplate model of Southwest Japan, sequential generation of large earthquakes including the 1995 M7.2 Hyogo-ken-nanbu earthquake can be interpreted. The Hyogo-ken-nanbu earthquake occurred in the middle of the 100-km long Takatsuki-Rokko-Awaji tectonic line (TATL). The aftershock distribution and the surface ruptures suggest that the main shock is caused by a 40-km long rupture of the central part of the TAIL. Thirty-kilometer long eastern and western parts of the TATL remain unruptured. The average moment-release rate of the 40-km central part of the TATL can be estimated from the slip-rate of constituent active faults. The magnitude of an earthquake generated at the central part can be predicted from the seismic moment, which is calculated from the moment-release rate and the elapsed time since the 1596 Keicho-Fushimi earthquake. The calculated magnitude of 6.8 to 7.2 is consistent with the actual magnitude of the Hyogo-ken-nanbu earthquake. If the eastern and western parts unruptured have the same slip-rate as that of the central part, the accumulated moments after the previous event amount to the earthquakes with a magnitude of 6.7 to 7.1. A two-meter right-lateral slip along the Nojima fault, which constitutes western-central part of the TATL, occurred at the time of the Hyogo-ken-nanbu earthquake. This slip may help a southward migration of the Kinki-Outer mircoplate. The movement may reduce normal stresses along the Median Tectonic Line and may increase stress concentration along the Nankai trough. The kinematics of the microplate and sequential generation of large earthquakes along the microplate boundaries suggest a high probability of generating an offshore earthquake along the Nankai trough in a period ranging from 2003 to 2040 A. D.View full abstractDownload PDF (2135K)
- |<
- <
- 1
- >
- >|