Classification of low-frequency earthquakes has been made quantitatively by using a diagram of seismic-moment factor M_e versus characteristic period T_c. T_c and M_e correspond to period of corner frequency and seismic-moment density at T_c of each earthquake. About 3, 000 earthquakes from 1926 to 1978 along the Kurile, Japan, and Ryukyu trenches have been analyzed and M_S's of those earthquakes cover the range from 3 to 8.3. Most of earthquakes beneath the inner trench slopes have been classified into low-frequency events which show large T_c for the same M_e, while ordinary earthquakes have been commonly found in the frontal arc regions. Relations among magnitudes and seismic moment: M_S-M_JMA, M_S-m_b, and M_S-M_o, for low-frequency events are also different from those for ordinary events. The relations, therefore, cannot be explained by a scaling model for ordinary earthquakes. A scaling model has been derived for low-frequency earthquakes in a statistical manner, taking into account constraints based on the relations among magnitudes and seismic-moment mentioned above. Corner frequencies of source spectra of the low-frequency model are always one half of those of the ordinary model with the same seismic-moment. This model has been also justified by the data of so called tsunami earthquakes in other subduction regions in the world, suggesting a similarity relation among destructive tsunami earthquakes and low-frequency earthquakes with small magnitude.

抄録全体を表示

Near Hakodate, the southernmost part of the Oshima peninsula, Hokkaido, a noticeable seismic activity of swarm type has continued since October 23, 1978. Up to the end of 1980, there occurred more than 6, 000 earthquakes larger than M 2, including numerous perceptible ones in and near Hakodate. Three maximum events were of M 4.2 and reached seismic intensity III on JMA scale. The activity is divided into two stages: the first from the beginning to the end of 1978 and the second since 1979. Frequency of earthquake occurrence reached the highest peak, more than one hundred per day, in late October, 1978, in the first stage. From a point of view of energy released by earthquakes, however, the most violent activity appeared in the second stage, especially from middle March to May 11, 1979.
In order to locate earthquakes precisely, a temporary seismic network using wireless telemetry was deployed surrounding the focal region. From this observation the following were revealed. In the first stage, earthquakes occurred exclusively off Hakodate (NW region). While in the second stage, earthquakes gave shape to one more focal region (SE region) about 10km southeast from NW region. Most of large events occurred in SE region. Activity in SE region decreased rather monotonously and vanished by April, 1980. On the other hand, activity in NW region has continued in some complicated mode, reviving in August, 1979 and January, 1980.

抄録全体を表示

The turning of the Fukozu fault associated with the 1945 Mikawa earthquake was investigated by the present geological survey. It was ascertained that the fault had two continuous bends designated the first and third bend, and two discontinuous bends designated the second and fourth bend. Consequently the Fukozu fault is considered to be composed of three separated earthquake faults.
Judging from the formation of these faults, their major features may be explained as the results of an east-northeastward upthrusting of three land blocks on the west-southwestern sides of the faults.
These three faults are all active ones and the left-lateral displacement along the WSWENE course of the first fault (from Atsumi Bay to Maeno) has presumably accumulated in the same direction, a total distance being about 450m. Our survey, moreover, revealed that the west-southwestern extension of the WSW-ENE course of the first fault might be probably an active fault.

抄録全体を表示

Active faults in the Kanto Plain have been studied by the geophysical prospecting method, such as the seismic and the gravity measurements. It is revealed that the active fault is accompanied with the basement fault whose displacement exceeds several hundreds meters in its deeper part. This evidence shows that the geophysical method is a usefull method to detect the hurried active fault which is covered by the alluvium. It is concluded that the following order of study, firstly detecting the basement fault by means of the geophysical prospecting method, and secondly carrying out the precise geomorphological, geological and test well surveys, is a practical method for the study of hurried active fault in the alluvial plain. Comparing the surface displacement or the displacement velocity of active fault with the displacement of its basement fault makes it possible to estimate the activity of active fault.

抄録全体を表示

The present paper is a macroscopic and qualitative research for the change of seismicity in and around dam sites before and after the impounding.
Dams sampled in this research are forty three dams with the height of dam crest more than eighty meters where the impoundment started during 1963-1976.
Seismic data used are taken from “The Seismological Bulletin of the Japan Meteorological Agency (January 1926 to October 1980)”.
Seismic activities within the hypocentral distance of fifty kilometers from the centers of dams are surveyed by dividing the periods into two before and after the impounding.
The nature of seismic activity for forty three dam sites can be divided into the following three classes:
(1) Seven dam sites where seismic activity became active after the impounding.
(2) Twenty one dam sites where seismic activity did not change through before and after the impounding.
(3) Fifteen dam sites where the change of seismic activity is hard to be clearly recognized by the impounding.
In some of these dam sites, earthquake swarm followed after the impounding. These earthquake swarms, however, are difficult to be connected to the impounding, because the hypocentral distances are too far as thirty to fifty kilometers. This is a subject to be solved in future.
It has generally been considered that induced seismicity is closely related to geological structure of the dam site. But, the present paper does not discuss about it and left it to a future study too.

抄録全体を表示

An observation was carried out at Osakayama hill, situated in the southwestern part of Shiga Prefecture, to obtain data needed for studying the topographic effects on seismic ground motion. The Osakayama hill exhibits a simple topography extending nearly from north to south and the linearity of the ridge lends itself readily to two-dimensional modeling. Three-component seismometers were installed both at the ridge crest and in the disused railway tunnel running through below the hill, and many records were obtained simultaneously at both observation points from local earthquakes occurring in the Kinki district.
Analyzing the records frequency-dependent amplification of the motion at the ridge crest relative to the motion in the tunnel was observed clearly for the component perpendicular to the trend of the ridge. All the spectral ratio curves estimated have a peak commonly over a narrow range of frequency round 5Hz, with the maximum amplification of about 15. On the other hand, no clear peak was found for the component parallel to the trend of the ridge and the amplification was relatively small. Taking into account these phenomena, the horizontal particle motions were examined with the use of bandpass filtered seismograms, and then a specific feature reflecting the effect of topography was discerned. It was found that the motions at the crest are polarized in the direction perpendicular to the trend of the ridge compared with those in the tunnel, especially in the range of frequency 4-6Hz corresponding to the peak amplification. It was also found that the linearlity of particle motion increase gradually with time. Judging from the fact that all the phenomena mentioned above are observed independently of either azimuth or wave type, they were considered to be caused by resonance of the hill.

抄録全体を表示

Spatio-temporal seismic pattern for shallow and small events which occurred during 1974-1980 in the Shikoku District, western part of the Median Tectonic Line in Japan, is investigated to elucidate similarity of precursory phenomena in seismic pattern between relatively small earthquakes and greater events. The biggest event during 1974-1980 in the district is on Sep. 8, 1977 and its magnitude is 4.6; this was followed by some aftershocks but not preceded by forerhocks in a narrow sense. The main event was obviously associated with “seismicity gap” around the focal zone before 3 months of the event. Size of the gap is 30-40km in diameter. Before the seismic gap, there was “precursory swarm” not only in an adjacent area of the impending main event but also around the gap area. The precursory swarm continued approximately 2 years before the seismic gap. Spatial migration of activity in the swarm area is also seen. In another word, a sort of “seismic gap” before the precursory swarm could also exist. A linearment of seismic activity along a strike of composite focal mechanism solution is found in a period during the time of precursory swarm.
Temporal variation of b-value of the seismic zone is also shown. b-value anomalously increased up to 1.8 during the first half of the precursory swarm period. The high b-value decreased in the last half down to 0.6, which followed the seismic gap that prepared the main event.

抄録全体を表示

Precise relocation and focal mechanism study for shallow earthquakes occurring near the junction between the Kurile and the northeastern Japan arcs during the period from 1979 to October 1981 are made by using 35 stations of the microearthquake observation networks of Hokkaido University, Tohoku University, Hirosaki University, and Central Research Institute of Electric Power Industry. Obtained distribution of earthquakes shallower than 60km shows being in spatial inhomogeneity and in clusters. The 24 seismic clusters may be classified into a group of the inter-plate and one of the intra-continental plate earthquakes.
The P-axes of composite focal solutions for the seven inter-plate seismic groups in the 40.60km depth near the Pacific coast and those of the five intra-plate seismic groups in northern side of Honshu are nearly horizontal and oriented in the approximately normal direction either to the Kurile trench or to the Japan trench axis. On the contrary, the two intra-plate seismic groups shallower than 40km occurring in or close to the arc-arc junction between Honshu and Hokkaido exibit the near horizontal P-axes approximately parallel to the trench axis. Those two P-axes may indicate an anomalous state of stress near the junction.

抄録全体を表示

Geomagnetic total force around the Median Tectonic Line in the north-eastern Shikoku district was observed by the proton magnetometer. The characteristic magnetic total force accompanied with the Sambagawa metamorphic belt, the Ryoke belt and the Median Tectonic Line was observed. The distribution of observed magnetic total force is consistent with the crustal structure around the Median Tectonic Line derived from the explosion seismic experiment in the north-eastern Shikoku district.

抄録全体を表示

The velocity ratio V_p/V_s is examined in relation to the P wave velocity structure or the seismic activities in central Shikoku, Japan. The following four layers distinguished from one another according to seismic velocities have been observed in Shikoku: the superficial, granitic, basaltic and mantle layers. The occurrence of earthquakes has been active mainly in the granitic layer and the uppermost mantle. A ratio V_p/V_s=1.73 is observed for both of these seismically active layers. The V_p/V_s of the ductile region of the mantle where earthquakes have scarcely occurred is inferred to be 1.80 or over, as was expected. A V_p/V_s=1.75 is applied for the basaltic layer, but the value is not necessarily reliable. It is remarkable that a V_p/V_s=1.58-1.65 observed for the superficial layer is unexpectedly low. This low value may reflect the high fracture of rocks and the resultant low seismic activity in that layer.

抄録全体を表示

The velocity distributions of P and S waves in a seismic dipping zone in Hokkaido and Tohoku regions are calculated by travel-time analysis based on a simple ray tracing method. The earthquakes analyzed are located in the middle part of Tohoku and west off Aomori Prefecture.
The velocity in the crust is specified for all the earthquakes in each of three regions; Hokkaido, Pacific coast of Tohoku and middle part of Tohoku, whereas the velocity is assumed to increase as a linear function of depth from the top of mantle toward the hypocenter.
Comparison of the velocity structures between the foci and Hiroo with those between the foci and Kushiro indicates that there are three percent velocity contrast for P waves and two percent contrast for S waves. The rays from the foci to Hiroo pass through a seismic active zone. As to the rays from the foci to Kushiro, however, two thirds of their lengths are below the seismic active zone in the seismic dipping zone.
The velocity structure for P and S waves is proved to be both vertically and laterally heterogeneous in the seismic dipping zone.

抄録全体を表示

Observations of the four permanent ocean bottom seismograph station (TKOBS) operated off the south coast of Tokai area, Central Japan are summarized here. Earthquake detection capability of local earthquake at each station is identical to those of sensitive land station, although wind generating micro-tremor and artificial noise due to ships disturbed the observation sometimes. Events of magnitude 3 (MJ) are detectable at ranges of up to 150km from the system.
Seismograms from near events resemble to those obtained at land station. However, features of seismogram from local events are generally much different. Separation of compressional waves on the vertical component from transverse waves on the horizontal components are so prominent that it is almost impossible to see S phase onset on the vertical record, on the other hand, P phase onset on the horizontal seismogram always indicates delay of few tenth to 2 second.
The seismograms especially from deep focus events are more impulsive for both P and S phases than those obtained at land stations. These features of seismogram imply that crustal structure under the station may be less heteorogeneous than those of land station.
Because of reflection from the sea surface the seismograms are much more reverberatory than those obtained on land-based system. In spite of these differences of seismogram in wave forms, observed wave amplitude is almost same as those obtained at land stations.
The four stations show prominent travel time anomalies up to 2 seconds for both P and S phases. P arrivals at TK1 station are generally 1sec faster than those expected from JMA standard travel time table. As distance from station to the coast becomes shorter, P arrival becomes later and delays about 1 second at the TK4 station. These travel time anomalies may be qualitatively explained by the shape of crustal structure that the Moho discontinuity becomes shallow and thickness of crustal layer decreases as continental crust approach the trench.
Micro-earthquake activities which are not detectable from land stations around the TKOBS are located. Earthquake swarm often occurs around the north eastern margin of aftershock area of Tonankai earthquake (1944) and micro earthquake occurs sometime in the region between the western margin of Nankai-Suruga trough and the TK3 station. Seismicity in the trough is quiescent and no earthquake is observed. On the opposite side of the trough, seismic activity along the Oshima Zenisu ridge, echelon ridge of Izu Ogasawara inner arc, extends to southwest of Nishi Shitito fault zone. Generally, observation confirmed that micro-earthquake activity is very low around the array where large inter-plate earthquake is expected by many seismologists.

抄録全体を表示