The survey of underground water was carried out at Yasutomi near Yamasaki fault in the western region of Hyogo Prefecture.
The purposes of the survey are the identification of the fault by underground water, investigation of distribution of the water and to make clear the chemical composition. The fault is expressed by sudden fall of water level. At the northern side along the fault the water head is higher than expectation. And this phenomenum is observed at subfault. Those phenomena are caused by gauge clay of the faults which plays as barrier to movement of the water. In the area between the main and subfaults the water level is low and stable. It shows this zone is composed of crashed rocks, and easily removes and accumulates the water.
The water above the fault has high values of pH, temperature and HCO₃^- content.

抄録全体を表示

As the first step of computer-based automatic processing of short-period seismic signals, distinctive features of signals and noise series must be well recognized. A parametric model of autoregressive (AR) process of the second order is successfully applied to microearthquake signals and noises in order to extract the features in frequency domain. Characteristic differences between P-waves, S-waves, coda waves and noises are well described by the two autoregressive coefficients (a₁, a₂) of the second order and residual variance (σ_Z²). The time evolution in terms of the estimated autoregressive coefficients (â₁, â₂) shows anticlockwise turn along the elliptic locus when seismic signals are added to noise series. It is proved that this evolution indicates the variation of temporal peak frequency vs. power diagrams showing roughly dynamic power spectrum of the events. On the basis of temporal variations of the autoregressive coefficients, we may discriminate between P-waves and S-waves in the seismic traces.

抄録全体を表示

For intermediate-depth earthquakes occurring in the eastern Hokkaido, two remarkable phases with sharp onsets and large amplitudes are frequently recorded on the seismograms at station KMU. The station KMU, one of the nine telemetering seismic stations of Research Center for Earthquake Prediction of Hokkaido University, is located on the western side of the Hidaka Mountain. The two remarkable phases are referred to as the K₁ and K₂ phases. The K₁ phase arrives in 1.0-4.9sec after the direct P arrival and the K₂, in 1.7-9.2sec after the direct S arrival. The deep structure under the Hidaka Mountain is inferred from analyses of wave form and travel times of these phases.
The principal component analysis of seismic waves shows that the prevailing wave motion of the K₁ phase is the longitudinal mode while that of the K₂ phase is the transverse mode. The analysis also shows that the K₁ phase arrives with much smaller incident angles than for direct P wave. The predominant frequencies are in the range of 5-10Hz for the K₁ phase, and in the range of 1-5Hz for the K₂ phase. These values are smaller than those for the preceding direct waves. The K₁ and K₂ phases have longer duration than the preceding direct waves. From the characteristics obtained above, the K₁ phase is interpreted as a reflected P wave and the K₂ phase, as a reflected S wave. Travel time analysis of the K₁ and K₂ phases shows that the reflector exists at depths of about 60km just beneath the station KMU and dips toward the northeast with the surprisingly large dip angle of about 40°. The reflector may correspond to part of the descending lithosphere.

抄録全体を表示

The structure of the lower crust and the uppermost mantle in Shikoku is examined mainly by using the apparent velocity of seismic P waves. The inclination of the layers is nearly perpendicular to the trend of the tectonic lines just as in the upper crustal layer. The correspondence of the structure to the focal distribution of earthquakes and the tectonic lines is suggested. The proposed structure is as follows. The Conrad discontinuity is estimated at about 15 and 8km in depth under the Mikabu and the Median Tectonic Lines, respectively. The basaltic layer becomes thinner southwards to the ocean and disappears in the vicinity of the Butsuzo Tectonic Lines. The Moho discontinuity is over 35km deep in the innerland area, but rises to the depth level of about 23km with the disappearance of the basaltic layer in the oceanic side. Consequently, the continental basaltic layer is separated from the oceanic one by the mantle which rises steeply to the depth level of about 23km. It is supposed that the P* phase is difficult to be detected as the first motion because of the reduced energy of the seismic waves propagating along the Conrad. This reflects a characteristic feature of the continental basaltic layer.

抄録全体を表示

On January 13, 1979, a shallow earthquake of magnitude 4.2 after the Research Center for Earthquake Prediction of Hokkaido University (hereafter called RCEP) occurred off Mukawa, southern Hokkaido. Some distinguished foreshocks and aftershocks with relation to this event were observed.
It is a remarkable event because in this region such shallow events have seldom occurred by this time.
After two days of the main shock a seismological observation had set up at Mukawa by means of a cassette tape recorder with a tape speed of 0.10mm/sec. From the view point of disasters, seismic intensity of the main shock was also investigated in the whole of Mukawa town.
From the above detailed seismological survey data, several important facts were found as follows.
1. Some stages of seismic activities were suggested. In the first stage, four small events took place in succession in the furthest places off Mukawa. This activity was completed in a relatively short span of time of half an hour. After a following aseismic period of seven hours, the dramatic main stage appeared near the southeast rim of the source region of the impending main event. At this second stage, two extremely large main event occurred successively in mere two minutes. After that, such a phenomenon of occurrence of earthquakes in a limited region and in a short time interval was sometimes recognized in a vaguely sense.
2. Focal mechanism solution of the main event suggested that type of fault movement was a left-lateral strike slip pattern as well as other shallow events occurred previously in Hidaka district. The fact was attributed to the field of east west trending compressional shearing stress in the whole of Hidaka district. It may be said in this connection that a fault plane with dip direction of 311.4° measured counter-clockwise from the north is in accord with a general direction of source area implied by epicentral distribution of a series of small events with relation to this main shock.
It is furthermore consistent with an edge of “old downed delta” [MOGI (1964)].
3. Based on the results of error analysis in hypocentral parameters with the prediction analysis after WOLBERG (1967), the standard errors of hypocentral coordinates due to random errors in the data and model parameters of velosity structure may be several kilometers at most.
4. Micro-seismic zoning map of earthquake intensity for the main shock suggested that local difference of seismic intensity depends upon variety of geological structures in places.
5. A T-phase, traveling through the ocearn with the velocity of sound in sea water was found on seismograms of coastal station ESH, which is one of RCEP stations. The path of propagation from the epicenter to ESH is mostly oceanic.

抄録全体を表示

A simple method is developed to determine seismic moments of earthquakes by using tabulated data in usual seismological bulletins. The method is qualified through the criteria such as simplicity of calculations, coverage of wide magnitude range, and insensitivity of the instrumental response: At first, characteristic period T_c of each earthquake is defined as the average value of apparent periods of wavelets which give maximum amplitudes of ground displacement at epicentral distances between 200 and 700km. Secondly, amplitude information is taken into consideration, making a product of maximum amplitude, its period and epicentral distance. Seismic-moment factor M_e for a given earthquake is defined at the characteristic period T_c as the average value of those products evaluated from horizontal components at stations within epicentral distance range from 200 to 400km. The narrow range of epicentral distance in evaluating M_e is taken so as to reduce the uncertainty due to seismic-energy attenuation into a permissible range and to be able to obtain equal number of observations for small earthquakes to that for large ones. The relation between the seismic-moment factors and characteristic periods for 163 intraplate earthquakes in Japan from 1926 to 1977 clearly demonstrates that M_e is proportional to the cube of T_c. A scaling model of earthquakes that satisfies the empirical relations among surface-wave magnitude, JMA magnitude, and body-wave magnitude facilitates the estimate of static seismic-moments from calculated M_e's. The seismic moments of 16 earthquakes determined by conventional analyses from near- and/or far-field observations are consistent with static seismic-moments thus estimated. This shows the potential in practice of the present method especially in routine processing of seismic data.

抄録全体を表示

As one of the series of field surveys on human response during and after an earthquake, an interview method was introduced and carried out at the occurrence of the 1978 Miyagiken-oki earthquake for elucidating details about time-dependent behavioral patterns under drastically changing environments.
A group of housewives living in apartment houses of the identical housing plan was chosen in Sendai city so as to reduce number of spatial factors which may diversify human behaviors. Behaviors and changes of indoor circumstances such as by scattering and overturning furniture were interviewed in detail. Number of children in home, fire sources in use and so on was noted. Simultaneously, a questionnaire survey for estimating seismic intensities was performed to the housewives and also to the inhabitants of the surrounding area.
Results obtained are as follows.
(i) Behaviors which range widely from very passive and do-nothing situation to restless movement, can tentatively be classified into 5 typical patterns in terms of 3 unit processes defined as single action, point-to-point movement, and stand-by posture.
(ii) Seismic intensity is higher by 0.5 on the 5th floor than on the 1st floor, and accordingly overturned furniture are larger in number on upper floors. For better understanding of occupants' behaviors, however, another factor such as distance to the ground surface should be considered.
(iii) Housewives behave more actively and restlessly, under the circumstance that there are children to be protected and/or fire sources in use. This was proved by counting number of unit actions and by measuring totally moved distance during shaking.

抄録全体を表示

Volcano Usu, which had been in a dormant state for 32 years since the formation of Showa Shinzan lava dome in 1943-45, became active beginning with local earthquake swarms and the following sequence of eruptions in August 1977. The volcanic activity has caused the remarkable upthrust deformation at the center of summit crater, accompanying numerous earthquakes. The uplift amounted to about 140m in relative height at the end of 1978. More than 150, 000 earthquakes including about 18, 000 events felt by inhabitants of Sobetsu Hot Spring Town (Δ=2-3km) had occurred by December 1978. In order to study the characteristics of this swarm activity, space-time-magnitude distributions are investigated in detail. The determination method of earthquake magnitude using entire coda amplitude traces is introduced for the 1023 data in August 1978. It is found that this method is powerful to estimate earthquake magnitude in wide range, when we can not utilize the maximum amplitude and/or the F-P time due to the inevitable continuous occurrence of a swarm.
The majority of the earthquakes is located within the small summit crater (1.7 by 1.8km), forming several clusters which encircle the major upthrust area. Distribution of shallow earthquakes (h<1.0km) in August 1978 shows a significant correspondence to the U-shaped zone of major fault traces which outline the upthrust area. Large earthquakes (M>3.8) are mainly located near the two corners of the U-shaped zone of major fault traces. Distribution of these large events are consistent with that of the localized stress concentration caused by a magma intrusion. In the period of October 1977 to February 1978, large earthquakes clustered mostly near the southeast corner, where the fault cut Ogariyama cryptodome and Ousu lava dome. This implies that these lava domes play an important role such as a barrier which can accumulate considerable stress in the upthrust movement. In July and August 1978, large earthquakes occurred near the northwest corner. This phenomenon may be caused by the stress redistribution due to the change of fracturing process in the development of the doming activity, and due to the eruptions from ‘Ginnuma Crater’. Magnitude-frequency distribution of earthquakes is quite distinct from cluster to cluster. In some cases where large earthquakes are clustered, small events are quite few in number and hence the distribution shows a remarkable peak. The considerable departure from Gutenberg-Richter's distribution may be interpreted by the concept of scattered barriers in different sizes and different strengths in the extraordinary heterogeneous medium beneath the summit crater.

抄録全体を表示