Great earthquakes along the Nankai Trough did not occur at random but occur quasiperiodically. The time intervals of their recent occurrences are 90-150 years. As the nearest past earthquakes were the Tonankai earthquake in 1944 and the Nankai earthquake in 1946, about the first third of the interval time have lapsed and now we are entering to the next third term. The “long-term” here used means this second term, i. e., about 40 years from now. The earthquakes which should be predicted are the intra-plate ones with M≥7.3. There were six earthquakes in this category during the four interval times of successive great earthquakes from 1498 to 1944. There were no earthquake in the first third term, two earthquakes in the middle third and four earthquakes in the last third. If the seismicity in the coming second term is the same as that in the past term, the possibility of the occurrence of earthquake is 50% (2 earthquakes in 4 interval times). Moreover, considering the high seismicity during the nearest past 100 years, this possibility may be raised. Next problem is to point out the place of the forthcoming earthquake. The one clue is to use the migration of seismicity and the other is the erasure of faults activated in historical ages, because these activations can be considered as recent ones compared with their reccurrence times of activities. After these operations, the active faults which lie along the border region between Chubu and Kinki districts are selected to be the most probable place which will have the chance of earthquake occurrence. One of the objects of this study is to be confirmed the reliability of the earthquake prediction studies. Therefore, the examinations of the hypotheses on which the conclusions of this paper based are also important. First, the migration patterns of seismicity during every interval times of great earthquakes along the Nankai Trough are generally different. We can say with considerable errors that the migration patterns of successive two interval times seem to be alike and that the seismicity pattern in the past 40 years is resemble to that in the first third of the interval of 1854-1944. Next question is on the degree of contributions of the known active faults to earthquake prediction. The intra-plate large earthquaks in the past 100 years showed that the earthquakes which occurred only at the known active faults were very few and that, in the cases of large majority, new earthquake faults came to besides the known ones. These facts show the limit of the reliability of the predicted results by the topographically known active faults.

View full abstract

Time and space distributions and b-values of the micro-fractures in the mine and acoustic emissions (AE) were investigated and compared with those of earthquake swarms.
The swarms of micro-fractures were observed at the depth of about 400m in the Nakatatsu mine. AE events were generated at the triaxial test on granite under the confining pressure at 173MPa. AE events used in this analysis were observed under the axial stress at 80-90 percent of the compressive strength.
In an earthquake swarm, the seismic energy of the largest shock (E_smax) is in proportion to the cube of the linear dimension of the hypocentral area (L_swm). This relation holds true in AE events. However, E_smax of the micro-fractures in the mine is much smaller in comparison with that inferred from L_swm. The b-values of earthquake swarms are in the range of 0.6 to 2.0. The b-values (b=m-1) of AE events are in the range of 1.2 to 1.7. However, the b-values (b=m-1) of the micro-fractures in the mine are evaluated as 2.53 and 2.29, which are considerably large. These properties of the micro-fractures in the mine and AE events may be explained by the difference of the fracture strength of the medium in which each fracture occurs.

View full abstract

The Shikoku Island is divided into four zones different in geologic nature by the three major faults called the Median-, the Mikabu- and the Butsuzo-Tectonic Lines, respectively. It has become obvious from the precise examination of the focal distribution of earthquakes that the four zones are characterized not only by the geological features but the seismological ones. That is, there exist aseismic zones along the three tectonic lines and the deepest levels of seismically active regions in the crust are discontinuous across the tectonic lines. The aseismic zone of about 2km wide in the northern side of the Butsuzo Tectonic Line approximately corresponds to the Kurosegawa Tectonic Zone. Since this tectonic zone is considered to be a suture zone characterized by the serpentine melange and also to be the transitional zone from the continent to the ocean, this correspondence is noteworthy.

View full abstract

The studies on double-planed intermediate seismic zone beneath some island arcs revealed that the earthquakes have the two different types of focal mechanism solutions: down-dip compression and down-dip extension. These features lead the studies on the stress distribution in the descending plate and several stress sources have been proposed. The purpose of present study is to clarify detailed characteristics of distributions of thermal stress and stress due to the olivine-spinel phase change in the descending plate.
The two dimensional model of descending plate with convergence velocity of 8cm/yr and dip angle of 30° is assumed in the estimation of stress: this model is a simulation of the descending plate beneath the Northeastern Japan Arc. The plate is assumed to be a perfectly elastic body with constant physical coefficients. Under the condition that the displacement at the boundary of plate being free, the thermal stress and the stress due to the olivine-spinel phase change are estimated by the finite element method. The stresses in the descending plate are distinguished by compression at the upper and at the lower parts and by tension at the central part in both the cases. The magnitudes of the maximum principal stresses are several to 10 kbars and the directions of principal axes are almost parallel to the dip of the descending plate. The depth distributions of stresses in the plate are, however, different between the two cases. The higher state of thermal stress appears at the depth range between 70 and 550km, while that of stress due to the olivine-spinel phase change does at the depth range between about 300 and 500km. The present study predicts that the thermal stress induced in the descending plate relates closely to the double-planed seismic activity. The effect of boundary condition that the displacement at the tip of plate is fixed is also examined for the case of thermal stress. In this case, the compressional stress dominates in the whole plate. We infer that the tip of descending plate is not constrainted by the mesosphere.

View full abstract

The movements of the fracture zone (strike N20°E, dip angle 20°W and thickness 1m) in the observational vault at Abuyama Seismological Observatory have been monitored by the two extensometers acrossing this zone. The movements of this zone are analyzed in this paper with special reference to the temporal variation of activity of small and microearthquakes occurring under the fracture zone.
The secular movement of this fracture zone is extension in approximately E-W direction with slightly left lateral component. Before the high seismic activity, however, it returns to contraction with slightly right lateral component. And after this temporal high seismicity, the fracture zone returns to extension.
The principal axis of compressional stress obtained by seismological and geodetical data lies constantly in E-W direction. The change of movement of the fracture zone may reflect the variation of stress in the high seismic layer (depth of 5-20km) under the fracture zone.

View full abstract

Stick-slip is a possible physical model of earthquake faulting, and hence stick-slip experiments can provide much needed information about dynamic processes of earthquake faulting. In the present experiments, dynamic strain and displacement during slip of stick-slip for Tsukuba granite were measured at different locations in the neighborhood of the fault, and typical strain and displacement records are presented in this paper. Local stress concentration resulting in strain change is observed near the fault prior to the onset of foreslip, and this premonitory strain change and the onset of foreslip, often accompanied by foreshocks, propagate slowly along the fault. If this occurs along natural faults, it can be a promising premonitory effect for earthquake prediction. High applied stresses and an increase in the amount of gouge particles contribute an increase in high frequency components of waves generated during slip.

View full abstract

An active earthquake swarm with the largest event of magnitude 4.3 occurred in and near Iwasaki, the west coast of Aomori Prefecture, the northeast of Honshu, Japan since Sept. 16, 1978, being accompanied with precursory earthquake swarm before 3 and 9 days.
Horiuchi et al. (1981) investigated in detail the distribution of hypocenter locations based on the data obtained by a temporary seismic network using wireless telemetry. Following the paper, we investigated time variation of m and b values, focal mechanisms, distribution of total magnetic force and crustal movements.
Total energy of earthquakes discharged by this earthquake swarm was about 4.8×10¹⁸erg. Time variation of b value of NIBETU station indicated meaningful differences before and after the largest event.
Focal mechanism solution of the largest event obtained from P-wave motion was of a reverse fault type with a compressive axis of the northwest-southeast direction, showing the consistency with focal mechanism of inland earthquakes occurring in the Tohoku district.
Survey of total magnetic force showed high anomalies at the epicentral region.
Continuous observation by a bubble type tiltmeter at Fukaura distant about 9km from the swarm region did not detect any abnormal variation.
Results of distance measurements and levelling survey were compared with those computed from earthquake fault models, indicating that observed crustal movements could not be explained by a simple elastic deformation model.

View full abstract

Magnetic data obtained in and around the Bonin region are compiled. Magnetic anomaly on the Bonin Rise shows that the rise was produced at about 19°N in latitude. The distribution of the differentials of magnetic anomalies indicates extension of subbottom structure of the Bonin Rise to the landward region of the Bonin Trench at this area.

View full abstract