Tidal fluctuations of groundwater temperature are recognized at Nishiawakura observation well in Okayama Prefecture, Japan. The tidal fluctuations of the temperature are similar to those of the discharge induced by the earth tides, and a little time lag is recognized. Our model which assumes heat transmission between fluctuated groundwater flow and surrounding rocks can qualitatively explain the observed tidal fluctuations of the groundwater temperature, but the model is not enough to explain the tidal amplitude of the groundwater temperature. A careful investigation of groundwater temperature variation at suitable wells would be useful to detect a crustal strain, which is important for earthquake prediction research.

View full abstract

We present a new ‘bending’ formulation of three-dimensional ray tracing in a spherical Earth. This is suitable for computer programming, and we have already corrected some errors in previously published formulations. By using this formulation we calculate a seismic ray traveling in a velocity model for the subduction zone around the Japan islands. The result is compared with those by appoximate ray-tracing methods, such as Ray Initializer, Circular Ray Tracer and Pseudo-Bending. The model includes the Conrad and Moho discontinuities, and the upper boundary of the subducting Pacific plate. To evaluate the effect of the discontinuities we also calculate exact rays for continuous models derived from the Lagrange or spline interpolation of the discontinuous model. The comparison shows that appoximate ray-tracing methods and smoothed continuous models may lead detailed seismic tomography of the subduction zone to a wrong result. In particular, the discontinuities should be taken into calculations to obtain precise trajectories of rays traveling through a subducting slab.

View full abstract

Uyeda et al., reported that the anomalous changes of the telluric potential observed at Izu-Oshima island using the telegraphic facilities are possibly precursors of earthquake swarms off the east coast of Izu Peninsula. The fact, however, that almost all their ‘signals’ appear from 6 to 22 o'clock implies that those ‘signals’ are of man-made origin. On the basis of the evidences, the author claims that those anomalous changes are not related with earthquakes but are caused by the leak current associated both with the use of the telephone and the precipitation. The ‘fact’ on which their conclusion was based that the number of occurrence of the anomalous changes of the telluric potential a day increases before the commencement of the earthquake swarms is not actual. Furthermore, the fact that we had big earthquake swarms off the east coast of the Izu Peninsula once a year in the rainy season in these three years might have lead them to the fault. Their observational results simply reveal that the noise level of the self-potential at Izu-Oshima Island rose due to the precipitation, as we often usually experience.

View full abstract

We derived three-dimensional P-wave velocity structure beneath the Japanese Islands by a tomographic inversion of travel time data. We used 37, 617 arrival time data from 581 earthquakes observed at 99 Japan Meteorological Agency stations by high-sensitive type seismometers (EMT, EMT76 and OBS) and estimated the velocity anomalies in a number of blocks and corrections to the source parameters simultaneously. In this study we used a block size of 0.5°×0.°×50km (for the uppermost three layers, 0.5°×0.5°×33km).
Down to a depth of about 300km, we obtained detailed features. High velocity anomalies corresponding to the Pacific slab from Hokkaido to Kanto district and the Philippine Sea slab from Aichi prefecture to Shikoku are well delineated. Especially, we found aseismic extension of the Philippine Sea slab beneath the Japan Sea coast of Chugoku district. Low velocity anomalies corresponding to the volcanic front and the wedge portion of the upper mantle are found. We also see low velocity anomalies on the oceanic side of the Pacific slab below the depth of 250km.
In comparison with the result from ISC data obtained by Kamiya et al. in 1989, we got clearer images in this study in the shallow regions down to a depth of about 300km, because of accurate data used in this study. The efficiency of this data set for travel time inversions like as this study is found to be comparable to the Japan University Network Earthquake Catalog data.

View full abstract

The present paper aims to estimate the regional distributions of maximum acceleration or intensity in the main area of Muroran regarding 1968 Tokachi-Oki earthquake as a future coming earthquake. The analyses were made by the method of Kobayashi and Midorikawa whcih first calculated response spectrum and maximum acceleration on the base rock and secondly evaluated response spectrum and maximum acceleration on the ground surface. The underground structure was classified into shallow and deep parts. The deep part, which is typically recognized at Horobetsu area east of Muroran, consists of base rock of Paleogene, four strata of Neogene and andesite of Quaternary. The shallow part, what is called soft ground, is made up of sand, clay, gravel and others. In the main area of Muroran city, especially in and around Muroran bay, the soft layer immediately lays upon the Muroran stratum. The main area of Muroran city, which has a 10km×10km area, was divided into about 340 segments and analyses were made for about 150 segments which include densely populated residential districts and therefore, for which estimations of earthquake motion are essential. S wave velocity and density for shallow parts were obtained on the basis of geological sections of boring or by utilizing a kind of seismic prospecting using surface waves. The deep underground structure was treated as a common one for all segments with the exception of changing the thickness of the Muroran stratum. The regional distributions of maximum acceleration which were obtained by the above mentioned procedures and data were consistent with the intensity distributioins of past earthquakes. The present result is expected to be useful for taking countermeasure against earthquake disasters due to the coming earthquake.

View full abstract

Rapid estimation method of fault parameters is proposed for the 1944 Tonankai earthquake, which has two fault planes with different stress drop in Aida's model. The strong motion accelerograms calculated from the stochastic propagating rupture model are used as the pseudoobserved data. The fault length estimated from them is shorter than that of the Aida's model because of neglected effect of fault with the smaller stress drop. Comparison of results by the tsunami simulation on Aida's and rapid estimation models shows that the tsunami height along the shoreline is affected not by this estimated error but the difference of total amount of sea bottom deformation. The real time forecasting of tsunami warning is performed in the eastern coast of the Kii peninsula very closed to the source region. The time available for numerical computation is thus severely limited. However, the numerical forecasting is possible in time at least by the arrival of maximum tsunami waves.

View full abstract

We compiled data on the direction of the crustal stress measured in-situ and estimated from shallow (depth≤35km) earthquake focal mechanisms in the central and western parts of Honshu. In the central part of Honshu, the stress data were compiled by Tsukahara and Ikeda in 1991. This report summarizes the stress state in a rather large area: the central and western parts of Honshu. The results are as follows. (1) The stress state estimated from the stress direction and the type of faulting in and around the Izu Peninsula are explained in terms of dynamic interaction between the Philippine Sea (PHS) plate and the Asian (ASA) plate. (2) In the central part of the Kanto plain and the Kanto mountains, the minimum horizontal compressive stress (S_Hmin) is equal to or smaller than the vertical stress (i. e., tensile stress), and the azimuthal direction is roughly east-west. This stress condition cannot be explained by the westward movement of the Pacific (PAC) plate or the northwestward movement of the PHS plate against the ASA plate. (3) In the areas of the northern part of Kanto, Shin'etsu, Gifu, Hokuriku, and Chugoku, the azimuthal direction of the maximum horizontal compressive stress (S_Hmax) is west-northwest=east-southeast. Although this direction is approximately the same as that of the movement of the PAC plate, the stress state in these areas is not attributed to the plate movement because (a) some provinces with quite different stress directions lie between the PAC plate and these areas, and (b) S_Hmin in these areas is tensile stress. It seems that the stress state of this area is dominated by a large scale crustal movement of the Japan Sea or Eastern Asia. (4) In the area of Lake Biwa to Osaka, the stress state is slightly different from that in the areas discussed in (3), i. e., (a) the azimuth of the S_Hmax direction is approximately east-west, and (b) the type of faulting is different. There are strike-slip and reverse faults as opposed to strike-slip faulting alone for the surrounding area. The reason why this area has its own unique characteristics is not yet understood. (5) In the central Shikoku region, the azimuth of the S_Hmax direction is approximately east-west and the stress state regime is of the strike-slip faulting type. The east to west S_Hmax azimuth agrees well with the maximum shortening direction obtained from geodetic surveys in the region during these 100 years, which is attributable to the large crustal deformation at the Nankai Earthquake (1946, M=8.1). In contrast, however, the maximum shortening direction after the earthquake has been north to south. Therefore, the east-west S_Hmaxazimuth will change to north-south in the future.

View full abstract

In view of the recent increase of tall or big structures with a natural period of 1 to 10sec, it is one of the most urgent and indispensable subjects to establish the effects of surface geological conditions around a site, i. e. site effects, on long-period strong ground motion. In order to elucidate the site effects at various sites in Japan, we have analyzed strong ground motions during intermediate-depth earthquakes, on which source and path effects may be simple. In this paper, we analyze JMA strong-motion seismograms from four intermediate-depth earthquakes in Kyushu district with a depth of 116 to 130km and a magnitude of 6.0 to 6.8 (JMA magnitude). These seismograms represent the site response to nearly vertical incident plane S waves; they are simply composed of the primary S wave on a rock site, while they show the long-period scattered waves after the amplified S wave on a basin site. These scattered waves cannot be explained by one-dimensional structure model. These facts suggest that scattered waves are excited by subsurface irregularity of structures. We define thee indices I_s, which represent the strength of scattered wave excitation due to incident S waves. Generally, I_s values are relatively large at stations on alluvium or diluvium, while they are small at stations located on a rock site. There exists, however, a wide range of I_s values at stations on alluvium or diluvium. This indicates the variety of the deep subsurface velocity structures around these stations. Unfortunately detailed velocity structures in Kyushu district are not available. Then, we examine relation between I_s values and Bouguer anomalies around the stations, which may reflect the velocity structures near the surface. The Bouguer anomalies around stations having large I_s values show a deep trough with a short wavelength of about 50km.

View full abstract

Spatial and temporal variations of seismicity in the eastern Yamanashi Prefecture during the period January 1, 1980 through September 30, 1990 are investigated in detail on a basis of precisely determined hypocenters and magnitudes. It is suggested that a fracture boundary exists in the Philippine Sea plate extending from northwest of Manazuru headland toward the mid of the eastern Yamanashi seismic zone. Seismicity in the eastern Yamanashi is high on its west side. In particular most earthquakes in the depth range below 20 km occur to the west of the extension of the fracture boundary. On the other hand the August 8, 1983 earthquake with an M6.0 and most of its aftershocks occurred on the east side of the fracture at relatively shallow depth around 20km distributing along an inclined plane to the north. We consider that the cause of the 1983 earthquake was rebounding of the land plate dragged by the subducting Philippine Sea slab in the east side of the fracture boundary. After the 1983 earthquake noticeable changes in seismicity were observed on the west side of the fracture. There has been recognized a seismic gap in the eastern Yamanashi seismic region. A part of the seismic gap was filled with aftershocks of the September 5, 1988 earthquake with an M5.6. In the rest of the gap, whose volume is about 3×4×6 cubic km, seismicity seems to have decreased further since the occurrence of the M5.6 earthquake in 1988, and it is expected that an M5-6 earthquake will occur in the near future to fill the gap.

View full abstract