The first arrival data from the explosions of 1968 in the sea off Cape Erimo and of 1969 in the sea off the Shakotan Peninsula and at the Teine quarry, have been analyzed to derive the crustal structure in the profile between Cape Erimo and the Shakotan Peninsula in the southern part of Hokkaido, Japan. In these explosions, the fixed-array moving shot point technique was applied and the records were obtained from observing stations on land with magnetic-tape-recording systems.
Most of the records from the shots off Cape Erimo were not used for the derivation of the crustal structure. The quality of records from the shots off the Shakotan Peninsula was excellent at the stations northwest of the Ishikari-Yufutsu plain, but it was fairly poor at the other stations. The time-term method was applied only to data of the highest possible quality, so that the crustal structure was obtained only in the northwestern half of the profile. Under the explosion site in the Sea of Japan, the crust, consisting of granitic and basaltic layers with nearly the same thickness, is about 17km thick at a point about 50km distant from the shoreline. The crustal thickness increases abruptly toward land and reaches about 30km at a point 40km distant from the shoreline, where the basaltic layer is about twice the thickness of the granitic layer. The velocities obtained in each layer in the crust and in the upper mantle are fairly low, but the gravity anomalies calculated by assuming the densities expected from these velocities agree well with the observed gravity anomalies.

View full abstract

The travel times of P waves to Japan from the underground nuclear explosion of November 6, 1971, at Amchitka Island, the Aleutian Islands, deviate systematically with region in comparison with the travel times by CARDER et al. (1966). Arrival times at stations on the Pacific coastal region from Hokkaido to the northern Kanto area, the northern part of west Japan, and at islands near Japan (Izu Is., Ryukyu Is. and islands in the Sea of Japan) are somewhat earlier than the Carder et al. times. On the other hand, at stations in the southern Kanto, southern Chubu, central and southern Kinki, Shikoku and Kyushu areas, the arrival times are later as a whole.
The body wave magnitude MB, estimated from the data in the Japan Meteorological Agency network is larger by about 0.7 than that reported by USNOAA. It is noticed that the discrepancy between them is probably caused by the difference of the frequency responses of seismographs used for the estimation.

View full abstract

In order to study the dependence of a seismic spectrum on earthquake magnitude, spectral analysis for forty-three earthquakes is made by using an analog band-pass filtering method. A number of event pairs from nine seismic regions located in continents and island arcs are selected, and spectral ratios for two earthquakes from the same epicenter region but with different sizes are studied. The spectral ratio differs among the epicentral regions. Most of the results are generally in good agreement with the ω-square model proposed by Aki, but, some of them show ω⁰ or ω^-1-dependece. This result is consistent with the observed regional variations of the mean Ms-mb, which was taken for the events that occurred for about four year. Though regional variations of the spectral ratio are not definitely detected, it is tentatively proposed that the main factor causing the difference may be due to the difference in the source time function among the event pairs.

View full abstract

The propagation of SH elastic waves is studied assuming a surface source of disturbance. The medium is assumed to be a half space with a free crack which extends from the surface to a finite depth. The method is based on the numerical simulation using the finite difference technique, and the interception effect, or in other words the weakening of wave energy, by the crack is studied by the use of Fourier transform. If the correspondence between the SH wave field and that of the sound wave is considered, the same result is interpreted as the interception of the sound by a rigid wall, and that will be an important subject in big cities suffering from noise problems.

View full abstract

Expressions for surface displacements due to a fault in a multi-layered medium were derived in Part I of this paper. In Part II, numerical approach to calculate theoretical seismograms in the near field is prsented with some examples.
Further extension of the problem for a two-dimensional moving fault will be studied in the forthcoming paper.

View full abstract

An earthquake source model studied in this paper predicts a higher value of P wave corner frequency than S wave corner frequency; the ratio of P to S wave corner frequency is about 1.3 on the average. This result owes mainly to the slip characteristics on the fault, that is, the center of the fault slips for a longer time than the edges and consequently a greater relative displacement takes place near the center. Relationships of source dimension with the corner frequencies for P and S waves are derived to estimate the source dimension of earthquakes from teleseismic body wave spectra. The far-field spectra from the present model of equidimensional rupture propagation demonstrate the spectral decay of ω^-2 at high frequencies. The seismic efficiency is found to be independent of the size of source dimension and expressed as a function of initial stress, stress drop, and rupture velocity. Being interpreted in terms of this model, the ratio of frictional stress to final stress can be estimated from observations of rupture velocity.

View full abstract

P-waves from the CANNIKIN explosion were recorded at sensitive stations densely distributed in Japan. The distribution of the travel-time residuals suggests a distinct heterogeneity of the upper mantle associated with the underthrusting of the Pacific and Philippine plates, The negative anomalies in Hokkaido are most prominent in the south and are reduced to normal in the north, making a remarkable contrast with the positive anomalies near the subduction region of the Philippine plate. These contradictory distributions may be caused by the predominances of the inner and outer low-Q zones beneath Hokkaido and beneath Southwest Japan, respectively. A narrow belt of negative anomaly is apparently developed from the central part of Japan to the west along the latitude of 35.5°N, which might be explained by the particular deformation of the plate near the junction of North Honshu and Izu-Bonin arcs.

View full abstract

Theoretical displacement seismograms are computed for the study of group and phase velocities of Love type surface waves in a dipping layer overlying an elastic body using the solution derived in a previous paper by the present authors. The result does not show that the group velocity in a dipping structure corresponds to the theoretical group velocity of horizontal structure having the thickness equal to an arithmetic average of depths between the source and receiver. The Sato's formula is found to explain the group velocity obtained by the analysis of theoretical seismograms for the short period waves in the cases of dip angles smaller than 2.0°, his formula being derived under the assumption that the phase velocity depends only upon the thickness at an observation station. The analyzed phase velocity agrees well with the phase velocity for horizontal layer with the thickness at an observation point for the short period waves when the dip angle is small.

View full abstract

The nuclear explosion Cannikin in the Central Aleutian islands detonated in 1971 was observed at about 150 seismometrical stations in Japan. These data were used with the purpose of investigation of lateral inhomogeneity in the upper mantle associated with the Japan Island arc by means of a three-dimensional-ray-tracing method recently developed. The upper mantle was divided into three zones in terms of velocity contrast based on the work by Utsu; 1) the low velocity zone, uppermost 275km in the mantle, on the continental side of the inclined deep and intermediate earthquake zone, 2) the high velocity inclined seismic zone of 100-km thick reaching to a 350-km depth, 3) the low velocity zone, uppermost 350km in the mantle, on the Pacific side of the inclined seismic zone. The best-fit velocity contrasts in each zone to a standard earth model were -4, +3, and -1% for the above mentioned zones 1), 2), and 3), respectively. This model is in good agreement with the observed P-wave traveltime anomalies in north-east Japan, however, inconsistent in detail with the observation in south-west Japan, although it explains a general trend of the observed traveltime anomalies in whole Japan. Suitability of the present model is discussed comparing with tho observation.

View full abstract