Seismic return motion

Possibilities of backswing of fault displacement are investigated on the basis of a simple crack model. Although it has been suggested that the backswing would not occur in real earthquakes because of friction, there is evidence that some earthquakes might have caused return motion “locally” on the fault surface. This paper attempts to derive conditions for such local return motion. Spatial extension of the crack beyond the stress-filled region into the stress-deficient region gives a favorable site for the return motion, the site being just outside the stress-filled region. In our model the return motion occurs at this site when the frictional stress is less than about 30% of the regional stress in the stress-filled region. The overall return motion, on the other hand, would not occur unless the frictional stress is less than 10% of the regional stress.     

Incessant excitation of the Earth's free oscillations: global comparison of superconducting gravimeter records

Records of superconducting gravimeters (SGs) at Canberra (Australia), Esashi (Japan), Metsähovi (Finland) and Syowa Stations (Antarctica) were analyzed to search for further evidence of background free oscillations of the Earth. Spectrograms for 1-year period and averaged power spectra for seismically quiet periods were obtained for each of the stations. Anomalous features of the oscillations observed at Syowa Station, such as an apparent seasonal variation and a high intensity at frequencies between 3 and 4 mHz, were absent at the other SG stations. Among the SG stations used in this study, the background free oscillations were detected most consistently and distinctly at Canberra, where the noise level was comparable to that at the IDA quietest station, while that at Syowa Station was close to the critical limit for detecting the oscillations. The background free oscillations provide a good reference to evaluate the noise level in the milliHertz band.     

Letter to the editor Complete maps of the aspect sensitivity of VHF atmospheric radar echoes

Using the MU radar at Shigaraki, Japan (34.85°N, 136.10°E), we measure the power distribution pattern of VHF radar echoes from the mid-troposphere. The large number of radar beam-pointing directions (320) allows the mapping of echo power from 0° to 40° from zenith, and also the dependence on azimuth, which has not been achieved before at VHF wavelengths. The results show how vertical shear of the horizontal wind is associated with a definite skewing of the VHF echo power distribution, for beam angles as far as 30° or more from zenith, so that aspect sensitivity cannot be assumed negligible at any beam-pointing angle that most existing VHF radars are able to use. Consequently, the use of VHF echo power to calculate intensity of atmospheric turbulence, which assumes only isotropic backscatter at large beam zenith angles, will sometimes not be valid.     

Gravity wave intensity and momentum fluxes in the mesosphere over Shigaraki, Japan (35°N, 136°E) during 1986–1997

Averaged seasonal variations of wind perturbation intensities and vertical flux of horizontal momentum produced by internal gravity waves (IGWs) with periods 0.2/1 h and 1/6 h are studied at the altitudes 65/80 km using the MU radar measurement data from the middle and upper atmosphere during 1986/1997 at Shigaraki, Japan (35°N, 136°E). IGW intensity has maxima in winter and summer, winter values having substantial interannual variations. Mean wave momentum flux is directed to the west in winter and to the east in summer, opposite to the mean wind in the middle atmosphere. Major IGW momentum fluxes come to the mesosphere over Shigaraki from the Pacific direction in winter and continental Asia in summer.     

Re-connexion of magnetic lines of force: Evolution in incompressible MHD fluids

Time-dependent incompressible MHD solutions in two dimensions are obtained numerically to study the evolutionary process involving a re-connexion of magnetic lines of force. Given an initial antiparallel magnetic field, or a current sheet, to which there is an injection of fluid in a transverse direction, we seek to see how the process of re-connexion builds up. In this numerical experiment, special considerations are given to the confirmation of reconnexion, the formation of X-type magnetic field, the speed of growth, conditions that control the evolution, acceleration of particles, the structure of the diffusion region and so forth. The findings are: magnetic lines of force can re-connect and grow to the X-type configuration in fluids of any finitely large hydromagnetic and hydrodynamic Reynolds numbers; the conditions local to the neutral point are less important than the boundary conditions that set up global flow patterns; acceleration of fluid in bulk only concerns whether the X-type configuration grows to the comparably large extent or not; the electric field at the neutral point due to the rapidly changing magnetic field is less efficient in accelerating charged particles.     

Stress drops, wave spectra and recurrence intervals of great earthquakes — implications of the Etorofu earthquake of 1958 November 6

Summary . The great Etorofu earthquake of 1958 November 6 is characterized by a relatively small aftershock area (70 × 150 km²) and an extremely large felt area. The felt area is more extensive than those of any other large earthquakes which have occurred in the southern Kurile to northern Japan arc since the beginning of this century. The mechanism is a pure thrust fault typical of most great earthquakes in island arcs. A body wave magnitude of m _b = 8.2 is obtained at periods around 6 s using more than 40 observations, although an m _b value of only 7.6–7.7 would be expected empirically from the observed surface wave magnitude of M _s= 8.1–8.2. Both an unusually large felt area and a high m _b indicate a dominance of high-frequency components in the seismic waves. A seismic moment of M _o= 4.4 × 10²⁸ dyne cm is determined from long-period surface waves from which a high stress drop of Δσ = 78 bar is obtained using a relatively small aftershock area. Historic data indicate an anomalously long time interval between the 1958 event and any earlier great earthquake from the same source region. The observed high stress drop can be interpreted as a consequence of this long intervening period through which strain built up. The dominance of the high-frequency seismic waves can then be interpreted as a result of this high stress drop. Stress drops, seismic wave spectra and recurrence intervals of great earthquakes are in this way closely related to each other. The 1958 event may represent a high strength extreme of stochastic fluctuation of fracture strength relevant to great earthquakes.