Seismic Wave Excitation Research Articles

1983年日本海中部地震の時間依存モーメントテンソル

A large complex earthquake is modeled as a series of multiple point sources. This is to evaluate the time-dependent moment-tensors of complex earthquakes in a subjective manner using the least-squares criterion. Since the earthquake sources are no longer a point, low pass filterting is applied to both synthetic and observed seismograms before the inversion in order to permit a multiple point-source representation. Numerical tests of inversion schemes and least-squares programings have been done by assuming theoretically the time dependent moment tensors. It is shown that the convergence of the solutions is usually achieved within two to three iterations to the assumed moment tensors.The theory is applied to the 1983 Japan Sea earthquake. Total seismic moment of 5.4×1027dyne·cm is obtained. The value is consistent with previous estimates from body-waves, surface-waves and tsunamis. It is revealed that the focal mechanism of the earthquake is time-changing. The focal mechanism solution as a function of time is consistent with the inferred fault plane configurations from aftershocks. A low-pass filtered focal-mechanism thus obtained would represent the mechanism for large scale fault segments. Since much more freedom is expected for the mechanisms of small scale fault patches, the mechanism must be highly at random. Such the fault patches are responsible for the short-period excitation of seismic waves due to large earthquakes.

Excitation of short-period body-waves by great earthquakes

Short-period seismograms from large earthquakes are generally complex and are hard to Fourier-analyze. An innovation of spectral analysis has been made for WWSSN short-period seismograms. The parameters necessary for the analysis are: dominant frequency ƒ 0; maximum amplitude A max; and a characteristic duration σ of complicated wave-packets. Spectrum of short-period seismograms, Y(ƒ), is shown to be approximated as Y(ƒ)=1.07 σ ƒ0 A max The spectrum, after correcting for the effects of geometrical spreading, unelastic attenuation instrumental response, and crustal magnifications, is used to retrieve a short-period seismic moment M 1 of earthquake sources. The method has been applied to more than 900 short-period seismograms from 79 large earthquakes throughout the world. These earthquakes cover the seismic moment, M 0, from 7.5 × 10 29 to 7.5 × 10 24 dyne.cm. M 1 would not be saturated in this range, although body-wave magnitude usually is. A relation of M 0 ∞ M 1 2 is obtained for the present data set. A relation between M 1 and redetermined body-wave magnitude m∗ b is also shown log M 1=1.24 m ∗ b+17.9 where m b ∗ is obtained from maximum amplitudes of P-wavetrains. The analysis method is based on an extension of algorithm of earthquake magnitude determination, however, it has a potential beyond the body-wave magnitude. As M 1 is related closely to the source excitation of short-period seismic waves, it is one of the key parameters in understanding the fracture mechanisms of irregular faults.

Excitation of low-frequency seismic body waves in an underground explosion

Excitation of low-frequency seismic body waves in an underground explosion

The excitation of long period seismic waves by a source spanning a structural discontinuity

Simple theoretical results are obtained for the excitation of seismic waves by an indigenous seismic source in the case that the source volume is intersected by a structural discontinuity. In the long wavelength approximation the seismic radiation is identical to that of a point source placed on one side of the discontinuity or of a different point source placed on the other side. The moment tensors of these two equivalent sources are related by a specific linear transformation and may differ appreciably both in magnitude and geometry. Either of these sources could be obtained by linear inversion of seismic data but the physical interpretation is more complicated than in the usual case. A source which involved no volume change would, for example, yield an isotropic component if, during inversion, it were assumed to lie on the wrong side of the discontinuity. The problem of determining the true moment tensor of the source is indeterminate unless further assumptions are made about the stress glut distribution; one way to resolve this indeterminancy is to assume proportionality between the integrated stress glut on each side of the discontinuity.

On the excitation of seismic waves by explosions in soft soils

On the excitation of seismic waves by explosions in soft soils

Excitation of seismic surface waves with periods of 15 to 70 seconds for earthquakes and underground explosions

Using data from high-gain seismographs operating in the deep (543-meter) mine observatory at Ogdensburg, New Jersey, we analyzed the excitation of long-period surface waves by earthquakes and underground explosions for four different regions of the world: western United States, the Aleutians, Novaya Zemlya, and Central Asia. The most important result of this study is that discrimination between earthquakes and underground explosions on an Ms-mb basis is enhanced when the amplitudes of Rayleigh waves with periods near 40 sec, rather than 20 sec, are used to determine Ms. A discriminant based on surface waves with periods near 40 sec is particularly advantageous because of a very stable and pronounced minimum in earth noise in this period range. Love waves also yield a distinct separation between earthquakes and explosions on the basis of Ms (Love) at 20 and 40 sec with mb and can be used for discrimination of small events for which no Rayleigh waves are recorded. The discrimination threshold at Ogdensburg is mb′≃3.8 for the western United States at a distance of about 30° and mb≃4.4 for the Aleutians at 70°. All the events studied could be discriminated by using 40-sec Rayleigh waves. Amplitude spectra of Rayleigh waves in the period range 15 to 70 sec were computed for events in the western United States, the Novaya Zemlya region, and the Aleutians. The spectral amplitudes decrease more rapidly with increasing period for explosions than for earthquakes in the western United States and Novaya Zemlya. These spectral differences are consistent with an impulsive source time function for explosions. In the Aleutians, however, the shape of the spectrum of long-period waves for the underground explosion Milrow is similar to spectral shapes for some, but not most, nearby earthquakes. Two plausible explanations for this spectral similarity are contamination of the Milrow spectrum by earthquake-generated Rayleigh waves, or the effects of focal depth and radiation pattern on the earthquake spectra.

The Relative Excitation of Seismic Surface and Body Waves by Point Sources

Summary Experimental evidence shows that earthquakes can be differentiated from explosions on teleseismic records by their relatively greater amplitude of Rayleigh waves compared with the amplitude of the direct P-wave; in effect the ratio of magnitude estimates mb/Ms, is greater for explosions than for earthquakes. In this paper we have tried to assess the contribution of source mechanism alone to this effect. We have used point dilatation and double-couple sources to represent explosions and earthquakes and derived the amplitudes of the waves generated from such sources in a uniform half space. The results show that source mechanism is an important factor governing the value of mb/Ms.

Excitation of seismic surface waves by atmospheric nuclear explosions

The Rayleigh waves generated by nuclear explosions in the atmosphere and by earthquakes are investigated to determine the relative effects of the source and the propagation medium on the observed amplitude spectrums. It is found that when the waves from a given source propagate along different paths the shapes of the spectrums change markedly. Comparison of earthquake- and explosion-generated waves leads to the conclusion that the source characteristics are obscured by the propagation effects. The theoretical expression for the spectrum of the Rayleigh wave displacement is derived for an explosion in a homogeneous atmosphere. The yield of the explosion introduces a frequency-dependent factor into the equations. Using this frequency dependence, the relative yields of explosions are computed from the amplitude-spectrum ratios of Rayleigh waves recorded at a distant station.

Relative excitation of surface waves by earthquakes and underground explosions in the California-Nevada region

The results of a study of the relative excitation of long-period and short-period seismic waves by earthquakes and explosions in the California-Nevada region are presented. The Richter magnitude M, which is based on the amplitudes of short-period waves recorded at sites less than 600 km from the epicenter is used as a measure of short-period waves. A new parameter AR—the sum of the areas of the envelopes of the surface waves on certain three-component long-period instruments—is defined and used as a measure of the long-period waves. The surface wave parameter was determined from long-period seismograph records from Pasadena and Berkeley, California, Ruth and Reno, Nevada, and Palisades, New York. The four West Coast stations provide good azimuthal sampling of most of the events studied. Surface waves recorded at Palisades from the Fallon-Stillwater and Fairview Peak-Dixie Valley aftershock sequences were also studied. There is a wide range in the excitation of surface waves for a given Richter magnitude, both for explosions and, especially, for earthquakes. For any given earthquake there may also be a wide range in the excitation of surface waves as a function of azimuth, an indication of an asymmetrical radiation pattern. Generally speaking, the relative excitation of surface waves by explosions is much smaller than by earthquakes of the same Richter magnitude. Most of the earthquakes studied generated surface waves 5 to 10 times greater than the maximum observed for explosions of the same Richter magnitude. A comparison of the relative excitation of surface waves in different regions indicates a very large regional variation. Shocks off the coast of northern California generate relatively large surface waves, whereas shocks near Laguna Salada in Baja California often generate very small surface waves. An empirical curve is derived for the variation of the surface wave parameter AR as a function of distance. A formula for estimating the surface wave magnitude MS is derived from the parameter AR.