The surface wave radiation patterns (amplitude and initial phase) were determined for several of the larger underground nuclear explosions in Nevada. Initial phases of surface waves from explosions and earthquakes were obtained by performing a Fourier analysis of the record and equalizing the phase to the source, using long‐period records from the temporary long‐range seismic measurement stations installed by the Geotechnical Corporation. Rayleigh waves from explosions in tuff and alluvium appear to have nearly the same initial phase at all azimuths, indicating an explosive force acting as a step function in time as the probable approximate source mechanism. However, the presence of Love waves, along with some asymmetry in Rayleigh wave amplitudes, indicates that some asymmetrical forces are also acting at the origin. The collapses following these explosions generate Rayleigh waves with polarities apparently reversed from those of the explosions. The collapses generate much weaker Love waves relative to Rayleigh waves than the explosions, and therefore the Love waves must be generated at or very near the source. The one explosion in granite (Hardhat) gave a surface wave radiation pattern with double‐couple symmetry in both phase and amplitude, a symmetry very different from that found for the explosions in tuff and alluvium. Such a radiation pattern cannot be explained by a simple, symmetric explosive force. Several explanations of this double‐couple symmetry are considered, including the possibilities that the surface waves were generated by cracking or forced motion along pre‐existing cracks, tectonic strain release resulting from cavity formation, or, more probably, triggering of tectonic motion. The latter hypothesis would explain the radiation pattern, and it is supported by the occurrence of an aftershock sequence, outside the cavity zone, similar to that common for earthquakes. Analysis of surface waves from earthquakes indicates that, in many cases, the earthquake sources are more complicated than the explosion sources. The information about source mechanisms obtained in this study is of fundamental importance to the problem of the identification of explosions and earthquakes.

No abstract

The surface‐wave magnitude (Ms) is used as the measure of the relative excitation of long‐period waves by underground explosions and earthquakes in five distinct geographical and tectonic regions of the world: the western United States, the Aleutian‐Kamchatka area, southern Algeria (presumed explosions), central Asia, and Novaya Zemlya. The Ms‐versus‐mb data indicate that the underground explosions generated smaller surface waves than earthquakes from the same region of comparable body‐wave magnitudes (mb). For the events (mb >5) and regions studied, the most significant result is that it is possible to separate the earthquake and explosion populations on the basis of their surface‐wave magnitudes if the events are analyzed on a regional basis. Theoretical considerations of the dimensions of the seismic source in space and time lead to Ms‐versus‐mb relationships that account for the general trend of the observational data. Further elucidation of the diagnostic capability awaits the extension of such studies to events of lower body‐wave magnitudes (mb <4).

Data are presented that address the problem of discrimination between earthquakes and explosions of small magnitude (3¾≤mb≤4¾). Data sources include various long‐period seismometer installations in the United States. The data suggest that long‐period discriminants useful at mb≥4¾ remain as useful at mb≥4. The conclusion is also drawn that detection of the surface wave of mb=4 earthquakes can be accomplished at distances of 6000–7000 km.

Wide‐band long‐period seismographs with peak magnification up to 500,000 at 30–40 sec have been installed in an airtight chamber in a deep mine. The rigid environmental control allows for relatively noise‐free output at high magnifications. The seismograms of the central South Pacific event of August 24, 1968, recorded by these instruments are compared with long‐period seismograms from, the WWSSN instruments at the same site. The high‐magnification systems recorded several body phases and the dispersed Rayleigh waves, whereas the WWSSN instruments recorded only the Rayleigh waves. The body waves on the high‐magnification system enabled the computation of an epicentral distance of 87°±5° for this small‐magnitude (CGS mb = 5.0) event. The surface wave magnitude (M8) computed by using the high‐gain records is 4.4.