On the basis of popular reports of animal agitation prior to earthquakes, a semiquantitative study has been carried out of the possibility of the emission of acoustic strain and fracture radiation prior to earthquakes. Such emission has been observed in the laboratory and in mines, and the bearing of these reports on the animal behavior reports is discussed. The comparative auditory/vibration response of humans and animals is discussed insofar as quantitative data are available.
Theoretical estimates are derived of the frequency of the sound emitted by preliminary fracturing in a highly stressed region. The results for rockburst conditions in mines are consistent with the frequency spectra of the observed signals. Application to earthquake conditions indicates some emission should occur over high audible to low ultrasonic frequencies. Estimates of the intensity of such emission are also obtained to ascertain if it might be detectable on a practical basis. These results lead to transmission range estimates which, for rock Q-values at normal pressure, indicate a very short range with at best a marginal possibility of detection under certain very favorable circumstances. Since, however, rock Q-values are known to increase with pressure, it is pointed out that this may lead to a substantial increase in local transmission ranges within regions of high stress. This effect might permit the practical observation of preliminary emission for shallow earthquakes whose strain fields extend to the surface.