Since June 1973, the California Institute of Technology Seismological Laboratory has been monitoring quarry blasts in southern California for the purpose of detecting possible velocity changes before earthquakes. On June 1, 1975, an ML = 5.2 earthquake occurred near Galway Lake, about 60 km southeast of Barstow, California. On November 15, 1975, and December 14, 1975, ML = 4.7 earthquakes occurred about 30 km southeast of Galway Lake near Goat Mountain. These three epicenters are close to Hector and Victorville quarries, which have been monitored by CIT.
First-motion data, the distribution of aftershocks, and ground breakage associated with the Galway Lake earthquake indicate right-lateral strike slip on a fault striking N20°W, dipping 70°SW. First-motion data and the distribution of aftershocks for the first Goat Mountain earthquake indicate normal dip slip on a plane striking north-northeast, dipping about 60° to the west-northwest.
Blasts at Hector and Victorville quarries were timed with an accuracy of ± 0.01 sec, and first arrivals at a number of stations of the USGS-CIT network can be read to an accuracy of ± 0.02 sec. The data are plotted in terms of residuals versus time at each station in such a fashion as to reflect trends in velocity. Origin times of all earthquakes ≳ 4.0 in our study area are plotted on these curves.
The most important results of this study are observations that are “negative” in character. These observations are: (1) no changes greater than about 0.1 sec (or about 1 per cent in average velocity) are seen at any station during the 2-year period of this study, (2) given the flatness of the curves, it is difficult to draw correlations between any larger earthquakes and changes in velocity. In particular, no unique change is seen before the Galway Lake earthquake along two paths that cross the epicentral region of this earthquake at right angles to each other. The data are such that only an anomaly less than 2 months in duration could have escaped detection. Similarly, no unique change is seen before the Goat Mountain earthquakes along two subparallel paths through the epicentral area. Only an anomaly less than 1 month in duration could have escaped detection.
One observation that is “positive” in character can be made from the curves; namely, slight but systematic changes in velocity can be seen. For Hector blasts, most stations show a systematic increase in velocity with time of as much as 0.8 per cent. For Victorville blasts, most stations show an opposite trend.
The results of this study are somewhat disappointing from the point of view of the standard dilatancy model, which predicts a 10 to 20 per cent decrease in P velocity over an area of several source dimensions in diameter before an earthquake. Before the ML = 5.2 Galway Lake earthquake, this decrease should occur over an area about 30 km in diameter over a period of 3 to 6 months. Before the Goat Mountain earthquake, this decrease should have occurred over an area about 20 km in diameter over a period of 2 to 4 months. Our data preclude the possibility of precursory changes this large before these earthquakes. It is still possible that dilatancy accompanied these earthquakes, but the effect must have been small. It is also possible that these earthquakes are not representative of other ML = 4.7 to 5.2 earthquakes; however, at least two different types of faulting are represented, namely strike slip and normal faulting.
The small systematic changes in velocity that are seen may have one of the following explanations: (1) there were systematic variations in local delays at the two quarries, or (2) there were regional changes in crustal velocity. The fact that shot points migrated in more or less systematic fashions in both Hector and Victorville quarries during the period of this study suggests that the first explanation may be correct. The second explanation is intriguing, but the opposite trends for the Hector and Victorville data are somewhat puzzling, unless adjacent regions, one surrounding Hector quarry and one surrounding Victorville quarry, are simultaneously undergoing opposite changes in velocity. This possibility is difficult to evaluate. One can observe, however, that during the 2-year period of this study, all larger earthquakes were concentrated in the region of the Hector quarry, and there was simultaneously an absence of larger earthquakes in the region of the Victorville quarry. Perhaps the occurrence of larger earthquakes is related to rising velocities near Hector, if they are indeed rising. Such a correlation is reasonable if the velocity increase is due to tectonic stress loading.