A field program to record microaftershocks was undertaken following the great Alaska earthquake of 28 March 1964. Three weeks after the main shock more than 1000 microaftershocks per day were recorded on a high-gain, high-frequency seismograph operating at a gain of more than 106 at 10 cps. The microaftershock data, together with the data for the teleseismically recorded aftershocks, furnish new information on the aftershock process.
This study, when considered with others, reveals certain general features of aftershock sequences that must be accounted for in any successful model of the aftershock process. First, over a magnitude range from at least m = 0 to m = 6, aftershocks follow the Gutenberg and Richter magnitude-frequency relationship with b values between 0.8 and 0.9. At present, departures from 0.8 or 0.9 in b values reported for aftershock sequences cannot be distinguished from the uncertainties in the estimates of b. For a particular aftershock sequence, the distribution of magnitudes is stationary in space and time. Second, over an interval of months, the frequency of occurrence of large aftershocks decays uniformly in time according to an inverse power law with an exponent somewhat greater than 1.0. In contrast, microaftershock activity sampled on a time scale of hours departs markedly from the inverse power law decay; microaftershocks occurring in a small volume tend to cluster in time. This difference suggests that aftershocks occur independently of one another except, perhaps, for those that are separated from each other by distances of a few kilometers or less. Third, prominent aftershock sequences typified by the above-mentioned statistical properties occur only at shallow depths within the Earth's crust. Fourth, secondary aftershock sequences occur occasionally. Characteristically, secondary aftershocks occur on the edge of the main aftershock zone and have focal mechanisms that are distinct from those of ordinary aftershocks. The statistical properties of secondary sequences are identical to those of ordinary aftershock sequences.
Within the accuracy of the data, aftershocks and microaftershocks of the 1964 earthquake occurred at depths less than 35 km. The few deeper shocks that occurred within or near the aftershock zone represent the usual subcrustal seismicity associated with this region. Assuming that the primary fault associated with the main shock is a low-angle overthrust, the aftershocks originated within a few tens of kilometers of the fault. A detailed investigation of microaftershocks reveals a clustering of events in space and time that is identified with small swarms or secondary aftershock sequences and is not readily attributed to triggering by tidal strains.