- With seismic bulletins from the National Earthquake Information Service, and the LASA and NORSAR arrays for the decade of the 1970's, a search was made for seismic precursors of various types. Using a cutoff of m(,b) = 5.8, 510 main shocks were identified to guide the search. The true rate of foreshock occurrence was statistically established to be less than 20%. Foreshocks were almost exclusively associated with shallow main shocks. Very weak trends were found for seismicity level, mean magnitude, and the distance to the main shock of events as the main shock approaches by a method of averaging data over main shock regions. A new variable is proposed to track departures from spatial-temporal randomness in seismic zones--it is the product of interevent time and distance of the events to the main shock. The cumulative curve of this variable was applied extensively to the mid-America seismic zone with only mild success in predicting main shocks.
An experiment using LASA recordings of nearly 400 events in the lower Kuril Islands revealed that measured stress drops did not appear to change prior to five main shocks in that region. Unrelated to precursory behavior, it was found that the stress drops increased significantly with magnitude (m(,b)) or with seismic moment. A search for other seismic precursors for the same Kuril main shocks failed to identify any strong precursory patterns.
A numerical one-dimensional fault model was employed to simulate seismic behavior. This model produced "seismic events" with statistical properties akin to true ones. It is shown that main shocks in the model occur without significant precursory seismic behavior, that they tend to occur under a uniform-stress, low-entropy condition, and that their recurrence time is highly random.
Altogether the empirical and model studies lead to a theory which emphasizes the random component of stress on fault planes. It is hypothesized that the wavenumber spectral content of the stress field is whitened after main shocks and that subsequent activity reddens this spectral content, thereby eventually making conditions favorable for another main shock. This theory is consistent with empirical data but does not predict any significant precursory seismic activity.
- Dissertation Note:
- Ph.D. The Pennsylvania State University 1982.
- Source: Dissertation Abstracts International, Volume: 43-08, Section: B, page: 2482.
View MARC record | catkey: 13611755