Karen R. Felzer, Rachel E. Abercrombie, Göran Ekström

Bull. Seis. Soc. Am., February 2004, volume 93 pp.88-99

Abstract

It has long been debated whether multiplets and foreshocks are a result of the same physical triggering process that causes aftershocks, or whether a special process is involved when clustered earthquakes are of similar size or a small earthquake precedes a larger one. If there is only a single physical triggering process involved in all short term tectonic earthquake clustering then we expect aftershock, multiplet, and foreshock rates to be correlated with each other world wide. We also expect to find the average number of earthquakes triggered by any earthquake to be a simple function of the magnitude of the triggering earthquake, independent of the magnitude of the earthquakes triggered. Finally, since we expect foreshocks to trigger their mainshocks and not the other way around, we do not expect mainshock magnitude to have any control over the magnitude, number, or spatial extent of foreshocks. Using correlation tests, Monte Carlo trials, and binomial probability we find agreement between the data and these predictions in the global CMT and NEIC catalogs and the California CNSS catalog.