Precursory seismic quiescence

Seventeen cases of precursory seismic quiescence to mainshocks with magnitudes fromM _L=4.7 toM _S=8.0 are summarized. The amount of rate decrease ranges from 45% to 90%. The significance of these changes varies between 90% and 99.99%. The assumption that the background rate is approximately constant is fulfilled in most crustal volumes studied. All quiescence anomalies seem to have abrupt beginnings, and the rate during the anomalous period is fairly constant. The duration of the precursors ranges from 15 to 75 months, and it is not clear what factors determine that time. At least three successful predictions have been based on seismic quiescence. These cases have shown that mainshocks can be predicted based on quiescence, but they have also shown that the interpretation of the data in real time is difficult and nonunique. If a false alarm is defined as a period of quiescence with a significance level larger than a precursory quiescence in the same tectonic area, then we estimate, based on searches in four areas, that the false alarm rate may be on the order of 50%. Failure to predict may be expected in perhaps 50% of mainshocks, even in carefully monitored areas. Quiescence cannot be used as a precursor in tectonic environments with low seismic activity. Most characteristics of the phenomenon are still poorly defined, but data exist which probably permit at least a doubling of the presently available data on case histories.