Faults usually are not isolated features but exist within a:population of faults which may interact through their stress fields, This poses two serious problems for seismic hazard analysis. The most severe such problem lies in estimating the likelihood of whether or not a future earthquake will be confined to a single fault (or fault segment) or will jump to adjacent faults and result in a larger earthquake. We review recent results which show that it is possible to determine the degree of fault interaction from geological data,alone. We propose that the probability of an earthquake jumping from one fault to another will increase with the degree of stress interaction between-the faults, and introduce a simple criterion to estimate the degree of interaction based on separation; and overlap of echelon normal fault pairs. This statics based criterion for normal faults agrees qualitatively with the limited dynamic modeling of (Harris, R.A,, Day, S.M., 1993. Dynamics of fault interaction: parallel strike-slip faults. Journal of Geophysical Research 98, 4461-4472) of the more complex case of strike slip faults and suggests that a more general criterion may be obtainable. The second-problem discussed is the hazard associated with earthquakes being triggered by earlier earthquakes on a different fault. This phenomena produces seismic hazards distinct from that associated with ordinary aftershocks. We point out that with rapid data-acquisition and proper preparation, it is feasible:to issue a short-term hazard assessment regarding:triggered earthquakes shortly after the occurrence of a potentially triggering event. (C) 1999 Elsevier Science Ltd. All rights reserved.
274JXTimes Cited:8Cited References Count:25