NEHRP Clearinghouse

Generalized Semi-Markov Process for Modeling Spatially and Temporally Dependent Earthquakes.
Lutz, K. A.; Kiremidjian, A. S.
National Science Foundation, Arlington, VA.; Electric Power Research Inst., Palo Alto, CA., cJuly 1993, 102 p.
Site-specific hazard estimation requires the modeling of the occurrences of earthquakes on any faults with the potential to impact the site. Previous earthquake occurrence models have assumed either spatial independence or temporal independence or both. However, for large magnitude earthquakes (approximately moment magnitude 6.5 and above) occurring infrequently on long faults, evidence indicates that the assumptions of temporal and spatial independence are not valid. A new fault behavior model incorporating temporal and spatial dependence is needed to estimate site-specific hazard in areas subject to such earthquakes. This research develops an earthquake occurrence model that is a generalized semi-Markov process (GSMP) and allows for the simulation of the fault behavior through time. The size of each simulated earthquake is related to the amount of slip that is released. In order to apply the model to a fault, the following data must be known for each cell along the entire length of the fault: the slip rate, the mean and standard deviation of the earthquake interarrival times, and the time of the last earthquake. Additionally, the time of the last earthquake that ruptured the entire fault must be known. The model can then simulate the sizes and locations of earthquakes occurring along the fault for the time period of interest.
Seismic hazards; Dynamic response; Earthquake engineering; Earthquakes; Faults (Geology); Site characterization; Earth movements; Forecasting; Markov processes; Structural geology; Mathematical models; Tectonics