Simulating earthquake ground motion at a site, for given intensity and uncertain source location

Following a companion article, ground motion acceleration time historiesduring earthquakes can be described as realizations of non-stationarystochastic processes with evolutionary frequency content and instantaneousintensity. The parameters characterizing those processes can be handled asuncertain variables with probabilistic distributions that depend on themagnitude of each seismic event and the corresponding source-to-sitedistance. Accordingly, the generation of finite samples of artificial groundmotion acceleration time histories for earthquakes of given intensities isformulated as a two-stage Monte Carlo simulation process. The first stageincludes the simulation of samples of sets of the parameters of thestochastic process models of earthquake ground motion. The second stageincludes the simulation of the time histories themselves, given theparameters of the associated stochastic process model. In order to accountfor the dependence of the probability distribution of the latter parameterson magnitude and source-to-site distance, the joint conditional probabilitydistribution of these variables must be obtained for a given value of theground motion intensity. This is achieved by resorting to Bayes Theoremabout the probabilities of alternate assumptions.Two options for the conditional simulation of ground motion time historiesare presented. The more refined option makes use of all the informationabout the conditional distribution of magnitude and distance for thepurpose of simulating values of the statistical parameters of the groundmotion stochastic process models. The second option considers allprobabilities concentrated at the most likely combination of magnitude anddistance for each of the seismic sources that contribute significantly to theseismic hazard at the site of interest.