Using ground-motion simulations within a Monte Carlo approach to assess probabilistic seismic risk

Accurate description of ground motion characteristics is a vital step in probabilistic seismic hazard and risk assessment. The growing number of ground motion models, and increased use of simulations in hazard and risk assessments, warrants a comparison between the different techniques available to predict ground motions. This research aims at investigating how the use of different ground-motion models can affect seismic hazard and risk estimates. For this purpose, a simple case study is considered, with a circular seismic source zone of a 100km radius, where earthquakes follow the Gutenberg-Richter relationship between magnitude 5 and 7.5. A stochastic ground-motion model is used within a Monte Carlo analysis to create a benchmark hazard output. This approach also allows the creation of a large number of records, removing the influence of potential deficiencies in the quality and quantity of empirical data when comparing models. Calculating the hazard with simulated ground motions helps to capture details of the ground-motion median and variability, which the fixed functional form of a ground motion prediction equation may fail to properly model. A variety of ground-motion models are then fitted to the simulated ground motion data. These include classic ground motion prediction equations (one with a basic functional form, another with a more complex form), and a model using an artificial neural network consisting of a single hidden layer of three nodes. The hazard is estimated from each of these models, with both fixed and magnitude-dependant standard deviations (sigmas) considered. This approach is extended to a risk assessment for an inelastic singledegree-of-freedom system. Results show the impact of ground-motion models on hazard and risk assessment estimates, with ground motions from larger events and from closer source-to-site distances having a larger influence on the results than expected.