Stochastic Ground Motion Simulation

Abstract

Strong earthquake ground motion records are fundamental in engineering applications. Ground motion time series are used in ▶ response-history dynamic analysis of structural or geotechnical systems. In such analysis, the validity of predicted responses depends on the validity of the input excitations. Ground motion records are also used to develop ▶ ground motion prediction equations (GMPEs) for intensity measures such as spectral accelerations that are used in▶ response-spectrum dynamic analysis. Despite the thousands of available strong ground motion records, there remains a shortage of records for large-magnitude earthquakes at short distances or in specific regions, as well as records that sample specific combinations of source, path, and site characteristics. The limited number of recordings has become problematic in the emerging field of ▶ performance-based earthquake engineering (PBEE), which considers the entire spectrum of structural response, from linear to grossly nonlinear and even collapse, and thereby requires ground motions with various levels of intensity for various earthquake design scenarios (e.g., a design scenario can be defined by the earthquake magnitude, distance, and site conditions). To obtain the desired ground motions for the purposes of PBEE, it is common engineering practice to scale or modify acceleration time series that were recorded during previous earthquakes to represent certain ground motion characteristics for the design of structural or geotechnical systems. However, scaling and modification methods can significantly alter other ground motion characteristics and result in unrealistic earthquake ground motion time series. Synthetic ground motions can be used instead to replace or supplement recorded motions when scarcity of previously recorded motions becomes a problem, provided they accurately capture characteristics of real earthquake ground motions and their natural variability. Synthetic ground motions can be based on deterministic or stochastic simulations. A determinist model is one in which variables are uniquely determined, and the model performs the same way for a given set of initial conditions. Conversely, in a stochastic model, randomness is present and variables are not described by unique values, but rather by probability distributions. In earthquake engineering, deterministic ground motion simulation is commonly referred to as “physics-based” ground motion simulation. These simulation models synthesize the earthquake source by defining a source model (e.g., kinematic or dynamic rupture models) and describe the seismic wave travel path by defining a material model (e.g., seismic velocity model); then, they utilize numerical methods (e.g., finite element or finite difference methods) to estimate the solution to the wave propagation