Stochastic Simulation of Spatially Distributed Ground Motions Using Wavelet Packets and Kriging Analysis

Considering spatial distribution of ground-motion is important in seismic hazard analysis of spatially distributed infrastructure systems such as long-span bridges, lifelines, railways. It is often the case that distributed ground-motion time histories are required to perform analyses at multiple locations where recorded time histories are not available. In this paper, we propose a geostatistics-based method to simulate spatially-distributed synthetic ground motions using wavelet packets and kriging analysis. In this model, thirteen wavelet parameters are used for time-frequency characterization of earthquake ground motions. The spatial correlation of these parameters is determined through semivariogram analysis using densely populated recordings from the Northridge and Chi-Chi earthquakes. It is observed that the spatial correlations of most wavelet parameters are closely related to regional site conditions. Kriging technique is then used to estimate the wavelet parameters at unmeasured locations using the spatial correlations. It is demonstrated that the simulated ground motions are in good agreement with the actual recordings. This method can be used for time history analysis of spatially distributed infrastructure systems and circumvent difficulties in traditional ground-motion selection and modification processes. It is also useful to derive time-histories at an unobserved location using recorded ground-motion data in the neighborhood. 1 Graduate Student Researcher, Dept. of Civil and Environmental Engineering, The Hong Kong University of Science and Technology, Hong Kong Assistant Professor, Dept. of Civil and Environment al Engineering, The Hong Kong University of Science and Technology, Hong Kong Huang D, Wang G. Stochastic simulation of spatially distributed ground motions using wavelet packets and kriging analysis. Proceedings of the 10 National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014. Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska 10NCEE Stochastic Simulation of Spatially Distributed Ground Motions Using Wavelet Packets and Kriging Analysis D.R. Huang and G. Wang