Application of the wavelet transform in filtering the noise of real earthquake accelerograms

In the deterministic analysis of building structures, real accelerograms are used as input data. Very often, only instrumental recordings of accelerograms that are not cleared of noise are publicly available. Such accelerograms cannot be used directly in the analysis of building structures. Various broadband filters are used to adjust the instrumental recordings of accelerographs. This article discusses the possibility of applying filters based on the wavelet transform. The technique of the algorithm of filtering from the noise of the source data is considered. As an efficiency of this filtering algorithm, a comparative analysis of the dynamic coefficient and the energy criteria of the pulse and Arias for the corrected and instrumental accelerograms is carried out.

Selection of earthquake ground motion accelerograms for structural design in Hong Kong

Hong Kong is situated in an intraplate area of low to moderate seismicity, where recorded strong-motion accelerograms are scarce. As part of the seismic design process, dynamic time-history analysis is sometimes required for assessing the seismic performance of structures. Hence, earthquake accelerograms representative of various local design hazard levels are needed. This study aims to develop a methodology for building up a database of earthquake accelerograms for the Hong Kong region which can be used for the analysis and design of a variety of structural and geotechnical systems. An outline of the previous probabilistic seismic hazard analysis studies conducted for Hong Kong is first presented and several design earthquake scenarios are determined, which contribute most to the local seismic hazard given a specific return period. The median response spectra and conditional mean spectra of the selected earthquake scenarios for reference rock sites are then derived based on the component attenuation model. The conditional mean spectra are then used as the target spectra for the selection and scaling of recorded accelerograms sourced from the available global database. Finally, a suite of conditional mean spectra–compatible accelerograms for rock sites is presented, from which event-specific and site-specific accelerograms are generated through dynamic site response analysis.