Effects of site conditions on earthquake ground motion and their applications in seismic design in loess region

The Loess Plateau is an earthquake prone region of China, where the effects of loess deposit on ground motion were discovered during the 2008 Wenchuan earthquake(Ms8.0) and the 2013 Minxian-Zhangxian earthquake(Ms6.6). The field investigations, observations, and analyses indicated that large number of casualties and tremendous economic losses were caused not only by collapse and damage of houses with poor seismic performance, landslides, but also amplification effects of site conditions, topography and thickness of loess deposit, on ground motion. In this paper, we chose Dazhai Village and Majiagou Village as the typical loess site affected by the two earthquakes for intensity evaluation, borehole exploration, temporary strong motion array, micro tremor survey, and numerical analysis. The aim is to explore the relations between amplification factors and site conditions in terms of topography and thickness of loess deposit. We also developed site amplification factors of ground motion for engineering design consideration at loess sites. The results showed that the amplification effects are more predominant with increase in thickness of loess deposit and slope height. The amplification mayincrease seismic intensity by 1 degree, PGA and predominant period by 2 times, respectively.

Analysis of the role of diffraction in topographic site effects using boundary element techniques

The role played by the diffraction field on the problem of seismic site effects is studied. For that purpose we solve and analyze simple scattering problems under P and SV in-plane wave assumptions, using two well known direct boundary-element-based numerical methods. After establishing the difference between scattered and diffracted motions, and introducing the concept of artificious and physically based incoming fields, we obtain the amplitude of the Fourier spectra for the diffracted part of the response： this is achieved after establishing the connection between the spatial distribution of the transfer function over the studied simple topographies and the diffracted field. From the numerical simulations it is observed that this diffracted part of the response is responsible for the amplification of the surface ground motions due to the geometric effect. Furthermore, it is also found that the diffraction field sets in a fingerprint of the topographic effect in the total ground motions. These conclusions are further supported by observations in the time-domain in terms of snapshots of the propagation patterns over the complete computational model. In this sense the geometric singularities are clearly identified as sources of diffraction and for the considered range of dimensionless frequencies it is evident that larger amplifications are obtained for the geometries containing a larger number of diffraction sources thus resulting in a stronger topographic effect. The need for closed-form solutions of canonical problems to construct a robust analysis method based on the diffraction field is identified.

Locating the Source Regions of the Single and Double-Frequency Microseisms to Investigate the Source Effects on HVSR in Site Effect Analysis

Evaluating the seismic site effect by the ambient noise based horizontal-to-vertical spectral ratio(HVSR)method is strongly affected by the spatial and temporal variations of the ambient noise sources.Therefore,it is necessary to locate the source regions of ambient noise and investigate the relationships between the source energy and HVSR values at the predominant frequency(HVSRf_(0))of the site.The generation mechanisms of the single-and double-frequency microseisms(SFMs,0.05-0.085 Hz and DFMs,0.1-0.5 Hz)in ambient noise are better understood than the noise in other frequency bands and they are dominantly composed of fundamental Rayleigh(Rg)waves.With this advantage,the recordings of SFMs and DFMs at 30 stations in the east coast region of the United States are used to demonstrate a study on locating their source regions with reasonable certainty and constructing the functional relationship between the HVSRf_(0) and the source energy of SFMs and DFMs.The recordings are processed in four sub-frequency bands(Fs)of SF and DF bands and a polarization analysis is carried out to select the ellipsoids approximating the particle motions of Rg waves.Then the probability density functions of the back azimuths of the ellipsoids’semi-major axes are computed for each F and station,and are projected on the ocean to determine their possible source regions.These regions are further constrained by(1)the correlation coefficients between the SFMs and the WAVEWATCHⅢ(WWⅢ)hindcasts of ocean wave spectra in the SF band,or between the DFMs and the modeled DF energy on ocean surface in the selected time windows in the DF band,(2)the energy contribution defined by(i)the average WWⅢocean wave energy and the ocean bottom topographical gradient in the SF band,or(ii)the average modeled DF energy on ocean surface and a frequency and water depth dependent coefficient measuring the conversion efficiency of DF energy from water to solid earth in the DF band,and(3)the percentile retained energy of Rg waves in both the SF and DF bands.Results of source regions reveal that(1)the SFMs recorded in eastern US result from the interactions of low frequency(0.05-0.085 Hz)ocean waves with the continental slope and shelf of western North Atlantic Ocean;(2)the source regions for long-(0.1-0.2 Hz)period DFMs are located in the deep ocean close to the continental slope;and(3)the short-(0.2–0.5 Hz)period DFMs are generated in the continental shelf.Finally,the correlation analyses between the simulated source energy and the HVSRf_(0) values at the stations whose f_(0) s fall in DF band are carried out revealing significant source effect on thick sediments at low frequencies.

Double-Frequency Microseisms on the Thick Unconsolidated Sediments in Eastern and Southeastern Coasts of United States: Sources and Applications on Seismic Site Effect Evaluation

This study presents a systematic analysis of double-frequency(DF) microseisms recorded on the unconsolidated sediments in the eastern and southeastern coasts of United States. For all recordings, the site effect parameters(predominant frequency(f_(0)), amplification factor and unconsolidated sediment thickness(UST)) are obtained by Nakamura method and the DF spectra are classified into five groups in terms of the DF peak patterns and the recording locations relative to the coastline. The frequencies and energy levels of the DF peaks in horizontal direction and the amplification factors are associated with the UST which is resulted from seismic site effect. By polarization analysis, the primary vibration directions of the DF peaks are identified and presented as great circles passing through the recording stations intersecting mainly along the continental slope. Correlation analyses of time histories of the DF energy and the ocean wave climate observed at buoys show that the low(<0.2 Hz) and high(>0.2 Hz) frequency DF microseisms are generated in the deep ocean and the continental shelf respectively. It is concluded that the continental slope plays a significant role in the generation of DF microseisms as it causes reflection of waves from the open ocean, initiating standing waves.