High-frequency Ambient Seismic Noise Simulation And Surface-wave Imaging

This dissertation presents the research results about surface-wave imaging using ambient seismic noise. I mainly use ambient seismic noise simulation as a research tool. Ambient seismic noise methods have been getting more and more attention in investigation of near surface geology, because people can use those methods to achieve subsurface shear-wave velocity models. Shear-wave velocity is an important parameter in investigating the subsurface geology and assessing geological hazards. Furthermore, people can achieve velocity in deep part (>100m) of subsurface geology due to low frequencies of surface waves (<5Hz) in ambient seismic noise. People mainly utilize surface waves in ambient seismic noise because those surface waves possess high energy. Therefore the research about ambient seismic noise surface-wave imaging will promote applications of those methods in near surface. Meanwhile, people have found some factors, like uneven distribution of noise sources, which will make the calculated shear-wave velocity biased from true ones. Therefore it is necessary to understand those factors about surface-wave imaging in ambient seismic methods. Wavefield simulation has been recognized as an important tool in seismology research. Ambient seismic noise simulation could be used to achieve synthetic data under certain condition, for example noise sources are distributed unevenly. Therefore I use ambient seismic noise simulation in the research. I focus on Rayleigh waves in the transverse component of ambient seismic noise and noise sources uneven distribution, where "transverse" is defined as the direction perpendicular to a great-circle path or a line in small scale through observation sensors.People nowadays extract Love waves from the transverse component of ambient seismic noise data using seismic interferometry. Most people assumed that other waves in the component could be negligible. Rayleigh waves, however, can exist in the component when Rayleigh waves propagate in other directions besides radial direction. I demonstrate that Rayleigh waves exist in the result of the transverse component in seismic interferometry using theoretical derivation and examination on synthetic and real data. I present the equation of Rayleigh waves in transverse component cross-correlations; extract Rayleigh waves from synthetic transverse component of ambient seismic noise using seismic interferometry and multichannel analysis of surface waves (MASW); also process the real data which were recorded in Xinjiang Province, and present Rayleigh waves in the frequency-velocity domain. I discuss the influence of those Rayleigh waves in this dissertation. Rayleigh-wave and Love-wave phase velocities are close each other in high frequencies (>0.1Hz), and hence those two kinds of surface waves might be merged in the frequency-velocity domain. Rayleigh-wave phase velocities may be misidentified as Love-wave phase velocities if Rayleigh waves are neglected in picking Love-wave dispersion curve. Therefore the inverse model of the picked Love-wave phase velocities may also be biased from real shear-wave velocity model. To get accurate surface-wave phase velocities from the transverse component data using seismic interferometry in investigating shallow geology, I suggest using MASW to calculate real Love-wave phase velocities.Seismic interferometry theory in ambient seismic noise assumes that noise sources are distributed evenly, but the assumption is hardly met in practice. The noise source uneven distribution will cause empirical Green’s functions and calculated phase velocities biased from real ones. I discuss the influence of noise source uneven distribution on passive source methods, including seismic interterometry, spatial autocorrelation method, refraction microtremor method and passive multichannel analysis of surface waves. I compare the calculated surface-wave phase velocities using those methods with theoretical phase velocities. Synthetic ambient noise data, while noise sources are unevenly distributed, are used in the discussion. It is indicated that people will calculate biased surface-wave phase velocities using those passive source method if noise sources are distributed in an offline direction of a survey line. The inverse model will also be biased based on those biased phase velocities. If main noise sources are distributed in the inline direction of a survey line, people will calculate accurate phase velocities using seismic interferometry, refraction microtremor method, and passive multichannel analysis of surface waves.