Contribution to Seismic Modeling and Imaging in the Presence of Reflector Roughness

Download or read book Contribution to Seismic Modeling and Imaging in the Presence of Reflector Roughness written by Guochao Gao and published by . This book was released on 2020 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: Due to various geological processes and crustal movements, rough interfaces widely exist within the Earth. The rough interface can strongly affect seismic wave propagation, manifested as changes in the amplitude, phase, scattering angle, frequency content, and even the wave-type conversion. Inevitably, the quality of seismic imaging or inversion is also greatly influenced. Despite the numerous works devoted to the interaction of waves with rough interfaces, this interaction remains to be better understood, as it is still quite challenging to model the seismic wave propagation and to properly reconstruct the subsurface. The thesis investigates the effect of rough interfaces on seismic wave modeling and imaging, and explores the potential of an electromagnetic method to remove this effect and to better image the subsurface.We use a spectral-element method, and more specifically the code SPECFEM2D, for modeling acoustic wave propagation in the time domain. First, we consider a sinusoidal grating and illustrate numerically the consequences of the grating equation on the temporal signals. Then, using f-k analysis, we show the location of the different diffraction orders in the frequency-wavenumber domain. After a sensitivity analysis, we select an appropriate configuration that allows for the separation of diffraction orders from a shot gather. Last, both roughness height and correlation length are shown to obviously influence the appearance of the diffracted wavefield. However, the correlation length has less effect on the energy of the diffracted waves than the interface roughness.We adopt a full-waveform inversion (FWI) scheme based on the software package DENISE to study the influence of different roughness heights and correlation lengths on seismic imaging results. When the roughness height increases up to the dominant wavelength or is greater, the random noise dominates in the seismic data, and the FWI results significantly deteriorate, especially for the reconstruction of a horizontal reflector located below the rough interface. In contrast, the correlation length has a much smaller effect on both random noise and quality of the inverted results than the roughness height. As shown here, the interface roughness has a major impact on both seismic wave propagation and imaging. When a rough interface is expected to be present in the subsurface, its effect should be critically considered in FWI, in order to properly reconstruct reflectors possibly located below, and then to properly interpret images of the subsurface. In this context, we perform some preliminary tests on the use of a selective extinction method to remove the impact of the roughness on the wavefields. The results are promising and show the potential of the method for better imaging. In addition, the standard deviation of the amplitude of the processed data may be used to evaluate the characteristics of the rough interface, which is also of interest for geophysicists and geologists.