Polarization Filtering Of Three-component Seismic Wave Field Separation Method

As the development of seismic exploration,more and more attention has been paid to the medium of complex structure. The new technique of using polarization characteristics of seismic wave to research seismic information has become an important aspect in seismic data processing .Particle motion can be used to describe the position and shape of the seismic wave's particle trace in space,for digital discrete three-component seismic recording (X, Y, Z). We can build a time-shift waveform filter through calculating the covariance matrix within a moving time window. By means of polarization analysis, ground roll can be suppressed. We can get shape parameters of ellipsoid from the eigenvalues of the covariance matrix , and discuss thoroughly about the physical meaning of the covariance matrix. After knowing the relationships among the matrix and geometry , parameter of polarization and matrix , body wave and polarization plane, we may construct the function of polarization filter picking up the linearly polarized wave. Key steps are as follow :1. We can ascertain the size of time window through particle motion diagram. For a time window with N sample points , these sample points can be looked upon as the aggregate of a random sample point sets, so the covariance matrix of these points can be determined.2. We can get the ellipticity and linearity from the eigenvalues of the covariance matrix.3. According to the sequence determined by eigenvalues , we can distinguish the energy origin of each group of eigenvector in each filter section and the position of main polarization plane.The characters of seismic wave can be distinguished by combing the relation between the polarized characteristic of seismic wave and spread direction. Via selectively choosing the parameters (such as p, q, N etc.) of the filter, the noisy waves can be suppressed, P and S-wave can be separated ,and the effective waves can be strengthened further.This thesis analyses the polarization characteristics of seismic wave, introduces several commonly applied polarization filter methods particularly, and gives moreattention to the Poline method.The main researches of this thesis includes:1. We present four methods of polarization filtering and compared them theoretically.2. Correlated algorithms of covariance matrix were researched and analysed. Thoroughly studying was given to the physical meaning of covariance matrix.3. The author compiles the software combining VC++, MATLAB and MATCOM to realize the automatic of computation and visualization of the polarization filtering.4. Several typical theoretical models are designed and practical datas are applied to testify the smoothness and robustness of polarization filter system and good results have been achieved.