| In traditional seismic exploration, due to the low dominant frequency of seismic records, the seismic profiles mainly reflect the large-scale tectonic structure and depositional unconformity surface, such as syncline, anticline, fault, and sedimentary facies. However, as a result of the dramatic variation of composition, density of rock, and temperature, pressure of the fluid in the pore, the elastic parameters of earth medium, such as elastic modulus, velocity, Poisson ratio, absorption characteristics, are inevitably heterogeneous. Because the scale of these heterogeneities is much smaller than that of the seismic wave length, these heterogeneities only generate few of incoherent scattering in seismic records. So these scattering have not been taken into account or have been ignored in traditional seismic data processing and interpretation. In seismic inversion, the subsurface medium has also been simplified as a series of uniform layered medium model and has not involved the influence of the seismic scattering.However, for the application of high resolution, wide-band seismic exploration method, the scale of seismic wave wavelength and that of the small scale heterogeneities of medium are closer and closer. So the impact of these heterogeneities on seismic wave field is more obvious, and generate a lot of scattering of seismic waves. These scattered seismic waves, to some extent, have a major influence on the accuracy of the seismic processing and interpretation. On the one hand, only if the seismic wave propagation theory which applied in heterogeneous medium has been systematically studied, we can find a method for diminishing those impact of scattered waves. On the other hand, disposing of or ignoring the seismic wave field caused by inhomogeneous medium, which, namely, makes us give up some potential valuable information sources of subsurface medium, such as sedimentation model, sedimentary microfacies, lithological changes. So, investigating on the characteristics of seismic scattering and finding a kind of method to detect small scale inhomogeneity are completely necessary. The investigation of this dissertation is a part of research and exploration contents mentioned above, and the purposes are as follows:1) provides a flexible, convenient, three-dimensional description and modeling methods of random medium for the study of the seismic wave field in complex medium; 2) find a kind of method to estimate the heterogeneity characteristics of subsurface using post-stack seismic data.The main contents and conclusions of this dissertation are as follows:1) The mathematical description methods of the random medium. This study established the autocorrelation function of the three-dimensional non-stationary random medium, and proposed the modeling method of three-dimensional non-stationary random medium based on stochastic process theory.2) The parallelization of modeling algorithm. Since modeling algorithm of three-dimensional non-stationary random medium is time consuming, this study designed an parallelized algorithm for modeling the three-dimensional non-stationary random medium. Tests show that the efficiency of the parallelized algorithm has been greatly improved.3) The parameter estimation of the random medium from post-stack seismic data. Based on the theory of non-stationary random medium, an estimation algorithm of random medium parameters is proposed from the two-dimensional post-stack seismic data. The algorithm can estimate the spatial distribution characteristics of the random medium parameters, including autocorrelation length of a and b and the relevant angle0. The test with synthetic seismic data shows the algorithm is correct and feasible,which has been extended to the parameter estimates in the three-dimensional stationary random medium.4) The estimation experiment of random medium parameters on the real seismic data. Taking one typical two-dimensional seismic profile which better reflects of delta sedimentary facies in Member1of Dongying Formation in Dagang Oilfield as for instance, this dissertation presents parameter estimation of random medium through the profiles guided by the theory of non-stationary random medium. And then according to the estimation results, the delta sedimentary facies in the profile has been analyzed and divided into three microfacies. This experiment shows that the parameters of random medium can be used as new seismic attributes in the interpretation of the seismic data which reflects the method that has a good prospect and can provides a very effective means to study the heterogeneity of the actual medium and to separate seismic facies and sedimentary facies. This dissertation is mainly divided into five chapters. The first chapter, a brief introduction to the research background, purpose, significance, research progress status at home and abroad, as well as the thesis’s main research content, main achievements and innovations. The second chapter, briefly introducing the basic concepts and different types of random medium, different parameters corresponding to the characteristics of the random medium. The third chapter discusses the description method and the modeling algorithm of the three-dimensional non-stationary random medium model, and the parallelized processing scheme of the modeling algorithm. The fourth chapter mentions parameters estimation algorithm, the estimation test of the theoretical model and actual seismic data in both two-dimensional non-stationary random medium and three-dimensional stationary random medium. The last chapter summarize the thesis’s main work and conclusions and put forward the problems existing in the thesis and the future research direction. The innovation of this dissertation is mainly represented in two aspects:1) the modeling algorithm of the three dimensional non-stationary random medium could build more complex three-dimensional models which random medium parameters change continuously in the three-dimensional space. So this is an innovation in random medium theory and modeling research;2) compared with the previous parameters estimation works of two-dimensional random medium, this thesis not only increases the estimation of autocorrelation angle θ, but also can estimate parameters of the non-stationary random medium and parameters of three dimensional random medium. These parameters of random medium is of great value to sedimentary microfacies analysis by using real seismic data and seismic facies controlled wave impedance inversion. So this is an innovation in application of random medium theory. |
PDF Full Text Request