Forward Modeling Analysis And Application Of Seismic Image

Based on seismic exploration of the best offset reflection wave, seismic image method is developed with the advantages of efficient, fast data acquisition, flexible acquisition methods and the relatively simple data processing. It has been widely used as a kind of new method of shallow seismic exploration, which also can detect the quality of holes, rocks, caves, bedrock surface of ups and downs, pipelines and large concrete members effectively. Seismic reflection data involves time domain and frequency, however, data processing is limited in the basic data of time domain with little information of frequency domain, which can effectively improve the accuracy of interpretation of seismic image data in comprehensive utilization of frequency domain. Therefore, by forward modeling of seismic image, this thesis analyzes and concludes time-frequency features of seismic image records under different modeling conditions with theoretical and practical significance.On the foundation of the theory of finite difference, this thesis utilizes the staggered grid method to obtain numerical solution and its difference scheme of two-dimensional elastic wave equation. By analysis and selection of two-dimensional the staggered-mesh finite difference of first-order elastic wave equation as the source processing method, and variable coefficient absorbing boundary for the artificial boundary as well, this thesis respectively establishes the isotropic half space model, the model of horizontal layered media and the model with low velocity anomalies.On the basis of the mathematical model, this thesis programs forward modeling of corresponding seismic image, and numerically simulates the mathematical model to obtain diverse models of seismic image waveform records as well as wave field snapshot at different moments. Meanwhile, the study selects relevant records to get the time-frequency distribution in the S transform.From the perspective of time domain and frequency domain, this thesis respectively analyzes and summarizes major findings of forward modeling to be listed as follow: Firstly, when seismic image is applied to the homogeneous medium, the seismic wave will spread from incentive point to every direction, energy gradually reduces, and the recorded in the time domain is given priority to direct wave. Secondly, when seismic image is applied to the medium of horizontal layered structure, the wave propagation layer interface generates a phenomenon of reflection, refraction and transmission, wave type conversions, and complicates the wave field. The time domain records the received strong refraction wave, reflection wave and surface wave. There are two strong energy groups in time-frequency distribution, and the difference between refraction wave and reflected wave is not big. It will form a group of energy in the process that S transform cannot be separated. The other energy group is formed by surface wave, its frequency attenuate step by step. Finally, when seismic wave encounters low velocity body in transmission, wave will produce a diffraction phenomenon when it spreads to the anomalous body points, and disturb the wave field around points. The wave cannot produce refracted wave on the top of interface of low velocity anomalies, but it can on the bottom of interface, which results in its longer time of wave arrival than both sides. When the surface wave in the anomalous body is missing, the frequency falls fast after wave propagating to the abnormal body, which shows that elastic wave attenuates quickly in low anomalous body.By application of the above rules to the explanation of the measured seismic image data, this thesis has achieved good effects and testified the validity of forward modeling results of seismic image. Meanwhile, features of time-frequency under conditions of different models are obtained from forward modeling, which provide certain guiding significance in explanation of measured data.