Forward Modeling Of Shear-wave Splitting In HTI Media And Study Of Fracture Property Identification

A higher demand in oil exploration has been proposed by the increasing shortage of oil and gas resources. The special reservoir exploration has increasingly attracted the attention of the people, and the fractured reservoir is our interesting point. The description of fractured reservoir lies in the quantitative description of the fracture properties such as fracture azimuth and fracture density. The wave field characteristics of seismic waves including longitudinal wave and shear wave propagating in the anisotropy media are anisotropy. The shear wave is particularly sensitive to anisotropy. When propagating in the anisotropy media, the shear wave will separate into two shear waves, one of which polarizes parallel to the fracture strike called fast shear wave and another polarizes perpendicular to the fracture strike called slow shear wave. This is the shear-wave splitting which is the most important characteristic of anisotropy media. Forward modeling is an important means of researching the propagating of seismic waves. The HTI media is an important model of fracture which is azimuth anisotropy. In the actual seismic exploration, the survey line bias with the fracture strike, and both of the x component and the y component contain the fast and slow shear wave field.In the thesis, firstly we derive the propagation equation in HTI media when the survey line bias with the fracture strike. We obtain the snapshot and seismic record of the HTI media by finite difference with flux-corrected transport technique. We can see the shear-wave splitting in the snapshot. We research the influence of fracture azimuth and fracture density to the shear-wave splitting.Secondly, we introduce the Pearson correlation coefficient into the seismic exploration, propose the Pearson correlation coefficient method to detect the fracture property. With the theoretical single trace seismic record, we verify the correctness of this method. Calculating with the2D3C seismic record which is forward modeled, we obtain a better result compared with the traditional cross-correlation method. The results calculated with the two methods are both high accuracy without noise. But the result calculated with the Pearson correlation coefficient method is better than that calculated with the cross-correlation method by adding noise into the seismic record. We study the relationship between the accuracy and the noise by adding different SNR of noise. The result is that the Pearson correlation coefficient method has a better accuracy and more stable than the cross-correlation method. The Pearson correlation coefficient of two vectors is a value while the cross-correlation of two vectors is a vector, so the computational efficiency of the Pearson correlation coefficient method is better than that of the cross-correlation method. Time window is an important factor that affects the accuracy of the Pearson correlation coefficient. The time window should contain the wavelength of both the fast and the slow shear wave, but it should not be too large, or the computational efficiency will reduce. So the choice of time window size is an important factor.Then, we rotate the x and y component records to obtain the fast and slow shear waves record. The common phase axis of fast and slow shear wave separated by the Pearson correlation coefficient method is more clearer and has higher resolution than that separated by the cross-correlation method. The next work we should do is that we should add the layer stripping technique into this method.