| Summary: | In this thesis I examine how the shear-waveform changes when it propagates through media with different crack strikes and how this affects our ability to interpret the shear-waveform in terms of rock properties. It is found that the polarization of the leading split shear-wave is dependent on the anisotropic structure of the last medium it passed through. However, if the anisotropy is weak, or the raypath short, visual interpretation of the polarization diagrams will be difficult. Data collected in reflection surveys are harder to interpret than those collected in vertical seismic profiles as the signal is not recorded in the layer of interest. The presence of changing crack orientations with depth will mean that the results of rotation type anisotropic estimation techniques will no longer give accurate estimates of crack strike and time delay. Field data from a vertical seismic profile in a region where changing crack strike with depth is suspected are examined as a case study. The amplitudes of shear-waves reflected from anisotropic interfaces are investigated to see if they contain useful information about the anisotropic properties of a reflecting layer. The relative amplitudes of reflected shear-waves polarized parallel and perpendicular to the strike of vertical cracks give information about the crack density. Field data are modelled to explain the significance of this effect. For thin cracks, the amplitudes of offset reflections can give information about crack content which maybe useful in crosshole work. Our ability to interpret offset reflected data is considerably reduced if the near-surface layers are also anisotropic or the acquisition line is not parallel or perpendicular to the crack strike. In the final part of this thesis vertical seismic profile data from three sites along the Austin Chalk trend in Texas are examined. Two of the data sets have multi-offset data which allows the anisotropic structure to be modelled. The anisotropic parameters of time delay and polarization angle are extracted from the data using a selection of currently available techniques. Most of the observed anisotropy is in the near-surface and polarization directions follow regional stress patterns. The anisotropic parameters are modelled and a combination of crack and thin layer anisotropy, giving an orthorhombic symmetry, is required to get a good fit.
|
|---|
