Imaging crustal diffraction zones and seismic tomography

A tomographic method has been developed to determine three-dimensional structure and velocity model from seismic reflection data. The interfaces separating layers of different physical parameters in the model are approximated by polynomials and the velocity in each layer is either a constant or a linear function of depth. Direct modeling uses a ray bending method and inversion employs an iterative damped least-squares method. The method is stable and produces good estimates of the model parameters for a three-dimensional synthetic data set.; A two-dimensional Vibroseis seismic survey was carried out on Vancouver Island under the Project LITHOPROBE-I to determine the crustal structure and to understand the tectonics of this convergent margin. The tomographic method was applied to a small portion of Line-1, and a three-dimensional structure with a velocity model was obtained by taking advantage of the crookedness of the line.; A new seismic processing method, called the Common Fault Point (CFP) stacking method, has been developed to image discontinuities in the subsurface acoustic impedance. The method uses the diffraction hyperbolae produced by such discontinuities on prestack traces and concentrates the diffracted energy at the apexes by applying appropriate moveout and amplitude corrections and stacking. Each CFP stacked trace for a particular surface location being obtained by assuming diffractors to reside directly beneath that surface location. The CFP stacked section is produced by repeating the process for each surface location. The discontinuities are identified on these sections by high amplitudes.; The CFP method was tested for synthetic data with and without noise. A velocity model obtained from the seismic data at different source-receiver distances is used for CFP stacking. The method was also applied to reflection data sets from southern Alberta and Melville Island. The vertical and lateral resolution limits of the method were investigated by using a step fault and a dyke model respectively. The resolution depends upon source pulse, the resolution criterion, the size of the Fresnel zone and the recording parameters.