The Optimizing Process Of Pre-stacking Seismic Gathers And Dispersion-parameter Inversion

With the difficulty of exploration and development of the oil and gas reservoir increased, which makes accuracy of the effective reservoir prediction higher required. It needs the continuous development and enhancement of new seismic exploration method and technology. Therefore, it is increasingly important for reservoir prediction to be taken effective measures for fluid identification. Post-stack seismic data can reflect the characteristics of strength and continuity of the events, but there are serious problems in describing the complex reservoirs without obvious characteristics of the wave impedance. And pre-stack seismic data can keep enough features of seismic reflection amplitude changing with offset or incidence angles, the optimization of pre-stack seismic data can better improve the seismic data processing and interpretation. Moreover, quality of pre-stack seismic data processing will have direct impact on the subsequent work of the seismic data processing and interpretation, so it is necessary to do a good job in the optimization of pre-stack seismic gathers.Based on the essential feature of the seismic reflection amplitude changing with offset(or incidence angles) and frequency, the paper expand the Russell reflection coefficient formula to the frequency-angle domain and mainly studies the frequency-dependent AVO inversion. However, the phenomenon of wavelet distortion will be caused by NMO correction in the process of pre-stack seismic gathers, which results in the shift to the lower frequency and the amplitude value becoming lager of the wavelet, especially for shallow seismic data of the long-offset. NMO stretch phenomenon will affect the resolution of seismic data and frequency-dependent AVO analysis. Based on the matched filtering algorithm, the paper studies the method of eliminating NMO stretch correction, and analysis the influence of NMO correction stretching to AVO analysis, and eliminated the amplitude distortion caused by NMO stretch and effectively kept AVO inversion of the seismic data, and then, combined the spectral-balancing with the matched filtering algorithm realized the purpose of improving resolution of seismic data. What's more, as to the practical application of the matched filtering algorithm and frequency-dependent AVO inversion method, and finally correctness of the frequency-dependent AVO inversion is verified. Process and progress of the research includes the following contents:(1) Expanding the Russell reflection coefficient formula. Based on Chapman's dynamic equivalent medium theory, frequency-dependent velocity dispersion is introduced into the new inversion formula to have set up the forward model, and frequency-dependent AVO pre-stack angle gathers is calculated with the method of wave equation forward modeling. Therefore, characteristics of the different frequency seismic reflection characteristic parameters(amplitude, velocity dispersion, attenuation and so on) varying with the frequency and incidence angles are to the greatest extent digged up, and seismic dispersion and abnormal attenuation related to the reservoir fluid are detected.(2) An effective algorithm of eliminating NMO stretch correction with the matched filtering is presented, it can make seismic data in the events more continuous eliminate the low-frequency problems caused by the conventional NMO, improve the resolution of seismic data, and can effectively keep the information of amplitude in original seismic data changing with offset. The algorithm, which is suitable for seismic data of any offset, does not require the seismic wavelet is known.(3) Application of matched filtering algorithm to models and practical data commendably eliminates the wavelet stretch phenomenon caused by NMO correction, effectively keeps the seismic data AVO effect and improves the continuity of the events and resolution of seismic data, doing good to seismic data processing and interpretation.(4) Method of frequency-dependent AVO inversion with Russell reflection coefficient formula is applied to practical data, results of inversion via different formula are comparatively analyzed, which both have made good effects. Information related to the hydrocarbon reservoirs can be effectively explained, providing a good method of reservoir fluid identification. Overall, a good impact has been made on the reservoir fluid identification.