The Study On Methods Of Avo Prestack Inversion

AVO inversion is an important approach of extracting lithology parameters which are hiddened in seismic information. The classical method of AVO inversion usually estimates parameters only using two-parameter equation, and with the inversion result, P-wave and S-wave information, the location of reservoir can be decided. But we know that the density information can be used in predict the saturation of reservoir fluid, so it has a very important significance in study on three-parameter AVO inversion for trusty density.Because of the unavoidable ambiguity in AVO inversion, this paper studys AVO forward mordling first, to improve veracity and definition of inversion with the result. Then, we research two three-parameter AVO inversion methods, one bases on model, and another bases on Bayes theory. The prestack AVO inversion method which bases on model uses Powell optimum arithmetic to estimate model parameters through the comparison of synthetic, which obtained from model parameters forward modeling, and real seismic data. The application used in model and real data shows that parameters extracted using the method are more stable and have a high resolution, and confirmes the validity of this method. But we found that the inersion result depends on initial model badly, so it is unfit for no-well inversion. The second inversion method which bases on Bayes uses Gidlow approximation to invert P-wave impedance, S-wave impedance and density. Bayes theory combines likelihood function, Gauss distribution function, with prior geological information, adjustable Huber distribution function, so that the distributing rule of model can be shown more nicety. The method gives credible inversion result in model test and real application, and can invert reasonable profile as it can adjusts information adopt quantity between well data and seismic data automatically. So it has a broad foreground.For describing properties of rocks and fluids intuitionisticly, people always invers elastic parameters and density to Laméparameters, because that Laméparameters can distinguish fluid more sensitivity than P-wave velocity and S-wave velocity. Many scholars have done a mass of studys in this area, which adopt indirect methods mostly. For more exact rock physical parameter volume and decreasing cumulative effect of miscalculation, on the basis of predecessors'work, this paper makes a study of a directing inversion method of Laméparameters elementary.which bases on Bayes. From the real application, we can find it is brighter in gas-bearing reservoir, and it cofirms that the method is effective.