Methodologies Of Model Driven Inversion With Pre-stack Seismic Data

Two important approaches are usually utilized in parameters estimating undergroundwith seismic data, model driven and data driven. The former comprises of modelparameterization, seismic analytic expression and seismic inversion. The latter estimates theparameters underground directly with related siginal processing theory by considering seismicdata as special signals. This paper focuses on the parameters estimating with model drivenapproach. In this paper, different types of methods of model parameterization, seismicresponse expression and seismic inversion are proposed for different purposes with practicalcomplicated geology and geophysics problems.Model parameterization and seismic analytic expression lay the foundation for seismicinversion. This paper initially relates the fluid factor (or fluid/pore term) directly with P-waveand S-wave modulus based on poroelasticity theory, and derived the novel seismic reflectioncoefficient in terms of P-wave modulus, S-wave modulus and density, which lays thefoundation for the direct estimation for P-wave and S-wave modulus and fluid discrimination.Furthermore, another novel seismic reflection coefficient in terms of Young’s modulus,Poisson’s ratio and density is proposed to lay the groundwork for rock brittleness evaluationand sweet spot detection of shale gas. Finally, combining with elastic wave inverse scatteringtheory, Born approximation and stationary phase condition, the methodology of yielding theseismic scattering coefficient equation for heterogeneous medium is developed, and with thismethod, different kinds of seismic scattering coefficient equations of complicatedheterogeneous medium are proposed.As for seismic inversion methods, the AVO and AVA simultaneous inversion methodswith smooth model constraint are proposed, these method have little dependence on the initialmodel, which make it possible to estimate parameters for heterogeneous medium. For the condition with elastic impedance equation as forward solver, the novel inversion strategynamed EVA (Elastic impedance variation with incident Angle) inversion in Bayesian schemeis proposed to enhance the stability of the elastic impedance inversion. For model estimationwith seismic scattering coefficient equation as forward solver, the heterogeneous seismicscattering inversion method is proposed with smooth initial model constraint. In this approach,the parameters of the perturbation medium are initially estimated, and then they can besuperposed to the parameters of the known background medium as the renewal parameters ofthe background medium in next iteration. The fluid factor and fracture parameters in theheterogeneous medium are estimated with the proposed method, respectively. Furthermore,we explore the application of this method in in stress evaluation in heterogeneous shalereservoirs.