Improved fixed-point seismic inversion constrained by instantaneous phase

Abstract

Some subtle reservoirs are hidden in the weak reflection areas and thin layers in seismic profiles, and they cannot be identified easily. Thus, the thin layer interpretation and the subtle reservoir prediction are the essential in reservoir management plan while field development and production stage. To improve the recognition capability in the weak reflection areas and resolution, the fixed-point (FP) algorithm is employed to obtain the inversion results since the conventional FP seismic inversion (CFPSI) is sensitive to the weak changes in seismic signals. However, the CFPSI suffers from low convergence. Therefore, the new iterative equations of pre- and post-stack FP seismic inversion are designed to improve the computational efficiency and convergence. Besides, a new constraint, which is derived from the instantaneous phase of the seismic trace, is also sensitive to the weak changes of the seismic signal. The instantaneous phase constraint is then involved in the objective function, contributing to a significant improvement of identification capability in weak reflection areas and the resolution of inversion. Additionally, the initial model constraint is introduced into the objective function to weaken the multi-solution and noise sensitivity. The synthetic examples demonstrate that the computational efficiency, recognition capability in the weak reflection areas and resolution of the proposed approach are higher than those of the CFPSI. The field examples suggest that the subtle reservoirs and thin layers, which are hidden in the weak reflection areas in the field data, are accurately identified using our method. Thus, the proposed method is useful for improving the recognition capability in weak reflection areas and resolution of the inversion results.