Strategy Study On Quasi-P Waves Full Waveform Inversion In VTI Media Based On Sensitivity Analyze

Taking the anisotropy of subsurface media into account,anisotropic acoustic full waveform inversion(FWI)has potential for obtaining velocity and anisotropic parameter.Velocity and anisotropic parameter both have significant influence on seismic migration and reservoir description.Because of the poor sensitivity of anisotropic parameter,conventional FWI methods fail to simultaneously invert for desirable velocity and anisotropic parameter.Quasi-P wave modeling is the foundation of full waveform inversion in acoustic approximation VTI media.In this paper,I compared a few methods of quasi-P wave modeling and have tested on synthetic model.To enhance the stability of simulation and the absorption of boundary,I deprive and achieve the staggered grid finite difference and PML absorb boundary condition.As the result of numerical simulation,the first order velocity-stress equation is choose for its stability.The complete process and formula of full waceform inversion in acoustic VTI media are studied and deduced in this paper.Based on sensitivity analysis,the sensitivity of parameters under different parameterization is analyzed.In this paper,I propose a new quasi-P waves FWI method in VTI(vertical transverse isotropy)media.We exploit the NMO(normal moveout)velocity parameterization to enhance the sensitivity of anisotropic parameter on the near-offset data.Crosstalk between parameters during multi parameter inversion raises the difficulty of inversion.Based on the sensitivity analysis,we develop a hierarchical inversion workflow to relieve the crosstalk between different parameters and further improve the accuracy of velocity and anisotropic parameter models.Simultaneously,cross-correlation objective functional reduce the influence of amplitude during the first step in hierarchical inversion.The preliminary of velocity result is more closed with global optimization.The application on synthetic and real data proves the effectiveness of our method.