Anisotropy Estimation and Image Remapping Using Model Based Moveout

Typically model parameter estimation is achieved through multiple iterations of linearized tomography and depth migration. In this paper we discuss a method to reduce the migration effort by applying local approximations to the imaging and modelling operators using a model based moveout, which is a mapping between imaged specular reflections in initial and updated models. This allows for very efficient imaging and model testing, so that multiple iterations of imaging and linearized tomography or generalized iterative inversion can be applied. The model based moveout and tomography is demonstrated for anisotropic inversion in the case of known subsurface control, where the vertical velocity is known and the task is to estimate the anisotropic Thomsen parameters for VTI media. The method requires an iterative joint inversion of these parameters by projecting residual depth errors onto model updates.