Least-squares Wave-Equation Migration for Broadband Imaging

We introduce an efficient Least-Squares Wave-Equation Migration (LS-WEM) for broadband imaging of seismic data. The procedure poses depth migration as an inversion problem. It is capable of producing high-resolution images of the subsurface corrected for overburden effects, variations in illumination, and incomplete acquisition geometries. The modeling engine for the inversion is an anisotropic one-way wave-equation extrapolator able to efficiently propagate high frequency seismic data using high-resolution earth models (e.g. derived from Full Waveform Inversion). The LS-WEM solves for the earth reflectivity in an iterative fashion through data space residual reduction. It integrates the anisotropic wavefield extrapolation operator with a fast linear inversion solver in an efficient migration/demigration algorithm. The algorithm produces high resolution images with high amplitude and structural fidelity. Applications of LS-WEM to synthetic and field data examples (from the Gulf of Mexico and the North Sea) consistently delivered higher resolution images compared to standard seismic migration.