This website uses cookies to improve your experience. If you continue without changing your settings, you consent to our use of cookies in accordance with our cookie policy. You can disable cookies at any time.



Beam migration methods rely on local slant stacks of the seismic data for constructing the beams. Slant stack resolution can deteriorate with limited aperture, and spatially sparse or irregular sampling of the seismic data. In the τ-p domain, energy from one angular p can leak onto others. Especially for the 3D problem, the full slant stack data can expand the original traces to an unacceptable size, depending on the number of chosen slowness values. The leaked energy can introduce noise and artificial structure in the image and also reduce the efficiency of the beam migration. We propose a high resolution 3D beam forming procedure by solving the slant stacking operation in a Least-Square Inversion (LSI) fashion. In each iteration, the method first solves for the most energetic angular component, inverses slant stack the selected p component to the original trace locations, subtracts the inversed data from the traces and continues for the next iteration. In order to enhance sparsity, the data traces are partitioned in time windows, and the LSI slant stack is conducted within each overlapping time window. A three dimensional field data example shows the proposed method reduce the original slant stack size dramatically and the final image S/N ratio is also improved, some subsurface structures are more continuous and enhanced comparing with the migration image with original slant stack data.