Improve seismic imaging with data-driven least-squares migration (LSM)
Least-squares migration (LSM) is a powerful data-driven approach to seismic imaging, which aims to recover the Earth’s true reflectivity by minimizing the misfit between recorded seismic and forward-modeled data. Applicable to all seismic acquisition environments and geometries, including land and ocean bottom, it is particularly valuable to challenging time-lapse 4D seismic projects with highly non-repeatable vintages or weak 4D signals. Amplitude distortions caused by absorption can be compensated for by extending the approach to include Q.
LSM offers many important benefits when compared to conventional migration. These include: reduced migration artifacts and swing noise, evenly balanced illumination, improved signal-to-noise ratio, better event continuity, and more interpretable seismic amplitudes. LSM also proves an ideal partner to full-waveform inversion (FWI) as it can take full advantage of the highly detailed velocity models.