Paper Summary
We developed an orthorhombic (ORT) multiarrival control laser-beam migration flow as a model-building engine, with a goal to attain faster computational speeds over standard Gaussian beam or Kirchhoff migration, without sacrifice of frequency, or accuracy. Common-offset data are decomposed and controlled to sparse tau-p domain events. ORT kinematic ray tracing is used to construct controlled-width beams (laser beams). The Gaussian beam multiarrival imaging condition is applied. Though the semblance and wavelet are picked and isolated as a small number of seismic elements, a many-to-many mapping from data domain to image domain is adopted, to allow different arrivals for different data slowness (Pm )contributing to a single imaging point. Cleaner migration images are provided in a reduced computation time cycle, for production tomography, by using the laser-thin beam spread approach and by saving the precomputed sparse tau-p elements and “timetables” to disk. Included case studies demonstrate the ORT multiarrival capability of this model-building migration flow using synthetic and real data examples from the Gulf of Mexico.