Joint Waveform Inversion with the Least-squares Migration

Full Waveform Inversion (FWI) estimate model with a non-linear optimization, and Least-Squares Migration (LSM) is its linearized form. We present a scheme to combine the FWI and LSM inversion together. After careful choosing the loop scheme, LSM result is used to produce an optimized gradient direction for FWI. The joint waveform inversion can provide a better result with higher resolution and faster computation efficiency. We use numerical test to explain the relation between the background macro-velocity and the image and prove this method’s effectiveness.