Multiscale Reflection Phase Inversion with Migration Deconvolution

Abstract

Reflection full waveform inversion (RFWI) can recover the low-wavenumber components of the velocity model along with the reflection wavepaths. However, this requires an expensive least-square reverse time migration (LSRTM) to construct the perturbation image and RFWI also suffers from cycle-skipping problems. As an inexpensive alternative to LSRTM, we propose the use of migration deconvolution (MD) with RFWI. To mitigate cycle-skipping problems, we introduce a multiscale reflection phase inversion (MRPI) strategy which boosts the low-frequency data and only needs to explain the phase information in the recorded data, not its amplitude spectrum. To mitigate cycle-skipping problems, we use the rolling-offset strategy which gradually extends the offset range of data with an increasing number of iterations. Numerical results show that the MRPI + MD method can efficiently recover the low-wavenumber components of the velocity model and is less prone to getting stuck in local minima compared to conventional RFWI.