Application of Image-Guided Full Waveform Inversion to a 2D Ocean-Bottom Cable Data Set

Image-guided full waveform inversion (IGFWI) was proposed and synthetically tested in Ma et al. (2011). Compared with conventional FWI, IGFWI yields models (e.g., velocities) that make better geologic sense because it takes into consideration the subsurface structures, which can be apparent in migrated seismic images. Moreover, IGFWI converges faster in fewer iterations, especially when reflection energy is used to invert for high-wavenumber details in the model. We test the IGFWI method on a 2D ocean-bottom cable (OBC) data set. In this test, we first use refraction data to update the lowwavenumber component of the model and then proceed to invert for high-wavenumber details with reflection data. In the reflection stage, we incorporate structural constraints into inversion; as a result, the estimated model make plausible geologic sense.