Search for a command to run...
A big part of full-waveform inversion’s (FWI) success can be attributed to the fact that it attempts to use most of the information present in the recorded data. We build on two main ideas, one, the total wavefield has a wealth of information and second, while FWI does a wonderful job in utilizing this information, occasionally, it falls short. The weak and poorly sampled part of the recorded wavefield often has little or no influence on FWI, mainly, because FWI is looking for a global solution that miuimizes a certain norm of the energy misfit. However, we sometimes want to direct FWl’s attention to a specific part of the data that is not well represented, but that has useful information about the zone of our interest. To address this challenge, we propose a workflow that conditions the data in a targeted manner. We demonstrate the potential opportuuities that arise when we perceive FWI beyond an inversion “black box”. Finally, we share two field data examples from sub-salt fields in the Gulf of America to demonstrate the effectiveness of our approach.