An application for null-space shuttling for 4D full-waveform inversion Available to Purchase

The effective null-space (all the models that produce an equal or smaller misfit) of the full waveform inversion objective function can be quite large. Once the inversion converges there are many models that yield a similar fit. In the context of 4D FWI, this space is even bigger. The shuttling technology al- lows us to navigate the null-space to find models with equal or better fit that follow a secondary goal (i.e., minimizing the 4D model difference). In this abstract, we show a way to imple- ment the 4D shuttling algorithm with two different goals and an application to a Gulf of Mexico dataset.