Seismic Interferometry

One of the most significant pioneering directions of work that inspired researchers in what was to become the field of seismic interferometry was experimental work in the mid- to late 1990s by Mathias Fink and coworkers on time-reversed acoustics. In particular, the group conducted experiments with piezoelectric transducers surrounding a distribution of a large number of scatterers (e.g., a forest of 2000 steel rods in a water tank; Derode et al., 1995). When the recorded signal from a first experiment was reversed in time and reemitted around the scatterers in a second experiment, the complex signal was focused back at the original source location even when the transmitters and/or receivers had been perturbed or removed. Furthermore, Derode et al. (1995) observed that the width of the time-reversed and refocused pulse appeared five to six times narrower compared to the same experiment without the scattering rods. The remarkable robustness and insensitivity to initial conditions of the experiment by Fink and coworkers offered new hope to exploration geophysicists, who are often faced with incomplete acquisition geometries or experiments that deviate from ideal conditions. In particular, the observations astonished many exploration geophysicists who, based on their attempts to image with reverse-time migration, believed time-reversal techniques to be inherently unstable. The apparent paradox was later resolved by Snieder and Scales (1998), who showed that the instability is caused by inaccuracies in the model and its nonlinear relation with the multiply scattered waves rather than by inaccuracies in the data or an incomplete recording aperture.