There is a great deal of strong, coherent energy in the ambient noise records. For the most part, this energy is due to pumps, which are located throughout the survey area. The noise covers a wide range of frequencies, from roughly 20 to 100 Hertz.
The pumps are strong surface noise sources, but we are interested in finding sources at depth. There are two possibilities. One is that we will be able to see secondary sources, structures in the subsurface that scatter pump noise energy. Another possibility is that seismic energy will be generated at depth due to the extraction of oil from the reservoir. In either case, the passive seismic processing methods we have previously developed for other datasets (Vanyan and Cole, 1990; Cole et al, 1989) may be useful.
Our principal processing methodology has been to look for evidence of sources at depth by summing along hyperbolic (for nearby sources) or linear (for more distant sources) trajectories. In the case of near-field sources, a 3D grid of possible source locations is constructed. Then, given an estimate of the velocity field, moveout trajectories are computed for sources at each of the grid locations. Finally the data are stacked or semblance is computed along these trajectories. This technique was first proposed by Nikolaev and Troitskiy (1987).
We supplemented this method in SEP-65 with a technique to suppress surface noise sources that might otherwise obscure buried sources. We will undoubtedly want to apply this technique to these data, where there are strong sources of noise on the surface.