Signal recovery beyond conventional Nyquist: The sample rates used for seismic acquisition do not need to limit the maximum recoverable frequencies

The seismic industry's conventional wisdom teaches that the temporal and spatial sample rates used when collecting seismic data place a hard limit on the maximum recoverable frequencies. This conventional wisdom is wrong. The data-collection Nyquist frequencies (time and space) do not need to limit the maximum frequencies in the final migrated stack volumes. This is not to say that the sampling theorem is wrong but that we as an industry have not optimally applied it since the advent of digital data collection and processing. Normal Neidell and Daniel Wisecup have previously told us of “the errors of our ways,” but few have listened. What I present here is similar to Wisecup's method. The goal of this article is to provide simple examples that “even a geologist could explain to a teenager” as to why the sample rate does not necessarily limit the maximum recoverable frequency. It is directed at geophysics students (and their professors) so that they might have a new, enlightened “conventional wisdom” compared to us “seasoned” geophysicists. With this insight, they should demand more from their seismic data providers.