Ambient Noise Imaging techniques and potential in warm shallow water

The class of acoustic imaging techniques known as Ambient Noise Imaging started with Acoustic Daylight, which has since been joined over the past decade by several other independent algorithms. To date, the algorithms described in the literature have generally been statistical approaches, treating the insonification by sources of opportunity as a random variable. In warm shallow waters, where snapping shrimp abound, there is the possibility of considering each locally-generated snap to be a deterministic source, and to beamform and search for scattering from possible targets in a coherent manner, using the direct received snap as a matched filter. Depending on the snapping shrimp density (number of snaps per square metre per second), there is usually a substantial gap in time between each locally-generated snap and the next (following a Poisson distribution in time), during which coherent matched filtering can be applied to search for scattered replicas of the snap. Within a range of 100-200 m, depending on the snap density, it is estimated that this method can produce more accurate and high-dynamic range images compared to statistical techniques, comparable to a multi-static active system. While the range is very short compared to most sonar systems, ANI is completely covert and can image completely silent targets. In shallow water, where the threat is often from smaller, very quiet vehicles and divers, this could be a useful tool in the shallow-water ASW armoury. Furthermore, using snapping shrimp implies a diversity of source locations and orientations to possible targets that potentially provides more information than conventional multi-static active systems, while remaining completely covert. Introduction The research area of forming images from underwater ambient noise began with preliminary ideas from Flatte’ and Munk (who considered ‘acoustic shadowing’ by large objects for a JASON report) [1]. Prof. Buckingham and colleagues advanced this early idea of ambient noise shadowing to a more complete concept of ‘Acoustic Daylight’ (AD) [2], which included the possibility of increased illumination from ‘front-lighting’ in addition to silhouetting by ‘backlighting’ in a statistical sense. These ideas were based on a very close analogy with optical vision, where the visual clues are all spatial, the human eye being too slow to interpret temporal variability in the illuminating signals. This analogy is unnecessarily restrictive for the ocean acoustic case, where the temporal statistics are of a longer timescale and the sensors’ response time much shorter than that of the eye. While the theoretical possibility of imaging using the AD technique became demonstrated fact with the first deployment of ADONIS in 1994 [3], it must be