Passive acoustic tracking of divers and dolphins

This thesis describes work performed in the analysis and development of positioning algorithms for self-noise of a known kind; it also describes the development of short base-line systems capable of positioning the sources. Many studies of wild cetaceans rely on tracking the movement of wild animals, often in hostile conditions and with limited contact with these animals. Advanced technology exists for satellite or radio tracking of marine wildlife, but this relies on an animal being first caught and tagged. In situations where random interactions with marine wildlife are to be analysed, it is not practicable to attach devices to an animal, so it is appropriate to use passive techniques, in which the animal's self noise is located and tracked. Present passive systems usually include a long base-line array, which can be difficult to deploy. The problem may be overcome by reducing the array in size, but it results in an increase in positioning errors. This study attempts to quantifY these errors and looks into the practicability of short base-line passive arrays. Two systems are described here, both for tracking impulsive sounds in real time. The first is for use on pelagic trawl nets, the other as a prototype high-speed system to prove the different algorithms developed before and during this study. The prototype systems, each having a minimum of four receivers positioned in various configurations, have been tested in a tank with a controlled sound source. The source is a 'pinger', which allows the systems also to be adopted for diver positioning and tracking. A survey of unclassified literature has shown characteristics of cetacean acoustic signatures, which have been utilised in the optimisation of the systems. The physiology of cetaceans has also been reviewed to help understand the physical limitations of the systems presented.