Localisation of Sperm Whales using Bottom-Mounted Sensors
P.R. White, T.G. Leighton, D.C. Finfer, C. Powles
Institute of Sound and Vibration Research, University of Southampton, Highfield, Southampton SO17 1BJ, UK
If the movie is not visible on this page , please click on the following link: ..\Cross-faculty UAUA centre\Gallery_files\sperm_whale_local_1.gif
It is important for vessels to know the whereabouts of any whales in their vicinity, to avoid collisions with them, or to avoid harming the whales when they use their active sonar. Active sonar operates by sending out a pulse of sound or ultrasound, and looking for the echoes as this pulse bounces off objects. This is the most effective technology for finding distant objects underwater. However we cannot use active sonar to find whales if that sonar is going to harm them.
This project looks at ways of using 'passive' sonar to find out whether there are any whales nearby. When we use passive sonar, we do not send out any pulses of sound or ultrasound. Instead we listen to the noises made by other objects to find out where they are. Clearly this method does is safer to use when whales are nearby. Whilst ships also keep watch out for whales using binoculars, this does not work well at night, or in rough weather conditions, or if the whale is diving beneath the sea.
The movie on this page shows a volume of sea water from the surface down to a depth of 1.6 km, with a 'footprint' of 4 km by 4 km. The cluster of dots on the screen show where our system has estimated a sperm whale (Phyester macrocephalus) to be (four dots not belonging to the cluster are errors). The cluster of dots shows how the whale swam along a path at a depth of nearly 1 km, probably as it looked for squid to eat. It sometimes make sharp turns during this hunt, perhaps as it 'locks onto' another target squid. The whale makes a series of clicks as it swims, and it uses these as a form of 'active sonar', because there is no light at these depths to see by.
In fact, every time the whales makes a click, it bounces around the whale's head a few times, to give subsidiary clicks. The larger the whale's head, the further apart are these subsidiary clicks, so our passive sonar system can even tell you how big the whale was that we were listening to.
Technical details:
The clicks made by a real whale, as recorded using an array of five bottom-mounted hydrophones (i.e. underwater microphones). The data analysed were supplied as part of the 2nd International Workshop on Detection and Localisation of Marine Mammals, Monaco, 2005.
The localisation technique employed is based on a weighted least squares optimisation procedure applied to estimated delays. The method by which these delays are estimated in a robust manner is detailed. Three analytic acoustic models for the environment are explored and the effects of noise and model mismatch are considered.
For further details click on the link to:
White, P.R., Leighton, T.G., Finfer, D.C., Powles, C. and Baumann, O.N. Localisation of sperm whales using bottom-mounted sensors, Applied Acoustics, 67, 2006, 1074-90
This page was last updated by TG Leighton,5 May 2006