The significance of passive acoustic array-configurations on sperm whale range estimation when using the hyperbolic algorithm
Vallarta Hernandez, Jonathan
MetadataShow full item record
In cetacean monitoring for population estimation, behavioural studies or mitigation, traditional visual observations are being augmented by the use of Passive Acoustic Monitoring (PAM) techniques that use the creature’s vocalisations for localisation. The design of hydrophone configurations is evaluated for sperm whale (Physeter macrocephalus) range estimation to meet the requirements of the current mitigation regulations for a safety zone and behaviour research. This thesis uses the Time Difference of Arrival (TDOA) of cetacean vocalisations with a three-dimensional hyperbolic localisation algorithm. A MATLAB simulator has been developed to model array-configurations and to assess their performance in source range estimation for both homogeneous and non-homogeneous sound speed profiles (SSP). The non-homogeneous medium is modelled on a Bellhop ray trace model, using data collected from the Gulf of Mexico. The sperm whale clicks are chosen as an exemplar of a distinctive underwater sound. The simulator is tested with a separate synthetic source generator which produced a set of TDOAs from a known source location. The performance in source range estimation for Square, Trapezium, Triangular, Shifted-pair and Y-shape geometries is tested. The Y-shape geometry, with four elements and aperture-length of 120m, is the most accurate, giving an error of ±10m over slant ranges of 500m in a homogeneous medium, and 300m in a non-homogeneous medium. However, for towed array deployments, the Y-shape array is sensitive to angle-positioning-error when the geometry is seriously distorted. The Shifted-pair geometry overcomes these limits, performing an initial accuracy of ±30m when the vessel either moves in a straight line or turns to port or starboard. It constitutes a recommendable array-configuration for towed array deployments. The thesis demonstrates that the number of receivers, the array-geometry and the arrayaperture are important parameters to consider when designing and deploying a hydrophone array. It is shown that certain array-configurations can significantly improve the accuracy of source range estimation. Recommendations are made concerning preferred array-configurations for use with PAM systems.