File:FEDLINK - United States Federal Collection (IA assessmentofperf1094511060).pdf

A series of 10 to 20 kHz, frequency-sweeping signals synthesizing whistles of vocalizing Odontocetes was transmitted from a J-9 sound projector suspended from the Research Vessel Pt Sur while over the U.S. Navy Southern California Offshore Range (SCORE) Underwater Acoustic Range from 11 to 13 August 2004. The transmissions were recorded by a group of seven near-bottom hydrophones of the Range. Using statistical analysis on ensembles of the repeated transmissions, the relationship between probability of detection p (D), probability of false alarm p (FA), signal-to-noise ratio (SNR) of the band-passed hydrophone data and detection range were derived for both a correlator and energy detector. To extrapolate the detection range for a different SL, a ray propagation model was employed. Additionally, the feasibility of using the near-bottom hydrophones of the Range for three-dimensional localization and for reconstructing the source signal waveform was assessed. While the experimental results show that accurate horizontal location estimates can be easily obtained through a minimization of the misfit between the observed and predicted differences in the signal arrival times at a cluster of hydrophones, a high-quality depth estimate is more difficult to accomplish. In order to choose a satisfactory depth estimator, simulated data were used to systematically quantify the sensitivity of the source depth estimates, produced by a set of commonly used frequency and time-domain processing methods to additive noise, sound-speed profile mismatch and hydrophone position errors. The simulation results suggest that a timedomain signal magnitude matching scheme consistently outperforms the other methods. The performance of this scheme was further demonstrated with experimental data. For source signal waveform reconstruction, the sensitivity of a frequency-uncorrelated, leastsquares technique to the same errors was investigated.