| Summary: | This study sought to experimentally quantify the sonar performance of omni-directional receivers as a means to passively detect vocalizing Odontocetes. To accomplish this objective, controlled experiments using a calibrated mid-frequency sound source were conducted on the San Clemente Island Underwater Range (SCIUR) in July 2002. Six Odontocete signals were selected for transmission: 2 orca and 2 pilot whale whistles, and sperm whale and Risso's dolphin clicks. Several hundred iterations of each signal were broadcast at stations 300 m to 12,000 m from the range's moored, three-element array. Statistical analyses were performed on the output of an energy and matched filter detector to quantify detection probability (P(D)) and range limits as a function of false alarm rate (P(FA)), signal type, and signal to noise ratio (SNR). The matched filter was generally the superior performer, detecting the orca2 and pilot1 whistles beyond 5000 m with a 90% P(D), 1% P(FA), and source level (SL) of 140 dB re 1 æPa. For the same conditions, the orca1 and sperm whale calls were detected at 1500 m, but the pilot2 and Risso's dolphin signals were not detected at the peak realized SNR of (-2) dB. The energy detector had no detections with a 90% P(D) and 1% P(FA) at this (-2) dB SNR, but all signals except one orca whistle were detectable beyond 1000 m with a 50% P(D) and 1% P(FA). The sperm whale was the exceptional energy detector performer, with detection ranges exceeding 7 km (140 dB re 1 æPa SL) at the 50% P(D) and 1% P(FA) Approved for public release; distribution is unlimited. Lieutenant Commander, United States Navy http://archive.org/details/assessingperform109451169
|
|---|