Trabattoni Alister, Barruol Guilhem, Dréo Richard, Boudraa Abdel
Institut de physique du globe de Paris, Université de Paris Cité, CNRS, F-75005 Paris, France.
Ecole Navale IRENav/Arts & Métiers ParisTech, BCRM, CC 600, 29240 Brest, France.
J Acoust Soc Am. 2023 Jan;153(1):260. doi: 10.1121/10.0016810.
We report in this study how ocean-bottom seismometers (OBS) can be used as passive sonars to automatically detect, localize, and track moving acoustic sources at the ocean surface. We developed single-station methods based on direction of arrival and on multi-path interference measurements capable of handling continuous erratic signals emitted by ships. Based on a Bayesian mathematical framework, we developed an azimuthal detector and a radial detector and combined them into a fully automatic tracker. We tested the developed algorithm on seismic and hydroacoustic data recorded in the Indian Ocean by an OBS deployed at 4300 m depth, 200 km west of La Réunion Island. We quantified the performances using archives of commercial-vessel trajectories in the area provided by the Automatic Identification System. Detectors demonstrate capabilities in the detection range up to 100 km from the OBS with azimuthal accuracies of a few degrees and with distance accuracies of a few hundred of meters. We expect the method to be easily transposed to any other kind of sources (such as marine mammals).
我们在本研究中报告了海底地震仪（OBS）如何用作被动声纳，以自动检测、定位和跟踪海洋表面移动的声源。我们基于到达方向和多径干涉测量开发了单站方法，能够处理船舶发出的连续不稳定信号。基于贝叶斯数学框架，我们开发了方位探测器和径向探测器，并将它们组合成一个全自动跟踪器。我们在留尼汪岛以西200公里、深度4300米处部署的一个OBS记录的地震和水声数据上测试了所开发的算法。我们使用自动识别系统提供的该区域商业船只轨迹档案对性能进行了量化。探测器展示了在距OBS高达100公里的探测范围内的能力，方位精度为几度，距离精度为几百米。我们预计该方法可轻松应用于任何其他类型的声源（如海洋哺乳动物）。
