Antartic blue whale localization with ocean bottom seismometers in southern indian ocean

• Main Authors: DREO, Richard, BOUFFAUT, Léa, GUILLON, Laurent, LABAT, Valérie, BARRUOL, Guilhem, BOUDRAA, Abdel-Ouahab
• Format: Other/Unknown Material
• Language: English
• Published: 2017
• Subjects: Sciences de l'ingénieur [physics]/Acoustique [physics.class-ph]; Sciences de l'ingénieur [physics]/Traitement du signal et de l'image; Antarctic; Indian; Antarc*; antartic*; Blue whale
• Online Access: https://oskar-bordeaux.fr/handle/20.500.12278/76727

While visual survey of whales requires substantial means for limited areas, passive acoustic monitoring (PAM) offers larger scale coverage for long periods and less costs. It usually provides information about species behavior, e.g. seasonal movements, but tools are needed to detail the individuals' behavior. From October 2012 to November 2013 as part of the German-French "RHUM-RUM" (Réunion Hotspot and Upper Mantle - Réunion Unterer Mantel) seismic experiment, a 70km by 40km array of 8 Ocean Bottom Seismometers (OBS) was deployed in Southern Indian Ocean in a mountainous area, with depths from 2500 to 5500 meters. The [0-50] Hz-frequency band covered by the OBS's hydrophone provides observations about whales. Each source-OBS path has its own acoustic propagation. Indeed, closest OBS can be reached by direct rays, while remote OBS can only be reached by multi-reflected rays. Therefore, the localization problem cannot be solved directly using a classical Time Difference Of Arrival (TDOA) algorithm. In this work, the TDOA problem is solved in the case of long range detection, even with mountainous relief, enabling localization and tracking of whales. For each point of the spatial matrix representing the area, Times Of Arrival (TOA) of signal on the OBS are computed with a ray tracing algorithm (BELLHOP), taking into account the bottom profile. The theoretical corresponding TDOA are then compared to measured ones using a loss function. The obtained results, using L1, L2, cross-correlation cost functions, show the effectiveness of the proposed strategy to track whales on their calls. For example, an Antarctic blue whale is tracked during 10 hours from 40 kilometers south of the array center to 40 kilometers north where the mean speed is close to 10 km/h on a straight trajectory.