Global passive acoustic monitoring of the oceans with MERMAID Argo floats

• Main Authors: Bonnieux, Sébastien, Hello, Yann, Rocca, Frédéric, Reinicke, Isabel, Sigloch, Karin
• Other Authors: Géoazur (GEOAZUR 7329), Institut national des sciences de l'Univers (INSU - CNRS)-Observatoire de la Côte d'Azur, Université Côte d'Azur (UniCA)-Université Côte d'Azur (UniCA)-Centre National de la Recherche Scientifique (CNRS)-Institut de Recherche pour le Développement (IRD France-Sud ), OSEAN S.A.S Le Pradet, Osean
• Format: Conference Object
• Language: English
• Published: HAL CCSD 2024
• Subjects: Ocean; Acoustic; Embedded Software; Domain Specific Language; [SDU.OCEAN]Sciences of the Universe [physics]/Ocean; Atmosphere; [INFO.INFO-ES]Computer Science [cs]/Embedded Systems; Orleans; Pacific; Fin whale
• Online Access: https://hal.science/hal-04505476; https://hal.science/hal-04505476/document; https://hal.science/hal-04505476/file/OSM_2024.pdf

International audience We introduce a multidisciplinary acoustic version of MERMAID floats, equipped with a hydrophone for listening to sounds up to 20 kHz, and capable of carrying several other sensors. The MERMAID float is currently used by seismologists for earthquake recordings; the Earthscope-Ocean organization has deployed almost 100 floats in the middle of the Pacific Ocean, Mediterranean Sea and China Sea. The new version with high frequencies will enable global passive acoustic monitoring for several fields of research such as the study of cetaceans, meteorology or anthropological activity.It is capable of embedded processing to reduce the amount of data to transmit, which is limited to a maximum of a few megabytes per month. New algorithms can be developed by scientists with a specific programming language called MeLa, similar to Python or Matlab languages, to allow the development of applications with a limited knowledge of embedded systems. The MeLa language also takes into account the processing and energy limitations of the instrument, to ensure that the applications will behave as expected (e.g. without missing samples), and will not drain the battery too fast for ensuring a lifetime of 5 years per instrument.We have implemented four applications that are 1) a seismic detection algorithm based on signal level triggering and power estimation between different wavelet scales, 2) a fin whale detection algorithm based on a deep neural network, 3) a rain and wave gauge using a power ratio between two frequencies, and 4) an ambient noise measurements with usual spectral density estimation. Over the air programming allows applications to be updated or installed even once the float is deployed at sea. We are searching for collaboration scientific data interpretation and/or other applications developments. First trial in the Mediterranean Sea will occur in 2024.