Tracking ships with submarine cables: Unsupervised arrival time detection and event localization using distributed...

Monitoring marine traffic is crucial for maritime safety, collision prevention, and the protection of seabed infrastructure such as pipelines, power, and telecommunication cables. This study investigates the use of distributed acoustic sensing (DAS) for ship detection in the Trondheimsfjord, Norway, using a fiber-optic telecommunication cable trenched at the seabed. By analyzing continuous ship-generated acoustic signals with persistent homology, direct arrival time curves are extracted and inverted for traveltimes, enabling ship localization. The results show that persistent homology effectively detects arrival time patterns in continuous signals in DAS data, with detected ships located up to 1.6 km away from the fiber. The localization results are in good agreement with automatic identification system (AIS) data, and an unidentified vessel absent from AIS records was also detected. These findings highlight the potential of DAS for real-time marine traffic monitoring, offering new possibilities for enhancing maritime situational awareness and mitigating the risk of damage to critical infrastructure. Additionally, this approach could contribute to collision avoidance systems aimed at reducing whale-ship strikes.