A real-time framework for mapping subsea cable burial state using Poincaré spectral coherence of DAS measurements

Distributed acoustic sensing (DAS) on subsea fibre-optic cables is emerging as a powerful tool for underwater acoustics, provid�ing dense, kilometre-scale measurements of sound propagation through the water column, the seabed, and the cable’s ambient environment. These observations enable new approaches to environmental acoustic monitoring and subsea-infrastructure assessment, including the detection of oceanographic processes, anthropogenic noise, and geophysical wavefields. However, a central challenge remains: fidelity of DAS measurements depends critically on acoustic coupling between the cable and its surroundings, i.e., variations in burial, exposure, and suspension alter the incident acoustic energy coupling into the fibre,
introducing inconsistencies or artefacts in environmental and structural interpretations. Detecting these coupling states directly from DAS data is difficult because the signatures are subtle and datasets are exceptionally large. We introduce a simple, scalable method based on Poincaré spectral coherence. It quantifies the consistency of neighbouring channels across selected acoustic frequency bands. Buried segments show smooth, coherent spectral behaviour, whereas exposed or suspended sections exhibit sharp spatial variability. Applied to two shallow-water deployments, including a 5.8-km coastal cable with
diver-verified burial, the method reliably identifies major coupling transitions. Its unsupervised, computationally efficient, real-time compatibility strengthens the case for DAS as a next-generation underwater vibrations sensing technology.