Estimation of seafloor seismic properties from ship noise detected by Distributed Acoustic Sensing

Distributed Acoustic Sensing (DAS) is a recent ground-breaking photonic technology allowing to transform existing fiber optic cables into dense arrays of sensors. It has proved highly performant for imaging seabed sediments and sub-surface properties due to its exceptional spatial density and its ability to acquire data in challenging environments such as the seafloor, boreholes, glaciers or volcanoes. In ocean applications, this technology leverages the existing network of submarine telecommunication fiber optic cables for seabottom monitoring, as well as for detecting noise radiated by vessels. In coastal areas, the submarine DAS cables are often buried in the seafloor to prevent damage from marine life or manmade objects such as anchors. The fact that the fiber is buried improves the coupling with the ground , although burial depths are typically unknown.

In this study, we analyze a rare dataset from a submarine optical fiber offshore Marseille, France, where we have access to the burial depth to estimate seismic properties of the sediments of the seabed, such as P wave velocity, using the detected noise radiated by ships. We compare theoretical strain at the seafloor induced by an incident pressure wave in the water column, with the real longitudinal strain recorded by the DAS technology in a buried section of the cable. We consider several vessels crossing the cable obliquely, with different crossing angles and vessel's characteristics. Based on this comparison and following physical theory of wave propagation, we obtain a first order estimation of seismic wave velocity within the sediments of the sea subsurface. These results are consistent with expected velocities for Plio-Quaternary sediments, which dominate the seafloor in this region. Our results demonstrate that anthropogenic noise from ships can be effectively used to provide quantitive information on the very shallow sediment properties.