A smart fiber-optic towed hydrophone array for open-sea cruising observation

Dear Editor,

Ocean observation has traditionally relied on surface and spaceborne platforms, including buoys, satellite altimeters, radar, optical imagery, and AIS-based vessel tracking, which provide broad coverage of surface waves and maritime activity.1,2 Beneath the sea surface, interferometric fiber-optic hydrophones provide passive, lightweight, multiplexed, and high-sensitivity acoustic sensing.3,4 Distributed acoustic sensing (DAS) on submarine cables has also enabled long-distance observations of seafloor earthquakes, ocean-solid Earth interactions, storms, ships, and marine bioacoustics.5,6 However, standard fibers and telecom cables have limited responsivity to weak underwater signals. Recent helically wound and sensitized DAS configurations improve pressure-to-strain coupling and support towed hydrophone and marine streamer applications.7,8 Nevertheless, the broader capability of DAS hydrophones in a moving towed cable remains underexploited. Here, we demonstrate a self-aware hybrid fiber-optic towed hydrophone array that integrates a DAS section with a high-signal-to-noise ratio (SNR) interferometric section, enabling depth and array-shape self-sensing, ocean-wave observation, and underwater acoustic detection within one intelligent composite cable.

In the present sea trial, the deployed towed cable had a total length of approximately 770 m, and the sensing array was towed about 500 m behind the vessel. Within the same cable, the interferometric and DAS sections used separate sensing fibers and independent optical interrogation paths. The interferometric section contained 128 hydrophone elements over approximately 104 m, while the DAS section contained 32 hydrophone elements with a 5 m spacing over approximately 160 m (Figure 1A). Both sections employ compact fiber-wound air-backed mandrels with a length of 42 mm and an outer diameter of 11.6 mm as pressure-sensitive transducer structures, but they differ in wound-fiber length and optical readout mechanism. Each interferometric hydrophone element contains 22 m of wound fiber in a push-pull configuration and forms the high-SNR acoustic backbone, whereas each DAS hydrophone element contains 55 m of helically wound sensing fiber and converts pressure- and temperature-induced strain into distributed optical phase responses. Embedded depth sensors and weak fiber Bragg gratings provide depth references, gauge-length marking, and temperature monitoring. This passive and highly multiplexed architecture assigns complementary functions to the two mechanisms: sensitive acoustic detection by the interferometric section and distributed depth, wave, and acoustic sensing by the DAS section.

Figure 1B shows the phase response of a representative DAS element during the sea trials. The subsequent functional demonstrations use frequency-partitioned interpretation: the quasi-static (<0.02 Hz) component for depth sensing and array-shape inversion, the 0.02–0.3 Hz band for ocean-wave observation, and the 5 Hz–2 kHz band for acoustic detection. This frequency-partitioned logic is central to the system: the same DAS cable acts as a distributed physical interface linking the towed array, the ocean environment, and the acoustic field.

We first examined whether DAS can sense depth comparably to a depth sensor (Figure 1C). When a towed element undergoes vertical motion, the hydrostatic pressure acting on its package changes, and this pressure perturbation modulates the DAS phase. In Figure 1C1, the DAS-derived depth, converted using a measured pressure sensitivity of −129 dB re 1 rad/μPa, agrees well with the depth variation measured by the depth sensor, with a coefficient of determination (R2) of 0.99. This result indicates that the quasi-static DAS phase contains a stable hydrostatic-pressure component and can therefore serve as a measurable observable for distributed depth sensing along the towed cable.

The time-domain DAS-derived depth also follows the depth-sensor record with an average relative deviation of 0.80% (Figure 1C2), demonstrating recovery of slow cable-depth variations. The DAS element has a measured temperature sensitivity of 1.61 × 104 rad °C−1, but the water-column temperature variation during the trial was only on the order of 10−2°C. The thermal phase contribution was therefore much smaller than the pressure-induced phase change, and pressure dominated the depth inversion. Under the uncommon condition of large temperature variations, the embedded weak FBG can provide real-time temperature monitoring to compensate for the temperature-induced phase contribution before depth inversion.

Using all valid DAS elements, we reconstructed the time-varying depth profile of the DAS section (Figure 1C3). The result shows that the actual towed cable was not an ideal horizontal array under towing and hydrodynamic forcing. Unlike conventional towed arrays that rely on separate and sparsely deployed depth-sensing units, each DAS hydrophone element provides local depth information at the acoustic-sensing position, yielding denser depth constraints for array-shape inversion and subsequent depth-aware beamforming correction.