Temperature is central for ocean science but is still poorly sampled on the deep ocean. Here, we show that Distributed Acoustic Sensing (DAS) technology can convert several kilometer long seafloor fiber-optic (FO) telecommunication cables into dense arrays of temperature anomaly sensors with milikelvin (mK) sensitivity, allowing us to monitor oceanic processes such as internal waves and upwelling with unprecedented detail. We validate our observations with oceanographic in-situ sensors and an alternative FO technology. Practical solutions and recent advances are outlined to obtain continuous absolute temperatures with DAS at the seafloor. Our observations grant key advantages to DAS over established temperature sensors, showing its transformative potential for thermometry in ocean sciences and hydrography.

<p><span dir="ltr" role="presentation">Ocean water temperature measurements are fundamental to atmospheric and ocean sci</span><span dir="ltr" role="presentation">ences. Obtaining them, however, often comes along with major experimental and logistic </span><span dir="ltr" role="presentation">challenges. Except for the uppermost ocean surface temperature, which can be measured </span><span dir="ltr" role="presentation">from satellites, temperature data of the ocean is often poorly sampled or nonexistent, espe</span><span dir="ltr" role="presentation">cially in deep-water regions. </span></p> <p><span dir="ltr" role="presentation">Although Distributed Acoustic Sensing (DAS) technology has become popular because </span><span dir="ltr" role="presentation">its high sensitivity to strains and mechanical vibrations, our work focuses on its usage on </span><span dir="ltr" role="presentation">tens-of-kilometer-long underwater fibre-optic (FO) telecommunication cables to measure </span><span dir="ltr" role="presentation">temperature anomalies at the seafloor at millikelvin (mK) sensitivity. This is possible because </span><span dir="ltr" role="presentation">of the lack of dominant strain signals at frequencies less than about</span> <span dir="ltr" role="presentation">∼</span><span dir="ltr" role="presentation">1 mHz, as well as the </span><span dir="ltr" role="presentation">poor coupling of the fibre with these signals while remaining highly sensitive to slow ambient </span><span dir="ltr" role="presentation">temperature variations that locally affect its optical path length. DAS allows us to observe </span><span dir="ltr" role="presentation">significant temperature anomalies at the continental shelf and slope of the Mediterranean </span><span dir="ltr" role="presentation">sea, South of Toulon, France over periods of several days, with variability remaining relatively </span><span dir="ltr" role="presentation">low at the deep ocean. By means of this approach, oceanic processes such as near-inertial </span><span dir="ltr" role="presentation">internal waves and upwelling can be monitored at unprecedented detail.</span></p> <p><span dir="ltr" role="presentation">Our observations are validated with oceanographic in-situ sensors and alternative Dis</span><span dir="ltr" role="presentation">tributed Fibre Optic Sensing (DFOS) technologies established for temperature sensing. We </span><span dir="ltr" role="presentation">outline key advantages of DAS thermometry over the aforementioned sensors in terms of </span><span dir="ltr" role="presentation">spatial coverage, sensitivity, versatility and highest attainable frequency. At the current state </span><span dir="ltr" role="presentation">of the art, DAS can only measure temperature anomalies as opposed to absolute temper</span><span dir="ltr" role="presentation">ature, a drawback that could be compensated via single temperature calibration measure</span><span dir="ltr" role="presentation">ments.</span></p>

Atmospheric phenomena and ocean waves have long been known to be intimately related, and the imprint of the latter in seismological records has been persistently pointed out (e.g.