Until now, if researchers wanted to record whale songs, they had to rely on individual underwater microphones. The innovative technology now makes it possible to detect the sounds of giant marine mammals on a much larger scale: using optical fiber cables already laid on the seabed. Cables are so sensitive to the slightest vibration that even the sound waves emitted by whales are reflected in the signal. In an experiment in the Arctic Ocean, scientists were able to use this data to locate whales to within a few meters.
Despite their size, whales are often difficult to track in vast oceans. While local observations and satellite images of surfacing whales allow scientists to draw conclusions about the distribution and migration of giant marine mammals, they only give an incomplete picture. Underwater microphones, the so-called hydrophones, which record whale singing, cover only a small area at a time. Especially with the changing activities of whales and humans in the face of climate change, it is important that researchers obtain more accurate data on the location of whale populations in order to be able to protect animals more effectively.
“Satellites” in the ocean
A team led by Léa Bouffaut from the Norwegian University of Science and Technology in Trondheim has successfully tested a new approach that could track whales across large areas of the ocean from their singing and even monitor them in real time. To do this, scientists used existing fiber optic cables laid on the seabed and connected them using a technique called Distributed Acoustic Sensing (DAS). The DAS technology enables the transmission of light pulses through unused optical fibers by means of an interrogator and the assessment of their transit time. Vibrations in the cable affect the propagation time of the signals. Therefore, the system is already used to detect earthquakes.
Bouffaut and her team focused on much weaker tremors: the ones that cause the whale’s call sound waves. “I believe this can change the field of marine bioacoustics,” says Léa Bouffaut. “Using hydrophones is extremely expensive. But fiber optic cables are found all over the world and are readily available. In my opinion, this system could become something like satellites in the ocean. ‘
A new kind of data
The research was carried out in the Svalbard archipelago in Norway. In this part of the Arctic Ocean, underwire whales like the blue whale foraging in summer. For 44 days, scientists recorded signals from the fiber-optic cables laid there and received a huge amount of data, about seven terabytes a day. “If something moved or made a noise near that fiber buried in the seabed, we could measure it,” reports Bouffaut’s colleague Martin Landrø. “So we saw a lot of ship traffic, many earthquakes of course, and we could even see storms that were far away. And finally: whales. In total, we recorded at least 830 whale calls. ”
The first challenge was to interpret the signals correctly. “We searched for signals without knowing exactly what to expect,” explains Bouffaut. “This is new technology and a new kind of data that no one has used before to find whales.” However, by analyzing the frequency, pattern and repetition, they were able to clearly identify the whale songs in the data generated. With this knowledge, it will be possible to train machine learning models in the future, and thus simplify and automate data analysis.
Contribution to the protection of whales
“Thanks to this system, we have the ability to cover a large area,” says Bouffaut’s colleague, Hannah Joy Kriesell. “And because we perceive sound from different angles, we can even determine the position of an animal. If we go further, which still requires additional work, it could be real-time, which would be a really big step forward in acoustic whale monitoring. ” Especially if ship traffic in the Arctic increases with the melting of the ice sheet, it could make a huge difference to the protection of whales.
Source: Léa Bouffaut (Norwegian University of Science and Technology (NTNU), Trondheim) et al., Frontiers in Marine Science, doi: 10.3389 / fmars.2022.901348