Nuclear Monitoring System Suggests Landslide Cut off Internet in West Africa

Hydroacoustic signals captured by the world’s international nuclear monitoring system suggest an underwater landslide may have broken communications cables in March 2024, disrupting internet traffic in west African countries for several weeks.

Researchers used data collected by hydrophones installed by the International Monitoring System of the Comprehensive Nuclear-Test-Ban Treaty Organization to determine the location of the possible landslide, placing it along the steep slopes of Trou Sans Fond Canyon offshore of Ivory Coast.

The proposed landslide corresponds with the timing and location of four broken cables in the canyon, according to Vaibhav Vijay Ingale, a postdoctoral scholar at UC San Diego’s Scripps Institution of Oceanography. Ingale and colleagues recently shared their findings in the journal  Seismological Research Letters.

“This detection off the Ivory Coast is particularly exciting because it demonstrates the potential of using existing hydroacoustic data to monitor submarine landslides more effectively,” said Ingale. “It suggests that there could be many more events like this happening that we’re simply not aware of, either due to a lack of monitoring infrastructure or because we haven’t been actively looking for them in the hydroacoustic data.”

Four communications cables broke on March 14, 2024 in the Trou Sans Fond Canyon, about 107 kilometers (66 miles) offshore from the city of Abidjan in the Ivory Coast. The extent of the service disruptions made it important to determine the cause of the breaks, said Ingale.

The researchers decided to look for “acoustic detections of any signal behind the incident, as these low-frequency waves can tell us about different sources like earthquakes, volcanic eruptions, submarine landslides and biological activities in the ocean water column,” he noted.

One of the closest sources of hydroacoustic data came from hydrophones that have been deployed near Ascension Island as part of the nuclear-test-ban treaty network. 

“When we examined the hydroacoustic data recorded between March 6 and March 22, 2024, a low-frequency signal on March 12 caught our attention,” said Ingale.

The signal was relatively short—lasting less than a minute and half—and was not detected before or eight days after the cable breaks. The signal was only detected by hydrophones, not any land-based stations. And when the researchers examined seismic data from the region, they found no events with arrival times that coincided with the low-frequency signal.

Ingale and colleagues concluded that the signal likely came from a submarine landslide, making this the first reported instance of detecting such a landslide using hydrophones.

“Since this was the first instance of detecting a submarine landslide with hydrophones that wasn’t associated with an earthquake or eruption, there was no precedent for how the signal should appear,” said Ingale. “We had to carefully scan the available data for anomalous patterns that didn’t resemble known tectonic or volcanic signals. The difficulty was compounded by the fact that hydrophone data can be noisy due to ocean sounds like marine life, vessel traffic and other anthropogenic interactions.”

Once the geophysicists and acousticians had confirmed the presence of a “true” landslide signal, they used the signal data to calculate where the signal originated, placing it in a location consistent with the cable breaks and the steep slopes of the underwater canyon.

Ingale said if hydrophones can reliably detect signals from submarine landslides, they could be used as part of an early warning system for cable operators, helping them identify threats and prepare for disruptions.

“Furthermore, insights from hydroacoustic monitoring can lead to better engineering standards, such as deeper burial of cables in sediment-prone areas or rerouting around historically unstable slopes,” he suggested. “In cases where a cable break occurs, analyzing hydroacoustic data can help determine whether a landslide was the cause, aiding forensic analysis, insurance claims and understanding broader risks to undersea infrastructure.”

In addition to Ingale, the study was co-authored by Ross Parnell-Turner and Wenyuan Fan from Scripps Oceanography, Peter Talling from Durham University, and Jeffrey Neasham from Newcastle University. The study was funded by UC San Diego’s Institute of Geophysics and Planetary Physics’ Green Foundation Fellowship.

– Adapted from the Seismological Society of America