Mar 31, 2021 8:00 AM

How Cargo Ships Could Help Detect Tsunamis

By using GPS data to monitor slight changes in elevation, the world’s fleet of commercial vessels could aid in forecasting incoming waves.

This March marked 10 years since the Tohoku earthquake triggered a devastating tsunami off the northeast coast of Japan. Major tsunamis are rare throughout history, but it’s often impossible to know exactly when and where one will occur. But to improve warning, the thousands of cargo ships that traverse the world’s oceans could offer critical, if unexpected, assistance.

While ferrying goods, commercial vessels could leverage their GPS systems to form a distributed network of sensors capable of picking up subtle elevation changes indicative of a passing tsunami, a series of large waves most often caused by earthquakes under the sea. This approach, which researchers simulated in a recently published paper, aims to significantly enhance detection and forecast abilities by simply turning existing ships into a floating array for sensing waves, all without having to spend a lot of money on new infrastructure.

“With just one ship, it would be difficult to tell if the signal you’re seeing is noise or a tsunami,” says study senior author Anne Sheehan, a professor in the University of Colorado Boulder’s department of geological sciences. “But if you have a bunch of ships, then they would be going up and down in a pattern that could be used to determine whether it’s a tsunami.”

Boats steadily bob amid the waves. But a passing tsunami slightly alters the sea surface elevation, changing the average elevation throughout the rhythmic up-down cycle. “The period of the tsunami, the time over which the wave passes the ship, is on the order of 10, 20, 30 minutes, whereas ocean waves have a period of maybe 15, 20 seconds,” says James Foster, a professor at the Institute of Geodesy, University of Stuttgart, who authored an early paper on commercial ships and tsunami warnings.

His paper focused on data collected in 2010, when a tsunami generated by an 8.8 magnitude earthquake happened to pass underneath a University of Hawaii research vessel en route to Guam that was equipped with high-accuracy GPS. In the deep ocean, a tsunami wave might be more than 300 miles long but only a meter or less in height, making it unnoticeable to the people on the ships floating above. Yet GPS can detect this sustained elevation change—as first shown by Foster’s analysis of the Hawaii ship’s data—a chance finding that alerted scientists to ocean vessels’ tsunami-detecting potential.

After seismometers detect an undersea earthquake, a tsunami warning may be issued to coastal regions by national or regional emergency services. As the waves advance, offshore seafloor pressure sensors can record and help forecast the tsunami’s activity. But the cost to install and maintain such sensors and buoys makes them nonviable for many countries. Instrument outages and coverage gaps between sensors can also pose challenges.

Video: Anne Sheehan; Jakir Hossen

Roaring at speeds up to 500 miles per hour in the open ocean, tsunamis can still take several hours to reach the shore, depending on their point of origin. However, in regions like the Cascadia subduction zone, which runs along the Pacific coast from Northern California up to British Columbia, a tsunami might take only 30 minutes to reach the coastline, leaving little warning time. “We need to be prepared, if something happens, so that we can get the warning [out] and evacuate people. We need information,” says M. Jakir Hossen, a researcher with the Cooperative Institute for Research in Environmental Sciences (CIRES) at CU Boulder.

Hossen, Sheehan, and their colleagues modeled how well a cargo-ship-based sensing array might actually work. Hossen is the first author on their paper published in Earth and Space Science in February, evaluating ship-borne GPS tsunami forecasting in the Cascadia subduction zone via a computer simulation. Given the region’s steady vessel traffic, the researchers used actual ship coordinates supplied by the global data and analytics provider Spire. While marine traffic typically follows similar routes, the number and spatial distribution of ships varies, which the simulation took into account. The study also simulated tsunami-produced variations in ship elevation and velocity. The team used data assimilation, a technique that combines observations with a numerical model to improve predictions, in order to forecast the virtual tsunamis.

Supposing each ship was equipped with a GPS sensor that could precisely measure elevation (and therefore detect a passing tsunami), the simulation indicated that a 20 kilometer gap—about 12 miles—between vessels in areas with high ship density would be sufficient to make accurate forecasts and that predictions can be made reliably within 15 minutes of tsunami onset.

And that matters because the Pacific coast is due for some sizable tectonic activity from pressure buildup, according to scientists. “In the Cascadia subduction zone region, many studies show that a big earthquake is coming,” says Hossen. “We don’t know when and where it could trigger a tsunami.”

But this system wouldn’t be ready to go right away. While commercial ships routinely use GPS, they don’t report their elevation data—exactly how much they’re bobbing. Around the globe, the Automatic Identification System (AIS) continuously tracks their latitude and longitude, but these broadcasts don’t include elevation, since boats presumably stay at sea level. To detect tsunamis, these slight changes in elevation would have to be relayed in real time, but given satellite navigation’s ubiquity, including this information may be feasible.

“What I really liked about this method is that the method is cheap,” says Anne Bécel, the Lamont Associate Research Professor in the Lamont-Doherty Earth Observatory at Columbia University, who was not involved with the CU Boulder study. “If this method is fully developed, it would become very affordable for many countries that are threatened by local tsunamis.”

Commercial vessels could complement, not replace, existing tsunami detection mechanisms, while offering a much more cost-effective approach than adding new seafloor pressure sensors. While ships using GPS could help predict a tsunami’s threat by recording wave height, which correlates with its damage potential, they wouldn’t necessarily sound the alarm that a tsunami had been generated, says the University of Stuttgart’s Foster. “This system is never likely to be the thing that triggers the alarm. It’s going to be the fact that there was a huge earthquake that triggers the alarm,” he says.

Still, other geologic events—such as submarine landslides and volcanic eruptions—can cause tsunamis. A warning system based only on wave observations, and not on what triggered them, would be advantageous, says Sheehan, who’s also a fellow at CIRES. “With this method, we’re not assuming really anything about the earthquake or the landslide or about whatever causes the tsunami. We’re just looking at the waves as they are recorded by the ships, so you’re using the actual observations,” she says.

Foster says that shipping companies have been very receptive to the idea of using their boats to help forecast tsunamis. But before that can happen, scientists will need to do more research on the extent of the floating network that will be needed, as well as the precision and processing of ship-based GPS data.

While the CU Boulder study relied on a simulation, adding further data from real ships could enhance the findings, Bécel says. “The next step will really have to show that, with high-precision GPS, [researchers] have the same results with high accuracy,” she says. “Right now it looks like it’s very promising.”