Google’s submarine cable shows potential for earthquake early-warning systems

February 25, 2021
Google has provided a new way to detect waves in water and earth. (Pixabay/Elias Sch.)

Google has provided a new way to detect waves in water and earth. (Pixabay/Elias Sch.)

Geoscientists harnessed data signals transmitting through a Google-owned fiber-optic cable that runs more than 6,200 miles under the Pacific Ocean seafloor to record dozens of earthquakes and storm swells, detecting seismic activity from changes in the signal’s polarization.

The measurements, detailed in a paper that will be published Friday in Science, offer a new way to someday improve early-warning systems for earthquakes and tsunamis without constructing new infrastructure. Few ocean-based seismic sensors exist because of the challenges in deploying and maintaining them.

Hundreds of cables on or below the ocean floor transmit 99% of international data using optical fibers, thin tubes of glass that carry signals in the form of lasers. These submarine cables have been increasingly investigated as potential seismic sensors in the last two decades, according to Zhongwen Zhan, a professor of geophysics at the California Institute of Technology and the paper’s lead author.

A technology named distributed acoustic sensing allows earthquake-created strain against a cable to be detected by observing changes in scattering light caused by the optical fiber’s natural imperfections. However, it requires “dark fibers” that are not transmitting data and therefore limits which preexisting submarine cables can apply the technology, given that new cables are very expensive.

Scientists from Caltech, Google and the University of L’Aquila in Italy tested a new approach that used signals from active cables to measure seismic activity. For nine months, the researchers monitored Google’s submarine Curie cable, which in early 2020 began delivering 72 terabytes per second between Los Angeles and Valparaiso, Chile. 

From one end of the cable, they monitored the laser’s polarization — the orientation of the light waves — and interpreted variations as earthquakes or waves that were pushing against the cable. It was sensitive to disturbances because most of its length was under the seafloor, where it remained at constant temperatures.

The team recorded roughly 20 earthquakes, including the 7.4 magnitude Oaxaca earthquake that struck Mexico in June. The researchers also recorded about 30 ocean storm swells, which they said demonstrated the technology's potential to detect tsunamis. No tsunamis formed along the east Pacific coastline during the monitoring — besides one-centimeter-tall waves created by the powerful earthquake, although those were not detected through the cable.

“The approach we come up with in this paper is … can we just use the telecommunication signals they are sending anyway to monitor earthquakes?” Zhan said. “The exciting conclusion we have is yes, we can.”

According to Zhan, Google was interested in the scientists' research and was already monitoring the polarization of its data signals. The company uses differently polarized signals to double the the amount of data it sends through its cables in a method called polarization multiplexing, which is also used in radio signals and satellite communications.

Beyond supporting its employees who were coauthors of the paper, Google and its parent company Alphabet did not provide funding for the study. According to the paper, the research was funded by the Gordon and Betty Moore Foundation, which was started by a cofounder of Intel, and one author was partially funded by an Italian government program.

The cable-turned-seismic sensor had its limitations under the polarization approach. The researchers did not know where it was along the thousands of miles of Curie cable that seismic disturbances they measured were happening — unlike in dark-fiber approaches, which can pinpoint the source of a signal to within a few meters on the cable.

The team also did not detect an earthquake with a magnitude lower than 4.4 because of ambient noise in the signal, which may have come from temperature variations in the less than 5 miles of cable running over land.

Yet this prototype of the new measurement method was broadly successful and would also be easily scalable, Zhan said, because it does not require new infrastructure and could instead be performed on any submarine cable with polarization measurements.

Applied widely, it could also flesh out the currently sparse ocean-based seismic sensors, which would teach scientists more about offshore tectonic activity and provide people better warning of an incoming earthquake or tsunami, Zhan said.

The polarization approach is the latest of several technologies that turn submarine telecommunication cables into useful ocean sensors that, with the right equipment, could track seismic activity and even the effects of climate change on deep-ocean temperatures and sea-level rise, according to William Wilcock, a marine geophysicist at the University of Washington.

“Where cables do cross ocean basins, they could transform academic seismology by providing new observations in regions that are presently poorly observed,” Wilcock, who was not involved in the study, wrote in a perspective article that will be published in the same issue of Science.

Zhan said there is more research to be done to better interpret the detected changes in polarization and reduce the noise that cables pick up, but he also recognized the method’s potential among preexisting technology.

“I believe there are technologies that are emerging in this fiber-sensing community, so the polarization one we presented is a new one,” he said. “All these methods can probably work together to get the best combinations of sensitivity and resolution.”

The article, “Optical polarization-based seismic and water wave sensing on transoceanic cables,” will be published Feb. 26 in Science. The authors of the study were Zhongwen Zhan and Jorge Castellanos, California Institute of Technology; Mattia Cantono, Valey Kamalov, Rafael Müller and Shuang Yin, Google; and Antonio Mecozzi, University of L’Aquila. The lead author was Zhongwen Zhan.

We use cookies to improve your experience on our site and to show you relevant advertising.