Fiber Optic Earthquake Detection: What Chile Proved and What Mexico Still Lacks
On Saturday, September 19, at 12:00 central time, the seismic alert will sound across all 32 Mexican states. The Second National Drill 2026 runs five scenarios at once, from a magnitude 7.7 near Tehuacán, Puebla, to a 7.6 southwest of Tonalá, Chiapas. One day later, in Málaga, the ECOC 2026 conference hosts a workshop on telecom fiber as a geophysical sensor. Both events point to the same question. How far along is fiber optic earthquake detection on existing cables, and what would it take to plug it into a national seismic network?
The short answer: it works, it already runs in the region, and it has limits worth understanding. Chile is turning a submarine cable into a seismic sensor. California has peer-reviewed evidence that 4 seconds of fiber data can help size an earthquake. Mexico, meanwhile, keeps expanding its seismic network with conventional stations, along routes where telecom fiber already lies in the ground. BiiSensing takes no part in any project described below.
How fiber optic earthquake detection works on a telecom cable
The technique is Distributed Acoustic Sensing (DAS). An interrogator at one end of the fiber sends laser pulses down the cable and reads the light that scatters back. Any strain along the fiber, including the ground motion of a seismic wave, shifts that backscatter. The instrument resolves that signal by distance, so one fiber behaves like a dense line of sensors. For a longer explanation, see our complete guide to distributed acoustic sensing.
Two properties make telecom fiber attractive for seismology. First, it already runs along roads, rail corridors, coastlines and seafloors, where dense coverage with conventional instruments costs the most. Second, it does not have to be dark. NEC Laboratories America reported validating DAS on 55 km of customer fiber while that fiber carried 400 Gb/s DWDM traffic. In other words, a live telecommunications link can sense and communicate at the same time. We covered the operational side in our article on dark fiber sensing.
The limits are just as real. Nobody routed a telecom cable for geophysics, so its coupling to the ground varies from one section to the next. Traffic, machinery and ocean swell add noise that the processing has to separate from seismic signals. Furthermore, the fiber measures strain along its own axis, so sensitivity depends on the angle between cable and incoming wave. However, none of this disqualifies the approach. It means every deployment needs calibration against reference seismometers before anyone trusts its output.
Chile turned a submarine cable into a seismic sensor
The most advanced Latin American case of fiber optic earthquake detection is the Cable Prat, a submarine telecommunications cable off Chile. In December 2025, operator Gtd and Universidad Técnica Federico Santa María (USM) signed an agreement to install DAS on the cable. The goal is to detect earthquakes, tsunamis and changes in waves and currents. Earlier work under the POST project showed the technology can detect earthquakes up to 10 seconds ahead of current systems.
According to the USM research center AC3E, the first phase covers 100 km from the Concón node, under researcher Marcelo Soto. The plan then extends to Cartagena, Constitución, Biobío and La Serena. The model is worth noting: an existing cable, one interrogator, a university team and the telecom regulator Subtel around the same table.
Chile is also designing sensing in from the start. Subtel commissioned a prefeasibility study from Salience Consulting and Pioneer Consulting, with CAF financing. It evaluates a 1,600 km submarine cable to Antarctica across the Drake Passage. The design would carry distributed sensors for temperature, pressure and seismic acceleration. Estimates run from about USD 370 million for the minimum configuration to USD 620 million for the optimal route.
From recording earthquakes to early warning in 4 seconds
Detecting a quake after the fact is useful. Sizing it while it is still happening is what an alert system needs. In July, researchers from the United States Geological Survey (USGS) and Cal Poly Humboldt published a study in Nature Communications (Sawi et al., vol. 17, 4776). The data came from DAS on the internet fiber between Arcata and Eureka, California. Their machine learning model reads the first 4 seconds after the P wave and classifies whether the event has reached magnitude 5.4 or more. They trained it on earthquakes from magnitude 3.5 to 7.1.
“One of the biggest challenges in earthquake early warning is determining how large an earthquake has become as quickly as possible,” said Theresa Sawi of the USGS in the Cal Poly Humboldt release.
What ECOC 2026 and the research community are debating
On Sunday, September 20, ECOC 2026 in Málaga hosts a workshop titled “Telecom Fiber as a Geophysical Sensor for Earthquake, Tsunami and Microseismic Monitoring”. Oskars Ozolinš of Riga Technical University, Steinar Bjørnstad of Tampnet and Guy Torfs of Ghent University organize the session. The program promises to “critically assess the realistic capabilities and limitations of using existing telecom infrastructure for seismic monitoring”.
Industry is reorganizing around the same idea. On September 1, NEC Laboratories America split its Optical Networking and Sensing group into two departments. Yue Tian heads the new Optical Sensing and Solutions department. Tian’s summary: “a telecom fiber already underground can also serve as a sensor for the city”. NEC’s ECOC presentations include earthquake and tsunami monitoring through telecom fiber.
Latin America is contributing research as well as test sites. On September 9, a team led by Arnaldo Leal-Junior at the Universidade Federal do Espírito Santo in Brazil published a comprehensive review of DAS in Opto-Electronic Advances. In geophysics, it lists earthquake detection, subsurface imaging and volcanic monitoring.
Mexico: 58 new stations and the fiber between them
Now back to the country where the alert sounds on Saturday. On July 29, UNAM announced 58 new stations for the Mexican Seismic Network. The National Seismological Service (SSN) runs 38 of them, and the Seismic Instrumentation Unit at the Institute of Engineering runs the other 20. The expansion targets Sonora, Chihuahua, Coahuila, Nuevo León, Tijuana and Baja California. So far, 8 of the 20 engineering stations are complete, and UNAM expects to finish the rest in 2026.
Every one of those stations is a point instrument. The data they produce travels, in UNAM’s own words, over fiber optic links, Red UNAM and radio. Between two stations there can be long stretches of cable that today only carry traffic. Chile and California show that the same cable could also produce data, filling the gaps between stations with continuous measurements. To be clear, this describes what the cable could do, not what any Mexican institution plans.
The five drill scenarios on September 19 span Puebla, Baja California, Nuevo León, Quintana Roo and Chiapas. Baja California and Nuevo León also appear where the network is growing. Therefore, a fiber optic earthquake detection pilot along one existing route, calibrated against the new stations nearby, would answer the ECOC question with Mexican data.
What to do with the fiber you already own
For an operator, a utility or a public agency, a first measurement is closer than it looks. The sequence below mirrors what the cases above share: existing fiber, one section first, and a research partner.
- Inventory the fiber. Which routes cross seismically active zones, who owns them, and whether a dark fiber or a live wavelength is available for sensing.
- Pick one section. A single stretch with a reference seismometer nearby is worth more than an ambitious map. Chile started with 100 km from one node.
- Run a proof of concept on the installed cable. Measure coupling, noise and detection performance on that specific fiber, not on a datasheet.
- Bring in the seismologists. Both cases above pair an interrogator with a university team that validates events against the official catalog.
- Decide with data. Only then does it make sense to discuss coverage, integration with an alert system, or interrogator specifications.
BiiSensing, the first Latin American company dedicated to fiber optic sensing, works on exactly those steps. We assess the feasibility of a specific fiber section and run proofs of concept on fiber that is already installed. We also support the project from the initial survey to commissioning. Our seismic and natural hazards solution describes what DAS detects in that context. If you have fiber along a route that matters, talk to an engineer about whether sensing applies to it.
