How Earthquake Early Warning Systems Work
An earthquake early warning system does not predict earthquakes. It detects one that has already started and races the news ahead of the damaging waves, buying people and machines a few seconds to react. That window sounds tiny, but it is long enough to stop trains, shut gas valves, halt elevators, flash alerts to phones, and tell surgeons to lift a scalpel away from a patient. The systems are now operational in several seismically active countries and are credited with saving lives and reducing secondary damage.
Why P-Waves Beat S-Waves
The principle relies on a basic fact about the waves an earthquake produces. A rupture sends out two main types of body wave. Primary waves, or P-waves, are compressional: they push and pull the ground in the direction of travel, travel fastest (around 6 kilometers per second in crustal rock), and are usually felt as a brief jolt. Secondary waves, or S-waves, are shear: they move the ground side to side, travel more slowly (around 3 to 4 kilometers per second), and carry most of the damaging shaking.
Because P-waves outrun S-waves, a sensor close to the epicentre can detect the P-wave and transmit an electronic signal - moving at the speed of light - to a city before the destructive S-wave arrives. The further the city is from the epicentre, the longer the warning. Stations up to about 60 kilometers away typically get tens of seconds; right next to the fault, the S-wave arrives almost with the P-wave, so the warning is zero. The system is therefore most useful for the short warning in the middle ring, not the immediate epicentral zone.
From Sensor to Alert
A modern network is built from hundreds or thousands of seismometers spread across a region, linked to data centres that estimate magnitude and location within seconds. Algorithms scan the first few seconds of the P-wave: its amplitude and frequency give a quick estimate of how big the eventual shaking will be. If ground motion is predicted to exceed a threshold, alerts go out by phone, broadcast, train signalling, and automated industrial controls.
This works because processing is fast, but false alarms are a constant trade-off. Operators tune the thresholds to balance useful warning against unnecessary panic, and most systems assign confidence levels or only alert above a magnitude that produces widespread damage.
Country Examples
- Japan: the nationwide J-ALERT and Earthquake Early Warning system pushes sirens and phone notifications via public broadcast and mobile carriers, and is wired into bullet-train braking.
- Mexico: SAS and its successor SASMEX issue alerts to loudspeakers, radio, and TV, designed for the long distance between the Pacific coast and inland cities.
- United States: ShakeAlert, run by the US Geological Survey, covers the West Coast and delivers through apps, wireless emergency alerts, and partnerships with utilities and transit.
Key Takeaways
- These systems detect an earthquake already in progress and warn ahead of the slower, damaging S-waves.
- More distance from the epicentre means more warning time, but zero warning right at the fault.
- Alerts power automated responses - stopping trains, closing valves, alerting phones - more than they give individuals time to flee.