How Do Earthquake Warning Systems Work?

Christian Sager


Human beings haven't discovered how to predict earthquakes yet... yet the common toad can detect seismic activity days before it strikes their habitat. But toads can't measure the intensity of earthquakes and locate their focal points can they? At least there we've got them beat. But after studying changes in rock structure, radon gas emissions, underground water flow and even the behavior of animals like toads themselves, we still can't forecast when these natural disasters may strike next.

If you live in an area prone to earthquakes, this uncertainty probably takes up at least a little space in the back of your mind. With the dangers involved, it would probably relieve your worry if there were a better way to know when your home was about to violently shake. In fact, earthquake warning systems do exist, used in both Japan and Mexico. They only provide a brief warning, but even that is enough to save many people's lives.

The first idea for an earthquake warning system came in 1868 from a physician in San Francisco who wanted to set up an alarm bell to warn the community. Over a century later, scientists are still looking for an answer, studying magnetic fields, gas seepage, computer simulated patterns and topographical changes recorded by radar-mapping satellites. In 1975, Chinese scientists predicted an earthquake hours before it happened based on the small foreshock tremors that preceded it. But since then, seismologists determined that foreshocks aren't actually a reliable predictor.

The systems currently in place gather data from hundreds of networked seismic monitoring stations. While still vague, these stations can alert communities to tremors, using the speed of telecommunications to outrun the quakes so people have the precious time they need to sound alarms, stop trains, control elevators, close bridges and tunnels, and to automatically turn off industrial systems like gas mains. This worked in 2011 when Japan's Meteorological Agency detected an earthquake and warned the city of Sendai that a 9.0 magnitude quake was headed their way, with only seconds to spare.

California is trying to get a system up-and-running similar to that in Japan. It works by detecting an earthquake's initiation through P-wave (primary wave) energy radiating from its center. These P-waves move faster (6 km per second) than an earthquake's S-waves, which generate the most damaging ground movements. The farther you are from the center, the more advanced warning you'll receive as the P-waves trigger alerts from various stations. The ShakeAlert system in California uses three algorithms to analyze seismic data and predict a quake's impact.

  1. Onsite: This system sends an alert when one station detects three seconds of P-wave vibrations, followed by a separate detection at a second station. Its alerts are faster than the other algorithms, but it also tends to send out false alarms.
  2. ElarmS: This alarm won't trigger unless four total stations detect P-waves, but it only takes half a second of vibration for it to activate.
  3. Virtual Seismologist: Like OnSite, this algorithm triggers after three seconds of P-waves. But it is slower than the others because it also calculates factors like regional fault hazards and station integrity.

Between these three algorithms, the ShakeAlert system attempts to estimate the size and location of the earthquake. A fourth algorithm called FinDer is also being tested, measuring ground shaking against pre-calculated values. Even though California Gov. Jerry Brown signed legislation requiring the state to develop a statewide earthquake warning system, it still requires funding before it's fully operational. Alerts aren't yet made public because California still needs more seismic stations in place, as well as an outreach program to educate citizens on how to act on the information they'll receive via email, radio, television and smartphone applications. Currently, only a select group of researchers and organizations (Disneyland for instance) receive warnings from ShakeAlert. To build a statewide system will cost at least $80 million over five years, with an additional $16 million a year if the system were to expand to cover the entire West Coast of the United States.

By combining the P-wave detecting algorithms with Global Positioning Systems (GPS), researchers think they can significantly improve the accuracy of these earthquake warnings. The inclusion of GPS data on the physical displacement of the ground will help estimate how an earthquake fluctuates as it travels hundreds of kilometers. On their own, GPS measurements have too low of a sample rate to accurately predict P-waves, but integrated with seismic detection both sensors can locate earthquakes more precisely.

Scientists also hope that additional GPS data will help reduce what's known as "the blind zone," the area closest to the quake's epicenter. In this zone, shaking occurs only a second after P-waves are detected, giving no time for warning. Regardless of how close nearby seismic stations might be, a recent study indicates that the typical California earthquake has a blind zone of 32 kilometers (19.88 miles).

So while we haven't quite caught up to the mysterious earthquake sensing powers of the common toad, human beings are making significant progress detecting these lethal events. As we saw in Japan, even a few seconds can make a huge difference in the cost of property damaged and the amount of lives lost.