Earthquake Monitor

Earthquake early warning vs detection: what your phone can and cannot do

People conflate two very different things: getting a warning seconds before the shaking reaches you, and getting a notification that an earthquake happened somewhere. They rely on completely different systems. Knowing which is which matters when you choose what to trust.

There is a moment of confusion almost everyone has the first time they install an earthquake app. They expect their phone to buzz before the shaking, like a smoke alarm. Then a notification arrives saying "magnitude 4.2, 600 km away, three minutes ago" and they wonder what the point was.

Both behaviours are legitimate. They come from two different systems that solve two different problems. Understanding the difference tells you what to expect and what to rely on.

Detection: telling you what happened

This is what most earthquake apps, including ours, primarily do. Seismometers around the world detect ground motion, the data flows to agencies like the USGS (United States) and EMSC (Europe-Mediterranean), and those agencies publish a feed of confirmed events: location, magnitude, depth, time.

An earthquake monitor app reads that feed and shows you the events, usually on a map, often with alerts you can filter by magnitude and distance.

The key fact: this happens after the earthquake. Detection, confirmation, and publishing takes anywhere from under a minute (for large, well-instrumented quakes) to several minutes (for small or remote ones). By the time you get the notification, the shaking is over - certainly if it was far away, and usually even if it was nearby.

What detection is good for:

  • Knowing whether the bump you just felt was an earthquake or a truck.
  • Checking on family in a region you saw mentioned in the news.
  • Following seismic activity near a fault, a volcano, or a place you are travelling to.
  • Building a sense of how seismically active your area is over time.

What it is not good for: warning you before the shaking. That is a different system entirely.

Early warning: buying you seconds

Earthquake early warning is a genuinely different technology, and it is remarkable when it works.

It exploits a quirk of physics: earthquakes produce two main kinds of waves. P-waves (primary) travel fast but are relatively weak. S-waves and surface waves travel slower but cause most of the damaging shaking. The gap between them grows with distance.

If a dense network of seismometers near the epicenter detects the P-wave, a system can calculate the earthquake\'s size and location in a second or two, then send a warning that races ahead of the slower, damaging waves. The further you are from the epicenter, the more warning you get - anywhere from a couple of seconds to a minute.

That is enough time to drop, cover, and hold on. Enough for surgeons to pause, for trains to brake, for gas valves to close automatically. It saves lives and reduces injuries.

Where early warning actually exists

Early warning requires a dense seismic network and government infrastructure. It exists in a limited set of places:

  • Japan: the world\'s most mature system, run by the JMA. Warnings appear on every phone and TV.
  • US West Coast: ShakeAlert covers California, Oregon, and Washington. Delivered through the MyShake app and, on supported phones, the OS-level alerts.
  • Mexico: SASMEX, one of the earliest systems, with public sirens in Mexico City.
  • Taiwan, parts of Chile, Italy, and a growing list of others.

If you are not in one of these regions, no app can give you early warning, because the underlying network does not exist where you are. This is not an app limitation; it is a "the sensors and the alert infrastructure are not deployed here" limitation.

The Android twist: crowdsourced detection

Google built an interesting middle ground into Android. Modern Android phones have accelerometers, and when many phones in an area detect shaking simultaneously, Google\'s system can infer that an earthquake is happening and issue an alert. It is effectively a giant crowdsourced seismometer.

This extends a form of early-ish warning to places without a formal network, though it is less precise and less fast than a dedicated system like ShakeAlert. Apple does not currently run an equivalent on iPhone, so on iOS you rely on official regional systems where they exist.

What this means for choosing an app

Be clear about which problem you are solving:

  • You want to be warned before shaking. Use the official early-warning channel for your region: ShakeAlert-powered apps on the US West Coast, the JMA system in Japan, SASMEX in Mexico, and so on. Make sure OS-level emergency alerts are enabled in your phone settings. No third-party "detection" app can replace this.
  • You want to know about earthquakes, track activity, and check on places. A detection app reading USGS / EMSC feeds, with customisable alerts, is exactly the right tool. That is what our Earthquake Monitor does.

The two complement each other. In an early-warning region, you would ideally have both: the official warning for the seconds-before alert, and a detection app for the broader picture - magnitude details, the live map, activity near family elsewhere.

Reading the numbers you get

When a detection alert arrives, three numbers tell the story:

  • Magnitude: the energy released. Logarithmic - a 6 is about 32 times the energy of a 5. More on magnitude scales here.
  • Depth: shallow quakes (under 70 km) are felt more strongly at the surface. A shallow 4.5 nearby can feel stronger than a deep 6.0 far away.
  • Distance: how far the epicenter is from you. Combined with depth and magnitude, this tells you whether you would have felt it.

A useful habit: set your alert threshold based on where you live. In a seismically quiet area, a low threshold (say M3.5 within 300 km) keeps you informed without spam. In an active region, raise it so you only hear about events likely to be felt, otherwise you will get pinged constantly.

The honest summary

No app can predict earthquakes. Nobody can; that is settled science, not a limitation we will engineer around. Detection apps tell you accurately what happened, fast. Early-warning systems give you seconds of notice where the infrastructure exists. Crowdsourced systems extend a rough version of warning to more places.

Use the official warning system for your region if you have one, enable OS-level emergency alerts, and use a detection app for everything else. That combination is the realistic best you can do today, and it is genuinely useful.

Keep reading