Earthquakes are detected primarily by seismometers, instruments that sense ground vibrations and convert them into electrical signals. More than 3,000 seismic stations operate across the United States alone, and thousands more span the globe, working together to pinpoint where earthquakes start and how strong they are. The basic principle behind detection hasn’t changed much in nearly two millennia, but the technology has become remarkably precise.
How a Seismometer Works
At its core, every seismometer relies on inertia. A heavy mass is suspended inside the instrument by a spring. When the ground shakes, the frame of the seismometer moves with it, but the suspended mass tends to stay still. That difference in motion between the frame and the mass is what the instrument actually measures.
Converting that physical movement into usable data typically involves a magnet and coil assembly. As the mass shifts relative to the coil, the changing magnetic field generates a small voltage. That voltage is proportional to the speed of the movement, giving scientists a continuous electrical signal they can amplify, digitize, and record on a computer. The output is a seismogram: a squiggly line that charts ground motion over time. Larger wiggles mean stronger shaking, and the pattern of those wiggles reveals a surprising amount about the earthquake itself.
Two Types of Waves Tell Different Stories
When rock fractures deep underground, the rupture sends out two main types of seismic waves. P-waves (primary waves) are compressional, meaning they push and pull the ground in the direction they travel, similar to how sound moves through air. They’re the fastest seismic waves and arrive at a station first. S-waves (secondary waves) move the ground side to side and travel more slowly.
This speed difference is the key to locating an earthquake. A seismometer first picks up the P-wave arrival by detecting a sudden change in the power and frequency of its signal. The S-wave shows up later, buried within the tail end of the P-wave signal. Scientists identify S-waves by examining the two horizontal components of ground motion and looking for the characteristic side-to-side polarization pattern. The longer the delay between the P-wave and S-wave arrivals, the farther away the earthquake occurred.
Finding the Epicenter
A single seismometer can tell you how far away an earthquake was, but not the direction it came from. That distance defines a circle on the map: the earthquake happened somewhere along that ring. Add a second station and you get two circles that intersect at two points. A third station produces a third circle, and the spot where all three circles overlap is the epicenter. This method, often called triangulation, is the standard approach the USGS and other agencies use worldwide.
In practice, dozens or even hundreds of stations may record the same earthquake. More data points mean a more precise location, and modern software can calculate an epicenter within seconds of the first waves arriving. The depth of the earthquake is estimated from additional wave arrival patterns, since deeper events produce slightly different timing relationships between stations.
How Earthquake Size Is Measured
Most people have heard of the Richter scale, but seismologists largely retired it decades ago. The replacement, called the moment magnitude scale, is now the global standard because it works reliably across the full range of earthquake sizes, from barely perceptible tremors to the most catastrophic events on record.
Moment magnitude is based on a physical quantity called seismic moment, which accounts for two things: how much the rock slipped along the fault and the total area of the fault surface that ruptured. Scientists estimate moment from seismograms and sometimes from ground deformation measurements taken by satellites. A standard formula then converts the moment into a familiar single number. Because it’s tied to the actual physics of the rupture rather than just the height of wiggles on a specific instrument, moment magnitude doesn’t “max out” the way the Richter scale did for very large earthquakes.
That saturation problem was real. Traditional broadband seismometers can become overwhelmed during intense nearby shaking. The feedback electronics that keep the instrument’s mass centered simply can’t keep up, and the signal clips, like a microphone distorting at a loud concert. When that happens, S-wave readings underestimate the earthquake’s true size. P-waves, which arrive before the most violent shaking, are less affected and can still provide accurate early magnitude estimates.
Newer Detection Technologies
Traditional seismometers aren’t the only tools anymore. High-rate GPS stations (part of a broader satellite navigation category called GNSS) can track the actual displacement of the ground during an earthquake with accuracy down to a few millimeters on the horizontal plane. Unlike seismometers, which measure how fast the ground is moving, GPS stations measure how far the ground has physically shifted. This makes them especially valuable during large earthquakes, where seismometers may saturate but GPS stations keep providing clean data. GPS-based detection now achieves timing accuracy within about one second of traditional seismic solutions for identifying when an earthquake begins.
Another emerging approach uses existing fiber optic cables as makeshift seismometers. A technique called distributed acoustic sensing (DAS) sends laser pulses through a fiber optic cable and measures tiny distortions in the reflected light caused by ground vibrations. Every meter of cable essentially becomes a sensor. This is particularly useful offshore, where installing traditional seismometers on the ocean floor is expensive and logistically difficult. DAS systems can also be tuned to avoid the saturation problems that plague conventional sensors, making them a promising option for early warning applications near large faults.
Earthquake Early Warning Systems
Detection speed matters most when it feeds into early warning. The ShakeAlert system, which covers the west coast of the United States, uses data from the USGS Advanced National Seismic System’s network of over 3,000 stations. The system detects the fast-moving but less destructive P-waves and quickly estimates where the earthquake is and how strong it will be. It then sends alerts to people in the path of the slower, more damaging S-waves before that shaking arrives.
How much warning you get depends entirely on your distance from where the earthquake started. If you’re close to the epicenter, you might get only a few seconds. Farther away, warning times can stretch to several minutes. Even a few seconds is enough to drop under a desk, step away from hazardous equipment, or trigger automatic systems that slow trains and open firehouse doors. Within minutes of a significant earthquake, the network also generates a ShakeMap showing the geographic distribution of ground shaking, which emergency responders use to direct resources to the hardest-hit areas.
The Oldest Earthquake Detector
The idea of using inertia to detect earthquakes goes back nearly 2,000 years. Around 132 CE, a Chinese inventor named Zhang Heng built a bronze vessel about two meters across. Eight dragon figures were mounted on the outside, each facing a cardinal or intercardinal direction, with a bronze toad sitting below each dragon’s mouth. Inside the vessel, a central pillar acted as a pendulum. When seismic waves tilted the device, the pillar would swing, triggering a mechanism that caused one dragon to release a ball into the toad’s mouth below it. The direction of the dragon that dropped its ball indicated which direction the earthquake had come from. The principle of inertia at its center is the same one that drives every seismometer built since.