TREMR

Every earthquake you see on Tremr — the M2.1 off the coast of Japan, the M4.8 in Turkey, the M6.0 in Peru — was detected by an instrument buried in the ground somewhere on Earth. That instrument is a seismograph, and the chain of events from ground tremor to the number on your screen is a remarkable piece of engineering that has been refined over more than a century.

The Basic Idea

The core principle of a seismograph is surprisingly simple: inertia. When the ground moves, a heavy mass suspended inside the instrument tends to stay still while the casing around it moves with the earth. The relative motion between the stationary mass and the moving casing is what gets recorded.

Think of it like a pen hanging from a spring inside a box. When the box shakes, the pen stays roughly in place (due to inertia) while the paper attached to the box moves beneath it. The result is a wavy line — a seismogram — that captures the motion of the ground over time. That wavy line is the raw material of seismology.

From Pendulum to Digital

The earliest seismographs, built in the late 19th century, used exactly this pendulum-and-drum mechanism: a heavy weight on a wire, a rotating drum of paper, and an ink pen tracing the movement. They were enormous instruments, sometimes weighing several tonnes, housed in specially built vaults to isolate them from vibration.

Modern seismographs are very different. Today's instruments use a coil of wire suspended inside a magnetic field. When the ground shakes, the coil moves relative to the magnet, generating a tiny electrical current proportional to the ground's velocity. That current is amplified, filtered, digitised, and recorded — continuously, 24 hours a day, 365 days a year — onto a computer hard drive or transmitted directly to a data centre over the internet.

Modern instruments also record motion in three directions simultaneously: up-down, north-south, and east-west. This three-axis data allows seismologists to fully characterise the motion of the ground at any point, which is essential for locating an earthquake's source and understanding its mechanism.

Reading the Waves

An earthquake generates several types of seismic waves, and a seismogram shows them arriving at different times. The first to arrive are P-waves (primary or compressional waves) — they travel fastest, pushing and pulling rock like a spring. Next come S-waves (shear or secondary waves) — slower, but they shake rock side to side and carry more destructive energy. Finally, surface waves roll across the outer layer of the earth like ripples on water; they're the slowest but often the most destructive at distant locations.

The time difference between P and S wave arrivals at a single station tells seismologists how far away the earthquake was. With readings from at least three stations, they can triangulate the earthquake's epicentre. With dozens or hundreds of stations, they can pinpoint location to within a few kilometres and calculate depth, fault geometry, and the amount of energy released.

The Global Network

No single seismograph can detect all earthquakes. What makes modern earthquake monitoring possible is the global network — thousands of stations spread across every continent, from Iceland to Antarctica, from the bottom of the Pacific Ocean to the summit of Hawaiian volcanoes.

The USGS operates the Global Seismographic Network (GSN), a backbone of around 150 high-quality stations worldwide, supplemented by thousands of regional and national networks. Every station streams data continuously to processing centres, where algorithms scan incoming waveforms around the clock for earthquake signatures. When a potential event is detected at multiple stations simultaneously, an automated system triggers an alert and begins computing a location and magnitude — often within 5–10 minutes of the earthquake occurring.

From Sensor to Tremr

Once USGS algorithms have processed a detection and computed a preliminary magnitude and location, the event is published to the USGS public GeoJSON feed — the same feed that powers Tremr. Tremr queries that feed every five minutes and displays new events on the map and in the list, usually within 10–20 minutes of the earthquake occurring.

The magnitude you see on Tremr is a USGS estimate that may be updated as more stations report in and seismologists refine their analysis. Large earthquakes often have their magnitudes revised in the hours and days after the event as data accumulates. The "reviewed" status flag in the detail panel indicates whether a human analyst has confirmed the automated estimate.

Next time you tap an earthquake dot on Tremr, picture the chain behind it: ground vibrates → coil moves in magnet → current flows → data streams to servers → algorithm detects pattern → magnitude computed → feed updated → your screen refreshes. That whole chain, from shaking ground to your fingertip, typically takes less time than it takes to brew a coffee.