How Earthquakes Form, Are Measured, and What Plate Tectonics Explains

Earthquakes are a routine part of how Earth's outer shell behaves, not a random or unpredictable anomaly. According to the United States Geological Survey (USGS), the planet's crust is broken into large pieces, called tectonic plates, that slowly move, slide past one another, and collide. Most earthquakes occur along faults — fracture lines in the rock — located at or near the boundaries between these plates.

What actually happens underground

The USGS describes the process as a three-step mechanism known as elastic rebound. First, as tectonic plates move, the rough, jagged edges of a fault get stuck against each other even though the rest of the plate keeps moving, so stress and energy build up in the surrounding rock. Second, once the accumulated force finally exceeds the friction holding the fault's edges together, the rock suddenly slips. Third, the stored energy is released outward as seismic waves, which shake the ground — the earthquake people feel.

Earthquake activity often comes in sequences. The USGS notes that smaller tremors called foreshocks can precede a larger mainshock, though they are only identified as foreshocks in retrospect, and that aftershocks — smaller quakes following the mainshock — can continue for weeks, months or even years. Despite this understanding of mechanics, the USGS states plainly that scientists cannot predict when an earthquake will occur, even in areas where they know earthquakes are likely.

Three kinds of plate boundaries, three kinds of quakes

Tectonic plates interact in three distinct ways, and each produces a different earthquake pattern, according to NOAA Ocean Exploration and the USGS. At divergent boundaries, plates pull apart and magma rises to form new oceanic crust; earthquakes are common there, with the Mid-Atlantic Ridge as an example. At convergent boundaries, plates collide — one often sliding beneath the other — producing powerful earthquakes along with mountain ranges, deep-sea trenches and volcanic chains, a pattern visible around the Pacific Ring of Fire. At transform boundaries, plates grind past each other sideways without creating or destroying crust, generating earthquakes through friction; California's San Andreas Fault is the standard example.

How seismologists record and size an earthquake

Earthquakes are detected through a global seismographic network, in which each station measures ground movement at its own site, the USGS explains. The core instrument is a seismometer — typically a pendulum or a mass mounted on a spring — housed inside a seismograph that records the vibrations as the rock slips and energy radiates outward in waves from the epicenter.

Seismologists distinguish between two separate measurements. Magnitude describes the size of the earthquake at its source and is the same number no matter where it is measured from. Intensity, measured on the Modified Mercalli Intensity Scale, describes the shaking and damage felt at a specific location, which changes depending on distance from the source and local ground conditions.

Richter scale versus moment magnitude

The magnitude scale most people have heard of — the Richter scale — was developed by Charles Richter in the 1930s to measure earthquakes in Southern California using nearby seismograph stations; it was technically known as ML, for 'local magnitude,' per the USGS. It measures the largest wiggle, or amplitude, recorded on a specific instrument called a Wood-Anderson seismograph.

As seismograph networks expanded worldwide, researchers found Richter's method was only valid for certain frequency and distance ranges, and it could not properly register the lowest-frequency waves produced by the largest earthquakes — a limitation known as saturation. Seismologists later introduced the moment magnitude scale (Mw), which the USGS says is based on the physical quantity of 'moment' — the amount of slip on the fault multiplied by the area of the fault surface that slipped — making it valid across the full range of earthquake sizes. The USGS now reports earthquake magnitudes using moment magnitude.

How common are earthquakes?

Earthquakes are far more frequent than most people realize. The USGS's National Earthquake Information Center locates roughly 12,000 to 14,000 earthquakes every year. IRIS (the Incorporated Research Institutions for Seismology) notes that the number of detected earthquakes has risen over time mainly because seismometer networks have become more sensitive, not because earthquakes themselves have become more frequent.

Because the moment magnitude scale is logarithmic, IRIS states that for each whole step up in magnitude, the annual number of earthquakes recorded worldwide decreases by roughly a factor of ten, while the energy released increases by roughly a factor of thirty.

Why this matters

Understanding that earthquakes are the predictable byproduct of ongoing plate motion — rather than isolated, unexplainable events — is the basis for hazard maps, building codes and early-warning systems used around the world. The plate-boundary framework from the USGS and NOAA also explains why certain regions, such as the Pacific Ring of Fire or the San Andreas Fault zone, experience earthquakes repeatedly over geological time, even though no single event can be predicted in advance.