Faults occur primarily along the edges of tectonic plates, where enormous slabs of Earth’s crust collide, pull apart, or grind past each other. Roughly 90% of the world’s earthquakes happen at these plate boundaries. But faults also form in the middle of continents, far from any plate edge, and some hide deep underground without ever breaking the surface. Understanding where faults develop starts with understanding the forces that create them.
Plate Boundaries: Where Most Faults Form
Earth’s outer shell is broken into about 15 major tectonic plates that constantly move relative to one another. The boundaries between these plates are where stress concentrates, rock fractures, and faults develop. Three types of boundaries produce three distinct styles of faulting.
At divergent boundaries, plates pull away from each other. The stretching thins the crust until it cracks like peanut brittle, creating normal faults where one block of rock drops down relative to the other. On land, this process carves out long rift valleys flanked by mountain ranges. East Africa’s Great Rift Valley and the Basin and Range Province across the western United States are classic examples of this landscape. Beneath the ocean, the same process builds the mid-ocean ridge system, a chain of underwater mountains that circles the globe for over 40,000 miles.
At convergent boundaries, plates push into each other. The compression forces rock to buckle and slide along reverse faults (also called thrust faults), where one block is shoved up and over the other. Subduction zones, where an oceanic plate dives beneath a continental plate, produce some of the largest and most dangerous faults on Earth. The Cascadia Subduction Zone off the Pacific Northwest coast and the megathrust fault system beneath Japan are both convergent boundary faults capable of generating massive earthquakes.
At transform boundaries, plates slide horizontally past each other. The grinding produces strike-slip faults, where rock on either side moves laterally rather than up or down. The San Andreas Fault in California is the most famous example. It marks the boundary where a sliver of western California, riding on the Pacific Plate, slides north-northwestward past the rest of North America. The San Andreas is actually just one of several parallel faults that accommodate this motion across a broad zone of crustal deformation. The Queen Charlotte Fault off the coast of British Columbia is another transform boundary between the same two plates.
Faults in the Middle of Continents
Not all faults sit neatly along plate edges. Some of the most surprising seismic zones are located hundreds of miles from the nearest boundary. The New Madrid Seismic Zone in the central United States, centered beneath Missouri, Arkansas, and Tennessee, is a prime example. In 1811 and 1812, this zone produced a series of earthquakes so powerful they temporarily reversed the flow of the Mississippi River.
These intraplate faults typically form in zones of ancient weakness buried deep in the crust. The New Madrid zone sits on top of the Reelfoot Rift, a fracture system created roughly 500 million years ago when a supercontinent broke apart. That ancient rifting thinned and weakened the underlying rock. Later geological events, including compression from a mountain-building episode and the passage of a volcanic hotspot, further altered the crust by pushing molten rock into the rift. The result is a patch of weakened lithosphere that still accumulates and releases stress today, even though the nearest plate boundary is thousands of miles away.
Other notable intraplate fault zones include the Wabash Valley Seismic Zone along the Illinois-Indiana border and faults beneath Charleston, South Carolina, which produced a devastating earthquake in 1886.
How Deep Faults Can Form
Faults only develop where rock is cold and brittle enough to snap rather than slowly bend. This brittle zone extends from the surface down to a depth that varies by location but generally ranges from about 10 to 25 kilometers (6 to 15 miles) in continental crust. Below that depth, temperatures and pressures are high enough that rock behaves more like putty, deforming slowly without fracturing. Geologists call this transition the brittle-ductile boundary.
This boundary matters because it sets the maximum depth an earthquake rupture can reach. Intermediate and larger earthquakes typically rupture from the surface all the way down to this transition, and the deeper that boundary sits, the wider the fault area that can break in a single event, and the more energy it can release. In subduction zones, where cold oceanic crust plunges beneath another plate, the brittle zone can extend much deeper, which is why subduction zone earthquakes can be extraordinarily powerful.
Faults That Never Reach the Surface
Some faults are completely hidden underground. A blind thrust fault is a shallow-angle reverse fault that terminates before it reaches the Earth’s surface. When it ruptures, it can produce significant uplift and shaking, but it leaves no visible fault line on the ground. This makes these faults especially dangerous because they can be difficult to detect before they cause damage.
Southern California has several known blind thrust faults. The Elysian Park Thrust runs directly beneath downtown Los Angeles. The Northridge Thrust Fault, which ruptured in the 1994 Northridge earthquake, was another blind fault. That magnitude 6.7 quake killed 57 people and caused roughly $20 billion in damage, all from a fault that had not been identified as a major hazard beforehand. Geologists suspect that many more blind thrust faults remain undiscovered beneath populated areas.
How Scientists Find Hidden Faults
Detecting faults that don’t break the surface requires specialized technology. Airborne LiDAR (a laser-based scanning system flown from aircraft) creates ultra-high-resolution topographic maps that can reveal subtle ground deformation invisible to the naked eye, with resolution as fine as 0.25 meters. Scientists look for faint lineaments, slight offsets in terrain, or unusual drainage patterns that hint at buried faulting.
When LiDAR identifies a suspicious feature, seismic reflection imaging can confirm whether a fault exists below the surface. This technique bounces sound waves off underground rock layers to create a cross-sectional image of the subsurface. Combining these two methods, researchers in California’s northern Walker Lane identified a previously concealed strike-slip fault system that extended continuously to the surface in one area but was completely buried in another. Satellite-based radar interferometry adds another tool, measuring tiny changes in ground elevation over time that can indicate slow fault movement.
Why Location Matters for Risk
The type of plate boundary determines the style of faulting, but the specific location of a fault relative to where people live is what determines risk. Transform faults like the San Andreas tend to produce shallow earthquakes with intense shaking along a narrow corridor. Subduction zone faults can rupture over enormous areas and trigger tsunamis. Intraplate faults, while less frequent in their activity, can catch populations off guard because they occur in regions not typically associated with earthquakes.
Faults also cluster in ways that compound risk. The transform boundary in California, for example, is not a single clean line but a broad zone of shearing with dozens of parallel and branching faults. Each one represents a surface where rock can slip. The San Andreas alone runs roughly 800 miles through the state, but it’s the network of smaller, less famous faults surrounding it that produces the majority of California’s day-to-day seismic activity.