What Happens Underground During an Earthquake?

When you feel the ground shake during an earthquake, you're experiencing the surface expression of a dramatic underground event that's been building for decades, centuries, or even millennia. Deep beneath your feet—sometimes just a few kilometers down, sometimes tens of kilometers—solid rock is suddenly fracturing and sliding in an explosive release of accumulated strain. In seconds, rock surfaces that have been locked together under immense pressure slip meters past each other, sending shock waves radiating through the Earth at thousands of meters per second.

But what exactly is happening down there? How can solid rock, seemingly as permanent and unchanging as anything on Earth, suddenly break apart? Why does stress build up for so long before being released so suddenly? How does energy travel from the source to shake buildings hundreds of kilometers away? And what determines whether an earthquake will be a barely noticeable tremor or a catastrophic rupture?

Understanding what happens underground during an earthquake requires piecing together evidence from seismic waves, geological observations of ancient faults exposed at the surface, laboratory experiments on rock fracture, and increasingly sophisticated computer models. What emerges is a picture of extraordinary violence happening in darkness and silence beneath our feet—a process both simple in concept (rocks breaking) and dizzyingly complex in detail.

This article explores the underground mechanics of earthquakes: how stress accumulates, why rocks suddenly fail, how ruptures propagate along faults, how energy radiates away as seismic waves, and what physical changes occur in the rocks themselves during these brief but violent events.

The Earthquake Cycle: From Locked to Broken

Understanding what happens during an earthquake requires first understanding what happens between earthquakes.

The Interseismic Period: Building Stress

The setup:

• Tectonic plates move steadily at 1-10 cm per year
• At plate boundaries, faults separate the plates
• Friction prevents smooth sliding—faults lock together
• But the plates keep trying to move
• Stress accumulates in rocks surrounding the locked fault

Elastic strain accumulation:

• Rocks behave elastically—they deform but don't break
• Like stretching a rubber band
• Strain energy builds year after year, decade after decade
• Rocks far from fault move with plates
• Rocks near fault deform but can't slip
• Creates increasing differential stress

How long does stress build?

• Depends on plate velocity and fault strength
• San Andreas Fault: 100-300 years between major earthquakes on given segment
• Cascadia Subduction Zone: 300-600 years
• Some faults: thousands of years
• During this time, no earthquakes—just quiet stress accumulation

Observable effects at surface:

• GPS measurements show slow ground deformation
• Strain meters detect changing stress
• Tilt meters measure ground tilting
• But these changes are tiny—millimeters per year

The Critical Point: Why Rocks Finally Break

The physics of rock failure:

• All materials have a breaking strength (failure stress)
• For rock under pressure, this is typically 50-200 megapascals
• When applied stress exceeds strength, rock fractures
• But the actual process is more complex

Why fracture happens suddenly:

• Rock doesn't gradually weaken as stress increases
• It remains strong until critical stress is reached
• Then failure is nearly instantaneous
• This is called stick-slip behavior
• "Stick" phase: locked, stress building
• "Slip" phase: sudden failure and sliding

The role of friction:

• Static friction: resists initial motion (strong)
• Dynamic friction: resists continuing motion (weaker)
• Once slip starts, friction drops
• This makes slip accelerate—the earthquake begins

Heterogeneous stress distribution:

• Stress isn't uniform along fault
• Some patches more stressed than others
• Highest-stress patch fails first
• This becomes the hypocenter (earthquake's starting point)
• Rupture then propagates from there

The Rupture: Seconds of Violence

When a fault finally fails, what happens underground is both simple (rocks sliding) and extraordinarily complex.

Rupture Initiation: The First Moments

The hypocenter (focus):

• Point where rupture begins
• Usually not at ground surface
• Typically 5-25 km deep for crustal earthquakes
• Can be deeper in subduction zones (up to 700 km)

Initial failure:

• Small patch of fault (perhaps 100m x 100m) exceeds failure stress
• Rock breaks and begins to slip
• This releases stress on that patch
• But increases stress on adjacent areas
• Like breaking one link in a chain under tension

The cascade begins:

• Stress transfer to adjacent fault patches
• These patches now closer to failure
• They fail in rapid succession
• Creates propagating rupture front
• Like a crack spreading through glass

Rupture Propagation: The Earthquake Grows

Rupture velocity:

• Rupture propagates at 2-3 km per second typically
• About 70-90% of the shear wave velocity in rock
• Can vary considerably along single fault
• Slower in some sections, faster in others
• Can approach or even exceed shear wave speed (supershear rupture)

Bilateral vs. unilateral rupture:

• Bilateral: Rupture propagates in both directions from hypocenter
• Unilateral: Rupture propagates primarily in one direction
• Direction affects ground shaking intensity
• Areas in rupture direction experience stronger shaking (directivity effect)

Controlling factors:

• Stress distribution: Rupture propagates into high-stress regions
• Fault geometry: Bends, kinks, or steps can stop rupture
• Material properties: Weak zones facilitate rupture; strong zones resist
• Previous earthquakes: Recently-ruptured sections have low stress

Fault Slip: The Main Event

Magnitude of slip:

• Small earthquakes (M4-5): Centimeters of slip
• Moderate earthquakes (M6-7): Tens of centimeters to 1-2 meters
• Large earthquakes (M7-8): 2-10 meters
• Great earthquakes (M8-9): 10-30 meters
• Megaquakes (M9+): 30-50+ meters (1960 Chile: up to 50 meters)

Slip rate during rupture:

• Meters per second of fault motion
• Extremely rapid compared to tectonic plate motion (cm/year)
• Explains the violent shaking
• Total slip happens in 1-2 seconds at any given point

Slip distribution along fault:

• Not uniform—some patches slip more than others
• Areas of maximum slip called "asperities"
• These release most energy
• Create strongest ground shaking
• Some patches barely slip at all

What the Rock Experiences

Stress drop:

• Difference between stress before and after earthquake
• Typically 1-10 megapascals (10-100 atmospheres)
• Represents energy available for seismic radiation
• Complete stress drop rare—residual stress usually remains

Frictional heating:

• Rapid sliding generates intense friction
• Converts mechanical energy to heat
• Temperature can spike to 1,000°C+ at fault surface
• But heat dissipates quickly into surrounding rock
• Evidence: pseudotachylite (rock melted by friction) found in ancient faults

Pressure changes:

• Fault slip changes pore pressure in rock
• Can squeeze fluids or create suction
• May trigger fluid migration
• Affects aftershock distribution

Pulverization and damage:

• Rock near fault is fractured and pulverized
• Creates fault gouge (crushed rock powder)
• Damage extends tens to hundreds of meters from fault
• Reduces strength for future earthquakes

Rupture Termination: Why Earthquakes Stop

Why doesn't rupture continue indefinitely?

• Low stress regions: Rupture reaches area already relaxed by previous earthquakes
• Fault geometry: Major bends or discontinuities stop propagation
• Strong barriers: Unusually strong rock sections resist failure
• Fault ends: Rupture reaches physical end of fault
• Energy depletion: Available strain energy exhausted

Segmented faults:

• Many faults divided into segments
• Segments separated by structural complexities
• Single earthquake usually ruptures one segment
• Rarely, rupture jumps between segments (creating larger earthquake)
• 2002 Denali earthquake ruptured three fault segments

Seismic Waves: Carrying Energy Away

The rupture generates seismic waves that radiate through Earth—this is what we feel as shaking.

Types of Seismic Waves

Body waves (travel through Earth's interior):

• P-waves (Primary/Compressional waves):
  • Fastest seismic waves (5-7 km/s in crust)
  • Particle motion parallel to wave direction (push-pull)
  • Can travel through solids, liquids, and gases
  • First to arrive at seismometer
  • Cause initial jolt of earthquake
• S-waves (Secondary/Shear waves):
  • Slower than P-waves (3-4 km/s in crust)
  • Particle motion perpendicular to wave direction
  • Cannot travel through liquids
  • Arrive second at seismometer
  • Cause most violent shaking
  • Most damaging to structures

Surface waves (travel along Earth's surface):

• Love waves:
  • Horizontal side-to-side motion
  • Faster of the two surface wave types
  • Can be very destructive
• Rayleigh waves:
  • Circular/rolling motion (like ocean waves)
  • Slowest but often largest amplitude
  • Can cause buildings to rock back and forth

Wave Generation at the Source

The radiation pattern:

• Seismic waves don't radiate uniformly in all directions
• Pattern depends on fault orientation and slip direction
• Some directions get stronger waves than others
• Four lobes of compression and four of extension
• Called the "focal mechanism"

Frequency content:

• Earthquakes generate wide range of frequencies
• Small earthquakes: mostly high frequency (>1 Hz)
• Large earthquakes: significant low frequency content (<1 Hz)
• Frequency affects what structures are most vulnerable

Wave Propagation Through Earth

Wave speed variations:

• Seismic waves travel faster through dense, rigid rock
• Slower through soft, loose sediments
• Speed increases with depth (rock gets denser)
• This causes wave refraction (bending)

Attenuation (energy loss):

• Geometric spreading: Energy spreads over larger area with distance
• Intrinsic attenuation: Energy converted to heat in rock
• Scattering: Energy redirected by heterogeneities
• Result: Ground shaking decreases with distance from epicenter

Amplification in soft soils:

• Seismic waves slow down in soft sediments
• Energy must go somewhere—increases amplitude
• Can amplify ground motion 2-10 times
• Explains why some areas shake much harder than nearby bedrock
• Critical factor in earthquake damage distribution

The Energy Budget: Where Does the Energy Go?

When a fault ruptures, the stored elastic strain energy is converted into several forms.

Energy Distribution

Typical breakdown for a large earthquake:

• ~95% converted to heat through friction
• ~5% radiated as seismic waves
• ~1% used to create new fracture surfaces

Why this matters:

• Only small fraction becomes shaking we feel
• But that 5% contains enormous energy
• M9 earthquake: seismic energy equivalent to ~25,000 atomic bombs
• Total energy (including friction): 50,000+ atomic bombs

Magnitude and Energy

The magnitude scale:

• Each magnitude unit represents ~32x more energy
• M6 releases ~32x more energy than M5
• M7 releases ~1,000x more energy than M5
• M9 releases ~1,000,000x more energy than M5

What controls magnitude?

• Rupture area: Larger area = larger magnitude
• Slip amount: More slip = more energy released
• Stress drop: Higher stress drop = more energy per unit area
• Formula: Magnitude proportional to log(Area × Slip)

Different Types of Faulting Underground

The style of underground rupture depends on the tectonic setting and stress orientation.

Strike-Slip Faults

Geometry:

• Vertical or near-vertical fault plane
• Horizontal sliding motion
• One side moves horizontally past the other

Underground characteristics:

• Rupture width limited to seismogenic zone thickness (~15-20 km)
• Can rupture hundreds of kilometers along strike
• Creates narrow zone of intense deformation
• Examples: San Andreas, North Anatolian Fault

Thrust (Reverse) Faults

Geometry:

• Shallow-dipping fault plane (10-30° typical)
• Hangingwall moves up and over footwall
• Compressional stress environment

Underground characteristics:

• Can rupture enormous areas (100,000+ km² for M9)
• Shallow dip means large vertical displacement
• Creates seafloor uplift (tsunamis if underwater)
• Capable of largest earthquakes (M9+)
• Examples: Cascadia, Japan, Chile subduction megathrusts

Normal Faults

Geometry:

• Steeply dipping fault plane (60° typical)
• Hangingwall moves down relative to footwall
• Extensional stress environment

Underground characteristics:

• Generally smaller than thrust or strike-slip
• Common in rift zones and mid-ocean ridges
• Typically M7 or smaller
• Examples: Basin and Range Province, East African Rift

Deep Earthquakes: A Special Case

Earthquakes deeper than 70 km present a paradox—they shouldn't exist.

The Deep Earthquake Paradox

The problem:

• At depth, temperature and pressure are extreme
• Rock should be too hot and plastic to fracture
• Should deform smoothly instead of breaking
• Yet earthquakes occur to 700 km depth

Where deep earthquakes occur:

• Only in subduction zones
• Within descending oceanic slab
• Trace slab as it sinks into mantle
• Create Wadati-Benioff zones

Mechanisms of Deep Earthquakes

Dehydration embrittlement (70-300 km depth):

• Water-bearing minerals in slab (serpentine, amphibole) break down
• Release water increases pore pressure
• Reduces effective stress, allows brittle failure
• Water release is localized, creates earthquakes

Phase transformations (300-700 km depth):

• Minerals undergo solid-state phase changes
• Olivine → spinel structure (denser packing)
• Phase change can be sudden and localized
• Volume change creates stress, triggers failure
• Called transformational faulting

Thermal runaway:

• Alternative hypothesis
• Localized shearing heats rock
• Heating weakens rock, allowing more shearing
• Positive feedback creates rapid slip
• Debate continues about relative importance

Characteristics of Deep Earthquakes

Differences from shallow earthquakes:

• Less damaging: Energy dissipates over long travel path to surface
• Felt over wider area: Wave propagation more efficient at depth
• Higher frequency: Smaller source dimensions
• No surface rupture: Far too deep to reach surface
• No aftershocks (usually): Different failure mechanism

Notable deep earthquakes:

• 1994 Bolivia M8.2 at 631 km depth—felt across South America but caused no damage
• 2013 Okhotsk M8.3 at 609 km—largest deep earthquake recorded
• Demonstrate that extreme magnitudes possible even at great depth

Aftershocks: The Aftermath Underground

After the main rupture, the underground environment remains highly disturbed.

Why Aftershocks Occur

Stress transfer:

• Mainshock doesn't relieve stress uniformly
• Some areas have increased stress after mainshock
• These areas fail in aftershocks
• Particularly at rupture ends and edges

Time-dependent processes:

• Viscoelastic relaxation: Deep warm rock slowly flows, transferring stress upward
• Pore pressure changes: Fluids redistribute, changing effective stress
• Afterslip: Continued slow slip on fault transfers stress to locked patches

Aftershock Patterns

Omori's Law:

• Aftershock rate decreases with time
• Rate proportional to 1/(time since mainshock)
• Many aftershocks immediately, declining over days to years
• Some aftershock sequences last decades

Båth's Law:

• Largest aftershock typically ~1 magnitude unit smaller than mainshock
• M7 mainshock → largest aftershock ~M6
• But significant exceptions exist

Spatial distribution:

• Cluster around mainshock rupture area
• Concentrate at rupture ends
• Define areas of increased stress
• Sometimes occur on different faults entirely

What We've Learned from Studying Earthquakes

Seismology: Reading the Underground

Seismic waves as information carriers:

• Wave arrival times locate earthquake depth and position
• Wave polarities determine fault orientation and slip direction
• Wave frequencies constrain rupture dimensions
• Wave amplitudes measure energy release

What seismologists can determine:

• Hypocenter location (within ~1 km for well-recorded events)
• Fault plane orientation
• Slip direction and amount
• Rupture velocity and direction
• Stress drop
• Energy release

Surface Evidence of Underground Processes

When faults rupture to surface:

• Ground cracks trace fault location
• Offset features (roads, fences, rivers) measure slip
• Scarp height indicates vertical displacement
• Rupture length constrains magnitude

Paleoseismology: Reading ancient earthquakes:

• Trenching across faults reveals past ruptures
• Offset soil layers record previous earthquakes
• Dating techniques determine timing
• Reveals earthquake recurrence intervals

Laboratory Rock Mechanics

What experiments reveal:

• Rock strength under different conditions
• Friction behavior during sliding
• Effects of fluids on faulting
• Temperature and pressure effects
• How rocks damage and weaken

The Bottom Line

When the ground shakes during an earthquake, the surface trembling is just the faint echo of violent processes unfolding kilometers beneath our feet. Underground, solid rock that has been locked together for decades or centuries suddenly fractures and slides in an explosive release of accumulated strain. In mere seconds, fault surfaces slip meters past each other at velocities of meters per second, releasing energy equivalent to thousands of atomic bombs.

The process begins at a single point—the hypocenter—where stress finally exceeds the rock's breaking strength. From there, rupture propagates along the fault at 2-3 kilometers per second, unzipping the locked interface like tearing fabric. As rock surfaces that have been pressed together under immense pressure suddenly slide, friction generates temperatures exceeding 1,000°C, literally melting some of the rock. The movement isn't smooth—it's violent, chaotic, and heterogeneous, with some sections slipping tens of meters while adjacent areas barely move.

The energy released radiates outward as seismic waves traveling at thousands of meters per second. P-waves arrive first, causing an initial jolt. S-waves follow, bringing the violent shaking that topples buildings. Surface waves arrive last but often carry the largest amplitudes, causing the rolling motion that can last minutes in large earthquakes. By the time these waves reach the surface, they've traveled through tens of kilometers of rock, been refracted, reflected, amplified by soft soils, and diminished by distance—yet still carry enough energy to shake cities hundreds of kilometers from the source.

What makes earthquakes particularly dangerous is the contrast between the time scales involved. Stress accumulates slowly and steadily—millimeters per year, year after year, decade after decade, sometimes for centuries. The release is catastrophically sudden—meters of slip in seconds. This is the stick-slip behavior that characterizes earthquakes: long periods of locked faults building stress, then sudden failure and violent slip.

Not all underground ruptures are equal. Strike-slip faults produce horizontal sliding that can rupture hundreds of kilometers along the fault but remains relatively narrow. Thrust faults, with their shallow dip angles, can rupture enormous areas—over 100,000 square kilometers in the largest earthquakes—and produce the massive vertical seafloor displacements that generate tsunamis. Deep earthquakes, occurring hundreds of kilometers down in subducting slabs, shouldn't exist according to simple theory yet demonstrate that brittle failure can occur even under extreme pressure and temperature through mechanisms we're still working to fully understand.

Even after the mainshock ends, the underground environment remains disturbed. Stress has been redistributed rather than eliminated. Some areas near the rupture now have increased stress and will fail in aftershocks—sometimes hundreds or thousands of them over months or years. The rock has been fractured, pulverized, and heated. Fluids have been squeezed and redistributed. The fault zone has been fundamentally altered, affecting how it will behave in future earthquakes.

We can't see these underground processes directly, but we can read their signatures in seismic waves, observe their surface expressions, simulate them in laboratory experiments, and model them in computers. What emerges is a picture of extraordinary violence happening in darkness—solid rock shattering, sliding, and releasing energy in seconds that took centuries to accumulate. Understanding these processes doesn't prevent earthquakes, but it enables us to predict where they'll occur, estimate their potential magnitude, design structures to withstand them, and develop warning systems to provide precious seconds of notice before the shaking arrives.

The next time you feel the ground shake, remember: beneath your feet, kilometers underground, rock is fracturing and sliding at velocities you could drive a car, releasing energy that has been building since long before you were born, sending shock waves through Earth at the speed of sound, all in seconds that feel like an eternity.

Additional Resources

Explore related earthquake topics: Learn about how tectonic plates create earthquakes, understand why earthquake depth matters, and discover why earthquakes cannot be predicted despite understanding these underground processes. Explore regional earthquake threats in California, the Pacific Northwest, Alaska, Chile, Turkey, New Madrid, and Mexico City. Learn about earthquake preparedness and earthquake swarms. Find safety basics in our comprehensive FAQ, and observe earthquakes in real-time on our earthquake map.