Seismic waves are pulses of energy that travel through the Earth (and other planetary bodies) whenever rock suddenly shifts, something heavy strikes the ground, or even when ocean waves pound the seafloor. They behave much like sound waves in air, except they move through solid rock and liquid metal, and their speeds and shapes change depending on the material they pass through. That sensitivity to material is what makes seismic waves so extraordinarily useful: they are the main tool scientists have for seeing inside a planet without digging.
Body Waves and Surface Waves
Seismic waves come in two broad families. Body waves travel through Earth’s interior, while surface waves travel along or near the surface. Within each family there are important subtypes, and the differences between them matter for everything from earthquake damage to deep-Earth imaging.
Body waves split into two kinds. P-waves (primary waves) are compressional: the rock alternately squeezes and stretches in the same direction the wave is moving, like a slinky pushed from one end. They are the fastest seismic waves, which is why they arrive first on a seismogram. S-waves (secondary, or shear, waves) move rock side to side, perpendicular to their travel direction. They are slower than P-waves, and they cannot pass through liquids because liquids have no rigidity to transmit sideways motion. That single fact turned out to be one of the most consequential discoveries in geophysics.
Surface waves are slower still, but they often carry the most destructive energy during an earthquake. Rayleigh waves roll the ground in an elliptical motion, a bit like ocean swells passing through rock. Their particle motion can be measured by comparing horizontal and vertical ground motion recorded by sensor arrays.1Geophysical Journal International. Estimating Rayleigh wave particle motion from three-component array analysis of ambient vibrations Love waves, the other major surface-wave type, shake the ground horizontally with no vertical component. Under certain conditions, the two types can actually exchange energy with each other, a phenomenon called Love-Rayleigh coupling, which has been observed at periods between roughly 20 and 30 seconds depending on the direction and the structure of the ground.2Geophysical Research Letters. 3D particle motion trajectories: Direct observation of Love‐Rayleigh coupling
How Waves Lose Energy and How Basins Amplify Them
As seismic waves travel outward from their source, they weaken. Some of that loss comes from geometric spreading: the energy fans out over an ever-larger area. But two other processes matter just as much. Intrinsic absorption converts wave energy into heat as rock grains slide against one another. Scattering happens when waves bounce off small-scale irregularities in the rock, redirecting energy in many directions at once. Studies that separate these two effects find that their relative importance depends on wave frequency. In tectonically active northeast India, for example, scattering dominates at frequencies below about 3 Hz, while intrinsic absorption takes over at higher frequencies.3Geophysical Journal International. Separation of intrinsic and scattering seismic wave attenuation in Northeast India Similar patterns have been documented in Alaska.4Bulletin of the Seismological Society of America. Separation of Intrinsic and Scattering Seismic Wave Attenuation in the Crust of Central and South-Central Alaska
While attenuation weakens waves in most settings, certain geologic structures can dramatically amplify them. Sedimentary basins, the broad bowls of softer rock and soil that underlie many cities, act like natural amplifiers. When seismic waves enter soft sediments from harder bedrock below, they slow down, and the energy piles up. Recordings from the Seattle basin showed that waves in the 0.2 to 0.8 Hz range were amplified by factors of 8 to 16 compared to nearby bedrock sites during a distant earthquake, with later-arriving surface waves pushing amplification even higher.5Bulletin of the Seismological Society of America. Amplification of seismic waves by the Seattle basin, Washington state Three-dimensional simulations of the Nenana basin in central Alaska found average horizontal amplification of about four and vertical amplification of about seven, driven mainly by the basin’s 3D geometry rather than by shallow low-velocity layers alone.6Journal of Geophysical Research: Solid Earth. Analysis of Seismic Wave Amplification in Sedimentary Basins Using 3D Wavefield Simulations: Nenana Basin, Central Alaska The Po Plain in northern Italy shows a similar story for longer-period waves above one second.7Bulletin of the Seismological Society of America. New Insights into Long‐Period (>1 s) Seismic Amplification Effects in Deep Sedimentary Basins: A Case of the Po Plain Basin of Northern Italy
This basin amplification is a major reason two neighborhoods in the same city can experience very different shaking from the same earthquake. If you live on deep sediment rather than solid rock, the waves you feel may be several times stronger than what your neighbor on bedrock experiences.
Mapping Earth’s Hidden Layers
Nearly everything we know about the deep interior of our planet comes from tracking how seismic waves change speed, reflect, or disappear as they dive into the Earth. The logic is straightforward: when a wave hits a boundary between two materials with different densities or stiffnesses, some energy reflects back and some bends, just like light passing from air into water. By recording thousands of earthquakes at hundreds of stations worldwide, seismologists have built up an increasingly detailed picture of what lies below.
The most dramatic early discovery was Earth’s liquid outer core. Because S-waves cannot travel through liquid, there is a wide “shadow zone” on the far side of the planet from any earthquake where direct S-waves never arrive. That shadow mapped out the outer core’s boundary. The existence of a solid inner core within the liquid core was later confirmed by studying normal modes, the whole-Earth vibrations that ring like a bell after very large earthquakes.8Seismological Research Letters. Revisiting Seismological Discoveries of the Inner Core
Work continues at the very base of the mantle, where thin patches called ultralow velocity zones sit right at the core-mantle boundary. These zones slow P- and S-waves dramatically and appear to be enriched in iron, possibly as a result of material carried down by subducting tectonic plates over billions of years.9Geophysical Journal International. Ultralow velocity zones at the core–mantle boundary in the subduction regions consistent with the iron-rich origins revealed by ScP waveform complexities
Another window into deep structure is shear-wave splitting. When an S-wave enters rock whose crystals are aligned in a preferred direction, the wave splits into a fast component and a slow component, much like light splitting in a polarizing crystal. Measurements at the East Pacific Rise found that the fast direction lines up with the spreading direction, and the amount of splitting differs between the two sides of the ridge, suggesting asymmetric flow patterns in the mantle.10PubMed. Shear-wave splitting and implications for mantle flow beneath the MELT region of the east pacific rise Continent-scale studies across Antarctica show broadly northeast-southwest fast directions that rotate clockwise from west to east, reflecting a mix of frozen-in structure within the rigid plate and flowing mantle beneath it.11Journal of Geophysical Research: Solid Earth. Shear Wave Splitting Across Antarctica: Implications for Upper Mantle Seismic Anisotropy
Earthquake Early Warning
Because P-waves travel faster than the more destructive S-waves and surface waves, there is a brief window between the first tremor and the strong shaking. Earthquake early warning systems exploit that gap. Sensors near the source detect the P-wave, quickly estimate the earthquake’s size, and send out alerts before the damaging waves arrive. Even a few seconds of warning is enough for trains to brake, surgical robots to pause, and people to take cover.
A key challenge is estimating magnitude fast enough to be useful. One well-tested approach measures the peak displacement within the first three seconds of the P-wave arrival. In southern California, that measurement agrees with catalog magnitudes to within about 0.2 units for earthquakes below magnitude 6.5.12Geophysical Research Letters. Magnitude estimation using the first three seconds P‐wave amplitude in earthquake early warning Newer methods that combine multiple ground-motion measurements from P-waves have improved reliability further, with one approach tested on Japanese strong-motion data showing roughly a 35 percent increase in successful alarms compared to single-parameter methods.13Geophysical Research Letters. A P wave‐based, on‐site method for earthquake early warning Other techniques use signal-to-noise ratios during P-wave detection to estimate magnitude and distance simultaneously, aiming for both speed and accuracy.14PubMed Central. A Synchronous Magnitude Estimation with P-Wave Phases’ Detection Used in Earthquake Early Warning System
Imaging Without Earthquakes
You do not need an earthquake to generate useful seismic waves. The ground is never truly still: ocean waves, wind, traffic, and industrial machinery all produce a constant hum of low-level vibrations called ambient noise. Over the past two decades, seismologists learned to extract meaningful signals from this noise by cross-correlating recordings at two different stations. Done over long enough time periods, the correlation approximates the signal that would be recorded if one station acted as a source and the other as a receiver.15Journal of Geophysical Research: Solid Earth. Empirical Green’s Function Retrieval Using Ambient Noise Source‐Receiver Interferometry This technique, called seismic interferometry, has opened up imaging in regions with few earthquakes, including stable continental interiors and ocean basins.16Geophysical Journal International. The relationship between cross correlations and Green’s functions in ambient noise interferometry with Bayesian constraints
For industry applications like oil and gas exploration or carbon-storage monitoring, active-source surveys fire controlled energy into the ground and record the reflections. A technique called full-waveform inversion tries to match every wiggle of the recorded signal with a computer-simulated one, iterating until the velocity model fits. Researchers have pushed this into three dimensions, aiming to recover detailed images of crustal rock beneath sediment.17Geophysical Journal International. Next-generation seismic experiments: wide-angle, multi-azimuth, three-dimensional, full-waveform inversion One persistent obstacle is “cycle skipping,” where the computer locks onto the wrong wiggle and produces a misleading model. Newer approaches use envelope-based measurements to avoid that trap.18Journal of Seismic Exploration. Direct envelope-based reflection full waveform inversion for shallow marine seismic data
Seismic Waves on the Moon and Mars
Seismology is not confined to Earth. The Apollo missions left seismometers on the Moon, and the recordings were unlike anything geophysicists had seen. Impacts produced signals that built slowly to a peak and then decayed over extraordinarily long periods, sometimes ringing for an hour or more.19PubMed. Seismic data from man-made impacts on the moon That prolonged reverberation points to intense scattering in a dry, fractured, airless crust, where there is essentially no mechanism to damp out the energy quickly.20Space Science Reviews. Lunar Seismology: A Data and Instrumentation Review
NASA’s InSight lander brought seismology to Mars starting in 2018. It detected hundreds of marsquakes and even picked up seismic waves generated by meteorite impacts. Those surface waves, measured for the first time on another planet, revealed that the Martian crust north of the equatorial dichotomy has a shear-wave velocity of about 3.2 kilometers per second and is structurally different from the crust beneath the lander.21PubMed. Surface waves and crustal structure on Mars Receiver-function analysis of marsquakes constrained the crustal thickness beneath InSight to either about 20 or about 39 kilometers, depending on which subsurface interface marks the true base of the crust.22PubMed. Thickness and structure of the martian crust from InSight seismic data InSight also detected reflections from Mars’s core-mantle boundary, pinning the radius of the liquid metal core at about 1,830 kilometers.23PubMed. Seismic detection of the martian core In a few years of data, seismology revealed more about the deep structure of Mars than decades of orbital measurements had.
Waves From Oceans, Glaciers, and Elephants
Earthquakes and explosions are not the only things that set the ground vibrating. The constant pounding of ocean waves against the seafloor generates a background hum detected by seismometers worldwide, called microseisms. Two mechanisms are at work: a primary one, where waves interact with sloping sea floor, explains the longer-period signals (roughly 13 to 300 seconds), while a secondary mechanism, involving nonlinear interaction between ocean waves traveling in opposite directions, dominates at periods shorter than about 13 seconds.24Geophysical Research Letters. How ocean waves rock the Earth: Two mechanisms explain microseisms with periods 3 to 300 s These microseisms are the “noise” that ambient-noise interferometry exploits for imaging.
Glaciers generate their own seismic signals. Small icequakes occur when a glacier lurches forward in stick-slip motion over rough bedrock. Recordings from Glacier d’Argentière in the French Alps showed that the signals were best explained by horizontal shear faulting with slip parallel to the glacier’s flow direction, consistent with the base of the ice catching and releasing against bumps in the rock.25Journal of Geophysical Research: Earth Surface. Basal icequakes recorded beneath an Alpine glacier (Glacier d’Argentière, Mont Blanc, France): Evidence for stick‐slip motion? Monitoring these icequakes can track changes in glacier dynamics as climate warms.
Even the biological world produces detectable seismic waves. Experiments with trained African elephants showed that their low-frequency vocalizations produce Rayleigh waves that could propagate through the ground up to about 2 kilometers under ideal conditions, potentially farther than the airborne sound itself.26Geophysical Research Letters. Seismic waves from elephant vocalizations: A possible communication mode? Elephants are known to respond to distant ground-borne vibrations, raising the possibility that seismic communication supplements their acoustic calls.
Human-Made Seismic Noise and Its Sudden Disappearance
Everyday human activity, traffic, factory machinery, construction, creates a constant buzz of high-frequency seismic waves that seismometers pick up as background noise.27PubMed. Global quieting of high-frequency seismic noise due to COVID-19 pandemic lockdown measures This anthropogenic noise usually sits at frequencies above about 1 Hz and is strongest in urban areas. During the early months of the COVID-19 pandemic in 2020, lockdown measures around the world led to a dramatic drop in this seismic background. Seismologists called it the “anthropause,” and it gave them an unprecedented view of subtle natural signals that are normally buried under human noise. The quieting was documented at stations on every continent and demonstrated just how pervasive our seismic footprint is.
Fault ruptures interact with this noisy crust in interesting ways. Simulations of earthquake ruptures show that when a fault tears through damaged, fractured rock rather than pristine material, the broken rock radiates more high-frequency energy above 1 Hz, while the rupture itself slows down.28Journal of Geophysical Research: Solid Earth. Delayed Dynamic Triggering and Enhanced High‐Frequency Seismic Radiation From Brittle Rock Damage in 3D Dynamic Rupture Simulations In an urban setting where buildings and infrastructure are tuned to resonate at certain frequencies, this kind of detail matters for predicting the actual shaking pattern.
When Waves Reach the Sky
Seismic waves do not stop at Earth’s surface. Large earthquakes push the ground upward, and that motion launches pressure waves into the atmosphere. These acoustic-gravity waves rise through thinning air, growing in amplitude as density drops, and can reach the ionosphere roughly 300 kilometers up. There, they disturb the electrically charged plasma enough to cause measurable ripples in GPS signals. The 2012 Sumatra double earthquake produced wave-like perturbations in the ionosphere’s total electron content that were clearly detected by ground-based GPS networks.29Journal of Geophysical Research: Space Physics. GNSS TEC‐Based Detection and Analysis of Acoustic‐Gravity Waves From the 2012 Sumatra Double Earthquake Sequence
Volcanic eruptions can do the same thing on a spectacular scale. The January 2022 eruption of Hunga Tonga–Hunga Ha’apai in the South Pacific sent atmospheric acoustic-gravity waves around the globe multiple times. Modeling of those waves, using techniques borrowed from seismological normal-mode theory, estimated the eruption’s equivalent energy at roughly 9 exajoules.30Journal of Geophysical Research: Solid Earth. Modeling Atmospheric Acoustic‐Gravity Waves From Violent Volcano Eruptions With Seismological Approaches The event underscored how tightly the solid Earth, the ocean, and the atmosphere are coupled through wave physics.
Engineered Defenses Against Seismic Waves
An emerging branch of earthquake engineering asks a provocative question: rather than just building stronger structures, can we redirect or absorb seismic waves before they reach a building? The answer, at least in simulations, is increasingly yes. Seismic metamaterials are engineered structures, often arrays of resonators buried in or placed on the ground, that exploit wave-interference effects to create frequency bands in which waves cannot propagate. Simulations of novel designs using hierarchical unit cells and bioinspired resonators have achieved up to a 40 percent reduction in seismic wave amplitude across low and broad frequency ranges relevant to buildings and bridges.31Materials Today Communications. Advancing infrastructure resilience through seismic metamaterials: Current insights and future prospects
Scaled-up versions of these ideas, sometimes called meta-barriers, have been modeled as walls of specially tuned resonant units placed in the ground between a fault and a city. One study analyzed a wide-bandgap meta-barrier capable of blocking about 85 percent of surface waves in the 4.5 to 30 Hz range, which covers much of the frequency band that causes the worst structural damage.32Structures. Large-scale engineered meta-barriers for attenuation of seismic surface waves The technology is still largely theoretical and lab-scale, but the physics works, and the gap between simulation and real-world deployment is narrowing as materials science catches up.
A Brief History of Listening to the Ground
Humans have been trying to detect earthquakes instrumentally for nearly two millennia. The earliest known seismoscope was built in 132 AD by the Chinese scholar Zhang Heng. It was a bronze vessel with dragon heads that would drop a ball in the direction of the shaking, giving a crude indication of where a distant earthquake occurred. But nearly 1,750 years passed before anyone built a device that could actually record ground motion over time. In 1875, Filippo Cecchi built a low-sensitivity seismograph in Italy. British scientists working in Tokyo in the 1880s independently developed instruments, and in 1889, Ernst von Rebeur-Paschwitz recorded the first teleseism, a distant earthquake detected on the other side of the world. Viscous damping, which keeps the recording needle from bouncing uncontrollably, was introduced by Emil Wiechert in 1898.33Bulletin of the Seismological Society of America. The early history of seismometry (to 1900) Modern broadband seismometers can detect ground displacements smaller than the width of an atom, a sensitivity that would have been incomprehensible to those early inventors but that follows directly from the engineering lineage they started.