How Earthquake Waves Travel Through Different Rock Types

On September 19, 1985, a magnitude 8.1 earthquake struck the Pacific coast of Mexico near the Michoacán coast — 350 kilometers from Mexico City. At that distance, the shaking should have been moderate: attenuated by the journey, diluted across the broad wavefront, reduced to a nuisance rather than a catastrophe. In most of Mexico City, that is exactly what happened. Buildings on the rocky hills and ancient lakebed edges of the city swayed uncomfortably but survived. But in the former lakebed zone — the central districts of the city built on the thick, water-saturated clay sediments of the drained Lake Texcoco — the shaking was amplified to 50 times the intensity measured on surrounding bedrock. Hundreds of buildings collapsed. Nearly 10,000 people died.

The difference between survival and catastrophe in Mexico City in 1985 was not the earthquake's magnitude, not the distance to the fault, not the design of the buildings in isolation — it was the rock beneath the foundations. The clay sediments of the old lakebed did something that no building code could prevent: they slowed the incoming seismic waves dramatically, forced the wave energy into a smaller cross-section, and then resonated at a frequency that matched the natural period of many mid-rise buildings, shaking them with extraordinary persistence until their structural capacity was exhausted. Physics, not bad luck, killed those people. And understanding that physics — how seismic waves behave as they travel through different rock and soil types — is the foundation of modern seismic hazard assessment.

The geology beneath a city can amplify earthquake shaking by factors of 10 to 50 compared to bedrock. It can extend the duration of strong shaking from seconds to minutes. It can trap wave energy in resonant basins that ring long after the direct waves have passed. And in the most extreme cases — when shaking exceeds the strength of water-saturated loose sand — it can convert apparently solid ground into a liquid that flows out from beneath foundations in a process called liquefaction. None of these effects depend on how close you are to the fault. They depend entirely on what is between the bedrock and your building.

The Nature of Seismic Waves: A Brief Primer

Before examining how different rock types affect seismic wave propagation, it is worth establishing precisely what seismic waves are and why they are sensitive to material properties in the first place. Seismic waves are mechanical waves — propagating disturbances in the stress and strain state of rock and soil — and like all mechanical waves, their behavior is governed entirely by the elastic and inertial properties of the medium they travel through.

Body Waves: P-Waves and S-Waves

Two types of body waves travel through the interior of the Earth. P-waves (primary waves, or compressional waves) are longitudinal waves: the particle motion is in the same direction as wave propagation, alternately compressing and dilating the rock as the wave passes, the same way sound travels through air. P-waves can propagate through any material — solid, liquid, or gas — because all materials resist compression. Their velocity is determined by the bulk modulus (resistance to compression) and shear modulus of the material, divided by its density.

S-waves (secondary waves, or shear waves) are transverse waves: particle motion is perpendicular to the direction of propagation, shearing the rock sideways as the wave passes. S-waves can only propagate through materials with shear strength — solids. They cannot travel through liquids or gases, because fluids have no resistance to shear. Their velocity is determined by the shear modulus alone, divided by density. Because liquids have zero shear modulus, S-wave velocity in a fluid is zero — S-waves do not exist in water, magma, or the Earth's liquid outer core.

Surface Waves: Love and Rayleigh

In addition to body waves, seismic energy is carried by surface waves — waves that travel along the Earth's surface rather than through its interior. Love waves involve horizontal shearing motion transverse to the propagation direction, trapped in a low-velocity surface layer by total internal reflection at its base. Rayleigh waves involve a retrograde elliptical particle motion — a combination of vertical and horizontal motion in the plane of propagation — that decays exponentially with depth. Surface waves travel more slowly than body waves but carry more energy per unit distance at teleseismic ranges, because they spread cylindrically (in two dimensions) rather than spherically (in three dimensions), losing amplitude more slowly with distance.

Surface waves dominate the later arrivals on seismograms at regional to teleseismic distances, and they are responsible for the longest-period, most destructive shaking in great earthquakes felt at large distances. Their behavior is particularly sensitive to near-surface geology because they are trapped in and guided by the upper layers of the crust — the very layers that vary most dramatically between hard bedrock and soft sediment.

Seismic Velocity in Different Rock Types

The velocity at which seismic waves propagate through a rock or soil is the single most important material property for understanding site effects. Velocity is controlled by three factors: the stiffness of the material (its elastic moduli), its density, and — for P-waves — the degree of fluid saturation of its pore spaces. Different rock types span nearly two orders of magnitude in seismic velocity, from loose, water-saturated sediment at 100–200 m/s to fresh crystalline basement rock at 6,000–7,000 m/s.

The pattern in this table carries profound implications. The S-wave velocity drops from 3,000+ m/s in fresh granite to as low as 50–150 m/s in soft saturated clay — a factor of 20 to 60. This velocity contrast is the fundamental driver of seismic amplification, through a mechanism called impedance contrast.

Seismic Impedance and the Conservation of Energy

Seismic impedance is the product of a material's density and its seismic wave velocity: Z = ρ × V. It is the acoustic analog of electrical impedance — the property that determines how much of an incoming wave is transmitted across a boundary between two materials and how much is reflected back. When a seismic wave crosses from a high-impedance material (fast, dense rock) into a low-impedance material (slow, soft sediment), something physically remarkable happens: the wave must conserve energy flux (energy per unit area per unit time) across the boundary, and it does so by changing amplitude.

Energy flux in a seismic wave is proportional to Z × A², where A is the wave amplitude. If the wave moves into a material with lower impedance Z — which happens when it moves from bedrock into soft sediment — the amplitude A must increase to keep the energy flux constant. The amplitude increase is proportional to the square root of the impedance ratio. For a wave moving from granite (Z ≈ 3,000 × 7,000 = 2.1 × 10⁷ kg/m²·s) into soft clay (Z ≈ 1,600 × 100 = 1.6 × 10⁵ kg/m²·s), the impedance ratio is about 130, and the amplitude increases by roughly a factor of 11. Ground motion amplitudes at the surface can be further enhanced by the free surface effect (a factor of 2 for vertically incident waves) and by resonance, bringing total amplification in extreme cases to 50× or more.

The Physical Intuition: Why Slow Means Tall

There is an intuitive way to understand why waves grow taller when they slow down. Imagine a series of ocean waves approaching a beach. In deep water, the waves travel fast and their crests are spaced far apart. As they enter shallow water, they slow down. The wave period — the time between crests — stays the same (it is determined by the source, not the medium), but as the wave slows, each crest catches up slightly on the one ahead of it, and the wavelength shortens. Because energy is conserved but is now concentrated in a shorter, slower wave, the wave must grow taller — the surf zone pileup of ocean waves is driven by the same impedance principle that drives seismic amplification in soft sediment.

In earthquake seismology, the incoming seismic wave from bedrock enters the soft sediment layer, slows dramatically, and its wavelength shortens proportionally. The same energy is now packed into a slower, shorter wave — and the amplitude grows. At the free surface (the ground surface), the wave reflects back downward, and the superposition of the upgoing and downgoing waves at the surface creates a standing wave pattern that further doubles the amplitude. The total amplification at the surface of a soft sediment layer can be large enough to turn a distant, moderate earthquake into a locally catastrophic event.

Hard Rock: Granite, Gneiss, and Crystalline Basement

Crystalline basement rock — granite, gneiss, quartzite, and similar metamorphic and igneous rocks — represents the reference condition for seismic wave propagation. Fresh, unfractured granite transmits seismic waves at 5,500–6,500 m/s for P-waves and 3,000–3,800 m/s for S-waves, with very low attenuation: the waves lose little energy to heat as they propagate, traveling efficiently over large distances. High impedance means incoming waves from depth are not significantly amplified at the surface — the impedance contrast between deep basement and shallow basement is small, and there is no soft layer to trap and resonate the energy.

Cities founded on crystalline bedrock experience earthquake shaking closest to what the source earthquake actually produced — unmodified by site effects. San Francisco's Nob Hill and Pacific Heights neighborhoods, founded on Franciscan Complex graywacke and chert, experienced far less shaking in the 1989 Loma Prieta earthquake than the Marina District built on bay fill just a few kilometers away. The difference in shaking intensity between those neighborhoods — not the difference in distance to the fault — determined where buildings collapsed and where they stood.

Fractures, Faults, and Velocity Anomalies in Hard Rock

Fresh crystalline rock is an idealization. Real crustal rock is fractured at multiple scales — from microscopic grain-boundary cracks to kilometer-scale fault zones — and the degree of fracturing dramatically affects seismic velocity. A highly fractured granite may have P-wave velocities 30–50% lower than its intact counterpart, because the cracks reduce the effective elastic moduli without proportionally reducing the density. Fluid-filled fractures are particularly effective at reducing S-wave velocities, because the fluid provides additional compliance (it cannot sustain shear) while adding to the mass that must be accelerated.

Fault zones are the extreme end of this spectrum. The principal slip zone of a major fault — where grinding and pulverization have reduced rock to fine-grained fault gouge — can have S-wave velocities as low as 200–500 m/s, surrounded by a damage zone of fractured rock at 800–1,500 m/s, embedded in intact rock at 3,000+ m/s. This velocity structure creates a low-velocity channel along the fault that can trap and guide seismic waves — the fault zone waveguide effect — concentrating energy along the fault trace and potentially amplifying shaking in the narrow corridor immediately adjacent to the fault. The San Andreas and North Anatolian faults both show this effect in high-density seismograph data.

Volcanic and Oceanic Rocks: Basalt and Ophiolite

Basalt — the dominant rock of oceanic crust and of many continental volcanic provinces — has seismic velocities broadly similar to granite in its fresh, dense form (5,000–6,500 m/s for P-waves), but volcanic rocks are rarely fresh and dense. Hawaiian-type basalt flows are typically highly vesicular in their upper parts, with 20–40% porosity from trapped gas bubbles, dramatically reducing seismic velocity. Pyroclastic deposits — ash, pumice, and volcanic tuff — are even more porous and can have P-wave velocities as low as 1,000–2,000 m/s despite being technically "volcanic rock."

In subduction zones and oceanic settings, altered basalt — greenschist-facies basalt with chlorite, epidote, and actinolite replacing the original pyroxene and olivine — has significantly lower velocities than fresh basalt, and serpentinized peridotite (mantle rock altered by reaction with seawater) can have P-wave velocities as low as 3,500–5,000 m/s compared to 7,800–8,200 m/s for fresh peridotite. These velocity contrasts within the subducting slab and the overlying mantle wedge create complex wave propagation effects that influence the distribution of shaking in arc and forearc settings.

Sedimentary Rocks: Limestone, Sandstone, and Shale

Sedimentary rocks span a wide range of seismic properties depending on their lithology, porosity, cement type, and degree of fluid saturation. This variability makes them the most complex rock class from a seismic wave propagation standpoint, and the most important from a practical hazard perspective — most cities in the world are built on sedimentary rock or sediment cover.

Limestone and Dolomite

Dense, well-cemented limestone and dolomite are among the stiffest sedimentary rocks, with P-wave velocities of 3,500–6,000 m/s — approaching crystalline basement in their best-cemented forms. Their impedance is high enough that cities founded on intact limestone experience relatively modest site amplification. However, karstified limestone — riddled with dissolution cavities and caves — is a very different material. Karst reduces both velocity and the effective elastic moduli dramatically, and the collapse of karst cavities during strong shaking is a distinct secondary hazard in regions like southeastern Europe, the Caribbean, and parts of the Middle East where karst limestone underlies urban areas.

Sandstone: The Fluid Saturation Effect

Sandstone illustrates one of the most important principles in rock physics: the Gassmann effect, or the dramatic influence of pore fluid on seismic velocity. Dry sandstone with 20% porosity might have a P-wave velocity of 2,000–3,000 m/s. The same sandstone fully saturated with water will have a P-wave velocity of 3,000–4,500 m/s — significantly higher — because the incompressible water stiffens the rock against compression. But the S-wave velocity changes very little between dry and saturated sandstone, because S-waves involve shear deformation and water cannot resist shear: the water adds mass without adding shear stiffness, slightly reducing V_S.

This differential response of V_P and V_S to fluid saturation is the physical basis of seismic exploration for petroleum reservoirs: a gas-saturated sandstone has a very different V_P/V_S ratio than the same sandstone saturated with brine or oil, and the change in this ratio — detectable in seismic reflection data — can identify hydrocarbon accumulations before drilling. The same physics also means that the seismic amplification properties of near-surface sandstone and gravel layers change with the water table depth: the saturated zone below the water table has higher P-wave velocity and impedance than the unsaturated zone above it, creating an additional impedance contrast at the water table that can trap and amplify seismic waves.

Unconsolidated Sediments: Where Amplification Is Greatest

The greatest site amplification — and the greatest seismic hazard from soft-ground effects — occurs in thick accumulations of unconsolidated or poorly consolidated Quaternary sediments: alluvial fans, river floodplains, bay muds, lacustrine clays, and estuarine deposits. These materials are geologically young, mechanically weak, and almost universally water-saturated — a combination that produces the lowest seismic velocities and the highest impedance contrasts with underlying bedrock of any natural material at the Earth's surface.

Bay Muds and Marine Clays

Marine and estuarine clays deposited in protected bay environments are among the most seismically amplifying materials known. San Francisco Bay mud — the Holocene-age soft clay that underlies the filled shoreline districts of San Francisco, Oakland, and the South Bay — has S-wave velocities of 50–150 m/s and densities of 1,500–1,700 kg/m³, giving an impedance of roughly 0.75–2.5 × 10⁵ kg/m²·s. The contrast with the Franciscan bedrock beneath it (impedance ~1.4 × 10⁷ kg/m²·s) is nearly two orders of magnitude — producing amplification factors of 5–15 for moderate-period ground motion.

During the 1989 Loma Prieta earthquake, the Marina District of San Francisco — built on bay fill and hydraulically placed sand from the 1906 earthquake rubble — experienced ground shaking three to five times more intense than nearby bedrock sites. Multiple apartment buildings collapsed, fires broke out from ruptured gas lines, and the Embarcadero Freeway was severely damaged — all in a neighborhood that was 50 kilometers from the earthquake's epicenter. The Cypress Street Viaduct in Oakland, built on similar bay mud, collapsed entirely, killing 42 people trapped in their cars.

Mexico City Lacustrine Clays: The Extreme Case

The Lake Texcoco sediments beneath central Mexico City represent the most extensively studied and most catastrophically consequential soft-soil seismic amplification site in the world. The ancient lakebed consists of exceptionally soft, high-water-content volcanic clay — montmorillonite-rich, with water content exceeding 300–400% of the dry soil weight in its natural state. S-wave velocities in this material are as low as 40–70 m/s in the shallowest layers, with the clay layer extending to depths of 20–50 meters before reaching stiffer transition zones and eventually competent bedrock.

In 1985, this sediment package amplified the incoming seismic waves from the distant Michoacán earthquake by factors of 30–50 in the frequency band of 0.3–0.5 Hz (periods of 2–3 seconds). This happened to match almost exactly the natural resonant period of the clay layer itself, and also the natural period of many 8–15 story reinforced concrete buildings in the city. The triple resonance — source waves, soft soil, and building period all aligned — produced an unusually prolonged and destructive shaking episode that exhausted the ductility of buildings that would normally have survived. The 1985 earthquake became the defining case study for the importance of site-specific seismic hazard assessment in urban planning.

Basin Resonance: When the Whole Valley Rings

Site amplification due to a single impedance contrast at the bedrock surface is amplified further when the soft sediment layer has a specific geometry: a flat-bottomed basin with parallel horizontal layers. In this configuration, seismic waves trapped in the sediment layer bounce back and forth between the free surface above and the bedrock below, and if the travel time for one round trip equals the wave period, the waves constructively interfere — resonance — and the amplitude grows with each bounce until attenuation limits the buildup.

The resonant period of a uniform sediment layer of thickness H and S-wave velocity V_S is T = 4H / V_S. For the Mexico City lakebed (H ≈ 30 m, V_S ≈ 40 m/s), T = 4 × 30 / 40 = 3 seconds — consistent with the observed peak amplification period. For the Seattle basin (H up to 7 km of Quaternary sediment with average V_S ≈ 500 m/s), the fundamental resonant period is much longer — T ≈ 56 seconds — primarily affecting the longest-period surface waves from distant great earthquakes rather than the short-period body waves from local moderate events.

3D Basin Effects and Edge Waves

Real sedimentary basins are not flat-bottomed uniform layers — they are three-dimensional bowls with irregular bedrock topography, lateral velocity gradients, and basin edges where the sediment pinches out against bedrock. At these basin edges, incoming seismic waves diffract, converting body wave energy into surface waves that travel laterally into the basin interior. These basin-edge-generated surface waves — sometimes called edge waves or locally generated surface waves — arrive after the direct body waves and extend the duration of strong shaking significantly beyond what a 1D vertical layer model would predict.

The Kobe, Japan, earthquake of 1995 produced a belt of severe damage along the northern edge of the Osaka-Kobe alluvial plain — a narrow zone just a few kilometers wide where the basin edge geometry focused and amplified ground motion to intensities far exceeding those on both the hard rock north of the damage belt and the deeper basin sediments to the south. The "damage belt" effect in Kobe, originally mysterious, was subsequently explained by 3D wave propagation modeling of the Osaka basin geometry — demonstrating that basin geometry, not just basin sediment properties, is a critical determinant of local seismic hazard.

Liquefaction: When Shaking Turns Ground to Fluid

The most catastrophic consequence of seismic wave propagation through saturated loose sediment is not amplification but liquefaction — the sudden, shaking-induced loss of strength in water-saturated granular materials that causes them to behave temporarily as a dense fluid rather than a solid. Liquefaction occurs when cyclic shear stress from seismic waves causes loose sand or silt grains to rearrange into a denser packing, forcing pore water upward faster than it can drain. The excess pore water pressure temporarily carries the full weight of the overlying soil, reducing the effective stress between grains to approximately zero and with it all frictional resistance to deformation.

A liquefied soil layer has essentially no shear strength: it flows like a liquid on gentle slopes, allows heavy structures to sink into it and light structures to float up through it, and can spread laterally on slopes as low as 2–3 degrees in a process called lateral spreading. The 1964 Niigata, Japan, earthquake produced spectacular liquefaction: entire apartment blocks tilted or sank into the ground as the loose river deposits beneath them liquefied. The buildings themselves were largely undamaged by the shaking — their structural integrity was intact — but they were tilted at angles of 30–80° from vertical, completely uninhabitable and economically total losses.

Attenuation: How Rock Absorbs Seismic Energy

Seismic waves do not only change velocity and amplitude as they travel through different materials — they also lose energy to the rock itself through a process called anelastic attenuation. Unlike geometric spreading (which reduces amplitude as the wavefront expands but conserves energy) or impedance effects (which redistribute energy between reflected and transmitted waves), anelastic attenuation converts seismic wave energy irreversibly into heat through the friction of grain boundaries, pore fluid viscosity, and microscopic crack surfaces as the rock deforms cyclically under the passing wave.

Attenuation is characterized by the quality factor Q — the ratio of the energy stored in one cycle of oscillation to the energy dissipated per cycle. High Q means low attenuation (the wave travels far without losing energy); low Q means high attenuation (the wave is rapidly absorbed). Crystalline basement rock has Q values of 200–1,000 — very low attenuation. Soft sediments have Q values of 10–50 — high attenuation. This explains why soft sediment amplifies long-period waves but attenuates high-frequency waves: the high attenuation of soft sediment preferentially absorbs the high-frequency energy while allowing lower-frequency energy to survive, skewing the amplified ground motion toward longer periods that are most damaging to medium-rise buildings.

The Paradox of Soft-Soil Hazard

Soft sediments simultaneously amplify ground motion at intermediate frequencies and attenuate it at high frequencies. This creates a situation where soft-soil sites experience greater shaking than bedrock sites at the periods that matter most for most buildings (0.3–3 seconds), while potentially experiencing less shaking at very high frequencies (above 10 Hz) where only very short, stiff structures are vulnerable. The net effect for most urban construction — 1–20 story buildings with natural periods of 0.1–2 seconds — is that soft-soil sites are significantly more hazardous than bedrock sites, and the hazard ratio increases with building height up to the point where the building period exceeds the dominant amplified period of the sediment layer.

Practical Implications: Why Your ZIP Code Matters

The physics described in this article has direct, practical consequences for anyone living in an earthquake-prone region. Your seismic risk is not solely determined by your distance from the nearest fault — it is determined in equal or greater measure by the geology beneath your building. Two houses a kilometer apart, one on bedrock and one on bay fill, can experience shaking intensities differing by a factor of ten in the same earthquake. That factor of ten is the difference between moderate cracking and structural collapse for many building types.

This knowledge is encoded in modern seismic hazard maps and building codes through site classification systems like the NEHRP categories (based on Vs30), probabilistic ground motion maps that incorporate site amplification factors, and microzonation studies that map the spatial distribution of soft ground conditions in urban areas. The USGS provides publicly accessible seismic hazard maps for the United States that include site amplification factors, and many earthquake-prone cities — Tokyo, Los Angeles, Istanbul, Bogotá — have produced detailed microzonation maps that identify the highest-risk soft-soil zones within the urban fabric.

Conclusion

Seismic waves do not simply travel from a fault to a city. They are transformed continuously by every geological boundary they cross — refracted at velocity gradients, reflected at impedance contrasts, amplified as they slow and narrow, attenuated as they lose energy to grain friction, resonated in sediment-filled basins, and in extreme cases converted into a liquefied substrate that destroys foundations without ever producing shaking intense enough to collapse the building above. The geology between the fault and the building is not a passive conduit — it is an active filter that can transform a distant, moderate earthquake into a local catastrophe or attenuate a nearby large one to manageable levels.

The 1985 Mexico City disaster is the canonical demonstration of what happens when this physics is ignored in urban planning. The 1989 Loma Prieta earthquake provided the same lesson for San Francisco. The 1995 Kobe earthquake illustrated the 3D complexity of basin edge effects. Each of these events added precision and urgency to a scientific understanding that was already complete in its essentials: where you build determines how badly an earthquake will hurt you, independent of how far you are from the fault. No engineering solution — no building code, no structural innovation — fully replaces the protection conferred by founded on hard rock. And no amount of distance from a fault fully compensates for the amplifying power of a deep, soft, water-saturated sedimentary basin.