Plate tectonics is the unifying theory of Earth science -- the framework that explains why continents are where they are, why earthquakes and volcanoes cluster along predictable belts, why the ocean floor is young while continental rocks can be billions of years old, and why mountain ranges rise and fall over geological time. It describes the outer layer of the Earth as divided into a small number of large, rigid plates that move relative to each other, driven by the planet's internal heat, at rates measured in centimeters per year.
The theory is less than sixty years old in its modern form, yet it has reorganized geology, geophysics, paleontology, and oceanography as thoroughly as evolution reorganized biology or quantum mechanics reorganized physics. It provides the deep context for understanding natural hazards -- earthquakes, tsunamis, volcanic eruptions -- and for interpreting the rock record that carries the history of the planet's surface over four and a half billion years. It also offers, in the supercontinent cycle, a view of Earth history on a timescale that dwarfs human civilization to a flicker.
Alfred Wegener and the Rejected Hypothesis
Alfred Wegener, a German meteorologist and polar explorer, presented the continental drift hypothesis in 1912. His core claim was that the continents had once formed a single landmass -- he called it Urkontinent, later named Pangaea -- and had since drifted apart. The evidence Wegener assembled was impressive and, in retrospect, essentially correct. The coastlines of South America and Africa fit together with remarkable precision. The Glossopteris seed fern, a distinctive Permian-era plant, appeared in fossil beds on five southern continents today separated by oceans -- implying a continuous land connection that its seeds could not otherwise have crossed. Matching sequences of Permo-Carboniferous glacial deposits appear on landmasses now sitting in the tropics, making sense only if those landmasses were once positioned near the South Pole. Mountain ranges that terminate abruptly at the Atlantic coast of Europe and Africa resume on the opposite shore of the Americas.
Yet the geological establishment rejected Wegener with something close to contempt, particularly in North America. The core objection was mechanistic: Wegener could not identify a force sufficient to move continent-sized slabs of granite through oceanic basalt. He speculated that centrifugal forces from Earth's rotation or tidal forces from the moon might be responsible, but calculations showed these were many orders of magnitude too weak. Without a plausible mechanism, most geologists treated the pattern matching as coincidence and preserved the existing paradigm of static continents connected and separated by land bridges that had since subsided.
"The continents must have once been united into a single land mass, and have since drifted apart like pieces of a cracked ice floe." -- Alfred Wegener, The Origin of Continents and Oceans (1915)
Wegener continued advocating for his hypothesis until his death on the Greenland ice sheet in 1930, never seeing its vindication. The irony is that the mechanism he lacked -- mantle convection and seafloor spreading -- was operating throughout the debate, waiting for the technology of ocean floor mapping to make it visible.
Seafloor Spreading: The Missing Mechanism Found
The confirmation of continental drift came not from the continents themselves but from the ocean floor, which was largely unmapped until midcentury naval surveys revealed an unexpected world. Wartime submarine operations had produced detailed bathymetric charts of the Atlantic, which in turn revealed the Mid-Atlantic Ridge -- a submarine mountain chain running the length of the ocean, with a distinctive rift valley along its crest.
Harry Hess at Princeton synthesized these surveys in a landmark 1962 paper titled "History of Ocean Basins," proposing that the ocean floor is continuously created at mid-ocean ridges and then spreads outward symmetrically in both directions before eventually descending back into the mantle at deep-ocean trenches. Hess called this process seafloor spreading.
The decisive confirmation came from paleomagnetism. As volcanic basalt cools and solidifies at mid-ocean ridges, iron-bearing minerals align with Earth's ambient magnetic field, recording its direction at the time of cooling. Earth's magnetic field periodically reverses -- North becomes South and vice versa -- on timescales of hundreds of thousands to millions of years. If seafloor spreading were occurring, the ocean floor should display symmetric stripes of normally and reversely magnetized rock running parallel to the ridge axis on both sides. In 1963, Fred Vine and Drummond Matthews published analysis of magnetic anomaly data from the Indian Ocean showing precisely this pattern of bilateral magnetic stripes.
The magnetic reversal chronology provided an independent clock that could be used to calculate spreading rates. The evidence converged: rock near the ridge axes is young, basalt becomes progressively older with distance from the ridge, and no ocean floor anywhere on Earth is older than about 200 million years -- confirming that all older seafloor has been recycled back into the mantle at subduction zones. By the late 1960s, the synthesis of continental drift, seafloor spreading, and paleomagnetism had consolidated into the theory of plate tectonics.
The Three Types of Plate Boundaries
Plate boundaries are zones where tectonic plates interact. The nature of the interaction determines the characteristic geological features produced.
Convergent Boundaries: Where Plates Collide
At convergent boundaries, plates approach each other. Three scenarios are possible:
When an oceanic plate converges with a continental plate, the denser oceanic plate subducts beneath the lighter continental crust, descending into the mantle at angles of 30 to 70 degrees. The descending slab carries seawater-saturated sediments and oceanic crust into the hot mantle, where the release of water dramatically lowers the melting point of the overlying mantle wedge, generating magma that rises to form volcanic arcs on the overriding plate. The Cascades of the Pacific Northwest are a volcanic arc produced by subduction of the Juan de Fuca plate beneath North America; the eruption of Mount St. Helens in 1980 was a direct product of this process.
Where two oceanic plates converge, the older, denser plate subducts, producing island arcs and deep oceanic trenches. The Mariana Trench, the deepest point on Earth's surface at 10,935 meters, marks active subduction of the Pacific plate beneath the Mariana plate.
Where two continental plates converge, neither sinks easily due to buoyancy, and the resulting collision crumples and thickens the crust into mountain ranges. The Himalayas are still rising from the ongoing collision of the Indian and Eurasian plates that began roughly 50 million years ago; Mount Everest is lifted by approximately 5 millimeters per year, partially offset by erosion.
Divergent Boundaries: Where Plates Pull Apart
At divergent boundaries, plates move apart. At mid-ocean ridges, seafloor spreading creates new oceanic crust -- the process Hess described. On continents, divergence produces rift valleys -- linear downfaulted basins -- such as the East African Rift System, which represents a continent in the early stages of splitting apart. The Red Sea and Gulf of Aden represent a more advanced stage, where rifting has proceeded far enough to create a narrow ocean. Iceland sits atop the Mid-Atlantic Ridge where it has built up above sea level, offering a rare surface exposure of a mid-ocean ridge; it is one of the most volcanically active places on Earth.
Transform Boundaries: Where Plates Slide Past
At transform boundaries, plates slide horizontally past each other with no creation or destruction of crust. The San Andreas Fault in California, where the Pacific Plate moves north relative to the North American Plate at about 5 centimeters per year, is the world's most studied transform fault. The horizontal sliding motion produces strike-slip earthquakes rather than the vertical displacement characteristic of other boundary types.
| Boundary Type | Plate Motion | Primary Feature | Example |
|---|---|---|---|
| Convergent (ocean-continent) | Toward each other | Volcanic arc, trench | Cascades, Andes |
| Convergent (continent-continent) | Toward each other | Mountain range | Himalayas, Alps |
| Convergent (ocean-ocean) | Toward each other | Island arc, trench | Japan, Mariana |
| Divergent (oceanic) | Away from each other | Mid-ocean ridge | Mid-Atlantic Ridge |
| Divergent (continental) | Away from each other | Rift valley | East African Rift |
| Transform | Horizontally past | Strike-slip fault | San Andreas Fault |
What Drives the Plates?
The driving mechanism of plate tectonics remains an active area of research, and the simple explanation -- mantle convection drags the plates like a conveyor belt -- has been substantially revised.
Ridge push is the mechanism Hess originally envisioned: hot, buoyant material rises at mid-ocean ridges, and the elevated topography creates a gravitational gradient that pushes plates downhill away from the ridge. The force is real but calculations suggest it is modest -- insufficient on its own to account for observed plate velocities.
Slab pull is now considered the dominant force. Cold, old oceanic lithosphere is denser than the underlying asthenosphere. When it subducts, its negative buoyancy creates a gravitational body force that literally pulls the trailing plate along. The correlation between subduction rate and the length of subducting slab margin is striking: plates with extensive subduction zones, like the Pacific Plate, move faster than plates with little subduction, like the African Plate.
Mantle convection remains important as the means by which heat is transported from Earth's deep interior to the surface, creating the thermal structure within which plates move. But the convection cells do not simply drag the plates passively -- the relationship between surface plate motion and deep mantle flow is bidirectional and complex. Seismic tomography, which images Earth's interior by analyzing variations in seismic wave velocities, reveals a mantle far more heterogeneous than simple convection models predict, with cold subducted slabs stagnating at the 660 km phase transition boundary and large low-shear-velocity provinces (LLSVPs) at the core-mantle boundary whose origin remains debated.
Hot Spots and Mantle Plumes
Some volcanic activity does not occur at plate boundaries. Hot spots are regions of persistent volcanism attributed to mantle plumes -- columns of anomalously hot material rising from deep in the mantle, possibly from near the core-mantle boundary. As a tectonic plate moves over a stationary hot spot, it creates a chain of volcanic islands progressively older in the direction of plate motion. The Hawaiian-Emperor Seamount Chain is the textbook example: the active Big Island of Hawaii sits directly over the current hot spot, while the older, eroded islands and submerged seamounts extending northwest to the 80-million-year-old Emperor Seamounts trace the direction and rate of Pacific Plate motion.
The Yellowstone supervolcano sits over a hot spot beneath the North American continent. The Snake River Plain in Idaho is a scar left by the continent moving southwest over the hot spot over the past 16 million years. Yellowstone's calderas are among the largest volcanic features in North America, and the hot spot drives the hydrothermal activity -- geysers, hot springs -- that makes Yellowstone National Park geologically distinctive.
The Supercontinent Cycle
The supercontinent cycle is the repeated assembly and dispersal of Earth's continental crust into single landmasses over timescales of hundreds of millions of years, driven by the long-term operation of plate tectonics.
Pangaea, the most recent supercontinent, assembled approximately 300 million years ago during the Permian period through the collision of the northern supercontinent Laurasia with the southern supercontinent Gondwana. At its peak, Pangaea extended nearly from pole to pole, with the Tethys Sea as a large embayment on its eastern side. Pangaea began breaking apart approximately 175 million years ago as the North Atlantic opened first, then the South Atlantic as South America separated from Africa.
Before Pangaea lay Rodinia, assembled approximately 1.1 billion years ago and dispersed between 750 and 600 million years ago during the Neoproterozoic. The fragmentation of Rodinia has been linked, controversially, to the Snowball Earth glaciations -- periods when global ice cover may have extended to the equator. The dispersal of continental fragments into lower latitudes increased weathering of silicate rocks, drawing down atmospheric CO2 and potentially triggering runaway glaciation.
Earlier supercontinents -- Nuna/Columbia (c. 1.8 billion years ago) and possibly others -- are reconstructed with progressively less confidence as the paleomagnetic record becomes more ambiguous and fewer rocks have survived.
Projections forward suggest another supercontinent will assemble in roughly 200 to 250 million years. Different models yield different configurations depending on which plate boundaries remain active: some predict the Pacific Ocean closing as the Americas converge with Asia (Amasia); others predict the Atlantic Ocean closing as the Americas drift back toward Eurasia (Pangaea Proxima).
Earthquakes: Mechanics and the Limits of Prediction
Tectonic plate motion is not smooth and continuous. Fault segments are locked by friction while stress accumulates over decades to centuries; when the accumulated stress exceeds the frictional strength of the fault, the locked segment suddenly slips, radiating energy as seismic waves. The elastic rebound theory, proposed by Harry Fielding Reid after the 1906 San Francisco earthquake, describes this cycle of elastic strain accumulation and sudden release.
Earthquake magnitude is measured on a logarithmic scale: the moment magnitude scale (Mw) quantifies the energy released from the seismic moment (fault area x average slip x rock stiffness). Each unit increase in magnitude corresponds to approximately 32 times more energy released. The 2011 Tohoku earthquake (Mw 9.0) released roughly 600 times as much energy as the 1994 Northridge earthquake (Mw 6.7).
Earthquake prediction -- specifying the time, location, and magnitude of future earthquakes with precision sufficient to enable evacuation -- remains beyond the reach of seismology. Researchers have investigated precursory phenomena including radon gas emissions, groundwater level changes, anomalous animal behavior, electromagnetic signals, and foreshock sequences. None has proved reliable or general. The fundamental difficulty is that earthquake nucleation involves processes at the scale of fault zone mineralogy -- fluid pressure changes, stress corrosion, frictional properties of fault gouge -- that cannot be monitored directly.
Probabilistic Seismic Hazard Analysis (PSHA) is the engineering and policy response to prediction failure. Rather than forecasting individual earthquakes, PSHA calculates the probability that ground motion will exceed specified levels at a given location within a given time period. The output is a hazard curve used to set building code standards. The 2011 Tohoku earthquake exposed limitations of PSHA in practice: the earthquake exceeded the design basis for the Fukushima nuclear plant, and post-event analysis showed that paleoseismic evidence of very large prehistoric tsunamis in the region had been underweighted in hazard assessments.
Geological Timescales and Why They Matter
Understanding plate tectonics requires comfort with timescales that have no human intuitive equivalent. The Himalayas -- which feel permanent and immovable -- are geologically young, the product of a collision that began a mere 50 million years ago. The Atlantic Ocean began opening roughly 175 million years ago; it will likely close again 250 million years from now. The rocks in Canada's Acasta Gneiss, the oldest known crustal rocks on Earth at approximately 4.03 billion years old, predate the emergence of complex life by more than 3 billion years.
The geological record preserved in rocks, fossils, and ocean floor sediments is the primary evidence base for plate tectonic reconstruction. Radiometric dating -- using the known decay rates of radioactive isotopes as natural clocks -- allows rocks to be dated with precision, providing the temporal framework within which plate motions can be reconstructed. The development of radiometric dating methods in the early twentieth century was as important to geology as the development of radiocarbon dating was to archaeology.
Plate tectonics is not merely an abstract framework for understanding the distant past. It is the driver of the active geological processes -- earthquakes, volcanic eruptions, tsunamis -- that shape human geography and impose real hazards on the 1.4 billion people who live within 100 kilometers of a plate boundary. Understanding the theory is directly practical: it informs building codes, land use planning, early warning systems, and the assessment of long-term geological risk. It also grounds one of the most important intellectual recognitions available from natural science: that the physical world we inhabit is not static but dynamic on timescales that vastly exceed any human frame of reference, and that the mountains and oceans that seem permanent to us are temporary configurations in a planet that has been rearranging itself for four and a half billion years.
Frequently Asked Questions
Who proposed continental drift, and why was the idea rejected for so long?
Alfred Wegener, a German meteorologist and polar explorer, presented the continental drift hypothesis in 1912. His core claim was that the continents had once formed a single landmass — he called it Urkontinent, later named Pangaea — and had since drifted apart. The evidence Wegener assembled was genuinely impressive. The coastlines of South America and Africa fit together like puzzle pieces with remarkable precision. The Glossopteris seed fern, a distinctive Permian-era plant, appeared in fossil beds on five southern continents that are today separated by oceans — implying a continuous land connection rather than the transoceanic dispersal its seeds could not have achieved. Matching sequences of Permo-Carboniferous glacial deposits appear on landmasses that now sit in the tropics, making sense only if those landmasses were once positioned near the South Pole. Mountain ranges that terminate abruptly at the Atlantic coast of Europe and Africa resume on the opposite shore of the Americas.Yet the geological establishment rejected Wegener with something close to contempt, particularly in North America. The core objection was mechanistic: Wegener could not identify a force sufficient to push continent-sized slabs of granite through oceanic basalt. He speculated that centrifugal forces from Earth's rotation or tidal forces from the moon might be responsible, but calculations showed these were many orders of magnitude too weak. Without a plausible mechanism, most geologists treated the pattern matching as coincidence and preserved the existing paradigm of static continents connected and separated by land bridges that had since subsided.Wegener continued advocating for his hypothesis until his death on the Greenland ice sheet in 1930, never seeing its vindication. The irony is that the mechanism he lacked — mantle convection and seafloor spreading — was operating throughout the debate, waiting for the technology of ocean floor mapping to make it visible. His case illustrates a genuine tension in the philosophy of science: empirical pattern matching, however compelling, faces resistance without mechanistic explanation, and rightfully so, since many compelling patterns are coincidental.
What is seafloor spreading and how did it confirm plate tectonics?
The confirmation of continental drift came not from the continents themselves but from the ocean floor, which was largely unmapped until midcentury naval surveys revealed an unexpected world. Harry Hess at Princeton synthesized these surveys in a 1962 paper titled 'History of Ocean Basins,' proposing that the ocean floor is continuously created at mid-ocean ridges — volcanic mountain chains running along the centers of ocean basins — and then spreads outward symmetrically in both directions before eventually descending back into the mantle at deep-ocean trenches. Hess called this process seafloor spreading.The decisive confirmation came from paleomagnetism. As volcanic basalt cools and solidifies at mid-ocean ridges, iron-bearing minerals align with Earth's ambient magnetic field, recording its direction at the time of cooling. Earth's magnetic field periodically reverses — North becomes South and vice versa — on timescales of hundreds of thousands to millions of years. If seafloor spreading were occurring, the ocean floor should display symmetric stripes of normally and reversely magnetized rock running parallel to the ridge axis on both sides. In 1963, Fred Vine and Drummond Matthews published analysis of magnetic anomaly data from the Indian Ocean showing precisely this pattern of bilateral magnetic stripes. The pattern was subsequently confirmed in ocean basins worldwide.The magnetic reversal chronology provided an independent clock that could be used to calculate spreading rates. The evidence converged: rock near the ridge axes is young, basalt becomes progressively older with distance from the ridge, and no ocean floor anywhere on Earth is older than about 200 million years — confirming that all older seafloor has been recycled back into the mantle at subduction zones. By the late 1960s, the synthesis of continental drift, seafloor spreading, and paleomagnetism had consolidated into the theory of plate tectonics, one of the most comprehensive unifying frameworks in all of science.
What are the three types of plate boundaries and what geological features does each produce?
Plate boundaries are zones where tectonic plates interact, and the nature of the interaction — whether plates are moving toward each other, away from each other, or sliding past each other — determines the characteristic geological features produced.At convergent boundaries, plates approach each other. When an oceanic plate converges with a continental plate, the denser oceanic plate subducts beneath the lighter continental crust, descending into the mantle at angles of 30 to 70 degrees. The descending slab carries seawater-saturated sediments and oceanic crust into the hot mantle, where the release of water dramatically lowers the melting point of the overlying mantle wedge, generating magma that rises to form volcanic arcs on the overriding plate. The Cascades of the Pacific Northwest are a volcanic arc produced by subduction of the Juan de Fuca plate beneath North America. Where two oceanic plates converge, the older, denser plate subducts, producing island arcs and deep oceanic trenches — the Mariana Trench, the deepest point on Earth's surface, marks active subduction. Where two continental plates converge, neither sinks easily due to buoyancy, and the resulting collision crumples and thickens the crust into mountain ranges; the Himalayas are still rising from the ongoing collision of the Indian and Eurasian plates that began roughly 50 million years ago.At divergent boundaries, plates move apart. At mid-ocean ridges, seafloor spreading creates new oceanic crust. On continents, divergence produces rift valleys — linear downfaulted basins — such as the East African Rift System, which represents a continent in the early stages of splitting apart, a process that will eventually produce a new ocean if it continues. Iceland sits atop the Mid-Atlantic Ridge where it has built up above sea level, offering a rare surface exposure of a mid-ocean ridge.At transform boundaries, plates slide horizontally past each other with no creation or destruction of crust. The San Andreas Fault in California, where the Pacific Plate moves north relative to the North American Plate at about 5 centimeters per year, is the world's most studied transform fault. The horizontal sliding motion produces strike-slip earthquakes rather than the vertical displacement characteristic of other boundary types.
What drives plate tectonics, and is it mantle convection, slab pull, or ridge push?
The driving mechanism of plate tectonics remains an active area of research, and the simple answer — mantle convection drags the plates — has been substantially revised by decades of subsequent work. The current consensus assigns primary importance to slab pull, with ridge push playing a secondary role and thermal convection operating as an overarching context rather than a direct driver.Ridge push is the mechanism Hess originally envisioned: hot, buoyant material rises at mid-ocean ridges, and the elevated topography of the ridge creates a gravitational gradient that pushes plates downhill away from the ridge. The force is real but calculations suggest it is modest — insufficient on its own to account for observed plate velocities.Slab pull is now considered the dominant force. Cold, old oceanic lithosphere is denser than the underlying asthenosphere. When it subducts, its negative buoyancy creates a gravitational body force that literally pulls the trailing plate along. The correlation between subduction rate and the length of subducting slab margin is striking: plates with extensive subduction zones, like the Pacific Plate, move faster than plates with little subduction, like the African Plate. Where the slab detaches (slab rollback), it can produce rapid extension of the overlying plate.Mantle convection remains important as the means by which heat is transported from Earth's deep interior to the surface, creating the thermal structure within which plates move. Hot upwellings carry heat from the core-mantle boundary to the base of the lithosphere; cold downwellings (subducting slabs) return material to depth. But the convection cells do not simply drag the plates passively like conveyor belts — the relationship between surface plate motion and deep mantle flow is bidirectional and complex. Seismic tomography, which images Earth's interior by analyzing variations in seismic wave velocities, reveals a mantle far more heterogeneous than simple convection models predict, with cold subducted slabs stagnating at the 660 km phase transition boundary and large low-shear-velocity provinces (LLSVPs) at the core-mantle boundary whose origin and significance remain debated.
What is the supercontinent cycle and what do we know about Pangaea and Rodinia?
The supercontinent cycle is the repeated assembly and dispersal of Earth's continental crust into single landmasses over timescales of hundreds of millions of years, driven by the long-term operation of plate tectonics. The evidence that such a cycle has operated throughout Earth's history is drawn from paleomagnetic records, the matching of ancient orogenic belts across modern continental margins, and the global distribution of certain rock types and fossil assemblages.Pangaea, the most recent supercontinent, assembled approximately 300 million years ago during the Permian period through the collision of the northern supercontinent Laurasia with the southern supercontinent Gondwana. At its peak, Pangaea extended nearly from pole to pole, with the Tethys Sea as a large embayment on its eastern side. Its existence is supported by the same evidence Wegener marshaled: Glossopteris fossils, Permo-Carboniferous glacial deposits, and matching coastlines. Pangaea began breaking apart approximately 175 million years ago as the North Atlantic opened first, then the South Atlantic as South America separated from Africa, processes that continue today.Before Pangaea lay Rodinia, assembled approximately 1.1 billion years ago and dispersed between 750 and 600 million years ago during the Neoproterozoic. The fragmentation of Rodinia has been linked, controversially, to the Snowball Earth glaciations — periods when global ice cover may have extended to the equator. The dispersal of continental fragments into lower latitudes increased weathering of silicate rocks, drawing down atmospheric CO2 and potentially triggering runaway glaciation. Rodinia's assembly is less well-constrained than Pangaea's because older rocks are fewer and the paleomagnetic record becomes increasingly difficult to interpret.Projections forward suggest another supercontinent — variously called Pangaea Proxima, Amasia, or Novopangaea, depending on the model — will assemble in roughly 200 to 250 million years. The exact configuration depends on which plate boundaries remain active and how the major ocean basins evolve. In some models, the Pacific Ocean closes as the Americas converge with Asia; in others, the Atlantic Ocean closes as the Americas drift back toward Eurasia. These are not idle speculations — they are constrained predictions from the same plate motion models that successfully retrodict the assembly of Pangaea.
Why is earthquake prediction so difficult, and what is probabilistic seismic hazard analysis?
Earthquake prediction — specifying the time, location, and magnitude of future earthquakes with precision sufficient to enable evacuation — remains beyond the reach of seismology. This is not for lack of effort. Researchers have investigated precursory phenomena including radon gas emissions, groundwater level changes, anomalous animal behavior, electromagnetic signals, and foreshock sequences. None has proved reliable or general enough for operational forecasting. The 2009 L'Aquila earthquake in Italy, which killed 308 people, became a cause celebre when several scientists were initially convicted of manslaughter (later acquitted on appeal) for reassuring the public after a swarm of foreshocks, illustrating the human stakes of communicating earthquake uncertainty.The fundamental difficulty is that earthquake nucleation involves processes at the scale of fault zone mineralogy — fluid pressure changes, stress corrosion of fault-sealing minerals, the frictional properties of gouge — that cannot be monitored directly and are sensitive to initial conditions in ways that make long-range prediction inherently problematic. Earthquake faults also interact through static and dynamic stress changes over large areas, so the state of any given fault depends on the history of the entire fault network.Probabilistic Seismic Hazard Analysis (PSHA) is the engineering and policy response to prediction failure. Rather than attempting to forecast individual earthquakes, PSHA calculates the probability that ground motion will exceed specified levels at a given location within a given time period — typically expressed as the probability of exceedance over 50 years, which maps to return periods relevant to building lifetimes. PSHA integrates a seismic source model (characterizing which faults exist, how fast they slip, and their historical earthquake record), a ground motion model (how shaking attenuates with distance), and uncertainty in both. The output is a hazard curve used to set building code standards.The 2011 Tohoku earthquake and tsunami, which killed approximately 19,000 people and triggered the Fukushima nuclear disaster, exposed limitations of PSHA in practice. The earthquake, at magnitude 9.0, exceeded the design basis for the Fukushima plant by more than an order of magnitude. Post-event analysis showed that paleoseismic evidence of very large prehistoric tsunamis in the region existed but had been underweighted in hazard assessments. The difficulty of incorporating low-frequency, high-magnitude events into models designed primarily around the historical instrumental record remains a fundamental challenge.