Plateaus form when large blocks of the Earth’s crust are pushed upward, built up by volcanic material, or both, and then maintained as broad, elevated flatlands because the forces that raised them outpaced the erosion working to tear them down. The process is rarely simple or singular. The Tibetan Plateau owes its enormous height to the collision of two continental plates, while the Altiplano in South America rose through a combination of crustal shortening, magma additions, and the removal of dense rock from beneath. Even the Colorado Plateau, which looks geologically calm on the surface, was lifted in at least three distinct episodes driven by very different mechanisms over tens of millions of years. Understanding how a plateau forms means following a story that starts deep in the Earth’s interior and ends with rivers, glaciers, and climate reshaping the surface.
When Continents Collide
The most dramatic plateau-building mechanism on Earth is the head-on collision of tectonic plates. The Tibetan Plateau, the highest and largest on the planet, exists because the Indian plate has been plowing into the Asian plate for roughly 50 million years. That ongoing collision crumpled and thickened the crust in a way loosely comparable to what happens when you push two stacks of paper toward each other on a desk: the material in between buckles, folds, and piles up. Field data from the western Qaidam Basin, a sedimentary depression on the plateau’s northern edge, confirm that the region formed in a compressive tectonic environment driven by this India-Asia collision.1Geoscience Frontiers. Cenozoic uplift of the Tibetan Plateau: Evidence from the tectonic–sedimentary evolution of the western Qaidam Basin The result is a crustal slab roughly twice the normal continental thickness, which floats high on the denser mantle rock beneath it, much like a thick wooden block rides higher in water than a thin one.
South America’s Altiplano-Puna plateau, the second largest on Earth, tells a related but distinct story. There, the oceanic Nazca plate is diving beneath the South American plate rather than colliding head-on. The plateau rose primarily because horizontal shortening thickened a crust that had already been softened by the heat of nearby volcanic activity. Uplift in the Altiplano region began around 25 million years ago, coinciding with a speedup in the rate at which the plates were converging and a period when the subducting slab flattened out beneath the continent.2Annual Review of Earth and Planetary Sciences. THE EVOLUTION OF THE ALTIPLANO-PUNA PLATEAU OF THE CENTRAL ANDES But the shortening observed at the surface accounts for only about 70 to 80 percent of the total crustal thickening, which means other processes like magma injection from below and thinning of the dense lower lithosphere contributed the rest.
Even within a single plateau, the timing of uplift can vary from place to place. In the Central Andes, different stretches of the Eastern Cordillera and Altiplano rose during different geological windows, driven at various times by crustal shortening, the removal of dense lower rock, and lateral flow of crustal material.3Annual Review of Earth and Planetary Sciences. Tectonic Evolution of the Central Andean Plateau and Implications for the Growth of Plateaus A plateau, in other words, is not simply punched upward in one event. It is assembled over millions of years through overlapping processes that shift in importance as conditions change.
Plateaus Built by Lava
Not every plateau owes its elevation to colliding or subducting plates. Some are constructed from the ground up by enormous volumes of volcanic rock. Continental flood basalt provinces form when repeated eruptions pour lava across a landscape over millions of years. The main eruptive phase is usually intense but geologically brief, lasting roughly one to five million years, during which stacks of thick, flat-lying lava flows accumulate into a layered sequence that can reach kilometers in thickness.4Lithos. The anatomy of Continental Flood Basalt Provinces: geological constraints on the processes and products of flood volcanism The Columbia River Basalt Group in the northwestern United States and the Deccan Traps in India are classic examples: flat-topped landscapes whose elevation comes largely from piled-up lava rather than crustal shortening.
Oceanic plateaus work on the same basic principle but beneath the sea. The Ontong Java Plateau in the western Pacific is the world’s largest oceanic plateau, and recent seismic data show that its total crustal thickness reaches about 23 kilometers. Only around three kilometers of that is layered basalt from surface eruptions; the remaining 12 to 15 kilometers were added by magma that intruded into the mid- and lower crust through underplating and deep injection during major plume eruptions roughly 116 to 108 million years ago.5PubMed Central. Probing the world’s largest oceanic plateau: from making to collision The surface lava, in this case, is almost a sideshow compared with the massive hidden thickening below.
Mantle Plumes and the Push From Below
Some plateaus owe a significant part of their elevation to hot material rising from deep in the Earth’s mantle. When a mantle plume, a column of abnormally hot rock, impinges on the base of the crust, it can dome the surface upward and trigger flood volcanism at the same time. The Ethiopian Plateau in East Africa is a good example: it sits above a region where the Afar plume punched into the base of the continental lithosphere, doming the crust upward and triggering massive Oligocene-age flood basalt eruptions. The plateau today is thought to represent the preserved remnant of that domed surface, shaped by both the plume’s push and the underplating of magma that ponded beneath the crust as the basalts differentiated.6Earth and Planetary Science Letters. Stability of the Upper Nile drainage network (Ethiopia) deduced from (U–Th)/He thermochronometry: implications for uplift and erosion of the Afar plume dome
The distinction between a plume-driven plateau and a collision-driven one matters because it affects the surface expression. Plume uplift tends to produce a broad, roughly circular dome that gets modified by later volcanism and erosion, whereas collision zones produce elongated plateaus that track the plate boundary. In reality, most large plateaus bear fingerprints of more than one mechanism, which is part of why geologists still argue over the relative contributions of each process.
The Colorado Plateau and Plateaus With Complicated Histories
The Colorado Plateau in the southwestern United States is a favorite puzzle in geology because it sits at high elevation (averaging roughly 2,000 meters) yet lacks the severe internal deformation you would expect from a collision zone. It was uplifted in at least three separate episodes spanning roughly 50 million years. The first, between about 70 and 50 million years ago, raised the region above sea level as a flat oceanic slab slid beneath the continent at an unusually shallow angle. A second pulse of uplift, roughly 38 to 23 million years ago, coincided with intense regional volcanism and the removal of that subducted slab. A third phase, less than 20 million years ago, tracked the inward spread of basaltic magmatism as convective erosion ate away at the dense core of the underlying lithosphere.7Annual Review of Earth and Planetary Sciences. Tectonics of the Colorado Plateau and Its Margins
Before any of that uplift, the region actually sank. With the arrival of the flat-lying Farallon slab beneath western North America, dynamic subsidence swept from west to east across the plateau and reached its maximum around 86 million years ago, when the region was flooded by shallow seas.8Geology. Dynamic subsidence and uplift of the Colorado Plateau The fact that the same chunk of crust first sank, then rose in multiple stages driven by entirely different deep-Earth processes, illustrates why “how is a plateau formed” rarely has a one-sentence answer. The mechanism that lifts a plateau can change completely over its lifetime.
Erosion as Sculptor
Once a plateau is elevated, erosion goes to work immediately. Rivers carve canyons into its surface, ice sheets and glaciers grind down its margins, and the slow retreat of cliff faces strips away overlying rock layer by layer. The Grand Canyon is the most famous product of this interaction: a plateau lifted high enough for a river to cut more than a kilometer downward through flat-lying sedimentary rock.
On the Colorado Plateau, scarp retreat has been one of the dominant erosional processes for tens of millions of years. Resistant caprock layers protect softer rock below, but once the caprock edge is breached, the cliff face retreats horizontally, sometimes for great distances. Research on retreat rates of scarps carved in Upper Cretaceous rocks shows that this process was fast enough to bring the current cliff faces to their present positions from the center of the Monument Uplift on the central plateau. The rates are controlled mainly by the thickness and hardness of the caprock, and because scarp retreat can operate simultaneously at different stratigraphic levels, it was capable of stripping away vast volumes of the plateau’s sedimentary cover during the Cenozoic.9Earth Surface Processes and Landforms. The significance of scarp retreat for cenozoic landform evolution on the Colorado Plateau, U.S.A.
In mountain ranges and high-elevation plateaus that extend into cold climates, glacial erosion can act as a ceiling on how high the land gets. In the Cascade Range of Washington State, glacial erosion linked to spatial gradients in the elevation where snow accumulates faster than it melts created a tilted surface of hundreds of cirques across the range. Peaks and ridges now rise no more than about 600 meters above this zone, effectively capping the range’s height. Slopes above the cirques hover near threshold steepness, meaning any further uplift is matched by faster glacial and hillslope erosion that shaves the surface back down.10Quaternary Research. Influence of a glacial buzzsaw on the height and morphology of the Cascade Range in central Washington State, USA Geologists sometimes call this the “glacial buzzsaw” effect, and it helps explain why many high plateaus and ranges have surprisingly uniform summit elevations rather than ever-increasing peaks.
How Scientists Reconstruct Past Elevations
Determining when and how fast a plateau rose is harder than it sounds, because the rocks themselves do not carry an altitude stamp. Geologists have developed several indirect methods. One inventive approach uses the bubbles trapped in ancient basalt lava flows. Because the pressure difference between bubbles at the top and bottom of a flow depends on atmospheric pressure at the time the lava solidified, measuring the size distribution of those vesicles gives a direct read on the paleopressure, and therefore the paleoelevation, at the eruption site. This technique has been applied to the Colorado Plateau and avoids the uncertainties that plague proxies based on temperature or moisture, which are affected by climate as well as altitude.11Geology. Timing of Colorado Plateau uplift: Initial constraints from vesicular basalt-derived paleoelevations
Biological proxies offer another window. On the northern Tibetan Plateau, pollen records of montane conifers have been used to reconstruct past elevations. During the Middle Miocene, about 15 million years ago, two parallel mountain ranges in the region stood at roughly 1,300 meters and 400 meters, respectively. Both then rose rapidly to around 3,600 meters by the Late Miocene, around 7 to 11 million years ago, depending on location.12PubMed. A new biologic paleoaltimetry indicating Late Miocene rapid uplift of northern Tibet Plateau That kind of rapid gain, more than two kilometers in a few million years, suggests punctuated episodes of uplift rather than a slow, steady climb.
How Plateaus Reshape Climate
A plateau does not just passively sit at elevation; it changes the atmosphere around it. The Tibetan Plateau is the most studied example. Numerical climate modeling shows that the uplift of Tibet was a major driver of the East Asian monsoon system. The East Asian monsoon appears to be more sensitive to Tibetan uplift than the South Asian monsoon, and the winter monsoon component is affected more than the summer one. In northern China, the monsoon climate essentially switched on when the plateau reached roughly half its current elevation, establishing the strong winter northerly winds that define the region’s climate today.13Palaeogeography, Palaeoclimatology, Palaeoecology. Sensitivity of East Asian monsoon climate to the uplift of the Tibetan Plateau
Uplift also creates rain shadows. As the northern Tibetan Plateau rose, it caused significant reductions in annual precipitation across a broad swath of inland Asia to its north, primarily by blocking moisture-bearing air masses and enhancing the rain shadow effect. At the same time, mountainous areas within the uplift zone saw increased rainfall during their respective rainy seasons because air forced upward over the new high ground shed its moisture on the windward side.14Quaternary Science Reviews. Impacts of uplift of northern Tibetan Plateau and formation of Asian inland deserts on regional climate and environment This is one mechanism behind the formation of Central Asia’s vast inland deserts: as Tibet grew taller, it progressively starved the land to its north of moisture.
Rain shadow effects are not confined to Tibet. In the eastern Himalayas, the uplift of the smaller Shillong Plateau redirected moisture patterns over the foreland basin. Once the Shillong Plateau reached a critical elevation, roughly 1.2 million years ago, it began forcing significant orographic rainfall on its windward side, and mean annual precipitation on the leeward side dropped by a factor of about 1.7 to 2.5 over the last one to three million years.15Geochemistry, Geophysics, Geosystems. Formation of a Rain Shadow: O and H Stable Isotope Records in Authigenic Clays From the Siwalik Group in Eastern Bhutan A plateau does not need to be enormous to reshape regional rainfall; even a modest rise can cross a threshold that rewrites the hydrological map.
Volcanic Plateaus and the Carbon Cycle
Flood basalt plateaus have an outsized influence on Earth’s long-term climate through a mechanism that has nothing to do with blocking wind. When basaltic rock is exposed to rain and atmospheric gases, it weathers chemically, and that weathering process pulls carbon dioxide out of the atmosphere. The CO2 consumed by basalt weathering worldwide is estimated at about 4 trillion moles per year, and the rate is strongly correlated with temperature: hotter climates weather basalt faster, drawing down more CO2.16Chemical Geology. Basalt weathering laws and the impact of basalt weathering on the global carbon cycle17Earth and Planetary Science Letters. Temperature dependence of basalt weathering
This creates a feedback loop. A large igneous province erupts, flooding the atmosphere with CO2 and warming the planet. But the newly exposed basalt then weathers faster because the climate is warmer, pulling CO2 back out. Modeling of Pangaea-era flood basalt events shows that on multimillion-year timescales, the weathering-driven carbon drawdown dominates over the eruptive CO2 input for most of the studied timeframe, meaning the net long-term effect of a flood basalt province is to cool the planet rather than warm it.18PubMed Central. Limited long-term cooling effects of Pangaean flood basalt weathering The volcanic plateaus that seem like agents of climate catastrophe in their initial eruption phase gradually become part of the planet’s thermostat.
Plateaus Beyond Earth
Plateau formation is not unique to our planet. Mars has the Tharsis rise, a volcanic province the size of North America that stands about 10 kilometers above the surrounding terrain. Unlike Tibet or the Altiplano, Tharsis was likely built primarily by volcanic construction rather than tectonic uplift. Mars lacks the plate tectonics that drives collision zones on Earth, so the massive topography of Tharsis is better explained by billions of years of volcanic output piling up over a stationary hot spot in the mantle, thickened further by the flexural response of the lithosphere beneath the load.19Journal of Geophysical Research: Solid Earth. Evolution of the Tharsis Province of Mars: The importance of heterogeneous lithospheric thickness and volcanic construction Without plate motion to spread the volcanic material laterally and without significant water erosion to carve it away, the Tharsis rise has remained a monumental feature for billions of years, far more static than any plateau on Earth.
Living on a Plateau That Keeps Changing
Plateaus are not frozen in time once they form. They continue to be reshaped by erosion, climate shifts, and human activity. On the southern Tibetan Plateau, pollen and fungal spore records spanning 3,600 years show a major shift in how people used the landscape. For the first 1,800 years of the record, cereal-type pollen dominated, indicating farming. Then, around 1,800 years ago, cereal pollen dropped sharply while indicators of livestock grazing, including dung-associated fungal spores, increased by roughly four-fold.20Nature / Communications Earth & Environment. 3,600 years of human adaptation to drought intensification on the southern Tibetan Plateau The shift tracks a period of intensifying drought, and it shows how human societies have had to continuously adapt to the environmental conditions that a plateau’s elevation and geography impose. The same geological forces that built the plateau millions of years ago still shape the lives of the people who inhabit it, through the climate patterns, soil types, and water availability that high elevation creates.