What Will Earth Look Like in 1 Million Years?

Earth one million years from now would still be recognizable at a glance: the continents would sit in almost exactly the same positions, the Sun would shine at nearly the same brightness, and the basic chemistry of oceans and atmosphere would remain familiar. But zoom in and the differences become striking. Ice sheets would likely have advanced and retreated multiple times, coastlines would have shifted, entire mountain ranges would have worn lower, and the species roaming every continent would be different from anything alive today. Perhaps most surprising, the chemical fingerprint of our own civilization, particularly the carbon dioxide we have released, would still be faintly detectable in ocean chemistry and rock.

Continents in Almost the Same Place

Tectonic plates move at roughly the speed your fingernails grow, typically a few centimeters per year. Over a million years, that adds up to somewhere between 20 and 70 kilometers of displacement for most plates. To put that in perspective, the Atlantic Ocean is about 5,000 kilometers wide. A million years of spreading would widen it by a distance roughly equal to the length of a marathon course. North America, South America, Africa, Europe, and Asia would all be in essentially the same places on a globe, and a time-traveling geographer could identify every continent without much trouble.

That said, the edges of plates are where the action concentrates. Subduction zones would have consumed a thin strip of Pacific seafloor. East Africa’s Great Rift Valley would have widened and deepened, inching the Somali Plate further away from the main African plate. Iceland would have grown slightly as the Mid-Atlantic Ridge continues to build new crust. None of these changes would redraw the world map in any dramatic way, but they would be measurable, and localized earthquakes and volcanic eruptions along these boundaries would continue at roughly their present pace.

Ice Ages and the Climate Pendulum

One million years is long enough for Earth to cycle through roughly ten full glacial-interglacial swings. These swings are driven by slow, periodic changes in Earth’s orbit and axial tilt, collectively known as Milankovitch cycles, which alter how much solar energy reaches high latitudes at critical times of year. The three main periodicities run at about 23,000, 41,000, and 100,000 years, and together they have been the dominant force governing the ice ages over the past several million years.¹

During glacial peaks, ice sheets several kilometers thick have historically blanketed much of Canada, Scandinavia, and northern Russia, locking up enough water to drop sea levels by well over 100 meters. During interglacials like the one we currently enjoy, ice retreats to Greenland and Antarctica, and seas rise. Earth’s climate system tends to settle into distinct states that researchers have labeled Hothouse, Warmhouse, Coolhouse, and Icehouse, each characterized by different greenhouse gas concentrations and polar ice volumes, with threshold-like transitions between them rather than smooth gradients.² Over the past 40 million years, these transitions have looked like abrupt jumps between quasi-stable states rather than gentle slides.³

Without the complicating factor of human greenhouse gas emissions, the default expectation for the next million years would be more of the same: a continuing series of glaciations punctuated by warmer interglacials, all under the broad Icehouse state Earth has occupied for the last few million years. Whether human activity has pushed the climate system past a threshold into a different state is one of the genuinely open questions, and it depends heavily on how much additional carbon enters the atmosphere in the coming centuries.

The Long Tail of Human Carbon Emissions

Even if all fossil fuel burning stopped tomorrow, the carbon dioxide already in the atmosphere would not simply wash out. The ocean absorbs a large share within centuries, but the final cleanup depends on silicate weathering: the slow chemical reaction between CO₂-rich rainwater and exposed rock, which converts atmospheric carbon into dissolved minerals that eventually become marine carbonate sediments. The timescale for that process to draw atmospheric CO₂ back toward its pre-perturbation level has been estimated at roughly 240,000 years, with a plausible range of 170,000 to 380,000 years.⁴

That means at the million-year mark, silicate weathering would have had time to complete its work two to four times over. Atmospheric CO₂ from our era would be largely reabsorbed. But “largely” is not “entirely.” The same modeling work found that about 10 percent of the peak CO₂ anomaly and roughly 21 percent of the peak temperature anomaly persist on the silicate-weathering timescale.⁴ A geochemist examining ocean sediments laid down a million years from now could, in principle, detect a thin anomalous carbon layer corresponding to our century, much as we detect the carbon spikes from ancient volcanic episodes today. Whether the cumulative warming was enough to delay or skip one or more of those expected glacial cycles is an area of active research, but the plausible answer is yes for at least the next few hundred thousand years.

Reshaped Coastlines and Landscapes

Even holding sea level constant, coastlines change. Rivers deposit sediment, waves carve cliffs, and deltas build and retreat. Over a million years, the Mississippi Delta would have shifted substantially, the Nile Delta would look different, and low-lying islands built from coral would have grown, shrunk, or migrated depending on whether the ocean was in a warm or cool phase at any given time. During glacial periods, the English Channel, the Bering Strait, and much of the shallow continental shelves around Southeast Asia would dry out as sea levels dropped, reconnecting landmasses that are currently separated by water. During interglacials, those connections would flood again.

Mountains erode over this timescale, but not as dramatically as you might think. The Himalayas, Alps, and Andes are all still being pushed upward by tectonic forces that roughly balance erosion. Computational work on mountain lifespan has shown that the interplay between river incision and landsliding creates a complex feedback: rivers cut deeper valleys, which steepen hillslopes, which trigger more landslides, which in turn deliver rock debris that can either armor the riverbed and slow erosion or provide abrasive sediment that speeds it up.⁵ Tectonically active ranges would still stand tall after a million years. Older, tectonically quiet ranges, like the Appalachians, would wear down further but still be recognizable as uplands.

The biggest landscape changes would come from glaciation itself. Each time ice sheets advanced across Canada or Scandinavia, they would scrape and reshape the terrain underneath, carving new lake basins, rerouting rivers, and depositing moraines. The Great Lakes region of North America was essentially sculpted by the last few glacial cycles, and future cycles would continue that remodeling.

What Would Be Living There

A million years is a significant stretch for evolution. To get a sense of the pace, consider that our own species, Homo sapiens, has existed for roughly 300,000 years. The genus Homo has been around for about two million. Over a million-year window, new species arise, existing ones go extinct, and ecosystems reorganize. The fossil record from the last several million years of the Pleistocene shows multiple turnovers in mammal faunas as ice ages shuffled habitats back and forth across continents.

One fascinating wrinkle is what would happen to domesticated species if human management disappeared. Dogs, cats, cattle, sheep, chickens, and crop plants have been shaped by thousands of years of artificial selection. Research on feralization, the process by which domesticated organisms return to living wild, shows that it is not simply a reversal of domestication. Feral populations face novel selection pressures and carry complex genetic histories that include both wild and domestic ancestry.⁶ Modern feral dogs in parts of the world already show traits that differ from both wolves and pet dogs. Over a million years, the descendants of today’s domesticated animals would have diverged into lineages that look nothing like their ancestors on either the domestic or wild side of the family tree.

We know that recovery from even catastrophic extinction events, while slow, does happen. After the end-Permian mass extinction roughly 252 million years ago, when over 90 percent of species were lost, ecosystems rebuilt themselves over the following 8 to 9 million years, progressing stepwise from simple, low-level food webs to complex ecosystems with large predators and diverse prey.⁷ That recovery also produced entirely new groups, including marine reptiles and the lineage that eventually led to dinosaurs. A million years is far shorter than a full post-extinction recovery, but it is more than enough time for substantial species turnover within existing ecosystems, especially under the pressure of rapidly shifting climates.

Volcanic Eruptions and Asteroid Strikes

Over a million-year window, the odds of at least one very large volcanic eruption are high. Explosive super-eruptions, the kind that create enormous calderas and blanket continent-sized areas in ash, are rare on a human timescale but expected on geological ones. These events devastate large regions with pyroclastic flows, and the sulfur gases they inject into the stratosphere can linger for years, reducing surface temperatures and disrupting weather patterns globally. Researchers describe this as a “volcanic winter,” though the consensus is that even super-eruptions would produce temporary climate disruptions rather than permanent shifts in Earth’s climate regime.⁸ The Toba eruption around 74,000 years ago may have caused a multi-year global cooling event, but Earth’s climate recovered within a few thousand years.

Asteroid impacts are another wildcard. Small impacts happen constantly and pose no global threat, but a strike from a body several hundred meters across could cause regional devastation. One specific case study is illustrative: orbital calculations for the 500-meter asteroid Itokawa suggest that an Earth impact from this particular object is likely within a million years, though even that probability is only about four times higher than the average background impact rate for asteroids of that size.⁹ A 500-meter impactor would be devastating regionally and could trigger short-term climate cooling from dust and aerosols, but it would be far smaller than the roughly 10-kilometer asteroid that ended the Cretaceous period. Earth would recover from such an event within centuries to millennia.

Earth’s magnetic field adds another dimension. The field reverses polarity at irregular intervals, and a million-year span would almost certainly include at least one reversal. During the transition, which can take thousands of years, the magnetic field weakens and becomes disorganized, reducing the shielding that protects the surface from solar and cosmic radiation. Whether this has measurable effects on the biosphere is still debated, but researchers have flagged it as an area where geophysics and ecology intersect in ways that are not yet well understood.¹⁰

The Sun, for Its Part, Would Barely Change

The Sun is getting brighter over time as hydrogen fusion gradually increases its core temperature, but the change over a mere million years is negligible. Modeling of Earth’s climate response to a brightening Sun found that the climate warms roughly linearly for solar constant increases of up to about 10 percent, with an abrupt shift to much hotter conditions appearing around a 12.5 percent increase. Conditions that would cause significant water loss to space do not appear until the solar constant rises by about 19 percent.¹¹ The Sun’s luminosity increases by roughly 1 percent every 100 million years, so in a million years the increase would be on the order of 0.01 percent, an amount far too small to matter for Earth’s climate. The Sun is effectively a constant over the timescale we are discussing.

The Fossil Record of Civilization

If future geologists, human or otherwise, excavated the rocks laid down in our era, what would they find? Quite a lot, it turns out. Humans produce artifacts from materials that are either extremely rare in nature in their pure form, like metallic aluminum and titanium, or entirely unknown naturally, like plastics, boron nitride, and tungsten carbide. Global plastic production alone runs to hundreds of millions of tonnes per year, and total aluminum production since the mid-twentieth century exceeds 500 million tonnes.¹² These materials escape into both marine and terrestrial sediments as litter, microplastics, and industrial residues, becoming what researchers call “technofossils”: human-made objects that function as time markers in the geological record, just as trilobites or ammonites mark older periods.

The densest concentrations would be found in the remnants of landfills and urban centers, but escaped materials are widespread enough in ocean sediments and soils to form a globally recognizable layer. Plastics like polyethylene and polypropylene would degrade substantially over a million years, but their chemical signatures, particularly their carbon isotope ratios, would persist in the sedimentary record. Metals like stainless steel would corrode but leave characteristic mineral assemblages. Glass, ceramics, and some synthetic minerals could survive more or less intact, compressed into thin strata.

One especially durable signature would come from nuclear waste. Germany’s repository-siting process, for instance, explicitly requires that a geological disposal site ensure safety for at least one million years.¹³ The fact that engineers are designing containment systems on a million-year timescale tells you something about how long certain radioactive isotopes remain hazardous. Whether those repositories hold or eventually leak, the concentrated deposits of heavy elements like plutonium and cesium would stand out in the rock record as sharply as an iridium layer from an asteroid impact.

Would Humans Still Be Around?

This is the question everyone really wants answered, and it is the hardest to address with evidence. No model can predict the trajectory of a technological civilization over a million years. What we can say is that the typical mammalian species lasts roughly one to two million years before going extinct or evolving into something different enough to be classified as a new species. By that metric, Homo sapiens is a young species with statistical room to persist, but a million years from now we would likely either be extinct, substantially evolved, or technologically transformed in ways that make the word “human” a stretch.

If humans vanished tomorrow, the planet would not stay empty for long in evolutionary terms. Ecosystems are resilient on geological timescales. The end-Permian example showed that even after losing over 90 percent of species, life rebuilt complex food webs within about 8 to 9 million years, and some groups diversified within just 1 to 3 million years.⁷ A million years after human disappearance, the planet would host a full and diverse biosphere, with novel species filling every ecological niche, including many of the niches we currently dominate. Our cities would be buried geology. Our plastics would be trace chemistry in sedimentary rock. And somewhere in the strata, a thin, strange layer of aluminum, concrete, and synthetic minerals would mark the brief moment when one species reshaped the entire planet.