Did a supervolcano nearly wipe out humanity?
Probably not.
About 74,000 years ago the Toba volcano in Sumatra erupted on a scale that dwarfs anything in recorded history, and a famous idea says it crashed the human population to perhaps only a few thousand people. That idea made a set of predictions. Researchers checked them. The evidence doesn’t support a global crash.
The click
Where the Toba crash should show up
If the eruption nearly ended us, three records should change at the orange line. Do they?
Why people believed it
Because the eruption really was monstrous. On the latest central estimate, Toba threw out roughly 5,300 cubic kilometres of rock, about 130 times the 1815 Tambora eruption, which already caused a “year without a summer.” (The size is uncertain: the latest range is about 50 to 150 times Tambora, and older estimates were lower, closer to 40 times.) Its ash is modelled to have fallen more than 5 mm thick over about 40 million square kilometres, and glass shards from it turn up in South African sediment roughly 9,000 km away.
An event that big feels like it must have been a catastrophe. In 1998 anthropologist Stanley Ambrose proposed that the volcanic winter it caused could have decimated human populations, leaving them small enough to split into genetically distinct groups. Popular retellings hardened that into a near-extinction, with survivors usually put at somewhere between about 3,000 and 10,000 people.
How it was checked
If Toba nearly ended us, three things should show up. Looking for them is how the idea was tested:
- People should disappear or change abruptly after the ash. They don’t. At Pinnacle Point in South Africa and at Jwalapuram in India, people were living and making tools both below and above the ash layer, with continuity across it.
- The climate should turn harsh where people lived. In East Africa, Lake Malawi sediment shows no major change in composition and no evidence of substantial temperature change.
- Our genes should show a population crash at about 74,000 years ago. Whole-genome studies do find dips in human numbers, but at other times: a possible decline roughly 150,000 to 130,000 years ago and a founder effect about 50,000 years ago. Not at Toba.
Three predictions, none confirmed. That’s the answer: the global-crash version of the Toba catastrophe hypothesis is no longer supported by the evidence.
What this doesn’t say
- “Nothing happened” is not the claim. Conditions may have been severe where ash fell thickest. At Jwalapuram, a 2025 study of seasonal ash layers reports about a year of muted cooling followed by roughly four years of warming and drier conditions. That is the authors’ interpretation, and they note they found no direct evidence of the El Niño mechanism they suggest behind it.
- Survivors leave the records. Archaeological sites show where people were; they can’t show who wasn’t there. The evidence rules out a global crash far better than it rules out local suffering.
- The big cooling numbers are models, not measurements. Climate simulations suggest severe cooling in places like Europe, North America and central Asia but a muted response in the Southern Hemisphere, and how much cooling they produce depends heavily on unknowns such as the amount of sulfur released and the size of the aerosol droplets.
- “Supervolcano” describes the past, not a countdown. Four giant eruptions in 1.2 million years is too few to define a schedule.
Go deeper
That’s the click. Everything else in the source article (exactly how the ash was dated, the glass chemistry, the sulfur isotopes in ice cores, the modelling debates) is real and interesting, and it lives here:
NatureBlogger, “Toba: Following the Ash from Sumatra to South Africa”. It carries the dating, the volume estimates, the climate modelling, and the full citations.