Transcript
On 30 June 1908, something exploded over the Podkamennaya Tunguska river in central Siberia — not on the ground, but in the air. The blast felled roughly 80 million trees across about 2,000 square kilometres of remote boreal forest, an area larger than greater London, stripped bare in seconds by a single atmospheric explosion. Seismic instruments detected the event as far away as western Europe. No crater was ever found. No significant meteorite fragments. No object. Just devastation pointing outward in all directions, and then silence.
That combination — enormous destruction, zero physical remnant — is the core puzzle of the Tunguska event. It is what made the event so hard to explain for so long, and what makes the explanation so instructive once you understand it. The question is not just what happened, but how science eventually figured it out, why it took so long, and what remains genuinely contested more than a century later.
Start with what the site actually looked like when the first scientific expedition arrived — and note that the expedition did not happen in 1908. It happened in 1927. Nineteen years after the explosion. Leonid Kulik, a geologist working for the Soviet Academy of Sciences, had attempted to reach the site as early as 1921, but the region's extreme remoteness defeated him. The reasons for the delay compound on each other: the area is extraordinarily isolated even by Siberian standards, and between 1908 and the mid-1920s, Russia passed through World War I, a revolution, and a civil war. These were not conditions suited to mounting remote scientific expeditions.
When Kulik finally reached the site in 1927, he found a landscape frozen in time. Trees were still down. And the pattern of how they had fallen told an immediate story. They were toppled radially — pointing away from a central point, like spokes of a wheel, uniformly across 15 to 30 kilometres in every direction from the epicentre. Walk toward the middle of that pattern and you find something even more revealing: trees still standing. Scorched, bark stripped, branches gone — but upright. Dead, but not felled.
That pattern is physically diagnostic. If a solid object had struck the ground, you would expect an impact crater and a very different debris field, with material thrown outward in ways shaped by the angle of approach and the nature of the ground. What Kulik found instead matches what fluid dynamics and atmospheric physics predict for a detonation high above the surface: a powerful downward and outward pressure wave that topples everything in the surrounding area, while directly below the explosion point, the blast is moving straight down. Trees there get the full force from above — scorched and stripped — but the lateral forces that would push them over are not dominant at the epicentre. An overhead explosion explains standing scorched trees. A ground impact does not.
Three lines of evidence converge on the same model: the radial tree-fall pattern, the scorched-but-standing trees at the very epicentre, and the complete absence of a crater or any significant meteorite fragments. Each of those facts alone might have other explanations. All three together point decisively to an airburst — an object that exploded in the atmosphere before reaching the ground. This is not a tentative conclusion. It is the scientific consensus, and Kulik's 1927 field data provided the foundation that every subsequent model has been built on.
That 19-year delay was not neglect. It was the result of war, revolution, and geography. But it matters, because Kulik's data — the radial tree-fall, the marshy bog where a crater might have been expected, the standing scorched trunks — became the dataset that all later calculations depend on. Without it, there would have been nothing to model.
Now, the impactor destroyed itself completely. That is the central physical fact. Whatever entered the atmosphere that morning vaporised before it reached the ground, dispersing its energy and gases into the air, leaving no recoverable fragment of any significant size. And that is precisely why, more than a century later, we still do not know for certain what it was made of.
The two main candidates are a stony asteroid and a comet. The debate is not purely academic. If the body was cometary, its structure and composition were fundamentally different from a rocky asteroid, and models of how it behaved in the atmosphere need to account for that. The distinction also matters for how we think about similar objects that might approach Earth in the future.
The case for a cometary origin rests partly on observations made in the days after the event. Unusually bright night skies were reported across Siberia and Europe in late June and early July 1908 — the kind of silvery, luminous nocturnal illumination associated with noctilucent clouds. Some researchers have pointed to this as possible evidence that cometary ice vaporised in the upper atmosphere, seeding the conditions for those clouds. A comet, after all, is largely ice and dust. If it disintegrated high enough, it might leave a trace in the atmosphere's upper layers.
But in 1998, astronomer Z. Sekanina published a peer-reviewed analysis directly addressing this question. His conclusion was unambiguous: hypotheses based on a cometary origin face, in his words, unsurmountable difficulties. Sekanina's argument is that a comet fragile enough to vaporise completely before reaching the ground would have begun disintegrating much higher in the atmosphere than the evidence suggests, and the physics of how it would have entered and traversed the atmosphere do not fit the observed pattern of destruction on the ground. That paper has been cited nearly 60 times in the subsequent literature — not a universal endorsement, but a measure of how seriously the argument has been engaged.
More recent modelling, informed by data from the 2013 Chelyabinsk event, favours a stony asteroid roughly 50 to 80 metres across, entering at approximately 55,000 kilometres per hour and exploding at an altitude of 10 to 14 kilometres. NASA's own summary describes the impactor as approximately 40 metres in diameter. The estimates vary slightly depending on the model, but the class of object is consistent: a mid-sized rocky body.
The honest qualifier here is important. No macroscopic fragment of the Tunguska object has ever been recovered. The compositional question cannot be settled by direct physical evidence, because there is none. The asteroidal case is stronger, but it rests on inference from the blast pattern and atmospheric physics, not on a piece of rock in a laboratory. The comet hypothesis has serious problems, but the debate is not entirely closed.
The energy released in the Tunguska explosion is estimated at somewhere between 3 and 15 megatons of TNT equivalent — that is the range most consistently cited across the key sources. Some older estimates pushed as high as 50 megatons, while more recent momentum-corrected models have pulled the number down toward the lower end of that range. Even so, 3 megatons is hundreds of times more powerful than the atomic bomb dropped on Hiroshima. The spread is not a rounding error. It reflects genuine methodological disagreement: different researchers using different physical models, different assumptions about the object's velocity and composition, and different interpretations of the seismic and barometric records from 1908. The number is unlikely ever to be pinned down more precisely, because the event left no direct physical trace to measure.
The 80 million trees across 2,000 square kilometres are worth holding in mind here. That destruction was the work of an explosion that left no ground scar — the energy was deposited into the atmosphere and converted into a pressure wave and thermal pulse. Every felled tree in that area registers what the energy did. The forest itself is a kind of indirect measurement.
There is also a minority strand of research that questions the entire airburst framework. A researcher named Ol'khovatov has argued that anomalous eyewitness accounts, possible tectonic activity on 30 June 1908, and complex meteorological conditions that day suggest the event may have had a different origin entirely — involving electromagnetic or tectonic factors rather than a cosmic object. This view is not mainstream, but it draws on a legitimate methodological concern: the earliest eyewitness testimony from the local Evenki people was collected in the 1920s and 1930s, and some researchers have argued that key testimonies may have been selectively edited or de-emphasised in early publications. How reliable those accounts are, and which of them were prioritised by Kulik and his contemporaries, is a question that affects all reconstructions that rely on witness evidence.
There is also the Lake Cheko question. Lake Cheko is a small body of water located about 8 kilometres from the estimated epicentre. A team of researchers proposed that it might be an impact crater formed by a fragment of the Tunguska object — meaning the event was not purely an airburst but included some surface impact. That hypothesis attracted significant attention. But it is described in the research literature as controversial. Dendrochronological analysis of trees around the lake has not produced clear evidence that the lake formed in 1908, and sediment studies have not settled the question. The wider scientific community remains sceptical. It is a genuinely open question, but not one that overturns the airburst model — even if a fragment did reach the ground, the dominant energy release was atmospheric.
To understand why Tunguska matters beyond its own explanation, place it alongside two other events: Comet Shoemaker-Levy 9 in 1994, and the Chelyabinsk meteor in 2013.
Shoemaker-Levy 9 struck Jupiter between 16 and 22 July 1994, breaking apart into more than 20 fragments that hit the planet in sequence. It was the first time humans directly observed a collision between two solar system bodies in real time. The impact scars on Jupiter were visible through amateur telescopes. The abstract possibility of cosmic impacts became, for the first time, viscerally concrete.
The Chelyabinsk event brought it closer still. On 15 February 2013, an asteroid 17 to 20 metres across — a fraction of the size of whatever hit Tunguska — entered the atmosphere over Russia and exploded. The energy release was roughly 500 kilotons. The immediate effect was a shockwave that shattered windows across a wide area. More than 1,200 people were injured — not by fire or by a crater, but by broken glass. An object that small, exploding over a populated region, sent over a thousand people to hospital. That is what Tunguska-scale means in practical terms: if an object one-quarter the diameter of the Tunguska impactor can cause that much harm, then a Tunguska-scale event over a city represents a civilisation-level threat.
Tunguska, Shoemaker-Levy 9, and Chelyabinsk together constitute the evidentiary basis for a formal planetary-defence programme. Each contributed something different: Tunguska established that large airbursts happen and can devastate enormous areas with no warning. Shoemaker-Levy 9 demonstrated that the solar system is still actively producing large-scale impacts. Chelyabinsk showed that even small objects cause significant harm, and that they can arrive entirely undetected.
That pattern — not a single scare, but a sequence of events — is what drove institutional change. NASA's Planetary Defense Coordination Office was formally established on 7 January 2016. Its mandate is to detect, track, and characterise near-Earth objects, and to coordinate a response if one is found to be on a collision course. In September 2022, NASA's DART spacecraft deliberately impacted an asteroid called Dimorphos. The collision shortened Dimorphos's orbital period by 33 minutes — demonstrating that kinetic deflection, hitting an asteroid hard enough to change its path, actually works in practice.
But the system still has gaps, and the Tunguska story is not finished. In 2023, an asteroid designated 2023 NT1 passed within one-quarter of the Earth-Moon distance. It was 30 to 60 metres across — roughly Tunguska-scale — and its potential impact energy was estimated at around 10 megatons. It was discovered two days after its closest approach. Two days after. An object that could have levelled an area the size of greater London was found only once it had already passed.
The detection gap for objects in that size range remains real. These are large enough to cause regional or urban-scale destruction, but small enough to be genuinely difficult to find in advance. Surveys that track larger objects are more complete, but Tunguska-class asteroids represent a category where warning time, if any, may be measured in hours or days rather than years.
So where does this leave us? The airburst model is solid. An object entered the atmosphere over the Podkamennaya Tunguska river on the morning of 30 June 1908 and destroyed itself in the air, releasing somewhere between 3 and 15 megatons of energy, felling 80 million trees across 2,000 square kilometres while leaving no crater and no recoverable fragment. That much is not seriously in dispute.
What remains open: the precise energy, which different methods estimate across a range too wide to treat as settled. The composition of the body — stony asteroid is the best-supported answer, but it is an answer derived from inference, not from a piece of the object itself. Lake Cheko's origin remains unresolved, with neither the impact hypothesis nor its rejection firmly established by the sediment and dendrochronological record. And the reliability of early eyewitness testimony, and what may have been lost or altered in its collection, is a methodological question that no subsequent analysis has fully answered.
What is not open is the significance of the event. A cosmic body capable of destroying a major city passed through the atmosphere undetected, left no warning, and was not scientifically investigated for nineteen years. Understanding it took the better part of a century. Building a formal response to the category of threat it represents took even longer. That work is ongoing — and 2023 NT1, found two days after it had already passed, is a precise measure of how much of it remains to be done.