Most meteors are grains of sand. They light up around 100 km and are gone in two seconds. By laboratory standards the air up there is a vacuum. This lesson answers how a grain of sand in a vacuum manages to be bright enough to light a whole street.
A grain of sand glowing in something close to vacuum
This lesson starts with the thin air 100 km up and works out what actually heats a meteor, what is actually glowing, and what it takes for anything to reach the ground.
Three names, one rock
Out in space it is a meteoroid, the streak of light it makes tearing into the atmosphere is a meteor, and whatever survives the trip and lands is a meteorite.
The middle one gets misread. A meteor is an event rather than an object, and the light is not the rock glowing. The rock is usually a grain of sand, invisible to the eye from 100 km away. The light is far larger than the rock.
Friction points at the wrong heat source
Meteors mostly light up between 80 and 120 km. The air density there is about a millionth of sea level, which by laboratory standards is a vacuum. A grain of sand in a vacuum lights up a whole street.
The trouble with friction is not whether there is enough air. It is that friction makes you picture heat being rubbed out of the surface. The molecules in your room are moving too, averaging 500 m/s, faster than an airliner. They do nothing to you because each is far too light and you have no speed relative to them.
A meteoroid does have speed relative to them, a great deal of it. Going through at tens of kilometres per second means every air molecule slams into it at that speed. Energy goes as the square of speed, so a hundredfold difference in speed is a ten-thousandfold difference in what one molecule carries. A blow like that does not merely push a surface atom aside; the atom leaves.
This is where the heat comes from. The meteoroid unloads its kinetic energy into the molecules it runs into, and those molecules carry it back to the surface. Thin air only makes that happen less often, and the force of each hit is unchanged.
The rock is not what glows
Past roughly 2,000 K the surface material goes straight from solid to gas. That is ablation, not combustion, since there is almost no oxygen up there and the rock is being vaporised.
What comes off is the metal inside: sodium, magnesium, iron and calcium. Those atoms hit the surrounding air at speed, electrons are kicked up and fall back, and specific wavelengths come out. That cloud of metal vapour and excited air is what we see.
Meteors therefore have colour, mapped onto elements. Sodium runs yellow, magnesium blue-green, iron yellow, calcium violet, atmospheric oxygen green. Fast ones often leave a persistent train hanging in the sky, the afterglow of plasma recombining.
Fireballs, and why they burst in mid-air
A meteor brighter than Venus is a fireball. Big ones usually break up in the air. The pressure difference across the body keeps rising as it falls, and most asteroid fragments carry cracks rather than being solid. The cracks are forced open, the body shatters, total surface area jumps, and the remaining energy comes out in a very short time. It looks like an explosion.
The 2013 Chelyabinsk object was about 20 metres and broke up near 30 km. It punched no crater at all, and the injuries came from glass shattered by the shock. That is the difference between an airburst and an impact, with the energy released while still in the air.
A fireball is worth writing down: the time, the bearing and elevation where it appeared and vanished, what you would compare its brightness to, whether it left a train, and whether you heard anything. Those together are what allows an orbit and a landing zone to be worked back.
Why there are more after midnight
Earth moves along its orbit at about 30 km/s and rotation carries you around it. The pre-dawn side faces the direction Earth is heading, so you are standing on the windscreen, and the evening side is the rear window.
On the windscreen, head-on arrivals show up and slower meteoroids that could never catch you get caught by Earth instead. The second half of the night therefore carries more meteors, and faster ones on average.
Showers take their name from the constellation their radiant sits in, so a radiant in Perseus gives the Perseids. The radiant itself is perspective, since those tracks are parallel and only meet at a point once projected onto the sky, the way railway tracks meet in the distance.
The night of November 1833
From late on 12 November into the dawn of the 13th, the eastern half of North America rained meteors all night. Later estimates put it in the tens of thousands per hour. People thought the world was ending, and newspapers ran eyewitness letters for weeks.
Denison Olmsted at Yale asked people across the region to write down which direction the meteors came from. Nobody much had thought to do that. When the records came back and were laid over each other, every track traced to the same spot in Leo, and that spot moved with the stars all night rather than with the ground.
If meteors were weather, that crossing point should have stayed fixed over some compass bearing on the ground. It did not. Olmsted concluded that what was falling came from beyond the Earth, and that the Earth was passing through something out in space.
Meteors became a subject for astronomy from then on. Working out what actually lights up a rock falling in from space took a good deal longer.
Lab: what is doing the heating
The three sliders are altitude, speed and diameter. The right-hand panel swaps to a different picture as you move them, because what happens in front of a meteoroid is not the same thing at different sizes and altitudes.
This lab computes orders of magnitude, not forecasts. Real meteor modelling integrates mass loss and deceleration together and needs the material's heat of ablation, and only the part that settles which picture applies is kept here. Above 86 km the air density comes from the standard atmosphere's upper extension, where real values swing with solar activity and time of day, and an order of magnitude either way is normal.
Start with these
Two millimetres, 59 km/s, 100 km up. In the right-hand panel the molecules arrive one at a time and there is no shock in front. Almost every meteor is in this state.
Same size, a little over half the speed, so it has to fall further before it lights up. Impact energy grows with the square of speed, and slowing down has to be made up for by thicker air.
Forty centimetres. At this size the meteoroid is much larger than the spacing between molecules, and a layer of compressed air starts building in front of it. This is where compression heating genuinely applies.
Twenty metres, fully into continuum flow, with a clear bow shock. The 2013 object broke up in this state, and the shock reaching the ground shattered a city's windows.
Meteoroid, meteor, meteorite: one rock, three lives
While it drifts in space it is a meteoroid, the streak of light it makes tearing into the atmosphere is a meteor, and whatever survives the trip and reaches the ground is a meteorite. Three stages of the same object.
The middle one gets misread. A meteor is an event rather than an object, and the light you see is not the rock glowing. By then the rock is usually the size of a grain of sand and invisible to the eye from 100 km away. The light is far larger than the rock, and its shape has nothing to do with it.
Friction is not what the textbooks should have said
Meteors mostly light up between 80 and 120 km. The air density there is about a millionth of sea level, which by laboratory standards is a vacuum. A grain of sand in a vacuum lights up a whole street.
The trouble with friction is not whether there is enough air. It is that friction makes you picture heat being rubbed out of the surface. The air molecules in your room are moving too, averaging 500 m/s, faster than an airliner. They do nothing to you because each one is far too light and because you have no speed relative to them.
A meteoroid does have speed relative to them, a great deal of it. Going through at tens of kilometres per second means every air molecule slams into it at that speed. Energy goes as the square of speed, so a hundredfold difference in speed is a ten-thousandfold difference in what one molecule carries. A blow like that does not merely push a surface atom aside; the atom leaves.
This is where the heat comes from. The meteoroid unloads its kinetic energy into the molecules it runs into, and those molecules carry it back to the surface. Thin air only makes that happen less often, and the force of each hit is unchanged. Air density therefore decides how high a meteor starts glowing, not whether it glows.
So what is actually glowing
Once the surface is driven past 2,000 K the material goes straight from solid to gas, a process called ablation. It has nothing to do with combustion: there is almost no oxygen at that altitude and no oxidation reaction driving it. The rock is being vaporised.
What comes off is the metal inside the meteoroid, sodium, magnesium, iron and calcium. Those atoms hit the surrounding air at tens of kilometres per second, their electrons are kicked to higher levels, and light of specific wavelengths comes out as the electrons fall back.
That cloud of metal vapour and excited air is what we see. Meteors therefore have colour, and the colour maps onto elements. Sodium runs yellow, magnesium blue-green, iron yellow, calcium violet, and atmospheric oxygen green. Fast meteors often leave a persistent train, the afterglow of plasma recombining and reacting, which can hang in the sky for seconds and occasionally minutes.
Compute the temperature of the compressed air in front of a meteor with school thermodynamics and you get a figure like 200,000 K. Real air begins coming apart and losing electrons in the low thousands, and those processes eat the energy, so a real shock layer only reaches somewhere over ten thousand. The same reason is why spacecraft re-entry cannot be estimated with the textbook formula.
Almost nothing reaches the ground
The overwhelming majority of material entering the atmosphere each day is dust-sized and vaporises completely near 100 km. Most of what you watch on a summer night started smaller than a pea.
Surviving takes size, toughness and ideally a low speed. Iron meteorites resist ablation best and stony ones tend to break up high. Extra speed hurts rather than helps, since the ablation rate climbs faster with speed than the mass can absorb.
Whatever survives goes through a stage called dark flight. By around 20 km the air has already dragged it below cosmic speed, the light goes out, and the rest of the trip is ordinary free fall. A meteorite therefore lands merely warm rather than red hot, and it does not punch the crater people imagine, unless it was tens of metres across to begin with.
Fireballs
A meteor brighter than Venus, around magnitude −4, is a fireball. One ending in a bursting flash is conventionally a bolide, and extreme cases past magnitude −17 are sometimes called superbolides. These boundaries are all observing conventions with no physical dividing line behind them.
Big ones often blow up in mid-air. The pressure difference between front and back keeps rising as the object falls, and most asteroid fragments already carry cracks rather than being solid through. The pressure forces the cracks open, the parent body shatters into many pieces, total surface area jumps, the ablation rate jumps with it, and nearly all the remaining kinetic energy comes out in a very short time. It looks like an explosion.
A fireball is worth writing down: the time to the minute, the bearing and elevation where it appeared and vanished, what you would compare its brightness to, whether it left a train, and whether you heard anything and how long afterwards. Those details together are what lets professional networks work back to an orbit and a landing zone.
Fireballs worth knowing about
Events that were witnessed, that left material behind, or that changed how this was understood.
Lab: why there are more after midnight
Earth moves along its orbit at about 30 km/s while rotation carries you around it. Sometimes you face the direction Earth is heading and sometimes you face away. Drag the time and see which side you are on.
The pre-dawn side faces the direction Earth is heading, so you are standing on the windscreen. Head-on arrivals show up, and slower meteoroids that could never catch you get caught by Earth instead, which is why counts climb noticeably after midnight. A head-on meeting tops out near 73 km/s and the Leonids are that kind; something overtaken from behind comes in far slower, around 17 km/s.
The evening side faces away from Earth's heading, so you are on the rear window. Only meteoroids fast enough to catch up with Earth get in, and the counts are naturally low. Seeing more of them means waiting until after midnight.
A radiant at the zenith shows the most, and down near the horizon most of the tracks hide below it. At this altitude only about 82% is left. The ZHR quoted in shower tables assumes the radiant overhead under a sky dark enough for sixth-magnitude stars, so what you count is always less.
Sporadic meteors and meteor showers
Everything above concerns sporadic meteors, arriving at random. There are a few an hour on any night. Showers work differently, with certain dates carrying far more, and the same dates coming round every year.
Comets are the reason. Each time a comet nears the Sun its surface ice sublimates and carries dust out with it. That dust keeps travelling roughly along the comet's orbit and over time spreads into a stream circling the Sun. Where Earth's orbit crosses such a stream we run into a batch every year at the same point. The dates are fixed because the date Earth returns to a given point on its orbit is fixed.
Not every parent is a comet. The Geminids come from 3200 Phaethon, an asteroid that sheds material, and the Quadrantids' 2003 EH1 is generally taken to be a burnt-out comet nucleus. The name comes from the constellation the radiant sits in, so a radiant in Perseus gives the Perseids. Where one constellation hosts several, a nearby bright star separates them, as with the Eta Aquariids and the Delta Aquariids. The Quadrantids are the odd one out, since the constellation Quadrans Muralis was abolished in 1922 and the radiant now counts as being in Boötes, though the name stayed.
Want to know when to watch this year
This lesson covers the mechanism. For which night each shower peaks and whether the Moon will be in the way, there is a year-by-year table.
Fireballs worth knowing about
Event
Date
Why it matters
The Leonid storm
1833-11-13
An all-night storm over North America. The radiant stayed fixed against the stars rather than the ground, the first strong evidence that meteors come from beyond Earth.
Sikhote-Alin iron fall
1947-02-12
A witnessed iron meteorite shower in the Russian Far East with a large recovered mass, a classic sample for studying how a fall scatters.
Jilin meteorite shower
1976-03-08
In Jilin, China. Yielded the largest known single stony meteorite, across a strewn field of several hundred square kilometres.
Peekskill fireball
1992-10-09
Filmed simultaneously by camcorders at several games across the eastern United States, one of the few cases with an orbit recovered from video. A fragment hit a parked car.
2008 TC3
2008-10-07
The first object ever detected before impact, with its fall predicted and samples then recovered, from the Sudanese desert.
Chelyabinsk
2013-02-15
A roughly 20-metre asteroid broke up near 30 km. The shock shattered windows and injured over a thousand people. The best-documented such event, thanks to dashcams.
FAQ
Do meteors really burn up from friction with the air?
A meteoroid does have speed relative to them, a great deal of it. Going through at tens of kilometres per second means every air molecule slams into it at that speed. Energy goes as the square of speed, so a hundredfold difference in speed is a ten-thousandfold difference in what one molecule carries. A blow like that does not merely push a surface atom aside; the atom leaves. This is where the heat comes from. The meteoroid unloads its kinetic energy into the molecules it runs into, and those molecules carry it back to the surface. Thin air only makes that happen less often, and the force of each hit is unchanged. Air density therefore decides how high a meteor starts glowing, not whether it glows.
Is the glowing part the rock itself?
Once the surface is driven past 2,000 K the material goes straight from solid to gas, a process called ablation. It has nothing to do with combustion: there is almost no oxygen at that altitude and no oxidation reaction driving it. The rock is being vaporised. What comes off is the metal inside the meteoroid, sodium, magnesium, iron and calcium. Those atoms hit the surrounding air at tens of kilometres per second, their electrons are kicked to higher levels, and light of specific wavelengths comes out as the electrons fall back.
Why do meteors come in different colours?
That cloud of metal vapour and excited air is what we see. Meteors therefore have colour, and the colour maps onto elements. Sodium runs yellow, magnesium blue-green, iron yellow, calcium violet, and atmospheric oxygen green. Fast meteors often leave a persistent train, the afterglow of plasma recombining and reacting, which can hang in the sky for seconds and occasionally minutes.
The lab's arithmetic is an order-of-magnitude teaching estimate and should not be used for any forecast. Published energies, masses and casualty figures for historical events vary between sources, and the figures here follow official or journal sources, so anyone quoting them should go back to the original.