How Was the Moon Formed? The Giant Impact and the Evidence for It

Capture, in which a wandering body was caught by Earth's gravity, needs a way of slowing it down, which gravity alone does not provide, and a captured moon would be expected to have a different composition from Earth. Fission, proposed by Charles Darwin's son George in 1879, has a molten early Earth spinning so fast that a lump of it flew off, leaving the Pacific basin as the scar; it requires the Earth to have spun once every two and a half hours, and there is no way to get rid of that spin afterwards. Co-formation, the two bodies condensing together from the same material, predicts a Moon with an iron core like the Earth's, and the Moon has almost none.

The three ideas were still competing in July 1969 when Apollo 11 brought back the first rocks, and the rocks were the first real evidence. Over six landings the astronauts collected 382 kilograms of them, and what the samples showed ruled all three out.

The lunar samples told four things at once:

• •The Moon is dry: it had almost no water and almost none of the volatile elements, such as sodium and potassium, that boil off at moderate temperatures, as if the material had been strongly heated
• •The Moon is iron-poor: its density and its tiny core show that it is made mostly of the kind of rock found in the Earth's mantle, not of a whole planet's worth of material
• •The Moon was once molten to great depth: the pale highland crust is a rock that forms by floating on a global ocean of magma
• •The Moon is made of Earth: the ratios of oxygen isotopes in lunar rock match Earth's exactly, while every meteorite from elsewhere in the solar system differs

In the mid-1970s two papers proposed the same idea independently: that late in the Earth's formation, when it was nearly full size, a body about the mass of Mars struck it a glancing blow. The impactor's iron core sank into the Earth and merged with its own, which explains why the Moon has so little iron. The rocky mantles of both bodies were partly vaporised and thrown into orbit as a disc of hot debris, which explains the missing volatiles and the magma ocean, and the disc coalesced into the Moon within a few years or decades. Computer simulations from the 1980s onward showed that the collision could produce a Moon of the right mass in the right orbit, and the hypothetical impactor was named Theia, after the Greek goddess who was mother of the Moon.

The impact also explains why the Earth spins as fast as it does though the tilt of the Moon's orbit remains harder for simple versions of the model to explain, and it gives a date. Lunar rocks and the decay of radioactive elements in meteorites put the event at about 4.5 billion years ago, some 30 to 100 million years after the solar system began to form, and the Moon has been drifting away from the Earth ever since at, currently, 3.8 centimetres a year.

The awkward part is the isotopes. In the simplest simulations the Moon is made mostly of Theia, not Earth, and Theia, having formed elsewhere in the solar system, should have had different isotope ratios, as every other body does. Yet the Moon's oxygen, titanium and tungsten match Earth's to within measurement error. Several solutions are being tested: that Theia happened to form near enough to Earth to share its composition; that the impact was so energetic that both bodies vaporised and mixed thoroughly before the Moon condensed; that the young Earth was spinning very fast, so that a smaller impactor could throw off mostly Earth material; or that there were two impacts rather than one. Seismic images of two continent-sized blobs of dense rock deep in the Earth's mantle have been proposed, since 2023, as the buried remains of Theia itself.

The Moon is why the Earth is habitable in the way it is. Its pull raises the tides, which may have helped life move from sea to land, and its mass steadies the tilt of the Earth's axis, which would otherwise wander chaotically over millions of years and swing the climate with it, as the axis of Mars does. The giant impact also stripped the early Earth of its first atmosphere and reset the planet's clock, and the Moon's cratered surface, unweathered since, is the best record of the bombardment the young Earth went through. The samples that decided the argument are still being studied; a vial sealed on the Moon by Apollo 17 in 1972 was opened for the first time in 2022, and the rocks China returned from the far side in 2024 are the first from there.

The Moon turns once on its axis in exactly the time it takes to orbit the Earth, so the same hemisphere always faces us. This is not a coincidence but the work of tides. Earth's gravity stretched the young Moon slightly out of round, and while the Moon spun faster than it orbited, the pull on that bulge acted as a brake, slowing the spin until the bulge stayed pointed at Earth. This state, called tidal locking, is common among moons. The far side is not permanently dark; it gets as much sunlight as the near side.

The same process is working on the Earth, only more slowly because the Earth is so much heavier. The Moon's pull on the ocean tides drags against the planet's rotation, lengthening the day by a tiny fraction of a second each century and handing that spin to the Moon, which is why it drifts outward. The evidence is written in rock: fossil corals from around 400 million years ago show daily growth bands suggesting a year of roughly 400 days, so each day was shorter. Records of ancient eclipses, whose visible paths depend on Earth's rotation, confirm the slowing.

The Apollo and Luna samples did more than settle the Moon's origin; they let scientists date surfaces they will never visit. The idea is simple. A surface collects impact craters over time, so an older surface has more of them per square kilometre. The pale highlands are saturated with craters, while the dark maria, lava plains that flooded huge basins, have far fewer. Once returned rocks from specific landing sites had been dated in the laboratory, those crater densities could be matched to real ages, turning a crater count anywhere on the Moon into an estimate of age.

That calibration is now used across the solar system. Planetary scientists estimate the ages of surfaces on Mars and Mercury by counting their craters and adjusting for differences in impact rates, all leaning on the lunar timeline. New samples sharpen it. China's Chang'e 5 mission in 2020 brought back basalts from a young-looking region that turned out to be about two billion years old, younger than any lava collected by Apollo. That showed volcanic activity on the Moon lasted longer than many models allowed, and it filled a gap in the middle of the crater-age curve.

The 3.8 centimetres a year is a measurement, and the equipment that makes it is still on the Moon. Apollo 11, 14 and 15 left panels of reflectors there, and two Soviet Lunokhod rovers carried similar French-built ones. Observatories fire short laser pulses at them and time the round trip, about two and a half seconds, which gives the distance to within millimetres. So few photons make it back from each pulse that the work needs large telescopes and very sensitive detectors, and the same data have been used to test Einstein's theory of gravity and to show that the Moon probably has a partly liquid core.

The measurement also raised a puzzle. Tidal forces weaken steeply with distance, so a closer Moon would have been pushed away much faster. Run today's rate of tidal braking backwards and the Moon would have been almost touching the Earth only about one and a half billion years ago, which cannot be right for a body three times that age. The usual explanation is that the present rate is unusually high: the shapes of today's ocean basins, especially the Atlantic, happen to resonate with the tides and waste their energy efficiently, and with the continents arranged differently in the past, the braking was weaker.

For all of history people knew only one face of the Moon, and the first look at the other came in October 1959, when the Soviet probe Luna 3 photographed it. The result was a surprise. The near side is blotched with the dark maria, lava plains that cover around a third of its surface, but the far side is almost all bright, heavily cratered highland, with maria covering only a few percent. Later missions found that the crust on the far side is also noticeably thicker, which would have made it harder for molten rock from below to reach the surface there.

Why the two halves differ is still debated. One proposal ties it to the Moon's violent birth: soon after it formed, the Moon was much closer to a young Earth that was still glowing hot, and the side facing it stayed warmer for longer, so crust built up more thickly on the cooler far side. Another suggests that a second, smaller moon formed from the same debris disc and later collided gently with the larger one, plastering extra material onto what is now the far side. Others point to the heat-producing elements concentrated under the near side, or to the vast ancient impact basin near the south pole.

The Moon formed about 4.5 billion years ago when a Mars-sized body struck the young Earth, merging its iron core with Earth's and throwing a disc of vaporised mantle rock into orbit that gathered into the Moon. That explains why the Moon is dry, iron-poor, once molten and chemically a twin of the Earth's mantle; why it is so close a twin is the part still being worked out.