Lunar Recession

The lag that lifts the Moon

Stand on a beach and the Moon's gravity is, very slightly, pulling the ocean toward it — and pulling the far side of the Earth away from it. That is the familiar two-bulge tide. The piece of the story almost everyone misses is timing. If Earth did not rotate, the tidal bulge would sit exactly on the Earth-Moon line and nothing interesting would happen. But Earth spins once every 24 hours, far faster than the Moon takes to crawl around its orbit in 27.3 days. Friction between the tidal flow and the ocean floor, plus the planet's own inability to deform instantly, means the bulge is carried ahead of the sub-lunar point. The high tide arrives early.

That small misalignment is the entire mechanism. The leading bulge is a lump of extra mass sitting slightly in front of the Moon. Its gravity pulls the Moon forward along its orbit — a tiny but relentless tangential tug. Adding energy and angular momentum to an orbiting body raises it to a larger orbit, so the Moon edges outward. By Newton's third law the Moon pulls back on the bulge, dragging against Earth's rotation and braking the spin. The result, distilled to one sentence: Earth is slowly handing its rotational angular momentum to the Moon, and the Moon spends that gift by climbing to a higher, slower orbit. We call the secular outward drift lunar recession.

Angular momentum is the conserved currency

The cleanest way to see why the Moon must move out is to track the conserved quantity. Ignoring the tiny torque the Sun exerts, the total angular momentum of the Earth-Moon system is fixed. It has two parts: Earth's spin and the Moon's orbit.

L_total = L_spin + L_orbit = I_Earth ω_Earth + m_Moon √(G M_Earth a)  =  constant

Here I_Earth ω_Earth is Earth's spin angular momentum, and the orbital term uses the specific angular momentum of a near-circular orbit, which grows as the square root of the semi-major axis a. Tidal friction siphons angular momentum out of the spin term, so ω_Earth falls — the day lengthens. To keep L_total constant, the orbital term must rise, which means a must increase. The Moon has to recede.

Energy, by contrast, is not conserved — and that is the point. The system continuously loses mechanical energy to frictional heat, mostly dissipated in shallow shelf seas. About 3.7 terawatts of tidal power is dumped on Earth, of which roughly 95 percent goes into the lunar tide. Crucially, raising the Moon to a higher orbit costs less energy than Earth's spin gives up; the difference is what is lost to heat. The system is sliding downhill in energy toward the minimum-energy, maximum-separation configuration consistent with its fixed angular momentum, which is a doubly synchronous state.

The torque and the tidal lead angle

The forward force on the Moon comes from the displaced bulge. If the bulge leads the Earth-Moon line by a small angle δ (the geometric lag angle, related to the dissipation parameter through the tidal quality factor Q as δ ≈ 1/(2Q)), the torque on the Moon scales as

N ≈ (3/2) (k₂/Q) (G m_Moon² R_Earth⁵) / a⁶

where k₂ is Earth's degree-2 Love number (how much the planet deforms, ≈ 0.30 for the solid body but effectively larger including the ocean), Q is the tidal quality factor (≈ 12 for the modern Earth — a remarkably lossy, low-Q system because of the oceans), R_Earth is Earth's radius, and a is the orbital distance. The steep a⁻⁶ dependence is decisive: the torque was vastly stronger when the Moon was close, and weakens rapidly as it recedes.

The recession rate follows from setting da/dt against this torque. Differentiating the orbital angular momentum gives

da/dt = 2N / (m_Moon √(G M_Earth / a))  ∝  a⁻¹¹ᐟ²

so da/dt falls even faster than the torque. Plugging in modern values reproduces the measured 3.8 cm/yr. Running the equation backward in time, the strong a⁻¹¹ᐟ² dependence means the rate climbs steeply as a shrinks — which is exactly why a naive linear extrapolation backward is so badly wrong (see the misconceptions below).

How we measure 3.8 cm per year

The recession is not inferred — it is directly ranged. The Apollo 11, 14 and 15 missions and the Soviet Lunokhod 1 and 2 rovers left corner-cube retroreflector arrays on the Moon. Observatories such as the Apache Point Lunar Laser-ranging Operation (APOLLO) fire short laser pulses at these arrays and time the round-trip of the returning photons.

d = (c × t_round-trip) / 2,   c = 299,792,458 m/s,   t ≈ 2.5 s

The round trip takes about 2.5 seconds. Modern Lunar Laser Ranging reaches millimetre-level precision on the one-way distance, so a 3.8 cm/yr drift accumulates to a clearly detectable signal within a single year and an unmistakable one over the half-century of data. The same data set tests the equivalence principle, the constancy of G, and geodetic precession to exquisite accuracy as a bonus. Independently, the lunar secular acceleration in longitude — the Moon appearing to run ahead of where a fixed-period model predicts — was first inferred by Edmond Halley in 1695 from ancient eclipse records, long before anyone understood tides were the cause.

The numbers, on one page

Quantity	Value	Note
Recession rate (semi-major axis)	3.82 ± 0.07 cm/yr	Lunar Laser Ranging
Mean Earth-Moon distance	384,400 km	Center to center
Day lengthening (lunar tidal)	~2.3 ms / century	Astronomical, from the torque
Day lengthening (observed long-term)	~1.7 ms / century	Tidal minus glacial rebound
Tidal power dissipated	~3.7 TW total, ~95% lunar	Mostly shallow seas
Earth tidal quality factor Q	≈ 12	Low Q = very lossy
Moon's orbital period	27.32 days (sidereal)	vs. 24 h Earth spin
Devonian day length	~22 hours	400 Myr ago, coral growth bands
Final locked day / month	~47 days each	Never reached; Sun dies first

One vivid way to feel the scale: 3.8 cm/yr is almost exactly the rate at which human fingernails grow. Over a human lifetime the Moon retreats by roughly three metres. Over the span of recorded human civilization, about two hundred metres.

The day grows longer — and corals remember

The recession of the Moon and the lengthening of the day are two faces of the same torque, so the past day length is a fossil record of past tidal braking. The astronomical tidal contribution slows Earth by about 2.3 ms per century. The observed slowdown is gentler, near 1.7 ms per century, because the solid Earth is still rebounding from the kilometres of ice that pressed it down during the last glaciation; that rebound makes the planet more spherical, lowers its moment of inertia, and — like a spinning skater pulling in their arms — speeds it back up, partly fighting the tides.

Integrate the slowdown over hundreds of millions of years and the day was dramatically shorter in the deep past. Remarkably, this is testable in the rock record. Some fossil corals and bivalves lay down a fine daily growth band modulated by an annual envelope; counting the daily bands per year reveals how many days the year held. Devonian corals from about 400 million years ago show roughly 400 days per year, implying a day of about 22 hours — in striking agreement with the integrated tidal slowdown. Tidal rhythmites, layered sediments that record the rise and fall of individual tides, extend this clock further back and confirm the Moon was closer and the day shorter.

Where tidal evolution shows up across the Solar System

The Earth-Moon pair is the textbook case, but tidal recession and its mirror image, tidal decay, sculpt satellite systems everywhere. The deciding factor is whether the moon orbits outside the synchronous radius (where orbital period equals the planet's spin) and prograde — in which case it recedes — or inside it, or retrograde — in which case it falls.

• Earth's Moon. Prograde, outside synchronous orbit, so it recedes at 3.8 cm/yr and brakes Earth's spin. The Moon is already tidally locked to Earth (one face always toward us); the Earth is not yet locked to the Moon, which is why recession continues.
• Phobos at Mars. The opposite case. Phobos orbits Mars in 7.7 hours, faster than Mars's 24.6-hour day, so the tidal bulge it raises lags behind it and tugs it backward. Phobos is spiralling inward at about 1.8 cm/yr and will be torn apart at the Roche limit or crash in roughly 30–50 million years.
• Triton at Neptune. Triton orbits retrograde, so tides relentlessly drag it inward over billions of years toward an eventual Roche-limit disruption that may build a new ring system.
• Io, Europa, Ganymede. Jupiter's tides push Io outward; the Laplace 1:2:4 mean-motion resonance locks the three moons so that this pumping forces Io's eccentricity, and the resulting tidal heating powers the most volcanically active body in the Solar System.
• Pluto-Charon. The endpoint of recession: this pair has already reached double synchronization. Both bodies keep the same face permanently toward each other, the day and the month are equal at 6.39 days, and the tidal torque has switched off. It is the future of the Earth-Moon system in miniature.

Deep time: where this ends

Lunar recession does not run forever. It halts when the bulge stops leading or lagging — that is, when Earth's rotation period equals the Moon's orbital period. In that doubly tidally locked state the Moon hangs motionless over one hemisphere, the same face of Earth always toward it, and the net tidal torque vanishes. Solving the angular-momentum bookkeeping for the Earth-Moon system gives a final day and month of about 47 days each, with the Moon roughly 1.6 times farther than today, reached in something like 50 billion years.

That future will not arrive. The Sun leaves the main sequence in about 5 billion years, swelling into a red giant whose envelope reaches near Earth's orbit; solar tides and mass loss will reorganize — and very possibly destroy — the Earth-Moon system long before the 47-day day. There is a further subtlety: solar tides act on Earth too, and once the lunar month nears the day length the Sun's torque can take over and start the Moon spiralling back in. The simple "Moon recedes to a stable 47-day lock" picture is the leading-order answer, not the final word.

Common misconceptions and edge cases

• "The Moon is escaping Earth." It is not. The Moon never gains escape velocity from this process; it merely climbs slowly until the torque shuts off at the synchronous lock. Recession is a redistribution of angular momentum within a bound system, not an ejection.
• "Extrapolate 3.8 cm/yr backward to find when the Moon formed." A classic creationist talking point and a classic error. Because da/dt scales as a⁻¹¹ᐟ², the rate was far slower when the Moon was farther out in the past and far faster when it was close. Today's rate is also anomalously high because the modern Atlantic sits near tidal resonance. Properly integrating the rate is fully consistent with a 4.5-billion-year-old Moon; the linear extrapolation that "puts the Moon at Earth's surface 1.5 billion years ago" simply uses the wrong physics.
• "The tidal bulge is two stationary lumps under the Moon." In reality the open-ocean tide is a system of rotating amphidromic waves steered by continents and the Coriolis force, not a simple ellipsoid. The clean "lead-angle bulge" is a useful idealization; the true geometry is why dissipation depends so strongly on the shape of ocean basins.
• "The day lengthens by exactly 2.3 ms per century, period." That is the lunar tidal contribution alone. The observed rate is smaller because glacial isostatic rebound is spinning Earth back up, and on shorter timescales core-mantle coupling and atmospheric angular momentum cause wobbles of milliseconds that the tidal trend rides on top of.
• "Recession and tidal locking are different phenomena." They are the same physics seen from two ends. The Moon's spin was tidally locked long ago by the very torque that now, acting on Earth, drives recession. The system is mid-way through a single relaxation toward mutual lock.