Moon: Origin, Properties, and Significance

• The Moon is Earth's natural satellite, characterized by its large relative size, low iron content, and formation via a giant-impact process.
• Advanced simulations and hydrodynamic models reconcile the Moon's mass, angular momentum, and volatile depletion with its early molten state.
• Observational studies of the exosphere, tidal dynamics, and radiometric effects highlight the Moon's crucial role in both planetary science and future astronomical research.

Searching arXiv for relevant Moon papers and ids to ground the response. {"query":"Moon formation origin Earth-Moon system arXiv (Canup et al., 2021, Asphaug et al., 2021, Malamud et al., 2024, Nakajima et al., 2018, Zahnle et al., 2015, Baraka et al., 2023, Glenar et al., 2019, Schneider et al., 2023, Robinson, 2011, Gupta, 2014)","max_results":10,"sort_by":"submittedDate"} {"query":"(Canup et al., 2021)","max_results":5,"sort_by":"relevance"} In the Earth–Moon system, the Moon is roughly $1/4$ the radius of the Earth, with mean radius $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$, a tiny core of perhaps $\sim1\%$ of its mass, and an origin tied to a system whose present angular momentum is $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$. Lunar samples and dynamical arguments indicate that the early Moon was hot and at least partially molten, with a deep magma ocean perhaps hundreds of kilometers deep. These properties make the Moon a stringent test case for planetary origin models, because any successful account must simultaneously explain the Moon’s size, iron depletion, thermal state, volatile record, and long-term tidal evolution (Canup et al., 2021).

1. Observational constraints of the Earth–Moon system

Four constraints dominate modern discussion of lunar origin. First, the Moon’s size is unusual relative to its primary: the lunar mass fraction is $M_M/M_\oplus=0.0123$, and the Moon is nearly one-quarter the size of Earth. Second, the Earth–Moon system has high angular momentum, commonly written for the present two-body configuration as

$L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$

with $a\approx384{,}000\ {\rm km}$ and $e\approx0.055$, giving $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$. Third, the Moon is iron-poor relative to Earth: the lunar core mass fraction is $\sim1\%$, whereas Earth’s core contains nearly $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$0 of its mass. Fourth, the young Moon possessed a global magma ocean, and crystallization timescales discussed in the literature span from $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$1 Myr to $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$2 Myr, depending on depth and heat-loss efficiency (Canup et al., 2021).

These constraints are coupled rather than independent. A model that reproduces the system angular momentum but leaves too much iron in orbit, or that produces an iron-poor Moon but fails to match the Earth–Moon compositional relation, is incomplete. This interdependence is central to why lunar origin remains a major problem in planetary science.

2. Competing origin models and their evidentiary status

Early non-impact models included capture, co-formation, and fission. In the modern synthesis, capture requires a mechanism to dissipate the orbital energy of a Mars-sized body without adding iron to the Moon, while co-formation fails to explain the Moon’s small core and the high system angular momentum. For that reason, the current consensus is the giant-impact hypothesis: a collision between proto-Earth and a roughly Mars-mass impactor, often labeled Theia, generated a silicate-rich circumterrestrial disk from which the Moon accreted (Canup et al., 2021).

Within the giant-impact family, the canonical scenario assumes an impactor mass ratio $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$3–0.2, impact velocity $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$4, and impact angle $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$5. Refinements include fast-spinning-Earth impacts, half-Earth grazing collisions, synestia scenarios, and multiple-impact models. A distinct variant is the hit-and-run return scenario, in which a first collision at $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$6 strips mantle material and leaves a surviving runner that returns after $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$7–1 Myr for a second giant impact. In smoothed-particle hydrodynamics calculations of the terminal collision, the protolunar disk typically has $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$8–1.5 $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$9, grazing returns can yield $\sim1\%$0, the disk iron fraction remains $\sim1\%$1, and common orthogonal returns produce disk inclinations of $\sim1\%$2–$\sim1\%$3 (Asphaug et al., 2021).

Multiple-impact models replace one terminal impact with several Mars-to-lunar-mass collisions. In a recent hybrid hydrodynamic/N-body treatment with self-consistent initial conditions and material strength, moonlet collisions span perfect merger, partial accretion, hit-and-run, and erosion/disruption regimes. From 20 self-consistent cases, $\sim1\%$4 of collisions resolved as perfect mergers, $\sim1\%$5 as partial accretion, $\sim1\%$6 as net erosion, and $\sim1\%$7 as restart cases; overall, $\sim1\%$8 of collisions did not reduce the mass of the largest moonlet. This supports the feasibility of accretionary lunar growth in a multi-generation framework (Malamud et al., 2024).

A contrasting non-impact account is Gupta’s proposal that Earth and Moon formed in parallel from the same region of the solar nebula. In that scenario, the Moon came into existence from the same gaseous cloud as Earth, moons can form only at the time of planet formation in a parallel and simultaneous process, and the present spins and orbits were imparted later by tangential impacts rather than by a single catastrophic Moon-forming collision (Gupta, 2014). This mechanism directly contrasts with late-impact models and is presented as a general account of large-moon formation in both the Solar System and extrasolar systems.

3. Disk accretion, Roche-limit physics, and volatile retention

In giant-impact models, the Moon forms from a circumterrestrial debris disk. Canonical calculations describe a disk of mass $\sim1\%$9–3 $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$0, mainly silicate melt and vapor, and N-body accretion from that disk yields a single Moon of mass $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$1 just outside the Roche limit $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$2 (Canup et al., 2021). In Gupta’s nebular co-accretion picture, Roche-limit sorting is likewise invoked, with a classical fluid Roche limit

$$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$3

and continued lunar growth parameterized by an order-of-magnitude accretion rate $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$4 (Gupta, 2014).

A central question is whether a hot Moon-forming disk necessarily implies strong water loss. Nakajima and Stevenson computed the disk’s vertical thermal structure for $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$5, bulk water abundances of 100, 500, and 1000 ppm, and mid-plane temperatures from 2500 K to 4000 K. Their result is that the upper parts of the Moon-forming disk are dominated by heavy atoms or molecules, with hydrogen a minor species. Hydrogen escape is therefore diffusion-limited rather than hydrodynamic. Using

$$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$6

they estimated total water-loss fractions $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$7, and in a representative 1.5-lunar-mass disk with 100 ppm $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$8, the lost fraction is $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$9. This implies that a giant-impact origin can be consistent with a water-rich Moon and that the observed depletion of K, Na, Rb, and related volatiles requires another mechanism (Nakajima et al., 2018).

The compositional problem remains broader than water alone. The Moon is depleted in K, Rb, Cs, Zn, and In by factors of 5–200 relative to Earth, yet some samples imply up to hundreds of ppm $M_M/M_\oplus=0.0123$0 in the lunar interior. This combination has made disk chemistry, vapor–melt exchange, and post-impact volatile processing central to current models of lunar accretion (Canup et al., 2021).

4. Early thermal history and tidal evolution

Post-impact Earth strongly constrains the Moon’s early orbit. In the “tethered Moon” framework, a steam–CO$M_M/M_\oplus=0.0123$1–rocky atmosphere capped Earth’s cooling so efficiently that the net radiative loss was limited to $M_M/M_\oplus=0.0123$2. Because the global magma ocean could not cool faster than this radiative bottleneck, Earth’s solidification time was $M_M/M_\oplus=0.0123$3–10 Myr, with longer cooling for higher-angular-momentum cases (Zahnle et al., 2015).

This atmospheric cap changes the tidal problem. As mantle viscosity increased during cooling, tidal dissipation became strong in partially solidified mantle, but the same atmosphere that slowed cooling also limited how much tidal energy Earth could radiate away. The consequence was a negative feedback between viscosity-dependent tidal heating and temperature-dependent viscosity. Orbital evolution of the Moon was therefore orders of magnitude slower than in constant-$M_M/M_\oplus=0.0123$4 models: instead of recession rates $M_M/M_\oplus=0.0123$5 at $M_M/M_\oplus=0.0123$6, the first few Myr admitted only $M_M/M_\oplus=0.0123$7–1 $M_M/M_\oplus=0.0123$8 (Zahnle et al., 2015).

The slower expansion made resonant capture likely. The lunar evection resonance occurs at approximately $M_M/M_\oplus=0.0123$9, and capture from a nearly circular orbit is almost certain if $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$0. In the tethered model, that condition is satisfied early, at $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$1–$L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$2 yr, while Earth is still molten. However, because $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$3 at that stage, lunar tides damp rather than excite eccentricity, so evection does not strongly pump $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$4. Substantial eccentricity growth is instead favored later, when the Moon is between $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$5 and $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$6, after Earth has cooled enough that $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$7 (Zahnle et al., 2015).

This suggests that present-day inclination and eccentricity need not be primordial in their current form. A plausible implication is that the Moon’s modern orbit encodes a sequence of post-formation resonant and dissipative episodes rather than a direct fossil of the impact geometry alone.

5. Exosphere, volatiles, outgassing, and plasma interaction

The Moon today supports a highly tenuous, surface-bound exosphere. Apollo mass spectrometry found daytime total neutral density $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$8 and nighttime density $L_{\rm orb}=\mu\sqrt{G(M_\oplus+M_M)a(1-e^2)},$9, with a total neutral-atmosphere mass $a\approx384{,}000\ {\rm km}$0 tonnes. Sunrise composition included $a\approx384{,}000\ {\rm km}$1, Ne, $a\approx384{,}000\ {\rm km}$2, $a\approx384{,}000\ {\rm km}$3, $a\approx384{,}000\ {\rm km}$4, and trace $a\approx384{,}000\ {\rm km}$5, $a\approx384{,}000\ {\rm km}$6, $a\approx384{,}000\ {\rm km}$7, $a\approx384{,}000\ {\rm km}$8, CO, $a\approx384{,}000\ {\rm km}$9, and $e\approx0.055$0. LCROSS excavated comparable masses of $e\approx0.055$1 and CO, with order-of-magnitude yields of $e\approx0.055$2 kg for each species (Crotts, 2012).

Several active processes are inferred. Apollo seismometers identified $e\approx0.055$3 epicenters, and deep moonquakes cluster along mare–highland boundaries with probability $e\approx0.055$4 of random alignment. Photoelectric charging near the terminator can generate surface potentials of $e\approx0.055$5 V, loft $e\approx0.055$6–$e\approx0.055$7 dust to heights up to $e\approx0.055$8 km, and sustain an estimated dust transport of $e\approx0.055$9. Episodic explosive gas release is modeled to occur when subsurface flow reaches $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$0, producing a typical $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$1 kg cloud, launching a regolith plug of $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$2, and forming a $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$3 m crater in minutes (Crotts, 2012).

Space weathering adds a second layer of volatile processing. A kinetic simulation of the coupled Sun–Earth–Moon system found that the Moon spends nearly $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$4 of its orbit in Earth’s magnetotail, but the magnetotail does not prevent the influx of solar-wind ions and ionospheric ions. Typical solar-wind proton fluxes $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$5 combined with empirical yields $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$6–$$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$7 imply $$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$8–$$L_{\rm EM}\approx3.5\times10^{34}\ {\rm kg\,m^2\,s^{-1}}$$9, and passage through the magnetotail lowers $\sim1\%$0 by only $\sim1\%$1. The same simulation yields dayside surface potentials of $\sim1\%$2–20 V, nightside potentials of $\sim1\%$3 to $\sim1\%$4 V, and electric fields of a few mV/m, sufficient to affect charged-dust transport (Baraka et al., 2023).

A separate L1 observatory concept emphasizes terrestrial oxygen delivery. In that treatment, Earth’s magnetic tail blocks $\sim1\%$5 of the solar wind and enables transport of $\sim1\%$6 from Earth’s upper atmosphere to the Moon during full-Moon passage, with an estimated flux $\sim1\%$7. The oxidation pathway is invoked to explain lunar hematite, with reported abundances of 0.1–1% at latitudes $\sim1\%$8 and an assumed conversion efficiency $\sim1\%$9–$R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$00 (Gore et al., 2021). The contrast between near-continuous ion delivery in global kinetic modeling and strong shielding in the observatory concept marks an active point of interpretation rather than a resolved consensus.

6. Earthshine, unresolved spectra, and astronomy from the Moon

The Moon is also a radiometric environment and an observational platform. Earthshine is the dominant source of natural illumination on the lunar surface during lunar night and in permanently shadowed regions. Near zero phase, the broadband hemispherical irradiance from Earth at the Moon is $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$01, with roughly equal contributions from solar reflectance and thermal emission, $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$02 and $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$03. Earth’s thermal irradiance at the Moon, $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$04–70 mW m$R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$05, exceeds the Moon’s internal heat flow of 9–13 mW m$R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$06 and should therefore be included in thermal models of permanently shadowed regions; simulations further show that a few-percent-area water frost mixed with regolith would be detectable in visible observations using earthshine illumination (Glenar et al., 2019).

From a remote observer’s perspective, the Moon can materially alter unresolved observations of the Earth–Moon system. In coupled Earth and Moon thermal models, a Moonlike satellite observed at full phase contributes about 20% of the total infrared continuum, $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$07 in Earth’s 9.6 $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$08 ozone band and 15 $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$09 CO$R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$10 band, as much as 80% in the 6.3 $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$11 water band, and more than 90% in the 4.3 $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$12 CO$R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$13 band. If unrecognized, this can bias brightness-temperature retrievals upward by about 20–40 K and distort inferences about atmospheric composition. The same phase dependence makes exomoon detection by gibbous-minus-crescent differencing feasible at signal-to-noise ratios of order 10–30 for a Moon-sized companion around an Earth twin at 10 pc (Robinson, 2011).

The lunar surface has correspondingly been proposed as a site for future astronomy. The absence of atmosphere gives access to the full UV/Vis/IR band and diffraction-limited seeing, while low gravity and absence of wind permit very large optics and substantial payloads. Proposed architectures range from a 0.3–1 m precursor telescope to large arrays, 30 m-class monoliths, and interferometers with baselines up to 10 km; at $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$14, a 10 km baseline corresponds to $R_M\approx1{,}737\ {\rm km}\approx0.27\,R_\oplus$15. Ambitious science cases include LOUPE-style spectropolarimetry of Earth as a single-pixel exoplanet, high-contrast imaging of exoplanets and exomoons, resolved studies of Einstein rings and gravitational arcs, and tests of quantum correlations across Earth–Moon scales (Schneider et al., 2023).

These observational roles broaden the scientific meaning of the Moon beyond origin studies alone. The Moon is simultaneously a product of planetary accretion, a laboratory for exospheric and regolith physics, a source of systematic effects in exoplanet spectroscopy, and a potential site for long-baseline astronomy.