Small Bodies
Earth's Quasi-Moons: The Rocks That Shadow Us
On April 27, 2016, a 15-second exposure from the Pan-STARRS telescope on Hawaii's Haleakalā caught a speck barely 40 metres across looping around our planet — yet the little rock does not orbit Earth at all. Kamo'oalewa circles the Sun in almost exactly one year, and from our moving vantage it seems to trace a lazy, tilted noose around us, drifting no closer than about 14.4 million kilometres (roughly thirty-eight times the Moon's distance) and staying tethered to us not by gravity but by coincidence of timing. It has shadowed Earth for centuries and will keep it up for a few hundred more — and it is only the best-known of at least seven such ghostly companions.
- Known Earth quasi-moons~7 confirmed (2025)
- Best-known469219 Kamo'oalewa (2016 HO3)
- Kamo'oalewa size~40 m (est. 18–100 m)
- Closest approach~14.4 million km (~38 lunar distances)
- Kamo'oalewa spin~28 min per rotation
- Longest-lived2023 FW13 — ~100 BC to AD 3700
- Earth's Hill sphere~1.5 million km (0.0098 au)
- Sample-return missionTianwen-2, arrived 6 July 2026
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
What you'd actually see: a wobble, not an orbit
If you could hover above the Sun and watch Earth's neighbourhood in a frame that rotates once a year — so that Earth appears to sit still — a quasi-moon would seem to swing around our planet in a stretched, tilted loop, month after month, year after year. That is exactly the picture that inspired the name Kamo'oalewa, drawn from the Hawaiian creation chant Kumulipo and meaning roughly "the oscillating fragment." To a naive eye the loop looks like an orbit. It isn't.
Switch to a non-rotating frame — the way the real solar system works — and the illusion dissolves. The rock is simply going around the Sun on its own independent ellipse, taking almost exactly one year, the same as Earth. Because the two periods match, our planet never laps the asteroid and the asteroid never laps us; the geometry between them repeats, so from Earth the companion appears pinned nearby. The apparent loop is a parallax artifact of two bodies marching in lockstep around a common star.
Crucially, the loop is retrograde: in the co-rotating frame the quasi-moon circulates the "wrong" way, tracing a backward sense that is characteristic of quasi-satellite geometry. That retrograde loop is a useful visual hallmark, but the definitive test of a quasi-satellite is dynamical: the body remains gravitationally unbound and outside Earth's Hill sphere, whereas a genuinely captured moon can circulate either way. And the loop is huge — for Kamo'oalewa the asteroid never actually comes closer to Earth than about 14.4 million kilometres, roughly thirty-eight times farther than the Moon, and can swell to tens of millions of kilometres. (The two orbits pass as close as about 4.6 million km — the minimum-orbit-intersection distance — but the bodies never occupy that crossing point at the same time.) No spacecraft-scale gravity binds it; only the arithmetic of matched orbital periods.
The mechanism: 1:1 resonance and the Hill sphere
The engine behind every quasi-moon is a 1:1 mean-motion resonance. The asteroid's orbital period equals Earth's to within a fraction of a percent — Kamo'oalewa's semi-major axis is 1.001 au against Earth's 1.000, giving it a year of about 365.8 days. What makes its path look loopy from here is its different shape: an eccentricity near 0.10 and an inclination of about 7.8° to the ecliptic. That eccentric, tilted ellipse crosses inside and outside Earth's near-circular one, so over each year the asteroid runs ahead, then falls behind, then catches up — weaving the closed loop we perceive.
The deciding boundary is Earth's Hill sphere, the region within which our planet's gravity, not the Sun's, dominates. Its radius is only about 1.5 million kilometres (≈0.0098 au) — less than four times the Moon's distance. A true moon lives deep inside this bubble. A quasi-moon spends its entire cycle outside it, which is precisely why it is never gravitationally captured: the Sun always has the stronger pull on it. The "companionship" is a resonance lock, not a leash.
- True satellite: inside the Hill sphere, bound, stable for aeons.
- Quasi-satellite: outside the Hill sphere, unbound, held only by matched periods; stable for centuries to hundreds of thousands of years.
- Minimoon: briefly captured inside the Hill sphere, then released.
Because the lock is delicate, quasi-moons are transient on cosmic timescales. Kamo'oalewa is expected to keep its quasi-satellite behaviour for only a few hundred years before drifting into other co-orbital patterns; dynamical models give it a total lifetime in this state of order 0.3–0.5 million years. It is a phase, not a permanent address.
Cousins of the loop: horseshoes, tadpoles, and minimoons
Quasi-satellite motion is one member of a family of co-orbital behaviours, and the same asteroid can slide between them over centuries. All arise from the same 1:1 resonance but differ in geometry:
- Tadpole (Trojan) orbits: the body librates around one of the stable Lagrange points, L4 or L5, staying 60° ahead of or behind the planet. Earth has two confirmed Trojans, 2010 TK7 (discovered in WISE data in 2010, confirmed 2011) and the larger 2020 XL5 (confirmed at L4 in 2022), both librating around the leading L4 point.
- Horseshoe orbits: the body drifts all the way from far ahead of Earth to far behind it, turning back before it reaches us — tracing a horseshoe that wraps around both L4 and L5. The best-studied example is 3753 Cruithne, sometimes miscalled "Earth's second moon."
- Quasi-satellite orbits: the tight retrograde loop described above, staying near Earth throughout.
Real objects switch modes. Kamo'oalewa, Cruithne, and others repeatedly transfer between horseshoe and quasi-satellite states on timescales of centuries to millennia — a quasi-moon today may be a horseshoe companion in a thousand years.
All of this is distinct from a genuine minimoon: a small asteroid that Earth's gravity briefly captures into a true, if temporary, geocentric orbit. Capture requires a slow encounter, under about 1 km/s relative to the Earth–Moon system. The car-sized 2006 RH120 orbited Earth from July 2006 to July 2007; 2020 CD3 was bound from roughly 2017 until early March 2020 (a ~2.7-year capture); and 2024 PT5 made a partial-orbit "mini-moon" pass in late 2024. These are the closest thing Earth has to extra moons — but they leave within months to a few years, unlike the durable resonance of a quasi-moon.
The numbers on Kamo'oalewa — and why the size is fuzzy
Kamo'oalewa (formally 469219 Kamo'oalewa, provisional designation 2016 HO3) is the smallest, closest, and most-studied Earth quasi-moon, discovered on 27 April 2016 by Pan-STARRS 1 on Haleakalā and formally named in 2019. Its precise diameter is genuinely uncertain because it is so faint — absolute magnitude about 24, near the limit of even large telescopes. Ground-based estimates put it around 40–100 metres; more recent analyses, including 2026 JWST thermal data, argue for something smaller — the JWST thermal modelling gives a mean diameter of about 18 ± 2 metres (roughly 16–20 m). Either way, it is a boulder, not a world — comparable to a Ferris wheel or a small office building.
What is well pinned down is its spin: it rotates roughly once every 28 minutes, one of the faster rotators known among near-Earth asteroids. Its orbit hugs Earth's: semi-major axis 1.001 au, eccentricity ≈0.10, inclination ≈7.8°, Earth minimum-orbit-intersection distance of about 0.031 au (4.6 million km, ≈12 lunar distances). It is decidedly not a collision threat — the geometry keeps it millions of kilometres away.
The other confirmed Earth quasi-moons are fainter and less characterised. Standouts include:
- 2023 FW13 — about 15 m across, discovered March 2023 by Pan-STARRS; models suggest it has been a quasi-satellite since roughly 100 BC and will remain so until around AD 3700, making it the longest-lived known.
- 2025 PN7 — about 20 m, announced in 2025, with an estimated tenure of decades rather than millennia.
- Earlier members such as 2004 GU9, (277810) 2006 FV35, 2013 LX28, and 2014 OL339 round out the small confirmed set.
Is Kamo'oalewa a chip off the Moon? A live debate
The most tantalising claim about Kamo'oalewa is that it might be a piece of the Moon. A 2021 study led by Ben Sharkey (University of Arizona / Lowell Observatory) found that its reflectance spectrum is unusually red — redder than typical inner-solar-system asteroids — and matches space-weathered lunar silicates better than ordinary asteroid types. That prompted a 2024 modelling paper suggesting the rock could be ejecta blasted off the Moon by the impact that dug the young crater Giordano Bruno, perhaps within the last few million years.
But the case is not settled. In 2026, JWST spectroscopy of Kamo'oalewa complicated the picture: the near-infrared spectrum looked more consistent with an S-, V-, or E-type asteroid — possibly an enstatite-rich, oldhamite-bearing body — than with genuine lunar mare soil. In other words, extreme space-weathering can make an ordinary silicate asteroid look lunar without being lunar. Reasonable specialists currently disagree, and the object sits at the centre of an active argument about how to read reddened asteroid spectra.
This is exactly why China's Tianwen-2 mission matters. Launched on 28 May 2025, the spacecraft reached Kamo'oalewa on 6 July 2026 after a roughly billion-kilometre cruise, imaging the asteroid from about 20 km away, and is designed to collect a sample and return it to Earth around 2027. Laboratory analysis of that sample — mineralogy, isotopes, exposure age — should decide whether Earth's quasi-moon is a wandering asteroid or a long-lost fragment of our own Moon. Until then, honesty requires saying: we don't yet know.
Misconceptions, and how these ghosts were found
Quasi-moons collect myths, so a few corrections are worth making explicit:
- "Earth has a second moon." No. A quasi-moon is not a moon; it orbits the Sun and lies outside Earth's Hill sphere. The headline-friendly phrase is convenient but physically wrong.
- "It could hit us." Not on its current path — Kamo'oalewa stays millions of kilometres away, and the loop geometry actively keeps quasi-satellites at arm's length.
- "It's permanent." No co-orbital state is permanent; these are transient phases lasting centuries to at most a few hundred thousand years, with objects shuttling between horseshoe, tadpole, and quasi-satellite motion.
- "Only Earth has them." No — Venus hosts the quasi-satellite 524522 Zoozve (2002 VE68), and quasi-satellites are known or suspected for other planets too.
Finding these objects is genuinely hard. They are tiny, faint, and — worst of all — usually near the Sun in our sky, spending long stretches lost in daylight glare. It took the wide-field, repeated all-sky sweeps of survey telescopes like Pan-STARRS on Haleakalā to catch them, and confirmation typically demands follow-up from facilities such as the Canada–France–Hawaii Telescope and observatories in Arizona before the Minor Planet Center lists the discovery. As survey depth improves — and with the Vera C. Rubin Observatory's sky survey now hunting exactly this kind of faint, fast mover — astronomers expect the roster of Earth's co-orbital companions to grow, and the once-exotic idea of a rock that shadows us to become almost routine.
| Property | True moon (the Moon) | Quasi-moon (Kamo'oalewa) | Minimoon (2020 CD3) |
|---|---|---|---|
| What it orbits | Earth, gravitationally bound | The Sun, in 1:1 resonance | Earth, but only temporarily |
| Inside Hill sphere (~1.5M km)? | Yes, deep inside | No — well outside | Yes, briefly |
| Typical distance | 0.384 million km | ~14–40 million km | < ~1.5 million km |
| Duration of the arrangement | ~4.5 billion years | Centuries to ~10⁵ years | Months to a few years |
| Held by Earth's gravity? | Yes | No — a shared-orbit illusion | Yes, weakly |
Frequently asked questions
Is a quasi-moon really a moon of Earth?
No. A quasi-moon orbits the Sun, not Earth, and stays outside Earth's Hill sphere (the ~1.5-million-km bubble where our gravity dominates). It only looks like it circles us because its orbital period matches Earth's almost exactly — a 1:1 resonance — so the two bodies keep pace and their relative geometry repeats. Nothing binds it to us gravitationally.
How many quasi-moons does Earth have?
As of 2025, about seven are confirmed, including Kamo'oalewa (2016 HO3), 2023 FW13, 2025 PN7, and a few earlier members like 2006 FV35 and 2014 OL339. The number is fuzzy on the edges: objects drift into and out of the quasi-satellite state, and new faint ones keep being discovered.
Could Kamo'oalewa or another quasi-moon collide with Earth?
Not on its current orbit. Kamo'oalewa never comes closer than about 14.4 million kilometres — roughly thirty-eight times the Moon's distance. (The two orbits pass within about 4.6 million km, but the bodies never meet at that crossing.) The resonant geometry that makes it a quasi-moon is precisely what keeps it at a safe remove; it is a companion, not a threat.
Why does China's Tianwen-2 mission care about a 40-metre rock?
Kamo'oalewa may be a fragment blasted off the Moon, or it may be an ordinary but heavily space-weathered asteroid — its spectrum is ambiguous, and 2026 JWST data actually pushed against the lunar idea. Tianwen-2, which reached the asteroid on 6 July 2026, aims to return a sample around 2027, so laboratory mineralogy and isotopes can settle the origin question that telescopes alone cannot.
What's the difference between a quasi-moon and a minimoon?
A minimoon is genuinely captured by Earth's gravity into a temporary geocentric orbit — like 2006 RH120 (2006–2007) or 2020 CD3 — but it escapes within months to a few years. A quasi-moon is never captured; it orbits the Sun in lockstep with Earth and merely appears to loop around us, a state that can last centuries to hundreds of thousands of years.
If Kamo'oalewa never gets captured, could it ever become a real moon of Earth?
Extremely unlikely for this object, but not impossible in principle for some quasi-satellite. True capture requires the body to pass slowly through Earth's Hill sphere at under about 1 km/s relative velocity — the condition that makes fleeting minimoons possible. A quasi-moon's loop keeps it millions of kilometres outside that bubble, so it would first have to be nudged into a much lower-energy encounter, most plausibly with help from a lunar close pass. Rather than becoming a moon, a quasi-satellite far more often just transitions into a horseshoe orbit and eventually drifts away entirely.