1I/'Oumuamua

What was seen of 'Oumuamua was a point of light, never resolved. Everything known about its size and shape is inferred from its brightness and how it changed. For a comet-like albedo of 0.04 its mean radius was about 100 m⁠[2]. Spitzer's non-detection at 4.5 microns sets an effective spherical diameter below 98, 140 or 440 m depending on the thermal model, and an albedo above 0.2, 0.1 or 0.01 to match⁠[8]. The surface was red, with a featureless spectral slope like that of outer Solar System bodies and irradiated organic-rich crusts⁠[14]⁠[15].

Its incoming velocity was within a few km/s of the local standard of rest, the average motion of nearby stars, which is unusual for any one star and suggests a young object or one that has been little disturbed⁠[16]. Its orbit is published by JPL as solution 16, fitted to 207 observations with a non-gravitational term⁠[3].

Micheli and colleagues tested and rejected, one by one, radiation pressure on an ordinary rocky body, the Yarkovsky effect, friction-like drag, forces from the solar wind on a magnetised body, a binary or split nucleus, and an offset between the centre of light and the centre of mass. They found outgassing the one physically viable mechanism if the body had comet-like thermal properties⁠[5].

Three radio searches looked for transmitters. The Green Bank Telescope observed on 13 December 2017 across 1.1-11.6 GHz and found no narrowband signal, a limit of 0.08 W for an omnidirectional transmitter on the object⁠[24]. The Allen Telescope Array found nothing between 1 and 10 GHz, with limits of about 10 to 100 W in clean bands⁠[25]. The Murchison Widefield Array, which happened to have the object in its field on 28 November, found nothing at 72-102 MHz⁠[26]. The Green Bank raw data are public⁠[27].

Two quantitative tests have been run on the lightsail reading. A tumbling sail would produce a light-curve amplitude like the observed one with probability below 1.5 per cent, and would have been visible in all 55 observations with probability of 0.4 per cent⁠[28]. A sail of the inferred mass and thickness could not have been driven to speeds that would bring it from a nearby star in less than about half a billion years⁠[29].

The origin has been tested by tracing the trajectory back through the Galaxy. Using Gaia DR2, Bailer-Jones and colleagues found four candidate encounters in the last few million years, the closest with the M dwarf HIP 3757 at 0.60 pc about a million years ago and a slower one with the G dwarf HD 292249 3.8 million years ago. The authors note that their sample was a small fraction of all stars and that finding the true home was unlikely from the start⁠[30]. Scattering by the Galactic disc limits such traces to about 15 pc accuracy at 10 million years and 400 pc at 100 million, so a single home star cannot be named; the Carina and Columba moving groups are the most plausible source regions⁠[31]⁠[32]. Kinematic ages from its small excursion above the Galactic plane point the same way, to a young population of about a billion years or less⁠[33].

An ordinary icy planetesimal outgassing water below detection. This was the discovery team's preferred reading⁠[5]. Its difficulty is that water sublimation driven by sunlight does not fit the outbound acceleration for a body of this size, and CO and CO2 were excluded by Spitzer⁠[34]⁠[8]. Since 2023 a Solar System analogue has been found: small near-Earth asteroids with no visible activity that show accelerations too large for the Yarkovsky effect, called dark comets, of which more than a dozen are now known⁠[35]⁠[36].

Molecular hydrogen ice. Seligman and Laughlin showed that hydrogen sublimation, which needs little energy, reproduces the acceleration without a visible coma, and that mass loss would sharpen the shape⁠[34]. Hoang and Loeb replied that starlight and collisions in molecular clouds would destroy such a body long before it reached the Sun⁠[37]. Levine and colleagues, including the reading's authors, found the required formation temperatures implausible⁠[10].

Nitrogen ice from an exo-Pluto. Jackson and Desch found that a fragment of nitrogen ice, like the surface of Pluto, matches the acceleration, the size, a high albedo and the absence of CO, CO2 and dust, giving a body of about 45 by 44 by 7.5 m at the time it was seen⁠[38]⁠[39]. Siraj and Loeb argued that the mass of exo-Plutos needed to make such fragments common enough would exceed the heavy elements available in stars⁠[40]; cosmic-ray erosion in transit worsens the budget⁠[41]; and cratering simulations do not produce nitrogen fragments that large⁠[10].

Hydrogen made by cosmic rays in water ice. Bergner and Seligman proposed that cosmic rays produced molecular hydrogen inside an ordinary water-rich body over its journey, and that warming near the Sun released it from annealing amorphous ice, giving the push with no dust⁠[42]. It needs no exotic ice and is grounded in laboratory work. Hoang and Loeb argued that cooling by the escaping hydrogen would cut the gas budget by a factor of five to nine⁠[43]; the model's code is public, so the dispute can be rerun⁠[44].

A fractal dust aggregate. If the body were extremely porous, sunlight alone could give the push. Moro-Martín found the density would need to be about 10^-5 g per cubic centimetre⁠[45]; Flekkøy and colleagues showed such a body could survive and could explain an apparent spin change, and proposed a formation route in the coma of a comet in another system⁠[46]⁠[47]. Optical and infrared observations of a future object could tell such a body from a solid one⁠[48].

A tidal fragment. Zhang and Lin showed that bodies passing close to their star can be torn into elongated shards with baked surfaces and buried ice, which could outgas without visible activity⁠[49]. This is a formation route compatible with several of the readings above rather than a separate composition.

An artificial lightsail. Bialy and Loeb showed that radiation pressure explains the acceleration if the body has a mass-to-area ratio of about 0.1 g per square centimetre, a sheet 0.3-0.9 mm thick, and suggested it might be an artificial sail⁠[50]. Loeb has argued since that every natural reading invokes an object of a kind never seen before, so an artificial origin must be considered⁠[51]. The ISSI team found no observation that requires it⁠[11]; the radio searches found nothing⁠[24]; and the light-curve and travel-time tests weigh against it⁠[28]⁠[29]⁠[52]. Elvis set out what the reading would predict and how it could be tested⁠[53].

2I/Borisov, found on 30 August 2019, was an active comet in every respect: a coma, a tail, a nucleus of a few hundred metres to a kilometre, and an unusually high ratio of carbon monoxide to water⁠[12]⁠[54]⁠[55].

3I/ATLAS was found by the ATLAS survey on 1 July 2025, on a steeply hyperbolic orbit (eccentricity about 6.1, excess speed 58 km/s), already showing a compact coma⁠[13]. JWST found a coma dominated by carbon dioxide, with a CO2-to-water ratio among the highest seen in any comet, and later the first methane detected on an interstellar object⁠[56]⁠[57]. Nickel vapour was found far from the Sun, as it has been in Solar System comets and in 2I⁠[58]⁠[59]⁠[60]. It passed perihelion on 29 October 2025 at 1.36 au⁠[61]. Early Hubble images limited the nucleus radius to between 0.22 and 2.8 km⁠[62]. Images after perihelion give about 1.3 km, which the authors find consistent with the size implied by the non-gravitational acceleration⁠[63]. That acceleration, measured from nearly 7,800 positions, is required by the data but its systematic errors exceed the formal ones⁠[61]. Mass-balance estimates from it give a diameter of about 0.8-1.05 km⁠[64] and, from Loeb's group, a radius of 0.42-1.15 km depending on the gas that drives it⁠[65], both with large uncertainties. Two radio searches found no signals⁠[66]⁠[67]. NASA treats it as a comet and lists more than a dozen of its spacecraft that observed it⁠[68].

Loeb and colleagues raised the possibility that 3I is alien technology, in a paper they described as largely a pedagogical exercise⁠[69], argued from early astrometry for a nucleus above 5 km⁠[70], and placed both 1I and 3I at level 4 of a ten-point scale of anomaly⁠[71]. Other astronomers answered the claims point by point, notably Jason Wright, who argued that the reported features are known in comets and that the size claim did not survive the Hubble measurement⁠[72]. A Galactic population model applied a week after discovery found that 3I's speed implies an age above 7.6 billion years and a high water fraction, and makes a shared origin with 1I or 2I very unlikely⁠[73].

'Oumuamua itself is beyond reach of telescopes. It can be reached only by a spacecraft, and every study of a chase finds a flight of decades: the most developed Project Lyra trajectory launches in 2028 and arrives about 26 years later, passing the object at about 18 km/s⁠[74]⁠[75]. A refuelled Starship or an expendable Falcon Heavy could launch it⁠[76]. No agency has adopted it.

The practical route is another object like it. The Vera C. Rubin Observatory began its ten-year survey in mid-2026; one estimate puts the number of 'Oumuamua-like objects it will find in ten years at between less than one and about 84, a measure of how poorly the population is known⁠[77]⁠[78]. For such an object, the tests are known. Infrared photometry at several wavelengths would give its size and albedo and distinguish a porous aggregate from a solid⁠[48]. Spectra of the escaping gas would identify it: molecular hydrogen, nitrogen or water⁠[10]. A well-sampled light curve would fix its shape and spin, and astrometry would measure the push again. ESA's Comet Interceptor, due to launch in late 2028 or early 2029 and wait at the Sun-Earth L2 point, could fly past one if it is reachable⁠[79]⁠[80]. A close image would settle the shape, and a spectrum of the surface its composition.

For 'Oumuamua's own data, the open questions are computational: a joint fit of the light curve and the astrometry with one physical model of outgassing and torque, which no published model yet achieves⁠[19].