A chunk of ice and rock roughly 500 kilometres across, orbiting the Sun nearly four billion kilometres beyond Neptune, carries a thin shell of gas that physics says it cannot hold. Findings published in Nature Astronomy in May 2026 place the surface pressure of the object designated 2002 XV93 at 100 to 200 nanobars, a small icy body in a gravitational lock with Neptune known as a plutino. That pressure sits 50 to 100 times below the current surface pressure of Pluto, but up to 100 times above the upper limits previously placed on every other similarly sized object in the outer solar system. No body this small, at this distance from the Sun, was supposed to carry measurable gas at all.
The detection came through a technique called stellar occultation, one of the most precise tools available for probing the edges of distant solar system bodies. When a known object passes in front of a background star, the starlight dims sharply as the solid body blocks it. If there is an atmosphere present, the light dims more gradually at the edges of the shadow, bending slightly as it passes through the gas layer before the body itself cuts it off entirely. That gradual bending, called a refractive signature, is what the observation campaign recorded on 10 January 2024, when 2002 XV93 passed in front of a star of magnitude 15.8 as seen from East Asia. Three stations in Japan captured usable data: Kyoto University’s rooftop using a 20-centimetre portable telescope, Kiso Observatory using a 1.05-metre Schmidt telescope, and a citizen astronomer in Fukushima operating a 25-centimetre instrument. A fourth station at Okayama Observatory, which houses a 3.78-metre telescope, was lost to bad weather.
The Kiso data provided the cleanest atmospheric signature. As the star approached the edge of the object’s shadow, the telescope recorded a gradual dimming lasting approximately 1.5 seconds on both entry and exit, rather than the instantaneous drop that an airless body would produce. The Fresnel scale, a physical limit set by the geometry of light diffraction at a distance of 37 astronomical units, corresponds to a dimming duration of roughly 0.05 seconds at a wavelength of 550 nanometres. The 1.5-second duration is 30 times longer than that diffraction limit, ruling out any purely optical or instrumental explanation for the gradual dimming. The citizen astronomer’s station at Fukushima, sitting outside the main shadow path, recorded a separate gradual flux drop of approximately 10 seconds’ duration at 4.0 sigma significance, consistent with a refractive skimming of the outer atmosphere at a closest-approach distance of 230 to 240 kilometres from the shadow centre.
To convert that light curve into a pressure reading, the team ran models assuming three possible atmospheric compositions: pure methane, nitrogen-dominant, and carbon monoxide-dominant. These are the only gases expected to sublimate, meaning transition from ice directly to vapour, at the temperatures 2002 XV93 experiences, sitting in the range of 40 to 50 Kelvin, equivalent to roughly minus 223 to minus 233 degrees Celsius. At those temperatures, most ices are locked permanently solid; methane, nitrogen, and carbon monoxide are the exceptions, loose enough in their molecular bonds to pass into a gas phase at the surface. All three composition models produced consistent best-fit results: surface pressures of 124, 177, and 159 nanobars respectively, against a body radius of 244 kilometres. The standard statistical quality metric for the atmosphere-bearing models came in at around 28.5 to 28.7 with 29 degrees of freedom, compared to 89.2 for the atmosphere-free model, which is a three-fold improvement in fit quality.
The problem is that 2002 XV93 is too small to keep any of those gases. Whether an atmosphere can be retained depends on a ratio called the Jeans parameter, which compares the gravitational energy binding a gas molecule to the surface against the thermal energy pushing it outward. When that ratio approaches 1, gas escapes rapidly in a hydrodynamic outflow, effectively a continuous fast wind blowing into interplanetary space rather than the slow trickle of a more massive body. For a body of 2002 XV93’s size, estimated bulk density of 1,500 kilograms per cubic metre, and surface temperature of 47 Kelvin, the Jeans parameter sits close to 1 for all three candidate gases. A surface atmosphere at 100 to 200 nanobars around an object of these dimensions loses its gas on a timescale of roughly 100 to 1,000 years for methane, nitrogen, and carbon monoxide under conservative thermal escape assumptions. The age of the solar system is 4.6 billion years. An atmosphere that survives for only a thousand years is not a permanent feature; it is a transient, and something produced it very recently in astronomical terms.
The two plausible mechanisms run in opposite directions in terms of what they say about the object’s interior. The first is cryovolcanism, a process in which internal heat melts subsurface material, which then seeps or vents to the surface and escapes as gas. Cryovolcanism, driven by water-ammonia or water-methanol slurries rather than silicate magma, has been inferred on larger Kuiper Belt objects including Sedna, Gonggong, and Quaoar, all of which show surface chemistry consistent with ongoing internal processing at their larger radii of 500 kilometres and above. James Webb Space Telescope isotopic measurements of methane ice on Eris and Makemake published in 2024 indicate that surface methane on those bodies is at least partly supplied by warm interior reactions rather than being primordial, frozen-in ice from the early solar system. For a body as small as 2002 XV93, sustained cryovolcanism is considered unlikely because small bodies cool quickly, their interiors going solid and cold over geological time, but under specific circumstances, unusually high concentrations of antifreeze compounds such as ammonia or methanol, or tidal forcing from an undiscovered satellite, preserve liquid pockets long enough for venting to persist.
The second mechanism is a recent impact. A small comet, approximately 100 metres in radius, striking 2002 XV93 at a low relative velocity, consistent with the typical collision speeds between plutinos in their shared orbital zone, delivers between 100 billion and 1 trillion kilograms of carbon monoxide, methane, or nitrogen from its own interior. That quantity of gas, spread across the surface of a 250-kilometre body, is enough to produce a global atmosphere at the 100-nanobar scale. The low collision velocity matters because if the impactor hits slowly enough, a substantial fraction of the released volatiles stay gravitationally bound to the target rather than blasting free into space immediately after the collision. Based on crater counts from the New Horizons mission at Pluto, Charon, and the contact binary Arrokoth, the probability that a body of 2002 XV93’s size has been struck by a 100-metre or larger projectile within the past 100 years sits at roughly one in 100,000, though occultation monitoring at sub-kilometre scales allows the possibility of a higher actual abundance of small impactors than crater counts alone capture.
Every other comparably sized or larger body in the Kuiper Belt has been checked and returned nothing. Eris, at a radius of approximately 1,163 kilometres, showed no global atmosphere down to 10 nanobars in a multichord occultation, with its hypervolatiles currently frozen out at its distance of 97 astronomical units from the Sun. Haumea, at roughly 800 kilometres radius, showed no atmosphere down to 50 to 100 nanobars despite sitting at an intermediate distance where sublimation should be active. Makemake, at 710 kilometres, showed the same absence of global atmosphere at the 50 to 100 nanobar level, though the James Webb Space Telescope has since detected methane fluorescence from its surface, possibly indicating comet-like outgassing or an extremely thin localised atmosphere at around 10 picobars, 10,000 times below what 2002 XV93 now carries. Quaoar, at 555 kilometres, shows surface methane ice but no measurable gas, capped at 1 nanobar. Three other objects in the 200 to 300 kilometre radius range, Ixion, Huya, and 2002 TC302, were all checked via stellar occultation at distances comparable to 2002 XV93’s current position of 38 astronomical units and returned no detectable atmosphere.
The question of which mechanism produced the atmosphere is not currently answerable from the January 2024 data alone. Near-infrared to mid-infrared spectroscopy from the James Webb Space Telescope, targeted at 2002 XV93, would identify the molecular species present and give constraints on whether the gas is rising from a warm interior or condensing back after an impact-driven pulse. The second diagnostic is time. If the atmosphere is impact-generated, surface pressure is dropping measurably right now, and repeated occultation observations over a five to ten year window will record a monotonic decline toward zero. If pressure holds steady or varies seasonally with the body’s orbital position relative to the Sun, endogenous outgassing from an active interior is the more likely driver. 2002 XV93 currently sits at 38 astronomical units, close to its perihelion, where surface temperatures are at their highest and sublimation rates are greatest. Active monitoring of this target is ongoing.
Source:
Arimatsu, K., Yoshida, F., Hayamizu, T., Takita, S., Hosoi, K., Ootsubo, T., & Watanabe, J. (2026). Detection of an atmosphere on a trans-Neptunian object beyond Pluto. Nature Astronomy. https://doi.org/10.1038/s41550-026-02846-1
