Most clouds need help to form. On Earth, and as far as anyone had confirmed anywhere else, water vapour does not readily turn into droplets or ice crystals by itself. It needs a seed, such as a dust grain or another small particle, to condense onto. A new study of one of Mars's strangest weather features suggests that there, at least some of the time, the vapour does not wait for one.
The feature is the Arsia Mons Elongated Cloud (AMEC), a thin streak of water ice that trails downwind of the 20-km-tall Arsia Mons volcano. According to the European Space Agency, a team led by Jorge Hernandez-Bernal of LMD/CNRS/Sorbonne Universite could reproduce the cloud in a Mars weather model only after adding what Hernandez-Bernal calls "exotic physics": homogeneous nucleation, in which water vapour freezes directly into ice with no particle to start from.
A cloud that keeps coming back
AMEC is unusual even before its microphysics come into it. It stretches up to 1,800 km (1,118 miles) from the volcano. Arsia Mons is the southernmost of the three Tharsis Montes, Sci.News notes. The cloud does not appear once and then disperse. It forms, grows and fades every day for several months during the southern dusty season. Mars Express first revealed it in 2018, and ESA calls it Mars's "oddest cloud."
Its daily regularity makes AMEC a good test case. A cloud that repeats on schedule gives modellers a target to match: when it appears, how far it extends and how quickly it goes. ESA says the researchers based their modelling on data from three instruments aboard Mars Express: the Visual Monitoring Camera (VMC), the High Resolution Stereo Camera (HRSC) and OMEGA.
What the model needed
The study's key mechanism is how fast air cools over the volcano. ESA reports that winds crossing Arsia Mons push moist air up by several kilometres within minutes, and the temperature falls by 30 degrees in about 10 minutes. Sci.News describes the same process as a wave lifting the air.
That rapid cooling sends relative humidity soaring. ESA puts it at more than 100,000 times the levels typically experienced on Earth. At that point, in the team's account, water vapour freezes directly into cloud particles, which is homogeneous nucleation.
The preprint describing the work, "Homogeneous Nucleation of Water Vapor Evidenced by Elongated Clouds on Mars" by J. Hernandez-Bernal, A. Maattanen, A. Spiga and F. Forget, was submitted to arXiv on 29 September 2026. Its abstract says that introducing homogeneous nucleation into a Mars meteorological model reproduces the unique characteristics of the cloud, which had remained impossible to model under conventional cloud microphysics. The authors say this gives the first evidence of homogeneous nucleation of water vapour in a planetary atmosphere.
How strong is the claim?
The argument works by elimination, and it is worth being clear about what that means. Nobody watched individual ice crystals form without seeds on Mars. The team found that its model reproduced the observed cloud only when the process was included. That is a meaningful result, because the cloud is well observed and its daily cycle is a demanding test for any simulation. But it is still evidence from modelling, not a direct measurement. ESA's own headline says the cloud "may be" even odder than thought, and Hernandez-Bernal acknowledges that some aspects of the modelled cloud do not exactly match the observations.
The arXiv listing itself carries a note that the preprint has not undergone peer review. ESA's release of 7 October 2026, about a week after the preprint went up, says the study is published in Nature Geoscience, and Mars Express project scientist Colin Wilson is quoted in it.
Why It Matters
Homogeneous nucleation is not a new idea. As Sci.News points out, theorists have proposed it for the upper atmospheres of Earth and Venus, but no one had confirmed it in either place. If the AMEC result holds up, Mars would be the first planetary atmosphere where the process has observational support. The evidence would also come from a cloud big enough to be seen from orbit and regular enough to study day after day.
The authors write that the finding challenges current assumptions about cloud formation processes and is also relevant for certain clouds on Earth and potentially on other planets. ESA likewise says the result shows that unlikely processes should not be discounted when exploring other worlds, including exoplanets.
The result also shows the value of long-running missions. Mars Express discovered the AMEC in 2018 and has viewed it repeatedly since. Colin Wilson says the spacecraft can track how the cloud changes over mere hours, and notes that it is one of the few Mars orbiters, alongside ESA's ExoMars Trace Gas Orbiter, able to observe during the morning hours when the cloud is present.
What comes next
Because the evidence is a model match rather than a direct observation of seedless ice formation, the natural test is whether other researchers and models can reproduce the result. Hernandez-Bernal says the finding strongly suggests that Mars's humidity can reach the extreme levels the process requires. For now, a cloud already known as strange has given planetary scientists a candidate first: ice forming from vapour alone, with no dust needed.
