This Week in Astronomy: Dark Energy Wobbles, and the Smallest Worlds Beyond Neptune
Some weeks hand you one enormous headline. This week handed us four quieter results, each chipping away at something we thought was settled: a fresh wobble in dark energy, the smallest icy bodies ever seen beyond Neptune, the birthplace temperature of an interstellar comet, and a search that found absolutely nothing and was all the more useful for it.
Dark energy still will not sit still
The standard model of cosmology treats dark energy as a constant: a fixed amount of outward pressure woven into every cubic metre of space, unchanging since the Big Bang. Over the past two years that assumption has developed a wobble, and a paper posted on 4 September gives it another nudge.
A team led by Ryan Camilleri at the University of Queensland, with Tamara Davis, Dan Scolnic and more than twenty colleagues, took on the unglamorous but essential job of stitching the two largest collections of Type Ia supernovae into one consistent dataset. Combining Pantheon+ with the Dark Energy Survey sample known as DES-SN5YR gives 2,884 supernovae calibrated on the same footing, and pins the matter density of the Universe at 0.310, give or take about 0.012.
The interesting part comes when those supernovae are combined with the cosmic microwave background and the clustering of galaxies. Allow the strength of dark energy to drift over cosmic time rather than holding it fixed, and the data prefer a dark energy that was weaker in the distant past and has been changing since. The combined dataset also sharpens what cosmologists call the dark energy figure of merit to 315, roughly 30 per cent better than the previous best.
How seriously should we take it? Here the paper earns real respect for its candour. Depending on how the statistics are handled, the preference for an evolving dark energy sits between 2.5 and 3.1 sigma, intriguing but short of the five sigma physicists treat as a discovery. More tellingly, under Bayesian model comparison, which penalises a model for every extra free parameter, that preference drops to weak. A longstanding tension in the data does ease if you let dark energy vary, but easing a tension is not the same as proving a case. The stakes are high: if dark energy genuinely changes, the cosmological constant is not constant.
Reference: Camilleri et al., "Supernovae Unite: Combining Pantheon+ and DES-SN5YR", arXiv:2609.05053
Hubble and Webb find the smallest worlds beyond Neptune
Out past Neptune lies a vast field of icy debris left over from the construction of the Solar System. The large members we know reasonably well. The small ones, which are the actual leftover building blocks, have been almost impossible to study, because brightness falls away with the square of size. The bodies carrying the most information are the ones we can barely see.
On 8 September, NASA announced results from two companion papers in The Astronomical Journal that push into that gap. Teams led by the doctoral researchers Anastasia Morgan at Northern Arizona University and Marielle Eduardo at the University of Victoria, with David Trilling at Northern Arizona, pointed Hubble and Webb at the same patches of sky, pairing visible light imaging with infrared. They picked out 27 previously unknown trans-Neptunian objects, the smallest roughly five kilometres across. Morgan's description of the challenge is hard to improve on: comparable to spotting fireflies on the Moon from Earth.
Two findings stand out. There are fewer small objects out there than several planet formation models predicted, which constrains how violently the early Kuiper Belt ground itself down. More surprisingly, the small objects share the same colours as their much larger counterparts, and the two distinct populations, the dynamically "hot" bodies and the "cold" ones, show strikingly similar size distributions despite forming in different regions of the disc. They appear to have remembered the conditions of their own origin rather than having their surfaces reset by a long history of knocks. If planetesimal formation produces much the same spread of sizes whether the disc was hot or cold, dense or diffuse, then the process is far less sensitive to its environment than anyone expected. That is worth knowing when you are working out how planets assemble around other stars.
Reference: NASA, Hubble and Webb observations of distant Solar System objects, 8 September 2026
An interstellar comet with a very cold childhood
3I/ATLAS, the third confirmed interstellar object to pass through our Solar System, has been among the most closely watched targets of the past year, including by remote observers following it from southern hemisphere sites like ours. A paper published in Monthly Notices of the Royal Astronomical Society on 7 September adds something new, by looking at the part of the comet that usually gets ignored: its tail.
Léa Ferellec at Northumbria University, with Cyrielle Opitom and Colin Snodgrass at the University of Edinburgh, used the WEAVE spectrograph on the William Herschel Telescope in La Palma to take spectra of the comet's plasma tail rather than its bright inner coma. WEAVE's large integral field unit collects spectra across an extended patch of sky at once, which suits a faint, diffuse ion tail far better than a narrow slit. They found five ion species: molecular nitrogen, carbon monoxide, carbon dioxide, water and a hydrocarbon.
The number doing the work is the ratio of molecular nitrogen to carbon monoxide, measured at greater than 0.023 with an uncertainty of 0.001. Molecular nitrogen is among the most volatile substances a comet can hold on to, so retaining that much is a strong constraint. The team concludes these ices must have frozen at a temperature colder than roughly minus 240 Celsius, placing 3I/ATLAS among the comets richest in nitrogen anyone has measured. The composition is a measurement; the conclusion drawn from it, that the comet formed far out from its parent star in a region resembling our own Kuiper Belt or Oort Cloud, is a reasonable inference rather than a certainty. Which star it came from, nobody can yet say.
Finding no moons at all, and why that counts as progress
David Kipping at Columbia University has spent much of his career hunting for moons around planets orbiting other stars, so far without a confirmed catch. His latest paper reports another absence, and makes a persuasive case that this one is worth more than it sounds.
The target was LP 890-9 c, a temperate rocky planet circling a very cool red dwarf roughly 105 light years away. Rather than relying on a single transit, Kipping combined twelve separate JWST observations of the planet crossing its star. Stacking is the key move: the noise that normally defeats a moon search, slow instrumental drifts and the restless variability of the star itself, does not repeat identically from one transit to the next, so it averages down while a genuine repeating signal would not. The outcome is a limit. Any moon larger than 0.1 Earth radii, a radius of about 640 kilometres, is excluded at 95 per cent confidence. That is the most sensitive exomoon search yet performed.
This is encouraging rather than deflating because astronomers had quietly worried that JWST might have a noise floor: some irreducible systematic error that would stop transit measurements improving however much telescope time you threw at them. This work suggests no such wall exists. Sensitivity keeps improving as observations accumulate, which is exactly the property you need if your goal is something small, faint and rare. The first confirmed exomoon is still ahead of us, but the road to it looks passable.
Four results, four small corrections to the picture. That is how most of astronomy actually proceeds. Clear skies, and we will see you next week.