The word doing all the work is 'constant'

Companion blog post for S05E193 — Monday, 14 September 2026 | astronomydaily.io

An Australian-led team has rebuilt three decades of supernova observations into a single consistent catalogue of 2,884 exploding stars. It is the most careful look yet at whether dark energy holds still — and the answer is getting less comfortable.

When two teams discovered in 1998 that the expansion of the universe was speeding up, they gave the cause a placeholder name — dark energy — and got on with measuring it. Twenty-eight years later the placeholder is still there. We know what dark energy does; we have no settled idea what it is. And the simplest guess, the one baked into the standard model, is that whatever it is, it is constant: a fixed energy density belonging to empty space itself, the same strength now as ten billion years ago. That assumption has been quietly wobbling for about two years, and new work out of the University of Queensland has just given it another push.
The paper — ‘Supernovae Unite: Combining Pantheon+ and DES-SN5YR’, published in Publications of the Astronomical Society of Australia — is the work of PhD candidate Ryan Camilleri and Professor Tamara Davis, with collaborators across Australia, the United States, the United Kingdom, South Africa, Spain and France. What they have produced is the largest and most internally consistent catalogue of Type Ia supernovae ever assembled: 2,884 of them.
The number is impressive. The method is the actual achievement. Type Ia supernovae work as distance markers because they explode at close to the same intrinsic brightness every time. But the ones we have were collected over thirty years by dozens of surveys on dozens of telescopes, with different detectors, filters and calibrations. Pour all of that into one bucket and the systematic errors will swamp any cosmology you try to do with it.
So Camilleri’s team didn’t pour. They took Pantheon+, the big historical compilation, and the Dark Energy Survey’s full five-year sample of roughly 1,500 additional high-redshift supernovae, and rebuilt both from the same starting assumptions in one framework. “We’ve rebuilt three decades of astronomical observations into a single, consistent framework,” Camilleri says — and this is reanalysis, not tidying: “Over the years we’ve learned a lot more about how supernovae behave, so we’ve been able to go back and apply that improved understanding to older data.” The two worst offenders — cosmic dust, which dims light in ways that mimic distance, and host-galaxy mass, which correlates with supernova brightness for reasons still not understood — had to be handled the same way across the whole sample. The team thought the second serious enough to publish a companion paper on it.
The first result is reassuring. Taken alone, under a flat universe with constant dark energy, the catalogue returns a matter density of 0.310 — squarely in line with everything else we know. Nothing is broken.
The second result is the interesting one. Fold in the cosmic microwave background and baryon acoustic oscillations — the two other great pillars of cosmological measurement — and under constant dark energy the three data sets start pulling against each other. That tension eases if dark energy is allowed to change with time. The team find a preference for evolving dark energy over the standard model at between 2.5 and 3.1 sigma, and their compilation tightens the uncertainties on the dark-energy parameters by around 30 per cent.
Sigma deserves care. Three sigma is roughly a one-in-seven-hundred chance of seeing data like this if the standard model is right. Worth chasing; not a discovery. Particle physics reserves that word for five sigma, and cosmology has been burned by three-sigma results before.
What makes this one land differently is company. “Our supernova data from DES in 2024 first showed hints that dark energy may be time varying, and this new compilation also sees a deviation from the standard model,” Davis says. “So, two completely independent measurements have found hints of time variation in dark energy.” Add DESI, reporting something similar from baryon acoustic oscillations rather than supernovae, and you have three approaches with different failure modes producing the same faint smell of something wrong.
If it holds, the consequences run deep. A cosmological constant is the simplest thing dark energy could be. A dark energy that changes strength over cosmic time is a dynamic field with a history and a future, and it changes how this universe ends. Davis goes further: all of this, she suggests, “may also hold the clue to explain how gravity and quantum physics fit together.”
There is a geography here that is easy to miss. The work was led from Brisbane, with ANU and Swinburne on the author list, and published in Australia’s own journal. The Dark Energy Survey data at its core came off the Blanco 4-metre telescope at Cerro Tololo in the Chilean Andes. And the Nobel-winning observation that started the argument was made, in part, by Brian Schmidt at Mount Stromlo. What dark energy is has been a Southern Hemisphere argument from the beginning.
It will not stay open much longer. The Vera Rubin Observatory, also in Chile, is about to begin finding these supernovae in industrial quantities, and the Nancy Grace Roman Space Telescope — launched a fortnight ago — was built in large part to pin this measurement down. Within a few years we will know whether dark energy has been quietly changing all along, or whether 2,884 beautifully calibrated supernovae simply made everyone nervous for a while.
Either way, somebody had to do thirty years of homework first.

Sources: University of Queensland news release, 8 September 2026. Camilleri, Lee, Davis, Rubin, Shah, Scolnic, Lidman et al., ‘Supernovae Unite: Combining Pantheon+ and DES-SN5YR’, Publications of the Astronomical Society of Australia (arXiv:2609.05053); companion host-galaxy-mass paper arXiv:2609.05321.