Dark Energy May Not Be Constant, New Supernova Survey Suggests

For decades, cosmologists have relied on a remarkably simple idea to explain the accelerating expansion of the universe: dark energy behaves like a cosmological constant. But a new analysis of one of the largest collections of exploding stars yet assembled is adding to growing evidence that the story may be more complicated.

A newly released study combining major Type Ia supernova datasets has found a preference for a model in which dark energy changes with cosmic time. The result does not prove that dark energy is evolving, but it strengthens a scientific mystery that has been building since earlier results from the Dark Energy Survey and the Dark Energy Spectroscopic Instrument (DESI).

Supernovae as cosmic mileposts

Type Ia supernovae are among astronomers' most useful tools for measuring cosmic expansion. Because their intrinsic brightness can be calibrated, researchers can compare how bright these explosions appear from Earth with how much their light has been redshifted by the expansion of space.

By assembling thousands of these explosions across different distances, scientists can construct a Hubble diagram—essentially a record of how the universe expanded over billions of years.

The new Unite compilation combines data from the Pantheon+ and Dark Energy Survey Supernova Year 5 (DES-SN5YR) samples. It contains 2,884 likely Type Ia supernovae, making it the most comprehensive internally consistent supernova dataset of its kind, according to the researchers.

A possible crack in the standard model

The standard cosmological model, known as ΛCDM, assumes that dark energy is represented by a cosmological constant, Λ. In its simplest form, its density does not change as the universe expands.

The new analysis instead finds that observations can be better accommodated when the dark-energy equation of state is allowed to vary with time. When the supernova compilation is combined with cosmic microwave background and baryon acoustic oscillation measurements, the researchers report parameters of approximately w₀ = −0.861 and wₐ = −0.60 for their evolving-dark-energy model. 

That is intriguing because independent observations have been pointing in a similar direction.

The DESI experiment has previously found indications that its measurements of the universe's large-scale structure are not perfectly described by constant dark energy when combined with certain supernova and cosmic microwave background datasets. Earlier Dark Energy Survey results also left open the possibility that dark energy could vary over time. 

But scientists are not declaring victory

The most important caveat is that this is not yet a discovery of evolving dark energy.

The new Unite analysis reports only a modest statistical preference over the standard model: depending on the statistical method used, the preference reaches roughly 2.5 to 3.1 sigma. Bayesian model comparison, meanwhile, indicates only a weak preference for an evolving model.

Other recent analyses have likewise urged caution. A 2026 study using DESI data and multiple supernova samples concluded that current observations remain compatible with several dark-energy scenarios and that it may still be premature to claim statistically significant evidence for evolution. 

That distinction matters. In physics, intriguing deviations from a well-tested model can disappear as datasets become larger, calibrations improve and previously overlooked systematic errors are understood.

Why the result matters

If future observations confirm that dark energy evolves, the consequences would be profound.

The cosmological constant has become the simplest explanation for the accelerating universe. If it is not constant, physicists would need to consider alternatives, such as a dynamic field sometimes called quintessence, modifications to gravity, or other physics beyond the standard cosmological model.

It could also force scientists to rethink how the universe's expansion has changed over its 13.8-billion-year history.

For now, however, the safest conclusion is more restrained: the evidence is becoming harder to ignore, but it is not yet decisive.

The next generation of observations

The question is likely to become clearer as new surveys deliver larger and better-calibrated datasets. DESI is continuing to map the distribution of galaxies and quasars, while NASA's Nancy Grace Roman Space Telescope is designed in part to investigate dark energy and the expansion history of the universe.

The coming years could therefore be crucial. If independent observations repeatedly point toward the same departure from ΛCDM, cosmologists may eventually have to abandon the assumption that dark energy is constant.

For now, the universe has not rewritten its rulebook—but it may be hinting that the rulebook is more complicated than we thought.