Could Supernova Ages Explain Away Cosmic Acceleration?

Type Ia supernovae are among astronomy’s most important distance markers. The current dispute does not mean dark energy has been disproved. It raises a narrower but consequential question: could a change in supernovae’s standardized brightness with the age of their progenitor stellar populations resemble the distance–redshift pattern normally interpreted as accelerating cosmic expansion?

The claim from the Pantheon+ reanalysis

The contested analysis applies a progenitor-age-dependent correction to 1,701 Type Ia supernovae in the Pantheon+ catalogue. With that adjustment, its authors report that the monopole component of the deceleration parameter becomes positive. In plain terms, the Universe would still be expanding, but its expansion would be slowing rather than speeding up.

The underlying idea is straightforward. Type Ia supernovae are made into reliable, or “standardizable,” distance indicators by correcting their observed brightness using features of their light curves and colour. But if a residual brightness effect depends on progenitor age—and if those stellar populations systematically differ at higher redshift—some of the observed trend could reflect astrophysical evolution rather than cosmic acceleration.

If confirmed, that would weaken the case for acceleration specifically from supernova distances. It would not, by itself, overturn the wider Lambda-CDM cosmological framework or invalidate other measurements used to test it.

Why Schmidt, Riess and colleagues disagree

The opposing analysis agrees that host-galaxy properties must be accounted for. Its disagreement is over whether the proposed age correction is large enough, and sufficiently well established, to remove the evidence for acceleration.

The researchers defending the robustness of supernova cosmology make three main points:

• A galaxy’s age is not generally the same as the age of the particular star that eventually exploded. Treating host-galaxy age as progenitor age may therefore create a misleading effect.
• The disputed method’s estimated difference in progenitor ages between nearby and distant supernovae—about 5 billion years—is said to be overstated by a factor of three to five.
• The usual host-galaxy mass correction already accounts for known environmental dependencies that are correlated with stellar age. After applying that correction, the authors report no dependence of standardized supernova brightness on host age in the sample they studied.

There is still a substantive scientific disagreement on that last point. The original work on progenitor-age bias argues that the commonly used “mass step” correction does not fully remove the effect, because progenitor age and host mass evolve differently with redshift. The issue, then, is not simply whether corrections should be made. It is which physical quantity is being measured accurately and which model of a supernova’s environment best represents the data.

An independent reanalysis still finds acceleration

A later reanalysis, which reports correcting a coordinate error, found that the uncorrected Pantheon+ data are consistent with Lambda-CDM. Even after applying the disputed age correction, it obtained a monopole deceleration parameter of (q_m=-0.267). That remains a negative value, meaning accelerated expansion—though weaker acceleration than without the correction. 17

That result does not end the debate. But it shows why it is premature to treat a decelerating Universe as an established finding. The conclusion is highly sensitive to calibration choices, the link drawn between galaxy age and progenitor age, and the treatment of host-galaxy corrections.

How Rubin Observatory’s LSST could provide a decisive test

The Vera C. Rubin Observatory’s Legacy Survey of Space and Time, or LSST, is expected to measure at least 500 Type Ia supernovae per observing season and to produce tens of thousands of well-measured light curves out to roughly (z\sim1) over its 10-year survey. 9 Other forecasts anticipate the discovery of nearly 1 million Type Ia supernovae overall. 4

A vastly larger sample will not automatically resolve a systematic-error problem. It can, however, enable much sharper tests.

1. Compare like with like at the same redshift. Supernovae at roughly the same redshift but in hosts with different ages, masses, metallicities and star-formation histories can reveal whether a brightness difference remains after standard corrections.
2. Fit astrophysics and cosmology together. Rather than assuming population evolution is either absent or already known, researchers can fit it alongside cosmological parameters and test whether the data still require acceleration.
3. Use a more uniform observing programme. A single large survey can reduce some of the cross-survey calibration differences that may otherwise masquerade as trends with redshift.

The crucial caveat is that more data do not erase systematic biases. Most LSST events are not expected to have individual spectroscopic classifications and redshifts. Reliable photometric classification, host-galaxy redshifts, selection-bias controls and precise photometric calibration will therefore be essential. 4

What is really at stake

The cleanest test will be an independently calibrated Hubble diagram with well-characterized host galaxies: does it still require acceleration when flexible, empirically tested age and environmental corrections are allowed?

If it does, progenitor-age bias cannot be the main explanation for the supernova signal. If it does not, interpreting supernovae as direct evidence for late-time acceleration would need major revision. For now, the conflicting analyses point to an active methodological debate—not the final collapse of dark energy. 17