Strong Progenitor Age-bias in Supernova Cosmology. II. Alignment with DESI BAO and Signs of a Non-Accelerating Universe

TL;DR

This paper tackles the problem that Type Ia SN luminosity standardization may depend on progenitor age, a bias that can mimic or erode cosmic acceleration. By adopting an age-bias slope of $s \approx -0.030 \pm 0.004$ mag Gyr$^{-1}$ and modeling the redshift evolution of SN progenitor ages from the cosmic star-formation history and SN delay-time distribution, the authors correct SN distances and assess the impact on cosmological inferences. In a flat-$w_0w_a$CDM framework, SN data corrected for age-bias align with DESI BAO+CMB results and reveal strong tension with the cosmological-constant model $\Lambda$CDM ($>9$–$11\sigma$ Mahalanobis distance), favoring a time-varying dark energy equation of state. The joint analysis across SN, BAO, and CMB shows a new concordance, with SN data pushing $w_0$ higher and $w_a$ more negative, and implying a present universe closer to non-accelerating expansion; an evolution-free test using only young host galaxies further supports the dynamical-DE scenario. These findings have significant implications for cosmology, including potential implications for the Hubble tension and the interpretation of SN as standard candles in a universe with evolving dark energy.

Abstract

Supernova (SN) cosmology is based on the key assumption that the luminosity standardization process of Type Ia SNe remains invariant with progenitor age. However, direct and extensive age measurements of SN host galaxies reveal a significant (5.5σ) correlation between standardized SN magnitude and progenitor age, which is expected to introduce a serious systematic bias with redshift in SN cosmology. This systematic bias is largely uncorrected by the commonly used mass-step correction, as progenitor age and host galaxy mass evolve very differently with redshift. After correcting for this age-bias as a function of redshift, the SN dataset aligns more closely with the w0waCDM model recently suggested by the DESI BAO project from a combined analysis using only BAO and CMB data. This result is further supported by an evolution-free test that uses only SNe from young, coeval host galaxies across the full redshift range. When the three cosmological probes (SNe, BAO, CMB) are combined, we find a significantly stronger (> 9σ) tension with the ΛCDM model than that reported in the DESI papers, suggesting a time-varying dark energy equation of state in a currently non-accelerating universe.

Figures (10)

• Figure 1: Correlation between population age and HR for SN host galaxies based on our new age measurements reported in Paper I 2025MNRAS.538.3340C. The HR is a measure of relative luminosity when the SN sample is confined to a narrow redshift range. The left panel shows the 2019ApJ...874...32R sample, for which the strong significance of the correlation was originally reported by 2020ApJ...903...22L and has been repeatedly confirmed by third parties at the $>5\sigma$ level 2021MNRAS.503L..33Z2023SCPMA..6629511W. The right panel shows a larger sample ($N \sim 200$) of host galaxies (from 2011ApJ...740...92G sample) over a broader redshift range ($z < 0.45$), confirming the universal nature of the age-bias. The results shown are based on the LINMIX analysis, while the slopes and significances obtained from the full age posteriors are given in parentheses (see 2025MNRAS.538.3340C).
• Figure 2: Evolution of stellar population age with redshift. The result is similar to Figure 6 of 2022MNRAS.517.2697L, but for the $w_0w_a$CDM model, as suggested by a combined analysis of DESI BAO and CMB 2025arXiv250314738D. The distribution functions (blue lines) at $z = 0.0$, 0.5, 1.0, and 1.5 represent the SN progenitor age distributions. The black solid line indicates the median age of SN progenitors, while the black dashed line shows the mass-weighted mean age of the stellar population, obtained from the cosmic star formation history and compared with the observed data (red and blue circles). In the right panels, the solid and dashed lines show, respectively, the redshift evolution of the median progenitor age and the mean stellar population age, relative to $z = 0$, and the corresponding variations in Hubble residual, obtained by multiplying the age difference with the age-bias slope $|s|$.
• Figure 3: The residual Hubble diagram before (top panel) and after (bottom panel) the age-bias correction. The corrections are applied to the observational data from the DES SN project 2024ApJ...973L..14D, using the $\Delta\mathrm{HR}/\Delta\mathrm{age}$ slope ($0.030\pm0.004$ mag Gyr$^{-1}$) and the redshift evolution of the median progenitor age relative to $z = 0.0$, as shown in Figure \ref{['f2']}. After the correction, the SN dataset no longer supports the $\Lambda$CDM model (red solid lines), but instead shows better consistency with a time-varying dark energy equation of state, as described by the flat-$w_0w_a$CDM model favored by the combined BAO and CMB analysis (blue solid lines). As shown in the top panel, this model from BAO+CMB alone deviates significantly from that based on the combined analysis of BAO, CMB, and uncorrected SN data (green dashed line). In both cases, the relevant cosmological parameters are adopted from 2025arXiv250314738D. A binning scheme similar to that of the DES SN project was implemented to ensure $\sim$50 SNe per each bin. In the bottom panel, an uncertainty in the age-bias slope was propagated into the total error budget of the binned data. A similar trend is also obtained when using the Pantheon+ dataset 2022ApJ...938..110B.
• Figure 4: Similar to Figure \ref{['f3']}, but showing the comparison between SN and BAO distance measurements. The BAO distances are taken from 2025arXiv250314738D and 2025arXiv250306712D, and have been converted to distance moduli. After correcting for the progenitor age-bias (bottom panel), the SN and BAO distance scales show good agreement. For a self-consistent comparison, we adopt the value of the product $H_0 \cdot r_d$ from the combined DESI BAO (DR2) and CMB analysis within the flat-$w_0w_a$CDM model (see text for details).
• Figure 5: Confidence contours (68$\%$ and 95$\%$) in the $\Omega_m - w$ plane for the flat-$w$CDM model. The DES5Y SN data, shown before and after the age-bias correction (left and right panels, respectively), is compared with BAO and CMB data. In the left panel, a small overlap near $w=-1$ weakly supports the $\Lambda$CDM model when all three datasets are combined. In the right panel, after applying the age-bias correction, the SN contours shift upward, favoring a significantly larger value of $w$ ($\sim-0.56$). The three probes no longer intersect near $w=-1$, but instead show a progression in the mean $w$ values: from the CMB ($w \approx -1.5, z\approx 1090$), to BAO ($w \sim -0.92, 0.4 < z < 4$), to SNe Ia ($w \sim -0.56, 0 < z < 1.5$), suggesting a redshift evolution of the dark energy equation-of-state parameter.