On the impact of the supernova subsamples in reducing the Hubble tension

TL;DR

This work addresses the Hubble tension by testing whether mismatches between calibration and HF SN Ia samples in light-curve and host properties bias the local $H_{0}$ estimate. Using Pantheon+SH0ES data and a Tripp-like standardization with a mass step, the authors generate paired subsamples and perform Bayesian fits with $H_{0}$, $M_{B}$, $\alpha$, $\beta$, $\Delta_{host}$, and $\sigma_{int}$, while quantifying sample-matching quality via KS tests and extended metrics like $p_{MD}$ and Wasserstein-based distances. They find that light-curve parameters are similar across samples, but host properties differ (notably $M$ and sSFR), yielding a full-sample $H_{0} \approx 73.8$ km s$^{-1}$ Mpc$^{-1}$; when SN subpopulations are disentangled by stretch, persistent differences in $H_{0}$ and $M_{B}$ emerge between high- and low-stretch SNe, and the mass-step tends toward zero in well-matched subsamples. Incorporating subpopulations drastically increases the inferred $H_{0}$-uncertainty and can reduce the Hubble tension to about $2.5$–$3\sigma$ when separating the two SN channels, suggesting the mass-step may arise from combining distinct SN subpopulations rather than a single universal correction. Overall, the study highlights the importance of carefully matching SN subpopulations and host-galaxy properties in local $H_{0}$ determinations and provides a path toward more robust, population-aware luminosity standardization.

Abstract

The persistent 4-6$σ$ difference between early- and late-time Hubble constant ($H_{0}$) measurements, known as the "Hubble tension," is a major problem in modern cosmology. We study how differences in colour ($c$), stretch ($x_{1}$), and host galaxy properties-stellar mass ($M$) and specific star formation rate (sSFR)-between calibration and Hubble Flow (HF) Type Ia supernova (SN Ia) samples used by SH0ES affect SN luminosity standardization and $H_{0}$ estimates. We generate subsamples from both, estimating $H_{0}$, $M_{B}$, $α$, $β$, $Δ_{host}$, and $σ_{int}$. We use Kolmogorov-Smirnov to assess the consistency between subsamples and reveal how parameter estimates change as sample matching improves. The calibration sample is not fully representative of the HF sample, especially in $M$ and sSFR. Improving sample consistency leads to changes in $H_{0}$, $M_{B}$, $α$, and $σ_{int}$, though overall values remain broadly stable. Better-matched subsamples tend to yield a mass step consistent with zero within 1$σ$. By disentangling SN subpopulations, we find persistent differences in $H_{0}$ ($\sim$2-3$σ$) and $M_{B}$ ($\sim2σ$) between low- and high-stretch SNe: $H_{0} = 75.27 \pm 1.18$ km s$^{-1}$ Mpc$^{-1}$ for low-stretch and $H_{0} = 71.25 \pm 1.59$ km s$^{-1}$ Mpc$^{-1}$ for high-stretch, resulting in Hubble tensions of 6.07$σ$ and 2.52$σ$. These differences suggest SNe Ia subpopulations with varying dust and intrinsic colour not captured by a single $β$, impacting cosmology. Estimating a single $H_{0}$ for both subpopulations yields $H_{0} = 73.78 \pm 2.17$ km s$^{-1}$ Mpc$^{-1}$, with a much larger uncertainty that lowers the Hubble tension from $5.87σ$ to $\sim2.86σ$. Our results suggest that the mass step may arise from an over-correction of more than one SN subpopulation associated to different environments.

Figures (17)

• Figure 1: Strech ($x_{1}$) as a function of the logarithm of the stellar mass ($M$) in solar units ($M_\odot$) for supernovae from both calibration (blue, open circles) and Hubble flow (red, filled circles) samples. The dashed horizontal lines correspond to the median value of the light curves parameters distributions and the dashed-dotted vertical lines to the median value of the host properties distributions for each sample depending on the colour. The K-S test $p$-values obtained by comparing the parameters distributions of the calibration and Hubble flow samples are also displayed at the top of the Figure.
• Figure 2: Similar to Fig. \ref{['fig:x1_vs_mass']} for colour ($c$) and the logarithm of the specific star formation rate (sSFR).
• Figure 3: Redshift ($z$) distribution for both calibration (blue, delineated by dashed lines) and Hubble Flow (red, delineated by solid lines) SNe samples. The redshifts are corrected for both the CMB and peculiar velocities (VPEC) and are provided in the Pantheon+SH0ES compilation Riess_2022Brout_2022. The vertical dash-dotted lines indicate the redshift medians of each sample.
• Figure 4: $H_{0}$ (left) and $\sigma_{int}$ (right) as a function of the generated subsample size, using paired subsamples drawn from the calibration and HF samples. Grey dots show the full subsample distribution, lilac dots highlight subsamples with $p_{MD}$ above the 60th percentile, and orange open dots indicate those above the 80th percentile.
• Figure 5: Left panel: $H_{0}$ as a function of the $p_{MD}$ obtained for each generated calibration subsample when compared to the corresponding HF subsample. The yellow line represents the rolling median of each parameter, computed using a window size of 100 subsamples, while the shaded area represents its rolling standard deviation. The red dots correspond to the median values of the parameter distribution using subsamples with $p$-values higher than 0.05, 0.1, 0.2 and 0.4 . The red error bars indicate the difference between the 16th and 84th percentiles relative to the median values estimated for each bin. The dashed red line marks $p_{MD}$ = 0.05 and $p_{MD}$ = 0.4 on the logarithmic scale. Right panel: Same analysis for $\Delta_{\rm host}/\sigma_{\Delta_{\rm host}}$, where the dark blue dots represent subsamples with $p_{MD}$ below 0.05, medium blue those with $p_{MD}$ between 0.05 and 0.4, and in light blue those with $p_{MD}$ above 0.4. The black dashed line indicates the value of $\Delta_{\rm host}/\sigma_{\Delta_{\rm host}}$ estimated using the full SNe sample, which is 2.22$\sigma$. The grey shaded area highlights the region where the mass-step is consistent with 0 within $1\sigma$.