Characterising the Standardisation Properties of Type Ia Supernovae in the z band with Hierarchical Bayesian Modelling

TL;DR

This paper establishes the first full Bayesian analysis of Type Ia supernova standardisation in the z band, leveraging a joint Foundation-YSE sample and BayeSN to quantify how z-band peak magnitudes, after dust and shape corrections, perform as distance indicators. It demonstrates a measurable, wavelength-spanning mass step from optical to z band, finds a robust z-band shape-luminosity relation, and shows that z-band data improve distance estimates when combined with optical data, albeit not matching the precision of near-infrared bands. The work provides a transparent, scalable framework for incorporating z-band information into cosmological analyses and underpins future rest-frame z observations from Rubin-LSST and Roman Space Telescope. It also reinforces that dust alone cannot fully explain the host-mass dependence of SN Ia brightness, motivating larger, multi-wavelength samples and forward-modeling comparisons.

Abstract

Type Ia supernovae (SNe Ia) are standardisable candles: their peak magnitudes can be corrected for correlations between light curve properties and their luminosities to precisely estimate distances. Understanding SN Ia standardisation across wavelength improves methods for correcting SN Ia magnitudes. Using 150 SNe Ia from the Foundation Supernova Survey and Young Supernova Experiment, we present the first study focusing on SN Ia standardisation properties in the z band. Straddling the optical and near-infrared, SN Ia light in the z band is less sensitive to dust extinction and can be collected alongside the optical on CCDs. Pre-standardisation, SNe Ia exhibit less residual scatter in z-band peak magnitudes than in the g and r bands. SNe Ia peak z-band magnitudes still exhibit a significant dependence on light-curve shape. Post-standardisation, the z-band Hubble diagram has a total scatter of RMS $ = 0.195$ mag. We infer a z-band mass step of $γ_{z} = -0.105 \pm 0.031$ mag, which is consistent within $1σ$ of that estimated from gri data, assuming $R_{V} = 2.61$. When assuming different $R_{V}$ values for high and low mass host galaxies, the z-band and optical mass steps remain consistent within $1σ$. Based on current statistical precision, these results suggest dust reddening cannot fully explain the mass step. SNe Ia in the z band exhibit complementary standardisability properties to the optical that can improve distance estimates. Understanding these properties is important for the upcoming Vera Rubin Observatory and Nancy G. Roman Space Telescope, which will probe the rest-frame z band to redshifts 0.1 and 1.8.

Figures (10)

• Figure 1: The distribution of redshifts (top) and host galaxy masses (bottom) for the combined YSE and Foundation sample (shaded blue) compared to that for the CSP-I Contreras_2010Stritzinger_2011, iPTF Johansson_2021, and CfAIR2 Wood_Vasey_2008Friedman_2015 NIR samples.
• Figure 2: The distributions of $\theta$ (left) and $A_{V}$ (right) for the sample of 150 SNe Ia by survey. The estimates of $\theta$ and $A_{V}$ for each object come from the BayeSN fit to the $griz$ light curve data, as discussed in §\ref{['sec:methods']}.
• Figure 3: The star formation rate (SFR) vs. stellar mass for the host galaxies of the final sample of YSE and Foundation SNe Ia. The markers show the mean and the error bars show the standard deviation of the posterior samples from the Prospector-$\alpha$ fit Leja_2017Johnson_2021 from BlastJones_2024.
• Figure 4: An SN Ia at $z=0.059$ observed by YSE. Each subplot shows the data in an individual filter (upper right: $g$ band, upper left: $r$ band, lower right: $i$ band, lower left: $z$ band). The BayeSN fit to the $griz$ data is shown with a blue solid line, the fit to the $gri$ data is shown with a green dashed line, and the fit to the $z$-band only data is shown with an orange dot-dashed line. The YSE light curves that pass the data cuts have a median of 5 $z$-band observations. The Foundation light curves that pass the data cuts have a median of 6 $z$-band observations.
• Figure 5: The $z$-band peak intrinsic absolute magnitude, $M_{z, \text{int}} = m_{z, \text{int}} - \mu_{\Lambda \text{CDM}}(z)$, as a function of the inferred shape parameter from a BayeSN fit to optical $gri$ data only, $\theta_{gri}$. The fitted shape-luminosity relation is shown as a solid line. The shaded region shows the fitted residual scatter around this relation.