CIGaRS I:Combined simulation-based inferencefrom SNæ Ia and host photometry

Konstantin Karchev, Roberto Trotta, Raul Jimenez · 2025

Read CIGaRS I:Combined simulation-based inferencefrom SNæ Ia and host photometry by Konstantin Karchev, Roberto Trotta, Raul Jimenez (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-15899.txt.

Introduction

Konstantin Karchev, Roberto Trotta, and Raul Jimenez build a joint model for type Ia supernovae and the galaxies that host them. These bright blasts help measure the cosmos, but their light needs fixes tied to the host. The model infers, from photos alone, how blast light depends on the age and metal mix of the parent stars. It also infers how blast rates change with age, plus basic facts about the cosmos.

The model folds in star birth, dust, and the odds that a blast gets seen. On fake data with about sixteen thousand blasts, a neural method read off all values with care. Photo-based distances to hosts were sharp, and limits on the cosmos grew about four times tighter. The authors aim the method at a large sky survey era.

Catalog record

CREATORKonstantin Karchev, Roberto Trotta, Raul Jimenez
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE79 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-15899

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

CIGaRS I:Combined simulation-based inferencefrom SNæ Ia and host photometry Konstantin Karchev, Roberto Trotta, Raul Jimenez Abstract Using type Ia supernovæ (SNæ Ia) as cosmological probes requires empirical corrections, which correlate with their host environment. We present a unified Bayesian hierarchical model designed to infer, from purely photometric observations, the intrinsic dependence of SN Ia brightness on progenitor properties (metallicity & age), the delay-time distribution (DTD) that governs their rate as a function of age, and cosmology, as well as the redshifts of all hosts. The model incorporates physics-based prescriptions for star formation and chemical evolution from Prospector - β, dust extinction of both galaxy and SN light, and observational selection effects. We show with simulations that intrinsic dependences on metallicity and age have distinct observational signatures, with metallicity mimicking the well-known step of SN Ia magnitudes across a host stellar mass of ∼ 1 × 10 10 M ⊙ \sim 1\text{\times}{10}^{10}\text{\,}\mathrm{M_{\odot}} . We then demonstrate neural simulation-based inference of all model parameters from mock observations of ≈ 16 000 \approx 16\,000 SNæ Ia and their hosts up to redshift 0.9 0.9 . Our joint physics-based approach delivers robust and precise photometric redshifts ( < 0.01 <0.01 median scatter) and improved cosmological constraints, unlocking the full power of photometric data and paving the way for an end-to-end simulation-based analysis pipeline in the LSST era. \unnumbered equation]ext:Moreno-Raya_2016a:eqn:nickel equation]ext:Alsing_2023a:eqn:tau2 equation]ext:Alsing_2023a:eqn:delta equation]ext:Salim_2018:eqn:Rv-dustd subsection]ext:Karchev-starnre:sec:detprob keywords cosmology, dark matter and dark energy, galaxies and clusters, transient astrophysical phenomena 1 Introduction It has long been known that the brightnesses, stretches, and colours of type Ia supernovæ (SNæ Ia) correlate with global properties of their host galaxies, e.g. morphology, stellar mass, star-formation rate (SFR), metallicity, and/or stellar age ( Sullivan et al. 2003 ; Sullivan et al. 2006 ; Aubourg et al. 2008 ; Neill et al. 2009 ; Brandt et al. 2010 ; Lampeitl et al. 2010 ; Sullivan et al. 2010 ; Childress et al. 2013 ; Johansson et al. 2013 ; Lee et al. 2022 ; Chung et al. 2023 ; Chung et al. 2024 ) . Similar dependences on the SNæ’s local environment have also been found ( Rigault et al. 2013 ; Rigault et al. 2015 ; Rigault et al. 2020 ; Jones et al. 2015 ; Jones et al. 2018 ; Moreno-Raya et al. 2016a ; Moreno-Raya et al. 2016b ; Kim et al. 2018 ; Kim et al. 2019 ; Roman et al. 2018 ; Rose et al. 2019 ; Kelsey et al. 2021 ; Kelsey et al. 2023 ; Ginolin et al. 2024a ; Ginolin et al. 2024b ) . In current cosmological analyses ( Brout et al. 2022 ; Rubin et al. 2025 ; Vincenzi et al. 2024 , e.g.) , which use SNæ Ia as standardisable candles, it is common practice to (attempt to) correct for all these effects en masse via an ad-hoc “mass step” : a magnitude offset between SNæ Ia hosted in galaxies with stellar masses above and below a given threshold, usually set to the sample median or fixed around {10}^{10}\text{\,}\mathrm{M_{\odot}} . The rationale is that stellar mass is a simple-to-estimate proxy for nearly all host properties. Unveiling the physical processes underlying host – SN Ia connections would thus not only enhance our understanding of SNæ Ia and their formation, but also deliver more accurate and precise constraints on dark energy by improving the standardisation procedure. A supernova’s environment also has extrinsic effects on its light, which is dimmed and reddened by dust along the line of sight, both in the immediate surroundings of the progenitor system and in interstellar space within the host. It can also be scattered into the line of sight, significantly affecting finite-resolution photometry ( Duarte et al. 2025 ) . Accounting for dust requires careful hierarchical modelling ( Mandel et al. 2017 ) since a) extinction has a similar effect (redder – dimmer) to intrinsic colour-related standardisation (bluer – brighter) and b) there is a well-known relationship between the amount and properties of the dust in a galaxy and its SFR and stellar mass: the so-called “star-forming main sequence” ( Speagle et al. 2014 , see e.g.) . Different studies have come to conflicting conclusions regarding this interplay: for example, Brout and Scolnic 2021 attribute the mass step to differences in dust in low- and high-mass hosts rather than any intrinsic/physical/causal effect, while Thorp and Mandel 2022 find a preference for a single dust law across stellar masses and a mass step of \approx 0.06\pm\pm 0.03\text{\,}\mathrm{mag} , with more massive galaxies hosting intrinsically brighter SNæ Ia. However, Bayesian model comparison using the same data and models ( Karchev et al. 2023 ) mildly disfavours a non-zero mass step with posterior odds of 2 : 1 . Resolving these degeneracies and arriving at a definite conclusion regarding host – SN Ia connections requires tackling all interrelated effects simultaneously, including the apparent dimming due to cosmological distance and selection biases. While Bayesian hierarchical modelling (BHM) provides the principled framework for doing this, traditional Markov Chain Monte Carlo (MCMC) methods require sampling jointly the vast number of latent parameters (tens per SN ) and explicit computation — at every step — of the selection probability, which is intractable in realistic scenarios. To date, analyses either resort to ad-hoc approximations ( Rubin et al. 2015 ; Rubin et al. 2025 ; Hinton et al. 2019 ) or are split in two separate steps ( Thorp et al. 2021 ; Thorp and Mandel 2022 , see e.g.) : first, the SNæ’s distance moduli (i.e. intrinsic brightnesses) are estimated and corrected for selection effects using a fiducial, possibly mis-specified, model ( Kessler and Scolnic 2017 ) ; then, they are correlated with galaxy properties obtained in a separate analysis (rarely considering the associated uncertainties ( Shariff et al. 2016 ) ) and/or used in a cosmological fit. A two-step approach cannot account for many important statistical effects. For instance, the ages of SN Ia progenitors are not representative of their hosts’ stellar populations, and so regressing using the latter may be misleading. Moreover, standardisation implicitly gives higher weight to more massive hosts, where a larger number of SNæ occur — an instance of Eddington bias ( Eddington 1913 , see also Karchev and Trotta 2024 ) . Likewise, a preference for detecting ( apparently ) brighter objects biases the selected sample towards intrinsically brighter SNæ to a greater extent in dustier hosts, which are also more massive. Lastly, an apparent correlation — non-existent intrinsically — between SN brightnesses and properties of the hosts may be introduced by the use of photometric redshifts in the standardisation process due to the mass – age – redshift degeneracy. Similarly, an apparent redshift dependence of SN properties ( Lee et al. 2020 ; Nicolas et al. 2021 ) might arise due to the evolution of the host galaxies, e.g. their dust content ( Popovic et al. 2024 ; Thorp et al. 2024 ) or metallicity or age. In this paper, we present the first framework for combined inference and galaxy-related standardisation (CIGaRS) of SNæ Ia and their hosts, which addresses all of the above conceptual and methodological issues and can unequivocally disentangle intrinsic and extrinsic effects through physics-based forward

Read the complete text