The effect of environment on type Ia supernovae in the dark energy survey

Abstract

Analyses of type Ia supernovae (SNe Ia) have found puzzling correlations between their standardised luminosities and host galaxy properties: SNe Ia in high-mass, passive hosts appear brighter than those in lower-mass, star-forming hosts. As key cosmological probes, it is vital to understand galaxy-SN correlations, to ensure they do not bias measurements of cosmological parameters. In this thesis, I examine the host galaxies of SNe Ia in the Dark Energy Survey (DES). I use both spectroscopically- and photometrically-confirmed SN Ia samples (the DES3YR and DES5YR samples, respectively), measuring griz photometry of their host galaxies in both ‘local’ apertures centred on each SN, and for the entire ‘global’ host galaxy. I study the differences in the properties of these environments, such as stellar mass and rest-frame optical U − R colours, and their correlations with SN Ia parameters including Hubble residuals. For both DES3YR and DES5YR, all environmental property correlations with Hubble residual are significant at > 3σ, with the majority significant at > 5σ for DES5YR, for both local and global environment properties. U − R correlations are larger than for traditional global host mass, indicating that colour may better account for environmental effects. By analysing the scatter in the Hubble residuals, I also show that SNe Ia in redder, high-mass environments have higher scatter than SNe in bluer, low-mass environments, with a noticeably small scatter for blue SNe Ia in blue/low-mass environments. Such a sample currently presents the most homogeneous sample for use in cosmology. I find that redder SNe Ia also have larger Hubble residual steps than the bluer SNe, suggesting that they drive the overall step size. Investigating this further, I find suggestion that the global host galaxy stellar mass has the strongest relationship with SN colour, and by fitting for a two-component colour-dependent Hubble residual relationship (a simple approximation of a dust model), I am able to remove the mass step from the data. However, a statistically significant (2.9σ) step in local U − R remains, indicating that multiple environmental corrections may be needed to account for the dispersion in SN Ia luminosity. Finally, I investigate the differences in local properties for SN Ia siblings (SNe Ia occurring in the same host galaxy) in DES. In all cases, the larger the difference in local environment properties between SNe Ia in the same galaxy, the larger the differences in their Hubble residuals, and SN light-curve shape and colour. This emphasises the importance of local environmental corrections for SNe Ia standardisation.

More information

Identifiers

ORCID for Mark Sullivan: orcid.org/0000-0001-9053-4820

Catalogue record

Export record