Reconsidering the effects of local star formation on type Ia supernova cosmology

Recent studies found a correlation with ∼3σ significance between the local star formation measured by GALEX in SN Ia host galaxies and the distances or dispersions derived from these SNe. We search for these effects by using data from recent cosmological analyses to greatly increase the SN Ia sample; we include 179 GALEX-imaged SN Ia hosts with distances from the Joint Light-curve Analysis (JLA) and Pan-STARRS SN Ia cosmology samples and 157 GALEX-imaged SN Ia hosts with distances from the Riess et al. H0 measurement. We find little evidence that SNe Ia in locally star-forming environments are fainter after light curve correction than SNe Ia in locally passive environments. We find a difference of 0.000 � 0.018 (stat+sys) mag for SNe fit with SALT2 and 0.029 � 0.027 (stat+sys) mag for SNe fit with MLCS2k2 (), which suggests that proposed changes to recent measurements of H0 and w are not significant and numerically smaller than the parameter measurement uncertainties. We measure systematic uncertainties of ∼0.01-0.02 mag by performing several plausible variants of our analysis. We find the greatly reduced significance of these distance modulus differences compared to Rigault et al. results from two improvements with fairly equal effects, our larger sample size and the use of the JLA and Riess et al. sample selection criteria. Without these improvements, we recover the results of Rigault et al. We find that both populations have similar dispersions in distance than found by Rigault et al. and Kelly et al., with slightly smaller dispersions for locally passive (log() dex) SNe Ia fit with MLCS, the opposite of the effect seen by Rigault et al. and Kelly et al. We caution that measuring the local environments of SNe Ia in the future may require a higher resolution instrument than GALEX and that SN Ia sample selection has a significant effect on local star formation biases.

Duke Scholars

Author Daniel M. Scolnic Physics