Properties of Type Ia supernova spectra up to z ~0.3

• Context: Distance estimates using Type Ia supernovae lead to the discovery of the accelerating universe a decade ago. Further high precision studies of the dynamics of the universe using supernovae are now limited by systematic uncertainties. Optical spectra of supernovae from follow-up observations of the SDSS-II Supernova Survey provide a natural testbed for investigating key uncertainties like a potential brightness evolution of the standard candle and the nature of the observed non-standard colour-brightness relation. Spectroscopic data also provides information needed to improve our understanding of the physical processes taking place in Type Ia supernova explosions. Aims: If supernova properties change with redshift, this may be detectable through comparisons of spectral indicators. Studies of individual supernovae are needed in order to determine whether explosionproperties, and not only the population distribution, change with redshift. Similarly, a possible intrinsic colour-brightness relation can be explored through correlations with spectral properties.Methods: We describe spectral indicator studies of spectra observed at the NTT and NOT during 2006 and 2007. These benefit from excellent multi-band SDSS photometry while being at cosmological distances (z=0.05-0.3) where evolution, limiting Type Ia supernovae as standarizable candles, could be present. We perform measurements of pseudo equivalent widths and line velocities as a function of redshift and lightcurve parameters. We also present multiple Monte Carlo studies designed to determine possible systematic errors in our spectral analysis arising when comparing local high S/N supernova spectra with lower S/N host-contaminated distant spectra.Reults: We do not detect any significant signs of evolution in individual Type Ia supernova spectra up to z=0.3. Better control of selection effects (both of the local and the distant sample) is needed in orderto determine whether a population drift exists. We can confirm correlations between pseudo equivalent widths and lightcurve shape (stretch) and see indications of correlations with supernova colours. We conclude that it is possible to compare individual low S/N supernova spectra with local ones, but that care must be taken during both noise filtering and host-galaxy subtraction.  Host-galaxy contamination is the single largest source of systematic errors in our study, especially in the presence of differential slit loss.

Publication and Content Type

vet (subject category)

ovr (subject category)