Host-selected supernovae found not to bias cosmological analysis
Type Ia Supernovae (SNe Ia) are “standard candles” that allow astronomers to measure the expansion of the Universe. By measuring distances at various cosmic times using SNe Ia, research in the late 1990s discovered that the Universe is expanding at an accelerating rate. The unknown force responsible for the observed acceleration is popularly known as dark energy. Other experiments have confirmed this remarkable finding, yet the nature of dark energy is elusive.
So far, cosmologists have been using a “traditional sample” of SNe Ia for distance measurements: SNe Ia that are classified from their spectral features with redshifts also determined from their spectra. In particular at higher redshifts, however, there is a problem obtaining SNe Ia spectra because the process is very inefficient in terms of telescope time, costs, and elected yield. Moreover, the spectra need to be taken when SNe Ia are still bright.
To overcome these issues, the next-generation SN surveys (such as the “Dark Energy Survey”, DES, and the Large Synoptic Survey Telescope, LSST) are taking a different approach. These surveys will classify SNe Ia from photometry, and the redshifts will be obtained from their host galaxy spectra – which can be done at any time. The downside of creating a SN Ia sample using this approach, called “host-selected”, is that the sample will be biased towards bright, hence massive, hosts because it is relatively easy to obtain redshifts from galaxies that are bright and massive. Former CAASTRO-Swinburne PhD student Dr Syed Uddin (now President’s International Fellow at Purple Mountain Observatory, Chinese Academy of Sciences) and colleagues set out to test whether these host-selected samples produce biased cosmological results.
In their recent paper, they created random traditional and host-selected samples and found that the light curve properties of these two samples differed. Host-selected samples appeared to be redder and have narrower and faster declining light curves. Moreover, these samples had a tendency to live in more massive hosts and to explode at relatively lower redshifts compared to traditional samples. However, when the researchers compared cosmological parameters derived from these two samples, they did not find any significant differences. The matter density of the Universe and the equation of state of dark energy remained consistent across traditional and host-selected samples. This is an important indication for future SN surveys that will have limited real-time follow-up for spectroscopical classification.
Syed Uddin, Jeremy Mould, Chris Lidman et al. in the Publications of the Astronomical Society of Australia (2017) “Cosmological Inference from Host-Selected Type Ia Supernova Samples”