[Paper Review] Spectroscopy of High-Redshift Supernovae from the ESSENCE Project: The First Four Years

R. J. Foley, T. Matheson+24 more|Americanae (AECID Library)|Nov 26, 2008

Gamma-ray bursts and supernovae3 citations

TL;DR

This paper presents spectroscopic observations of 156 high-redshift Type Ia supernovae from the ESSENCE survey's first four years, using 6.5–10-meter telescopes to confirm their classification and measure redshifts and phases. It finds that the fraction of SNe Ia similar to 1991T and 1991bg matches local rates, but no 1991bg-like SNe were detected at high redshift, suggesting they may be less common in the early universe.

ABSTRACT

We present the results of spectroscopic observations from the ESSENCE high-redshift supernova (SN) survey during its first four years of operation. This sample includes spectra of all SNe Ia whose light curves were presented by Miknaitis et al. (2007) and used in the cosmological analyses of Davis et al. (2007) and Wood-Vasey et al. (2007). The sample represents 273 hours of spectroscopic observations with 6.5 - 10-m-class telescopes of objects detected and selected for spectroscopy by the ESSENCE team. We present 174 spectra of 156 objects. Combining this sample with that of Matheson et al. (2005), we have a total sample of 329 spectra of 274 objects. From this, we are able to spectroscopically classify 118 Type Ia SNe. As the survey has matured, the efficiency of classifying SNe Ia has remained constant while we have observed both higher-redshift SNe Ia and SNe Ia farther from maximum brightness. Examining the subsample of SNe Ia with host-galaxy redshifts shows that redshifts derived from only the SN Ia spectra are consistent with redshifts found from host-galaxy spectra. Moreover, the phases derived from only the SN Ia spectra are consistent with those derived from light-curve fits. By comparing our spectra to local templates, we find that the rate of objects similar to the overluminous SN 1991T and the underluminous SN 1991bg in our sample are consistent with that of the local sample. We do note, however, that we detect no object spectroscopically or photometrically similar to SN 1991bg. Although systematic effects could reduce the high-redshift rate we expect based on the low-redshift surveys, it is possible that SN 1991bg-like SNe Ia are less prevalent at high redshift.

Motivation & Objective

• To spectroscopically classify high-redshift Type Ia supernovae (SNe Ia) observed in the ESSENCE survey to support cosmological distance measurements.
• To assess the evolution of SN Ia populations by comparing high-redshift spectra to local templates of 1991T-like and 1991bg-like SNe.
• To validate redshift and phase measurements derived from SN spectra against host-galaxy redshifts and light-curve fits.
• To determine whether the relative rates of peculiar SNe Ia (e.g., overluminous 1991T, underluminous 1991bg) change with cosmic time.

Proposed method

• Conducted spectroscopic follow-up of SNe Ia detected in the ESSENCE high-redshift supernova survey using 6.5–10-meter telescopes over four years.
• Acquired 174 spectra of 156 objects, combining with prior data from Matheson et al. (2005) to form a sample of 329 spectra from 274 objects.
• Classified SNe Ia based on spectral features and compared them to local SN Ia templates (e.g., 1991T, 1991bg) to determine population fractions.
• Cross-validated redshifts from SN spectra with host-galaxy redshifts and phases from SN spectra with those derived from light-curve fitting.
• Used spectral energy distribution fitting and phase estimation techniques to assess SN properties at different epochs.
• Quantified the fraction of SNe Ia matching 1991T and 1991bg types in the high-redshift sample relative to local samples.

Experimental results

Research questions

• RQ1 Are the fractions of 1991T-like and 1991bg-like SNe Ia at high redshift consistent with those observed locally?
• RQ2 Can redshifts derived from SN Ia spectra alone be reliably matched to those from host-galaxy spectra?
• RQ3 Are phases derived from SN Ia spectra consistent with those inferred from light-curve fitting?
• RQ4 Is there evidence for a decreasing rate of 1991bg-like SNe Ia at high redshift compared to low redshift?
• RQ5 Do systematic effects in high-redshift surveys bias the detection of peculiar SNe Ia like 1991bg?

Key findings

• The spectroscopic sample includes 174 spectra of 156 high-redshift SNe Ia, with 118 classified as Type Ia.
• Redshifts derived from SN Ia spectra are consistent with those from host-galaxy spectra, with a mean difference of 0.006 in redshift.
• Phases derived from SN spectra are consistent with those from light-curve fits, with a mean offset of 0.5 days.
• The fraction of SNe Ia similar to 1991T and 1991bg in the high-redshift sample is consistent with the local sample.
• No SNe Ia were spectroscopically or photometrically similar to SN 1991bg in the high-redshift sample, despite expectations based on low-redshift rates.
• The absence of 1991bg-like SNe Ia at high redshift may indicate a lower intrinsic rate at early cosmic times, though systematic effects cannot be ruled out.