Superluminous supernovae from the Dark Energy Survey

Abstract

We present a sample of 21 hydrogen-free superluminous supernovae (SLSNe-I), and one hydrogen-rich SLSN (SLSN-II) detected during the five-year Dark Energy Survey (DES). These SNe, located in the redshift range 0.220 < z < 1.998, represent the largest homogeneously-selected sample of SLSN events at high redshift. We present the observed g, r, i, z light curves for these SNe, which we interpolate using Gaussian Processes. The resulting light curves are analysed to determine the luminosity function of SLSN-I, and their evolutionary timescales. The DES SLSN-I sample significantly broadens the distribution of SLSN-I light curve properties when combined with existing samples from the literature. We fit a magnetar model to our SLSNe, and find that this model alone is unable to replicate the behaviour of many of the bolometric light curves. We search the DES SLSN-I light curves for the presence of initial peaks prior to the main light-curve peak. Using a shock breakout model, our Monte Carlo search finds that 3 of our 14 events with pre-max data display such initial peaks. However, 10 events show no evidence for such peaks, in some cases down to an absolute magnitude of < −16, suggesting that such features are not ubiquitous to all SLSN-I events. We also identify a red pre-peak feature within the light curve of one SLSN, which is comparable to that observed within SN2018bsz.

Observational properties of extreme supernovae

“…Due to the incredible discovery pace of SLSNe, especially the hydrogen-free type which have been observed ten times more than the hydrogen-rich, we have roughly 110 candidates at the time of writing, spanning from those in neighbouring galaxies (SN2018bsz at 111 Mpc) [40] up to the distant Universe (DES16C2nm at z~2.0) [41]. Thanks to such a wealth of discoveries, augmented by the advent of all-sky surveys [42,43,44], the original concept of using an arbitrary fiducial luminosity threshold to define such a population 17 has now been reconsidered. From large sample studies it seems that the SLSN I luminosity function extends down to the luminosities of broad-line type Ic SNe (MAB ~ −20 mag) [42] or even fainter at the level of a more prosaic type Ic SN (MAB ~ −19 mag) [44].…”

“…Thanks to such a wealth of discoveries, augmented by the advent of all-sky surveys [42,43,44], the original concept of using an arbitrary fiducial luminosity threshold to define such a population 17 has now been reconsidered. From large sample studies it seems that the SLSN I luminosity function extends down to the luminosities of broad-line type Ic SNe (MAB ~ −20 mag) [42] or even fainter at the level of a more prosaic type Ic SN (MAB ~ −19 mag) [44].…”

“…2), referred to as 'bump' [59]. It has been speculated that such a bump could be ubiquitous for all types of SLSN I [60], but this has been recently disproved [44]. The overall evolution of the prototypical SLSN I Slow SN2015bn revealed the presence of undulations in the light curve, possibly due to the interaction of the expelled material with small clumps of CSM [53].…”

SN 2010kd: Photometric and Spectroscopic Analysis of a Slow-decaying Superluminous Supernova

“…Superluminous supernovae (SLSNe) are the most luminous SNe, having a mean absolute magnitude of ∼ −21.7 mag (Gal-Yam 2012; Quimby et al 2013;Nicholl et al 2016a;De Cia et al 2018;Angus et al 2019;Inserra 2019).…”

Models and Simulations for the Photometric LSST Astronomical Time Series Classification Challenge (PLAsTiCC)

“…During the Kaggle competition, a recent analysis of 21 SLSN-I light curves from DES (Angus et al 2019) suggests that the magnetar model is not sufficient to describe all of these events. To describe the full SLSN-I population, other models may be needed such as interactions with circumstellar material (e.g., Chevalier & Irwin 2011;Chatzopoulos et al 2013Chatzopoulos et al , 2016.…”