An outburst from a massive star 40 days before a supernova explosion

Abstract

Various lines of evidence suggest that very massive stars experience extreme mass-loss episodes shortly before they explode as a supernova.[1−4] Interestingly, several models predict such pre-explosion outbursts. [5−7] Establishing a causal connection between these mass-loss episodes and the final supernova explosion will provide a novel way to study pre-supernova massive-star evolution. Here we report on observations of a remarkable mass-loss event detected 40 days prior to the explosion of the Type IIn supernova SN 2010mc (PTF 10tel). Our photometric and spectroscopic data suggest that this event is a result of an energetic outburst, radiating at least 6 × 10 47 erg of energy, and releasing about 2 Ofek et al. 10−2 M ⊙ at typical velocities of 2000 km s −1 . We show that the temporal proximity of the mass-loss outburst and the supernova explosion implies a causal connection between them. Moreover, we find that the outburst luminosity and velocity are consistent with the predictions of the wave-driven pulsation model [6] , and disfavour alternative suggestions [7] . An outburst from a SN progenitor 3 Spectra of the supernova, showing a blue continuum with Balmer emission lines, are presented in Figure 2. The continuum becomes redder with time, and its slope corresponds to an effective temperature of over 16,000 K at day five and drops to about 8,000 K at day 27. The Hα line has an initial width of ∼ 3×10 3 km s −1 at day 6, decreasing to ∼ 10 3 km s −1 at day 14. A broad (10 4 km s −1 ) P-Cygni profile emerges by day 27. The spectra also show He I lines with decreasing strength, presumably due to the drop in temperature.The nature of the precursor bump is very intriguing and can potentially tell us a great deal about the SN explosion and the progenitor. The only interpretation that is fully consistent with the photometric and spectroscopic evidence is that the first bump represents an outburst from the SN progenitor about one month prior to explosion, while the brighter bump is initiated by a full explosion of the star a few weeks later. Below we analyse this model in the context of the photometric and spectroscopic data. In SI §6 we discuss some alternative models and conclude that they are unlikely.The mass ejected by the precursor burst can be estimated in various independent ways.By requiring that the precursor integrated bolometric luminosity, E bol,prec , is lower than the kinetic energy of the precursor outburst (moving at velocity v prec ) which powers it, we can set a lower limit on the mass ejected in the precursor outburstThe outburst velocity is estimated from the line widths of 1000-3000 km s −1 , seen in the early-time spectra of the SN. As this mass was presumably ejected over a period of about one month (i.e., the outburst duration), the annual mass-loss rate is about 10 times higher. A similar order of magnitude argument can be used to put an upper limit on the mass in the precursor outburst. If some of the SN bolometric energy, E bol,SN , is due to interaction between the SN ejecta, moving a...

Choked jets and low-luminosity gamma-ray bursts as hidden neutrino sources

“…Motivated by the CJ-SB model for LL GRBs, we consider an extended material with mass M ext ∼ 10 −2 M and radius r ext ∼ 3 × 10 13 cm. WR stars have been observed with such unusually massive envelopes in the months leading up to their SN explosion [77][78][79]. Nakar [31] suggested similar envelope parameters for LL GRB 060218, but without strong constraints on the density profile.…”

Type Ia Supernovae Strongly Interacting with Their Circumstellar Medium

“…These rise times can also be used to estimate the mass-loss rate of the progenitor system that gave rise to the CSM. Using typical SN Ia-CSM ejecta and wind velocities of a few thousand and 100 km s −1 , respectively, and equations in Ofek et al (2013), we find that SNe Ia-CSM have mass-loss rates of a few times 10 −1 M yr −1 .…”

“…The upper limits on the X-ray luminosity implied by these null detections are ∼10 39 -10 43 erg s −1 ; observed X-ray fluxes of SNe IIn fall in this range (e.g., Pooley et al 2002;Zampieri et al 2005;Immler et al 2007). Using the upper limits determined from the X-ray nondetections of SNe Ia-CSM and equations in Ofek et al (2013), we calculate mass-loss rate upper limits of a few times 10 −1 M yr −1 (which matches the mass-loss rates calculated above using the rise times of SNe Ia-CSM). Prieto et al (2007) found that the host galaxies of the first four discovered SNe Ia-CSM (SNe 1997cy, 1999E, 2002ic, and 2005gj) are all late-type galaxies (dwarf irregulars and late-type spirals) with star formation likely occurring within the last few hundred Myr.…”

FLASH SPECTROSCOPY: EMISSION LINES FROM THE IONIZED CIRCUMSTELLAR MATERIAL AROUND <10-DAY-OLD TYPE II SUPERNOVAE

“…3. In the case of PTF10tel (SN 2010mc), which had a precursor (Ofek et al 2013), we adopt as the SN explosion the mean between the date of the last precursor measurement and the date of the first measurement of the SN (main peak).…”

“…Intense mass loss via eruptive events sometimes manifests itself as precursors prior to the SN explosion , as in SN 2006jc (Pastorello et al 2007), SN 2010mc (Ofek et al 2013), PTF11qcj (Corsi et al 2014), and possibly SN 2009ip (whether an SN was finally produced is controversial; e.g., Mauerhan et al 2013;Graham et al 2014;Margutti et al 2014;Ofek et al 2014;Martin et al 2015). Mass loss from…”