The existence of multiple subclasses of type Ia supernovae (SNeIa) has been the subject of great debate in the last decade. In this work, we show how machine learning tools facilitate identification of subtypes of SNe Ia. Using Deep Learning for dimensionality reduction, we were capable of performing such identification in a parameter space of significantly lower dimension than its principal component analysis counterpart. This is evidence that the progenitor system and the explosion mechanism can be described with a small number of initial physical parameters. All tools used here are publicly available in the Python package DRACULA (Dimensionality Reduction And Clustering for Unsupervised Learning in Astronomy) and can be found within COINtoolbox (https://github.com/COINtoolbox/DRACULA).

The properties of the Supernovae discovered in coincidence with long-duration Gamma-ray Bursts and X-Ray Flashes are reviewed, and compared to those of SNe for which GRBs are not observed. The SNe associated with GRBs are of Type Ic, they are brighter than the norm, and show very broad absorption lines in their spectra, indicative of high expansion velocities and hence of large explosion kinetic energies. This points to a massive star origin, and to the birth of a black hole at the time of core collapse. There is strong evidence for gross asymmetries in the SN ejecta. The observational evidence seems to suggest that GRB/SNe are more massive and energetic than XRF/SNe, and come from more massive stars. While for GRB/SNe the collapsar model is favoured, XRF/SNe may host magnetars.

Studying a multi-dimensional structure of supernovae (SNe) gives important constraints on the mechanism of the SN explosion. Polarization measurement is one of the most powerful methods to study the explosion geometry of extragalactic SNe. Especially, Type Ib/c SNe are the ideal targets because the core of the explosion is bare. We have performed spectropolarimetric observations of Type Ib/c SNe with the Subaru telescope. We detect a rotation of the polarization angle across the line, which is seen as a loop in the Q - U plane. This indicates that axisymmetry is broken in the SN ejecta. Adding our new data to the sample of stripped-envelope SNe with high-quality spectropolarimetric data, five SNe out of six show a loop in the Q - U plane. This implies that the SN explosion commonly has a non-axisymmetric, three-dimensional geometry.

The Supernova Working Group was re-established at the IAU XXV General Assembly in Sydney, 21 July 2003, sponsored by Commissions 28 (Galaxies) and 47 (Cosmology). Here we report on some of its activities since 2005.

While Type Ia supernovae are widely used as distance indicators, the reasons for the correlation between luminosity and light curve width that allows SNe Ia to be calibratable standard candles is not yet fully understood, and in particular the details of the explosion mechanism are still the subject of heated debate. We present the results of a systematic approach that uses the high-quality data collected by the European Research Training Network “The Physics of Type Ia Supernova Explosions” to map the supernova ejecta and to infer the properties of the explosion.

The earliest spectrum available of a well-observed, local type Ia SN, the UV-optical spectrum of SN 1990N obtained 14 days before B maximum, was modelled with a Monte Carlo code in order to determine its epoch and consequently the rise time from explosion to B max, which is important for fitting template light curves. If a standard density distribution and composition for the SN ejecta are used, and the Cepheid distance modulus to the host galaxy, NGC4639 in the Virgo cluster, μ = 32.0, is adopted, the spectrum is best reproduced for an epoch t = 5.5 ± 1 days, implying for the SN a rise time of 19.5 ± 1 days. For a shorter μ, t is smaller, but for a T-F distance μ = 31.4 the solution is not consistent.

The properties of the best-observed peculiar, SN 1998bw-like Type Ic supernovae (sometimes called “hypernovae” ) are reviewed, starting from SN 1998bw itself and including SNe 1997ef and 2002ap. Analysis of the light curves and the spectra shows that, while these SNe display a range of properties (kinetic energy, mass of the ejecta, mass of 56Ni synthesized in the explosion), they have in common a larger-than-normal explosion kinetic energy, giving rise to the characteristic broad-lined spectra. Also, they all come from the collapse of bare CO cores of massive ( ≳ 20M⊙) progenitor stars. Some of the properties of these SNe, such as kinetic energy and mass of 56Ni, are probably correlated with the mass of the progenitor. Evidence that these powerful events are intrinsically asymmetric, suggesting that a correlation with at least some gamma-ray bursts can be expected, is also discussed.

We review the characteristics of nucleosynthesis in ‘hypernovae’, i.e., core-collapse supernovae with very large explosion energies (≳ 1052 ergs). The hypernova yields show the following characteristics: (i) the mass ratio between the complete and incomplete Si burning regions is larger in hypernovae than normal supernovae. As a result, higher energy explosions tend to produce larger [(Zn, Co, V)/Fe] and smaller [(Mn, Cr)/Fe], which could explain the trend observed in very metal-poor stars; (ii) because of enhanced α-rich freeze-out, 44Ca, 48Ti, and 64Zn are produced more abundantly than in normal supernovae. The large [(Ti, Zn)/Fe] ratios observed in very metal poor stars strongly suggest a significant contribution of hypernovae; and (iii) oxygen burning takes place in more extended regions in hypernovae to synthesize a larger amount of Si, S, Ar, and Ca (‘Si’), which makes the ‘Si’/O ratio larger. The abundance pattern of the starburst galaxy M 82 may be attributed to hypernova explosions. We thus suggest that hypernovae make important contribution to the early Galactic (and cosmic) chemical evolution.

A series of early-time optical spectra of the peculiar SNIa 1991T, obtained from 2 weeks before to 4 weeks after maximum, have been computed with our Monte Carlo code.

The earlier spectra can be successfully modelled if 56Ni and its decay products, 56Co and 56Fe, dominate the composition of the outer part of the ejecta. This atypical distribution confirms that the explosion mechanism in SN 1991T was different from a simple deflagration wave, the model usually adopted for SNe Ia.

As the photosphere moves further into the ejecta the Ni Co Fe fraction drops, while intermediate mass elements become more abundant. The spectra obtained 3–4 weeks after maximum look very much like those of the standard SN Ia 1990N. A mixed W7 composition produces good fits to these spectra, although Ca and Si are underabundant. Thus, in the inner parts of the progenitor white dwarf the explosion mechanism must have been similar to the standard deflagration model.

The fits were obtained adopting a reddening E(B – V) = 0.13. A Tully-Fisher distance modulus μ = 30.65 to NGC 4527 implies that SN 1991T was about 0.5 mag brighter than SN 1990N. At comparable epochs, the photosphere of SN 1991T was thus hotter than that of SN 1990N. The high temperature, together with the anomalous composition stratification, explains the unusual aspect of the earliest spectra of SN 1991T.

The model results allow us to follow the abundances as a function of mass. In particular, spectroscopic evidence is found that about 0.6M⊙ of 56Ni must have been synthesized in the outermost 1M⊙ of the exploding white dwarf. This implies that almost twice as much 56Ni was produced in SN 1991T than in normal SNe Ia, and explains the unusual brightness of this SN.

Medium resolution (2Å/px) but high s/n spectra of approximately twenty of the brightest blue stars in the young open cluster NGC 330 in the SMC have been obtained with EFOSC1 on the ESO 3.6m telescope, and analyzed in order to determine the atmospheric parameters and the evolutionary status of the stars. LTE and NLTE model atmosphere calculations were used to determine the stellar parameters. The Teff values were derived from fits of the UV continua for all stars where these were available, using Robertson's (1974) B and V photometry to scale the Kurucz model fluxes for metallicity Z = 0.1Z⊙. Luminosities of the sample stars lie in the range 4.0 < log(L*/L⊙) < 5.0 and spectral types between B0 and late-B.