| dc.description.abstract | Massive stars and supernovae are not only remarkable objects on their own but they are closely related to many other topics in Astrophysics, such as nucleosynthesis, star formation, and gravitational waves. However the properties of massive stars at late stages and their links to supernovae are not well understood.In our work we aim to improve the knowledge on these topics, by studying the spectra of evolved massive stars and the early spectra of supernovae that interact with the winds/atmospheres of massive stars, as they provide valuable information about supernova progenitors, such as mass-loss rates, wind velocities, and surface abundances.We use the radiative transfer code for expanding atmospheres in non-local thermodynamic equilibrium, CMFGEN. The code makes no assumptions for the source of radiative energy at the inner boundary and hence can be employed in modelling both stars with dense atmospheres/winds and eject a from explosive events interacting with the circumstellar material. Supernova progenitors are usually constrained from post-explosion data, but in exceptional cases they have been directly observed, mainly photometrically, and on even rarer occasions spectroscopically. Such is the case of SN 2015bh,a transient whose post-explosion fate is unknown, with a spectrum taken 1.5 yrs pre-explosion. In the first part of this thesis we aim to determine the progenitor properties of SN 2015bh using CMFGEN. Modelling the pre-explosion spectrum of SN 2015bh shows that the star had an effective temperature between8 700 and 10 000 K, luminosity of 1.8−4.74×106L, mass-loss rate of 0.6−1.5×10−3Myr−1, a wind terminal velocity of 1000 km s−1, and contained at least 25% H in mass at the surface, and half-solar Fe abundances. Therefore we conclude that the progenitor of SN 2015bh was a warm luminous blue variable star with an extended wind. Given the high wind velocity there is also the possibility the star was an inflated Wolf-Rayet star. If SN 2015bh was an impostor, we expect late-time spectroscopy to reveal either a similar luminous blue variable star or a Wolf-Rayet star, depending on how much H it retained in its envelope. If it was a genuine supernova, its minimum mass of 35M at the pre-supernova stage makes it a remarkable case of a successful explosion with black-hole formation.In the second part of this thesis, we investigate the post-explosion proper-ties of massive stars. We built an extensive library of spectra simulating the interaction of supernovae with their progenitor’s circumstellar medium at early times. We considered a range of progenitor mass-loss rates ( ̇M=5×10−4to10−2Myr−1), abundances (solar-like, CNO-processed, and He-rich), and explosion luminosities (L=1.9×108to 2.5×1010L). The diversity of massive star properties at the pre-supernova stage causes a variety of early-time interacting supernovae. We recognise three main classes of early-time spectra based on the ionisation stages of the species present, i.e. high (e.g. HeIIand OVI), medium(e.g. CIIIand NIII), and low-ionisation (e.g. HeIand FeII/III). Additionally, our modelled spectra respond strongly to changes in surface abundances, allowing well constrained measurements of H, He, and CNO. These can be used to obtain the progenitor type and mass, if it followed single star evolution. We also show that if ̇M.5×10−4(υ∞/150km s−1)Myr−1no interaction signatures will be observable in the spectra.Using our library of models described above, we then constrain the proper-ties of a sample of 17 observed early-time interacting supernovae. We show that these events cover a wide range of explosion and progenitor properties. They exhibit supernova luminosities from 108to 1012L, and temperatures from 10to 60 kK. The progenitor mass-loss rates are all higher than a few 10−4Myr−1,even up to 1Myr−1, with wind velocities spanning 100 to 800 km s−1. These values suggest that many progenitors of supernovae interacting with circumstellar material have significantly increased mass loss before explosion compared to what massive stars show during the rest of their lifetimes. While the surface abundances range from solar-like to H-depleted, we find that the majority of the events in our sample have CNO-processed surface abundances. In the single star scenario, this result points to a preference towards high-mass red supergiants as progenitors of interacting supernovae.Supernovae showing interaction with dense circumstellar material are extremely diverse and our grid of models can be extended to include all types of events, from the dimmest to the brightest supernovae observed to date. | en |
