| The evolution of massive stars, as well as their endpoints as supernovae (SNe), is important both in astrophysics and cosmology. While tremendous progress towards an understanding of SNe has been made, there are still many unanswered questions. The goal of this thesis is to study the evolution of massive stars, both before and after explosion. In the case of SNe, we synthesize supernova light curves and spectra by relaxing two assumptions made in previous investigations with the the radiative transfer code cmfgen, and explore the effects of these two assumptions. Previous studies with cmfgen assumed γ-rays from radioactive decay deposit all energy into heating. However, some of the energy excites and ionizes the medium. A new solver is developed to include these non-thermal excitation and ionization processes. Non-thermal excitation and ionization are crucial for forming some lines, especially Há in the nebular phase. To investigate non-thermal effects, a comparison is made between models with, and without, the non-thermal solver. Benchmarking the solver is done by comparing the non-thermal models with observations of SN 1987A. Satisfactory agreement is achieved and possible problems are discussed. With the new solver, future studies will shed light on the mixing of material between layers of different composition in supernova explosions and put further constraints on supernova explosion models.;Hubble expansion is a good approximation for most types of SNe, except Type II-P. Red supergiants are widely accepted to be the progenitors of Type II-P SNe and they have radii of hundreds to thousands of times larger than that of the Sun. Type II-P SNe “memorize” their large radii at the time of explosion for several weeks and material is still being accelerated. A time-dependent fully relativistic solver is developed to handle such cases. |
