A series of optical and one near-infrared nebular spectra covering the first year of the Type Ia supernova SN 2011fe are presented and modelled. The density profile that proved best for the early optical/ultraviolet spectra, rho-11fe, was extended to
lower velocities to include the regions that emit at nebular epochs. Model rho-11fe is intermediate between the fast deflagration model W7 and a low-energy delayed-detonation. Good fits to the nebular spectra are obtained if the innermost ejecta are dominated by neutron-rich, stable Fe-group species, which contribute to cooling but not to heating. The correct thermal balance can thus be reached for the strongest [FeII] and [FeIII] lines to be reproduced with the observed ratio. The 56Ni mass thus obtained is 0.47 +/- 0.05 Mo. The bulk of 56Ni has an outermost velocity of ~8500 km/s. The mass of stable iron is 0.23 +/- 0.03 Mo. Stable Ni has low abundance, ~10^{-2} Mo. This is sufficient to reproduce an observed emission line near 7400 A. A sub-Chandrasekhar explosion model with mass 1.02 Mo and no central stable Fe does not reproduce the observed line ratios. A mock model where neutron-rich Fe-group species are located above 56Ni following recent suggestions is also shown to yield spectra that are less compatible with the observations. The densities and abundances in the inner layers obtained from the nebular analysis, combined with those of the outer layers previously obtained, are used to compute a synthetic bolometric light curve, which compares favourably with the light curve of SN 2011fe.
The spectral properties of type Ia supernovae in the ultraviolet (UV) are investigated using the early-time spectra of SN 2001ep and SN 2001eh obtained using the Hubble Space Telescope (HST). A series of spectral models is computed with a Monte Carlo
spectral synthesis code, and the dependence of the UV flux on the elemental abundances and the density gradient in the outer layers of the ejecta is tested. A large fraction of the UV flux is formed by reverse fluorescence scattering of photons from red to blue wavelengths. This process, combined with ionization shifts due to enhanced line blocking, can lead to a stronger UV flux as the iron-group abundance in the outer layers is increased, contrary to previous claims.
The nebular spectra of the broad-lined, SN 1998bw-like Type Ic SN 2002ap are studied by means of synthetic spectra. Two different modelling techniques are employed. In one technique, the SN ejecta are treated as a single zone, while in the other a de
nsity and abundance distribution in velocity is used from an explosion model. In both cases, heating caused by gamma-ray and positron deposition is computed (in the latter case using a Monte Carlo technique to describe the propagation of gamma-rays and positrons), as is cooling via forbidden-line emission. The results are compared, and although general agreement is found, the stratified models are shown to reproduce the observed line profiles much more accurately than the single-zone model. The explosion produced ~ 0.1 Msun of 56Ni. The distribution in velocity of the various elements is in agreement with that obtained from the early-time models, which indicated an ejected mass of ~ 2.5 Msun with a kinetic energy of 4 x 10^{51} erg. Nebular spectroscopy confirms that most of the ejected mass (~ 1.2 Msun) was oxygen. The presence of an oxygen-rich inner core, combined with that of 56Ni at high velocities as deduced from early-time models, suggests that the explosion was asymmetric, especially in the inner part.
