Abundance Tomography of Type Ia Supernovae

Many uncertainties about the physics of Type Ia Supernovae have
been revealed in the recent past, and numerous pieces are puzzled
together to achieve a complete description of the phenomenon of
thermonuclear explosions in the sky. However, very important parts
are still missing. In particular, the concept lacks a proper
connection between the various evolutionary steps, namely the
progenitor scenario, explosion theory, nucleosynthesis from the
burning, and the observations. Early time spectra of Type Ia
Supernovae naturally contain information about all of these
processes and are at the centre of the entire scenario. Appropriate
models of that phase can provide the missing link and improve our
understanding of this field enormously. The goal of this thesis is
to advance new methods to calculate synthetic spectra in order to
extract the information contained in the observations more
efficiently. Based on a well established radiation transfer code, a
new technique called "Abundance Tomography" is developed to derive
the abundance distribution of Type Ia Supernovae ejecta. While
previous approaches were limited to the determination of the
abundances of specific species in restricted regions of the
supernova envelope, here a complete stratified distribution of all
major elements is obtained. This method is applied to the very well
observed normal SN 2002bo. Combining the early spectra with those
of the nebular phase leads to a coverage of the entire ejecta from
the centre out to the highest velocities. The abundances derived
are used to compute a synthetic bolometric light curve to test the
radial distribution of Fe group and intermediate-mass elements. The
sampling procedure of the incident radiation field at the lower
boundary is modified to obtain a better description of the real
situation in Type Ia Supernovae. This improves the overall flux
distribution significantly, especially in the red part of the
spectrum, where almost no real line opacity is found. Synthetic
spectra with this new procedure reproduce the observations much
more accurately, as is shown by models of SN 2002er. Hydrogen lines
have never been detected convincingly in Type Ia Supernovae
spectra. However, using spectra that were observed more than 10
days before maximum light, it is shown that small amounts of
hydrogen in the outer parts of the ejecta can explain high velocity
line absorptions, seen rather frequently in various objects, e.g.
SN 2002dj, SN 2003du, and SN 1999ee. The hydrogen is not claimed to
be primordial to the white dwarf but it is rather the effect of the
supernova ejecta interacting with circumstellar material, namely
the white dwarf's accretion disk build up prior to the explosion.
Finally, UV spectra of Type Ia Supernovae are discussed. The
ability of the Monte Carlo technique to deal naturally with this
wavelength region is proven. Applications are presented by
modelling spectra of SN 2001ep and SN 2001eh obtained with the
Hubble Space Telescope. The results are discussed in the broader
context of Type Ia Supernovae physics: What causes the diversity in
the nearby sample? What are the progenitors and how does the
explosion work? What is the influence on cosmological models? A
detailed knowledge of the abundances, their distribution in the
Supernova ejecta, and their ultimate causes delivers the key to
these fundamental issues.