We summarize various explosion models of Type Ia supernovae and their nucleosynthesis features for both Chandrasekhar and sub-Chandrasekhar mass white dwarf models. These models provide different predictions of the photometric and spectroscopic variations among Type Ia supernovae, which are compared with observations. Some attempts to model the peculiar SNe 1991T and 1991bg are shown.

Introduction

Type I supernovae (SNe I) are spectroscopically defined by the absence of hydrogen in their optical spectra and further subclassified into Ia, Ib, and Ic (see, e.g., Branch et al. 1991). The early–time optical spectra of SNe Ia are characterized by the presence of a deep absorption Si II line near 6150 Å, and their late–time spectra are dominated by strong blends of Fe emission lines (Harkness & Wheeler 1990). Relatively uniform light curves and spectral evolution of SNe Ia have led to the use of SNe Ia as a standard candle to determine cosmological parameters, H0 and q0 (Branch & Tammann 1992).

Recent attention has been paid to variations of light curves and spectra among SNe Ia. SNe 1991T and 1991bg have clearly revealed the presence of both spectroscopically and photometrically peculiar SNe Ia. Photometrically, maximum brightness and the decline rate of the light curve show some systematic variations, where SNe 1991T and 1991bg are situated at the two extreme ends of the brighter–slower tendency (Phillips 1993; Branch et al. 1993). Spectroscopically, the pre-maximum spectra of SNe Ia reveal a significant variation of the composition and expansion velocities of the outermost layers, whereas the post-maximum spectra are relatively uniform except for SN 1991bg.