SUPERNOVAE, TYPE I, THEORY AND INTERPRETATION KEN'ICHI NOMOTO Supernovae are stellar explosions that release energies of ****** erg and shine as bright as a whole galaxy. Supernovae are classified into two major types: type I and type II, where type I supernovae are identified from the absence of hydrogen lines in the maximum-light spectra, in contrast to their presence in type II supernovae. Type I supernovae are further subclassifled from the helium features in the spectra, namely, type Ia (no helium), type Ib (helium- rich), and type Ic (helium-poor). The lack of hydrogen lines implies that the progenitor of a type I supernova has lost its hydrogen-rich envelope before the explosion. The candidates for the progenitors of type I supernovae are white dwarfs, Wolf-Rayet stars, and helium stars in close binary systems. The currently popular models are the carbon deflagration of accreting C+O white dwarfs for type Ia supernovae and the explosion of Wolf-Rayet stars and helium stars for type Ib and Ic supernovaeù TYPE Ia SUPERNOVAE For type Ia supernovae, accreting white dwarfs have been considered to be promising candidates for the progenitor stars. The explosion mechanism originally suggested by Fred Hoyle and William A. Fowler, that is, the thermonuclear explosion of electron-degenerate cores, basically has been confirmed by extensive numerical modeling and comparison with observations. WHITE DWARF PROGENITORS Isolated white dwarfs are simply cooling stars that eventually end up as invisible frigid stars. The white dwarf in a close binary system evolves differently, however, because the companion star expands and transfers matter to the white dwarf at a certain stage of its evolution. This mass accretion can rejuvenate the cold white dwarf. The mass accretion onto the white dwarf releases gravitational energy at the white dwarf surface. Most of the released energy is radiated away from the shocked region as ultraviolet light and does not contribute much to heating the white dwarf's interior. The continuing accretion compresses the previously accreted matter and releases gravitational energy in the interior. A part of this energy is transported to the surface and is radiated away from the surface (radiative cooling) but the rest goes into thermal energy of the interior matter (compressional heating). Thus the interior temperature of the white dwarf is determined by the competition between compressional heating and radiative cooling; that is, the white dwarf is hotter if the mass accretion rate M is larger, and vice versa. The scenario that possibly brings a close binary system to a type I supernova explosion is as follows (although the exact evolutionary origin is not yet understood): Initially, the close binary system consists of two intermediate-mass stars [********(solar masses)]. As a result of Roche lobe overflow, the primary star of this system becomes a white dwarf composed of carbon and oxygen (a C+O white dwarf). When the secondary star evolves, it begins to transfer hydrogen-rich matter to the white dwarf. When a certain amount of hydrogen is accumulated on the white dwarf surface, hydrogen shell burning is ignited. Its outcome depends on M: For slow accretion (***********M* yr*), hydrogen shell burning is unstable and tends to "flash," which leads to the ejection of most of the accreted matter from the white dwarf; the strongest flash grows into a nova explosion. For these cases, the white dwarf does not become a supernova because its mass cannot grow. In other words, novae are not the precursors of supernovae. For intermediate accretion rates (************************* yr**), on the other hand, the hydrogen flashes and the subsequent helium flashes are of moderate strength, thereby increasing the C+O white dwarf mass toward the Chandrasekhar mass. When the white dwarf mass becomes 1.4 M* and the central density reaches ***************; explosive carbon burning starts at the white dwarf's center. If the accretion rate is higher than ********** yr**, the accreted matter is too hot to be "swallowed" by the white dwarf. The matter forms a common envelope, which is eventually lost from the system. As a result of mass and angular-momentum losses from the system, some binaries form a pair of C+O white dwarfs. Further evolution of such a double white dwarf system is driven by gravitational-wave radiation and leads to a Roche lobe overflow of the smaller mass C+O white dwarf. The fate of these merging white dwarfs is not clear yet but would be either a type I supernova explosion or a collapse to form a single neutron star. CARBON DEFLAGRATION When carbon is ignited at the white dwarf's center, carbon burning is so explosive as to incinerate the material into iron-peak elements; the central temperature reaches ******K. The resulting shock wave is not strong enough to ignite carbon in the adjacent layer; in other words, a detonation wave that propagates at supersonic speed does not form. Instead, the interface between the burned and unburned layers becomes convectively unstable. As a result of mixing with the hot material, fresh carbon is ignited. In this way, a carbon-burning front propagates outward on the time scale for convective heat transport. This kind of explosive burning front that propagates at a subsonic speed is called a convective deflagration wave. In the standard model, the propagation speed of the convection deflagration wave is on the average about one-fifth of the sound speed. It takes about one second for the front to reach the surface region, which is significantly slower than the supersonic detonation wave. Hence, the white dwarf expands during the propagation of the deflagration wave. Behind the deflagration wave, the material undergoes explosive nuclear burning of silicon, oxygen, neon, and carbon, depending on the peak temperatures. In the inner layer, nuclear reactions are rapid enough to incinerate the material into iron-peak elements, mostly **Ni. When the deflagration wave arrives at the outer layers, the density it encounters has already decreased due to the expansion of the white dwarf. At such low densities, the peak temperature is too low to complete silicon burning and thus only Ca, Ar, S, and Si are produced from oxygen burning. In the intermediate layers, explosive burning of carbon and neon synthesizes S, Si, and Mg. In the outermost layers, the deflagration wave dies and C+O remain unburned. The composition structure after freeze-out is shown in Fig. 1. In the standard carbon deflagration model, the amount of **Ni produced is ********** M*, and the explosion energy is E=(nuclear energy release) -(binding energy of the white dwarf) = ********** erg. The nuclear energy release is large enough to disrupt the white dwarf completely and no compact star is left behind. The outcome of carbon deflagration depends on its propagation speed, which involves a parameter such as the mining length of convection. The preceding standard model has been chosen because It accounts well for the observed light curves and spectra at both early and late times of type Ia supernovae. LIGHT CURVE The explosion energy goes into the kinetic energy of expansion, and without a late-time energy source the exploding white dwarf would not be bright. However, during the expansion phase, **Ni decays into **Co with a half-life of 6.6 days and **Co decays into **Fe with a half-life of 77 days. These radioactive decays produce gamma rays and positrons, whose energies power the light curve as follows. Gamma rays originating from radioactive decays are degraded into x-rays by multiple Compton scatterings. The photoelectric absorption of x-rays and the collisional ionization due to energetic electrons eventually heat the expanding materials and produce the optical light. The light curve powered by the radioactive decays reaches its peak at about 15 days after the explosion and declines because of the increasing transparency of the ejecta to gamma rays and due to the decreasing number of radioactive elements. The calculated curve is in good agreement with the observed bolometric light curves of SN 1972E and SN 1981B. SPECTRA Because type Ia supernovae do not have a thick hydrogen-rich envelope, elements newly synthesized during the explosion can be observed in the spectra; this enables us to diagnose the internal hydrodynamics and nucleosynthesis in type Ia supernovae. Synthetic spectra are calculated based on the abundance distribution and expansion velocities of the standard model and are found to be in excellent agreement with the observed optical spectrum of SN 1981B as seen in Fig. 2. The material velocity at the photosphere near maximum light is -10,000 km/sec and the spectral features are identified as P-Cygni profiles of Fe, Ca, S, SI, Mg, and O. At late times, the spectra are dominated by the emission lines of Fe and Co. The outer layers are transparent and the inner Ni-Co-Fe core is exposed. Synthetic spectra of emission lines of [*****] and [Co *] agree quite well with the spectra observed at such phases. The agreement implies that both explosion energy and nucleosynthesis in the carbon deflagration model are consistent with the observations of type Ia supernovae. TYPES Ib AND Ic SUPERNOVAE The difference of the maximum-light spectra of types Ib and Ic supernovae from those of type Ia supernovae was first recognized in terms of the lack of a Si feature at 6100 *. A more fundamental difference is the presence of a He line feature around 5800 * in Spectra of types Ib and Ic, which type Ia spectra do not have. The He feature of type Ib is strong whereas that of type Ic is fairly weak. Another important difference is found in the late-time spectra; the broad emission lines of oxygen appear in types Ib and Ic, whereas iron features dominate the type Ia spectra. The exponential tails of the light curves imply that the decays of **Ni and **Co power the light curves of types Ib and Ic supernovae. The peak luminosities are lower than those of type Ia supernovae by a factor of roughly 4, which implies that the amount of **Ni produced is about 0.15 M* in types Ib and Ic. Most type Ib and Ic supernovae are associated with star-forming regions. This fact has led to the currently popular idea that the progenitors of type Ib and Ic supernovae are helium stars more massive than *3 M*. HELIUM STAR MODEL Helium stars considered here are formed from stars more massive than 12 M* that have lost their hydrogen-rich envelope by strong wind as in Wolf-Rayet stars or by Roche lobe overflow in close binary systems. Such helium stars evolve in the same manner as helium cores in massive stars, thereby initiating a supernova explosion by the iron core collapse as in type II supernovae. Theoretical light curves and spectra for the helium star models are basically consistent with observations of type Ib supernovae. The light curves of type Ic supernovae tend to decline faster than type Ib; this suggests that the progenitors of type Ic supernovae are somewhat less massive than those of type Ib. Additional Reading Branch, D.(1987). Supernovae. Encyclopedia of Physical Science and Technology 13 507. Nomoto, K.(1985). Explosive nucleosynthesis in carbon deflagration models for type I supernovae. In Nucleosynthesis: Challenges and New Developments, W.D. Arnett and J.W. Truran,eds. University of Chicago Press, Chicago, p. 202. Trimble, V.(1982). Supernovae. Part I: The events. Rev. Modern Physics 54 1183. Wheeler, J.C. and Harkness, R.P.(1990). Type I supernovae. Rep. Prog. Phys. 53 1467. Woosley, S.E. and Weaver, T.A.(1986). The physics of supernova explosions. Ann. Rev. Astron. Ap. 24 205.