Classical nova (CN) and dwarf nova (DN) systems have the same binary components (a low-mass main sequence star and a white dwarf) and the same orbital periods. An important question that therefore arises is: are these systems really different ? (and if so, what is the fundamental difference ?) or, are these the same systems, metamorphosing from one class to the other ?

The first thing to note in this respect is that the white dwarfs in DN systems are believed to accrete continuously (both at quiescence and during eruptions). At the same time, both analytic (e.g. Fujimoto 1982) and numerical calculations show, that when sufficient mass accumulates on the white dwarf, a thermonuclear runaway (TNR) is obtained and a nova outburst ensues (see e.g. reviews by Gallagher and Starrfield 1978, Truran 1982). It is thus only natural, to ask the question, is the fact that we have not seen a DN undergo a CN outburst (in about 50 years of almost complete coverage) consistent with observations of DN systems ? In an attempt to answer this question, we have calculated the probability for a nova outburst not to occur (in 50 years) in 86 DN systems (for which at least some of the orbital parameters are known).

This slow nova has been observed spectroscopically at Observatoire de Haute-Provence in two wavelengths regions: 330–460 nm (plate factors: 4 and 8 nm), and 750–1100 nm (plate factors: 23 and 26 nm/mm).

Time coverage includes the following stages: close to maximum of light, early nebular phase and advanced nebular stage.

Spectra are similar to those of classical novae: strong continuum, wide emission bands with double structure, flanked by a moderate strength absorption component at −1400 km/s.The following ions are identified: H I, O I, C I, N I, Mg I and II, Ca I and II, Fe II, Ti II, Si II and Sr II.

Recently a number of studies have been published on the nuclear abundance of nova ejecta, as summarized by Truran and Livio (1986). H is always underabundant (compared to solar) and He is overabundant except for the cases where the heavier elements are far overabundant. The abundances of C, N, and O range from nearly solar to highly overabundant. A few novae are very rich in Ne and Mg as well as O, which has led to the discovery that these novae occur on O/Ne/Mg white dwarfs (Williams, et al., 1985). We will assume that the abundances are an accurate and consistently determined set of data for our purposes. The nova ejecta is a combination of original white dwarf material, remnant material, remaining on the white dwarf from the previous outburst, and accreted material, all of which has undergone thermonuclear processing during the outburst. The question we address here is “Can we untangle the observational abundances to determine the contributions of each source?” A positive answer would allow us to tell whether the white dwarf’s mass is increasing or decreasing and thus have implications on the accreting white dwarf model for a SNI.

We describe a mechanism that promises to explain how classical nova outbursts take place on white dwarfs of 1 M⊙ or less and for accretion rates of 4 × 10−10 M⊙ yr−1 or greater.

SU UMa stars are one of subclasses of dwarf novae. Dwarf novae are semi-detached close binary systems in which a Roche-lobe filling red dwarf secondary loses matter and the white dwarf primary accretes it through the accretion disk. The main characteristics of SU UMa subclass is that they show two kinds of outbursts: normal outbursts and superoutbursts. In addition to the more frequent narrow outbursts of normal dwarf nova, SU UMa stars exhibit “superoutbursts”, in which stars reach about 1 magnitude brighter and stay longer than in normal outburst. Careful photometric studies during superoutburst have almost always revealed the “superhumps”: periodic humps in light curves with a period very close to the orbital period of the system. However, the most curious of all is that this superhump period is not exactly equal to the orbital period, but it is always longer by a few percent than the orbital period.

The FU Orionis objects show us the abrupt brightening by ~5 map. The two best studied examples, FU ori and V1057Cyg, brightened and have remained very luminous for years and began to fade gradually (cf. Herbig 1977). On the other hand, a large number of young stellar objects have been discovered with energetic molecular bipolar outflows (Lada 1985, for a review). On high-resolution radio observations, the disks have been detected around the central infrared sources (cf. Kaifu et al. 1984, Hasegawa et al. 1984). Most of the disks seem to be perpendicular to the bipolar outflows. These observations suggests that the disk is strongly related with the energetics phenomena like bipolar outflow in star forming regions. (cf.Okuda and Ikeuchi 1986, Hanami and Sakashita 1986, Pudritz 1985, Shibata and Uchida 1985)

We investigate the unsteady motion of mass reservoir formed by the accretion onto the magnetosphere around rotating neutron stars. The unsteady motion of the reservoir induces secondary accretion to neutron star by R-T instability. The nonperiodic or quasiperiodic phenomena of X-ray bursters seems to be related to this property of mass reservoir on the magnetosphere. We classify the typical dynamical state of the reservoir into three types with the parameters which are accretion rate acc and angular velocity of neutron star Ωs. They are nonsequential oscillation, sequential periodic (quasi-periodic) oscillation, and chaotic oscillation states.

We have simulated the helium shell flashes on an 1.3 M⊙ white dwarf and estimated the amount of mass loss. Our results may suggest a serious difficulty for the theories of the formation of Type I supernovae and of the accretion-induced formation of neutron stars because a significant amount of envelope mass is ejected during a helium shell flash.

It is now almost certain that the progenitor of SN 1987 is the B3 supergiant Sk −69°202 (e.g. Walborn et al. 1988). This provides us with valuable constraints on the SN precursor: According to West et al.(1987) the star has Mv=−6.8. Using the calibrations of Humphreys and McElroy (1984) we find that the precursor had Teff = 13000 − 16000K, −Mbol= 7.7 - 8.1 (log L/L⊙ = 5.04) and a radius of 48 - 60 R⊙.

The SN precursor was not a red supergiant, but a rather blue supergiant, which also explains the peculiar form of the light curve (Woosley, 1988; Schaeffer et al. 1987). However, standard evolutionary calculations would expect such stars to explode as red supergiants.

Mass loss from the B3 Ia progenitor star for SN 1987A is revealed by the recent emergence of narrow ultraviolet emission lines. The emitting gas is nitrogen-rich, has low velocity, and may be located a light-year from the supernova. This gives every sign of having been ejected from the SK −69 202 progenitor when it was a red supergiant, prior to its brief and ultimately violent life as a blue supergiant. Changes in the hydrogen line profiles during the early evolution provide a way to estimate the density distribution in the supernova atmosphere, and the mass of hydrogen it contains. A preliminary estimate is that the power-law index of density in the envelope goes as V−11 and the mass that lies above a velocity of 6, 000 km s−1 is between 1 and 6 solar masses.

The dedication of at least 24 local and visiting observers has enabled the SAAO to accumulate a large body of spectroscopic and photometric observations of SN1987a. Broad band photometry has been obtained on every possible photometric night while spectra at 7A(FWHM), including the first spectrum in the world, were obtained every night for the first 14 nights and thereafter on a weekly basis. These data, which have been more fully presented elsewhere (Menzies et al. 1987, Catchpole et al. 1987) are briefly discussed here for the time interval 24 February until 31 August 1987. Throughout this paper we adopt a distance modulus of 18.5 and a reddening of Av=0.6 for SN1987a.

The explosion of SN 1987A in the LMC during the early hours of February 23 1987, has presented southern observers opportunity, unique in our lifetime, to gain a new insight into the supernova phenomenon. Fortunately, SN1987A has risen to the challenge, and has shown us that many of our comfortable pre-conceptions about Type II events will have to be revised in fascinating and sometimes unexpected ways. This paper is an attempt to describe what has been learnt so far wth an emphasis on the research being carried out in Australia. All references are dated 1987 unless otherwise specified.

Existing calculations of s-process nucleosynthesis in massive stars such as Sk −69° 202 show enhancements in the He-burning shells which, when mixed to the surface, lead to Sc, Sr, and Ba enrichments of factors of 2 to 15. Abundances derived from the supernova absorption lines using a simple scattering model yield enhancements of this magnitude. Thus, the observed and predicted s-process abundances are roughly in accord, although the Ba/Sr ratio derived for the supernova may be higher than the s-process can account for.

The issue I want to discuss is this: supernovae observed during their photospheric phases display three distinct kinds of spectra - but there are four distinct chemical compositions that should be considered for the outer layers of exploding stars. What is the correspondence between the observed supernova type and the chemical composition?

The three kinds of observed spectra are Type II, which by definition shows optical hydrogen lines, Type Ia, which has neither hydrogen nor helium lines, and Type Ib, which has helium but not hydrogen.

The compositions to be considered are hydrogen-rich, or, speaking loosely, “solar”; helium-rich, or, loosely, “Wolf-Rayet”; deflagration - initially a mixture of heavy elements from carbon to radioactive nickel, which decays through cobalt to iron; and detonation - initially a mixture of just helium and radioactive nickel.

We present synthetic spectra for atmospheres in order to interpret the observed spectra of the supernova 1987A during the first few months after the initial event. Spherical symmetry and density profiles are assumed which are given by the homologous expansion of the stellar structure of a B3 supergiant. For hydrogen, up to eight levels and, for helium, 16 levels are allowed to deviate from LTE. The radiation transport is calculated consistently with the rate equations both for the continua and for the lines. Radiative equilibrium is assumed. The observed spectra and colours in the optical wavelength range can be well reproduced by pure hydrogen models during the first few weeks. The behaviour of the UV flux is due to changes in the effective temperature and the photospheric radius. For later stages, the influence of elements other than hydrogen and helium must be taken into account in order to compare the calculated and observed spectra in the optical wavelength range. Reasonable agreement between calculations and observations can be obtained with the assumption of half solar abundances for all elements but for the s-process elements Sc, Ti and Ba which are overabundant. To explain the small changes in the spectra after about 3 weeks up to 4 months, we need a total hydrogen mass of about 8 to 10 M⊙.

Among the many historic opportunities provided by the recent supernova in the LMC is that to improve our understanding of the physical conditions in the neighborhood of supernova photospheres, even though 1987A was initially characterized by radial and time scales smaller (by a factor 5–10) than “standard” more luminous SNII. Two consequences of this understanding, which we shall focus on in this contribution, are (a) an estimate of the (frequency-dependent) location and thickness of the photosphere and (b) the only direct determination of the distance of the supernova (via the generalized Baade method). We find that the photosphere is sharp enough to allow the use of plane-parallel geometry in the calculation of the emergent continuum spectral flux, if we confine our attention to those epochs (temperature T ~ 5000−6000 K) at which hydrogen is recombining at the photosphere. We also find that the distance to this supernova is 43 ± 4 kpc. The reliability of this determination should improve when accurate spectrophotometric data for dates other than March 1 become available to us.

SN 1987A gives a unique chance to study both the precursor star and the subsequent dynamical evolution of the explosion. Comparison of the light curves shows that either H0 ∼ 100 km/s/Mpc, or SN 1987A ejected significantly less 56Ni than ordinary Type II supernovae. Investigation of the stellar structure pertinent to SK -69 202 reveals multiple solutions. For given luminosity, effective temperature, core mass and core radius, there are two families of envelope mass, one with large envelope mass and one with small envelope mass. The small envelope mass solutions can be ruled out by considerations of kinematics and the light curve. Envelopes of moderate mass may avoid each of these problems, but must be helium rich to be structurally self-consistent.

The spectrum in both the optical and the ultraviolet at about two days is fairly well represented by a hydrogen envelope with a power law density profile (ρ∝r-11 ) of one-quarter solar metallicity in LTE. Theoretical spectra at this early epoch tend to favor luminosities on the high side of observational estimates in order to ionize Ca II and prevent excessively strong lines at H and K and the infrared triplet, with some ramifications for distance estimates.

The spectra of SN1987A present an interesting contrast to other SN II events. A McDonald Observatory spectrum of SN 1985H in NGC 3359 of uncertain epoch shows a very close resemblance to that of SN 1987A at about two month’s age, including the strong line at 607 nm attributed by Williams to Barium. SN 1985H may have been of the same class of event as SN 1987A.

Presupernova evolution of the progenitor of SN 1987A, hydrodynamics of explosion (shock propagation, explosive nucleosynthesis), optical light curve due to shock heating and 56Co decay, and X-ray and γ-ray light curves are calculated and compared with the observations of SN 1987A. Constraints on the mass of the hydrogen-rich envelope Menv (i.e., mass loss history) and the helium abundance in the envelope are obtained from the progenitor’s blue-red-blue evolution as well as from the light curve. The explosion energy E and the mass and distribution of 56Ni are inferred from the light curves. Models and observations are in reasonable agreement for E/Menv = 1.5 ± 0.5 × 1050 erg/M⊙, Menv = 5 - 10 M⊙, and MNi ∼ 0.07 M⊙. Mixing of 56Ni into the envelope is indicated.

Light curves of exploding bare helium stars are also calculated to see whether the observed Type Ib supernova light curves can be accounted for.

A neutrino burst was observed in the KAMIOKANDE-II detector on 23 February, 7:35:35 UT ( ±1 minute) during a time interval of 13 seconds. The signal consisted of 11 electron events of energy 7.5 to 36 MeV, of which the first 2 point back to the Large Magellanic Cloud (LMC) with angles 18° ± 18 ° and 15° ± 27°.

A new analysis of the special low threshold trigger counts strongly indicates that KAMIOKANDE did not observe any signals at 2:52:36 which is the time claimed by the Mont Blanc group.

In this report the result of old calculations (Mayle 1985; Woosley, Wilson, Mayle 1986; Mayle, Wilson, Schramm 1987) of collapse driven explosions and new calculations of the kelvin-Helmholtz proto-neutron star cooling will be compared with the neutrino observations of supernova 1987a. The calculations are performed by a modern version of the computer model of Bowers and Wilson 1982. (See Mayle 1985 for more recent improvements).

First we give the results of the old calculations. In the collapse, bounce and cooling of the central iron core of a massive star, about 0.1% of the binding energy of the eventual neutron star is emitted in a short deleptonization burst as the bounce shock passes through the photosphere; 5% is emitted in the total deleptonization process; and 95% is released as thermal emission in all neutrino species. In a survey of a wide range of stellar masses, stars in the range 20 to 30 Mθ are found to have the most energetic antineutrino spectra . In calculations where black holes were formed (Woosley, Wilson, Mayle 1986 and Wilson 1971) very little neutrino emission was found associated with black hole formation. The neutrinos associated with BH formation also have low energies. The time history of the neutrino pulse is sensitive to the explosion mechanism. If the mechanism is a prompt exiting through the star of the bounce shock wave, the pulse has a high peak as the shock wave passes near the photosphere. It falls rapidly for the next first few tenths of a second and then declines slowly over several seconds to effectively zero. If no prompt explosion occurs then the shock becomes an accretion shock and matter continues to fall onto the proto-neutron star keeping up the luminosity. After the late time mechanism ejects the envelope the luminosity drops several fold to the luminosity associated with the bare proto-neutron star. During the accretion phase the energy rises to about 15 mev.