Observing supernova remnants provides important clues to the nature of supernova explosions. Conversely, the late stages of stellar evolution and the mechanism of supernova explosions affect supernova remnants through circumstellar matter, stellar remnants, and nucleosynthesis. The elements of supernova classification and the connection between supernova type and remnant properties are explored. A special emphasis is placed on SN 1987a which provides a unique opportunity to learn the connection between the star that exploded (whose name we know) and the remnant that will develop in our lifetimes.

The time dependent ionization and temperature structure of the circumstellar medium around supernovae has been calculated, in order to interpret recent supernova radio observations. For a stellar wind origin of the circumstellar medium, we relate the time of radio turn-on to the progenitor mass loss rate. We also show that large column densities for the UV resonance lines are expected. The results are applied to SN 1979c, SN 1980k and SN 1987A.

The supernova SN1986j resembles the prototypical Type V supernova SN1961v in the relatively slow ∼1000km/s expansion velocity, the slow light curve, and also in the Hα dominated spectrum. The optical spectrum is similar to the spectra of some novae, and some OB stars with massive winds, being characteristic of a nebular plasma at about 1010cm−3 and 104K. What makes SN1986j exceptional is its tremendous radio luminosity, the brightest radio supernova observed to date. The radio emission indicates the presence of a massive circumstellar wind, with which the SN ejecta are now colliding. Since the cooling time of the optically emitting gas is about an hour, a heat source is required to power the light curve. Shocks moving back into the ejecta offer a natural heat source, and account quantitatively for the observed luminosity and spectral character of SN1986j. The large Hα/Hβ ratio is attributed to trapping of Ly α, which pumps the n = 2 level of hydrogen, causing a finite optical depth in Balmer lines, and converting Hβ to Pα and Hα. The ratio of the derived H(n = 2) density and column density yields a size for the Hα emitting region consistent with the thickness of a cooling shock, but less than 10−7 of the 1017cm VLBI size. An important discriminant between shock models and photoionization models of the spectrum is that shocks predict Lyman 2-photon emission. The mass of the optically emitting material in SN1986j is about 1M⊙, substantially less than the 2000 M⊙ argued in the case of SN1961v by Utrobin. However, there may be, and probably is, considerably more unobserved ejecta. This material should reveal itself as the remnant of SN1986j continues to evolve.

We report the radio detection of supernova 1961v in NGC 1058. SN 1961v has a spectral index of -0.4 ± 0.1. At the distance of NGC 1058, the absolute monochromatic luminosity of this source is comparable to that of Cas A. A second nonthermal source with a spectral index of -0.3 ± 0.1 was also detected in NGC 1058 and is likely to be a remnant of a supernova that was not optically detected. The two radio sources, and two optically faint H II regions that coincide with the radio sources, are separated by only 3.5".

With the recent Supernova 1987a in the LMC, new and interesting possibilities have arisen for the solution of a problem relating to the explosion mechanism of supernovae. The presupernova was probably a B3Ia, the blue supergiant, and not a red giant as it was earlier thought likely. Calculations of the evolution, which have been made recently, show that the loss of hydrodynamical stability may be connected with carbon burning in the stellar core during the blue giant stage. This loss of stability of the CO core is the main factor in our explanation of the recent event of SN 1987a.

Before we consider the thermonuclear model of a supernova based on a thermal flash in the degenerate CO core, let us dwell on the present situation in the theory of supernovae. The main problem in supernova theory is the simultaneity of births of a compact remnant and an expelled envelope. The compact remnant later becomes a neutron star. The expelled envelope determines the curve of brightness. Now, it is clear that a supernova explosion may result in either the total disruption of a star or in the simultaneous production of a compact remnant and an expelled envelope.

The density distribution of the supernova ejecta and that of the surrounding medium are the most important parameters for the early evolution of supernova remnants. The distribution of the ejecta depends on the detailed hydrodynamics of the explosion, but the outer parts of a supernova can probably be represented by a steep power law density distribution with radius. Self-similar solutions are especially useful for modeling the interaction of a supernova with its surrounding. The supernova first interacts with mass loss from the progenitor star. Evidence for circumstellar interaction is present in a number of extragalactic supernovae, including SN1987a. The explosions of massive stars probably interact with circumstellar gas for a considerable time while Type Ia supernovae interact more directly with the interstellar medium. X-ray spectroscopy is a good diagnostic for the physical conditions in young supernova remnants and for the composition of the supernova gas.

Adiabatic models of supernova remnants (SNRs) show very large temperature gradients. The effect of thermal conduction on the Sedov solution was studied by a number of authors (Solinger et al., 1979; Cox and Edgar, 1983; Cowie, 1977).

The observations show, however, that young SNRs such as SN 1006 (Hesser and van den Bergh, 1981) and SN 1572 (Strom, Goss and Shaver, 1982) are in an intermediate state between free expansion and the Sedov phase. In these cases stellar matter cannot be neglected. Following Chevalier (1982), the freely expanding ejecta of a Type I SNR can be modeled in such a way that the inner 4/7 of the mass have constant density and the outer 3/7 have a ρ∝r−7 profile (the "ramp"). The interaction of the ejecta with the uniform circumstellar medium (CSM) gives rise to a pair of shocks. As long as the reverse shock is within the r−7 part of the density profile, the interaction region is described by a self-similar solution (Chevalier, 1982). Such a solution holds for an adiabatic single fluid; the temperature gradient, however, is so large that it may give rise to a quite high heat flux.

We present radio and optical images of the shell-like remnant of the 1901 outburst of Nova GK Persei. The behaviour of this object is remarkably similar to supernova remnants. The synchrotron radiation-emitting shell is polarized with the magnetic field oriented radially, as in young SNR’s. This similarity plus extensive data we have acquired on the expansion and the interstellar environment of GK Per indicate that the nova shell is colliding with ambient gas whose density is substantially higher than the ISM.

Furthermore, there is strong evidence that the ambient gas is circumstellar rather than interstellar, and that this material is the shell of an ancient planetary nebula associated with the white dwarf companion of GK Per.

Comparison, in the visibility plane, of radio observations of Cassiopeia A made at 151 MHz over a 2.3 yr interval indicate that the bulk of the radio emitting material has not been decelerated strongly.

Optical spectroscopy on nearly 50 filaments of 3C58 indicates a maximum expansion velocity of 1100 ± 100 km s−1. A considerable range in radial velocity with projected distance from remnant center is found suggesting that the remnant’s optical emission is not confined to a thin shell. Typical filament electron densities are between 200 - 500 cm−3 with Hα/[N II] ratios in the range 0.2 - 0.5. Optical extinction to 3C58 is modest with E[B-V] = 0.68±0.08. Preliminary radial proper motion measurements for a few outlying filaments suggest values of order 0.05” - 0.07” yr−1.

The fortuitous positioning of the Schweizer and Middleditch OB subdwarf behind SN1006 has permitted the detection and subsequent confirmation by IUE of broad (±5000 km/s) Fe II absorption features which probably arise from unshocked iron ejecta in the center of SN1006. The mass of detected Fe II, ∽0.012 M⊙, is however only 1/25 of the ∽0.3 M⊙ of Fe within ±5000 km/s predicted by carbon deflagration models. IR and optical observations exclude any appreciable iron in grains or Fe I, but high ion stages, Fe III and up, oould be present. Promising mechanisms for ionizing the unshocked iron in SN1006 include the radioactive decay of 44Ti, and photoionization by UV and X-ray emission from the reverse shock. Although the photoionization model works, insofar as it permits as much as 0.2 M⊙ of unshocked iron in the center of SN1006, agreement with the IUE data requires that the ejecta density profile be flatter, less centrally concentrated, than the W7 deflagration model of Nomoto, Thielemann, and Yokoi.

We have measured proper motions for fast, oxygen-rich knots in Puppis A, which we demonstrate are probably uncontaminated ejecta from the progenitor star’s core. Typical fast knots show motions of 0.1-0.2 arcsec yr-1 diverging from a point 4’ northeast of the center of the radio shell. A model assuming constant expansion fits the data well and gives an age of 3700 ± 300 yr for Puppis A. We also present new spectra which indicate the presence of neon along with oxygen in the fast knots.

We perform one dimensional numerical simulations with a Lagrangian hydrodynamics code of the adiabatic expansion of a supernova into the surrounding medium. The early expansion follows Chevalier’s analytic selfsimilar solution until the reverse shock reaches the ejecta core. We follow the expansion as it evolves towards the adiabatic blast wave phase. Some memory of the earlier phases of expansion is retained in the interior even when the outer regions expand as a blast wave. We find the results are sensitive to the initial configuration of the ejecta and to the placement of gridpoints.

Establishing that the interstellar pressure is higher than usually realized and that the magnetic component is probably dominant, I propose a drastic revision for our understanding of the interstellar landscape.

A global model for the ISM is described in which separate fluids of stars and gas interact through star formation, mass loss, stellar heating, and gaseous cooling processes. In all simulations, a steady state develops in which rarefied ionized regions separate ridges of neutral gas.

Observations of ring nebulae around WR and Of stars, including those in OB associations, have been used to give information on the pre-supernova environment.

A review is given of recent observations of the X-ray emission from supernova remnants carried out on the Einstein, Tenma and EXOSAT satellites as well as from a few sounding rocket experiments. Our current interpretation of the high resolution images, high resolution energy spectra and the first few spectrally resolved images is discussed.

Einstein observations of supernova remnants have been reviewed and analyzed. Images of 44 galactic remnants have been reprocessed, merged when necessary, and collected into a catalog. Some bright remnants were viewed with both moderate and high resolution instruments (IPC with 1’ resolution and HRI with 4” resolution). Some IPC images of nearby remnants have been separated into 2 energy bands, 0.2–0.6 keV and 0.6–4.5 keV; whereas most images cover the band 0.2–4.5 keV. The catalog consists of 72 images of the 44 remnants.

These images will be published in the form illustrated here. Contour levels are spaced geometrically as indicated below the figures and show the faintest observable features. The pictures are more linear and generally show only the brighter regions. Images are available now, however, in FITS format, on magnetic tape and may be obtained by writing to the author.

In this work we study the 2 – 10 keV spectra of eight relatively well known SNRs. These spectra were obtained with the medium energy experiment on-board the X-ray satellite EXOSAT. Details of the EXOSAT mission can be found in Taylor et al 1981 and the ME experiment is described in Turner, Smith and Zimmerman 1981. The ME experiment is well suited for this study since it provides a high sensitivity and a narrow field of view (45 ' FWHM). These spectra are considered in the context of a model in which the 2–10 keV continuum arises from shock heated interstellar material.