Recently, a new theme, i.e. the physics of supernova explosions, has been added to this list by a discovery derived from ROSAT survey observations of the Vela supernova remnant. The constellation Vela was scanned by the ROSAT X-ray telescope during the all-sky survey between October 1990 and January 1991. The image including the Vela SNR region is shown in Fig 1. Prior to the ROSAT observations only the X-ray brightest parts, marked in red and white in Fig 1, were known. The ROSAT image, now, reveals that the SNR is much larger at lower surface brightness levels. The faint western parts are crossed by a long arc-shaped filament, which has been noticed in Einstein-Observatory observations, but Figure 1 shows that Vela extends significantly even beyond the filament towards the west. At an even lower surface brightness level, a factor ~500 less than the brightest patch in the north of the remnant, a faint shell appears, the contour of which is well matched by a circle of 8.3° or 72 pc diameter. Between position angles p.a. 240° and 330° - measured from north over east centered on the pulsar position - the remnant extends further out to 5.2° (45 pc) along p.a. = 283°. The Vela pulsar PSR 0833-45, which powers a small surrounding synchrotron nebula is significantly offset by about 1.4° towards the west from the centre of the bright area, which led some authors to speculate about a non-association of the pulsar with the remnant or an unrealistically huge space velocity of the pulsar. However, within the new boundary the pulsar's current position is offset from the centre of the best-matching circle by just (25 ± 5) arcmin, which is about one tenth of the remnant's radius.
Figure 1: ROSAT all-sky survey image (0.1 - 2.4 keV) of the Vela supernova remnant; north is up and east is left; the angular resolution is 1 arcmin half power radius; mean exposure is 993 s. North is up and east is left. The present location of the Vela pulsar PSR 0833-45 is marked by the small dark cross. The direction of the proper motion vector of the pulsar is marked by the long arrow pointing to the north-west. The geometric centre of the remnant defined by the circle best-matching the circumference is marked by the big dark cross. Close to the north-west corner the bright Puppis-A SNR appears, which lies behind the Vela SNR. Surface brightness increases from light blue over yellow and red to white by a factor of 500. The blue stripe in the lower left-hand corner is due to unremoved scattered solar X-rays. Six objects coasting outside the remnant's boundary are found, labelled (A - F). Their symmetry axes intersect between the remnant's centre and the present pulsar position.
Figure 1 also shows at least 6 extended features, labelled A - F, well outside of the remnant's boundary. The area containing feature D shows evidence for the superposition of actually two bow-shaped objects. The feature D shows a bright head like feature A, whereas D' further west looks like a diffuse arc. Five (B, C, D/D', F) of the protruding objects show a distinct 'boomerang' type structure which opens towards the SNR centre. The other two objects A & E look like truncated cones, opening towards the remnant's centre as well. They extend from the general SNR boundary by 1.2° and 2.4°, or for an estimated distance to Vela of 500 pc, by ~10 pc and ~22 pc, respectively. The symmetry axes of these 6 structures intersect each other close to the remnant's geometric centre. The intersection of their symmetry axes with the known proper motion vector of the pulsar of (µ(alpha), µ(delta)) = (-0.040±0.004, 0.028±0.002) arcsec/yr, gives the objects' origin at (14.9±7.2) arcmin away from the present pulsar position along the proper motion axis towards the south-east. The proximity of the remnant's geometric centre and the origin of the 6 protruding objects to the pulsar position strongly favors their common origin; i.e. one single event gave rise to the pulsar, the remnant and the precursors. Furthermore, the angular offset of the objects' origin divided by the proper motion velocity of 0.049 arcsec/yr can be used to derive a new and independently determined age of the Vela SNR of (18000±9000) years, assuming the objects' origin marks the explosion site of the progenitor star. Within the error range this datum is consistent with the pulsar's spin-down age t(s) = 11200 years; t(s) = P/(2·P'), with P the pulsar period and P' its time derivative.
We suggest that the X-ray emission associated with the protruding objects is produced by shock-heating of the ambient medium by supersonic motion of the objects. This is strongly supported by the recent observation of non-thermal radio emission from the leading edge of object A using the Very Large Array in the USA. Four of the protruding objects are trailed by wakes of X-ray emission, which we propose to be Mach cones, extending back to the blast wave front of the Vela SNR. The Mach number is related to the opening angle of the cone, and Mach numbers between 2.4 and 4.0 are found. These low Mach numbers imply that the temperature of the ambient medium through which the objects are moving must be high, close to X-ray temperatures. Outside the general Vela SNR boundary we have found in the ROSAT all-sky survey data such an excess emission of almost constant surface brightness in a circular area with a diameter of 20° and the center within the Vela SNR boundary. The analysis of the ROSAT PSPC spectrum shows the temperature of this region to be kT(amb) = (0.10±0.02) keV.
Given the Mach number and kT(amb) we can determine the current velocity v and the expected temperature kT(exp) of each of the objects from standard shock relations. Of course, we can determine - in a totally independent way - the temperature kT of each of the protruding objects from its ROSAT PSPC spectrum. A comparisionof kT(exp), inferred primarily from geometry, with kT, determined from measured spectra, shows very good agreement within better than 20% for each of the objects, which strongly supports our view that we see shock-heated gas confined in Mach cones.
With the spectral data of the objects available a coarse estimate of the mass of each object can be derived by equating the loss of kinetic energy with the thermal energy content, including the wake contribution. Using the currently observed velocity, the mass is in the range between 0.01 - 0.5 solar masses per object.
We mention two explanations for the origin of the protruding objects: They could be due to interstellar clouds which have been accelerated at early times by the Vela blast wave to comparable speed. However, it is doubtful that clouds of the size and mass observed can survive such an encounter. Alternatively, and much more likely, the objects represent blobs of matter formed in the collapse of the progenitor star and expelled in the subsequent supernova explosion. They will travel largely unaffected through the hot, tenuous, post-shock plasma behind the blast wave and will pass its front at a late stage, when the blast wave has significantly decelerated. The suggestion for such a scenario, in which the explosion of some supernovae might resemble more that of a splinter bomb than that of a pressure bomb - giving rise to long-living confined explosion fragments -, has been made in the past. However, only very recently the results of two-dimensional hydrodynamical calculations of supernova explosions of massive stars have indicated that supernova explosions are not spherically symmetric. Asymmetries arise from instabilties excited by the passage of the explosion shock wave through the stellar interior. Rayleigh-Taylor instabilities form in the outer layers of the progenitor star near chemical transitions zone, e.g. the H/He interface and the He/CO interface. Convective instabilities form close to the mantle of the nascent neutron star. Highly heterogeneous clumps of matter so-called 'mushrooms' are formed by these instabilities. However, the two types of instabilities grow on different angular scales: whereas hundreds of clumps are produced by the Rayleigh-Taylor instabilities only about 40 - 60 clumps result from the convective instability deep inside the star.
The fact that we have observed objects in an SNR which is associated with a neutron star lends support to the presence of these types of instabilities in type Ib or type II supernova explosions of massive stars. Whether we observe the effects which went on in the outer or deep inner layers remains to be seen. However, the spectrally resolved ROSAT image shows a tendency for the bulk of the Vela SNR emission to break up into about 50 individual regions, wich supports the idea that the clumps formed deep inside the star drive the star into pieces conserving their number. This present hypothesis is now being followed by hydrodynamical calculations, which are not terminated at the instance of the explosion but continue to follow the growth and fate of the instabilities throughout the star and the ambient medium. By comparison with the ROSAT observations the range of the currently unbounded parameters of the explosion process will be known much better then, so that we eventually get a major step closer in understanding why and how massive stars explode.
(Bernd Aschenbach, from the Annual Report 1994)
Aschenbach, B., R. Egger, and J. Trümper, Nature 373, 587-590 (1995)
Get abstract from ADS service.
Get gzipped images: high-resolution figure 1 - brightness image - energy image
All rights reserved.
© Max-Planck-Institut für Extraterrestrische Physik, Postfach 1603, 85740 Garching, Germany.