SN 1987A is distinctive among X-ray detected supernovae. It is the only detected Type IIL supernova, and has by far the lowest luminosity of any X-ray detected SN. The detection can be ascribed to its proximity, at a distance of 50kpc in the LMC. As can be seen in Fig. 17.14, its low flux demonstrates why similar, more distant supernovae have not been detected. As the nearest supernova in 400 yrs, it has been the subject of intensive observation in all wavelengths. While several comprehensive reviews provide detailed descriptions of the phenomenology and theoretical interpretation (e.g., [101]), the object continues to evolve rapidly. In the X-ray band, the a O

1 Ne ' O Mg

Ar 1 Ca

Fe Fe

Fig. 17.16 The ACIS spectrum of SN 1998S shows that the spectrum arises in shock-heated gas and shows evidence of ejecta [121]

Energy (keV)

Fig. 17.16 The ACIS spectrum of SN 1998S shows that the spectrum arises in shock-heated gas and shows evidence of ejecta [121]

object has changed dramatically since the most recent publications, and it continues to be observed intensively by Chandra and XMM-Newton.

SN 1987A was discovered optically on January 27, 1987. At the time, the only orbiting X-ray observatories were the HEXE instrument on the Mir space station and Ginga. Observations over the first 5 months after the explosion detected no prompt signal [31]. Models suggested that X-rays downscattered from the y-radiation from decaying Ni thought to power the declining optical light curve would eventually be detected as the outer expanding envelope decreased in density. Nevertheless, it was a surprise when hard X-rays were independently detected by HEXE and Ginga 6 months after the explosion [31,153]. This surprising result was interpreted as evidence that the ejecta were mixed early during the expansion. Ginga also detected an iron line from the end product of the radioactive Ni decay. Ginga and HEXE followed the hard light curve for several months until SN 1987A became undetectable.

While initial ROSAT observations in June of 1990 yielded only an upper limit, SN 1987A was detected 8 months later, in February 1991 [12,40]. The supernova was initially detected at a 0.2-2.4 keV band luminosity of — 1034ergs_1, several orders of magnitude lower than any previous or subsequent supernova. This flux was interpreted as arising from the interaction of the supernova ejecta with circumstellar material from the blue giant wind.

ROSAT and ASCA monitored SN 1987A throughout their lifetime. From 1991 through 1995, ROSAT detected a nearly linear flux increase, by a factor of 5-6 from the initial detection [48]. Contemporaneous ASCA observations also showed a linear brightness increase. The average 0.5-10.0 keV spectrum during this interval was consistent with a 1.8 keV plasma with sub-solar metal abundances [75].

The object's light curve steepened markedly around the time of the launch of Chandra and XMM-Newton, allowing the full power of these observatories to be brought to bear. The initial Chandra observation resolved the emission into a ring consistent with the inner bright optical ring thought to be a equatorial band of circumstellar material, indicating that the shock was now encountering this inner ring [18]. An early grating spectrum revealed lines of N, O, Ne, Mg, and Si. Broadening of the composite line leads to an inferred shock velocity of (3400 ± 700) km s-1 [103].

Subsequent observations have followed the spectacular evolution of the supernova [112]. As the interaction between the forward shock and the band has evolved, Chandra has observed the ring to grow in diameter and overall brightness, and has resolved individual features along the ring brightening and dimming. The flux has increased exponentially; as of day 6 500 the supernova was approaching a 0.52.0 keV flux 50 times higher than the initial ROSAT detection. The most recent observations suggest the development of a new emission component, presumably associated with circumstellar material not in the equatorial band. At the same time, a decrease in the overall expansion rate by a factor of two is observed, suggesting strong interaction with the dense ring material. No evidence for a central compact object has yet been found.

A deep Chandra grating observation reveals the presence of numerous lines from multiply-ionized O, Ne, Mg, S, Si, and Fe [183]. Many lines are broadened, and inferred expansion velocities range from 300 to 1700km s-1, much smaller than the global radial expansion measured from images. The lower velocities are interpreted as arising in shocks produced as the supernova blast wave encounters dense protrusions on the ring.

17.5.2 SN1993J

SN 1993J, in the nearby spiral M81 (3.4Mpc), is the most intensively monitored X-ray SN after SN 1987A. It was initially detected 6 days after its discovery in the visible band, at the time the earliest X-ray detection of a SN [185]. Its early evolution was monitored extensively using ROSAT [70] and ASCA [163]. More recently, it has been observed using both Chandra and XMM-Newton [155,184]. It is the only SN besides SN 1987A for which a substantial number of X-ray spectral observations have been performed.

The early detection indicated the presence of dense circumstellar material in close proximity to the progenitor star. For the first month after the explosion, the X-ray spectrum was too hot to be characterized by ASCA (kT > 30keV; [83]), and the detection by OSSE instrument on the Compton Gamma Ray Observatory suggests an initial temperature on the order of 109 K [92].

ASCA observations between days 8 and 572 show dramatic spectral evolution [163]. A second, cooler component emerges (kT — 1keV), with a higher column density than the hot component. The two components are consistent with the expected emission from the forward and reverse shocks. The spectra soften dramatically with time as temperatures of both components decrease. Additionally, the column density to the reverse shock decreases, making the reverse shock the dominant contributor to the soft flux. Contemporaneous ROSAT PSPC data, which cover a narrower band, reflect this softening, with the 0.2-2.4 keV spectra prior to day 60, too hot to be constrained, and the spectra between days 200 and 400 consistent with a 1keV temperature [70]. The most recent high quality spectra from XMM-Newton and Chandra show that this trend continues. The Chandra spectrum (taken on day 2594) requires three components, two with low temperatures (0.35 ± 0.06 and 1.01 ± 0.05 keV) absorbed by a relatively high column density (4 x 1021 cm-2), and a high temperature component (kT = 6.0 ± 0.9 keV) with Galactic absorption (5 x 1020 cm-2). The low temperature components have some metal abundances deviant from solar. The XMM-Newton spectrum is fit by a two-component model with temperatures 0.34± 0.04keV and 6.54 ±4keV. The column density for both components is consistent with the larger of the two Chandra values (4x 1021 cm-2). Metal abundances are consistent with solar.

The 0.3-2.4 X-ray light curve between 1993 and 2001 shows remarkable structure, with a constant decay proportional to t-0 3 for the first 200 days, followed by dips and rises relative to this constant decay. The t-0 3 slope suggests an average CSM density profile p-s, where s — 1.65. The variations about this average have been interpreted as the result of mild density variations. From both the average slope and the variations it can be inferred that the mass loss rate of progenitor star was not constant, decreasing slowly on average as the explosion approached, but with brief episodes of lower and higher mass loss rates a few thousand years before the explosion.

Was this article helpful?

0 0

Post a comment