A new analysis of the Schweizer-Middleditch star, a hot subdwarf which lies along the same line-of-sight as the centre of the SN1006 SNR, has allowed us to place tighter constraints on its atmospheric parameters, and re-assess its distance. Since Wellstein et al. (1999) have demonstrated that the remnant of the donor star in a pre-SNIa binary system could appear as a hot subdwarf, albeit with an abnormally low mass, we can now re-address Schweizer & Middleditch's original question: is the SM star the stellar remnant of one component of the SNIa progenitor binary?
In order to begin answering this question, we need to convince ourselves that the SM star lies at the same distance as the SN1006 SNR. Unfortunately, there is a large range in the SNR distance estimates quoted in the literature. In Table 1, we list the various distance estimates to the SN1006 SNR itself and the method used to obtain that distance. Early estimates, based for example on the historical record of its brightness (e.g. Minkowski 1966) and early models of the X-ray emission, gave distances 1kpc. Most of the more recent estimates, based on a variety of theoretical models or measurements of e.g. the expansion velocity or proper motion of optical filaments, place the SNR at a distance of 1.5-2.0kpc. The one glaring exception is the estimate of Willingale et al. (1995), 0.70.1kpc, based on an analysis of the ROSAT PSPC X-ray image of the SNR.
Table 1. Distance estimates to the SN1006 SNR from the literature
We find the distance to the SM star 1050d2100 pc, assuming that it is an ordinary hot subdwarf. If Willingale et al's distance estimate is correct, then the SM star would lie a long way behind the remnant. In order for it to lie within the remnant, it would have to be of unusually low mass. A mass of 0.1-0.2 gives a distance compatible with Willingale et al's estimate, and in that scenario the SM star could indeed then be a remnant of the donor star in an SNIa progenitor system.
However, if Willingale et al's SNR distance estimate is wildly inaccurate, and the more conservative estimates of 1.5-2.0kpc are correct (Winkler & Long 1997), then the SM star cannot be a low mass remnant of the donor star in a pre-SNIa binary.
In fact, there are two more compelling arguments against the SM star having any relation to SN1006. Firstly, it is located 2.5' south of the projected centre of the remnant, and would have to possess a proper motion of 0.15" per year and a velocity of 800km sec-1 to have reached its current location. Unfortunately, the star simply does not possess this motion or velocity. Secondly, the presence of red-shifted metal absorption lines superimposed on the SM star's UV spectrum strongly indicate that the star is behind the remnant, since these features almost certainly originate at a shock front on the remnant's far side. Confirmation of this may come from observations of other nearby objects with strong UV fluxes and generally featureless far-UV continuums. Indeed, P.F. Winkler has an HST/STIS program to observe four such objects behind SN1006 during Cycle 8 (two QSOs and two A0 stars, program ID 8244), and one of these objects is even closer to the projected centre of SN1006 than the SM star. These targets are not scheduled to be observed until June-July 2000, but the detection of the same red-shifted features as seen in the SM star (and the non-detection of any additional features with separate velocities) would effectively rule out any exotic origin for these lines, and confirm the location of the SM star behind the SN1006 SNR.
Thus, the SM star can only be the remnant of the donor star in a pre-SNIa binary, such as might have produced SN1006, if the following four criteria are fulfilled: (1) The star has an unusually low mass for a hot subdwarf (0.1), (2) the low distance estimate to the SN1006 SNR of Willingale et al. (1995) is correct, (3) the red-shifted metal lines seen in the SM star's far-UV spectrum originate somewhere other than on the far side of the SNR, and (4) the SM star has a high proper motion and transverse velocity. Unfortunately, at the time of writing, none of these conditions can convincingly shown to be true. However, the tighter constraint we have been able to place on the distance to the SM star in this analysis can now be used to place an upper limit on the distance to the SN1006 SNR itself, and hence constrain the models and methods used to estimate the distances of supernova remnants.
© European Southern Observatory (ESO) 2000
Online publication: April 10, 2000