In general, supernova come in two flavors: The core collapse Type II supernovae, and the white dwarf over the Chandrasekhar limit, Type I supernova. The two are distinguished by the elemental composition of their spectra: Type II stars still have a hydrogen envelope and thus, hydrogen lines are prominent. Type I supernova, being the burned out cores of stars, don't have that envelope, so heavier elements, like silicon are present.
However, the Supernova 2002bj defies classification. Initially, it was classified as a Type II, but observation of its spectra has shown it to be more similar to a Type I (ie, it had the Silicon lines present), but it also had helium lines present which is uncharacteristic of that class. SN 2002bj also faded far more rapidly than should be expected for stars of the Type I classification.
So what are they?
The authors suggest this may be a new class of supernovae, called .1a supernovae, which was predicted theoretically, but had not yet been observed (or if it had, it wasn't realized).
The idea is that the system, instead of being composed of a white dwarf accreting mass from a red giant companion, the companion star was instead a fellow white dwarf. Models of this show that two white dwarfs in a sufficiently close binary orbit can transfer mass. Such systems are known as AM CVn stars.
If this sort of system were to have one star pass its Chandrasekhar limit, several predictions about the resulting supernova could be made:
- The supernova should be (relatively) faint and evolve quickly.
- The short timescale would allow for easier detection of short lived isotopes like 52Fe or 48Cr in addition to the normal isotopes we see in the afterglows of supernovae (I mentioned this decay process briefly in this post.
- These events should make up a few percent of all Type I supernovae observed.
Of these three criteria, 2002bj fits the first to a tee. The authors don't say whether or not these additional isotopes were observed. However they did hint that the higher than expected luminosity than that of this theoretical kind may be due to such short lived elements, but at the very least, this paper reinforces the idea that these should be something we keep an eye out for in the future. In regards to the last criteria, in the local region (to a distance of 60 Mpc) only 31 are Type I known, so this is still on track to meet the "few percent" criteria, but we shouldn't place bets with such a small sampling.
It will be interesting to see how this new class pans out in the future and just what new things we find to put in it.
Poznanski, D., Chornock, R., Nugent, P., Bloom, J., Filippenko, A., Ganeshalingam, M., Leonard, D., Li, W., & Thomas, R. (2009). An Unusually Fast-Evolving Supernova Science DOI: 10.1126/science.1181709