There is, it seems, more than one way to create an exploding star.
That's what scientists studying the origins of type 1a supernovae - important because they help to measure the accelerating expansion of the Universe and dark energy - have found.
A team, including Mark Sullivan of Oxford University’s Department of Physics, has reported in Science observations that suggest weaker stellar explosions from giant stars contribute to some of these bright supernovae. I asked Mark about the ‘second star’ mystery and dark energy…
OxSciBlog: Why study this particular supernova?
Mark Sullivan: This supernova ('PTF11kx') was found by the Palomar Transient Factory (PTF) in a galaxy 600 million light years away, quite close by cosmological standards, although still far too distant to directly view the star that exploded. The supernova was quickly identified as a 'type 1a Supernova', the same type that can be used as standard candles to measure astronomical distances, and that were used in the 1990s to detect the effects of dark energy.
PTF has found more than 1000 Type 1a Supernovae, and, at first, we thought this supernova was just like any other of its type. However, our data soon showed this was not the case - we saw dramatic evidence that the supernova was surrounded by significant amounts of gas arranged in shells, containing hydrogen, calcium, sodium and other elements. These shells were moving at different speeds, with the outermost shells travelling the slowest.
Such evidence is extremely unusual in these types of supernovae, and it prompted us to study this explosion in considerable detail to try and understand where this material had come from.
OSB: How do we think it formed?
MS: We think that this supernova, like all type 1a supernovae, formed from the explosion of a very compact white dwarf star in a binary system.
The white dwarf steals material from its companion until it cannot grow any bigger, at which point it explodes as a supernova some 10 billion times brighter than our sun. For many decades the nature of the 'second star' has been debated - is it another white dwarf star, or is it a much bigger star such as a red giant? This supernova helps answer this question. Our hypothesis was that the gas surrounding the supernova had been cast off in previous 'nova' eruptions, decades before the supernova explosion itself occurred.
Novae are much more frequent, weak explosions that do not destroy the star (unlike the supernova). Material blown off the red giant in a stellar wind lands on the white dwarf, and, as the material builds up, periodically explodes as a nova eruption. Such systems are well known - for example, a star in our own galaxy, RS Oph, has these explosions every 20 years or so.
This naturally explained the presence of the gas surrounding PTF11kx, and even the shells of material - the different shells correspond to different historical nova eruptions, with the outermost, oldest shells having been slowing down for longer than the innermost, younger shells. Two months after explosion of PTF11kx we saw the most compelling evidence of all - the supernova ejecta itself slamming into material left over from one of these previous nova eruptions.
Theoretical studies indicate that white dwarfs lose more mass in these nova eruptions than they gain from the red giant, and hence many astronomers concluded that novae could not produce type 1a supernovae. This new study is the first observational evidence that they can.
OSB: What does it tell us about how supernovae in general form?
MS: Previous results from PTF on the closest type 1a supernova for 25 years, have shown that that event could not have been a nova before it went supernova - the red giant star would survive the explosion, and this was not seen. So it is very unlikely that novae could explain all type 1a supernovae. So this new observation is compelling evidence that nature has more than one way to make a type 1a supernova explosion!
Predicting the exact number of supernovae that may arise from novae is difficult, as many of the signatures of the novae will depend on the angle at which the supernova explosion is seen from the Earth. If we had seen this explosion from a different perspective, it might well have been missed, or looked very different. But we estimate that novae give rise to more than 0.1 percent of all type 1a supernovae, but less than 20 percent.
OSB: How might these findings affect the search for dark energy?
MS: This result does not diminish the current evidence that we have for dark energy and the accelerating universe. But it does point to how we might make better measurements in the future. If there really are two (or more!) ways to make a type 1a supernova, then techniques that can identify these in astronomical data will be extremely valuable.
For example, some recent studies have shown that type 1a supernovae are not perfect standard candles, but instead their brightness depends on the type of galaxy in which they explode. At the moment, there is no good theory as to why this might occur - but the idea that these supernovae may come from different progenitor systems, and hence potentially have slightly different brightnesses, could help understand this mystery.
The Palomar Transient Factory is an international collaboration of scientists and engineers from the California Institute of Technology, DOE’s National Energy Research Scientific Computing Center at Lawrence Berkeley National Laboratory, NASA’s Infrared Processing and Analysis Center, the University of California at Berkeley, Las Cumbres Observatory Global Telescope Network, the University of Oxford, Columbia University, the Weizmann Institute of Science in Israel, and Pennsylvania State University.