Astronomers used to believe that all Type 1a supernovae were essentially the same brightness. That’s because they explode with the same amount of fuel. But now a supernova has been discovered that’s twice as bright as all the other Type 1a supernovae. This is a problem, since this kind of supernovae are used as standard candles, to determine distances across the Universe. Most recently, these supernovae have been used to calculate the mysterious force called dark energy that seems to be accelerating the expansion of the Universe.
A group of scientists affiliated with the SuperNova Legacy Survey (SNLS) have found startling evidence that there is more than one kind of Type Ia supernova, a class of exploding stars which until now has been regarded as essentially uniform in all important respects. Supernova SNLS-03D3bb is more than twice as bright as most Type Ia supernovae but has much less kinetic energy, and appears to be half again as massive as a typical Type Ia.
The lead authors of the report, which appears in the September 21 issue of Nature, include Andrew Howell, formerly of the Physics Division at Lawrence Berkeley National Laboratory and now at the University of Toronto, and Peter Nugent, an astrophysicist with Berkeley Lab’s Computational Research Division. Other lead authors are Mark Sullivan of the University of Toronto and Richard Ellis of the California Institute of Technology. These and many of the other authors of the Nature paper are members of the Supernova Cosmology Project based at Berkeley Lab.
Because almost all Type Ia supernovae found so far are not only remarkably bright but remarkably uniform in their brightness, they are regarded as the best astronomical “standard candles” for measurement across cosmological distances. In 1998, after observations of many distant Type Ia supernovae, the Supernova Cosmology Project and the rival High-Z Supernova Search Team announced their discovery that the expansion of the universe is accelerating – a finding that would soon be attributed to the unknown something called dark energy, which fills the universe and opposes the mutual gravitational attraction of matter.
“Type Ia supernovae are thought to be reliable distance indicators because they have a standard amount of fuel – the carbon and oxygen in a white dwarf star – and they have a uniform trigger,” says Nugent. “They are predicted to explode when the mass of the white dwarf nears the Chandrasekhar mass, which is about 1.4 times the mass of our sun. The fact that SNLS-03D3bb is well over that mass kind of opens up a Pandora’s box.”
Why Most Type Ia Supernovae are the Same
Classification of supernova types is based on their spectra. Type Ia spectra have no hydrogen lines but do have silicon absorption lines, a clue to the chemistry of their explosions. The white dwarf progenitors of Type Ia supernovae, typically about two-thirds the mass of the sun, are thought to accrete additional mass from a binary companion until they approach the Chandrasekhar limit. Increasing pressure causes the carbon and oxygen in the center of the star to fuse, producing the elements up to nickel on the periodic table; the energy released in this process blows the star to pieces in a titanic thermonuclear explosion.
Some variations have been observed in Type Ia supernovae, but these are mostly reconcilable. Brighter Type Ia’s take longer to rise to maximum brightness and longer to decline. When the time-scales of individual light curves are stretched to fit the norm, and brightness is scaled according to the stretch, Type Ia light curves match.
Brightness differences could be due to differing ratios of carbon and oxygen in the progenitors, resulting in differing final amounts of nickel in the explosion. The radioactive decay of nickel to cobalt and then iron powers the optical and near-infrared light curves of Type Ia supernovae. Differences in apparent brightness could also be products of asymmetry; an explosion viewed from one angle may be slightly dimmer than from another.
None of these possible differences are enough to explain supernova SNLS-03D3bb’s extreme brightness – which is much too bright for its light-curve “stretch.” Moreover, in most brighter supernovae, the matter ejected from the explosion travels at a higher velocity; that is, these explosions have more kinetic energy. But the ejecta of SNLS-03D3bb were unusually slow.
“Andy Howell put two and two together and realized that SNLS-03D3bb must have super-Chandrasekhar mass,” says Nugent.
The Mass of Evidence
One clue was the elements needed to produce the extra brightness. “All the power in a Type Ia comes from the burning of carbon and oxygen to heavier elements, notably nickel 56,” Nugent says. “A Type Ia of normal brightness makes about 60 percent of a solar mass worth of nickel 56, the rest being other elements. But SNLS-03D3bb is more than twice as bright as normal; it must have more than twice as much nickel 56. The only way to get that is with a progenitor that’s 50 percent more massive than the Chandrasekhar mass.”
The other factor is the slowness of SNLS-03D3bb’s ejecta, as detected in the shifting of elemental lines in its spectrum. The velocity of supernova ejecta depends on the kinetic energy released in the explosion, which is the difference between the energy released in thermonuclear burning minus the binding energy that acts to hold the star together, a function of the star’s mass. The more massive the star, the slower the ejecta.
But how could a carbon-oxygen progenitor ever accumulate mass greater than the Chandrasekhar limit without exploding? It’s possible that a very rapidly spinning star could be more massive. It’s also possible that two white dwarfs, with a combined mass well over the Chandrasekhar limit, could collide and explode.
Nugent says, “One clue came from our coauthor Mark Sullivan, who in the SNLS data had already found two distinct rates for the production of Type Ia supernova. They can crudely be broken into those that come from young star-forming galaxies and those from old, dead galaxies. So there’s an indication that there may be two populations of Type Ia’s, with two types of progenitors, and two different paths to explosion.”
In old, dead galaxies even the biggest stars are small, Nugent explains. The only kinds of Type Ia supernovae possible in these galaxies are likely to be the binary-system, mass-accreting, Chandrasekhar-mass type. But young star-forming galaxies produce massive objects and could be rich in white-dwarf plus white-dwarf binary systems, so-called “double-degenerate” systems.
“If the double-degenerate model is right, such systems will always produce super-Chandrasekhar explosions in these very young galaxies,” Nugent says.
Young galaxies are more likely to be found in the early universe, and thus at greater distances. Since distant Type Ia supernovae are crucial to the effort to measure the evolution of dark energy, it becomes essential to clearly identify Type Ia supernovae that do not fit the Chandrasekhar-mass model. This is easy to do with a Type Ia as odd as SNLS-03D3bb, but not all super-Chandrasekhar supernovae may be so obvious.
“One way to detect super-Chandrasekhar supernovae is by measuring ejecta velocity and comparing it with brightness. Another way is by taking multiple spectra as the light curve evolves. Unfortunately, taking spectra is the biggest expense in the whole pursuit of dark energy studies,” Nugent says. “The designers of these experiments will have to find efficient ways of eliminating super-Chandrasekhar supernovae from their samples.”
Modeling the Variations
It was partly in hopes of developing a quick and dependable way to identify candidate Type Ia supernovae for cosmological research that Nugent and coauthor Richard Ellis initially approached Sullivan and other members of the SNLS, with its large data base of supernovae. Working at the National Energy Research Scientific Computing center (NERSC) based at Berkeley Lab, Nugent developed an algorithm that could take a handful of photometric data points early in the evolution of a candidate supernova, positively identify it as a Type Ia, and accurately predict its time of maximum brightness.
One of the first Type Ia’s studied this way turned out to be SNLS-03D3bb itself. “It had such a high signal-to-noise ratio given its redshift that we should have suspected from the beginning that it was going be an unusual supernova,” Nugent says.
Nugent regards the discovery of the first demonstrable super-Chandrasekhar supernova as an exciting prospect: “For the first time since 1993” – when the brightness versus light-curve shape relationship was developed – “we now have a strong direction to look for the next parameter that describes the brightness of a Type Ia supernova. This search may lead us to a much better understanding of their progenitors, and the systematics of using them as cosmological probes.”
This understanding is one of the major goals of the Computational Astrophysics Consortium, headed by Stan Woosley of the University of California at Santa Cruz and supported by the Department of Energy’s Office of Science through the Scientific Discovery Through Advanced Computing (SciDAC) program, with Nugent and John Bell of the Computation Research Division and NERSC among the leading partners.
“Chandrasekhar’s 1931 model of stellar collapse was elegant and powerful; it won him the Nobel Prize,” says Nugent. “But it was a simple one-dimensional model. Just by adding rotation one can exceed the Chandrasekhar mass, as he himself recognized.”
With 2-D and 3-D models of the supernovae now possible using supercomputers, Nugent says, it’s possible to study a broader range of nature’s possibilities. “That’s the goal of our SciDAC project, to get the best models and the best observational data and combine them to push the whole ball of wax. At the end of this project, we’ll know the most we can know about all kinds of Type Ia supernovae.”
“A type-Ia Supernova From a Super-Chandrasekhar Mass White Dwarf Star,” by D. Andrew Howell, Mark Sullivan, Peter E. Nugent, Richard S. Ellis, Alexander J. Conley, Damien Le Borgne, Raymond G. Carlberg, Julien Guy, David Balam, Stephane Basa, Dominique Fouchez, Isobel M. Hook, Eric Y. Hsiao, James D. Neill, Reynald Pain, Kathryn M. Perret, and Christopher J. Pritchett, appears in the 21 September issue of Nature and is available online to subscribers.
Berkeley Lab is a U.S. Department of Energy national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California. Visit our website at http://www.lbl.gov.
Original Source: LBL News Release