Measuring Polarization	Modeling Geometries	Resolving Spectra

Where do we come from? What are we? Where are we going?

The title of Paul Gauguin's well-known painting makes a handy list of questions that researchers hope to answer by studying supernovae.

At least in part, we are -- and we come from -- supernovae, which created all atoms more massive than iron and seeded the universe with almost everything heavier than helium. As it happens, our best chance to find out where we're going also lies with supernovae.

For where we (and the rest of the universe) are going depends on what dark energy is, the mysterious something that is stretching space at an accelerating pace. Many models of dark energy have been proposed; to distinguish among them, it's essential to know whether the expansion of the universe has been smooth and steady in the past or whether it has proceeded in a more complex fashion.

The keys to cosmic expansion are Type Ia supernovae, the standard candles of cosmic distance measurement. Their brightness tells us how far away they are and thus how long ago they exploded. Their redshift tells us how much the universe has expanded while their light was travelling to us.

"Where are we going?" The questions posed by Gauguin's painting may soon be answered by supernova studies.

Eliminating uncertainty

"Good as Type Ia supernovae are as standard candles, there is a residual uncertainty of a few percent in brightness measurements, and thus distance measurements," says Greg Aldering, head of the Nearby Supernova Factory (SNfactory) based at Berkeley Lab. "If that uncertainty is due to something that drifts with redshift, it would be a big problem for the proposed SuperNova/Acceleration Probe, the satellite now being designed for dark energy research."

On the other hand, if the uncertainty is due to asymmetry in Type Ia supernovae explosions, no other physical explanation is needed. The superiority of Type Ia supernovae for measuring the expansion rate of the universe would be confirmed.

The SNfactory originally set out to reduce systematic uncertainties by addressing questions like the effect of intervening dust on a Type Ia supernova's intrinsic colors and the effects of different degrees of "metallicity" in a supernova (metallicity is the presence of elements heavier than helium). The SNfactory's goal is to collect 300 to 600 supernovae, close enough to measure with great precision but far enough away to be "in the smooth Hubble flow," so that their redshifts are unlikely to be disturbed by neighboring galaxies.

A mosaic of supernovae captured by the Nearby Supernova Factory's automated search.

The SNfactory depends on automated sky searches that image thousands of random galaxies each night; the same regions are imaged again two and a half weeks later, and by comparing images, computer programs and human scanners quickly find any supernovae that have appeared in the interim.

Next, detailed spectrographic studies will be made of each supernova, its home galaxy, and the surrounding sky, using an instrument called SNIFS (Supernova Integral Field Spectrograph). Built largely by the French members of the SNfactory collaboration, SNIFS will soon be mounted on the University of Hawaii's 2.2 meter telescope on Mauna Kea.

The SNIFS spectrograph, newly assembled by the SNfactory's French colleagues at the Centre de Recherche Astronomique de Lyon (CRAL).

A great deal can be learned about Type Ia supernovae from their light flux alone. A striking example is SN 2002ic, found by Michael Wood-Vasey, a UC Berkeley graduate student who wrote much of the software the SNfactory uses to scan for supernovae.

"The definition of a Type I supernova is that it has no hydrogen lines in its spectrum," says Wood-Vasey. "SN 2002ic's spectrum has broad hydrogen emission features -- but in everything else it's a perfect Type Ia."

Follow-up studies by astronomer Mario Hamuy of the Carnegie Observatories and his colleagues quickly explained the seeming anomaly: the hydrogen is matter sucked from a giant star, the first incontrovertible evidence of a Type Ia supernova originating in a binary system. Says Aldering, "This is a smoking gun."

The promise of polarimetry

"A giant companion is a good explanation for the hydrogen lines in the spectrum of SN 2001ic," says Aldering, "but that supernova also shows an interesting high-velocity calcium feature, similar to the one in 2001el with the peculiar polarimetry. What if this turns out to be a common thread?"

As rewarding as studies of flux alone can be, says Aldering, "the biggest problem for the SNfactory is that lots of things can change spectral lines. For example, we don't know enough about atmospheres: if an explosion's atmosphere is not in perfectly spherical layers, like an onion, it can throw us off."

When Lifan Wang joined the SNfactory, he encouraged a heightened awareness of the role that supernova shapes may play in the appearance of spectral lines, and a new appreciation for the potential of polarimetry in settling thorny questions. Aldering says that "at first we were planning to do everything with our optical spectrograph only. Through Lifan's lobbying, we are moving to include polarimetry as well."

Polarimetry requires large and exceptionally stable telescopes -- multimirror telescopes just aren't steady enough -- and the competition for telescope time is as fierce as ever. But the rewards promise to be great.

One line of pursuit is the relation of supernova brightness and polarization. The brightness of an asymmetric explosion depends on the angle of view, and so does the degree and angle of polarization. Thus, says Aldering, "brightness should vary with polarization." If a consistent relationship of this kind can be established, it will be a big step toward eliminating systematic uncertainty in brightness measures of Type Ia supernovae.

"What fascinates me is trying to understand the largest explosions around," says supernova theorist Peter Nugent, but he stresses that studying stellar explosions is no simple matter. "In the vast majority we detect only subtle differences in flux. Spectropolarimetry is a hundred times better; we can really see the differences."

"Only recently have telescopes and computers become powerful enough to do these studies," says Dan Kasen. "It's fun detective work, but we don't have enough data yet, and it's not easily interpretable."

Making the best use of polarimetry will depend on gathering data from a great many supernovae. These will have to be the kind the SNfactory is searching for, "nearby," their distances measured in the millions, not billions, of light years.

The search is just getting underway. As it progresses, a garden of strangely shaped explosions will bloom in the night skies.

Additional information

• More about the Nearby Supernova Factory
• More about the Nearby Supernova Factory's recording-breaking "rookie year"