BERKELEY -- An exploding star first observed last September is the largest and most luminous supernova ever seen, according to University of California, Berkeley, astronomers, and may be the first example of a type of massive exploding star rare today but probably common in the very early universe.

The brightest stellar explosion ever recorded may be a long-sought new type of supernova, according to observations by NASA's Chandra X-ray Observatory and ground-based optical telescopes. Credit: Illustration: NASA/CXC/M.Weiss; X-ray: NASA/CXC/UC Berkeley/N.Smith et al.; IR: Lick/UC Berkeley/J.Bloom & C.Hansen Download larger image version here

Unlike typical supernovas that reach a peak brightness in days to a few weeks and then dim into obscurity a few months later, SN2006gy took 70 days to reach full brightness and stayed brighter than any previously observed supernova for more than three months. Nearly eight months later, it still is as bright as a typical supernova at its peak, outshining its host galaxy 240 million light years away.

UC Berkeley post-doctoral fellows Nathan Smith and David Pooley estimate the star's mass at between 100 and 200 times that of the sun. Such massive stars are so rare that galaxies like our own Milky Way may contain only a dozen out of a stellar population of 400 billion.

"This was a truly monstrous explosion, a hundred times more energetic than a typical supernova," said Smith, who led a team of astronomers from UC Berkeley and the University of Texas. "That means the star that exploded might have been as massive as a star can get, about 150 times that of our sun. We've never seen that before."

"Of all exploding stars ever observed, this was the king," said Alex Filippenko, UC Berkeley astronomer and leader of the ground-based observations at the University of California's Lick Observatory in California and the W. M. Keck Observatory in Hawaii. "We were astonished to see how bright it got, and how long it lasted."

This artist's illustration shows what the brightest supernova ever recorded, known as SN 2006gy, may have looked like. The fireworks-like material (white) shows the explosive death of an extremely massive star. Before it exploded, the star expelled the lobes of cool gas (red). As the material from the explosion crashes into the lobes, it heats the gas in a shock front (green, blue and yellow) and pushes it backward. Credit: NASA/CXC/M.Weiss Download larger image version here

Based on the Lick and Keck observations, plus data from the Chandra X-ray Observatory, Smith, Pooley, Filippenko and their colleagues argue that the stellar explosion was not your run-of-the-mill supernova, but a possible pair-instability supernova. They have submitted a paper describing the discovery and their conclusions to The Astrophysical Journal.

Stars with masses at least 10 times greater than our sun end their lives after burning hydrogen to helium, helium to carbon, and on to larger elements until they reach iron, when fusion stops. Toward the end of this process, the heat produced in the core of the star becomes insufficient to support the outer layers, which collapse inward, finishing the fusion process and crunching the core to a neutron star or black hole. The outer layers of the star are blown off in a bright flare-up we observe as a supernova.

For stars much more massive than this, ranging from 140 solar masses to as many as 250, the temperature at the core becomes so great that before the fusion cascade is complete, high-energy gamma rays in the core start annihilating one another, creating matter-antimatter pairs, mostly electron-positron pairs. Since gamma radiation is the energy that prevents collapse of the outer layers of the star, once the radiation starts disappearing, the outer layers fall inward. The net result is a thermonuclear explosion that, theoretically, would be brighter than any typical supernova. In this type of supernova, the star is blown to smithereens, leaving behind no black hole.

"This discovery forces us to go back to the drawing board to understand how the most massive stars die," Smith said. "Instead of just winking away into a black hole, they apparently can suffer these brilliant explosions that can be seen far across the universe. The fact that this thing is so bright, and stayed bright for a long time, makes our chances of detecting them in the early universe much better."

This graphic gives a summary of our best current understanding of the evolution of stars, showing their birth, middle age and eventual demise. Credit: NASA/CXC/M.Weiss Download larger image version here

Such pair-instability supernovas should theoretically produce a greater percentage of heavy elements. According to Smith, the radioactive decay of nickel-56 produces most of the light of a supernova, and this pair-instability supernova produced about 20 solar masses of nickel, compared to maybe 0.6 solar masses in a Type Ia supernova. Astronomers think that a large proportion of the universe's first stars were supermassive stars like this that, upon exploding, seeded the early universe with the heavy elements from which planets and later, humans, were made.

"We may have witnessed a modern-day version of how the first generation of the most massive stars ended their lives, when the universe was very young," Filippenko said.

The star that produced SN 2006gy apparently expelled a large amount of mass prior to exploding, reminiscent of the star eta Carinae, a so-called luminous blue variable which, at 100 to 120 solar masses, is the most massive star in our galaxy.

"This is also very exciting because it suggests that eta Carinae, only 7,500 light years away, might possibly explode in a similar manner, becoming a spectacularly bright star in our sky," Filippenko said.

"We don't know for sure if Eta Carinae will explode soon, but we had better keep a close eye on it just in case," added Mario Livio of the Space Telescope Science Institute in Baltimore, Md., who was not involved in the research. "Eta Carinae's explosion could be the best star-show in the history of modern civilization."

Estimated to be 100 times heftier than our Sun, Eta Carinae is one of the most massive stars in our galaxy. It may also turn out to be one of the shortest-lived, because results for SN 2006gy suggest that it may be destroyed by a supernova at any time. This Hubble telescope image of Eta Carinae shows a huge, billowing pair of gas and dust clouds, caused by a giant eruption about 160 years ago, when it became one of the brightest stars in the southern sky. Credit: NASA/N.Smith & J.Morse Download larger image version here

University of Texas graduate student Robert Quimby first observed the supernova on Sept. 18, 2006 in the galaxy NGC 1260, located in the constellation Perseus. Filippenko's team immediately began observing it with its dedicated supernova search and monitor telescope at Lick, the Katzman Automatic Imaging Telescope.

Filippenko and his graduate student Ryan Foley subsequently obtained spectra of the star using the Lick 3-meter Shane telescope and the DEIMOS spectrograph mounted on the Keck II telescope.

Pooley led the Chandra observation, which allowed the team to rule out the most likely alternative explanation for the supernova, namely that it was an explosion of a white dwarf star into a dense, hydrogen-rich environment.

"If that were the case, this supernova would have been 1,000 times brighter in X-rays than what we detected with Chandra," said Pooley. "This must have been an extremely massive star."

"In terms of the effect on the early universe, there's a huge difference between these two possibilities," said Smith. "One pollutes the galaxy with large quantities of newly synthesized elements, and the other locks them up forever in a black hole."

"One exciting repercussion of this is that, if pair-instability supernovas really are this bright, it gives us hope that the James Webb Space Telescope might actually be able to detect these explosions from the first stars, thereby verifying that they may actually exist," he added.

The results from Smith, Pooley, Filippenko and their colleagues, including Weidong Li, Ryan Chornock, Jeffrey M. Silverman, Joshua S. Bloom and Charles Hansen of UC Berkeley and J. Craig Wheeler of the University of Texas, will appear in The Astrophysical Journal.

NASA's Marshall Space Flight Center, Huntsville, Ala., manages the Chandra program for the agency's Science Mission Directorate. The work also was supported by the National Science Foundation and the W. M. Keck Foundation.