The results were published in the Astrophysical Journal Letters on October 20 by Alexei Khokhlov, Elaine Oran, and Almadena Chtchelkanova of NRL and Peter Hoeflich, Lifan Wang, and J. Craig Wheeler of the University of Texas.

"Combining Texas observations with the cutting-edge numerical techniques at NRL has pointed the way to a new idea," says Wheeler, the Samuel T. & Fern Yanagisawa Regents Professor in Astronomy at the University of Texas at Austin. "We think that jets cause a major class of supernova explosions."

Supernovae are caused by the explosion of a massive star, and the explosions have been thought to arise through one of two mechanisms. In the first type, called Type Ia, massive stars can explode like a stick of dynamite, leaving no collapsed remnant.

Astronomers use Type Ia supernovae as "standard candles" to measure distances in the universe, and studies of Type Ia supernovae have suggested that the expansion of the universe is accelerating.

Other types of supernovae involve the collapse of the center of an especially massive star to form an extremely dense object, either a neutron star or, perhaps in some circumstances, a black hole. The formation of a neutron star is thought to be more common. These types of supernovae are called Type Ib and Ic and Type II.

Astronomer Lifan Wang, a Hubble Postdoctoral Fellow at the University of Texas at Austin, has studied all types of supernovae for several years, primarily using the 2.1-meter Otto Struve Telescope at McDonald Observatory.

Wang's work has focused on determining whether the light of supernovae is polarized - that is, if the light waves given off by supernovae are aligned in certain directions. If a supernova's light is expanding uniformly in all directions, there is no polarization. There will be measurable polarization if light from the parts of the supernova is spreading asymmetrically.

All the supernovae Wang has examined that are thought to arise from core collapse-the Type Ib and Ic and Type II supernovae-have been substantially polarized, and hence substantially "out-of-round." At the same time, all the Type Ia supernovae have shown little or no polarization.

For the polarized supernovae, Wang has identified a trend suggesting that the closer one looks to the center of a supernova explosion, the larger the asymmetry found. In many cases, his data suggest, the explosion must be occurring strongly along a preferred axis. The explosion must be bipolar. "These observations cannot be explained by current theory," says Wang, "so a new theory was needed."

When the core collapses, a neutron star forms before any explosion can occur. Up to now, the theory of core-collapse supernovae has been focused on the production of neutrinos that are generated within the newly formed neutron star.

These ephemeral particles carry off more than a hundred times the energy required to trigger the explosion of the star. The question has been whether they carry too much and spoil the explosion, or leave enough energy behind to cause the explosion.

To help with a new theory that explains supernova formation and takes polarization into account, Wang and his Texas colleagues turned to Khokhlov, Oran, and Chtchelkanova of NRL, who used computer modeling to test scenarios that could explain the newfound polarization of these supernovae. Their models tested the idea that collapsing supernovae begin by expelling mass and energy from the new neutron star in a strongly directional process.

"Moving mass and energy in a single direction is the operational definition of a jet," says Wheeler. "These are jet-induced explosions."

If the new jet theory is right, the traditional questions about neutrinos and supernovae may be irrelevant. In their calculations, Khokhlov and his associates found that the jet punches out of the star, but also sends shock waves sideways, sharing some of the energy throughout the star. The result is that the entire star is blown up by the jet and the neutrinos do not need to play any obvious role. The ejected matter is sent out in the jet and in a pancake containing other star material.

"The result is just what we need to explain the polarization," says Peter Hoeflich, a Research Scientist at the University of Texas at Austin, who is an expert on the flow of radiation from supernovae.

The numerical techniques to compute the effect of a jet on a star were developed by Khokhlov when he was at the University of Texas at Austin and have been refined and applied to this problem at the Naval Research Laboratory, where he is currently a Research Scientist.

The computer code developed by Khokhlov is fully three dimensional and has an "adaptive-mesh" capability, so that it automatically computes most carefully just where the need is greatest. This code was used by Khokhlov and his colleagues to compute the propagation of a jet from near the surface of a newly formed neutron star to its eruption into space.

"The next task is to better understand the origin of the jet," says Wheeler. "The most plausible cause is the rapid rotation of the neutron star and its strong magnetic field. We have begun to look into how the newly formed neutron star can channel its energy up the rotation axis by magnetic jets or intense pulsar radiation."