Supernova's light curve baffles scientists.

More than four years after light from supernova 1987A first reached Earth, radiation remains the key tool for investigating the hidden energy sources powering this exploded star. Soon after the supernova appeared, emissions of ultra-violet, infrared and visible light grew steadily fainter, following a predicted decay curve. But changes in the supernova's "light curve" over the past year now leave astronomers puzzled.

The changes hint at two dramatic possibilities: the abundance of elements in 1987A may differ widely from that in our solar system, or a new energy source -- perhaps a dense, spinning sphere of neutrons known as a pulsar -- lies hidden at the core of the object.

"It's a very exciting time to observe the supernova," says Alistair R. Walker of the Cerro Tololo Inter-American Observatory in La Serena, Chile. "We're getting to the stage where we have no comparable data from other supernovas."

The object's brightness began declining 85 days after astronomers first witnessed its stellar outburst in the Large Magellanic Cloud galaxy. Since then, emissions from 1987A have matched the output expected from the decay of cobalt-56, one of many radioactive elements produced during the explosion of the object. As cobalt-56 decays, it emits energetic photons called gamma rays. Some of the gammas excite atoms in the cloud of debris surrounding 1987A, causing the atoms to emit infrared and visible light observable from Earth. Based on the abundance of cobalt-65 as well as its half-life, scientists believe that until recently it provided the supernova's chief fuel.

But as observations of 1987A hit the three-year mark, little cobalt-56 remained, and the light curve flattened, reports Walker, Nicholas B. Suntzeff and their colleagues in the September ASTRONOMICAL JOURNAL. The flatter curve matches the slower decay of another isotope, cobalt-57, which the supernova produced in smaller amounts, the group notes. Another team, at the European Southern Observatory in La Silla, Chile, reports similar results.

So far so good. But although the shape of the light curve mimics the decay of cobalt-57, the magnitude of the curve -- indicating the amount of light now emitted by 1987A -- exceeds that predicted by theory, both teams say. One way to explain the greater emissions, note Suntzeff and his colleagues, is to assume that the supernova produced a ratio of cobalt-57 to cobalt-56 five times the ratio typical in our solar system. They will report these results in an upcoming ASTROPHYSICAL JOURNAL LETTERS.

The unusual ratio may pose a problem, several astronomers assert, even though the turbulent environment of 1987A -- located 160,000 light-years from Earth -- differs from that of the solar system. While nuclear burning inside stars creates the lighter elements, researchers believe it requires the violence of a supernova explosion to produce the heaviest materials, such as radioactive nickel, which then decays to cobalt, and ultimately to iron. Over time, thousands of supernovas spew out their contents, thus determining the abundance of heavy elements in our galaxy and others. According to this model, the abundance of isotopes created by individual supernovas] should not differ radically from the ratio found near Earth.

Suntzeff's team suggests another explanation for the new findings, one that no longer requires 1987A's ratio to conflict with our solar system's. A constant energy source lurking at the core of the supernova could also account for the larger light output -- perhaps a pulsar, long sought but never observed in 1987A, or a black hole. While such sources generally produce a totally flat light curve rather than the slowly declining one observed, Suntzeff says the latest data indicate 1987A's curve appears to be flattening.

Many astronomers caution that the findings provide only sketchy evidence for a pulsar. And absorption of the supernova's far-infrared emissions by Earth's atmosphere complicates efforts to measure the supernova's total brightness.

A study last month with NASA's Gamma Ray Observatory (GRO) may answer the cobalt ratio question, says Mark Leising of Clemson (S.C.) University. GRO measured the spectra of gamma rays from 1987A, which should allow researchers to calculate the amount of cobalt-57 produced by the supernova. In 1988, the Solar Maximum Mission satellite precisely calculated 1987A's quantity of cobalt-56. Comparing both isotopes will directly determine the relative abundance of cobalt-57. Very early results, Leising notes, suggest that GRO did not find the large increase inferred from the ground-based measurements.