Comet impact poses intriguing riddles.

Comet Shoemaker-Levy 9 is gone, its fragments exploded last week in Jupiter's atmosphere. But they leave behind more than a trail of dusty debris and giant smudges marring the solar system's largest planet. Astronomers must now contend with the task of wresting meaning from the barrage of data from this cosmic-collision.

As it turned out, the 6-day-long event contained enough surprises to delight -- and perplex -- just about everyone. Some astronomers even doubt that Shoemaker-Levy 9 was a comet.

"It's astonishing to me that these faint little fragments we've tracked for 14 months have produced these big scars on Jupiter," says Paul Chodas of NASA's Jet Propulsion Laboratory in Pasadena, Calif.

Scientists had expected to find at least traces of water and an abundance of compounds rich in oxygen and carbon in Jupiter's upper atmosphere. They reasoned that the explosion of a dusty ice ball -- the usual model for a comet -- should readily release such material.

Moreover, if the comet chunks plowed deeply enough, they might also excavate water f rom a layer of water clouds thought to reside lower in the Jovian atmosphere. But telescopes found no direct evidence of water or a definitive enhancement of oxygen- or carbon-bearing compounds. Astronomers have now retracted a preliminary report that spectra taken with the U.K. Infrared Telescope after one of the first impacts reveal methylene and the hydroxyl ion in the Jovian atmosphere (SN: 7/23/94, p.55).

Instead, ultraviolet spectra from the Hubble Space Telescope reveal strong emissions from ammonia, sulfur, and hydrogen sulfide. Observers had seen ammonia before, but no one had ever detected the other two gases on Jupiter.

Surprisingly, the compounds thought to be abundant in comets were missing. This led some researchers to propose that Shoemaker-Levy 9 might actually be a rocky body -- either an asteroid or a burned--out comet completely stripped of its ices -- rather than an active comet. Such distinctions are far from academic: The different bodies would profoundly affect the types of chemical reactions triggered by the impacts.

Scientists may solve these and other riddles as they address several key questions: How big were the Shoemaker-Levy 9 fragments? How deep did they penetrate into Jupiter? What were their chemical compositions?

Several lines of evidence suggest that the exploding fragments, despite the towering plumes and dark clouds they created, didn't burrow deeply into the Jovian atmosphere. In particular, the chemical fingerprints -- spectra taken by the Hubble Space Telescope and other instruments -- show no signature of material exhumed from regions well below Jupiter's visible cloud tops.

Researchers believe that the upper reaches of the Jovian atmosphere consist of three distinct cloud layers. The highest are the visible clouds of ammonia. Beneath them, according to models, lies a layer of ammonium hydrosulfide. The final layer is thought to be water.

The abundance of sulfur and ammonia and the lack of water may indicate that the fragments of Shoemaker-Levy reached no farther than the ammonium hydrosulfide layer, notes Keith S. Noll of the Space Telescope Science Institute (STSI) in Baltimore.

An entirely different phenomenon -- observed within the dark cloud created by fragment G, the largest of the 20-odd chunks -- also supports the idea of a shallow strike into Jupiter.

When Andrew P. Ingersoll of the California Institute of Technology in Pasadena studied Hubble images of this cloud, which resembles a black eye, he found evidence of a sound wave. Hubble pictures of the G impact site, taken during a 20-minute period beginning about 90 minutes after the explosion, show a sharply defined dark circle moving outward at a speed of about 800 meters per second.

Several researchers had predicted that the impacts might cause Jupiter to ring like a bell, producing sound waves deep within the planet that would then refract up to the visible surface. "But that's not the effect this comet is producing," says Drake Deming of NASA's Goddard Space Flight Center in Greenbelt, Md. Instead, he and Ingersoll believe, the sound waves come from Jupiter's tropopause, a region just below the stratosphere and above the ammonia clouds.

Because the temperature rises both above and below the tropopause, sound waves remain trapped in this region, expanding horizontally rather than bending upward from a lower depth, Deming adds. The generation of such waves "is consistent" with the notion that the fragments exploded at or just below the visible cloud tops, he says.

Disappointed that the sound wave didn't originate from a deeper, more intriguing part of Jupiter, Deming notes that features of the wave may still offer important clues about the nature of the Jovian troposphere.

Ingersoll also found evidence of another type of wave nearer the center of the G impact site. He has tentatively identified this fainter, slower-moving dark circle as a gravity wave, whose expanding ripples cause material in Jupiter's upper atmosphere to bob up and down. Ingersoll believes that the slower speed of the gravity wave indicates that it comes from the water-cloud layer, which lies deeper in the atmosphere. At the same time, the faintness of the wave indicates that the G fragment triggered the ripple when it exploded higher in the planet's atmosphere, he says.

Both sound waves and gravity waves produce tiny changes in temperature as they travel through the atmosphere. So why didn't infrared telescopes, which can directly detect such changes, find the waves? Ingersoll proposes that the ripples were so close together that even the highest-resolution infrared instruments couldn't distinguish a hotter-than-average ripple from an adjacent, colder one.

He ascribes their detection in visible light to both Hubble's ability to image small features "and luck." The luck, Ingersoll notes, came because the material surrounding the G impact acted as a visual tracer for the waves, condensing as a dark solid around colder ripples and remaining as a relatively transparent gas around adjacent, warmer ones.

Debate continues about whether each fragment consisted of a single solid body or a loosely bound agglomeration of much smaller pieces. The latter model, known as the rubble pile, seemed to fall into disfavor as astronomers witnessed the Jovian fireworks. But Erik Asphaug of Nasgs Ames Research Center in Mountain View, Calif., who helped develop the model, responded to critics in a widely circulated electronic-mail message entitled "Rubble Piles Are Not Wimps."

Kevin Zahnle of NASA Ames claims that the model matches the observations and that even the G fragment may have measured no more than I kilometer across, one-third the size estimated by Hubble scientists. Images of the actual impacts, taken by the Galileo spacecraft and expected to be radioed in 2 weeks, may help settle part of the controversy.