Small comet theory melts under scrutiny.

In science's version of an old-fashioned pillory, seven independent teams assailed the hypothesis that thousands of house-size snowballs plow into Earth's atmosphere each day.

At a meeting of the American Geophysical Union last week, the chairman of a session on this controversial topic issued a warning only half in jest. "I don't want any hitting in the trenches. No fighting, no biting," said Thomas M. Donahue of the University of Michigan in Ann Arbor.

Despite the admonition, the numerous critiques drew flashes of temper unusual at such conferences.

The small comet hypothesis was first proposed in 1986 by space scientist Louis A. Frank of the University of Iowa in Iowa City. Frank, along with Iowa's John B. Sigwarth, raised the idea after finding what he called "atmospheric holes" in images of Earth taken by the Dynamics Explorer I satellite. The holes were dark blemishes in the otherwise bright background of ultraviolet radiation emitted by the upper atmosphere.

To explain the spots, Frank proposed that 25,000 small, fluffy comets of almost pure water were bombarding Earth each day and disintegrating in the atmosphere. These clouds of water vapor show up as dark blotches on the images because they block atmospheric radiation from reaching the satellite, he said.

Although most researchers originally dismissed the idea, Frank captured both positive and negative attention last year when he presented corroborating evidence from new instruments on NASA's Polar satellite (SN: 5/31/97, p. 332; 12/20 & 27/97, p. 389). Ultraviolet images taken by Polar showed the presence of atmospheric holes, but they bypassed some problems in the original data.

At the meeting in Boston last week, Frank offered additional observations to bolster his hypothesis. Analysis of the satellite images suggests that the number of dark spots varies with the time of day and the season, following the same cycle as meteors. Other researchers quickly challenged the new findings.

Frank reported that small comets strike the atmosphere most often in the morning. That timing fails to match their purported path in space, says Alan W. Harris, an astronomer at NASA's Jet Propulsion Laboratory in Pasadena, Calif. "That's just completely backwards."

According to earlier reports by Frank, the speedy comets move in the same direction as Earth and overtake it. If that were true, the comets should plow into Earth's trailing edge and therefore hit in the evening, Harris says.

He and his colleagues hunted for direct evidence of the comets but came up empty-handed. The team used the Spacewatch telescope in Arizona, an instrument built especially to look for objects near Earth. A decade ago, one Spacewatch researcher reported observing bodies fitting Frank's descriptions, but other investigators associated with the project dismissed these observations as noise.

The detector on the present telescope is considerably more sensitive than the older version, so the hypothesized comets should show up much more clearly now, if they exist, says Harris. The telescope has found none thus far, he reports.

The comets have also eluded the Navy's string of radar installations spanning the southern United States. Stephen H. Knowles and his colleagues at the Naval Research Laboratory in Washington, D.C., examined data collected by this radar fence over 6 weeks. The system should have captured the comets at a rate of one every few minutes, but it failed to detect any of these snowballs among the thousands of objects observed, reports Knowles.

While Frank steadfastly supported his hypothesis at the meeting, several teams offered more mundane explanations for the black splotches in the satellite images. Larry J. Paxton of the Johns Hopkins University Applied Physics Laboratory in Laurel, Md., proposed that the spots could come from clouds in the lower atmosphere.

Although the satellites' ultraviolet imagers were designed to capture only ultraviolet radiation, tests on similar systems show that they often absorb some visible light. As a result, they detect some of the sunlight bouncing off clouds, says Paxton. In fact, Frank and his colleagues last year reported seeing outlines of clouds in a few Polar ultraviolet images.

Paxton hypothesized that breaks in the clouds could show up on the satellite images as dark spots. Their numbers vary precisely with the daily and seasonal cycle of cloud cover, he says.

"I think the simplest answer is that, in all likelihood, these [small comets] don't exist. But it's going to take some more tests," Paxton told Science News.

Two teams offered evidence that the dark spots could come from static in the camera systems themselves. Scientists from the University of Washington in Seattle and the University of California, Berkeley argued that Frank's analyses have been misled by electrical noise in the camera, creating the illusion of atmospheric holes where none exist.

Frank angrily rejected the criticisms, calling some of the rhetoric "detestable." In his rebuttals, he ridiculed his opponents, triggering strong responses.

The assault on the small comet hypothesis continued into another meeting session, which explored the atmospheric implications of the idea. According to Frank, if small comets exist, the tremendous amount of water they carry would force scientists to reconsider the origin of water on Earth.

Satellite measurements of the middle atmosphere, however, show it to be extremely dry. If the small comets do exist and are moistening the atmosphere, they would have to be much rarer than Frank has hypothesized, calculate researchers from the University of California, Irvine; Hampton (Va.) University; and the Naval Research Laboratory.

The weight of the current criticism is swaying some researchers who last year voiced cautious support for Frank. "I think Frank's been seriously challenged," says Donahue. "There's a strong likelihood that it is instrumental [noise]. That's the way the wind is blowing."