A constant rain of comets pelting the Earth is probably just noise common to many scientific instruments, according to a scientist at NASA's Marshall Space Flight Center. In a paper to be presented today at the fall meeting of the American Geophysical Union, Dr. James Spann says that images of comet impacts are random events common to instruments using image intensifiers like night-vision goggles.
Earlier this year, Dr. Louis Frank of the University of Iowa announced that he had discovered evidence of small comets plowing into the Earth's upper atmosphere. He first made this claim in 1986, and has renewed it with data from a new satellite and a new instrument.
The debate centers on images from two instruments - the Ultraviolet Imager (UVI) and the Visible Imaging System (VIS) - aboard the Polar spacecraft launched in 1996. Polar carries several instruments designed to study activities in the magnetosphere, the region where Earth meets space.
UVI and VIS take pictures of the aurora borealis, the northern lights. The aurora is like a TV picture tube where nature paints images in the upper layers of the atmosphere that depict activities back in the tail of the magnetosphere.
Spann is a UVI co-investigator; Dr. George Parks of the University of Washington in Seattle, is the UVI principal investigator. Other key team members are Mitchell Brittnacher of the University of Washington and Glenn Germany of the University of Alabama in Huntsville.
Frank is principal investigator for the VIS. He first made this claim in 1985 based on images taken by a camera aboard the Dynamics Explorer satellite. His claim was not widely accepted.
The claim was renewed this year when Frank said that images from VIS also showed dark smears which would indicate comets striking the atmosphere. Frank says that 5 to 30 a minute strike Earth, and over the eons have made a significant contribution to the world oceans.
Specifically, Frank bases his claims on images taken through Polar's VIS's Earth camera, one of three cameras making up the instrument. The Earth camera is sensitive to light in the 124-149 nm range (ultraviolet, above the human eye's blue-violet limit at about 300 nm) with 130.4 nm being the most sensitive line.
Frank found dark spots in the Earth camera images that he believes show hydroxyl being neutralized by water when cometesimals - house-size chunks of water ice - boil away as they hit the atmosphere. Hydroxyl is usually found where one atom of hydrogen has been stripped from a water molecule, leaving just an oxygen-hydrogen (OH) radical. The upper atmosphere glows in the light of hydroxyls.
But the nature of science is to challenge claims and discoveries. A challenge may provide independent verification, and move a claim from hypothesis to fact or scientific law. Or, it may refute the claim and redirect the investigation. And the more significant the claim, the more significant the challenge.
Thus, Spann, Parks, and their team members looked for signs in UVI data that comets were arriving. What they found, Spann says, was the same noise they saw on the ground.
"We believe it's native to the instrument," Spann said. "The VIS and UVI detectors are similar. And in calibrating the UVI on the ground, we saw dark pixels. We don't fully understand the mechanisms of dark pixels, but we're going to continue studying it with flight data and in the laboratory."
During ground tests, UVI provided images while in complete dark and with its sensor array fully illuminated. In the latter case, UVI sometimes returned darkened pixels (individual points on the image). VIS, however, was not tested in the same manner.
Spann has not examined Frank's data from the Dynamics Explorer, but expects that the a similar effect may have happened there. (Data from the Christa-Spas instrument carried by the Space Shuttle earlier this year measured higher-than-expected hydroxyl levels in the upper atmosphere, but did not see comet impacts. Thus, its data can be taken as supporting Frank's claims, but are not confirmation.)
Noise, small variations, and random errors are normal in most scientific instruments. That's why most investigations require dozens, perhaps hundreds, of data points to ensure that a discovery is real. The famous Hubble Space Telescope deep field image, for example, shows thousands of galaxies in a piece of "empty" sky. The striking image actually is a refined compilation of dozens of messy images apparently showing nothing of interest.
The challenge with Frank's images is that the impact effects, if real, are transient, lasting just a few minutes. That rules out multiple exposures and high speed to eliminate instrument effects and noise. VIS and UVI were designed primarily for wide views of the aurora so scientists could estimate how much energy is being pumped into the polar regions.
The Polar spacecraft provides another complication. It rotates so its other instruments sweep through space and sample conditions from all directions (the cameras are on a despun platform). As it rotates, it also wobbles, just a little, because a balancing device does not have enough mass to balance it completely.
As a result, images taken by UVI and VIS are slightly smeared; not enough to impair auroral studies, but enough to cause a slight double vision when looking at a point source. This double source indicates that the object is outside the camera. VIS claims that they see these double visions, but UVI analyses show they are random and accidental. No preference is seen in the so-called wobble direction
Frank has also reviewed the UVI data, Spann said, and bases his finds on points where UVI and VIS see dark spots in the same place and time. Spann says that is difficult to support for several reasons.
First, while the two instruments point in the same direction, they are not precisely aligned to ensure that the two sets of images match up. Second, they have different integration times. Every picture - even in high-speed photography - is a time exposure where the film or sensor is exposed so enough light arrives. Both UVI and VIS take several seconds to built their images. But they don't do it at the same time.
Finally, Spann continued, UVI produces enough dark spots that occasional matches between it and VIS are inevitable.
"Only rarely are two dark spots [one from each imager] near each other," Spann said, "and not all have double features" indicating smearing caused by the satellite's wobble.
The UVI team's presentation likely will stir a longer discussion. Again - the nature of science is to question. Spann expects that it will take an independent investigation to decide the issue one way or the other.
Meanwhile, the scientific literature has plenty of materials for consideration. Frank and his colleagues published four articles in the Oct. 1, 1997, issue of Geophysical Research Letters (Vol. 24, No. 19, pp 2423-2438). Parks and his team will publish one article in GRL on Dec. 15. Richard Kerr writes "Tiny Comets' Spots Called Artifacts" in Science magazine, Nov. 14 issue (pp 1217-1218)