With astonishing speed and candor, NASA has presented a compelling picture of how the space shuttle Columbia was destroyed.
In an open process that has invited Americans to play the part of forensic scientist, NASA officials have laid out a host of data that suggest a catastrophic failure of the system designed to protect the spacecraft's left wing from the searing heat of reentry. What's more, they have thrown into doubt their own estimates of the importance of an errant piece of foam.
Now, as they warn against a rush to judgment, they're fighting the momentum of a public that has already pronounced its verdict.
It may well turn out that a fragment of tile or a computer dispatch confirms that the foam knocked off crucial heat-resistant tiles shortly after launch. But this is an investigation of unusual complexity.
The piecing together of clues from one of the most intricate machines humanity has yet devised, which broke apart at 200,000 feet and has perhaps scattered its many fragments from Arizona to Louisiana, defies immediate and conclusive answer. And while NASA's openness about every shred of data - in contrast to its Challenger investigation - has encouraged just that, experience suggests that there could be some unexpected twists before the end.
"It's like taking a fistful of jigsaw puzzle pieces and trying to put together the whole puzzle," says James Oberg, a former member of the space shuttle Mission-Control team. "The way [the foam] may or may not have an impact is baffling."
The evidence is enough to convince many scientists of the cause of the crash: a failure of the left wing's thermal system. But to assume the failure was caused by the foam, without any direct evidence, flies in the face of science.
Eugene Covert and Robert Rummel won't even discuss it. Both members of the commission that investigated the Challenger disaster say that there is not enough data to know what happened. "The best thing is to be open-minded," says Dr. Covert of the Massachusetts Institute of Technology in Cambridge. "As human beings, we make up hypotheses and subconsciously try to reinforce them. You have to resist that temptation."
Theories for the explosion of TWA Flight 800 included mistaken military missile firings before officials laid the blame on a fuel tank. Investigators of Pan Am Flight 103 had to find a piece of copper wire embedded in a seat before they could link the crash to a bomb.
Paul Czysz, professor emeritus of aerospace engineering at St. Louis University, thinks there's already one explanation that's more plausible than the foam. "I think it was water ice," he says. "If it were foam hitting the wing, it would have just glanced off, but when you see this shower of particles, that's ice."
The Challenger disaster offers a curious parallel. The first theories that emerged after Challenger exploded in a fork of smoke and fire over the Atlantic Ocean focused on the massive orange fuel tank. Perhaps it leaked, some said, or perhaps some of the fuel lines were cut.
On the fifth day, though, unnamed sources behind NASA's veil of secrecy began to speak of footage that showed flames leaping from the right booster rocket. Ten days later, Congress held hearings on the O-rings that were eventually pinpointed as the cause of the flames. Within a month, the presidential commission dissected the entire disaster down to the tenth of a second.
This time, of course, things might be tougher. Investigators will lack detailed photographs of the shuttle's disintegration. For that reason, NASA's massive manhunt for the first pieces to fall off Columbia - which could be as small as a 6-by 6-inch silicon tile, falling as far west as Arizona - has taken on an air of urgency.
For scientists seeking answers with absolute certainty, such evidence could be a "missing link" that reveals precisely how the heat-shield breach occurred.
For many others, though, the pictures from liftoff are convincing in themselves. Even some scientists suggest that the images of the foam exploding against the underside of Columbia's left wing, combined with data that show the craft's problems began in the left wing, is likely more than circumstantial evidence.
"It does seem that [thermal damage] is the most likely culprit given the evidence we've seen so far," says John Olds, an aerospace engineer at Georgia Tech in Atlanta. "That's what they should be focusing on first."