WASHINGTON - It's too soon to say why the space shuttle Columbia exploded, but the early signs suggest that a heat problem enveloped the left wing and side of the spacecraft before it disintegrated more than 39 miles above north central Texas.
What caused the heat and subsequent explosion is still unknown, but a prime suspect is a slight mishap that occurred when the vehicle was launched on Jan. 16. At that time, some debris from the shuttle's external fuel tank slammed against the left wing, experts said.
No matter what the cause, safety experts have warned for years that a problem was coming. They told Congress that the shuttle program needed more money and newer equipment or else it faced rising safety risks. Six outside consultants on a safety panel issued such warnings; they were fired in March 2001.
NASA officials began sketching out the fragmentary details of what they knew about the crash at a press conference Saturday afternoon.
In the last seven minutes before NASA lost communication with the 21-year-old ship, Columbia seemed to be flying normally. There was no sign of loss of flight control before it exploded at Mach 18.3, about 13,000 mph, said shuttle program manager Ron Dittemore. The shuttle had performed "absolutely flawlessly" during its 16-day mission, he said.
Then temperature sensors and tire pressure sensors on the left side were lost "as if someone had cut the wire," said Dittemore.
NASA lost contact with the shuttle just when Columbia was enduring extreme heat from scorching through the Earth's atmosphere. Its intricate tile system is supposed to shield the vehicle from harm by heat.
"We were at our peak heating," Dittemore said. "We were at about 3,000 degrees Fahrenheit at wing edge. If we did have a thermal problem or expected problem, you would expect to get it at peak heating . ... And that's where we lost the vehicle."
The first indication of a problem came at 8:53 a.m. EST, when temperature sensors on left-wing hydraulic systems stopped sending data.
"We just lost it," said chief flight director Milt Heflin. Three minutes later, the wheel well and brake-line temperature indicators also failed. And at 8:59, eight temperature gauges went off line.
"We think they were acknowledging that measurement (of things going wrong) that they saw," Heflin said.
For the seven astronauts aboard Columbia, the explosion occurred when they were "still in a very nose-up attitude, so you're kind of coming in bottom-first; you're not flying like an airplane yet," said George "Pinky" Nelson, a retired three-time shuttle astronaut.
"Re-entry is the one thing the movies always get wrong. It's very smooth and quiet and gentle. The (g-forces) are just starting to build up, so they would have been still basically weightless. They're just watching the instruments."
That's when a tire pressure alarm appeared to go off, according to communication from Mission Control at the Johnson Space Center in Houston. For the astronauts, "there's nothing you can do at that point," Nelson said.
That tire alarm, some experts said, may have been the first sign that something had gone wrong with the delicate heat protection envelope that keeps astronauts alive as they plummet through the atmosphere. The heat may have caused the tire to rapidly inflate, setting off the alarm.
In trying to figure out what caused the ship to disintegrate 207,135 feet above north central Texas, aerospace experts instantly highlighted two broad possible causes:
_The most likely explanation is that something damaged the shuttle's delicate insulation - ceramic-like tiles and fluffy blankets - that then led to the fatal problem. A hole in the insulation could allow the heat of re-entry to burn a hole in the shuttle and explode the craft's on-board fuel cells or the highly explosive hydrazine that powers its auxiliary power units and the thrusters that maneuver it. A hole also could allow heat that would then trigger structural damage, such as the breaking off of a wing.
_ A defect also could have caused an explosion on board. Suspicion centers on the hydrazine tanks.
NASA Administrator Sean O'Keefe said two panels would investigate - one internal to NASA, the other consisting of outsiders.
"Our best experts are trying to understand what went wrong," Dittemore said. "We cannot say what caused the loss of Columbia. It's still very early in our investigation."
Federal officials were quick to discount terrorism as a cause.
When Columbia blasted off from Cape Canaveral, Fla., television cameras recorded bits of insulation from its massive, disposable external tank hitting a large section of the left wing, but NASA had discounted it as a problem during the shuttle's 16-day flight.
But those bits of insulation could have punctured the delicate, black, ceramic-like tiles that protect the shuttle.
"If I was guessing I would say it's a tile problem," said Norm Carlson, former shuttle operations chief at the Kennedy Space Center at Cape Canaveral. "They had some tile that was missing perhaps, and it burned through and caused an explosion of a fuel cell tanks or hydrazine tanks."
If debris hit the right place, one small hole could be fatal, Carlson said. "If it is on the wing, it could burn through and cause structural damage. A wing could come off and whatever, break up. What happens, it literally melts the aluminum."
"Not that the stuff is heavy and sharp, but once you get up to a certain speed, a piece of Styrofoam traveling at thousands of miles per hour can do a lot of damage," Carlson said.
NASA's Dittemore said: "We can't discount that there might be a connection. But we can't rush to judgment on this. There are a lot of things in the business that look like the smoking gun, but are not even close."
If heat penetrated the shuttle's protective tile system or if anything forced the shuttle to change its precise position coming through the atmosphere, the results would be catastrophic, said Jose Garcia, a retired senior shuttle engineer at Kennedy Space Center.
"It's gliding or kind of cutting through the atmosphere, and if it ever goes non-aerodynamic" (such as a wing bending or breaking off), "it'll just break apart like it hits something hard," Garcia said.
After Columbia's launch, NASA officials pored over videotape "and it was judged that the event did not represent a safety concern,'" Dittemore said. A similar problem had happened two flights earlier, but was not a safety issue.
Even if they had thought it was a problem for Columbia, "we do not have the capability to perform a space walk and do tile repair," Dittemore said.
But Garcia, who has been a NASA safety whistleblower in the past, said the space agency could have figured out a way to fix the tile.
Seymour Himmel, a former top NASA rocket engineer and a long-time former member of the quasi-independent Aerospace Safety Advisory Panel, offered another possible cause.
"I lean more and more to a failure that was structural in origin," he said.
For years, a special safety panel set up after the fatal 1967 Apollo 1 fire, which killed three astronauts, had harped on the need for safety upgrades and the lack of money to perform them.
In April 2002, Richard Blomberg, the safety panel's exiting chairman, told Congress: "In all of the years of my involvement, I have never been as concerned for space shuttle safety as I am right now. That concern is not for the present flight or the next or perhaps the one after that."
Since that time, three shuttles flew safely. Columbia was the fourth.
"One of the roots of my concern is that nobody will know for sure when the safety margin has been eroded too far," Blomberg told Congress "All of my instincts, however, suggest that the current approach is planting the seeds for future danger."
Last year, the Aerospace Safety Advisory Panel's first finding said:
"The current and proposed budgets are not sufficient to improve or even maintain the safety risk level of operating the Space Shuttle and ISS (International Space Station).
The panel said: "When risk reduction efforts . . . are deferred, astronauts are exposed to higher levels of flight risk for more years than necessary."
Testimony at that hearing highlighted $229 million in budget cuts to shuttle safety upgrades in the last three years' budgets. The annual shuttle budget is about $3.2 billion.
Six of those panel members were dismissed in March 2001 after making such complaints for years.
Garcia, who retired in late 1999, said budget problems have been serious and threatening safety for at least a decade.
"I spent my whole life at Kennedy minimizing risk; (since 1992) they've taken actions that have increased risk, not minimized it," Garcia said. "You ought to be doing everything you can to make that risky situation as safe as possible and they weren't doing that."
For investigators, two key areas will provide clues: Flight data that been collected in Houston, and recovered pieces of Columbia.
Getting clues from the leftover pieces of Columbia could be very tough because the pieces will have burned severely on the way down. It will be much harder than after the 1986 Challenger accident because that one took place only 10 miles up and the debris landed in the ocean, which helped preserve it.