One week after the space shuttle Columbia ripped apart above Texas, the clues that could solve the mystery of the disaster remain shrouded in scrambled computer code, shredded wreckage and smeary photographs.

Hundreds of engineers are scrutinizing screens of data, programming virtual collisions, designing high-impact tests and enhancing digital images. Search teams are picking through forests and suburbs for charred fragments, cracked ceramic tiles and scattered engine parts. Like fortunetellers poring over tea leaves, analysts are hunting for patterns in puzzling sensor readings, temperature spikes, wing flap movements and subtle shifts in aerodynamic forces.

Did a chunk of insulation inflict lethal damage when it smacked into Columbia's protective shell during liftoff? Did a shred of space junk bore a hole in a critical spot that ripped open during the fiery descent? Did a computerized guidance system malfunction, spiraling the spaceship into a suicidal roll?

It could be weeks, or months, before an answer becomes clear. Or the answer may remain forever elusive.

"It seems like we know less and less as time goes by. On Saturday morning I did a simple Web search and told my wife, 'I know what happened.' Now we have no idea," said Lauren Thompson, an aerospace analyst with the Lexington Institute in Arlington. "When the obvious solution looks less and less obvious, it raises the possibility we may never know what happened."

As shocked, exhausted and grief-stricken NASA officials struggled to uncover the cause, questions intensified about the expertise and independence of the panel investigating Columbia's destruction 200,000 feet above Texas. Although the space agency won praise for openness and candor, in contrast to the response after the Challenger disaster, aerospace experts and others said they are becoming increasingly concerned about how the probe is proceeding.

"NASA does one thing very well: It's a space agency involved in advancing the development of space technology. But it is not in the business of conducting forensic investigations into major air disasters or crashes," said Michael Slack, an aerospace attorney who was a NASA engineer from 1974 to 1980.

That lack of expertise appeared to show, said Slack and others, when officials vacillated about whether the disaster could have been set in motion by a 26-by-12-by-7-inch piece of foam insulation. The chunk of foam ripped off the shuttle's external tank at launch, glancing off Columbia's left wing.

"It hurts them a little bit when they start by saying the insulation wasn't a problem, then it was a problem, and then it's not a problem again," said Paul S. Fischbeck, a professor of engineering and public policy at Carnegie Mellon University in Pittsburgh who has studied the space shuttle for NASA. "I think they've cycled two or three times on that."

An agencysuch as the National Transportation Safety Board, which has expertise gained by scrutinizing hundreds of plane crashes, is better suited for this kind of inquiry, Slack and others said.

"I think the investigation should have immediately been turned over to someone who can do the job and doesn't have a vested interest," said John Macidull, who worked for the NTSB, the Federal Aviation Administration and the presidential panel that unraveled the 1986 Challenger disaster.

Some argue that the Columbia probe demands a similar presidentially appointed commission.

"Public perception is as important as everything else. I frankly think it should be more independent from NASA," said Allan J. McDonald, a Morton Thiokol Corp. engineer who raised questions about the safety of launching Challenger.

Others fear the inquiry is too narrowly focused on finding the specific flaw that doomed Columbia, avoiding a more important, wide-ranging exploration of how the space agency makes decisions.

"What we know from the past is that there's often a connection between the technical failure and the organization itself," said Diane Vaughan, a sociology professor at Boston College who wrote a book about the Challenger investigation. "So far, no one has begun to look at how the organization may have been involved."

Many experts defended NASA and the investigation, saying it is proceeding carefully and efficiently, has all the independence and expertise it needs, and that it is far too early to produce results.

"I think that the knowledge of the very complex shuttle system lies within NASA and those that work for NASA," said John M. Logsdon, director of the Space Policy Institute at George Washington University. "So bringing in another group that would have to learn all that for what essentially would be political reasons would introduce unnecessary delay."

The space agency launched its investigation hours after Columbia's obliteration, which occurred 16 minutes before the shuttle was due to touch down in Florida to end what until then had been a highly successful 16-day scientific mission.

"These folks who are doing the work are carrying a huge burden with them. They haven't had time to grieve. It's their friends and their program, and it's a terrible tragedy. They haven't had a chance to look up," said George "Pinky" Nelson, a former astronaut.

Investigators impounded all information about the mission at NASA centers and contractors around the country to begin painstakingly building a "fault tree," in which all possible failures are methodically eliminated until, hopefully, the culprit is revealed.

NASA formed teams to pick apart every scenario and summoned former officials for advice.

"They're taking a lot of us gray bears, as they call us, and asking us to come and review what they're doing," said Donald Emero, a retired shuttle manager called to Houston. "They're evaluating the data looking for any little twitch that is out of the ordinary and performing 'what if' kinds of analysis."

Last Sunday, officials revealed they were focusing on a curious succession of sensor readings scattered around the shuttle's left wing. They showed temperature rises and odd sensor outages. The data also showed unusual drag on the left side, and the computerized navigation system compensating by tugging in the opposite direction eight minutes before Columbia lost contact.

But none of the recorded temperature spikes was enough to destroy Columbia. By week's end, shuttle manager Ronald D. Dittemore speculated that the seemingly random pattern of sensor readings may have followed a common wiring trail that could lead back to the focal point of the disaster.

The explanation that seemed most obvious was that two parts of the shuttle that have long been troublesome combined to set up NASA's oldest shuttle for destruction. The 2.67-pound-foot chunk of foam insulation from the shuttle's massive external tank, which has frequently shed pieces during liftoff, flew into the fragile ceramic tiles that keep the ship from burning up during reentry, or perhaps hit the somewhat tougher carbon coating on the wing's edge. That could have left Columbia's aluminum skin vulnerable to the 3,000 degree cloud of gas that envelops shuttles during reentry. Dittemore seemed to go back and forth on that possibility all week.

"I probably wouldn't have thrown away the foam argument so quickly," said Paul Czysz, a professor emeritus of aerospace and mechanical engineering at Saint Louis University and a veteran NASA consultant. Czysz calculated that the foam might have hit the shuttle with the force of seven shots from a .30-caliber rifle. "That's a lot of energy," Czysz said.

As the week wore on, details began to dribble out about NASA's deliberations while the shuttle was aloft. The decision that the foam impact posed no threat seemed to come after relatively little debate, even though great uncertainty remained about the seriousness of the damage.

By week's end, initial hopes that key clues could be extracted from 32 seconds of garbled data from Columbia in its final moments seemed to fade, and the debris that showered down on Texas and Louisiana, and possibly as far west as California, seemed to offer the best hope for answers.

"You've got to find the first parts that came off," Czysz said. "After that, it's a chain reaction. If you find a whole bunch of chain links busted, you've got to go find the first one that broke."

More than 12,000 pieces of Columbia's remains, from slices of tire, a helmet and an astronaut patch to large sections of tiles, are being collected. Experts will try to reassemble as much as possible at the Kennedy Space Center. The initial focus was on Texas, where the most wreckage fell, hitting school roofs, back yards and lakes.

"It seems NASA did not come to the realization that some of the most critical evidence would be further up range than Dallas until Monday or so. That was one of the first clues that I had that NASA was not getting the desired input on how to conduct a forensic investigation into a tragedy like this," said Slack, the engineer-turned-lawyer.

By Wednesday, NASA expanded the search westward, dispatching teams to California and Arizona.

A break may have come Friday when searchers found a two-foot-long section of wing near Fort Worth. If the wreckage is from the left wing, the find could provide pivotal evidence. "You have a million pieces scattered from California to Louisiana and maybe beyond. They need to be looked at by eyes that are used to looking at material that has been tormented by high loads. NASA doesn't really have the eyes trained to look at something like that," Slack said.

Experts are also reviewing any photographs and video they can find. The launch pad camera that had the best view of where the insulation struck was out of focus. But the Air Force gave NASA a ghostly silhouette captured by a tracking camera at Kirtland Air Force Base in New Mexico, seconds before Columbia lost contact with Mission Control.

It appeared to show a jagged spot on the crucial front edge of the left wing, and perhaps debris trailing from the left rear of the craft as it plummeted at 12,500 mph. Or maybe the picture showed nothing. A weary Dittemore said it was too soon to know for sure.

On Thursday, 1,700 people -- members of Congress, the Cabinet, astronauts, relatives and others -- gathered at the Washington National Cathedral to honor the seven lost astronauts. The next day, NASA turned over the investigation to the panel led by Adm. Harold W. Gehman Jr. Responding to complaints from Congress, NASA Administrator Sean O'Keefe announced he was changing the panel's charter to give it more independence and adding more outside members.

There is the possibility that Columbia's demise may never be solved, which would put NASA in the difficult position of having to decide whether it is safe to continue flying the shuttle fleet.

"It is conceivable that with no close-up photographs, with no eyewitnesses and incomplete data, we'll never fully understand what caused the sequence of events that led to the disaster. The cause that is never found is not unheard of in the aerospace industry," Thompson said. "It's still not clear what caused the Boeing 737s in Colorado Springs and Pittsburgh to crash. And those planes were not 39 miles from the ground."