High-resolution images taken from a ground-based Air Force tracking camera in the southwestern U.S. show serious structural damage to the inboard leading edge of Columbia's left wing, as the crippled orbiter flew overhead about 60 sec. before the vehicle broke up over Texas killing the seven astronauts on board Feb. 1.

According to sources close to the investigation, the images, under analysis at the Johnson Space Center in Houston, show a jagged edge on the left inboard wing structure near where the wing begins to intersect the fuselage. They also show the orbiter's right aft yaw thrusters firing, trying to correct the vehicle's attitude that was being adversely affected by the left wing damage. Columbia's fuselage and right wing appear normal. Unlike the damaged and jagged left wing section, the right wing appears smooth along its entire length. The imagery is consistent with telemetry.

The ragged edge on the left leading edge, indicates that either a small structural breach--such as a crack--occurred, allowing the 2,500F reentry heating to erode additional structure there, or that a small portion of the leading edge fell off at that location.

Either way, the damage affected the vehicle's flying qualities as well as allowed hot gases to flow into critical wing structure--a fatal combination.

It is possible, but yet not confirmed, that the impact of foam debris from the shuttle's external tank during launch could have played a role in damage to the wing leading edge, where the deformity appears in USAF imagery.

If that is confirmed by the independent investigation team, it would mean that, contrary to initial shuttle program analysis, the tank debris event at launch played a key role in the root cause of the accident.

Another key factor is that the leading edge of the shuttle wing, where the jagged shape was photographed, transitions from black thermal protection tiles to a much different mechanical system made of reinforced carbon-carbon material that is bolted on, rather than glued on as the tiles are.

This means that in addition to the possible failure of black tile at the point where the wing joins the fuselage, a failure involving the attachment mechanisms for the leading edge sections could also be a factor, either related or not to the debris impact. The actual front structure of a shuttle wing is flat. To provide aerodynamic shape and heat protection, each wing is fitted with 22 U-shaped reinforced carbon-carbon (RCC) leading-edge structures. The carbon material in the leading edge, as well as the orbiter nose cap, is designed to protect the shuttle from temperatures above 2,300F during reentry. Any breach of this leading-edge material would have catastrophic consequences.

The U-shaped RCC sections are attached to the wing "with a series of floating joints to reduce loading on the panels due to wing deflections," according to Boeing data on the attachment mechanism.

"The [critical heat protection] seal between each wing leading-edge panel is referred to as a 'tee' seal," according to Boeing, and is also made of a carbon material.

The tee seals allow lateral motion and thermal expansion differences between the carbon sections and sections of the orbiter wing that remain much cooler during reentry.

In addition to debris impact issues, investigators will likely examine whether any structural bending between the cooler wing structure and the more-than-2,000F leading edge sections could have played a role in the accident. There is insulation packed between the cooler wing structure and the bowl-shaped cavity formed by the carbon leading-edge sections.

The RCC leading-edge structures are bolted to the wing using Inconel fittings that attach to aluminum flanges on the front of the wing.

The initial NASA Mission Management Team (MMT) assessment of the debris impact--made on Jan. 18, two days after launch--noted "The strike appears to have occurred on or relatively close to the "wing glove" near the orbiter fuselage.

The term "wing glove" generally refers to the area where the RCC bolt-on material is closest to the fuselage. This is also the general area where USAF imagery shows structural damage.

The second MMT summary analyzing the debris hit was made on Jan. 20 and had no mention of the leading-edge wing glove area. That report was more focused on orbiter black tiles on the vehicle's belly. The third and final summary issued on Jan. 27 discusses the black tiles again, but also specifically says "Damage to the RCC [wing leading edge] should be limited to [its] coating only and have no mission impact." Investigators in Houston are trying to match the location of the debris impact with the jagged edge shown in the Air Force imagery.

Columbia reentry accident investigators are also trying to determine if, as in the case of Challenger's accident 17 years ago, an undesirable materials characteristic noted on previous flights--in this case the STS-112 separation of external tank insulation foam debris--was misjudged by engineers as to its potential for harm, possibly by using analytical tools and information inadequate to truly identify and quantify the threat to the shuttle. As of late last week, NASA strongly asserted this was not the case, but intense analysis on that possibility continues.

The shuttle is now grounded indefinitely and the impact on major crew resupply and assembly flights to the International Space Station remains under intense review.

Killed in the accident were STS-107 mission commander USAF Col. Rick Husband; copilot Navy Cdr. William McCool; flight engineer Kalpana Chawla; payload commander USAF Lt. Col. Michael Anderson; mission specialist physician astronauts Navy Capt. Laurel Clark and Navy Capt. David Brown; and Israel Air Force Col. Ilan Ramon.

"We continue to recover crew remains and we are handling that process with the utmost care, the utmost respect and dignity," said Ronald Dittemore, shuttle program manager.

No matter what the investigations show, there are no apparent credible crew survival options for the failure Columbia experienced. With the ISS out of reach in a far different orbit, there were no credible rescue options even if wing damage had been apparent before reentry--which it was not.

If, in the midst of its 16-day flight, wing damage had been found to be dire, the only potential--but still unlikely--option would have been the formulation over several days by Mission Control of a profile that could have, perhaps, reduced heating on the damaged wing at the expense of the other wing for an unguided reentry, with scant hope the vehicle would remain controllable to about 40,000 ft., allowing for crew bailout over an ocean.

Reentry is a starkly unforgiving environment where three out of the four fatal manned space flight accidents during the last 35 years have occurred.

These include the Soyuz 1 reentry accident that killed cosmonaut Vladimir Komarov in 1967 and the 1971 Soyuz 11 reentry accident that killed three cosmonauts returning after the first long-duration stay on the Salyut 1 space station.

The only fatal launch accident has been Challenger in 1986, although Apollo astronauts Gus Grissom, Ed White and Roger Chaffee were killed when fire developed in their spacecraft during a launch pad test not involving launch.

No other accident in aviation history has been seen by so many eyewitnesses than the loss of Columbia--visible in five states.

Telemetry and photographic analysis indicate the breakup of the historic orbiter took place as she slowed from Mach 20-to-18 across California, Nevada, Arizona and New Mexico with the loss of structural integrity 205,000 ft. over north-central Texas where most of the debris fell.

The science-driven STS-107 crew was completing 16 days of complex work in their Spacehab Research Double module and were 16 min. from landing at Kennedy when lost. Landing was scheduled for 8:16 a.m. CST.

Abnormal telemetry events in the reentry began at 7:52 a.m. CST as the vehicle was crossing the coast north of San Francisco at 43 mi. alt., about Mach 20.

The orbiter at this time was in a 43-deg. right bank completing its initial bank maneuver to the south for initial energy dissipation and ranging toward the Kennedy runway still nearly 3,000 mi. away.

That initial bank had been as steep as about 80 deg. between Hawaii and the California coast, a normal flight path angle for the early part of the reentry. The abnormal events seen on orbiter telemetry in Houston indicate a slow penetration of reentry heat into the orbiter and damage on the wing, overpowering the flight control system. Key events were:

* 7:52 a.m. CST: Three left main landing gear brakeline temperatures show an unusual rise. "This was the first occurrence of a significant thermal event in the left wheel well," Dittemore said. Engineers do not believe the left wheel well was breached, but rather that hot gases were somehow finding a flow path within the wing to reach the wheel well.

* 7:53 a.m. CST: A fourth left brakeline strut temperature measurement rose significantly--about 30-40 deg. in 5 min.

* 7:54 a.m. CST: With the orbiter over eastern California and western Nevada, the mid-fuselage mold line where the left wing meets the fuselage showed an unusual temperature rise. The 60F rise over 5 min. was not dramatic, but showed that something was heating the wing fuselage interface area at this time. Wing leading edge and belly temperatures were over 2,000F. While the outside fuselage wall was heating, the inside wall remained cool as normal.

* 7:55 a.m. CST: A fifth left main gear temperature sensor showed an unusual rise.

* 7:57 a.m. CST: As Columbia was passing over Arizona and New Mexico, the orbiter's upper and lower left wing temperature sensors failed, probably indicating their lines had been cut. The orbiter was also rolling back to the left into about a 75-deg. left bank angle, again to dissipate energy and for navigation and guidance toward Runway 33 at Kennedy, then about 1,800 mi. away.

* 7:58 a.m. CST: Still over New Mexico, the elevons began to move to adjust orbiter roll axis trim, indicating an increase in drag on the left side of the vehicle. That could be indicative of "rough tile or missing tile but we are not sure," Dittemore said. At the same time, the elevons were reacting to increased drag on the left side of the vehicle, the left main landing gear tire pressures and wheel temperature measurements failed. This was indicative of a loss of the sensor, not the explosion or failure of the left main gear tires, Dittemore believes. The sensors were lost in a staggered fashion.

* 7:59 a.m. CST: Additional elevon motion is commanded by the flight control system to counteract left side drag. The drag was trying to roll the vehicle to the left, while the flight control system was commanding the elevons to roll it back to the right.

But the rate of left roll was beginning to overpower the elevons, so the control system fired two 870-lb.-thrust right yaw thrusters to help maintain the proper flight path angle. The firing lasted 1.5 sec. and, along with the tire pressure data and elevon data, would have been noted by the pilots.

At about this time, the pilots made a short transmission that was clipped and essentially unintelligible.

In Mission Control, astronaut Marine Lt. Col. Charles Hobaugh, the spacecraft communicator on reentry flight director Leroy Cain's team, radioed "Columbia, we see your tire pressure [telemetry] messages and we did not copy your last transmission."

One of the pilots then radioed "Roger," but appeared to be cut off in mid-transmission by static. For a moment there were additional static and sounds similar to an open microphone on Columbia but no transmissions from the crew.

All data from the orbiter then stopped and the position plot display in Mission Control froze over Texas, although an additional 30 sec. of poor data may have been captured.

Controllers in Mission Control thought they were experiencing an unusual but non-critical data drop-out. But they had also taken notice of the unusual buildup of sensor telemetry in the preceding few minutes.

About 3 min. after all data flow stopped, Hobaugh in Mission Control began transmitting in the blind to Columbia on the UHF backup radio system. "Columbia, Houston, UHF comm. check" he repeated every 15-30 sec., but to no avail. In central Texas, thousands of people at that moment were observing the orbiter break up at Mach 18.3 and 207,000 ft.

Milt Heflin, chief of the flight director's office, said he looked at the frozen data plots. "I and others stared at that for a long time because the tracking ended over Texas. It just stopped. It was then that I reflected back on what I saw [in Mission Control] with Challenger."

The loss of Challenger occurred 17 years and four days before the loss of Columbia.

"Our landscape has changed," Heflin said. "The space flight business today is going to be much different than yesterday.

"It was different after the Apollo fire, it was different after Challenger."

Columbia, the first winged reusable manned spacecraft first launched in April 1981, was lost on her 28th mission on the 113th shuttle flight.