RADAR DATA GIVE CONFUSING CLUES IN FLIGHT 990 CRASH.

Byline: Matthew L. Wald The New York Times

Now that crash investigators have used radar data to lay out a picture of a Boeing 767 on a wild roller coaster ride in the last few minutes of EgyptAir Flight 990, people familiar with the airplane are busy trying to concoct a chain of events that would match the data.

Some speculate that the pilots' ability to work the controls was reduced or eliminated by a hijacker or by an accident in the cockpit. Others think the autopilot might have mistakenly triggered a dive and the pilots tried too abruptly to pull out. Still others theorize that a key part, like the tail elevator, broke. A few think the problem is on the other end, with the radar data itself, which they say cannot be correct.

All say their theories could be just wild conjecture, and the cause could be something no one has mentioned yet. Whatever the cause, the radar data, released on Wednesday by the National Transportation Safety Board, added a piece to a puzzle that will be difficult to solve because most of the pieces are still missing.

Nor did the radar data offer a clue as to what touched off the disaster, which could have been anything from mechanical failure to pilot error to a terrorist attack.

In interviews on Thursday, a variety of experts described a number of plausible explanations of why, in its last few minutes, Flight 990 plunged sharply, then climbed more than a mile before falling again. Natural aerodynamic tendencies designed into the airplane could help explain part of the up-and-down, porpoising course, but possibly not all of it. Parts breaking off in flight could be part of the explanation.

Ocean holds answers

But the experts said they had very little information to fill in the blanks. Some crucial information, as on the condition of the crew and the plane, may be on the data recorders that the Navy is trying to recover from the ocean floor. At times on Thursday, the experts sounded like the proverbial blind man trying to describe an elephant by touching one part of it, but not feeling the whole animal.

``We don't know if the crew (was) still alive,'' one longtime 767 captain said about those puzzling minutes over the Atlantic early Sunday. The captain, who spoke only on the condition that he not be named, was sitting in his office with another jet pilot talking over the large variety of possibilities. ``We don't know who (was) flying the airplane, if anybody,'' he said, during the last moments.

Tim Forte, a former official of the safety board who is now a safety expert at Embry Riddle Aeronautical University, said the fact that there was no Mayday call was ``a strong indication of some sort of crew incapacitation.''

``If you're talking about the tail falling off or something like that,'' Forte said, ``then there would be nothing else to do but call.''

While there was no evidence for or against it, several experts said a hijacking attempt could have caused erratic or improper commands from the cockpit to the plane's control systems. On an EgyptAir flight on Oct. 20, an Arabic-speaking passenger kicked in the cockpit door on a flight from Istanbul and forced the crew to fly to Hamburg rather than Cairo.

400 feet a second

Without pilots at the controls, experts said, the natural tendency of the airplane would be to pull out of a dive if the speed became so great that the lift increased. It is not clear whether that could account for a climb as steep as the one reported by radar analysts.

The data show a dive from the assigned altitude of 33,000 feet down to 16,700 feet at a rate of more than 400 feet a second, triple the rate recommended by Boeing and about 10 times as fast as the speediest normal descent. But then the plane rolled to the right and dashed back up to 24,000 feet before it fell again. At 10,000 feet, almost two miles above the ocean, it took a turn to the left and apparently broke up.

But from the initial descent to 16,700 feet, about 37 seconds after the first dive began, the plane's transponder quit working and sent no more altitude information to controllers on the ground. From that point forward, altitude data came from the U.S. Air Force, which has radar installations in Riverhead on eastern Long Island and at Truro on Cape Cod that can estimate the altitude of incoming objects by recording the angle of elevation of the radar when the signal hit the object.

But some investigators are dubious that the plane could have climbed 1-1/2 miles and suggested that the interpretation of the radar data was wrong.

``Error increases with distances,'' one radar expert said. ``I don't think anybody's going to have a clue until we get the airplane off the bottom.'' Even then, he said, the plane's instruments could have been thrown off if the plane was spinning or skidding through the sky and its nose was not pointed forward.

Might be '50 replay

To other experts, a climb, commanded by the pilots, seems perfectly reasonable. A retired aeronautical engineer said that the EgyptAir accident could resemble the crash of a B-45 bomber in a government test in 1950.

In that case, the plane flew a profile that looked like peaks and valleys, but at the top of one peak it reached zero gravity and the pump for the hydraulic system, which was needed to move the flight-control surfaces, drew in air instead of hydraulic fluid and stopped working. The plane went into a dive until the system started working again, at which point the pilot pulled up so sharply that the resulting forces tore off most of the wings.