Pilots of the Pan American World Airways jetliner that crashed on takeoff near New Orleans last week received two radio warnings of potentially dangerous "wind shear" shortly before acceleration and discussed ways to abort the takeoff if necessary, the cockpit recorder indicates.

Preliminary findings from the jet's two flight recorders were released yesterday by Patricia Goldman, vice chairman of the National Transportation Safety Board, with the warning that they might change as analysis continues.

Sound quality on the cockpit tape is reported to be very low, and many of the pilots' words remain unintelligible.

According to the tape, seconds before the Boeing 727 crashed in Kenner, La., killing 154 persons, Capt. Kenneth McCullers said the plane was "sinking." He spoke about how to stop the descent as he apparently gave directions to copilot Donald Pierce, who was flying the jet.

Shortly before impact, an alarm horn sounded, followed by a recorded voice saying "pull up, pull up," warning the crew that the plane was too close to the ground, sources close to the investigation said.

The plane's second "black box," the flight data recorder, showed that the plane had reached an airspeed of 158 knots at liftoff, well above the 138 knots that 727 pilots normally try to attain in similar conditions.

The three-engine jet gained altitude for 14 seconds, but airspeed dwindled to 140 knots. Then the plane began descending, gaining speed before it smashed into a residential area.

This evidence from the recorders is compatible with theories that the plane was caught in violent wind shear, rapid changes in wind speed and direction, investigators said. Rain was falling, and thunderstorms were in the area at the time of the accident.

Precisely what information on weather conditions was received by the pilots and how they reacted to it is emerging as a central issue in the investigation.

Describing findings on the cockpit tape, Goldman said yesterday: "The crew performed the preflight checklist and requested weather information from air traffic control. They were advised by air traffic control that a low-level wind shear alert existed in all quadrants," the directions from the airport's center.

This was the first public confirmation that a wind shear alert had reached the cockpit of Flight 759.

Wind shear is detected by devices that compare air speed and direction inside and around an airport. Pilots generally say a wind shear alert is not an outright warning not to take off but is one more piece of information to be used in deciding whether to take off.

The voice recorder indicates that pilots also heard "an air traffic control advisory made to another departing aircraft regarding wind shear experienced by yet another aircraft on approach," Goldman said. " . . . The crew of Flight 759 discussed emergency procedures, including abort procedures," she said.

Aviation specialists said such discussion does not necessarily mean that the crew was unusually concerned about the weather. Planning who will do what in an emergency is standard procedure in preparing for takeoff in many airlines, they said.

"Windshield wipers were used on takeoff," Goldman said. After liftoff, she said, "the crew noted a positive rate of climb and called for [landing] gear retraction. The captain stated the aircraft was 'sinking' and discussed action to counteract the descent. The ground proximity warning device . . . was sounded until interrupted by initial impact."

It remains unclear whether the pilots knew they were in trouble while the plane was accelerating on the runway. However, stopping the plane on a wet runway after reaching high speed might have been dangerous in itself. The 9,200-foot runway used by Flight 759 has no safety overrun area.

In the seconds after liftoff, airplanes are supposed to maintain a steady increase in altitude and airspeed, the speed at which they are passing through the air around them as opposed to speed in relation to the ground.

If air around a plane suddenly changes direction, the plane's airspeed may drop rapidly, affecting its ability to remain airborne.

One explanation for Flight 759's airspeed at liftoff being about 20 knots faster than the target speed may be that the jet faced a strong headwind as it accelerated, raising its airspeed quickly, then passed through a wind shear area and into a tailwind that slowed its airspeed.

Another possible explanation for 759's loss of speed after takeoff is loss of power in an engine. However, technicians tearing down the jet's engines in New York have found no evidence of engine failure, investigation sources said.