History of Flight 191

Delta Airlines Flight 191 was a regularly scheduled passenger flight between Fort Lauderdale, Florida and Los Angeles, California, with an en route stop at the Dallas/Fort Worth International Airport, Texas (DFW). The airplane, a Lockheed L-1011 with 152 passengers and a crew of 11 on board, departed Fort Lauderdale at 1510 Eastern Daylight Time (EDT). The DFW airport terminal weather forecast showed a possibility of widely scattered rain showers and thunderstorms, becoming isolated after 2000 Central Daylight Time (CDT). The flight's dispatch package also contained company Metro Alert No. T87, valid to 2100 CDT, which stated that "an area of isolated thunderstorms is expected over Oklahoma and northern and northeastern Texas...a few isolated tops to above FL450."

The flight was uneventful until the airplane passed New Orleans. A line of weather along the Texas-Louisiana border had intensified, encouraging the flight crew to change their route to the more northerly Blue Ridge arrival route. This change necessitated a 10 to 15 minute hold northeast of DFW for arrival sequencing.

At 1735:33, Fort Worth Air Route Traffic Control Center (ARTCC) cleared Flight 191 to the Blue Ridge, Texas VORTAC for the Blue Ridge Nine arrival and to begin its descent. Flight 191 had received the current Automatic Terminal Information Service (ATIS) which reported scattered clouds at 6,000 and 2,000 feet, 10 miles visibility, temperature 101°, and wind calm. Visual approaches were in progress to the south. At 1743:45 center control cleared Flight 191 to descend to 10,000 feet and suggested that the flight turn to a heading of 250° to join the Blue Ridge zero one zero radial inbound. Looking at the weather radar, Flight 191 indicated that they were painting a "pretty good size" weather cell "at a heading of two five five..." and stated, "...and I'd rather not go through it, I'd rather go around it one way or the other." Center control gave another heading and stated that they would turn Flight 191 into Blue Ridge when able. At 1746:50 center control cleared Flight 191 direct to Blue Ridge and to descend to 9,000 feet. At 1751:19 the flight engineer said, "Looks like it's raining over Fort Worth." At 1751:42 Fort Worth Center instructed Flight 191 to contact DFW Approach control. Flight 191 checked in with DFW Approach control indicating it was descending through 11,000 feet and was in receipt of the latest ATIS information. At 1756:28, Regional Approach Control's Feeder East controller transmitted an all aircraft message which was received by Flight 191. The message stated in part, "Attention, all aircraft listening... there's a little rain shower just north of the airport and they're starting to make ILS (Instrument Landing System) approaches... tune up one oh nine one for one seven left."
At 1759:47, the first officer remarked, "We're gonna get our airplane washed," indicating he was aware of rain between Flight 191 and the runway 17L at DFW.

Flight 191 was sequenced third to land on runway 17L, behind Flight 351 (a Boeing 727) and a Learjet 25. American Flight 539 (a McDonnell-Douglas MD-80) was to land behind Flight 191.

Flight 191 switched over to Regional Approach Control's Arrival Radar-1 frequency at 1759:47 and reported their altitude at 5,000 feet. At 1800:36, the approach controller asked American Airlines Flight 351 if it was able to see the airport. Flight 351 replied "as soon as we break out of this rainshower we will." The controller told Flight 351 that it was located four miles from the outer marker, and to join the localizer at 2,300 feet; the controller then cleared the flight for the ILS approach to runway 17L. All of these transmissions were recorded on Flight 191's Cockpit Voice Recorder (CVR).

Flight 351 was on its second approach after being directed to execute a missed approach due to inadequate spacing with preceding traffic still on the runway. The captain of Flight 351 noted after the accident that during the first approach he saw only scattered clouds and one thunderstorm northeast of the field. He could see the cell from the cockpit and "it looked harmless...like showers." The cell did not contour on his weather radar. After passing the outer marker, inbound Flight 351 did not encounter any rain or turbulence, and the captain did not see any lightning. After the missed approach, Flight 351 was then sequenced into the traffic flow for another approach. On downwind at 2,500 feet, the flight encountered a minor windshear which was not considered a safety hazard and was not considered reportable. Flight 351's captain testified that on final approach he did not go through any weather cells, the nearest one being two miles east of the aircraft. The captain reported that after departing the outer marker inbound he encountered heavy rain which lasted until he descended through 1,000 feet. He did not encounter any turbulence or windshear, and he did not see any lightning during the approach. Flight 351 landed at approximately 1804, about two minutes prior to Flight 191's crash.

At 1800:51, the approach controller asked Flight 191 to reduce its airspeed to 170 knots indicated, and to turn left to 270°. Flight 191 acknowledged receipt of the clearance. Flight 191 had been sequenced behind a Lear 25 for landing.

The Learjet 25 pilot stated after the accident that he had used his weather radar until he was about 25 miles from DFW and "nothing looked bad." He was able to see the cells visually. The Learjet pilot did not encounter nor did not perceive a windshear while on final approach. He did encounter "light to moderate turbulence" after passing the outer marker and flew at higher than approach airspeed and above glide slope to compensate. Otherwise he had nothing to report, testifying "the only thing that we encountered was the heavy rain." The Learjet landed approximately one minute prior to Flight 191's crash.

At 1802:35, the approach controller told Flight 191 that it was six miles from the outer marker, requested that it turn to 180° to join the localizer at or above 2,300 feet, and cleared Flight 191 for the ILS. At 1803:03, the approach controller requested Flight 191 to reduce speed to 160 knots. Thereafter, at 1803:30, he broadcast, "And we're getting some variable winds out there due to a shower...out there north end of DFW." This transmission was received by Flight 191 and an unidentified flight crew member remarked, "Stuff is moving in." At 1803:46, the approach controller requested Flight 191 slow to 150 knots and to contact the DFW tower. Flight 191 switched to tower frequency and at 1803:58 stated, "Tower, Delta one ninety one heavy, out here in the rain, feels good." The tower cleared the flight to land and informed it, "wind zero nine zero at five, gusts to one five." The aircraft was then configured for landing with gear down and flaps at 33°, the landing flap setting.

At 1804:18, the first officer said, "Lightning coming out of that one." The captain asked, "What," and the first officer repeated "Lightning coming out of that one." The captain asked, "Where," and at 1804:23, the first officer replied, "Right ahead of us." Flight 191 continued descending on the final approach course.

At 1805:05, Flight 191, at 1,000 feet and encountering the northern gust front of a microburst, experienced a headwind increase that accelerated the airspeed of the aircraft to 173 knots. In response, the flight crew retarded the throttles to near flight idle in an attempt to preserve the approach speed of 150 knots.

Diagram of a thunderstorm
Diagram of a thunderstorm
At 1805:19, at 800 feet, Flight 191 entered the rainshaft directly beneath the convective cell. The captain cautioned the first officer to watch his airspeed and a sound, identified as rain, was recorded on the CVR. At 1805:21, the captain warned the first officer, "You're gonna lose it all of a sudden, there it is." At 1805:26 the captain stated, "Push it up, push it way up." At 1805:29, throttles were full forward and the sound of engines at high rpm was heard on the CVR, and the captain said "That's it." During these ten seconds, the headwind decreased by 25 knots and the downdraft increased from about 18 feet per second to more than 30 feet per second (1800 feet per minute). Thrust was near flight idle for the first six seconds of this period and, combined with the loss in headwind component, aircraft airspeed dropped 30 knots. Despite being at full, or near full power, airspeed continued to decrease over the next four seconds to 129 knots, for a total loss of 44 knots in ten seconds. At this point, the decreasing trend of the headwind reversed itself, and along with the high thrust condition, resulted in a rapid increase in airspeed from 129 to 147 knots. At 1805:31, thrust was reduced and by 1805:35 the airspeed decreased to 140 knots. The airplane was just above 600 feet, still in the rainshaft, approximately ten knots slow, but had essentially maintained the glide slope despite airspeed fluctuations of +23/-21 knots and downdrafts ranging from 15 to 40 feet per second during the preceding 32 seconds.

At 1805:35, Flight 191 hit the southern gust front of the microburst and encountered an atmospheric disturbance which could best be described as severe and localized. The disturbance included three horizontal wind vortices which were the product of the interaction between the microburst outflow and the inflow that was feeding the convective cell. Within one second, large variations in wind components along all three axes of the aircraft were noted. Indicated airspeed decreased from 140 to 120 knots, the vertical wind reversed from a 40 feet per second downdraft to a 20 feet per second updraft, and a severe lateral gust struck the airplane. This gust resulted in a very rapid roll by the airplane to the right, requiring almost full lateral flight control authority to level the wings. Angle of attack increased significantly from 6° to at least 23°. It is assumed that the stickshaker activated though the CVR did not capture this sound.

Full power was applied at 1805:36 and by 1805:44 all three engines were producing full power. Throttles would remain at full until the first contact with the ground. In response to the extreme angle of attack and the pilot's nose-down control column movement, a rapid nose-down pitch rate developed. Also, the flight again experienced a reversal in the vertical wind component and the airplane began a rapid departure from the glide slope. A vertical acceleration of +0.3g was recorded by the flight data recorder. Vertical wind reversals induced large angle of attack excursions as the airplane descended away from the glide slope.

Beginning at 1805:40, a 30 knot increase in the tailwind component was recorded which resulted in an inability to accelerate beyond approximately 135 knots despite maximum thrust and a steepening flight path. By 1805:44, the airplane was at 420 feet above the ground, its descent rate was about 3,000 feet per minute, its airspeed began to increase, and it was in a strong downdraft. The CVR recorded the first Ground Proximity Warning System (GPWS) alert, and one second later the captain called "TOGA", presumably referring to his selection of the "TakeOff/Go-Around mode for the flight director. At 1805:46, with the airplane at about 280 feet above ground, its descent rate was close to 5,000 feet per minute. Two more GPWS alerts were recorded, and despite the rate of descent being arrested by a 2-g pull up, the aircraft touched down in a field at 1805:52, approximately 6,000 feet north of the approach end of runway 17L. About the same time, the airplane emerged from the rain shaft, airborne, crossing state highway 114 which runs east and west on the north side of DFW. The L-1011's left engine struck a car, instantly killing the driver. This impact and an impact with a light pole on the highway caused a fire to break out on the left side of the airplane in the vicinity of the wing root. Witnesses generally agreed that the airplane then struck the ground in a left-wing-low attitude, careening toward and striking two water towers on the airport property. The fuselage rotated counter-clockwise after the left wing and cockpit area struck the water tanks. A large explosion obscured the witnesses' view momentarily, and then the tail section emerged from the fireball skidding backwards. The tail section finally came to rest on its left side with the empennage pointing south and was subsequently blown to an upright position by wind gusts.

Eyewitness accounts

Witnesses to the accident and the environmental conditions at the time of the accident were in agreement that the storm was located just north of State Highway 114, or about 1.5 to 2 miles north of the approach end of runway 17L. The eastern and western edges of the storm were two miles east and one mile west of the extended centerline of runway 17L. Those witnesses who had encountered rain that evening described the rainfall as heavy to intense. Witnesses on the highway who saw Flight 191 emerge from the rain described it as coming out of a "wall" or "curtain" of water. Many witnesses reported seeing lightning and some witnesses heard thunder. Both were reported to have occurred when the storm was near the airport.

Witnesses who commented on the wind indicated that the wind flow was outward from the storm. One witness reported that several highway traffic signs had been uprooted and blown over. Another witness, about 3 to 4 miles north of the airport, reported that a trailer containing 1,200 pounds of fertilizer was overturned during the passage of the storm.
Several other flight crews in the area saw lightning and even a funnel cloud, but none of the crews reported any of the sightings to the airport control tower.

The Windshear Hazard

Windshear has long been identified as a flight hazard, and one that can be extremely dangerous during takeoff and landing operations. Windshear is defined as a change in wind direction and/or speed in a very short distance in the atmosphere. Under certain conditions, the atmosphere is capable of producing dramatic shears very close to the ground. One of the atmospheric conditions capable of producing dramatic shears is the downburst from convective or cumuliform clouds. A downburst is a strong downdraft which induces an outburst of highly divergent, damaging winds on or near the ground. Downburst diameters have been observed to range from as little as 0.5 mile to 10's of miles in diameter. The specific hazard to aviation has been extensively researched and analyzed to be related to the scale of the downburst. The smaller spatial scale of a small, intense downburst results in tighter windshear gradients that are experienced in the penetrating aircraft as more rapid changes in wind vectors, perhaps well in excess of the performance capabilities of the airplane. In order to identify and differentiate the various downdraft phenomena and to draw attention to the serious hazard posed to aircraft, the term microburst was coined by Dr. Theodore Fujita of the University of Chicago and describes a downburst with a diameter less than four km (2.2 nautical miles).

A microburst is a very localized column of sinking air which produces damaging, divergent, and straight line winds at the surface. The severe downward acceleration of a microburst occurs when a relatively cold parcel of air experiences negative buoyancy while sinking through a relativelly warm environmnet. The "wet" microburst is accompanied by shafts of strong precipitation reaching the ground.

A "dry" microburst differs in that there is no surface precipitation. In this case, when rain falls below cloud base or is mixed with dry air, it begins to evaporate, and this evaporation process cools the air. The cool air descends and accelerates as it approaches the ground. The presence of a "dry" microburst may be indicated by generally unstable air and "virga" (rain that evaporates before it hits the ground). But the dangerous acceleration of the air parcel is not visible to the naked eye and may not be sensed with Doppler radar systems. The hazard to aviation is the same. As air reaches the ground it spreads out in all directions, and this divergence of the wind is the signature of the microburst.

An airplane traversing a microburst initially encounters the outflow on the front side, which increases the headwind and/or updraft component, causing the indicated airspeed to increase and/or the airplane to rise. When exercising tight control of airspeed and flight path, as can be expected when flying an ILS approach, the pilot will retard throttle and/or lower the nose in an effort to maintain or re-capture the glide slope. If encountered at takeoff, the airplane may become airborne prematurely due to increased headwinds.

Several seconds later, the headwind component begins decreasing as the airplane traverses the central core downdraft which can be very strong, causing the airplane to descend in the column of air. During an approach, responding to the downdraft, the pilot will be required to add power and pull the nose up to maintain the glide slope. Here, pilot recognition and reaction times and engine "spool up" from a low-power setting may conspire to leave the airplane in a lower than desired energy state. During takeoff, the downdraft and reduced headwind will decrease climb rate, forcing the airplane into a low-energy state as it is forced to increase angle of attack in order to avoid the ground.

Finally, the airplane encounters the back side of the microburst, the outburst, where the tailwind component can increase dramatically, causing the airplane to descend even further in response to the resultant indicated airspeed decrease. Dependent on the intensity of the microburst, the airplane may also experience one or more severe horizontal vortex rings caused by the interaction of the outburst with the surrounding atmosphere and convective weather inflows. During an approach or a takeoff, the pilot may not have enough energy and/or performance capability to maintain control and avoid ground contact.

In the case of Flight 191, the airplane, already in a low-energy state from its transit through the middle of the microburst, encountered not just a strong tailwind but a series of three strong wind vortices which were parts of vortex rings which circled the main downdraft. It is estimated that Flight 191 would have needed to make the decision to execute a go around prior to 800 feet above the ground in order to have a chance to avoid inadvertent ground contact.

View Delta L-1011 Microburst Animation below: 

At typical takeoff and approach speeds for airplanes, the time to traverse a microburst may be only 20-40 seconds. It is not unusual for the wind differential across the microburst to exceed 50 knots and differentials of close to 100 knots have been observed in field studies. Flight 191's microburst horizontal wind differential was determined to be 73 knots.

A high speed video of an actual microburst, the result of the processes described in the previous video, is available below. (Tucson Microburst).  This copyrighted video was provided by Mr. Bryan Snider, a Phoenix, Arizona professional photographer, and is used here with his permission.

Mitigating the Windshear Hazard

At the time of this accident, numerous strategies had been implemented to combat the windshear hazard. This accident highlighted their shortcomings.

Crew Training

At the time of Flight 191's accident, windshear training had been mandated for air transport operations. Air carrier operators were required to implement windshear awareness training into their pilot training curriculum which included education on windshear recognition and avoidance of low-altitude windshears. The mandate did not, however, require flight training, and in the aftermath of Flight 191 it became clear that the current training was inadequate and possibly taught flight crews to attempt to cope with a windshear phenomenon when they should be escaping the situation.

Crew training now requires flight training in windshear recognition and escape maneuvers. If utilizing a simulator for this training, the simulator must be specifically equipped and authorized for windshear training. Windshear training now provides specific guidelines to allow for rapid recognition and escape maneuvers are taught that utilize all of the airplane's performance capability.

Low Level Windshear Alert System (LLWAS)

The LLWAS, which uses an array of wind sensors distributed around an airport to detect and alert Air Traffic Control personnel to the presence of low altitude windshear, had been operational for a number of years prior to the Flight 191 accident. However, due to the latency inherent in such a ground based system, the time required to transmit the information to interested airplanes, and the sudden, violent, nature of a hazardous microburst, information gained from these systems are of little practical use to an airplane in the final stages of landing or about to begin their takeoff roll. It was determined that the LLWAS at Dallas Fort Worth could not have provided any timely windshear warning to Flight 191. LLWAS continues to provide useful information for pilot decision making but is not very useful as a predictive tool for rapidly changing wind conditions.

Terminal Doppler Weather Radar (TDWR)

The TDWR system detects and reports hazardous weather in and around airport terminal approach and departure zones. Doppler radar technology, which can detect the presence of windshear by measuring the velocity and direction of wind-driven precipitation and other particles suspended in the air, was in a mature state of development but had not yet been deployed at airports at the time of the Flight 191 accident. In the wake of Delta Flight 191, ground-based Doppler radar development was accelerated and was fully deployed at 45 high-risk United States airports by 1997.

Airborne Weather Radar

Delta Flight 191, like most airplanes of the time, had a weather radar system which was primarily designed for en route weather avoidance. With a minimum range scale of 50 nautical miles and with the necessary multiple manual adjustments of antenna tilt to "filter" out ground returns at low altitude, the system was of little use in characterizing weather when in the terminal area and was not capable of detecting windshear.

Airborne windshear detection and avoidance systems were in their infancy at the time of the Flight 191 accident and had only recently been introduced for transport category aircraft. At the time of this accident, these systems were limited in that they were reactive and only capable of warning a flight crew of a shear condition after the airplane had entered the condition. Flight 191 was not equipped with such a system.

Predictive windshear systems had not yet come of age at the time of the Flight 191 accident but were introduced into commercial operations in 1994, their development spurred on largely by the Flight 191 accident. These systems use technology to alert flight crew to the presence of windshear ahead of their airplane. Coupled with flight guidance information and appropriate flight crew procedures, these systems are very effective in avoiding the "wet" microburst threat.

In 1988, the FAA mandated that all air carrier turbine aircraft be equipped with an airborne windshear warning and flight guidance system, an approved airborne detection and avoidance system, or an approved combination of these systems.

Residual Hazards

Application of Airborne Doppler Radar Predictive Windshear Systems has provided a necessary detection tool to allow pilots to "see" and avoid hazardous windshears when precipitation or airborne particulates are of a sufficient size to be detected by radar. However, there remains hazardous "clear air" phenomena that cannot yet be detected by commercially available systems nor the naked eye. These "clear air" phenomena, including "dry" microbursts, may be generated by convective storm outflows, aircraft wakes, jet stream confluences, or may be terrain-induced as in the case of mountain waves and rotors. Most of the non-convective phenomena have been classified as Clear Air Turbulence (CAT) while wake encounters have earned their own classification as a causal or contributing factor in several recent accidents. All of these phenomena are, arguably, examples of windshear in that they exhibit "a change in wind direction and/or speed in a very short distance in the atmosphere."

Back to top

Between 1752 and 1800, a thunderstorm cell positioned off the north end of the DFW airport intensified from a VIP Level 1 (Weak) to a VIP Level 4 (Very Strong).

On final approach to runway 17L at DFW, Flight 191 continued the approach under a convective storm cell which was observed to contain lightning and was depositing a heavy rain shower. This convective storm cell produced an outflow containing a microburst. The microburst touched down just north of the DFW airport. The microburst diameter was 2.1 miles and was centered two miles north of the approach end of the intended runway, approximately 1,000 feet west of the runway's extended centerline. The horizontal windshear across the microburst was at least 73 knots, and the maximum updraft and downdraft were 25 feet per second (1500 feet per minute) and 49 feet per second (2940 feet per minute), respectively. There were six distinct reversals of vertical wind components along the southern side of the microburst, indicating that a horizontal vortex structure had formed along the boundary between the descending air and the ambient environment. Flight 191 penetrated the microburst and the vortex flow in the southern side of the microburst and initially touched down approximately 6,000 feet short of the runway at 1805:52. The flight then crossed a highway where the left engine struck a car and a light pole. Ultimately, the flight impacted water towers on the airport property where the tail section separated from the rest of the wreckage.

The board found that the flight crew of Flight 191 had sufficient information to assess the weather north of the approach end of Runway 17L. The lightning observed and reported by the first officer was adequate, combined with the other data known to the flight crew and captain, to determine that there was a thunderstorm between the airplane and the airport.

The captains' decision to continue the approach beneath the thunderstorm did not comply with the airline's weather avoidance procedures; however, the avoidance procedures did not specifically address thunderstorm avoidance in the airport terminal area.

The first officer successfully transited the first part of the microburst encounter by rotating the airplane above 15° nose-up pitch attitude and by increasing engine thrust to almost takeoff power. After penetrating the first part of the microburst, the engine thrust, which had been increased, was then reduced, and at just under 600 feet above ground the airplane had re-stabilized momentarily on the glide slope. It was the board's conclusion that the captain evidently believed they had successfully flown through the worst of the microburst windshear, and the approach was continued.

Seventeen seconds before initial impact, the airplane encountered rapid reversals in the lateral, horizontal, and vertical winds, causing the stickshaker to activate. The first officer exerted a 20- to 25-pound push force on the control column in response to the stickshaker. The flight director was placed in TOGA (Take-Off Go-Around) mode during the initiation of a missed approach seven seconds before initial impact. The flight director's TOGA mode did not command the optimum pitch attitudes required to transit a low-altitude windshear, opting instead to capture a reference airspeed. It is undetermined whether the first officer was following the pitch commands provided by the flight directors' TOGA mode during the final seven seconds of the flight.

The first officer exerted a 20- to 25-pound pull force on the control column in order to avoid ground contact which arrested the sink rate. However, the stickshaker activated momentarily, and the first officer, presumably in accordance with his past training, relaxed the pull force on the control column, which made ground contact inevitable.

The airline had not provided guidance to its flight crews concerning specific limits on the excursions of airplane performance and control parameters during low-altitude windshear encounters that would dictate the execution of a missed approach.

Although the captain did not audibly express his decision to execute a missed approach until he called for the selection of the TOGA mode on the flight director seven seconds before initial impact, maximum thrust had been applied before the airplane's rapid departure below the glide slope.

View NTSB Accident Report for Delta Flight 191

Back to top

Low altitude windshear had long been identified as a significant hazard in the aviation community. The National Transportation Safety Board (NTSB), who investigated the crash of Flight 191 along with multiple accidents that preceded it, used the accident as a forum to reiterate previous recommendations and urge that the findings from extensive windshear research and development activities that had been conducted over the previous 10 to 15 years be transferred into tangible benefits for aviation safety:

The most significant recommendations were:

• Development and installation of Doppler radar equipment at airports located in areas of high microburst risk must be expedited to provide improved real-time detection of windshear conditions. Improvement of the LLWAS (Low Level Windshear Alert System) to ensure optimum placement of the anemometer array and optimum software alarm logic.
• Pilot training must be improved to discuss the meteorological conditions conducive to the development of windshears, particularly convective windshears. Moreover, the training must stress avoidance of windshear.
• Pilot training must be improved to discuss the aerodynamic performance problems associated with windshear penetrations as well as simulations of windshear encounters during all low-altitude phases of flight. Pilot training must also stress the need for rapid recognition and response by using all of the airplane's performance capability, and address the effect of an out-of-trim speed condition on the control forces needed to use the airplane's performance.
• Development, certification, and installation of airborne equipment, which can provide the pilot early warning of windshear encounters and optimize the logic of command guidance instruments to enhance the pilot's response to an inadvertent windshear encounter.
• Windshear forecasting needs to better define the conditions conducive to microburst development as well as informing dispatchers and pilots when these conditions are present. Weather forecasting also needs to define when there is a windshear potential involving non-frontal systems.
• Improved communications between the weather service, air traffic controllers, and pilots is required to ensure that flight crew are provided the most current forecasts and existing conditions for planning flights, landing approaches, and departures.

Additional recommendations included:

• Air carrier operations manuals and training should be required to record pilot's specific windshear simulator training during initial and recurrent training sessions.
• Air carrier operations should establish thunderstorm avoidance procedures to verify that flight crews clearly understand the policy that no aircraft should attempt to land or take off if its flight path is through, under, or near (within a specified distance) a thunderstorm.
• In the event of an inadvertent windshear encounter, air carrier operations should not follow flight director guidance unless such systems incorporate windshear logic.
• ATIS (Automatic Terminal Information System) should broadcast a message whenever weather conditions conducive to thunderstorm or microburst development exist in the terminal area or when such actual conditions have been observed or reported. In the absence of an ATIS update, tower controllers should issue thunderstorm, microburst, and windshear reports when conditions differ from the ATIS.
• Develop a position in major terminal facilities, to be staffed with NWS (National Weather Service) meteorologists or FAA personnel trained for meteorological observations, to be the focal point for weather information coordination during periods of convective weather activity that adversely affects aircraft and air traffic control system operations.
• Eliminate gaps in surveillance and improve weather information coordination during periods of convective weather activity in major terminal facilities.

Back to top

At the time of the Delta Flight 191 accident there were no specific regulations regarding windshear. An FAA Advisory Circular; AC 00-50A, entitled Low Level Windshear existed which provided information on meteorology, windshear detection, airplane performance in windshear, and procedures for coping with windshear.

Back to top

Delta Flight 191 was but the latest in a large number of accidents attributed to the windshear hazard. During the period from 1964 to 1985, over 30 accidents and incidents had occurred in which windshear was identified as a contributing factor. These accidents and incidents had resulted in over 500 fatalities. Windshear had been studied in the wake of several previous accidents and the concept of "downburst" and the relationship of a downburst's spatial scale to the energy and difficulty in detecting the hazard had led to the term, "microburst." Though the hazard was becoming better understood, the only tool in use was the Low Level Windshear Alert System which had severe limitations as a detection tool for microbursts. Doppler radars were not yet operational at worldwide airports. At the time of this accident, a reactive windshear detection system, intended for installation on aircraft, had been approved for incorporation into the airline fleet, but had not yet undergone widespread installation. The FAA was in the process of making incorporation mandatory.

At the time of the Flight 191 accident, pilot low-level windshear training programs were in effect which discussed the necessity of avoiding microburst windshears, but also recognized that in some instances a pilot might inadvertently encounter one. As a result, typical simulator curricula taught procedures for coping with windshear, and this procedures training could have been a factor in the decision of the Flight 191 captain to continue the approach. Some training appeared to advocate a philosophy that the retrieval of the approach profile was the desired end result, not escape from the windshear environment. Delta and other operators of the time generally taught the windshear procedure of using maximum thrust, increasing the airplane nose-up pitch attitude, and allowing airspeed to decrease to near stickshaker speed if necessary to avoid ground contact, and lowering the nose slightly if the stickshaker was actuated. Although the captain's and first officer's training records did not show that they received this training, they probably received it during their Line Officer Flight Training (LOFT) and recurrent training periods. The captain's instructions to the first officer concerning the impending loss of indicated airspeed after they penetrated the microburst's downdraft and his subsequent commands to apply full power tend to corroborate that he, at least, had received this training.

By 1985 a large amount of airport terminal weather research had been conducted that had yielded significant information and understanding of microburst windshears. In 1982 the National Center for Atmospheric Research (NCAR) and the University of Chicago conducted field experiments near Denver, Colorado as part of the Joint Airport Weather Studies (JAWS) project. This research involved the use of Doppler radars among other data sources and was followed up by project CLAWS (Classify, Locate, and Avoid Windshear). The FAA, multiple agencies, and research groups had participated in this work, and the knowledge gathered had resulted in an operational windshear detection system , and new procedures for windshear encounters. At the time of this accident, installation had not yet been made mandatory.

Back to top

The Low Level Windshear Alert System (LLWAS) was inadequate to alert flight crews in the terminal area of a hazardous windshear situation. This inadequacy had been acknowledged since the system's inception but more sophisticated equipment that could provide forecasting capability to flight crews had not yet reached a level of development despite ongoing research.

Flight crews were not equipped with specific parameters to define the magnitude of an inadvertent windshear encounter and were ill equipped to make decisions on whether to continue an approach or attempt a go-around. Specific windshear escape maneuvers were not defined.

Gaps existed in the ability to gather and disseminate operationally useful hazardous weather updates to flight crews in the Dallas Fort Worth International terminal area.

Back to top

Most low altitude windshear encounters could be dealt with by detecting the presence of the phenomenon then using proper piloting technique to traverse the windshear.

Back to top

July 23, 1973, Ozark Airlines Flight 809

Ozark Airlines Flight 809, a Fairchild-Hiller 227B, crashed near St. Louis Airport, killing 38 of the 44 persons on board. Probable cause was the aircraft's encounter with a downdraft following the captain's decision to initiate and continue an instrument approach into a thunderstorm. Shortly before the crash, the flight crew was informed by the tower controller of a heavy rain shower moving right across the approach end of the runway. The crew acknowledged that they could see the shower but elected to continue the approach. After passing the outer marker, the aircraft began to descend below the glide slope until it struck the ground.

December 17, 1973, Iberia Flight 933

Iberia Flight 933, a McDonnell Douglas DC-10, approaching Boston's Logan International airport in bad weather (rain, fog, two-mile visibility) struck approach lights 500 feet short of the threshold and collided with a dike. The right main gear was sheared off, and the aircraft skidded and came to a rest 3,000 feet from the threshold. The accident caused injuries but no fatalities. Information from the aircraft's flight data recorder helped investigators establish the presence of windshear. Subsequent study of this accident led to a new understanding and awareness of the windshear hazard. In this accident the pilot did not recognize, and may have been unable to recognize, the increased rate of descent in time to arrest it before the aircraft struck the approach light piers. The increased rate of descent was induced by an encounter with a low-level windshear at a critical point in the landing approach where the pilot was transitioning from automatic flight control under instrument flight conditions to manual flight control with visual references. The pilot's ability to detect and arrest the increased rate of descent was adversely affected by a lack of information as to the existence of the windshear and the marginal visual cues available.

June 24, 1975, Eastern Airlines Flight 66

Eastern Airlines Flight 66, a Boeing 727, struck approach lights and crashed during an attempted landing at New York's Kennedy airport. Of the 124 occupants on board the aircraft, 113 were killed. Despite being aware of an L-1011 crew who had just abandoned their approach to the same runway due to "severe windshear," the tower controller cleared the 727 to land and the crew elected to continue their approach despite the presence of a very strong thunderstorm located over the final approach course. During transition from instrument to visual conditions, the aircraft experienced a high descent rate due to windshear which caused the aircraft to contact the approach light towers, ultimately severing the left wing which ignited a severe fire and caused the aircraft to roll into a steep left bank, ultimately leading to destruction of the aircraft. The flight crew's delayed recognition and correction of the high descent rate were probably associated with their reliance upon visual cues rather than on flight instrument reference. However, the adverse winds might have been too severe for a successful approach and landing even if they had relied upon, and responded rapidly to, the indications of the flight instruments. Contributing to the accident was the continued use of runway 22L when it should have become evident to both air traffic control personnel and the flight crew that a severe weather hazard existed along the approach path.

See accident module

August 7, 1975, Continental Flight 426

Continental Flight 426, a Boeing 727, attempting to takeoff from Denver's Stapleton airport runway 35L, reached approximately 100 feet altitude then crashed near the departure end of the runway. At the time of the accident, a thunderstorm with associated rain showers was moving over the northern portion of the airport. The National Transportation Safety Board determined that the probable cause of this accident was the aircraft's encounter, immediately following takeoff, with severe windshear at an altitude and airspeed that precluded recovery to level flight. The windshear caused the aircraft to descend at a rate which could not be overcome, even though the aircraft was flown at or near its maximum lift capability throughout the encounter. The windshear was generated by the outflow from a thunderstorm which was over the airplane's departure path. The 134 persons on board the aircraft survived the crash though 15 were severely injured.

July 9, 1982, Pan Am Flight 759

Pan Am Flight 759, a Boeing 727, attempted to depart the New Orleans International Airport during a severe thunderstorm. The aircraft climbed to approximately 100 feet when it began to descend, striking trees past the end of the runway and crashing into a residential area. All 145 people on board the aircraft along with eight people on the ground were killed in the accident. Probable cause was identified as "the airplane's encounter during the liftoff and initial climb phase of flight with a microburst-induced windshear which imposed a downdraft and a decreasing head wind, the effects of which the pilot would have had difficulty recognizing and reacting to in time for the airplane's descent to be arrested before its impact with trees. Contributing to the accident was the limited capability of current ground-based low-level windshear detection technology to provide definitive guidance for controllers and pilots for use in avoiding low-level windshear encounters."

Back to top

Section 121.358 was added to the Federal Aviation Regulations requiring operators to equip airplanes with either an approved airborne windshear warning and flight guidance system, an approved airborne detection and avoidance system, or an approved combination of these systems.

Section 121.424 was added to the Federal Aviation Regulations requiring operators to implement a low-altitude windshear flight training program. The training must include windshear maneuvers and procedures and, if utilizing a simulator for this training, the simulator must be specifically equipped and authorized for windshear training.

In addition to the above regulations, AC00-50A, Low Level Windshear was superseded by a Windshear Training Aid package; a multimedia training presentation which included a new Advisory Circular; AC00-54; "Pilot Windshear Guide." This package was developed by an FAA-contracted consortium and stressed detection and avoidance as the best defense against low-level windshears.

Back to top

None.

Back to top

Airplane Life Cycle:

• Operational

Accident Threat Categories:

• Incorrect Piloting Technique
• Windshear

Groupings:

• Loss of Control
• Approach and Landing

Accident Common Themes:

• Human Error
• Flawed Assumptions

Back to top

Human Error

The decision of the flight crew to continue the approach underneath a convective cell that was observed to be producing lightning was a clear human error. The separation of electrical charges that lead to lightning in a convective cell are brought about by the collisions between upward rising warm moisture particles that cool as their altitude increases and falling ice particles whose weight can no longer be sustained by the convective cell updrafts. Lightning is a visual indicator that severe updrafts and downdrafts are present. Flight under a thunderstorm exposes the aircraft to an increased risk of severe turbulence, hail, and downdrafts.

Flawed Assumptions

As evidenced by the Cockpit Voice Recorder from the Flight 191 accident, it was apparent that the flight crew assumed they could out-perform the environmental conditions they were faced with during this approach. Perhaps this assumption was reinforced by the successful approaches of other aircraft that were landing at Dallas-Fort Worth at the time. As mentioned in the previous section, the flight crew would have needed to execute a go-around before descending below 800 feet in order to avoid inadvertent ground contact for this windshear event.

Back to top

June 24, 1975, Eastern Airlines Flight 66

Eastern Airlines Flight 66, a Boeing 727, struck approach lights and crashed during an attempted landing at New York's Kennedy airport. Of the 124 occupants on board the aircraft, 113 were killed. Despite being aware of an L-1011 crew who had recently abandoned their approach to the same runway due to "severe windshear," the tower controller cleared the 727 to land and the crew elected to continue their approach despite the presence of a very strong thunderstorm located over the final approach course.  The aircraft experienced a high descent rate due to a severe downdraft which caused the aircraft to contact the approach light towers short of the runway, resulting in the destruction of the aircraft.  Meteorological analysis of this accident led to the discovery and characterization of the “downburst.”  This accident became the catalyst for further research which led to the definition of a specific danger to aviation known as a “microburst.”

This accident highlighted the inadequacies of terminal area weather analysis related to the recognition and reporting of severe weather information. It also highlighted the need for flight crews to recognize and avoid low-altitude hazards associated with thunderstorms along or near the approach path.  This accident resulted in the development and deployment of the Low Level Windshear Alert System (LLWAS) at over 100 airports over the next decade.

See accident module

March 3, 1991, United Flight 585

United Flight 585, a Boeing 737-200, crashed while attempting to land at Colorado Springs, Colorado after a short flight from Denver, killing all 25 occupants. After being cleared for a visual approach to runway 35, and after descending below 1,000 feet, the airplane suddenly rolled to the right and impacted the ground in a near vertical attitude. After a 21-month investigation, the NTSB issued a report stating that it could not identify conclusive evidence to explain the loss of' the aircraft but indicated that the two most likely explanations were a malfunction of the airplane's directional control system or an encounter with an unusually severe atmospheric disturbance.

The accident investigation board later amended the accident report based on two accidents of similar airplanes, that the probable cause of this accident was a loss of control of the airplane resulting from an uncommanded movement of the rudder surface. In the amended report the board concluded that weather conditions in the Colorado Springs area at the time of the accident were conducive to the formation of a horizontal axis vortex (rotor) and that some witness observations supported the existence of a rotor at or near the time and place of the accident. However, the board determined that too little was known about the characteristics of rotors to conclude whether a rotor was a factor in the accident.

July 2, 1994, USAir Flight 1016

USAir Flight 1016, a Douglas DC-9, crashed after experiencing a thunderstorm-induced microburst while attempting to land at Charlotte-Douglas airport killing 37 of the 57 persons on board. The weather information provided to the flight crew from USAir dispatch indicated that the conditions at Charlotte were similar to those encountered when the crew had departed there approximately one hour earlier. The only noted exception was the report of scattered thunderstorms in the area. Approaching Charlotte, the DC-9 flight crew noted visually and on airborne weather radar that rain and cloud buildups had moved into the vicinity of the airport. With visual contact of the runway, the flight crew turned from base to final and followed a Fokker 100 who had reported a smooth ride on their approach. The cloud buildup had evolved into a thunderstorm and the DC-9 entered an area of rainfall and experienced a 20-knot deviation in airspeed on final approach to which the crew reacted by executing a missed approach. The first officer initially rotated the airplane to the proper 15 degrees nose-up attitude during the missed approach. However, the thrust was set below the standard go-around EPR limit of 1.93, and the pitch attitude was reduced to 5 degrees nose down before the flight crew recognized the dangerous situation. The flight crew failed to establish and maintain the proper airplane attitude and thrust setting necessary to escape the windshear, and the airplane struck the ground after transitioning from a headwind of approximately 35 knots to a tailwind of 26 knots (a change of 61 knots), over a 14 second period.

For a list of accidents that occurred prior to Delta Flight 191, see the Precursors section of this module.

Back to top

Technical Related Lessons:

There are some natural phenomena that must be detected and avoided, as their severity can be such that they will exceed the performance capability of any pilot or airplane. (Threat Category: Windshear)

• At the time of this accident, the preponderance of windshear training was focused on maintaining flight path and continuing the approach. Despite observing lightning between the airport and their location, Flight 191 continued an approach and experienced a severe microburst as it attempted to traverse a heavy rain shower directly underneath a very strong convective cell. The microburst created multiple, severe horizontal vortical wind flows and a strong tailwind close to the ground that forced the airplane into the ground 6,000 feet short of the runway. The development of windshear alerting systems provided another means for a flight crew to assess the severity of an encounter and elect to terminate an approach rather than continue into a situation that exceeds the capabilities of the airplane.

In the event of an inadvertent entry into a severe windshear, an escape maneuver must be adopted promptly and correctly. Flight crews should be trained and equipped to recognize windshear and use all of their aircraft's available performance should the need arise. (Threat Category: Incorrect Piloting Technique)

• Delta and most major air carriers at the time of this accident taught their flight crews to trade airspeed for altitude if they inadvertently encountered low-altitude windshear. As it entered the microburst, Flight 191 encountered low altitude headwinds and downdrafts. At 600 feet on final approach, the flight remained on the glideslope despite having experienced a greater than ±20 knot windshear. As the flight descended below 600 feet, it was subjected to a strong tailwind shear and strong horizontal vortices from which the airplane was incapable of escaping. Subsequent to this accident, alerting systems have provided an aid in recognizing windshear, and training programs have emphasized the necessity to avoid the windshear by terminating the approach at an earlier phase.

Common Theme Related Lessons:

Due to the sudden, violent, and highly localized nature of a microburst, a pilot's decision to continue a landing or perform a takeoff should not only be predicated on the success of a preceding aircraft, regardless of the physical size or performance capability of the aircraft involved. (Common Theme: Human Error)

• Delta Flight 191 was one in a long line of airplanes approaching Dallas Fort Worth and was preceded by a Boeing 727 that encountered heavy rain until descending through 1,000 feet. However, the 727 did not encounter any turbulence or windshear nor did it see any lightning during the approach. The 727 landed two minutes before Flight 191's crash. A Learjet landed less than a minute before the Flight 191 crash. The Learjet pilot did not encounter (or did not perceive) a windshear while on final approach. He did encounter "light to moderate turbulence" and heavy rain after passing the outer marker and flew at higher than approach airspeed and above glide slope to compensate, but otherwise had nothing to report.

Back to Top