Blown Away

With thunderstorms, a minute or two can make a huge difference

June 1, 1995

There were ill winds last July when a DC-9 and a Beech Bonanza became victims of thunderstorm wind shear. The two accidents recounted here are flip sides of the same coin. The phase of flight and the circumstances are different, but the aerodynamic and avoidance principles are the same.

It was two decades ago that we developed the terms "downburst" and "microburst." They aptly describe the big downdrafts that make up the outflow of all thunderstorms. Fly into one and you may not have enough lift to overcome the downward rush of air or the rapid change in wind direction emanating from the base of the storm.

There was no large outbreak of strong thunderstorms in North Carolina as the DC-9 was approaching Charlotte/Douglas International Airport from the southwest. The crew discussed a shower sitting at the edge of the airport and expected a visual approach. Following is a transcript of the communications in the cockpit and with Charlotte Approach and Tower. Conversation not pertinent to the weather has been omitted. The first officer was the pilot flying.

US1016: Radio transmission from the flight
CAPT: Captain cockpit communication
FO: First officer cockpit communication
APPR: Charlotte Approach Control
TWR: Charlotte Tower

18:36:59 APPR: "Tell you what, USAir 1016, may get some rain just south of the field. Might be a little bit just coming off the north. Just expect the ILS now. Amend your altitude...maintain 3,000."
18:39:02 CAPT: "If we have to bail out.... (unreadable)"
18:39:06 CAPT: "It looks like we bail out to the right."
18:39:09 FO: "Amen."
18:39:09 CAPT: "Ten miles to the VOR which is off the end of the runway. 'Bout a mile off the end of the runway."
18:39:14 FO: "Yeah."
18:39:16 CAPT: "So I think we'll be all right."
18:39:20 CAPT: "Chance of shear."
(The dialog in italics occurred on the tower frequency before USAir 1016 switched over at 18:39:30.)
18:39:12 US806 (on the ground, waiting to depart from Charlotte): "And 806, looks like we've gotten a storm right on top of the field here."
18:39:16 TWR: "USAir 806, affirmative."
18:39:20 US806: "We'll just delay for a while."
18:39:44 TWR: "...Charlotte Tower, Runway 18 Right, cleared to land. Following an F[okker] 100 (an airliner about the same size as a DC-9) on short final. Previous arrival reported a smooth ride all the way down final."
18:39:49 US1016: "USAir 1016, I'd appreciate a pirep from the guy in front of us."
18:40:10 FO: "Yep, laying right there this side of the airport, isn't it?"
18:40:14 CAPT: "Well."
18:40:15 FO: "The edge of the rain is, I'd say."
18:40:15 CAPT: "Yeah."
18:40:42 TWR: "USAir 1016, company FK 100 just exited the runway, sir; he said smooth ride."
18:40:48 TWR: "USAir 916, wind is showing 100 at 19."
18:40:56 FO: "One hundred at 19, eh?"
18:40:59 TWR: "USAir 1016, wind now 110 at 21."
18:41:05 CAPT: "Stay heads up."
18:41:06 TWR: "Wind shear alert, northeast boundary winds 190 at 13."
18:41:18 TWR: "Carolina 5211, Charlotte Tower, runway 18R, cleared to land, wind 100 at 20. Wind shear alert, northeast boundary wind 190 at 17."
18:41:32 TWR: "USAir 806, you want to just sit tight for a minute, sir?"
18:41:35 US806: "Yes, sir, we'd like to just sit tight."
18:41:37 TWR: "USAir 797, company aircraft in front of you is going to sit and wait a while, sir. Do you want to go in front of him?"
18:41:43 US797: "No, no, it wouldn't sound like a good plan. We'll, uh, it didn't look like a whole lot to us on the radar taxiing out, so it shouldn't be, uh, shouldn't be too many minutes."
18:41:54 CAPT: "Here comes the wipers."
18:41:56 FO: "All right."
18:41:57.6 (sound similar to rain concurrent with sound similar to windshield wipers starts. The sound continues until impact)
18:41:58.9 FO: "There's, ooh, 10 knots right there."
18:42:06.4 CAPT: "OK, you're plus 20."
18.42.14.0 CAPT: "Take it around; go to the right."
18:42:16.1 US1016: "USAir 1016's on the go (the DC-9's altitude is approximately 200 feet agl)."
18:42:17.7 CAPT: "Max power."
18:42:18.5 FO: "Yeah, max power.... "
18:42:18.5 TWR: "USAir 1016, understand you're on the go sir, fly runway heading. Climb and maintain 3,000."
18:42:19.4 FO: "Flaps to 15 [degrees]."
18:42:20.8 (clicks similar to flap handle being moved)
18:42:22.0 CAPT: "Down, push it down."
18:42:25.5 US1016: "Up to three, we're takin' a right turn here."
18:42:27.9 TWR: "USAir 1016, understand you're turning right? (US1016's altitude begins decreasing below 350 feet agl)."
18:42:28.4 ("Whoop whoop terrain" sound begins and continues to first sound of impact)
18:42:28.5 CAPT?: " — power."
18:42:32.7 (vibrating sound similar to aircraft stick shaker begins)
18:42:35.6 (sound of impact)

The NTSB determined the probable cause to be the flight crew's decision to continue an approach into weather conducive to a microburst, the crew's failure to recognize a wind shear situation in a timely manner, its failure to establish and maintain the attitude and thrust necessary to escape the wind shear, and the lack of real-time adverse weather and wind shear hazard information dissemination from air traffic control.

Several aircraft on the ground delayed takeoff despite the pirep of a smooth ride from the Fokker 100. When the tower issued the wind shear alert at 18:41:18, it was probably time to abandon the approach. The crew was focused on landing, but the weather was changing rapidly as the runway visual range dropped to 500 feet in heavy rain. It was determined later that there was a microburst one mile east of the approach end of 18R, with maximum winds of 50 knots. The net wind change was calculated at 86 knots during the DC-9's missed approach.

The flight crew's coordination could have been better, and the investigation concluded that the flight could have survived this wind shear encounter if several factors had been different:

  • Maximum go-around power was not used. Approximately 2,000 more pounds of thrust were available. The first officer used the missed approach procedure instead of the wind shear recovery profile.
  • Missed approach calls for max normal power, a maximum 15-degree nose-up attitude, flaps retracted to 15 degrees, and gear up. The wind shear profile mandates full throttle, 15 degrees nose up initially followed by two-degree increases as needed to stick shaker, and the aircraft is left in its current configuration (gear and flaps down in this case).
  • The captain may have experienced vestibular illusion that gave him a false sense of the aircraft's being over-rotated to an abnormally steep pitch attitude. This is caused by the strong acceleration of the aircraft and must be overcome by reference to the flight instruments. The first officer, who was flying, was on the gauges and situationally focused. The captain was not, which led him to call for a reduction in pitch attitude.
  • The tower was cited for not reporting observed lightning to the aircraft and for failure to mention the rapidly diminishing visibility in heavy rain and other multiple low-level wind shear alerts.

The crew's adherence to the wind shear escape procedure could have saved the day, but the margins would have been razor thin and dependent on doing it all exactly right. The best decision would have been to delay the landing. Thunderstorms mutate rapidly; and even with a recent pirep, a minute or two can make a huge difference.

Light aircraft seem to have fewer encounters with wind shear for two reasons: They are able to adapt to changing wind conditions more rapidly because there is less inertia to overcome than with a heavier aircraft; and they probably don't spend as much time as do airliners near potentially severe thunderstorms.

The Beech A36 Bonanza pilot was under self-induced pressure to start on a July vacation. The passenger manifest included his three children and a friend. At 5:37 a.m. the pilot called Wichita Flight Service, requesting information for an IFR flight from Wichita's Colonel James Jabara Airport to Colorado Springs, Colorado. There was an approaching cold front, with most local stations reporting good VFR. But thunderstorms were active.

According to the flight service specialist: "There is a severe thunderstorm watch out north of your route for north central Kansas until 10 o'clock this morning. The front is lying east of Grand Island through western Kansas. Right now there is a pretty severe line of thunderstorm activity ahead of the front. There is a convective sigmet running from Pawnee, Nebraska, to Hays, Kansas, to 40 [miles] east of Garden City. A line of thunderstorms 20 miles wide, moving east at 30 knots, tops to 45,000. Looking at my live radar...it looks like you should be able to slide around the southern edge, south of Dodge City, to get around that area. Then once past, there is nothing on the radar all the way to Colorado."

The specialist then provided sequence reports, forecasts, and winds aloft. The pilot filed an IFR flight plan for a proposed departure time of 1400Z (9 a.m. local). It looked like a reasonable trip with some minor deviations needed to get around the convective activity.

By 8:30 a.m. the thunderstorms were approaching east Wichita and Jabara Airport. There were several witnesses who were pilots or line personnel for the FBO. The Bonanza took off on Runway 18 and had just become airborne when a strong northerly gust front arrived at the airport. The aircraft reached a maximum altitude of 50 to 75 feet, pitching and banking severely.

One of the pilot witnesses said the nose pitched up and the left wing dropped as if the airplane were going to enter a spin. The engine sounded as though it was producing full power, and the aircraft descended to the ground without spinning, left wing low. There were four fatalities and one serious injury.

National Weather Service Doppler radar measured a level-six cell and changes in wind magnitude of nearly 90 knots in the vicinity of the airport. At 7:50 a.m., Jabara was reporting winds of 120 degrees at 14 knots. At 8:35, just a few minutes after the accident, the winds were from 320 degrees at 51, gusting to 58 knots.

With full fuel, five passengers, and more than 300 pounds of camping gear on board, the Bonanza was 184 pounds above maximum takeoff weight and the center of gravity was just slightly behind the aft limit. This translated into marginal climb performance and potential instability, even under the best of circumstances. Being overtaken from behind by a 60-knot gust of wind was the coup de grace.

The pilot was reputed to be cautious, with nearly 1,000 hours in the Bonanza. He had attended safety seminars that included information on the hazards of thunderstorms. One of the witnesses stated, "I could not believe that anyone could be trying to take off at that moment." The only rational explanation is that frequently the weather looks good away from the storm and there is a temptation to make a run for it. According to the radar plot, the flight could have proceeded south and been out of danger in a few minutes. However, that action needed to be taken before the storm got within 20 miles of the airport.

The common lesson in both of these accidents is that patience is needed when dealing with thunderstorms. A cell on or near the airport — not the wishful thinking of pilots in a hurry — controls arrivals and departures. In most cases, 30 minutes will make the difference between an uneventful ride and a potentially fatal one.


See also the index of "Safety Pilot" articles, organized by subject. Bruce Landsberg is executive director of the AOPA Air Safety Foundation.