USAir Flight 1016: Crash, NTSB Findings, and Aftermath

USAir Flight 1016 was a scheduled flight from Columbia, South Carolina, to Charlotte, North Carolina, that crashed on July 2, 1994, after flying into a microburst on final approach to Charlotte-Douglas International Airport. The McDonnell Douglas DC-9 struck trees and slammed into a residential neighborhood just west of the airport, killing 37 of the 57 people on board. The investigation that followed exposed serious gaps in wind shear detection, air traffic control communication, and pilot training, and it remains the last wind shear crash in U.S. commercial aviation.¹

The Flight Into the Storm

Flight 1016 pushed back from Columbia Metropolitan Airport at 6:10 p.m. Eastern time and was airborne by 6:23 p.m. Captain Mike Greenlee and First Officer Phil Hayes were joined by flight attendants Richard DeMary, Shelley Markwith, and Karen Forcht.² As the aircraft descended toward Charlotte, a thunderstorm was developing rapidly at the approach end of Runway 18R. By 6:40 p.m., ceilings were overcast at 4,500 feet, visibility had dropped to one mile, and heavy rain was falling.³

The storm was producing an unusually violent microburst, a column of cold air plunging downward. NASA later characterized it as a small-diameter, rapidly intensifying event with peak radar reflectivity of 65 dBZ and surface rainfall above four inches per hour. Its wind shear hazard factor peaked near 0.3, roughly three times the FAA’s alert threshold and a level rarely observed in microburst events.⁴

At 6:39 p.m., the captain noted the rain was “sitting right on” the airport. Tower controllers issued wind reports and a wind shear alert for the airport’s northeast boundary, but the alert went out on a frequency the crew of Flight 1016 was not monitoring, because tower controllers for the two parallel runways worked separate frequencies.⁵

At 6:41:58 p.m., as the plane entered the rain shaft at roughly 200 feet above the ground, the first officer called out a ten-knot airspeed increase. Seconds later the captain saw a twenty-knot gain, the classic initial headwind signature of a microburst, and called for a go-around: “Okay, you’re plus 20 … take it around, go to the right.” At 6:42:16 the captain radioed the tower: “USAir 1016’s on the go,” and called for maximum power. Within fourteen seconds, the aircraft went from a 35-knot headwind to a 26-knot tailwind, a 61-knot swing, as it passed through the core of the microburst.⁶

The Ground Proximity Warning System sounded. The flight data recorder showed the engines spooling up, but by 6:42:32 the stall warning activated. At 6:42:35 the DC-9 struck the ground.⁵

Impact and Casualties

The aircraft broke through the airport perimeter fence and into a stand of trees near Old Dowd Road and Wallace Neel Road, just west of the airport, breaking into three major sections and igniting a post-impact fire.⁷ The aft cabin and tail section struck the carport of a house at 4228 Wallace Neel Road.⁸

Of the 57 people on board, 37 passengers were killed. Two flight attendants and 14 passengers were seriously injured. The captain, the first officer, one flight attendant, and one passenger received minor injuries.³

What the NTSB Found

The NTSB adopted its final report, NTSB/AAR-95/03, on April 4, 1996. The board identified four probable causes: the crew’s decision to continue the approach into severe convective activity conducive to a microburst; late recognition of the wind shear; failure during the attempted go-around to establish and maintain the pitch attitude and thrust needed to climb out of the downdraft; and inadequate real-time wind shear and adverse weather information from air traffic control.³

Air Traffic Control Breakdowns

The approach controller failed to relay ASR-9 radar data showing a strong precipitation cell near the airport, instead telling the crew they “may get some rain just south of the field,” language the NTSB called improper and non-standard. The tower supervisor did not ensure controllers were reporting wind shear alerts that had fired in multiple quadrants. Visibility had dropped to one mile, but no one passed that to the crew, and the runway visual range equipment was never activated. Most critically, the wind shear alert broadcast by the controller for the parallel runway went out on a frequency Flight 1016 never heard.⁵

Detection Failures on Board and on the Ground

The DC-9’s Honeywell Standard Windshear Detection System was reactive by design, alerting only after severe wind shear had already been detected. Its software included a delay meant to suppress nuisance warnings while the wing flaps were in transition. On the evening of the crash, that delay meant the system issued its warning only three to four seconds before impact, roughly twelve seconds later than it would have triggered without it.⁹

Charlotte’s Low-Level Windshear Alert System had its own problems. Internal FAA communications from 1993 had flagged inaccurate reporting of wind conditions, but no upgrades were made before the crash. Post-accident evaluation found that two sensors, to the northeast and northwest, were sheltered by obstacles that degraded their performance. An MIT Lincoln Laboratory engineer testified that about a minute before the crash, the northwest sensor missed the alarm threshold by just 0.7 knots, a gap attributable to tolerances built in to prevent false alerts.⁵

A Training Mismatch

The microburst at Charlotte arrived without the turbulence pilots had been trained to expect. Standard simulator scenarios paired wind shear with significant turbulence and abrupt airspeed swings. Flight 1016’s crew instead encountered an initial performance boost from the headwind, then a smooth, devastating transition to a tailwind, with no turbulence to cue them. The board concluded that repetitive simulator training may have conditioned pilots to look for the wrong warning signs.¹⁰

What Changed After the Crash

The NTSB’s recommendations, numbered A-95-40 through A-95-51, targeted three areas. Controllers were to display and issue all levels of ASR-9 precipitation data to flight crews, update ATIS immediately when wind shear, lightning, or rain was observed, and continue issuing advisories directly to pilots until acknowledged; blanket broadcasts in tower cabs were to be replaced by individual acknowledgment from each controller. The FAA was told to review all Low-Level Windshear Alert System installations nationwide and fix siting deficiencies. And simulator training was to include wind shear scenarios without prior turbulence, with a dedicated wind shear escape maneuver replacing the standard missed approach when the encounter happens below 1,000 feet in microburst-prone conditions.¹⁰

In November 1994, the board had already issued urgent recommendations on the Honeywell wind shear system, asking the FAA to warn operators about the flap-transition delay, review the system’s certification, require modifications for prompt warnings during flap changes, and amend the governing technical standard (TSO C117) to require testing with flaps in transition.⁹

The most consequential change was already in motion. Charlotte-Douglas did not have Terminal Doppler Weather Radar operational on the night of the crash. TDWR, developed by the FAA with MIT Lincoln Laboratory, provides automated microburst and gust front detection with enough lead time for controllers to warn pilots.¹¹ By December 1995, the FAA had completed TDWR installation at the 45 largest airports in the United States.¹ The FAA says that no wind shear accidents have occurred at any TDWR-protected airport since those systems became operational. NASA simulations conducted after the Charlotte crash suggested a forward-looking airborne wind shear radar could have detected the microburst and issued a warning roughly 60 seconds before impact.⁴

Microbursts had been killing airline passengers for two decades. The phenomenon was defined after Eastern Airlines Flight 66 crashed at JFK on June 24, 1975, killing 113 of 124.¹² Pan Am Flight 759 crashed in Kenner, Louisiana, in 1982, killing 145 aboard and 8 on the ground.¹³ Delta Air Lines Flight 191 crashed at Dallas-Fort Worth on August 2, 1985, killing 137, and became the primary catalyst for TDWR. Flight 1016 was the last entry in that sequence. TDWR deployment, predictive airborne wind shear systems, better engine performance on newer aircraft, and revised training and ATC procedures have together kept another U.S. commercial wind shear crash from happening since.¹¹

Litigation

Families of the victims pursued claims against USAir. Aviation attorney Jamie Lebovitz represented multiple families and secured confidential settlements; the amounts have not been publicly disclosed.¹⁴ The central factual issue tracked the NTSB findings: a tower controller had issued a wind shear warning, but it went out on a frequency the flight was not monitoring, a systemic failure beyond any single pilot’s decisions.

The board’s emphasis on pilot error drew criticism. John Cox, who served on the NTSB’s Air Traffic Control Group during the investigation and later became executive air safety chairman for the Air Line Pilots Association, said the captain experienced significant emotional distress and felt deep personal responsibility for the crash.¹ Others argued the microburst’s intensity exceeded the performance capabilities of the DC-9-31, and that the cascade of ATC and detection failures left the crew with almost no warning of what they were flying into.³

The Memorial

For years, members of the Charlotte community left flowers, poems, and small wooden crosses at the crash site. A granite memorial was later established at the Charlotte Douglas Airport Overlook after a headstone was relocated from the original site, which had become inaccessible.¹ The design represents the Morse code for CAE and CLT, the airport codes for Columbia and Charlotte. Its plaque honors those who died, those who were injured, and those who helped in the rescue.⁷

1
Forbes. USAir Flight 1016 Crashed 30 Years Ago. It Was the Last Wind Shear Crash
2
APFA. 27 Years Later: In Memory of USAir Flight 1016
3
NTSB. Aircraft Accident Report NTSB/AAR-95/03
4
NASA. NASA Technical Report on USAir 1016 Microburst
5
Flight Safety Foundation. Accident Prevention – USAir 1016
6
Aviation Safety Network. USAir Flight 1016 Accident Description
7
Axios Charlotte. USAir Flight 1016 Crash Charlotte
8
NTSB. NTSB Docket – DCA94MA065
9
NTSB. NTSB Safety Recommendations A-94-208 Through A-94-210
10
NTSB. NTSB Safety Recommendations A-95-40 Through A-95-51
11
FAA. Terminal Doppler Weather Radar
12
FAA. FAA Lessons Learned – Eastern Airlines Flight 66
13
64 Parishes. Pan Am Flight 759 Crash
14
NPHM. Airline Pays Confidential Settlements to Families of USAir Flight 1016