The Terminal Doppler Weather Radar, the Delta One Ninety-One Microburst, and the Wind Shear Alert System Built After the Worst Final Approach of the Decade
The Terminal Doppler Weather Radar system was built after Delta Flight 191 killed 137 people at DFW in 1985 - here's how it works and what pilots need to know.
The wind shear alert you hear on final approach is the direct result of one of the worst accidents in American aviation history. On August 2, 1985, Delta Flight 191 - an L-1011 TriStar with 163 people on board - encountered a microburst on long final for Runway 17 Left at Dallas/Fort Worth International Airport and struck the ground short of the runway. 137 people were killed. The technology built in response, the Terminal Doppler Weather Radar (TDWR), has since prevented dozens of similar accidents.
What a Microburst Does to an Airplane on Final Approach
A microburst is a small but intense downdraft. When the descending air column hits the ground, it spreads outward in all directions. For a crew on approach, the encounter unfolds in a specific, dangerous sequence.
First, entering the leading edge of the outflow produces an airspeed increase of roughly 10 to 15 knots. The airplane feels normal - good, even. The crew pitches up slightly, reduces power, and corrects back toward the glideslope. Then the aircraft crosses the core of the downdraft and encounters a severe sink rate. Exiting the trailing edge, the airspeed drops dramatically as a tailwind pushes from behind.
The entire energy exchange - headwind increase to severe sink to tailwind loss - can occur in approximately 20 seconds at typical approach speeds. The first warning sign is not turbulence. It is a brief, false sense that the approach was fine.
The Accidents That Preceded Delta 191
The FAA and NTSB had been tracking this problem for a decade before 1985.
Eastern Airlines Flight 66 crashed at Kennedy Airport in June 1975, killing 113 people on approach in a wind shear encounter. The accident investigation produced recommendations and funded further research. Pan Am Flight 759 went down at New Orleans International in July 1982, killing 153 people just after takeoff in the same basic phenomenon.
Both accidents generated recommendations. Neither generated a working detection system. The core problem was that the radar infrastructure at airports was not designed to detect wind shear.
Why Standard Radar Cannot Detect a Microburst
Airport surveillance radars are designed to track aircraft positions. They are not Doppler-capable and cannot measure the velocity of objects moving toward or away from the antenna.
The older National Weather Service weather radars could detect precipitation intensity - they could tell you a storm was present. They could not tell you that the wind inside that storm at the surface was diverging at 40 knots across a radius of half a mile. That distinction is everything when it comes to microburst detection.
How the Low Level Wind Shear Alert System Worked - and Where It Failed
After the Eastern 66 accident, the FAA developed the Low Level Wind Shear Alert System (LLWAS): a network of anemometers positioned around an airport at defined intervals. If a southeast sensor reads wind from the south at 15 knots while a northwest sensor reads wind from the north at 20 knots, the system recognizes a divergence and automatically sends an alert to the controller.
LLWAS works, but it has a critical limitation. It only measures conditions at the surface, at sensor locations. It cannot detect a microburst until the outflow actually reaches a ground sensor. By that point, an aircraft may already be on short final. The window between warning and impact can be measured in seconds.
How Doppler Radar Measures Wind Velocity
Doppler radar uses the same physics as a police speed gun. When a radar pulse returns from a raindrop, the frequency of that return shifts slightly depending on whether the raindrop is moving toward or away from the antenna. That Doppler shift is directly proportional to velocity.
By measuring thousands of these returns simultaneously, a Doppler radar produces a velocity field - an actual map of wind direction and speed throughout the scanned volume. This is fundamentally different from standard precipitation radar. The green, yellow, and red blobs on a weather display measure the power of the return signal (precipitation intensity). Doppler adds a phase measurement, tracking how the returned signal’s phase shifts between consecutive pulses to encode velocity. Same antenna principle, entirely different physical quantity being measured: motion, not mass.
How Delta 191 Accelerated the TDWR Program
The Terminal Doppler Weather Radar program had been in development since the late 1970s, progressing through the typical bureaucratic cycle. Delta 191 ended that pace. Congressional attention and sustained NTSB pressure accelerated the program into full deployment.
TDWR operates at C-band frequency, approximately 5.6 gigahertz. That frequency was chosen deliberately. Higher frequencies offer better resolution but attenuate heavily in rain - the water absorbs the signal. The 5.6 GHz band is an engineering compromise optimized for high-moisture convective conditions: sensitive enough for the detail required, far enough down the spectrum to maintain useful range in heavy precipitation.
What Makes TDWR Different from Standard Weather Radar
The antenna completes a full volume scan in approximately 60 seconds. The WSR-88D NEXRAD radars operated by the National Weather Service cycle every four to six minutes depending on operating mode.
That difference is operationally significant. A microburst can develop and reach peak intensity in roughly two minutes. At a four-minute scan interval, you may catch it on the way in, at peak, or miss the most dangerous phase entirely. At 60 seconds, the system has a reasonable chance of detecting the signature as it forms and issuing a warning before the first aircraft on approach enters the outflow.
How the Alert System Works in Practice
The processing algorithms are where the core engineering lives. TDWR uses automated detection algorithms tuned to specific signatures. A microburst signature is a Doppler velocity divergence of a defined magnitude within a specific altitude band and horizontal extent.
When the algorithm recognizes that pattern, it automatically computes the wind shear value in knots of airspeed change per nautical mile along the approach corridor and generates an alert - no human evaluation required. That alert transmits directly to the air traffic control tower.
When you hear “Wind shear alert, loss of twenty knots, three-mile final, Runway One Seven Left” on approach, that is the TDWR algorithm speaking through the controller. The knot value is the computed airspeed change you would experience flying through that wind shear on your specific corridor.
45 TDWRs were ultimately installed at the busiest airports across the United States, with priority given to major hub airports in convective weather environments: Orlando, Dallas/Fort Worth, Chicago, Denver, Miami, and Atlanta.
What the Accident Record Shows
Between 1969 and 1992, wind shear events at American airports caused or contributed to roughly two dozen fatal accidents or significant hull losses. After TDWR became operational across the major hub network, that number dropped dramatically. The technology did exactly what it was designed to do.
What Pilots Need to Understand About TDWR Alerts
The alert is a computational estimate, not a measurement. The algorithm calculates the wind shear value based on radar data from the most recent scan, projected onto your approach corridor. Conditions can evolve between the scan that generated the alert and the moment you reach that segment of the approach.
TDWR is optimized for the low-altitude terminal environment. The system monitors out to roughly 47 nautical miles but is specifically designed for the approach and departure corridors close to the airport. It is not a general weather surveillance radar and does not cover you transiting at altitude through a cell 30 miles from the field.
The absence of a TDWR alert is not clearance to continue an approach into a developing cell. TDWR is excellent at detecting the precipitation-laden microbursts common in high-moisture convective environments. It is less reliable in one specific scenario.
The Dry Microburst Problem
A dry microburst occurs when precipitation forms aloft but evaporates completely before reaching the surface. The cooling from evaporation drives a downdraft just as powerfully as a wet microburst. But because no precipitation reaches radar-detectable altitudes near the surface, the TDWR return is weak or absent - the system may not generate an alert.
Dry microbursts are most common in high-desert environments where the air between cloud base and the surface is very dry. Denver and the Front Range are the classic example, where summer dry microbursts can go underdetected by radar alone.
The FAA addressed this partially by integrating LLWAS with TDWR at airports most exposed to dry microburst risk. When the surface sensor network detects a wind shear signature the radar did not predict, the systems cross-reference and the alert fires regardless. The combined system is more robust than either operating alone.
What Comes Next: Phased-Array Radar
The FAA and NOAA have been researching phased-array radar as a potential TDWR replacement. Phased-array antennas steer the radar beam electronically rather than through physical antenna rotation, reducing scan times from 60 seconds to under 10 seconds.
At that update rate, the evolution of a microburst becomes visible nearly in real time. The National Severe Storms Laboratory in Norman, Oklahoma has operated a phased-array research radar for years, and the velocity data quality is significantly better than current rotating-antenna systems produce. Whether the FAA can fund a wholesale network replacement is an open question. The technology itself is ready.
Why This Matters for Pilots
The chain that produced the wind shear alert runs from Eastern 66 in 1975 through Delta 191 in 1985 through years of algorithm development and hardware engineering. When you hear a wind shear alert on the radio on a hot summer afternoon in Atlanta or Dallas, you are hearing the accumulated lesson of those accidents.
The alert is a warning. It deserves to be treated that way.
Key Takeaways
- Delta Flight 191 at DFW on August 2, 1985 killed 137 people in a microburst encounter and directly accelerated the TDWR program into full deployment
- TDWR detects microbursts by measuring Doppler velocity divergence - wind motion, not precipitation intensity - and automatically issues quantified alerts to tower controllers
- The system scans in ~60 seconds, fast enough to detect a developing microburst before the outflow reaches the surface
- 45 TDWRs are installed at major U.S. hub airports; the system is optimized for low-altitude approach and departure corridors, not en-route weather
- Dry microbursts in high-desert environments like Denver can go underdetected by radar; LLWAS surface sensor integration provides a partial mitigation
Radio Hangar. Aviation talk, built by pilots. Listen live | More articles