Unusual Attitude Safety · Case studies

How it actually kills you

Fifteen NTSB accidents, every one read in the original report. What they have in common is not bad luck. It is a small number of mechanisms that repeat, and a set of numbers almost nobody is taught.

The number that matters most

40 miles

Required separation between echoes before you fly between them. FAA Aviation Weather Handbook §22.8.2 item 14: avoid by at least 20 miles any thunderstorm giving an intense or extreme radar echo, and "such echoes should be separated by at least 40 mi before flying between echoes."

Forty miles is roughly Orlando to Daytona. Look at the gap on your tablet and ask whether it is that wide. Almost none of them are. That is the whole lesson of the sucker hole, and it is a number, not a judgement call.

ERA09LA3921 fatal
He flew between two bands, then turned into one

Sanderson FL, 8 July 2009, 09:05 local. Vans RV-7A, N774US. Two bands of thunderstorms across north and central Florida. The radar track shows him approaching, turning northeast to run parallel to the band, passing between the two weather areas, then turning back northwest and penetrating. The vertical stabiliser and rudder separated and were never found. Spar fractures showed ductile overload.

Probable cause: continued visual flight into IMC with convective activity. Contributing: failure to obtain an adequate weather briefing.

Your tablet is older than it says it is

NTSB Safety Alert SA-017, In-Cockpit NEXRAD Mosaic Imagery:

The age indicator associated with the mosaic image on the cockpit display does not show the age of the actual weather conditions. Weather conditions depicted on the mosaic image will ALWAYS be older than the age indicated on the display. In extreme latency and mosaic-creation scenarios, the actual age of the oldest NEXRAD data in the mosaic can EXCEED the age indication in the cockpit by 15 to 20 minutes.

A thunderstorm's entire life cycle is about thirty minutes. So the picture you are navigating by can be older than half the life of the thing you are navigating around, while the display tells you it is a minute old.

The current FAA handbook has a typo that inverts this instruction. FAA-H-8083-28A §22.8.2 item 7 reads:

Use data-linked weather radar (i.e., NEXRAD) mosaic imagery as the sole means for negotiating a path through a thunderstorm area (tactical maneuvering).

The "Do not" was dropped when the list was converted from Advisory Circular 00-24C, whose item 7 reads "Don't use data-linked weather NEXRAD mosaic imagery as the sole means for negotiating a path..."

Read as printed, the handbook instructs the exact behaviour that killed the pilots in SA-017. Teach the Advisory Circular's sense. Do not paste the handbook line onto a slide.

Having radar is not the same as being safe

Airborne radar suffers attenuation: a heavy cell absorbs the signal and hides whatever is behind it. The handbook calls the result a radar shadow, and adds that airborne radar is "for avoiding severe weather, not for penetrating it" and "does not provide assurance of avoiding turbulence."

ERA09FA3895 fatal
A radar-equipped aeroplane is not an immune aeroplane

Gulf of Mexico, 25 nm northwest of Port Richey FL, 8 July 2009, 14:52 local. Cessna 421C, N4467D. He had onboard weather radar, weather from Jacksonville Center, and PIREPs. At 14:46 he reported turbulence and asked how to get out; the controller said continue straight ahead about two minutes and he should be clear. He reported 2,000 fpm downdrafts, requested a reversal, declared an emergency, and said the aeroplane was upside down. Two debris fields. The storm had a defined overshooting top.

NTSB: near the heavier echoes the airborne radar "may have been unable to provide an accurate representation of the radar echoes along the aircraft's flight path; therefore the final penetration of the intense portion of the storm was likely unintentional."

The cell that forms behind you

A thunderstorm's full life cycle is about thirty minutes and updrafts can exceed 3,000 fpm. So the hazard that traps you on the way home did not exist when you looked at the weather on the way out. This is the mechanism that catches careful pilots, and the record shows it catching one who did everything right.

ERA11GA0663 fatal
He turned back. It was the correct decision. He died anyway.

Avon Park FL, 17 November 2010, 20:53 local. Cessna M337B, N1309. The briefed weather contained no showers or thunderstorms. Convection developed over south-central Florida during the mission. He discontinued and turned back — the right call, made in time — and on the return leg met the edge of an echo whose core reflectivity reached 55 dBZ. The right wing separated; every fracture surface showed overstress. No onboard weather radar.

Probable cause: inadvertent encounter with an unexpected intense rain shower with severe turbulence at night. He touched the edge, not the core.

ATC is not a weather-avoidance service

NTSB Safety Alert SA-011, Thunderstorm Encounters, uses four accidents as examples. Two of the four are Florida. Controllers have other duties, their radar is optimised for aircraft rather than weather, and intensity information is generally provided when asked for rather than volunteered.

DEN06MA1191 fatal
"No adverse weather along the route"

Argyle FL, 1 September 2006, 11:15 local. Mitsubishi MU-2B-35, N6569L. The preflight briefer noted no adverse conditions. Two hours later, descending toward Panama City, he asked to "get down lower so we can get underneath this stuff." The left wing separated inboard of the engine; both spars failed in catastrophic static up-bending overstress. Local authorities said the storm "materialized very quickly."

Probable cause names both halves: the pilot's inadvertent flight into thunderstorm activity, and the failure of air traffic control to use available radar information to warn him.

Being told, twice, and continuing

ERA22LA4042 minor · everyone lived
Central Florida, 3 PM, September

Saint Cloud FL, 9 September 2022, 15:02 local. Cirrus SF50 Vision Jet, N77VJ. He took multiple preflight briefings and delayed departure into the afternoon. ATC advised him twice of heavy to extreme precipitation on the final approach course. He continued the RNAV approach in VMC, entered IMC near the final approach fix, and hit extreme precipitation and turbulence. The aeroplane then climbed at a rate "well beyond the performance capability of the airplane, likely caused by updrafts from the mature thunderstorm." Envelope protection pitched it down, oscillation followed, and he pulled the parachute into a marsh.

Every element of the sea-breeze regime is in this one: central Florida, mid afternoon, September. The information was all available and all correct. He had more weather knowledge than most and it did not save the approach; the parachute did.

ERA16FA3092 fatal
Told to turn left. Acknowledged. Continued on course.

Hollywood FL, 3 September 2016, 09:50 local. Cessna 172, N6091E. No weather briefing. He asked to fly below 500 ft along the shoreline. The controller approved, advised heavy precipitation twelve o'clock and four miles, and told him to turn left and go three miles offshore. He acknowledged and continued on course. Three minutes later he reported reversing direction. Nothing after that. GPS track over radar shows entry into extreme intensity precipitation, then a rapid descent into the ocean.

What the fifteen have in common

MechanismWhat it means in the cockpit
Development rateThirty-minute life cycle, 3,000 fpm updrafts. Absent at briefing means nothing.
The gap that is not a gapForty miles between echoes, or it is not a gap.
NEXRAD latencyUp to 15 to 20 minutes older than the display admits.
Radar shadowThe first cell hides the second. Having radar is not immunity.
The cell behind youYour escape route is weather that has not formed yet.
ATC silenceNot a weather-avoidance service. Ask, and ask for intensity.

On sources. Every accident above was verified by reading the NTSB final report, not by remembering it. Report numbers, dates, registrations and probable-cause wording come from the original documents. Where the researcher could not verify something, it is not on this page.

These reports are public because the NTSB intends them to change what the next pilot does. That is the only reason they are here. There are no wreckage photographs on this site, not out of squeamishness but because the wreckage is the outcome and the lesson is in the decisions.