Lesson 6 -- Thunderstorms and Aviation Accidents

If there was ever a lesson that you should be reading, it is this one. Here we discuss several accidents that have happened to other pilots. It could happen to me. It could happen to you. The whole purpose of this course has been to keep us off this page. Please pay attention. Please learn the lessons from each of the accidents we present in this, the final chapter and end of the National Weather Association's "Thunderstorms and Flying" internet course in the public domain.
It is my privilege to introduce our "guest speaker" for this lesson. He is Major General Timothy Peppe, the Air Force Chief of Safety and Commander of the Air Force Safety Center.

It’s easy to be on the ground watching a passing thunderstorm and be awed by the display of lightning, the strong winds that appeared with no warning, and the invariably accompanying downpour. Providing there was is no associated loss of life or damage to property, most people enjoy watching a thunderstorm. There is nothing enjoyable about encountering a thunderstorm while in flight. The awesome lightning display observed from the ground becomes blinding and disorientating in the air. Unless taking unnecessary risks, the ground observer at worst will usually only lose television reception or home electrical service from a lightning strike. Aircraft flying in the vicinity of thunderstorms attract electricity - lightning strikes are common and can destroy electronic devices such as navigation equipment or radios and can also explode fuel tanks. Turbulence from winds associated with thunderstorms has killed passengers and aircrew who were thrown about the cabin. Turbulence can also cause an aircraft to exceed its structural limits and literally rip off the wings. Wind shear from thunderstorms has plummeted aircraft into the ground, causing some of the worst mishaps in the history of aviation. The rain associated with thunderstorms has been linked to aircraft accidents; the shear volume of water can stall a turbojet engine, denying the pilots the power required to counter wind shear. Every year the USAF suffers aircraft damage in flight and on the ground from hail.

Universally, military flight training stresses the danger posed by thunderstorms, not only due to cloud penetration, but also from the associated phenomena. The following is an excerpt from the USAF Weather Handbook: “WARNING: When you fly through a thunderstorm, the hazards that face you are extreme. You will be betting the aircraft, your life, and the lives of your crewmembers on the forces of nature. This must be the only remaining alternative!”

In the past, lack of weather forecasting capabilities and autonomous detection systems such as on-board radar often committed aircraft to routings where fuel limitations forced them into, or too close to, cumulonimbus cloud formations. Fortunately, technology has refined our ability to forecast and detect severe weather. Being surprised by severe weather is a rarity, however, upon occasion aircraft still encounter thunderstorms imbedded in other cloud formations. This usually occurs because the aircraft are operating in a part of the world where forecasting is not as advanced as that to which we are accustomed, and because the aircraft are not equipped with weather radar.

The most common reason for encounters between USAF aircraft and thunderstorms is poor decision-making by the people either flying or controlling the operation of the aircraft. These decisions range from a commander deciding to authorize the local daily flying program with thunderstorms in the vicinity of the airfield, to a pilot deciding not to turn on or monitor the weather radar, even though the meteorological briefing forecast cumulonimbus activity along the route of flight. Frequently, these faulty decisions resulted from lack of understanding of the indirect dangers associated with cumulonimbus clouds. Hail has severely damaged USAF aircraft over 20 miles from the source thunderstorm. Landing mishaps have occurred because of gust fronts crossing an aerodrome, also more than 20 miles from the source thunderstorm.

The importance of our nation’s military cannot be stressed enough and the potential loss of capability due to encounters with severe weather could easily affect national security. Aviators and anyone in the business of controlling aircraft movements need to be aware of all of the dangers associated with thunderstorms. The best way to learn is normally through direct experience, but in the case of severe weather the preferred way to learn is through the experiences of others, which is exactly the purpose of lesson plan 6.

Maj. Gen. Timothy A. Peppe
Chief of Safety
Headquarters, United States Air Force

Are you interested in studying aircraft accidents? You can go to the National Transportation Safety Board's website and roam the aviation section. It's not a pretty place. If you go to the airplane section and search the data base of accidents, you would find a number with the keyword "thunderstorm."

One member of the National Weather Association's Aviation Weather Committee has spent lots of time with that database. Here are his observations...

It has been said that those who do not learn the lessons of history are doomed to repeat it. All too often aircraft accident investigation sometimes brings this message to light as the same errors continue to be made. Weather related accidents continue to account for approximately 25 percent of the air carrier and general aviation accidents.
Better aircraft design and engineering has made aircraft safer with improved performance and permitting operations in all kinds of weather that years ago would have been impossible. Aircraft today are also able to fly higher and operate over most weather systems. Avionics improvements have made many aircraft able to operate into almost all kinds of weather conditions, and to navigate point-to-point in near zero-zero conditions. However, there is no true all-weather aircraft when severe weather is considered. Every pilot and aircraft has weather limitations.

Everything we do in aviation is influenced by the weather. To determine whether we go or not-go, or to select a route and altitude should be based upon weather considerations. We make a runway selection to take advantage of favorable winds. Our destination minimums, the approach we plan to fly, and the selection of an alternate should be based on weather decisions. Avoiding weather hazards is critical for safe operation and passenger comfort.
Every spring and summer brings some of the most hazardous weather conditions in the form of thunderstorms. Unfortunately, a lack of planning, limited training, and poor judgment will likely cause several thunderstorm related accidents to occur. Thunderstorms can produce the full spectrum of aviation hazards including convectively induced turbulence (CIT), lightning, icing, hail, low ceilings and visibilities, altimeter errors, engine water ingestion, tornadoes and wind shear/microburst. Any thunderstorm in the mature state can produce these hazards and must be avoided when at all possible. A stronger environmental wind shear and instability produces a stronger updraft and an increased threat of severe weather.

In addition, thunderstorms also cause additional problems to the National Airspace System. Depending on it’s organization and severity, there will be delays and cancellations as thunderstorms block routes, saturate air traffic control sectors, and impact terminal arrival rates.

In civil aviation, the lack of a pre-flight briefing often becomes one of the initial chain of events that begins the accident sequence. Even when a briefing is obtained pilots fail to properly plan for things like appropriate alternate routes in case the flight can not be completed as planned. Sometimes they they continue VFR operations into IMC conditions. Finally, if convective activity was predicted, we try to find out if it was avoidable and what action did the flight crew perform to minimize the impact from the weather.

Professional pilots often have flight dispatch or flight following services to help the pilot-in-command with the weather, performance, and planning activity. However, the pilot in command must still utilize the information, know where and when to ask additional questions, and review the provided data. As weather changes, they must know that regular updates are required to remain knowledgeable about the environmental factors affecting the flight. Failure to obtain the most current information, failure to have an alternative plan of action, failure to deviate around weather, and accepting a route or approach into the weather can put the flight into a dangerous position and threaten safety.
When obtaining a brief flight briefing, any mention of severe thunderstorms from the Convective Outlook (AC), Convective SIGMET (WST), or Severe Weather Forecast Alerts (AWW) should alert you that avoidance of any convective activity is a priority to ensure safety. Watch the Terminal Aerodrome Forecast (TAF) for amendments, for thunderstorms with high winds associated with them, and for reduced visibility in heavy rain. If its forecast, you have been warned and continuing into the terminal area with a thunderstorm approaching or impacting is now a potential safety-of-flight issue.

Recent accidents show the destructive power of thunderstorms. They also show how critical are a pilot’s knowledge of his environment, the correct use of resources available on the ground and in-flight, and a knowledge of aircraft and personal limitations. Thunderstorms are powerful entities. They will continue to impact flight safety.
In an earlier lesson, you were introduced to Mr. Terry Lankford, a co-chair of the NWA Aviation Weather Committee, a retired FSS specialist, and author. For this course, Mr. Lankford has offered us a look at Chapter 6 of a new book that will be published soon by McGraw-Hill Publishers. McGraw-Hill does own the copyright to this material. The book looks at aircraft accidents in a variety of weather conditions and is a wonderful addition to the course. (As course director, "cd", I've made a few comments to help tie the material to what we've learned.)


Thunderstorms contain just about every weather hazard known to aviation—turbulence, icing, precipitation (including hail), lightning, tornadoes, gusty surface winds, low-level wind shear (LLWS), effects on the altimeter, and low ceilings and visibilities.

Convective low-level wind shear (LLWS)

Case Study

In August 1985 a Delta Air Lines L-1011 crashed at the Dallas/Fort Worth Airport. The NTSB was unable to determine if the crew had been using airborne weather radar at the time of the crash. The NTSB report did state however: "The evidence concerning the use of the airborne weather radar at close range was contradictory. Testimony was offered that the airborne weather radar was not useful at low altitudes and in close proximity to a weather cell...;" although, "At least three airplanes scanned the storm at very close range near the time of the accident."

The accident was probably caused by a microburst from a single severe storm cell, which illustrates how weather can develop rapidly, often without any severe weather warning.

Pilots should avoid takeoff or landing when a thunderstorm is within 10 to 20 miles of the airport. This is the region of the strongest and most variable winds. Caution must also be exercised following thunderstorm passage. A strong, gusty outflow boundary can follow the storm. Along with these winds are downbursts and microbursts which produce severe low-level wind shear.

The FAA, along with a group of aviation specialists, has developed AC 00-54 Pilot Windshear Guide. Although, primarily for the airlines, much of the information can be applied to general aviation. The Department of Commerce publication, Microbursts: A Handbook for Visual Identification is for sale by the Superintendent of Documents It contains an in-depth, technical explanation of the phenomena along with numerous color photographs depicting microburst activity. Avoidance is the best defense against a microburst encounter.

Weather radar

Although an oversimplification, radar displays an image dependent on reflected energy or back scatter. Intensity depends on several factors; among them are particle or droplet size, shape, composition, and quantity. NEXRAD radars are capable of displaying both precipitation as well as cloud size particles. Note that radar detects precipitation, not turbulence.

Pilots have access to NWS radar and airborne weather radar, and to a limited extent ATC radar. The FAA is currently providing Air Traffic Controllers with either separate or overlay NEXRAD products. Each system has a specific purpose and its own application and limitations.

NWS radars are ideal for detecting precipitation-size particles. NWS radars can detect targets up to 250 nautical miles (nm), however, due to range and beam resolution, which is the ability of the radar to distinguish individual targets at different ranges and azimuth, an effective range of 125 nm is used.

NEXRAD has been a quantum leap in providing early warning of severe weather. Since NEXRAD is a Doppler radar it detects the relative velocity of precipitation within a storm. It has increased the accuracy of severe thunderstorm and tornado warnings, and has the capability of detecting wind shear.

Airborne weather radars are low power, generally with a wave length of three centimeters. Precipitation attenuation, which is directly related to wave length and power, can be a significant factor. Precipitation attenuation results from radar energy being absorbed and scattered by close targets and the display becomes unreliable in close proximity to heavy rain or hail. Intensity might be greater than displayed, with distant targets obscured. An accumulation of ice on the aircraft's radome causes additional distortion.

Case Study

According to the National Transportation Safety Board (NTSB), precipitation attenuation was a contributing factor in the crashes of a Southern Airways DC-9 in 1977 and an Air Wisconsin Metroliner in 1980. Precipitation attenuation is not significant with NWS 10-cm high-power units, however, it can be a serious problem with units of five centimeters or less, especially in heavy rain. The NTSB recommends: " the terminal area, comparison of ground returns to weather echoes is a useful technique to identify when attenuation is occurring. Tilt the antenna down and observe ground returns around the radar echo. With very heavy intervening rain, ground returns behind the echo will not be present. This area lacking ground returns is referred to as a shadow and may indicate a larger area of precipitation than is shown on the indicator. Areas of shadowing should be avoided."
When using an airborne weather radar it is imperative to understand the particular unit, its operational characteristics, and limitations. "Just reading through the brochure that comes with the equipment is certainly not enough to prepare a pilot to translate the complex symbology presented on the [airborne] scope into reliable data. A training course with appropriate instructors and simulators is strongly recommended," according to the March-April 1987 FAA Aviation News.

Case Study

I was the first officer and had just awoken from my rest break in first class when we hit severe turbulence. I had my seat belt on and was forced violently upwards into it. Everything in front of me went up into the air, scattering pillows and other service items throughout first class. Duration was maybe 10 seconds. When I got up to the cockpit, they were trying to clean up because water had spilled on some personal effects.

We contacted dispatch and started trying to determine the extent of damage to passengers and flight attendants. There was only one episode and there had been virtually no turbulence prior, nor was there any significant turbulence afterwards. Since I wasn't in the cockpit, I was unaware of what conditions we were flying in.

It was determined that the airplane flew into a thunderstorm during an overwater night operation that resulted in 10 seconds of severe turbulence. There was no "painting" of precipitation on radar, but lightning was observed along with St. Elmo's fire on the windshield.

Since radar detects precipitation, not turbulence, and thunderstorm related turbulence can extend up to 20 miles form the storm, pilots should be prepared for a significant turbulence encounter in the vicinity of any convective activity. The pilot's observation of lightning was a clue to the threat of turbulence.

Case Study

We were descending from FL290 to 15,000 ft with the autopilot on. We were deviating around a large cell at 10 o'clock. There were lower cumulus below us with tops 12,000 to 15,000 ft. We were flying south of track to avoid a large cell and miss smaller buildups. With the landing lights on we saw clouds ahead and got a short period of rain and light turbulence and one large jolt. We turned further south and broke into the clear. Four people were taken to area hospitals to be checked and released due to the one severe jolt.

The reporter went on to say: "Our radar showed nothing along our route of flight. After the encounter the controller asked how our ride was, as she had some indication of bad weather on our route." Controllers typically ask for flight conditions. The fact that a "ride report" was requested does not necessarily mean the controller was aware of hazardous weather along the aircraft's route.

Case Study

While cruising at FL330 the aircraft encountered severe turbulence and lost several hundred feet of altitude. There were injuries to two flight attendants and one passenger.

We were deviating around weather on a route suggested by ATC and our own flight dispatcher. We also agreed with their suggestions, as we had a good picture on our airborne radar. Approximately 10 minutes prior to the encounter, we had been visual, but at the actual time of the encounter we were IFR in cirrus type clouds. At the time of the encounter we were approximately 25 miles from the nearest contouring cell as depicted on our radar.

Since we were in clouds and radar showed us to be on a clear path, we can only assume we encountered a wind shear situation or flew into a rapidly developing buildup that did not contain enough moisture to give a return on our radar.

During a Callback conversation the reporter restated the fact that the turbulence was totally unexpected. The pilot stated that the color radar that was on board did not paint the smaller cells and that might have had a bearing on the incident.

Case Study

Rain showers were in the area. I received a heading from ATC for weather avoidance and was advised a previous aircraft had flown the prescribed route without incident. We had no excessive radar return and no contouring on the radar scope. Flight conditions were IFR, light rain, intermittent light chop. About 70 ml from destination we began experiencing excessive banking and pitching which I reported to ATC as severe turbulence. I requested immediate deviation and descent as the aircraft was barely controllable. This was approved. The intensity of the turbulence remained severe intermittently for a period of about seven minutes. During this period we reversed direction of flight and landed at an alternate.

The reporter went on to say that this incident occurred because aircraft and center radars are unable to detect wind direction changes associated with this type of turbulence. A possible solution would be Doppler radar. Certainly Doppler radar can detect changes in wind direction and intensity, but only if precipitation is present. Doppler radar is unable to directly detect turbulence and is of no use in clear air.

Case Study

Along our route we would encounter numerous thunderstorms and lines of cumulonimbus. We successfully navigated through the lines of cumulonimbus. We could see cumulonimbus south of us and north of our course on radar. Our radar showed several strong echoes over the arrival fix. We informed ATC and they said they could not see them. We asked for a ride report, and ATC said they had no aircraft in that area for awhile. Our radar showed a clear area south of our present position. We asked and received clearance to deviate and then pickup the rest of the arrival.

I was flying the aircraft on autopilot. I initiated a right turn. We were in the clouds and a relatively smooth flight. As the turn continued we temporarily broke out of the clouds and saw a rapidly developing cumulonimbus cloud in front of us. There was no indication of this on radar. We had excellent returns on the radar throughout the flight to this point. When I saw the cloud I immediately disengaged the autopilot and increased the angle of bank to avoid. We penetrated the side of the cloud and experience severe turbulence for about four to five seconds. The aircraft climbed rapidly, even through I applied full forward stick. This was a brief encounter.

The reporter concluded that this encounter was unavoidable since we were unable to detect it on radar. Pilots must remember the limitations on radar. If you fly in the vicinity of convective activity, be prepared for a possible encounter with severe turbulence.

Case Study

While enroute we descended from FL310 to FL220 for clouds. Deviating north of course, we rounded the corner to proceed direct. Radar showed one more area of thunderstorms for us to navigate around. Radar and visual cues showed an approximately 30 ml gap between storms with sunshine between. As we approach this area we again received clearance to deviate. As we passed north of the first area we turned left, the clearest direction with nothing on radar. As we finished our turn we were in and out of cirrus. At 12 o'clock, less than 1/2 ml was a small thunderstorm with tops estimated to FL230. With no room to turn, the captain put the ignition to override just prior to entering. The aircraft started to climb. The autopilot could not keep us and kicked off about the time of our one big jolt of turbulence. The captain was following on the controls and in a few seconds we popped out in smooth air 200 to 300 ft above assigned altitude.

The rule of thumb is 40 miles between severe storms. The problem is that it may be next to impossible to determine the severity of a storm, especially during the development stage. Like the general limitations on radar, if we fudge on the 40 mile rule, be prepared for a significant turbulence encounter.

Case Study

While on vectors with approach control we were issued a clearance to descent from 7000 to 6000 ft. The ride was rough, moderate chop to moderate turbulence and rain, with no thunderstorm activity noted on radar. Shortly after initiating the descent we encountered a "downdraft" that, while not violent, required higher than normal power settings to counter. We reported "unable to hold 6000 ft" to ATC and subsequently leveled off at 5500 ft. Approach control acknowledge and advised us to return to 6000 ft when able. We regained the altitude in about a minute.
I believe we may have passed through a squall line. However, the severity was not indicated by radar presentation, PIREP, or ATC.

The reporter states they may have passed through a squall line. However, the severity of the encounter does not appear to reflect a line of severe convective activity. It does reinforce the limitations of airborne and ATC radars, and the need for additional PIREPs on such activity.

Case Study


This incident occurred over New Orleans in thunderstorm weather. Both crew of a Cessna Citation were injured when the aircraft encountered severe turbulence in an area where their airborne weather radar indicated no precipitation.

Lightning detection systems

Lightning detection equipment, trade named Stormscope, was invented in the mid 1970s by Paul A. Ryan as a low-cost alternative to radar. Stormscope and similar lightning detectors sense and display electrical discharges in approximate range and azimuth to the aircraft. Like radar, Stormscope also has limitations. One misconception proclaims that in the absence of dots or lighted bands there are no thunderstorms. However, NASA's tests of the Stormscope differed. Precipitation intensity levels of three and occasionally four (heavy to very heavy) would be indicated on radar without activating the lightning detection system. A clear display only indicates the absence of electrical discharges. This does not necessarily mean convective activity and associated thunderstorm hazards are not present. Even tornadic storms have been found that produced very little lightning. The lack of electrical activity, as with the absence of a precipitation display on radar, does not necessarily translate into a smooth ride.
Many authorities agree that a combination of radar and Stormscope is the best thunderstorm detection system. It cannot be overemphasized that these are avoidance, not penetration devices. Thunderstorms imply severe or greater turbulence and neither radar nor lightning detection systems, at the present, directly detect turbulence.

Case Study

I was flying level with my Stormscope working, talking to approach control, when without warning from my Stormscope or the controller I experienced a sudden 800 foot increase in the altitude, very heavy rain, and heavy lightning. I at once grabbed the carb heat and the throttle. Suddenly the airplane stated to descent and I lost about 1200 ft. I added power and climbed to assigned altitude and leveled off OK. The controller didn't say anything, but this was a serious altitude excursion.

We have already discussed the limitations of ATC radar. In the next section we will specifically address this issue of ATC radar.

[A] Air traffic control (ATC) radar
ATC radar is specifically designed to detect aircraft, its wavelength reduces the intensity of detected precipitation. Additional features reduce the radar's effectiveness to see weather.

Case Study

The weather briefer gave me all available information for the flight, including thunderstorm activity in and along my original route of flight. At this point I decided to go east around the thunderstorm activity as to avoid any encounters with them. I established contact with ATC for advisories and asked for radar vectors around the thunderstorm activity. They asked if I wanted to go IFR, I chose to do so, and they issued a clearance.

Within a few minutes I was in IMC and reduced to maneuvering speed. Cruise was continued with occasional light chop. Rain was encountered along with increased turbulence. I requested a lower altitude, which was denied due to "towers in my area." Cruise was continued at present altitude and heading until I saw lightning ahead of my position, also the turbulence increased in intensity. I reported this to ATC, and they asked if I wanted to turn around. I answered affirmatively and stared a standard rate turn to the left to reverse my course. At this point the plane was hit by a violent updraft that registered at near 2000 ft/min. This was followed by a series of vertical shears, up and down. All gyro instruments were virtually unreadable. All control movement were for airspeed control. We were ejected from the cloud base in the fourth oscillation and in a nose down attitude.

I opted to make a precautionary landing rather than subject the aircraft, my passengers, or myself to any further danger. I felt at the time, and still do content, that this was the only safe option. The nose gear was held off as long as possible on landing, however, the effort was stifled by a drainage ditch at the edge of the field. Considering the alternative it was far less serious than it could have been, as there were no injuries.

The ARSR report went on the say that the reporter knew there was a line of thunderstorm and felt he could go around it. He accepted an IFR clearance because he was too close to chance getting into Class B airspace. When ejected from the cloud he was only a few hundred feet above the ground. Since he had carefully reviewed the route he knew there were power lines, hills, and towers in the area.

This pilot's first assumption was that he could circumnavigate the thunderstorm activity. Without electronic storm detection equipment a pilot's only option is to use the "mark one eye ball" to avoid convective activity. Why did the pilot accept an IFR clearance? This eliminated the pilot's only means of storm detection and avoidance. The pilot expected ATC to vector the aircraft "...around the thunderstorm activity." ATC is a resource, and may be able to assist the pilot around hazardous weather; but, pilots can never expect the controller to keep them completely free of adverse weather or assume pilot in command responsibility.

The reporter states: "I opted to make a precautionary landing rather than subject the aircraft, my passengers, or myself to any further danger." Even beginning the flight appears unadvisable. There have been many situations when I have had to park my Turbo-Cessna 150 (That's an attempt at a little humor.) and take a Boeing 757. When you don't have the equipment to handle the weather, don't go!

Case Study

On an IFR flight I was switched from approach to center after being cleared to deviate west of course. I cloud hear center talking to other aircraft, but they failed to respond to my check-in for some six to eight minutes. During this time I entered heavy weather which did not show on the passive (lightning detection) weather indicating system. Weather included heavy rain and severe turbulence. After autopilot disconnect and slowing of airspeed I was unable to hold altitude. Center finally responded to my calls. They told me they could not help me and switched me to approach control. After about a minute with approach, I broke into the clear.

I realize not all center radar has weather graphic overly. If center had been able to see the weather in my flight path and if they had not been so overloaded that they could not talk to me, I would have had a safer, more comfortable flight. I also realize the limitations of the passive weather indicating system equipment.

We've mentioned the limitation of lightning detection system and ATC radar. ATC's primary responsibility is the separation of known aircraft and the expeditious flow of traffic. Again, pilots can never expect ATC to keep them out of convective activity.

Aviation forecasts

Forecasters can predict within one to two hours, the onset of thunderstorms, with radar available. However, forecasters cannot predict with an accuracy that satisfies operational requirements the onset of a thunderstorm which has not yet formed. This is perfectly illustrated by most Center Weather Advisories (CWA) for convective activity. Some tout this product as a "nowcast." In my experience it is a "hindcast." I have yet to see one issued before the convective activity exists. This further reflects the difficult of predicting this awesome aviation hazard.

Case Study

The synopsis described a moist unstable air mass. Thunderstorms were not forecast for the time of flight, but were expected to develop; thunderstorms, however, were already being reported along the route. The pilot, without storm detection equipment, encountered extreme turbulence inadvertently entering a cell. The pilot, with three passengers, filed an IFR flight plan based on the fact that there were no advisories. After the encounter the pilot could not understand why a precaution or advisory regarding that system was not provided. There were no advisories in effect because, at the time of the briefing, none were warranted.

The pilot had the clues—moist unstable air; thunderstorms already reported—but put complete thrust in a forecast that included no precautions or advisories.

Avoidance is the operative word with thunderstorms, microbursts, and wind shear. A pilot's proper application of many resources—training, experience, visual references, cockpit instruments, weather reports and forecasts—make avoidance possible.

Another member of the NWA's Aviation Weather Committee is Mr. Tom Horne, Editor-at-large from AOPA Pilot magazine. Here are two of his published articles he wanted to share with you.

Convective Calamities
Case studies of how things go wrong in thunderstorm season
By Thomas A. Horne (From AOPA Pilot, June 1998.)

For the time being, let's do away with any deep theoretical discussions concerning thunderstorms. Besides, we should all know by now that thunderstorms are caused by moist, unstable air rising under the influence of frontal, terrain, or upper-air lifting forces. We should also know to look for advance warning about thunderstorms from National Weather Service products such as the convective outlook, radar summary charts, radar reports (rareps), TAFs, area forecasts, PIREPs, and low-level significant weather prognosis charts, among others. Of course, the diligent pilot will also obtain a complete preflight weather briefing from flight service or DUATS. That should go without saying.

With those sound government-issue bits of advice out of the way, it's time to look at a few examples of how things went terribly wrong for several general aviation pilots who flew in convective situations. No theory here. We're talking about the realities surrounding thunderstorm accidents.

The following accident synopses are instructive because they are so typical in their unfolding. And the lessons are so clear that there's no need for pontification.

• June 15, 1982
. A 2,500-hour commercial, noninstrument-rated pilot flying a Piper PA-32 Saratoga on a cross-country flight near Hartshorne, Oklahoma, was advised in flight of a severe thunderstorm and tornado watch along his route of flight. The pilot had not received a preflight weather briefing. At 4:13 p.m., flight watch told him of a line of radar returns 10 miles ahead. The pilot said that it "looks like a soft spot through there just to the north of our position," and flew on. Eight minutes later the pilot radioed that he had an emergency. He said that the engine had quit and that he had lost an aileron; 25 seconds later he said that the airplane was hard to control, and that he was at 2,500 feet and descending. Then he reported the loss of the right wing. An examination of the wreckage indicated that the airplane hit the ground in a flat spin. The pilot and his two passengers were killed.

July 13, 1994. An instrument-rated Cessna Skyhawk pilot received two preflight weather briefings — one on the day before the flight, the other prior to departing — before taking off on a cross-country flight from Janesville, Wisconsin, to Mount Pocono, Pennsylvania. During the second briefing the pilot was told to expect conditions in Pennsylvania to be 4,000 to 5,000 broken, with tops to 10,000 feet. Widely scattered thunderstorms, with tops to 35,000 feet, were also forecast. Where thunderstorms were forecast, the pilot was told to expect ceilings and visibilities in the 1,000-and-one mile range. While en route, the pilot made a fuel stop at the Wood County Airport in Bowling Green, Ohio. But he didn't receive a weather briefing prior to leaving Bowling Green. By 5:23 p.m., the pilot and his passenger were over western Pennsylvania, and a radio call to flight watch was made. The pilot was told of a severe thunderstorm watch in effect for parts of eastern Pennsylvania, as well as thunderstorms in progress northeast of Wilkes-Barre. In a second transmission, flight watch told the Skyhawk pilot, "...You are definitely going to have problems with that area [Wilkes-Barre, Mt. Pocono] with weather." A discussion about route deviations ensued. The pilot then called the New York Air Route Traffic Control Center, saying, "I'd just like vectors for the closest airport...sounds like we're going to be running into some storms here and I'm going to wait those out." Vectors for an approach to the Williamsport (Pennsylvania) airport were issued. While flying toward Williamsport, the pilot radioed, "...This is pretty bad for us right here. Do you have any vectors for us?" The controller said that his radar didn't paint weather very well, then asked for radar information from another airplane in the area. Seconds later, the Skyhawk pilot said, "We're flying into some heavy lightning here, and, ah, decreased visibility." The controller then said, "There's an airplane at your twelve o'clock and ten miles.... He says it's pretty good to the southwest if you want to take a turn to the south or get away from the weather." The pilot acknowledged that transmission, then was told to switch frequencies. He checked in on the new frequency, but the controller didn't acknowledge the callup at first. When he tried to call the Skyhawk pilot, there was no reply. The airplane crashed 14 nm northwest of the Williamsport airport, where the latest surface weather observation included a report of 6,000 broken and a visibility of seven miles. In the remarks section of the report it was noted that there was a thunderstorm northwest of the field, moving southeast, with occasional lightning in clouds. The pilot and passenger were killed.

• September 9, 1985. During his preflight weather briefing, a 2,000-hour, instrument-rated Mooney 201 pilot learned of the possibility of thunderstorms along his planned route from Beaumont, Texas, to New Orleans. On the morning of the accident flight, the pilot had a Stormscope installed in the Mooney. That evening, while en route, New Orleans approach control warned the pilot that convective activity with VIP (Video Integrated Processor) Level Three radar returns was along his flight path. The pilot didn't acknowledge this information. During his descent into the New Orleans area, ATC radar logs showed, the Mooney's descent was normal at first, but then went vertical. At about 10:05 p.m. the airplane crashed in a swamp near Kraemer, Louisiana, in a vertical attitude, killing the pilot.

• October 31, 1995. The pilot and three passengers in a Beech A36 Bonanza were killed during a takeoff in a thunderstorm from the Colonel James Jabara Airport in Wichita. A seven-year-old boy survived the crash with serious injuries. Before the flight, the instrument-rated pilot received a weather briefing and filed an IFR flight plan to Colorado Springs. During the 5:37 a.m. briefing, the pilot was told of a severe thunderstorm watch north of his route, a "pretty severe line of thunderstorm activity," and a convective SIGMET. The briefer said, "...It looks like you should be able to slide around the southern edge, uh, south of the Dodge City area to get around that area." At 8:30 a.m. a severe thunderstorm with Level Six radar returns and gust front winds of up to 63 knots was approaching Jabara from the northwest. Linemen busy tying down aircraft noticed the Bonanza taxiing out for takeoff. Just as a northwesterly gust arrived, the Bonanza took off. As several airport employees watched, the Bonanza climbed to about 50 to 75 feet, pitching and banking severely. One witness said that the nose then pitched up and the left wing dropped as though the airplane were entering a spin. The airplane crashed 1,000 feet from the departure end of Jabara's Runway 18.
Most accident reports provide a number of lessons. The common thread weaving through the previous four is a foreknowledge of convective weather. Although these pilots knew what might await them, they pressed on. Sure, the pilot in the Oklahoma crash wasn't instrument-rated, and apparently attempted to continue flying VFR into deteriorating weather — a major cause of many, many thunderstorm accidents. But he knew to call flight watch for inflight updates, and he was accurately warned of severe weather ahead. In the Pennsylvania crash, the pilot became well informed of the convective situation he was entering, and was executing a diversion to an alternate airport. This case may have been one of just plain bad luck.

It happens. The pilot was trying to avoid storms but appears to have flown into a cell that developed suddenly as he approached his alternate airport. He was acting in great diligence but got trapped. It would be easy to blame him for not doing a 180-degree turn, but in this case — as in so many others — that may not have worked. The pilot was working with, and trusting, all the information he could get and was doing just what he should have done. But he crashed anyway. What's the lesson here? Stay on the ground if storms are ever mentioned in a briefing? That would work, but so many times it's just not practical. And in any event, the go/no-go decision is composed of a blend of factors: your qualifications, experience, and comfort level; your airplane's performance; your weather avoidance equipment (if any); and the region you're flying in.

The New Orleans crash carries a special warning about the risks of flying at night in thunderstorm season. At night it's more difficult to see and avoid clouds with convective potential. A bright moon certainly helps, but not if you're picking your way along through cloud conditions that favor embedded thunderstorms. Flying VFR in IMC in areas with Level-Three weather? Not a good idea.

Finally, the Jabara accident. Here we enter the realm of mystery that surrounds so many weather-related crashes. Why did the pilot decide to take off? We'll never know. He must have known, on a purely intellectual level, about the dangers he was about to face. But he never put that knowledge into action. Figure out the disconnect between pilot knowledge and judgment, and we'll be well on the way to ending these sorts of crashes.
And Mr. Horne's second article on an aircraft that encountered hail...

A lucky pilot flies too close to a monster storm

When it comes to close encounters with bad weather, sometimes fate deals you a dirty hand. On August 30, 1996, at 4:15 p.m., Darrell Surface, a farm implement dealer in Kiowa, Kansas, obtained a weather briefing for his flight from Albuquerque, New Mexico, to Alva, Oklahoma. The briefer spoke of thunderstorms building in intensity from Raton, New Mexico, to Garden City, Kansas -- an area to the northwest of the route of flight. Scattered thunderstorms were active in the panhandle of Oklahoma, the briefer said, and the entire thunderstorm complex was moving east and southeast. This would affect the intended route. However, no SIGMETs were in effect.

Surface, who holds a private pilot certificate with instrument and multiengine ratings, has more than 2,900 hours of flight experience.

He filed an instrument flight plan and took off with two passengers in his Cessna 421B. At 5:55 p.m., Convective SIGMET 70C was issued for isolated severe thunderstorms 10 miles north-northeast of Las Vegas, with 20-mile diameters and tops above 45,000 feet. Controllers should have warned Surface about this SIGMET, but when he checked in with ATC at 6:05 p.m. and reported level at Flight Level 210, it was not mentioned. About 14 minutes later, he was cleared to descend to FL190.

Surface saw clouds ahead of him and turned on his radar to have a look. He saw two "curved, softball-shaped" radar returns on either side of his display screen and "no echoes behind" the roundish returns. He also saw what appeared to be a 10-mile-wide, echo-free gap between these two echoes. He decided to fly through the gap to make his way through the storms. By this time, the convective SIGMET was about a half-hour old. Surface hadn't been listening to HIWAS (he was navigating using GPS), nor had he contacted flight watch.

As the 421 approached the gap, Surface recalled thinking that the clouds around him "weren't ominous, but I couldn't see the tops." In a National Transportation Safety Board statement, he said that the storms around him "did not appear to be dark thunderstorms, only clouds like I had flown through many times before." He also said that the radar "showed rain on both sides but just a small ring of rain in front."

Then, Surface said, he encountered rain for 5 to 10 seconds, and light hail for another 5 to 10 seconds. Then all hell broke loose.

Apparently, that "small ring of rain" was anything but. Three waves of heavy hail assaulted the airplane. The windshield in front of him was blown out, and Surface was pelted with 2-inch-diameter hail. It dazed him and hit him all over his body, shredding his clothes, bloodying his skin, and detaching the iris from one of his eyes. The leading edges of the wings were beaten back "a good 6 inches." The cabin partition behind him was blown out by hail chunks that "left a hole big enough to put two footballs through," according to Surface. Both passengers were also injured by the hail. Surface put his head into the front passenger's lap to try to avoid further injury, and the passenger (who was a private, non-instrument-rated pilot) said that he "took over the controls" and "attempted to keep the wings level."

Now for the unusual part of the story. The next thing all three aboard the airplane recall is waking up in the 421's wreckage. All three had survived the experience, although the airplane was destroyed. The airplane had crashed in a level attitude on the desert floor, and no one knows why or how. As Surface and the front-seat passenger came to, the rear passenger opened the airplane's door and walked three-quarters of a mile to see if anyone was nearby. Ultimately, an Air Force C-130 homed in on the 421's ELT, and a medevac helicopter rescued all three.
"The Lord had his hands on the airplane that day," Surface said in an interview. "And He put us on the ground. I have no memories of the crash, only trying to protect my eye."

Today, Surface is flying again. His eye healed after surgery, he regained his medical certificate, and he bought a turbocharged Cessna 210. Like the 421, the T210 is used in Surface's business flying.

The accident emphasizes everything that we'll be saying in this series of weather articles: that storms are worst late in the day, that radar can lie, that storms can grow quickly, and that you must do your best to stay in touch with sources of current weather information. ATC was in the wrong for not advising Surface about the convective SIGMET. But with the luxury of hindsight, it also appears that the 421's radar picked up a hail shaft (not a "small ring of rain"), which is one of the most highly reflective phenomenon a radar can see, and that Surface definitely flew too close to a fast-moving monster thunderstorm. The NTSB has yet to issue an official probable cause for the accident. The information presented here is, in large part, derived from the NTSB preliminary report.
Thank you, Mr. Horne, for sharing your work with us. By the way, we've just added three other articles from Mr. Horne into previous lessons. Here are the links in case you want to read them.

Radar Realities

Convective SIGMETS

Summer Weather

And so we come to the end of the last lesson. As course director it has been my honor to be part of this effort. Thank you for your continued interest in aviation weather and in flying safety.
It is now my privilege to introduce Mr. Les Lemon, the President of the National Weather Association.

On behalf of the entire membership of the National Weather Association (NWA), I thank you for letting us share our keen interest in weather and its impact on professional and recreational flying activities. Every member of the NWA Aviation Weather Committee who put this outreach training/education program together for you is a volunteer. Each person was motivated solely by the desire to have you know as much as possible about thunderstorms and their many hazards, and the aviation weather support system so that your flying is the safest possible. I sincerely hope you caught a bit of their enthusiasm for the subject and will carry it with you for as long as you continue to fly. If you do that, then this project will have been successful. Thank you also for your questions and comments. They helped improve the course and many will lead to improvements in the weather support system and flying safety.

I also want to publicly thank the guest speakers that allowed us to use their comments and photographs in this course. We are gratified that these leaders took the time to give their full support to our efforts in the cause of flying safety. Thanks also to the many agencies and associations that allowed us to use their logos to show their interest and support.

Already, the NWA Aviation Weather Committee is thinking about other Web-based courses/tutorials to produce. Later this year there may be a "winter flying" course. Please give the committee your feedback on its work so that your needs are best served. I also invite you back to the NWA Web site ( We want you to return often to learn more and to check out our new initiatives to increase weather awareness and safety in all endeavors and to promote excellence in operational meteorology and related activities. Should you be inclined, we would also like to encourage you to consider membership in the NWA.

Again, thank you for letting the National Weather Association be part of your flying experience and continuing education.


Leslie R. Lemon
National Weather Association

Study Questions:
1. According to the reading, are there any "true all-weather aircraft" when it comes to severe weather?
2. What are some of the hazards of thunderstorms?
3. What are four most common deadly errors pilots make when it comes to thunderstorm flying?
4. According to the reading, what are some signs of thunderstorms in the terminal area?
5. According to the reading, what are three things a pilot needs to know when flying near convective activity?
6. Mr. Lankford quoted the NTSB's recommendations to avoid radar attenuation. What is that recommendation?
7. What must every pilot using a radar know?
8. Even if outside a thunderstorm cloud, what must every pilot be concerned about?
9. When using a lightning system, does a clear scope mean there are no thunderstorms in the area?
10. Can a pilot depend on ATC radar to keep them clear of thunderstorms?
11. According to Mr. Lankford, what is the operative word for thunderstorm flying?
12. In Mr. Horne's article, what was the common thread to all the accidents?
13. What is the problem with night flying?
14. What are the main points that Mr. Horne wants every pilot to know about thunderstorms?
15. Will you fly into a thunderstorm deliberately?
IF YOU ANSWERED "yes" and you are not a research pilot or hurricane hunter, please go to lesson one and start this course over! Thank you.