Most of the time American aviators learn and study about American accidents, but accident reports and studies from other countries can also pack useful lessons. For example, one interesting difference between flying in the United States and many other countries is the freedom to fly VFR at night. Most countries require pilots to be instrument rated and to be on a IFR flight plan to fly at night. Many others do not allow single engine flying at night.
Within the U.S. and a handful of other countries, single engine night VFR flying is accepted without question. However, pilots and investigators from countries that do not allow single engine night VFR flying can cite relevant accident statistics that show the rate of night takeoff accidents to be four to five times more frequent than day takeoff accidents.
Indeed, taking off into the black hole of a night sky does have some notable pitfalls. The pilot who isnt scared when taking off into a black hole at night probably should be. The average mishap pilot crashes within one minute after takeoff in this scenario. An accident report from a crash in Australia that occurred during dark night conditions is a good example to illustrate the numerous hazards involved in night flight, regardless of the country in which youre flying.
Australian Lesson Learned
At 00:32 hours, the pilot of a Piper PA-32 advised flight service that he was taxiing at the Mount Gambier, South Australia airport. Conditions at the time were moonless and very dark. Like much of Australias open country, sparse population meant that ground reference lighting was almost nonexistent. The only available ground reference lighting was more than 13 kilometers to the north of the airport. Between the lack of ground lighting and the moonless night, this left no visible horizon. In addition, low cloud layers at about 2,000 feet may have been sufficiently thick to obscure any starlight. The pilots night flying experience was limited, particularly when it came to dark night takeoffs without nearby ground lights.
The aircraft took off and was banked slightly right when it initially struck the tops of tall pine trees about 800 feet from the end of the runway, and about 550 feet to the left of the extended centerline. The aircraft struck the ground at an angle of about 40 degrees with the landing gear extended and low propeller rpm.
The Australian Bureau of Air Safety Investigation concluded that the circumstances of the accident were consistent with the pilot suffering the effects of somatogravic illusion, a very subtle form of disorientation. The Bureau also noted that the pilot failed to establish appropriate climb attitude by reference to the flight instruments.
For The Record, American Style
Night takeoff accidents such as this are fairly common. Within the United State, NTSB data indicates a total of 291 take-off accidents had occurred at night during a recent 10-year period. The vast majority, 281, involved fixed-wing aircraft. Of these, 112, or 40 percent, involved pilots who were not instrument rated.
The most common factor quoted in night take-off accidents was dark night conditions. Dark night is a special condition in which there is no illumination from the moon, making any distinction between the ground and horizon virtually impossible to discern. A total of 153 accidents, 54 percent of the total, had the dark night listed as a possible cause or factor.
Flying in such conditions requires the pilot to fly by reference to instruments. That creates the potential for spatial disorientation, and indeed, we find that spatial disorientation was listed as being a possible cause in 42 (14%) occurrences, and as the most probable cause in another 36 (12%) accidents. Combined, this implies that spatial disorientation was a factor in at least 78 (26%) night take-off accidents in the US over those 10 years.
One of the special illusions that we are exposed to during the takeoff, especially at night, is created by the combination of climb and acceleration, and the misinterpretation of those cues by our bodys internal balance mechanisms. Whenever a pilot is subjected to acceleration and climb at the same time while deprived of external visual cues (dark night or IMC takeoffs), the pilot senses a strong sensation of the nose being steeper than it actually is. This tempts the pilot to lower the nose of the aircraft. If the altimeter and vertical speed indicator lag, the loss of height can go unnoticed until it is too late to avoid a crash. Our body becomes disoriented when it receives conflicting messages from the various organs (inner and middle ears, muscular and skeletal systems, visual systems) and can create a situation in which the pilot is functionally disabled. It can be severe enough that a pilot cant tell if the aircraft is right side up.
Since we started out looking at an Australian case, lets continue with Australias analysis. A review of Australian accident records indicates 35 accidents occurred in Australia at night between the years of 1979 and 1992.
Fifteen (42%) of the 35 accidents were attributed to the presence of the somatogravic illusion, spatial disorientation, or some other form of visual or sensory illusion. In many cases, it was determined that the flight was correctly planned and performed in an aircraft with no signs of mechanical malfunction. Nine of the 15 accidents were likely to involve the somatogravic illusion as the principal factor in the accident. Ten of the 15 involved fatalities, which is a very high figure. Only two of the 15 accidents resulted in no injuries.
Twenty percent of the night takeoff accidents involved an engine malfunction, failure, or fire. Eleven percent (four accidents) involved a loss of control, with three of those occurring on the ground.
Weather was a factor in only three of these accidents. Two of these involved a change in wind direction or strength.
Finally, two accidents occurred when unlicensed personnel attempted to pilot a stolen aircraft. From the point of view of accident prevention, it hardly seems worth including these into a database.
Susceptibility to Illusions
The Australian Bureau of Air Safety Investigation attempted to identify factors that would increase a pilots susceptibility to visual or sensory illusions. The investigators hypothesized that a pilots physiological condition, age, experience, qualifications, environmental surroundings, and operational details were all factors in mishaps.
The data, however, fell short of giving a clear answer to the question. The Australians had hoped to increase the size of their sample by including British and American accidents into their study, but the reports lacked sufficient detail to be used in conjunction with the Australian files. In fact, the British noted only two accidents in the last 25 years in which somatogravic illusion may have played a role, perhaps because Britain does not have the wide open spaces common in the U.S. and Australia where visual horizons are frequently absent. The Australians were unfortunately left with a small database for their study.
Given these limitations, they still made a very in-depth analysis of common factors from their data, and the lessons learned from their study apply to us equally well. Of the 18 Australian pilots involved in the 15 accidents, 13 (72%) held instrument ratings. Four (22%) held instructor ratings. Four (22%) had valid VFR night privileges. Twelve (67%) held commercial pilot certificates, while six had private pilot licenses. Quite clearly the training and experience required for the commercial pilots license did not ensure immunity from sensory illusions.
Pilot experience seemed to have no effect on decreasing the susceptibility to sensory illusions. Pilots with over 15,000 hours and those under 200 hours were equally involved in mishaps. The statisticians plotted the experience level versus the number of accidents, but due to the small population of pilots under consideration, there was too much spread in the results to make any real inference. Pilots of all levels of experience were represented in the accidents.
Stress was found as a factor in 8 (53%) of the accidents. Interviews with witnesses in three of the eight accidents revealed that the pilots had conveyed some urgency that the flight be completed. Two of these involved air ambulance operations. Five of the remaining accidents involved stresses associated with family, business or financial problems. Australian investigators noted the skills and concentration involved in night flying and the added pressure involved in the take-off phase of flight make it possible that stress would increase the chance of a night take-off accident.
Fatigue was found in a number of the cases. Eleven (60%) of the pilots were awake for more than 13 hours before their accident, and 13 (72%) of the accidents had definite indications of pilot fatigue. The amount and adequacy of rest a pilot has prior to flying was considered. However, attributing an accident to fatigue is difficult because there may be no reliable indicators of fatigue and even if it can be proved, the investigator still has to establish a causal link between fatigue and the accident.
It may seem that weather would be somewhat to blame in a higher proportion of night take-offs. However, the Australian data does not bear this out. Twelve of the 15 accidents occurred in clear conditions. Only three occurred in IMC, usually involving fog shortly after takeoff. Weather and the proximity of clouds were not found to be common factors in night take-off accidents in this database.
The amount of light available to a pilot at night can vary significantly. Influences include cloud cover, weather, color of the terrain, and the phase of the moon. In all 15 accidents, the night was subjectively described as being extremely dark. In 12 accidents, the moon was absent. Fourteen accidents occurred in conditions that would have limited the availability of a useful, visible horizon, although airport lighting was suitable in nearly every case.
Interestingly enough, investigators found that in every case the doomed aircraft struck the ground within 3 nm of the airport, and within one minute of becoming airborne. In 87 percent of the crashes, the aircraft impacted to one side of the extended centerline with wings showing little or no bank at the time of impact. In every crash, the wreckage indicated that the aircraft was under control at the time of impact. Specifically, the investigators noted that the aircraft damage was consistent with the pilot being unaware of any unusual attitude.
When you look at the night accidents, you begin to notice certain airports mentioned more frequently than others. You would expect busy general aviation airports to have more night accidents just due to the rate of activity, but that isnt quite the case. A little bit of statistical analysis would suggest that some airports have a disproportionate number of accidents. With a little more insight into the problem, it begins to make perfect sense when you analyze the environment around some of these airports. We seem to call these places black holes.
Wendover, Utah, is a perfect example in this case. California City, Calif., is another. As you would expect, the rest of the desert seems to dwarf the landscape. The nearest civilization is several mountain ranges away. All of this means that on a moonless night, there is no horizon. Approaches and landings into these locations can make the pilot feel like they are landing in a black hole. An excellent study examining the black hole approach was done years ago by aviation researchers and is widely quoted now, but a night takeoff from such an airport deserves just as much attention.
The Fine Art of Instrument Takeoffs
A takeoff into night conditions should start with the thorough preflight plan. You must know the terrain and know your aircrafts capabilities. You need to know if there are obstacles that are unmarked and unlit. Sectionals can sometimes help, as can the Airport/Facility Directory. The A/FD will point out powerlines and trees that are within close proximity to the airport. Sectionals may give you an idea of the local terrain, however, that is fraught with some deception. I was briefing a younger pilot on some of the hazardous airports for air ambulance operations, such as Salmon, Idaho, He quickly pulled out a sectional and correctly interpreted the mountains in very close proximity to the airport. Even that, however, doesnt prepare you for just how close the mountains are.
Most general aviation aircraft dont have stellar climb out gradients. A heavy load can compromise the aircrafts performance so much that the terrain often becomes a factor.
The air carriers are required by law to compute the aircraft performance and to ensure that the aircraft can meet the climb-out gradient for that airport under those weather conditions. If an engine fails during takeoff, they have the climb performance to guarantee that they can fly the instrument departure and clear the terrain.
The same policy should apply to general aviation. You must ensure that your aircraft can out-climb the terrain and other obstacles you cant see. This means that you should know the terrain and know the favorable departure path that will give you the best safety margin.
Stack the Odds
After thorough preflight planning, you should continue with a complete aircraft preflight, to include a complete cockpit check. Everything needs to be working in your favor, because if a malfunction occurs on a dark night takeoff, the margin for error is extremely slim. Knowing that the instruments are reliable is a must. Unreliable attitude indicators are not acceptable in this environment. After engine start-up, make certain that the instruments read correct values and show proper trends while taxiing.
Check that all of the instruments cross-check correctly with each other. During a turn to the right, the turn coordinator should show a right turn, and the heading indicator should rotate counterclockwise. Hopefully the attitude indicator doesnt lean during a turn on the ground. If it does you certainly dont want to take off with it. Make certain that the trim is set for takeoff. Properly setting the trim knobs will help the control surfaces to streamline into proper positions, thus helping you to maintain control of the aircraft. The lights in the cockpit should be adjusted so that you can clearly read the instruments, but the lighting should not be so bright that it diminishes your vision outside.
Once you are ready for takeoff, taxi onto the runway and make certain that the nose of the aircraft is aligned straight down the centerline. On a real instrument takeoff, this would be the time to set the heading indicator to the nearest five degree mark. Make certain that the attitude indicator is straight and level.
A Matter of Technique
A night takeoff is not the time to gun the throttle. That will result in the nose veering off to the left side of the runway. It is better to advance the power smoothly, and make appropriate right rudder inputs as the throttle is advanced. Avoid the temptation to move your head because this will create rotational movements in your semi-circular canals which will be interpreted as aircraft movements, further increasing the chance that you will become disoriented. Your scan should go from inside to outside the cockpit.
During the takeoff roll, make certain that the attitude instruments seem to correlate with the aircrafts motion, and that the airspeed indicator is reacting normally. You would be surprised to hear how many pilots have neglected to remove pitot tube covers and taken off with inoperative pitot-static instruments.
As the airspeed approaches the normal rotation speed, apply necessary back pressure on the elevator to smoothly rotate the nose up. If you dont know the proper pitch attitude for rotation, the five degree pitch setting is a good average value in most light aircraft. If your attitude indicator does not have those nice pitch lines, then rotate the aircrafts nose about the equivalent of two pitch bars above the horizon. As the aircraft approaches flying speed, and especially after leaving the ground, the attitude indicator is your best friend. You must maintain that five degrees nose high and wings level attitude.
The altimeter and vertical speed indicator are secondary instruments to confirm a positive climb attitude. Until you climb high enough above the ground, dont move your head, and dont move your upper body. At this point, the most important goal you want to accomplish is to safely maintain a climbing attitude away from the ground. Gear and flap retraction can wait until a safe altitude. When you have safely gotten the aircraft established in a wings level climb away from the runway and at a safe height, dont forget to fly on your previously selected course towards favorable terrain. Turns should be done smoothly and with bank angles seldom steeper than 15 degrees. Remember, during this time your sole attention is on the instruments.
The Australian study was well done and reminds us that dark night takeoffs are a special risk that require careful preflight planning and careful execution. In the United States, we have the freedom to take off VFR at night in a single engine aircraft. With such freedom comes the responsibility of proper preparation and making the correct judgments. The safety margin during dark night takeoffs can be too slim to do otherwise.
-by Patrick Veillette
Patrick Veillette is an ATP with more than 11,000 hours. He directs a research program studying human behavior in high-risk environments.