The weather was terrible, it was getting dark, he barely knew how to fly his complicated new airplane properly. But almost every decision John F. Kennedy Jr. made on his fateful trip was sound. Events conspired against him. The only mistake he made was not to realize that the deck was stacked and he wasn't dealing.
Tom Wolfe wrote in "The Right Stuff" about the denial of pilots, the eagerness to find fault with a dead pilot as a means of protecting yourself against disaster. "What a stupid idiot," we say when one of us dies. "I'd never take off and fly into thunderstorms, of course he died." He was stupid, you're not, therefore you will survive. In an accident like this that's a dangerous game to play. Which of us pilots hasn't taken off a half-hour later than planned; struggled to get comfortable in a more complex airplane the first time we soloed in it; discovered that the weather along our route was worse than forecast?
Admittedly, I'm more experienced than Kennedy was. Part of that experience comes from studying accidents and learning from them. I've been thinking about that, thinking about how this flight must have gone, how pieces of it went like so many flights I've taken. And how I can learn from it, learn to do things differently if I'm faced with similar circumstances.
No single decision or factor caused this crash. It was what safety investigators call a "cascade" of events -- a series of decisions and problems that the pilot couldn't necessarily have foreseen. No single element was truly decisive, they piled up on each other, leading up to an inevitable disaster. It started with the late takeoff.
It's hell to sit and wait for passengers to show up, especially when a weather window is quickly closing. People say, "Oh, the luxury of a private plane! You can come and go as you please!" But of course, you can't. A pilot is always at the mercy of detailed weather reports that mostly guess at what's up there in the clouds.
I have waited in airports, anxiously scanning the skies, trying to divine the moods of the clouds, wondering how much daylight is left. I have made last-minute calls to flight service for the latest weather updates. I ask for pireps, or pilot reports, filed by fliers actually up there, reporting on real visibilities, winds and cloud heights, looking at my watch and wondering where the hell the passengers are.
Taking off into dusk isn't a death sentence. Night flying over land is beautiful, with the cities of light fanning out into the highways that connect them. The radio chatter dies down and often I'm the only pilot talking to a traffic controller, who might share a friendly joke and ask me about the view. The navigational instruments help pinpoint my location as I follow along on a map, the instrument needles pointing to a beacon on the map, telling me how far away it is.
Haze isn't a lethal condition either, by itself. Flying in haze is like driving with a piece of gauze thrown over your head. You can see light, you can see enough to make out intersections, traffic signals. But everything is a little blurry, behind a white film. One scary thing about haze is that there's no telling how bad it is. Aviation weather forecasts simply say. "Visibility: 5 miles and haze." Haze, in effect, decreases visibility because it makes the far away stuff too blurry to make out. So what am I really going to have? Even if some other pilot has kindly filed a pirep, guessing visibility distance is subjective and his 3 miles may look like 2 to me.
I have played the scene over and over in my head. Based on flights I've made, I can imagine the takeoff. Carolyn and her sister belatedly rushing out to the airplane. The pilot, limping, making one last walk-around, moving the flaps and looking in the fuel tanks to make sure they're full. It's a quick trip, just an hour flight. In an airplane that burns about 16 gallons per hour, 62 gallons will keep the engine running for about four hours, as long as the pilot remembers to switch tanks.
Switching tanks was just one more thing for me to remember when I transitioned to my Bonanza. I already had a dozen hours in a Piper Arrow, a
retractable-gear, high-performance airplane that's a bit smaller than the
Saratoga. I had learned how to handle the separate propeller adjustment and the knob for landing gear, but the big Bonanza still shocked me. The things to remember on such a complex airplane.
The Kennedy plane took off while it was still light outside, maybe hoping to get there before the sky completely blackened. Many pilots don't like monkeying around with night landings. The runways are brilliant strips of light surrounded by dark, they tend to look like they're sitting down in black holes. In aiming for them pilots often come in too low.
Kennedy planned the flight carefully, sensibly. He followed the coastline until the last possible moment, only then turning out to sea. The plane was always within gliding distance of the shore -- which was smart because he didn't have life vests in the plane. I don't blame him, I can't imagine having enough time in an emergency to wriggle into a vest while also trying to control a panicked plane, so I don't always carry them either.
Being able to see land also gave him a better horizon. It was starting to get dark, but he could probably see lights on the shore. I think about him peering through the dark, squinting for the Vineyard lights before turning out to sea. Maybe he asked the passengers to start looking for the airport as he turned to the right, heading toward the navigational beacon that sits on the Vineyard airfield.
Maybe he used the autopilot. When it's linked to the navigational radios an autopilot can fly the plane straight over the airport. Some small aircraft autopilots, like mine, can fly the airplane down to within a few hundred feet of the runway. But it took me almost a year to figure out how to use my autopilot for much more than stable cruise flight. I had never used one before, so whenever I started a descent, I turned it off to "hand-fly."
Maybe he wasn't precise about calculating the rate of descent. It's a simple mathematical formula for pilots, but it took me years of flying to be able to do it quickly, in my head. He had to lose about 4,500 feet. At the standard rate of descent, 500 feet per minute, let's see, divide 4,500 by 500, and that's nine minutes. The airplane cruises at about 180 miles per hour, so that's (180 divided by 60) 3 miles a minute. So it's going to take, uh, 9 times 3 and it'll take 21 miles to get to pattern altitude.
Whew! Doing that calculation always takes me a few minutes. By that time, at 180 miles an hour, he's already 34 miles out. It's a good idea to reach pattern altitude a little before actually entering the pattern, so he descends slightly faster than 500 feet per minute, not dangerous, at 700 feet per minute. Just trying to get down and level off, and leave enough time to go through the landing checklist without rushing.
At 2,300 feet, 20 miles away from the airport, something happened. Maybe he simply leaned forward, looked down at a switch, at a map, then looked up. That simple movement might have been enough to confuse him into making the climbing right turn, gaining 300 feet. Before I got my instrument rating, I once flew the Vertigon, a flight simulator machine created to induce vertigo in pilots, to teach them about the dangers of ignoring the instruments. As I sat in the machine, which was turning, the instructor told me to reach forward, as if I were picking up a dropped pencil. When I sat back up, I was gone. My head was spinning, I felt pressure on my body but couldn't tell which way I was turning. I felt sick to my stomach, the spinning instruments made no sense.
At 18 miles out, it looks like Kennedy recovered. The plane banked back toward the left, correcting for the climbing right turn. It returned to the 700-foot-per-minute descent established earlier. He was heading back to the airport. While he turned to correct is probably when he reached down for a pencil, or for the book with the control tower radio frequency or to flip the gear switch and look down at the indicator lights.
And all hell broke loose.
Thirty seconds later, the plane turned to the right and fell out of the sky. Perhaps the plane stalled and went into a spin, as the pilot battled the confusion in his inner ear. To stall it, Kennedy would have had to fight the controls hard, pulling back on the yoke strongly to correct for an imaginary dive. I never was able to get a full power-on stall in my Piper Arrow or Cherokee, I'm not strong enough. Or he could have pulled the power back sharply, perhaps confused by the whistling wind of the descent, the grinding, whiny noise of the gear unfolding, thinking that the sound of an overspeeding engine. To recover from a spin, if he recognized it, he would have had to stomp hard on the rudder with his injured right foot, if the plane were spinning to the left.
Or maybe Kennedy simply overcorrected on the turn, diving and turning until he could no longer wrestle it level, out of the spiral. Possibly something failed on the airframe, but this is a new plane, one that low-time pilots love because the wings tend to stay level, the nose tends to stay on the horizon.
A few seconds later, it's gone.
Did a late start cause the crash? No. Did haze cause the crash? No. Did nighttime flying cause the crash? No. Did the fancy new airplane cause the crash? No. Did disorientation cause the crash? Not by itself. They all piled up, like the 10 cars that hit an overturned tractor trailer on the highway. What caused the 10th car's crash? The inexperience of a driver who simply can't drive through a highway blocked by nine cars and a truck.