Drilling hills

April 1, 2002

Bruce Landsberg has served as executive director of the AOPA Air Safety Foundation since 1992.

Moving mountains with aircraft has never proven to be successful. The size of the hills is as irrelevant as the size of the aircraft.

The FAA categorizes this type of accident as CFIT, or controlled flight into terrain. CFIT is a leading cause of air carrier accidents, particularly in developing countries. More people and aircraft are lost to this category of flight than any other phase of flight in commercial aviation. Sometimes it happens on approach and sometimes en route.

In general aviation, as well, we have more than our share of CFIT accidents. There was a "classic" that occurred on the East Coast last August. It so perfectly describes the genre that it is worthy of study. The accident report is still preliminary so some of the finer points may change. The main issues, though, are clear, even if the weather was not.

The aircraft departed Annapolis, Maryland, en route to Indiana with a proposed return a week later. The private pilot was rated for single-engine aircraft but held no instrument privileges. His third class medical certificate was issued on February 6, 2001, and he reported 2,000 hours of flight experience then. The pilot completed a 1.9-hour annual proficiency check in a rental Piper Arrow with 180 horsepower in June. In July, he received an additional 50 minutes of dual during an initial checkout in a PA-28RT-201, the accident aircraft. He logged one additional flight of 1.8 hours in the PA-28RT-201 in mid-July before undertaking the long trip.

The preliminary accident report did not reveal how much cross-country time the pilot had or if he had received much instrument training. Two thousand hours is a respectable amount of flight time, and I'm always curious when someone is this heavily involved in aviation and yet does not invest in an instrument rating. Unless you are just doing local fair- weather hops—and it's tough to accumulate big hours that way—the opportunity to tangle with serious weather increases in direct proportion to exposure.

It is well known in insurance underwriting circles that VFR-only pilots in complex aircraft constitute an increased risk. Retractable-gear aircraft are likely to spend more time in cross-country flight than fixed-gear types. The faster and farther an aircraft ranges, the more weather systems it will face. However, it's one thing to spout platitudes about not flying while relaxing in your easy chair and quite another to be waiting to take off with an aircraft that is due back that day and a boss who wants you at work the next. Add some pressure from passengers who want to get home and a few good METARs to entice you. This is the definition of temptation.

Our capacity to rationalize under these conditions is remarkable and can be extremely dangerous. What's hard to square is that many VFR pilots with broad experience have been successful for years in dodging the big one. Only they know how many times they've run the scud. It's not something that most people record in logbooks. Some get religion after a bad experience and either get an instrument rating or resolve never to have such a close encounter again. However, as time passes some have a tendency to backslide, and then the cycle repeats itself or comes to an abrupt and violent end.

Few people touch a hot stove twice—the outcome is swift and certain. People would be less careful with stoves if they only got slightly singed on rare occasions. That's how it is with weather. Forecasts are uncertain, and pilots proceed cautiously—testing at first and gradually becoming bolder. Entanglements tend to be mild, encouraging more forays. Weather that looks the same often isn't, and terrain changes the dynamics. According to the safety investigator's report, an FBO employee stated that the airplane arrived in Indiana on August 8 and departed on August 11. When the airplane arrived, the female passenger stated that they had "run into a cell" around Terre Haute, Indiana, earlier in the day and that they had to spiral through "a hole" in the overcast to "get into the airport." That statement indicates this pilot had a rather casual approach to remaining in visual conditions. It takes some effort to fly into a thunderstorm and then dive through a small opening in a lower cloud deck while maintaining VFR minimums.

On the day of the return trip, an FBO employee at a fuel stop in Ohio stated that the airplane arrived at 11:30 a.m. The occupants ate lunch at the airport and "hung out" until 4:30 p.m., when the pilot purchased 26.9 gallons of fuel. This delay was presumably to let weather that was ahead of them move farther east. The female passenger stated to the employee that she had previously gotten over her fear of flying, and now had gotten over her fear of flying in thunderstorms. She again mentioned that they had encountered a storm on a previous flight. The employee closed the terminal building around 4:30 p.m. and noticed that the airplane was gone when he returned around 6 p.m.

The weather this summer evening was not good for VFR flight. A frontal system was slowly working its way east carrying scattered thunderstorms and ample moisture. The following reporting stations were within roughly 15 miles of the accident site and formed a triangle around it. The weather at Frederick Municipal Airport in Frederick, Maryland, at 7:20 p.m. included calm wind, three and one-half miles' visibility, scattered clouds at 1,200 feet, broken clouds at 2,600 feet, and an overcast cloud layer at 3,500 feet. The temperature was 24 degrees Celsius; dew point was 23 degrees C.

Weather reported at the Eastern West Virginia Regional Airport in Martinsburg, West Virginia, at 7:22 p.m. included calm winds, 10 miles' visibility, scattered clouds at 6,000 feet, and a broken cloud layer at 9,000 feet. The temperature and dew point were 23 degrees C. Lightning was reported in the distance to the east, southwest, and west of the airport.

Weather reported at the Hagerstown Regional Airport in Hagerstown, Maryland, at 7:19 p.m. included wind from 100 degrees at 7 knots, four miles' visibility in haze, broken cloud layers at 2,000 feet and 3,600 feet, temperature of 25 degrees C, and dew point of 21 degrees C. At 7:38 p.m., a thunderstorm was reported in the vicinity with lightning in the distance to the southeast and southwest of the airport.

There are traps for the optimist in these METARs. Hagerstown and Frederick were just above VFR visibility minimums, while Martinsburg reported good VFR. The dew point tells another, more sinister story. The air was saturated, or close to it, in all locations, which meant that clouds, mist, rain, and thunder were very much a possibility. The forecast of frontal activity certainly carried some weight since the pilot delayed for several hours in Ohio. The front's location was clear and the danger was present.

A search of air traffic control and flight service station records revealed no record that the pilot obtained a weather briefing or filed a flight plan. There are many unofficial sources of weather, especially with the proliferation of computers at FBOs, so it's entirely possible that the pilot knew exactly what he was getting into. That remains an unknown. However, FAR 91.103 requires us to be familiar with weather reports and forecasts, and if questioned by the FAA it would be very helpful to have a printout or written record of the information. Several witnesses observed the airplane flying in the vicinity of the accident site. One witness stated: "I heard the airplane flying overhead from west to east, about 7:15 p.m. It was a single-engine, low-wing, T-tail airplane with a red stripe. The airplane was flying straight and level and appeared to be flying in a direct line from the Martinsburg airport in West Virginia to the Frederick airport. The weather was drizzly, and because the ceiling was so low, I could not see the top of the mountain. The airplane was flying in and out of the clouds, and it looked like he was just trying to get home. I thought the airplane would just barely make it over the mountain.... The engine sounded smooth and normal. I never heard it sputter or quit."

A second witness stated: "I saw the airplane under fog and mist, and then saw it pull up real quick to try to get over the mountain. I could not see the top of the mountain from ground level because of the heavy mist and fog. When I saw the airplane, I knew he would not make it over the mountain. The engine sounded normal until I saw the airplane pull up, and then I heard the engine 'rev up.' I heard what sounded like an impact a few seconds later."

The wreckage was located early the next morning by two hikers in a heavily wooded area at an elevation of 1,355 feet msl. The peak of the ridgeline, three-quarters of a mile east of the accident site, was 1,800 feet msl, so the flight did not just clip the top of this hill—they weren't even close to clearing it. There were no survivors. At this writing, there were no identified malfunctions of the engine, instruments, or flight controls.

All of which brings us to just one question. Assuming the pilot was not suicidal, was his idea of weather such that, based on prior experience, despite some fog, some rain, and maybe even a "cell or two," it was still safe to go? There are some indications, judging from the passenger's remarks, that this pilot may have allowed his experience to cloud his judgment. There is no harm in "taking a look" as long as the VFR pilot commits to turning back before the clouds envelop the aircraft. If you're unable to make that commitment, even though the destination is only a few miles ahead, you're not acting as pilot in command but as passenger in command. The aircraft will continue on the path that was set.

When you're tempted to scud run, try some visualization: imagine rain, fog, turbulence, and add in the confounding factor of terrain or a tower that suddenly appears. Now imagine the aftermath of CFIT. There is no need to be graphic, but pilots should be very clear about what's at stake. When CFIT is involved, the score will always be mountains, one; pilots and passengers, zero.