Landmark Accidents: Midlevel Mayhem

It's hard not to feel sorry for yourself when getting bumped by building cumulus clouds at 10,000 feet and longing for the flight levels. I've often thought that another 10,000 feet would make a flight much easier and safer. Many times it does. However, aircraft flying in the low-flight-level altitudes (arbitrarily flight levels 180 to 290) are subject to some of the worst thunderstorm hazards imaginable. FL250 is just not high enough to get over the big ones, and it's a long way down to the cloud bases where the rain shafts may be seen and sometimes avoided. (That's assuming the veiling clouds down low are not so thick as to preclude some visual identification.) At the higher altitudes there's no quick bailout to a nearby airport — an option that is sometimes open to the bottom feeders. At FL250 the tops may be as far above you as the ground is below, and that's the time to proceed with great respect for Thor's anvil.

On June 14, several years ago, the pilot of a Piper Malibu contacted the Macon, Georgia, flight service station (FSS) at 4:52 p.m. Eastern Daylight Time and requested a briefing for a flight from Rowan County Airport in Salisbury, North Carolina, to Raleigh-Durham International Airport in Raleigh-Durham, North Carolina, and then on to Florida's Marco Island Airport. The area forecast for the Carolinas, Georgia, northern Florida, and coastal waters predicted thunderstorm activity with tops as high as FL450 (45,000 feet). The specialist stated there were "looming thunderstorms" in that area.

On the IFR flight plan from Raleigh-Durham to Marco Island, the routing was over Myrtle Beach, South Carolina, and Orlando, with a proposed departure time of 6 p.m., an altitude of FL260, true airspeed of 250 knots, and an estimated time en route of two hours, 45 minutes with four hours of fuel on board. If that airspeed seems higher than normal for a Malibu, it's because the aircraft had been modified with a Pratt & Whitney PT6 turboprop, known as a JetProp DLX.

Several convective sigmets were applicable. They describe the occurrence or expected occurrence of thunderstorms within two hours of issuance time and are expressed in geographical terms that are usually indecipherable when heard over the radio or telephone. If you don't have a screen in front of you the paramount question is, "What's the best way to avoid the entire area?"

The flight departed Raleigh-Durham at 6:28 p.m. EDT and contacted the Jacksonville, Florida, air route traffic control center to request a clearance direct to Orlando. The Jacksonville controller offered direct to Marco Island. All was routine and, at 8:02 p.m., the pilot requested permission to leave the frequency to check weather.

The Gainesville, Florida, FSS flight watch specialist advised the pilot of "cells east of St. Augustine; they continue to move east at around two-zero knots" and advised of convective sigmet 05E in effect for Southern Florida. The specialist suggested that a routing "toward the Tampa-St. Pete area and then southward" would avoid an area of thunderstorms. At this point, the airplane was about 60 miles east-northeast of Jacksonville.

At 8:06 the pilot reported back on frequency to the Jacksonville controller, who approved deviations as needed. The pilot replied that no deviations were necessary at that time. At 8:27:36 the pilot requested to deviate west of course to fly through "a little hole." The PA-46 was just east of Daytona Beach, Florida. When asked how far he needed to go, the pilot said, "About 10 or 12 miles." ATC then asked if the flight could fly 170 degrees. The pilot said that he could not and that the aircraft was "blocked in on the east side." The controller acknowledged, approved the deviation, and instructed the pilot to proceed direct to Marco Island when able. The flight turned about 20 degrees to the right.

At 8:29:44, the flight was handed off to the next sector. The pilot acknowledged and stated, "OK...(unintelligible) and a little hole here." Radar indicated the Malibu was in the vicinity of some heavy weather returns. The controller stated that he observed the weather area and was aware of other pilots deviating away from it on the west side.

At 8:29:49 the ground track plot showed the pilot had observed a three- to five-mile gap between two thunderstorm clusters and attempted to fly through an area of light radar echoes between the two large areas of heavier echoes. At 8:29:53 the pilot made an unintelligible transmission to the controller.

At 8:33:36 the aircraft departed level flight and began an uncontrolled descent from FL260. The radar showed that one cluster of thunderstorms had moved east-northeast, but other level 3 and 4 thunderstorms were present in the vicinity.

Witnesses observed the airplane come out of the clouds, about 300 feet above the ground, in a spiral, with the right wing missing. The NTSB data showed that the PA-46 descended rapidly in a left turn with a maximum descent rate of 20,700 feet per minute. According to ballistic trajectory estimates, the radar pod and wing panel separated at around 26,000 feet just prior to the airplane's rapid descent. The private pilot and two passengers were fatally injured. The right wing and horizontal stabilizer had separated from the aircraft while the main wing spar was broken inside the cabin area. The forces of nature are not to be trifled with.

Pilot decision making

The pilot held a private certificate with an instrument rating that was issued almost 10 years earlier. The last dated entry in his logbook was on April 10, six years prior. FAA records indicated that he completed a 16-hour Piper PA-46T Malibu ground and flight recurrent course less than a month before the accident. The pilot's flight time was estimated at 2,800 total flight hours, with 380 hours in the Malibu.

The Malibu flight instructor who last gave the pilot training also provided initial training in the PA-46-310 Malibu in 1996. The pilot was upgrading from a Mooney to the Malibu, his first pressurized aircraft. The CFI noticed during this training and subsequent refresher courses that the pilot "pushed himself dangerously close when making weather decisions in this class of airplane." He seemed to "lack a healthy respect" for the destructive forces of thunderstorms, and took "delight" in how close he could come in pushing the envelope. The CFI cautioned him only two weeks prior to the accident that his decision making was deficient and he needed to exercise "greater care" when flying his JetProp in and around "adverse weather systems."

It's instructive to look at the weather that existed that June evening. The nearest weather reporting station was about 12 miles to the southwest of the accident location. The reported weather at 8:19 p.m. was winds 180 degrees at 9 knots; visibility seven miles; sky condition scattered 2,000 feet, broken 5,500 feet, and overcast 7,000 feet; temperature 24 degrees Celsius; dew point 23 degrees C; altimeter setting 29.85 inches of mercury.

Sounds pretty benign at the local level, but a larger view showed a prefrontal trough extending northeast to southwest across northern Florida ahead of a cold front located over southern Georgia and Alabama. There was a broad southwesterly flow over the state with plenty of heat and moisture, as indicated by the narrow temperature and dew-point spread from the local station.

More to the point, there were no less than six convective sigmets that were in effect either just before or just after the accident, all indicating a large area of thunderstorms moving from the west at 10 to 25 knots with tops to FL450. It comes as no surprise that in late afternoon and early evening there will be fairly large areas of thunderstorms in summertime Florida, and many other parts of the country as well. Isolated air-mass storms are one thing — widespread areas of convection are something else. It's been said many times that sneaking between cells is not a good idea and, until we are able to see turbulence, not precipitation as is shown currently on radar, a wide berth is the only sensible way to handle organized systems.

This pilot clearly knew about the weather, but obviously did not heed the recommendation of the flight watch specialist to deviate to the western part of the state and avoid the entire area. It would have extended his flight by perhaps 30 minutes and might have necessitated a fuel stop.

The Malibu has been particularly susceptible to thunderstorm-related accidents (see " Convection at Altitude — PA-46," page 69) but, I hasten to add, through no fault of the aircraft. The FAA did a special certification review of the aircraft some years ago when there were quite a number of in-flight breakups. Nearly all were associated with convective weather and pilot decision making was assigned as probable cause. Radar, datalink, ATC, and flight watch recommendations can all help us to know where the bad stuff is. It's up to us to avoid it, even in a high-flying aircraft.

Bruce Landsberg is executive director of the AOPA Air Safety Foundation.

Understanding One Another

Do you understand air traffic control (ATC)? Does ATC understand you? During convective season, understanding one another is essential. The AOPA Air Safety Foundation has prepared an online course for pilots on flying in thunderstorm season. Visit the Web site ( www.aopa.org/safetycenter/courses/trw/).

Convection at Altitude — PA-46

According to the AOPA Air Safety Foundation accident database, the past 20 years show that the number of PA-46 accidents that are likely related to convective encounters has slowed somewhat from an almost annual occurrence in the late 1980s and early 1990s to roughly every other year beginning in 1999. There was a total of nine accidents during the 20-year period that have all the earmarks of a thunderstorm encounter. It is worth noting that about half occurred at flight levels 220 to 240, with the balance in the mid- to low teens.

Pilot experience varied from very high-time aviators to some with little experience in high-performance aircraft and high-altitude operation. Ignorance and overconfidence are equally dangerous and a combination of the two is lethal.

The FAA rightly gives us responsibility for accepting risk but provides regulatory guidance under FAR 61.31. Pilots operating PA-46s in the flight levels must be instrument-rated and have complex- and high-performance-aircraft endorsements from a CFI if they have not had prior experience in such aircraft before August 1997.

There is an additional requirement for high-altitude training including aerodynamics and high-altitude meteorology for pressurized aircraft that have a service ceiling above 25,000 feet. This does not apply to the factory PA-46 since FL250 is the maximum operating altitude and, therefore, it is incumbent upon smart pilots to voluntarily participate in such training. It is required for the JetProp DLX and is included in the transition course.

In addition to the usual physiology training, considerable study should be devoted to thunderstorm detection and avoidance, negotiating with ATC, use of airborne weather radar, high-altitude winds, turbulence, and icing. Some newly qualified "flight level" pilots (and a few old-timers) do not have a full understanding or, perhaps, appreciation of the high-altitude environment. This type of training is not easy to get and some practical experience is highly recommended. The airlines and most corporate captains and first officers spend years and hundreds, if not thousands, of hours learning the finesse required when operating airborne radar and of not dancing too closely with thunderstorms. — BL