September 20, 2007
Most ILS approaches take from three to five minutes to complete after crossing the final approach fix (FAF). What you're about to read is the story of a pilot who was cleared for an ILS approach just before his aircraft disappeared from radar, only to pop back up on the screen approximately 27 minutes later. For another seven minutes, controllers, while watching the aircraft, attempted to make contact with the pilot without success. Then, the aircraft disappeared from the radar screens for a final time.
|Piper PA-32 Lance
The pilot's day began just before 5:50 a.m., when he departed Arkadelphia, Ark., in his Piper Lance for the 60-mile flight to the Downtown Airport (F43) in El Dorado, Ark. This being an early January 1999 morning, it was still dark and IFR weather conditions existed along the route of flight.
The accident report does not mention what type of weather briefing the pilot received, nor does it go very deeply into the weather conditions. Downtown Airport is a small, general-aviation field with a single, 3000-foot-long, north-south runway and is surrounded mostly by woods, with a neighborhood to the north and a highway south of the runway. Because of its small size, the airport has neither an instrument approach nor any official weather reporting capabilities. The nearest airport with weather is El Dorado's South Arkansas Regional Airport (KELD), located about eight miles to the west. At 6:23 a.m., it reported a broken ceiling at 400 feet, an overcast ceiling at 1,000 feet, seven miles visibility in light rain and winds from 300 degrees at 12 kts. gusting to 19. A weather report at 6:50 a.m. showed similar conditions, this time with an overcast ceiling of 400 feet and a remark that the rain had ended 20 minutes prior. The temperature and dew point were both 9 degrees C.
The route south to El Dorado took the flight through Forth Worth Center's airspace and, at 6:07 a.m., the pilot asked the Center controller for a lower altitude. The controller cleared the pilot to descend to 2,000 feet and read him an earlier weather report -- although he did not identify when it was issued -- that stated that the El Dorado weather was seven miles visibility in light rain showers, the winds were from 320 degrees at 12 kts. and that the ceiling was 2,500 feet broken and 3,500 feet overcast. He then told the pilot to report the Downtown Airport in sight.
Based on the 6:23 and 6:50 a.m. weather reports from KELD, the weather that the controller had given the pilot would appear to have been outdated, as conditions around El Dorado were deteriorating. So it was no surprise that the pilot called the controller at 6:11 a.m. and told him that he could not see the airport. Faced with limited options, the pilot asked for an instrument approach into KELD. The controller instructed the pilot to fly a 270-degree heading and join the ILS approach for Runway 22.
A minute later, the controller advised the pilot that radar contact with the aircraft was lost and asked if he was able to navigate directly to the outer marker. The pilot responded that he could and the controller cleared the aircraft for the ILS approach to Runway 22. He asked for a report when the aircraft was established on the approach, which the pilot made a minute later.
The controller cleared the pilot from the frequency with instructions that he was to cancel on the present frequency, or with flight service. The pilot acknowledged the instructions in what would be his last transmission controllers would ever hear from him.
|Sectional Chart with South Arkansas Regional Airport (KELD), Downtown Airport (F43) and the town of Smackover (top).
Nothing was heard or seen from the pilot until 6:39 a.m., when the Lance popped back up on Fort Worth Center's radar screens. This time, however, the airplane was south of Downtown Airport, nowhere near KELD or the ILS 22 approach. The controller began a series of radio calls to the aircraft but never received a response. The controller then asked the pilot of a commuter airplane if he could communicate with the Lance pilot. The commuter pilot called twice but never got a response.
According to radar returns, the airplane made a climbing right turn to the north at the time the controller was trying to make contact with the pilot. The airplane then made a gradual left turn to the west over the town of El Dorado. At that point it was at 3,200 feet. The airplane then made a slow turn to the north. The last radar plot was at 6:48 a.m. and indicated that the aircraft was at 3,000 feet and still in a right turn.
Nine minutes after the aircraft disappeared from the radar for second and final time, the controller called the Jonesboro Automated Flight Service Station and asked if the Lance pilot had cancelled his IFR clearance. He had not. The controller then contacted other controllers in adjacent sectors and advised them that he had no idea where the aircraft was or where it was going.
At 7:14 a.m., the controller asked the pilot of an airliner flying in the vicinity to attempt to contact the Lance. Again, there was no response. Eventually, search and rescue was notified and the aircraft was located two days later in a heavily wooded area in Smackover, Ark., about six miles northeast of the South Arkansas Regional Airport, and about 10 miles northwest of the Downtown Airport. The pilot, the only person aboard the aircraft, was killed in the crash.
NTSB investigators conducted telephone interviews with witnesses in an area about two miles west of the Downtown Airport and about six miles southeast of the Regional Airport who reported hearing an airplane fly over their houses "numerous times at a very low altitude" around 6:30 a.m. Many of the witnesses said that the airplane woke them from their sleep. A pilot-rated witness stated that he thought the plane was going to "fly through his house" so he woke his wife. They heard the airplane pass several times while "climbing and descending" and thought "the pilot was in trouble."
According to FAA medical records, the 69-year-old pilot was issued a third-class medical certificate on Jan. 24, 1997. The last entry in the pilot's logbook was dated April 19, 1998. The most recent biennial flight review was dated April 16, 1988. There was no record of an instrument proficiency check or any record of instrument currency. According to an application for aircraft insurance, the pilot had accumulated 3,800 hours of flight time, but it could not be determined how much of that time was in the accident aircraft make and model. No other pilot records were located during the investigation.
Examination of the aircraft at the accident site and after its removal provided no indication of onboard mechanical or instrument failures.
The NTSB cited the pilot's spatial disorientation that resulted in a loss of control and an ensuing stall/spin as the cause of the accident. Other factors were the low ceilings and the nighttime conditions.
What really happened aboard the Lance, we'll never know. It's possible that the pilot simply got lost or disoriented on the approach, wandered around for 27 minutes before climbing high enough to be seen by radar, then lost control. But that doesn't seem likely.
A more plausible scenario is that the pilot shot the approach to KELD and when he broke out, turned back towards Downtown Airport, probably with the hopes that he could easily find the airport and land. But things didn't turn out that way.
The reported weather at KELD varied between a 400-foot broken ceiling and a 400-foot overcast ceiling, with the weather trend deteriorating as the morning went on. The visibility below the cloud base was good at seven miles, although there had been reports of rain in the area. Given that the Lance pilot was not able to locate the Downtown Airport in the existing weather conditions, it's possible that the weather at the Downtown Airport or between the two airports was worse than that at KELD.
It should be remembered that the weather report the pilot received from the controller suggested much more favorable conditions. It's impossible to determine if the pilot ever listened to the ASOS at KELD. At the very least, though, the pilot should have suspected that the weather report he received from the controller was not accurate, or that the weather had deteriorated, when he initially flew over the Downtown Airport at 2,000 feet and was unable to locate it.
The rules are clear regarding descent on an IFR clearance into an airport that does not have an instrument approach. ATC can only authorize a descent to the minimum vectoring altitude in the vicinity of the airport. If the pilot is unable to make visual contact with the airport, he must then proceed to an airport where the weather is VFR or there is a valid instrument approach.
It should be noted that there's nothing wrong with the basic intent of what the pilot was probably trying to do, which is to fly an instrument approach into one airport, break it off when entering VMC and proceed on to a VFR-only airport.
To do that, however, one must cancel the IFR clearance first before breaking off the approach, which the pilot never did. One must also be able to maintain the legal VFR cloud clearance and visibility requirements at all times. With a 400-foot ceiling and a route that would have taken him over a congested area, this would have been impossible to do.
The temptation to scud-run may explain how the airplane ended up back on radar south of Downtown Airport. But it doesn't explain what transpired in those 27 minutes between radar hits, because the two airports are only about eight miles apart.
The Piper Lance was equipped with an IFR-approved GPS, so it is likely that the pilot was using the GPS to position the aircraft over the Downtown Airport. The terrain between the two airports is relatively level, and there is a highway that runs between the south end of KELD and the south end of Downtown Airport. FARs aside, if the visibility had been as advertised, once the pilot broke out of the clouds and using the GPS, he should have been able to fly directly to the Downtown Airport. Is it possible that the pilot never broke out of the clouds or that the visibility between the two airports was not very good?
If the aircraft didn't break out of the clouds, it may be that the cloud deck was even lower in some areas than what was reported at KELD. Or perhaps the pilot did not fly a good ILS approach. It was still dark and it may be that he did break out of the clouds but was unable to locate anything recognizable on the surface.
Then there's the question of the pilot's instrument proficiency level. Just because the NTSB couldn't find any records attesting to his currency, that doesn't automatically mean his proficiency was lacking. It's interesting to note that witnesses reported hearing what sounded like an aircraft in trouble at 6:30 a.m. but the airplane didn't crash until around 6:48 a.m. It doesn't make sense that the pilot became spatially disoriented at 6:30 or before and continued to fly the airplane for a considerable time after that until it crashed. NTSB accident reports seem to indicate that most pilots who become spatially disoriented tend to lose control of their aircraft within about five minutes.
Avoiding The Trap
While we will never know what happened in this particular accident, especially in those 27 minutes when the airplane was off the radarscopes, it is entirely possible for you to prevent this type of accident from happening in the future.
First, if you file an IFR flight plan to an airport that does not have an instrument approach, you are required to file a legal alternate. Pick one where the weather is expected to be above your personal limitations. If weather reports in the vicinity of your destination indicate that you will not be able to make visual contact with the airport, consider going directly to your alternate.
Second, you should never plan on breaking off an approach at one airport and fly to another unless you know the weather conditions are at least VFR along the entire route. In some areas where terrain is a factor, you will need more than basic VFR minimums.
Under no circumstance should you ever break off an approach to fly to another airport without canceling your instrument clearance. If the weather conditions are marginal and you don't want to cancel your instrument clearance, then you must land at the airport where you are making the instrument approach. As you can see, floundering around in the clouds while trying to locate an airport that does not have an instrument approach invites all kinds of problems.
More accident analyses are available in AVweb's Probable Cause Index. And for monthly articles about IFR flying including accident reports like this one, subscribe to AVweb's sister publication, IFR Refresher.