Deadly decisions

As the snow steadily accumulates, the hunter hears the sound of a very loud engine that lasts for several seconds, then goes silent. The noise is distinctly different from the usual sounds of airplanes taking off and landing at the airport. Wondering what it could have been, he attempts to look in the direction of the sound he heard, but he can’t see anything because of the blowing snow.

That evening, the hunter learns that an airplane may have crashed on the 140-acre property he owns adjacent to the airport. He is asked to assist first responders in their search for the airplane and he helps them navigate a cleared trail across his land. They search deep into the cold, dark night, but then must suspend their efforts because of the inclement weather. Early the following morning, the search party locates the wreckage of a Grumman GA–7 twin-engine airplane about a half mile from the threshold of Runway 25 at Jamestown Airport. All aboard have died in the accident.

Three hours earlier

It was early afternoon when the pilot of N791GA, his sister, and a family friend returned from their lunch ready to board their rental airplane parked at Burlington/Alamance Regional Airport (BUY), located three miles southwest of Burlington, North Carolina.

The pilot had filed an instrument flight plan from Burlington to Jamestown. He had listed Franklin’s Venango Regional Airport (FKL) in Pennsylvania, located about 54 nautical miles southwest of Jamestown, as an alternate. The route had been planned at a cruise altitude of 9,000 feet msl with an estimated time en route of two hours and 19 minutes before landing at Jamestown, where the airplane was based. He had also called his flight instructor to discuss the planned flight, weather conditions, and making good decisions.

N791GA, a twin-engine Grumman American Cougar, had been refueled during lunch and was ready to depart to the north into a partly cloudy and cool sky. It was 2:58 p.m. local time.

Once the flight proceeded on course, it leveled off at its filed cruise altitude. At 5:09 p.m., N791GA began a descent from 9,000 feet msl, and 10 minutes later, it reached 4,100 feet msl. The pilot checked in with Buffalo Approach at 5:22 p.m. When asked for a pirep, the pilot mentioned being in and out of clouds, updrafts, downdrafts, snow, and light to moderate turbulence. He did not report ice accumulation. The pilot then requested and was cleared for the RNAV GPS Runway 25 approach at Jamestown Airport. It was 5:30 p.m. The Grumman turned to a northeasterly heading to set up for the approach, then banked to the west to intercept the initial approach fix.

At 5:40 p.m., when N791GA was flying inbound on the approach, the controller instructed the pilot to change to the airport’s advisory frequency and report back to cancel his IFR flight plan. It was the last transmission with the Grumman pilot before radar contact was lost at 5:47 p.m., about one mile northeast of the Runway 25 threshold.

Analysis

In its final report, the National Transportation Safety Board (NTSB) concluded that the accident’s circumstances were consistent with the pilot continuing to descend below the minimum descent altitude (MDA) in IMC, which resulted in controlled flight into terrain. But some key factors contributed to this tragic outcome.

During the NTSB investigation, the pilot’s flight instructor mentioned that the pilot called him before the accident flight, and they had specifically discussed making good weather decisions and avoiding icing conditions. The pilot had assured him he would. On October 28, he had departed on the trip from Jamestown to Burlington a day early—specifically for weather reasons.

The pilot, who held a commercial certificate with a multiengine instrument rating, had obtained a weather briefing and filed an IFR flight plan via ForeFlight. However, it appears he did not view any in-app weather imagery before the flight.

The graphical forecasts for aviation products issued before the accident flight and valid at 7 p.m. indicated VFR surface visibilities, a greater than 60 percent chance of snow showers, and a surface wind from the west at 15 knots, gusting to 25 knots. A graphical airmet (G-airmet) was valid for IFR conditions and the cloud forecast valid for the accident time indicated scattered clouds around 2,300 feet msl, overcast skies at 3,500 to 4,500 feet msl, and cloud tops to 13,000 feet msl. In addition, G-airmets were valid for mountain obscuration and icing conditions.

Jamestown Airport’s terminal aerodrome forecast (TAF), issued at 12:38 p.m. and valid for a 24-hour period beginning at 1 p.m., initially specified marginal VFR conditions with 6 miles visibility in light rain, winds out of 270 degrees at 17 knots, gusting to 29 knots, and an overcast ceiling at 1,500 feet agl. At 6 p.m., conditions were forecast to be IFR, with 2 miles visibility in light showers and snow, winds out of 280 degrees at 15 knots, gusting to 27 knots, and a broken ceiling at 2,000 feet agl.

While N791GA was en route to Jamestown, the National Weather Service in Buffalo, New York, issued an area forecast discussion at 3:44 p.m. It indicated that rain would change to snow between 3 p.m. and 5 p.m. near Jamestown and winds were expected to be gusty with low IFR visibilities in snow and blowing snow.

At 4 p.m. that Sunday, the pilot’s flight instructor checked the FlightAware flight tracking website and he noticed that N791GA was inbound to Jamestown, New York. He had hoped that the pilot would divert, but unfortunately, that was not the case. The instructor even texted him in-flight to encourage a turnaround for weather but to no avail. Alarmed about the situation, the instructor then drove to the airport and listened to ATC communications and the local CTAF. He was deeply concerned about the weather conditions and the pilot’s choice of the Runway 25 GPS RNAV approach.

A special METAR issued three minutes before the accident showed winds from 280 degrees at 16 knots with gusts to 21 knots, one-and-three-quarter miles visibility in light snow and mist, a few clouds at 500 feet agl, an overcast ceiling at 1,200 feet agl, and a temperature of 0 degrees Celsius with a dew point of minus 1 degree Celsius.

Two minutes after the accident, a special METAR indicated winds from 280 degrees at 20 knots, gusting to 30 knots, three-quarter mile visibility in light snow and mist with scattered clouds at 500 feet agl, and a broken ceiling at 1,000 feet agl. Skies were overcast at 1,700 feet agl with a temperature of minus 1 degree and dew point of minus 2 degrees Celsius.

N791GA’s last ADS-B target was recorded at 5:47 p.m. about one mile from Jamestown’s Runway 25 threshold, indicating an altitude of 2,250 feet msl. The RNAV GPS approach to Runway 25 stipulates a 2,100-foot msl minimum descent altitude at 1.1 nautical miles from the airport with one-half mile minimum visibility, a local altimeter setting, and lateral guidance but no vertical guidance using LNAV minimums. Using vertical guidance, the LVP MDA is 1,921 feet msl with one-half mile minimum visibility and a local altimeter setting.

In the end, it’s unclear why the pilot decided to embark on this flight and press on in worsening weather, especially after carefully consulting with his flight instructor about the weather en route and at his destination.

Not only were conditions prime for icing in western New York, but the initial approach fix (IAF) for the Runway 25 RNAV GPS approach was located far out at 12.4 miles from the Jamestown airport, prompting an early descent from smooth conditions above and increasing the flight’s time in inclement weather. Had the pilot opted for vectors to the Runway 25 ILS approach, the distance from the IAF would have been cut almost in half and could possibly have even prevented this tragedy. 

In addition, the pilot reported being in and out of clouds, updrafts, downdrafts, snow, and light to moderate turbulence. Flying in and out of clouds, he may have seen the runway and decided to not lose sight of it, a classic mistake in low IMC that can lead to diving below the descent profile.

Finally, the dark terrain below the descent path may have contributed to a black hole illusion on the long, straight in approach, which combined with turbulence and low visibility could have led the pilot to think he was high on the descent, therefore descending prematurely and flying into the darkness below.

[email protected]

airsafetyinstitute.org