September 1, 1999
By Bruce Landsberg
Fog has a way of materializing out of thin air. It doesn't happen that frequently, but it can catch unwary pilots. In the 1998 Nall Report, the AOPA Air Safety Foundation found that VFR flight into instrument meteorological conditions (IMC) continues to create the most dangerous accidents. The fatality rate is nearly 90 percent in these cases. Many involved pilots who took high risks and an optimistic view of conditions. In this review, we'll look at what appear to be two cautious pilots who didn't anticipate sudden changes in weather conditions. They failed to heed some subtle clues and made a logical decision that turned out to be wrong.
A weather briefing at 8 a.m. reported VFR conditions, and the two pilots filed a local VFR flight plan. A weather observation at 8:55 a.m., taken about 35 minutes before departure, reported a few clouds at 500 feet with ceilings broken at 4,300 feet and 20,000 feet. The 24-hour forecast was for visibility greater than three miles in mist with ceilings of 3,500 feet broken and 5,000 feet overcast. At 9 a.m., the forecast was amended for the hour between 9 and 10 a.m. to be visibility four miles in mist with scattered clouds at 200 feet and ceilings overcast at 500 feet. It is unknown whether the pilots received this update. If they had, it might have changed their plans.
The humidity was high with only a two-degree spread between the temperature and dew point. According to the tower controller, "The weather was nice prior to the [pilot's] calling and stating the weather was not what they expected. It was nice and clear in all quadrants with a little overcast/light fog to the east."
At 9:33:48, the Cessna 172 was cleared for takeoff from Runway 8. At 9:35:02, the pilot transmitted, "Tower, this is Cessna [call sign omitted], uh...we're going to come back for landing...uh...the ceiling is lower than we thought." The tower controller acknowledged the call and instructed the pilot to report when the airplane reached the base leg of the traffic pattern.
The tower controller observed the airplane over the departure end of Runway 8. It then entered a climbing left turn and disappeared from view. When radar indicated that the airplane was at 500 feet, the controller attempted to regain visual contact with the airplane by alternately scanning the radar display and the sky over the northern end of the airfield. At 9:35:40, the controller heard unintelligible cries over the tower frequency and stated, "A 'whoo' sound was made by the [aircraft]. I attempted communication with the [pilot] on tower frequency and on VHF emergency frequency. I still could not physically see the [aircraft], and the target on the radar went from 500 feet to 300 feet, to no target at all, almost instantly."
Several witnesses near the crash site reported that they heard the airplane overhead but could not see it through the low cloud cover. One witness in the parking lot adjacent to the crash site stated, "I heard the noise of a plane much louder than a normal plane. It was very low, but I couldn't see it. The fog was maybe four or five floors above a nearby 10-story building. When the plane came out of the clouds, she was in somewhat of a vertical drop."
According to another witness, "The clouds were really dense, really thick. The engine was running when it came out of the clouds. It sounded like it was too loud. I knew it was going to crash because it was in an awkward position. It never rotated — it came right on down. They must have been totally dependent on their instruments because they couldn't have seen anything."
Recorded radar data evaluated by the NTSB's Office of Research and Engineering revealed that the aircraft [took] off from Runway 8 at approximately 9:34:50, climbed through 300 feet msl while maintaining runway heading, then entered a climbing left turn at approximately 70 knots' groundspeed. The left turn continued until a west-southwest heading was established; then the airplane climbed abruptly to 800 feet msl, followed by a rapid descent to the accident site.
The Cessna struck terrain and an electrical substation building about one mile from the center of the runway, near the downwind leg of the traffic pattern. Both pilots were killed, and the aircraft was destroyed. Examination of the wreckage showed no malfunction of the instruments, flight control system, or the engine.
The pilot in the left seat held a private pilot certificate but was not instrument-rated. Her medical certificate had expired, and her logbook showed 146 hours of total flight experience, four hours of simulated instrument experience, and no actual instrument flight. She had 39 hours of flight time in the 172.
The pilot in the right seat held a commercial pilot certificate with an instrument rating. She had a valid medical certificate; her logbook revealed that she had 448 hours of total flight experience, including 105 hours in the 172. She had 133.6 hours of simulated instrument flight experience, including 2.5 hours in the 30 days prior to the accident. The log also showed 17.4 hours of actual instrument flight experience. In the 90 days preceding the accident, she recorded one hour of actual instrument flight experience, of which 0.2 hours were logged in the previous 30 days.
Seven months before the accident, the right-seat pilot had enrolled in a flight instructor certification course. Her training folder showed that before the accident she had received 68.3 hours of flight instruction during the course. Two of the training flights were devoted to instrument-flight maneuvers. The first flight took place about five months before the accident and was 1.7 hours long. The second one occurred about two weeks before the accident and lasted 2.0 hours.
In many VFR-into-IMC accidents, the forecast is substantially correct, and the accident occurs miles away from the observation point. In this case, onsite weather was available, and the clues that the field was about to go IFR were there — but subtle. Weather reported at the time of the accident was measured ceiling 300 feet overcast and four miles' visibility with mist. Winds were from 060 degrees at 10 kt. The temperature was 50 degrees Fahrenheit, and the dewpoint was 48 degrees. In a written statement, the tower operator said, "The weather was really foggy almost immediately after the [aircraft] went down." The airport's weather observer, making a 9:30 a.m. local observation, said that he observed clouds approach from the northeast and, "I was aware of the forecast and knew these conditions were unexpected."
Local geography plays a significant role in understanding how microclimates behave. The airport has a 10,000-foot concrete runway that is 150 feet wide and oriented 080 to 260 degrees magnetic. It is located on a peninsula bordered by a large body of water immediately to the east and a river to the south and west. With that amount of moisture nearby, fog banks would tend to form quickly or blow in off the water. Most airports in close proximity to water experience similar conditions. Note that in this case the wind was coming off the water, and the air was saturated. The weather may be clear, but it can change quickly.
We asked a local pilot, familiar with the area and this accident, what he would have done differently. He had been flying earlier in the day and noted, "The weather was not that great. There were layers of clouds about, and even though it was VFR west of the field, it wasn't the best of conditions." One could say that in hindsight the flight should not have launched, although many of us might have believed it was good enough to fly locally.
So, what do you do when surprised by instrument conditions? That depends. En route, at altitude, a 180-degree turn, possibly with a slight descent, should get one safely back to VFR in the shortest time.
But what about during initial climb after takeoff? The element of surprise should not be underestimated, and there would be a strong desire to return to VFR conditions just below and behind. Sometimes the best answer is counterintuitive. Turn and go down seems logical, but straight ahead and up is safer. Transitioning from visual to instruments typically takes a few seconds. The eyes must shift from the windshield to the gauges, and the brain must begin to process abstract instrument images that were, just seconds ago, the real world. A touch of vertigo would be possible, especially with multiple control inputs such as an abrupt, climbing left turn. The NTSB radar track showed the Cessna turning beyond the downwind heading and climbing steeply.
Now consider a VFR pilot in the left seat and a CFI-in-training in the right seat. Not only is there a transition from visual to instrument flight, but there is also a probable transfer of aircraft control. There would likely be confusion in the cockpit and a need to settle into an instrument scan — quickly. When en-countering such a situation, the only viable option is to minimize the distractions and fly the airplane. Climb straight ahead — in this case there were no obstacles over the water. Get the scan settled down first and the aircraft trimmed if needed. This may take 30 seconds or so. Then call the tower and advise them that you're in the clouds and need an IFR clearance with vectors for an approach or to clear weather. This is not a time for chart reading or radio navigation tuning. It is not a time to worry about the legal ramifications — those will all be worked out in due time. Just maintain control — shed the distractions and focus on keeping the aircraft upright and moving away from the ground. Talk to yourself if needed, and get a grip. Conflict with another IFR aircraft is possible, so don't take too long to advise ATC that you need help — but positive aircraft control is first.
I belonged to a flying club that was based at an airport on the West Coast. The club had procedures for dealing with the local microclimate. Fast-forming fog was common, and I was not instrument-rated. If we got caught close to the ground, the drill was as described above. If the fog formed while we were away from home base, any one of several nearby inland airports could serve as an alternate. We just called the club, and it would send someone out to retrieve the aircraft as soon as the weather cleared — no penalty, no foul. The fact that there was a plan to follow took a lot of the angst out of our flying and to my knowledge, there was never an incident, although I did have to work out some ground transportation on occasion. That is an acceptable outcome of an unexpected IMC encounter. It's not necessarily convenient, but it sure beats the alternative.
See also the index of "Safety Pilot" articles, organized by subject. Bruce Landsberg is executive director of the AOPA Air Safety Foundation.
Pilot Training and Certification,
Safety and Education,
VFR into IMC,
FAA Information and Services
March 7, 2014 ePilot Training Tip: 'Arrival or through flight'
Garmin has announced an upgrade making new features and options available to operators of G1000-equipped King Airs in the 200/250/300/350 series.
With a closing speed of about 900 knots, Air Force pilots on a training mission have seconds to aim and shoot heat-seeking and radar guided missiles at a drone target. Their success came from repeated rehearsals. But as author Larry Brown writes, “there is nothing like the real thing to gain experience.”
AOPA thanks our members for their continued support in protecting the freedom to fly.