Recently, a former airline colleague was killed when his Experimental aircraft impacted rapidly rising terrain while in level flight. It is easy to know exactly where the airplane struck the mountain with publicly available radar tracking. Through such data, friends and family awaiting the overdue pilot determined the location of the accident and alerted search and rescue authorities. The wreckage was spotted the next day and eventually reached through an hours-long hike through steep and primitive terrain.
The pilot was talented, highly trained and heavily experienced. He had a long military fighter pilot career, a long second career as an airline pilot and a lifelong passion for general aviation, even achieving medal status at the Reno Air Races.
All we know is that the aircraft struck the mountain. Understanding why the accident happened—if, indeed, it’s possible to know why—will have to await a formal investigation. I will resist speculation here with one exception: The accident has all the hallmarks of controlled flight into terrain—CFIT.
Training Starts Early
Ever since pilots started flying air machines from Point A to Point B, there have been instances where airworthy aircraft under full pilot control met an early and tragic ending involving terrain not designated as a runway. The term terrain in this instance is defined as “the ground, a mountain, a body of water or an obstacle.” Obstacle is further defined as natural or artificial permanent or temporary structures such as towers, transmission lines, bridges, buildings and so on.
CFIT accidents can occur in level cruise flight, in maneuvering flight and in the course of the approach and departure phases of flight. In many scenarios, the pilot or crew is unaware of the impending disaster until too late. Once airline and most corporate flying moved to cruise levels far above the highest terrain, the CFIT exposure window migrated to primarily the approach and departure phases of flight. The safety record steadily improved, but the danger was unfortunately not eliminated. CFIT continues to be at or near the top of fatal accident causes as it has throughout the history of powered flight. Airline currency training almost always includes discussion and review on CFIT causes and prevention.
Common and often repeated elements of CFIT occurrences fall into broad categories such as navigational/procedural errors and pilot misjudgment, as well as distraction or incapacitation. Like any tragic accident, the causality often turns out to be an insidious combination of several factors. Just when investigators think they’ve seen it all, some new accident gives birth to entirely new issues that had yet to have been considered. With just about every accident, we attempt to learn from the tragedy to improve “the system” in a process that is often referred to in the gallows humor of the industry as “tombstone engineering.”
Even though aviation in general and airline operations, in particular, are incredibly safe and generally moving in the right direction, there have still been several notable CFIT accidents. Early in my career, we moved to St. George, Utah, where just a few years previous, a U.S. Air Force B-52 had crashed into a mountain about 20 miles away. I was able to see the site of the crash several times by air and even flew with a pilot who had hiked to the site and wore a belt buckle fashioned from the debris he found there.
Later we moved to the foothills of Davis County, Utah, just a few miles away from where a United DC-8 had flown into a mountainside during a holding pattern that went awry while trying to troubleshoot an electrical problem. Even though the aircraft was under radar control, the crew were off frequency talking to their maintenance while the controller frantically tried to get them to turn and climb. The controller only managed to do so a few seconds before impact. For many years after this accident, when the seasonal vegetation and ambient lighting were just so, a ghostly image of the impact scene, including imprints of the wings and all four engines, could be distinctly seen.
From our foothill home there, we could also see the exact spot of another CFIT accident. This one was one of the very early flights of a United B-727, Flight 227, where a crew transitioning from props to turbines failed to account for the inherent turbine power lag and subsequently failed to arrest an overly steep descent. The Boeing crashed short of the runway, killing 43 of the 91 aboard.
These accidents and other similar ones all contributed to additional training and procedural changes as well as early generation terrain warning ground proximity warning systems (GPWS) designed to avoid if not eliminate CFIT.
The progeny of those early warning systems, now mostly called EGPWS systems (for enhanced), continue today in operation in airliners, military and corporate aircraft worldwide. Technology has undoubtedly improved safety, and accidents are declining. Still, as tragic events like the American 965 Boeing 757 crash near Cali, Columbia; the fairly recent UPS 1354 Airbus A300 crash on approach into Birmingham, Alabama; and the similar timeframe Asiana 214 B-777 crash into the seawall in San Francisco portend, even the most advanced systems and avionics won’t prevent CFIT if proper procedures aren’t also correctly employed.
On a smaller scale and closer to home, in November of 2011, a Turbo Commander departed my home base of Falcon Field in Mesa, Arizona, for a quick flight to Safford. Just minutes after takeoff, the airplane flew straight and level into a nearly sheer cliffside in the Superstition Mountains, where they all perished. Speculation was that the pilots leveled off to stay under the Class B airspace and lost situational awareness of the impending danger. Additionally, it appeared that the left-seat pilot left the aircraft under the control of his partner while he went back to address an issue with passengers. His remains were found in the back and not in the left-front seat. Since that accident, the FAA changed the boundaries and altitudes of the Class B airspace east of Phoenix to allow a larger window for aircraft trying to avoid the airspace to climb earlier and fly higher prior to the abrupt terrain changes further east.
One example that strikes particularly home to me involved SkyWest 5855, a Fairchild Metro III that was executing a challenging non-precision VOR/DME approach into Elko, Nevada. The approach was based on the Bullion VOR, which is situated atop a mountain south of the airport and serves as a critical stepdown point on the approach. For whatever reason, the crew descended early and impacted the top of the mountain very close to the VOR. The Metro went airborne again briefly and then tobogganed in deep snow down the north slope of the mountain until impacting a large boulder, which while injurious to both the aircraft and occupants, may have actually saved their lives by preventing the aircraft from careening further into the icy river at the bottom.
What is particularly notable about this event is the extreme rarity of an aircraft impacting a mountain to the degree that it terminates the flight, yet every occupant survives. Several guardian angels earned their wings on that one.
I knew both of the pilots personally, inspected the wreckage when it was eventually relocated to a hangar in Elko and was involved in producing a training supplement about the accident for ground school classes. In all, it was an experience that I will never forget.
Every time a tragedy happens, lessons are learned, and those lessons often spur changes to technology and procedures. New tech, however, can be both a blessing (often) and a curse (occasionally). The danger comes if new tech inspires overconfidence in iffy flight decisions or abandonment of traditional safety practices.
Student Days
I still remember decades ago as a teen student pilot receiving ground instruction for my first cross country flight. The instructor had me take my World Aeronautical chart, draw my route and then had me circle in bright red marker the highest terrain point within a half-hour flight in any direction anywhere along my planned route. That turned out to be about 50 miles in the mighty Cessna 150 we were to fly.
That early teaching and practice may have saved my life just about a year or so later. As a new low-time private pilot, I was flying three friends back from a high school event that got delayed into a night flight with barely legal visibility. Previous instruction spurred me to go back to study the chart and to overcome the temptation to fly direct to save time. Instead, we flew a longer route using established airways over lower terrain that offered many more diversion opportunities. It turned out to be a successful flight, but one that, in hindsight, I still should never have taken. Learning to fly in Idaho gives a lot of opportunities to learn to respect high terrain.
Now that most of us, especially new pilots, are becoming children of the magenta line and paper charts have become uncommon, some of those old school practices might deserve revisiting. I remember reading about a CFIT accident not too long ago where the pilot inserted the wrong identifier into his FMS destination, leveled off and never noticed the error until it was too late.
Of all the advances in technology over the decades since the jet age, in my opinion, synthetic vision, especially when tied to aural alerts, holds the most potential for reducing CFIT events. For most of industry history, advances in civilian aviation technology have originated in the airlines and trickled down to general aviation. The exact opposite has happened with syn-viz. The relative regulatory freedom of Experimental aviation combined with the significant cost advantages have caused synthetic vision to start at the “bottom” and work its way up. With syn-viz, an entry-level aircraft can be affordably equipped with useful safety tech that most new airliners lack.
Not that they are totally analog. Most airliners delivered today are equipped with a newish technology called Vertical Situation Display, which is a 2D side view display presenting a depiction of the aircraft’s current and projected path in reference to terrain and obstacles. While useful and better than not having such a reference available, a lot of pilots don’t find it very intuitive. The aircraft is moving forward, yet the display is depicted side to side. It is also located at the bottom of the map display and outside of a normal scan pattern. The nicety of synthetic vision is that by design it is an inherent part of the natural instrument scan and much more useful in its three-dimensional form. Many pilots who have used both, myself included, much prefer syn-viz to VSD.
Airliners with EFIS displays do have a form of terrain display, but it is still only two-dimensional and heavily pixelated with low-resolution clusters of dots or symbols that at a glance can be challenging to distinguish between similar weather radar returns. In the E/A-B world, we are very fortunate to have syn-viz available and at an amazingly low cost. Nevertheless, like other tech tools like NEXRAD, we still have to understand and use them properly to take full benefit while resisting the temptation for the cool tech to coax us into flying flights that should still not be flown. Fly into the wrong storm, and you can flutter out the bottom. Fly into a mountain, and the flight ends.
Looking Ahead
Finally, I would like to end this treatise with a few common-sense suggestions to avoid CFIT. It is intended to be a starting point for study and discussion and is by no means complete.
Don’t fly fatigued. Simulator studies have shown that sleep-deprived pilots can actually perform worse than impaired pilots.
Don’t fly ill. Even a common cold can vastly lower a pilot’s tolerance to hypoxia, cause severe and distracting sinus pain on the descent and increase the temptation to use a medication that shouldn’t be used in flight.
Don’t fly impaired. Not just the commonly assumed drugs and alcohol, but there are many other environmental issues that can cause impairment. Pilots should be aware of the signs, preventions and solutions to such things as hypoxia and carbon monoxide poisoning.
Don’t fly unprepared. If there is the slightest chance that a flight will require oxygen, then make sure that all the elements required to get that oxygen into your bloodstream are available for use. Portable pulse-ox sensors are relatively cheap and very handy for checking for changes. Always have appropriate charts—paper, electronic or both—readily available and current.
Look for opportunities not to fly solo. Another pilot is obviously preferred, but even a nonpilot can be useful in keeping you aware and alert with a non-distracting conversation, monitoring progress and pointing out things that don’t look right, like mountain goats seen through breaks in clouds.
Still use crew-type procedures when solo—things such as double verifying navigational entries, checklist statuses, altimeter settings and so on.
Incorporate a status check regularly. Items to check are arrival times and fuel estimates that make sense, the status of switchable things such as pitot heat, fuel pumps, exterior lights and so on.
Know your equipment and features before you fly. One example of many is to always restore a slewed screen back to its normal navigation status before moving on to the next task. On most devices, slewing a screen, such as to take a look ahead or to the side, will freeze on a position. With the passage of time, that alternate reality to aircraft position can be a trap to an inattentive pilot that the aircraft is safely in a spot where it actually isn’t. Never fly with a screen that shows a fixed lat/long position. Look at what you want to see, and then always return the screen to its active tracking position.
Always preflight a planned flight by studying the entire route and making mental notes on such things as terrain, restricted areas, MEAs, alternates and so on.
In the end, avoiding CFIT hazards falls back on the simple priority of situational awareness. Keeping our heads on a swivel, both inside and outside, can help keep us from dangerous complacency. Recognizing and respecting the potential traps of CFIT and focusing on good habits and procedures can help protect us from tragedy.
Photos: Shutterstock, Myron Nelson.
Thank you for the excellent article. As it is declared open for discussion and I still get adrenalin rush remembering my first inadvertent flight into IMC, I would suggest another consideration to avoid CFIT. It is my rationalization about why too many experienced and qualified pilots end up this way.
Controlled flight implies ability to aviate, while “into terrain” means failure to navigate. Most experienced and qualified pilots would agree that aviating is primary to navigating as we can’t navigate too far stalled, wrong side up or structurally apart . Aviating in IMC can take up to 100% of the pilot’s processing capacity so nothin is left for navigating.
One important factor is the actual control characteristics of the airplane. It may vary not only among different types but among different configurations of the same airplane. The less stable, the more pilot capacity required. Unfortunately this can’t be practiced in a simulator and experience from heavier types could even be misleading.
Another factor is the curvature of the desired flight path as it affects both aviating and navigating workload. Flying a straight line trough IMC could pose absolutely no problem and give us false confidence , while aggressive maneuvering may one day tip the sales.
Nice article. I live here in Nome, Alaska. I know of at least several CFIT accidents into mountains, where
all occupants survived.
NE of Nome, commuter flight, Navajo Chieftain; Cape Smythe airlines. All survived, plane was dismantled
and brought to Nome.
NE of Nome, Cessna Caravan. Also Cape Smythe, flew into the ground. All survived. Dismantled and
brought in
North of Nome. Cessna 310, State Senator Olson, flew into mountain. Claimed solo flight. Dismantled and
brought in. I have a web page with pictures of this. See http://www.xalaska.com/donny091510/donny.html
NW of Nome, Stinson 108, State Senator Bob Blodget and his son, Ricky. Flew into mountain at head
of Honey Creek. Abandoned, was there for years. The wind eventually got rid of it.
NW of Nome, north of Brevig Mission. Nome Flying Service Cessna 207, 7 occupants. Flew into mountain,
all survived. Pilot TJ Yeoman. This was my airline at the time (Now Bering Air, Inc). Plane was
dismantled and taken by sleds to Brevig Mission and barged to Nome.
NW of Nome, Port Clarence, Cessna 206, same pilot TJ Yeoman, One seat empty. Not a mountain but
flew into frozen bay. Plane desintegrated, but all survived. Plane sank into bay when spring thaw came.
This is not a rare occurrence. Depends upon the angle of impact. If there is any hint of what lies ahead,
sometimes the pilot can pull up and pancake into terrain. All of the accidents above, there was no hint.
One minute everything was fine, next it all came apart.
There were several other such accidents that only SOME survived.
Ramon Gandia, Nome, Alaska
The comment on inadvertent flight into IMC struck home with me. On an early flight shortly after getting my license, I loaded my young family of four into a Cessna 150 with two small back seats, children were 2 and 3 years old. All legal and within the flight envelope with proper weight and balance. At about 3000 AGL shortly after take off on a very humid summer day with a strong haze, I flew into the side of a small thunderhead without ever seeing in against the haze. I did exactly as I was taught, Get your head inside the cockpit, trust the instruments and make a 180 and fly level and straight until you can see outside. again. It worked and my family and I lived to a ripe old age.
In a career of 60 years in the air I have made other stupid mistakes, flying VFR on top, flying into a box canyon, etc. but in each case I just followed my training (instructor was an instructor in WWI and WWII and a crop duster). Following training, assuming you had a good instructor, usually works. The Box Canyon was handled by my instruction on ending and starting a new row while crop dusting…… But I think some of the instruction pilots get today is too automated and not based on situation and recovery, just how to fly, not how to think.
On one occasion I was checking out in a new aircraft. My last landing was a disaster, almost taking out several runway lights. But I got it on and on the way to the ready room I said, “You are going to fail me on that landing, aren’t you” The response was, “No, you recognized the error and corrected it, that was what I was looking for.” That is good instruction. I never forgot the confidence it gave me instead of beating me down.
I noticed YouTube pilots seem more aware because when solo, they act as they are teaching. This appears to make them more aware of check lists etc…
Fly as if you are teaching.
Hi Myron,
Great article! and good comments!
I was a controller for 38 years. I could write a book. Great advances come out of accidents, such as the eastern L-1011 into the miami swamp. Atc got better software for the ARTS automated radar tracking system. Such as MSAW. minimum safe altitude warning, and LA low altitude alert , if on a ifr beacon transponder code.
Living in palm springs now , reminds me of Dean Martins son in an F-4, and Frank Sinatra’s mother
in a learjet. They hit almost in the same spot on Mt San jacinto. mountains. no radar at the time, LA center was dept. control. The dept sid is a north east turn. much lower terrain. Both planes flew rwy heading in imc into the mountain. Enough political pressure got radar at palm springs and their own radar room. Now moved to socal appch in san diego.
Keep up your good articles!
best,
michael.
Thanks for reaching out and for the kind comments. That’s an oldie, but one of my favorites.
I used to enjoy it when controllers could ride jumpseat. I felt like a lot of healthy interaction took place on those flights (although it seemed like it was always flights going into Vegas on Fridays and out of Vegas on Mondays that they happened. 😄).