June 2010 NOT ALL THAT long ago even the best full-motion jet simulators had very basic visual presentations that were restricted to nothing more than a view straight ahead through the windshield.
Most simulators had a television-style screen mounted in each pilot’s windshield, and you couldn’t really see anything except the display on your side. And there were no visual displays for the cockpit side windows. It was the best technology the 1970s and early ’80s could provide.
This straight-ahead view only, with no peripheral clues — plus the lack of texture and detail in the only visual display available — made it very difficult to control the simulator flying visually. Because of the lack of details on the single-window visual displays, most pilots would do a perfectly fine job of flying the instrument approach procedure down to minimums, and then would wander all over the place once they looked up and tried to complete the landing.
What these simulators taught you with their limited visual cues was to not look up, at least not on an ILS approach. If you just kept your eyes on the instruments and tracked the ILS signal all the way to the runway, everything worked out fine. That was not the intended lesson, but that’s how you learned to get through simulator training.
Though the simulators lacked realism, they unintentionally reinforced a lesson all instrument pilots need to learn, and that is that the visual segment at the end of every IFR approach can be very difficult and, because you are so close to the ground, is very risky.
Today’s full-capability Level C and D simulators have excellent visual displays that wrap around for nearly a 180-degree field of view. The images on the displays are detailed with texture in terrain and even pavement, so our eyes can detect distance and relative motion. And most importantly, the images accurately portray visibility reduced by precipitation and fog so that you can never really get a crystal clear view of the runway all the way to touchdown when visibility is low.
Flying a modern simulator visually is realistic, and the fidelity of these simulators is one of several reasons the safety record of the jet pilots who train in them regularly is so good. When jet crew members encounter minimum visibility conditions at the end of an approach, it looks familiar because they have been there before in the sim.
The situation for the pilots of piston-powered airplanes, and other airplanes that cannot support the multimillion-dollar expense of a full-capability simulator, is not so good. The simulators for piston airplanes that do exist can’t match the visual authenticity of the jet sims, so training in them is better than not but still doesn’t match actual conditions.
Learning to fly IFR, and then practicing approaches under a hood, is totally useless in preparing a pilot for that final visual segment of an approach and landing in real instrument meteorological conditions (IMC). The hood limits your view, forcing you to fly by the instruments until reaching the end of the approach procedure, and then you take off the hood and are in instant VFR. Mother Nature doesn’t work that way. The irony is that instrument approaches end with a scud run to the runway, sometimes in very reduced visibility with a low ceiling.
What The Rules Say
The FAA is well aware of the risk involved in the visual portion of instrument approaches and has made rules to try to minimize the hazard. FAR 91.175 is one of the longest rules in the book and describes the conditions under which a pilot may descend below the minimum descent altitude (MDA) or decision height (DH) during an instrument approach.
Many IFR pilots get wrapped up in the list of features that must be clearly visible to continue down from decision height on an ILS approach, and that is important, but the more fundamental requirements for descent are summarized at the very top of FAR 91.175. That part of the rule demands that we do not descend below minimum authorized altitudes unless the runway is clearly visible and that, at all times, the airplane is in a position to make a normal approach and landing without need for unusual steep banks or dives.
As is often the case, the FAA rule writers don’t always talk to each other. The guys who laid out the basics of FAR 91.175 many years ago were dealing in common sense. The last thing any pilot should do is be yanking and banking in low visibility while flying close to the ground. That is obvious.
The other part of the FAA rule-making process, the group that lays out the requirements for instrument approaches, ignores FAR 91.175 in some cases by establishing minimum altitudes and decision points from which you cannot reach the runway without radical and unusual maneuvering.
For example, at Williamsport, Pennsylvania, the minimum visibility requirement for the localizer-only approach to Runway 27 is one mile, pretty typical for a straight-in nonprecision approach. However, the minimum descent altitude is 1,400 feet above the runway. So if you see the runway clearly at one mile, as the visibility minimum allows, you need to descend 1,400 feet in a mile. I don’t know of any normal airplane that can lose 1,400 feet of altitude in one mile without the extreme maneuvering the FAR 91.175 outlaws.
Williamsport is a radical example of how impossible it can be to get down to the runway from the minimum descent altitude, but it is not alone. A more typical example would be to have a visibility minimum of one mile with the MDA at 500 or 600 feet above the runway. To transition from level flight at the MDA to a normal approach to the runway that requires descending 500 feet in just one mile is a challenge. If you have a groundspeed of 90 knots, you will cover the distance to the runway in about 45 seconds. That translates into an average descent rate of about 750 fpm. That is a pretty high descent rate close to the ground while flying with only a mile of visibility.
The point is that what is legal in terms of the minimums on the chart might not be legal in terms of FAR 91.175 and its requirements for a normal visual approach to landing. And what’s legal may not always be wise. It’s vital to understand that there is no margin built into the minimum altitudes and visibilities on an approach, particularly a nonprecision approach.
Where to Look
While you’re still in the clouds, the question of where to look has only one obvious answer — the flight instruments. But once you emerge into visual conditions, especially in the crummy visibility permitted by minimums on most approaches, you have to split your attention between the view outside and the critical information on the instruments.
When flying as a crew, we are all taught to divide the duties, with one pilot looking out the windshield and the other monitoring the instruments. Typically the pilot flying stays on the gauges until the pilot not flying, who has been looking out the windshield, spots the runway. At that point the roles can reverse, and the pilot flying looks up and continues visually to the runway while the pilot not flying assumes the instrument monitoring.
The reason it’s necessary to monitor the instruments during the visual portion of the approach is that the view through the windshield can be very misleading. In low visibility and flying close to the ground, it can be very difficult to judge rate of descent, and it’s easy to develop a higher than desired sink rate. A visual-approach slope guidance device at the runway is a huge help for staying on the glidepath, but many runways at general aviation airports don’t have such aids, so paying attention to vertical speed is important during the final approach.
Another reason to check the instruments frequently is that airspeed is virtually impossible to estimate by looking out of the windshield, and it’s easy to fly too fast or too slow on the way from the landing decision point to the runway. Airspeed management is made even more challenging because you probably had only approach flaps extended until seeing the runway. So now you have the configuration change of extending landing flaps and adjusting attitude to maintain an appropriate glidepath, both of which can require significant power changes to stay on target approach speed.
In a crew, the pilot monitoring the instruments pays careful attention to the airspeed, of course, and verbally tells the pilot flying of any trends away from Vref. He also monitors vertical speed to be sure that a dangerous sink rate doesn’t develop. If the approach is an ILS, the pilot watching the instruments monitors the localizer and glideslope all the way to the runway and alerts the pilot of any deviation.
It’s easy to see how a trained crew has a safety edge over the single pilot at the end of an approach, but the single pilot can compensate some by using a capable autopilot. The best way to fly an approach single-pilot is with the autopilot coupled. The autopilot becomes the “pilot flying,” and the human pilot is then freed up to divide his attention between the instruments and the view ahead.
You should leave the autopilot coupled until you have a solid view of the runway. On an ILS with everything stabilized, I am a little slow to uncouple the autopilot at DH and use a few more seconds to be certain I can see enough to land and that airspeed and sink rate are on target.
Flying a nonprecision approach is more complicated and involves more risk than flying the ILS does, but again, it pays to be slow to punch off the autopilot once you gain sight of the runway. A few seconds of time looking out the windshield and analyzing your position relative to the runway are well spent while the autopilot holds altitude and course. Often I will even disengage the altitude hold mode, but not the autopilot, and command the autopilot to begin a descent. This permits even more time to look out the windshield while still glancing at the instrument to be sure my speed and sink rate are OK.
Should You Circle?
The riskiest instrument approach of all ends with a circle to the runway instead of continuing straight in. Some circling approaches involve a complete traffic pattern to the other end of the runway, for example. Others may involve a final approach course that doesn’t line up with any runway, so you have to make some kind of base-to-final turn maneuver to line up after you acquire the airport visually.
The reason there is so much risk in a circling approach is that you are flying very close to the ground — typically 500 feet or maybe even less — in visibility as low as one mile, and those conditions really screw up our normal sensations of flying a traffic pattern. It’s easy to bank too steeply, or to lose track of airspeed as you look out the windshield at the runway through the murk, as the ground rushes past much closer than we are accustomed to seeing. You must not lose sight of the runway at any time, but you also need to carefully monitor altitude, airspeed and sink rate.
The risks in a circling approach in minimum weather conditions are so great that many operators don’t allow their pilots to fly them. And a circle in low visibility at night is about the highest risk maneuver I can think of in instrument flying.
I was based in Kansas City, Kansas, for Flying in the 1980s, and there is a pretty common winter and early-spring weather phenomenon in the Great Plains that creates very strong surface winds from the south with widespread low ceilings. At the time, New Century Airport had no approaches to the south, so we had to circle out of the ILS approach to Runway 36. Even with better than the one-mile visibility minimum, it was very uncomfortable to be flying downwind at 500 feet or a little less above the ground with a rip-roaring tailwind blowing during my turn to base, and then final, up toward Iowa. I flew the circle at near minimums only once in the dark and swore never again. And the terrain around that airport is very flat. Imagine a circling approach in hilly country and you can see why some pilots just won’t fly them.
One of the many benefits of GPS is that the need for circling approaches is vanishing at many airports, including at New Century, where there is a GPS approach now to Runway 18, so the need for a terrifying circle is gone. The same is true at thousands of other airports, and with luck, GPS will make a circling approach very rare. However, GPS can’t move hills or obstructions, so at some airports a straight-in approach won’t be possible to all runways even though GPS guidance is available.
The Essential Autopilot
If you must circle, the best way is, again, with split duties in a crew. One pilot looks out to maintain visual contact with the runway while the other monitors the instruments and calls out deviations in airspeed or altitude.
A good autopilot can be the other crew member for the single pilot in a circle. With the altitude hold and heading modes selected, the autopilot will not go below minimum altitude, and the pilot can rotate the heading bug to command the necessary turns. With the autopilot holding altitude and restricting bank angles, you can take quick glances at the airspeed to make sure it’s not trending up or down. In a real circle in low weather, I would leave the autopilot coupled until turning onto final, when you can begin the descent.
If you think I believe an autopilot is essential for safe single-pilot instrument flying, you’re right. And I’m in the good company of the FAA and major training providers such as FlightSafety and Simuflite. The FAA requires that a full-capability autopilot be operating for all single-pilot flights in jets — no autopilot, no legal flight without a copilot. And the training companies in conjunction with the FAA demand that pilots fully understand and use the autopilot for all critical portions of the training and for checking flights in the simulator. Of course, the autopilot will fail occasionally in the sim, but it must be used at other times in single-pilot flying.
If you don’t have a capable auto- pilot, you are at higher risk while flying instruments single-pilot. You can try to mitigate that risk by avoiding near-minimum weather conditions, but that doesn’t always work. The major issue is lack of accurate weather reporting at many general aviation airports. You may believe, based on available observations from other airports, that you are flying an approach with lots of margin only to find out that, when you descend to the minimum altitude, visibility conditions are worse than you expected.
The bottom line to completing an instrument approach safely is really no different than it is in other aspects of flying — it takes discipline. If you apply the preamble of FAR 91.175 and follow the requirements that you see the runway clearly and be in a position to make a normal approach and landing at all times during the visual portion, the risk is very small. If you instead opt for the zero margin for error that the approach minimums allow, you are going to be flying the most risky kind of scud running imaginable.