Around 6:10 that morning, a chartered Learjet 35A departed Republic Airport (FRG) in Farmingdale, Long Island, on a short positioning flight to nearby GON. The crew was to pick up two passengers in Groton and fly them to a business meeting in Buffalo, N.Y. Weather in the region was basically VFR, but with localized areas of low clouds and fog. The automated weather observation at Groton airport reported 4,100 scattered and 9 miles visibility. In reality, conditions to the south and west were clear, and a 10-knot breeze from the southeast combined with a 2° temperature/dew point spread churned up the usual low clouds and scud just to the north and northeast of the airport.

Approaching from the shoreline about 7 miles west of the airport, the pilots reported they had the airport in sight and canceled their IFR flight plan. The aircraft entered a left downwind for Runway 23 at about 1,800 feet and continued its descent as it maneuvered for the runway. Reading the narrative in the preliminary NTSB report, it would seem the pilots were having difficulty maneuvering under VFR for a landing on Runway 23. According to the report, the aircraft turned base about 2.3 miles northeast of the airport. About 1.5 miles from the runway -- and while south of the extended centerline -- the aircraft turned left and then back right again. About one-eighth of a mile south of the runway threshold, the aircraft made a 60-degree right turn back toward the runway, crossed the runway at about 200 feet agl, and then turned back left toward the center of the airport.

The left turn continued, and the aircraft re-entered a left downwind at about 300 feet agl -- well below the 640-feet circling minimums for the airport. The last radar contact put the aircraft about one-eighth of a mile from the end of the runway.

A witness preflighting an aircraft at the time watched as the Learjet entered the downwind but lost sight of it as it skimmed or passed behind the low clouds. The aircraft reappeared about 200 feet agl and appeared to overshoot the extended runway centerline. According to the witness, the bank angle increased to about 90 degrees, and as it turned final at low altitude, the aircraft disappeared behind the trees. It was about 6:39 a.m.

The crash scene was horrendous. The Learjet clipped the first housetop, careened through a line of pines and hardwoods, and slammed into two more homes before tumbling into the Poquonnock River. Fuel sprayed through the air, igniting in an intense fireball. Almost miraculously, nobody on the ground was killed, but one woman was injured while she jumped from a bedroom window as her house was suddenly engulfed in flames. Another family survived only because they had spent the night away.

The area was closed off as firefighters battled the flames and later as crash investigators scoured the scene and sifted through the wreckage. By the time I arrived from my home, barriers had been erected to keep the public at bay, but the charred remains sent shivers up my spine. The glaring question, of course, is how and why did two highly experienced professional pilots get themselves into such a tight corner?

Circling Maneuvers

While it is unclear precisely why the pilots elected to land on Runway 23 when the approach to Runway 5 -- which has an ILS -- was reportedly clear of clouds, the fact is that they entered the downwind and attempted a "standard" traffic pattern to the usual no-wind runway. Unfortunately, the pattern was anything but standard.

When reverting to visual flight rules in generally sunny and clear conditions, it's easy to disregard the standard procedures that would be followed when conducting a circling maneuver at the end of an IFR approach. Although the flight crew in this case had canceled IFR and was making a VFR approach to the field, they might have fared better had they abided by the published circling minimums and followed standard operating procedures (SOPs) for a bona fide circling approach.

Circling approach procedures may vary from one operator to another, but a number of common threads run through the fabric of various operators' SOPs. Although not specifically required by regulation, some operators require that the MDA be maintained until the aircraft has maneuvered to within a 30-degree angle of the extended runway centerline. Some

 operators require the use of an autopilot during a circling maneuver, with "altitude hold" mode selected to maintain the MDA and "heading select" used for course tracking. Restrictions or even prohibitions on the use of circling approaches at night -- a visual condition that heightens risk -- are also common among many professional operators.

"According to our SOP," explains one flight manager for a fractional jet operator, "we can perform a circling approach only in daylight, and then the conditions must be at least 1,000 feet and 3 miles."

Other aspects of the circling approach are more clearly defined and universally accepted. According to FAA Order 8400.10, Air Transportation Operations Inspector's Handbook, which details the content and performance standards of FAA flight checks, "The circling MDA must be maintained until an aircraft [using normal maneuvers] is in a position from which a normal descent [less than 1,000 fpm] can be made to touchdown within the touchdown zone."

Even a proper circling approach has its drawbacks. While a circling maneuver as defined in 8400.10 implies a stabilized approach from the circling MDA, some view the circling procedure as running contrary to the stabilized approach philosophy. Especially when the decision to circle comes late, pilots may end up playing catch-up as they set up and perform the circling maneuver, and the ideal stabilized approach falls apart at the seams. At the very least, the circling maneuver requires a departure from a constant rate of descent as the aircraft levels off to complete the circle.

The Stabilized Approach

The importance of a stabilized approach has long been recognized in the turbine aircraft world. By definition, a stabilized approach is one in which the aircraft is properly configured and maintains a constant airspeed and descent rate as it follows a constant vertical flight path to the touchdown zone.

The underlying philosophy of the stabilized approach as outlined in FAA Order 8400.10 is that it allows pilots to "determine displacements from the course or glidepath centerline, to mentally project the aircraft's three-dimensional flight path, and then to apply the control inputs necessary to achieve and maintain the desired approach path." More specifically, the stabilized approach technique allows a pilot to more readily recognize when problems such as wind shear cause the aircraft to deviate from the desired flight path or airspeed. Perhaps just as important is that flying a stabilized approach keeps the workload to a minimum during a time when a pilot's attention must be keenly focused on the world around him.

Most operators, as recommended by the FAA, have SOPs that require pilots to be stabilized 1,000 feet agl for an instrument approach, or 500 feet agl for a VFR approach. As described in the Flight Safety Foundation Approach and Landing Accident Reduction (ALAR) Risk Awareness Tool Kit, the recommended elements of a stabilized approach include a number of criteria, including correct flight path, airspeed, configuration, descent rate and completion of all briefings and checklists.

Despite the emphasis placed on stabilized approach procedures, some operators still don't embody all the principles in their SOPs. Especially in the world of light turbine aircraft and small jets, "stabilized" approach characteristics can involve slowly reducing airspeed throughout a lengthy portion of the procedure. As one pilot notes, "This requires a lot of extra trimming and attention to airspeed control throughout the procedure, but it's the way we've all been taught."

The truth is that not all pilots religiously follow the edict to fly a stabilized approach, and the results show. According to the Annual Turbine Aircraft Accident Review 2002 by Robert E. Breiling Associates, 8 percent of worldwide jet aircraft accidents in the past five years occurred during the approach phase. Turboprops fared worse, with 18 percent of the accidents occurring in the approach phase. As Breiling notes, "Unstabilized approach conditions were implicated in numerous landing accidents, which account for roughly half of the total accidents."

In an article recently published by the Federation of Indian Pilots, Capt. Vivek Kulkarni of Air India states, "A detailed analysis of ALAs [approach and landing accidents] determined that CFIT, landing overruns, loss of control, runway excursion and non-stabilized approaches accounted for 76 percent of all occurrences. Indeed the non-stabilized approach was probably the initiating event in most incidents." In an Obstacle Clearance Panel working paper presented at a meeting in St. Petersburg, Russia, in 1996, author J. W. Gregory makes a connection between unstabilized approaches and CFIT accidents. It seems likely that the lack of a stabilized approach contributed to the sad end of the Learjet 35 crew this past summer in Groton.

Maneuvering Factors

As professional pilots, we like to think we're well beyond the point at which we could fall victim to a stall-spin situation, especially in the benign environment of a circling maneuver in VFR conditions. But the statistics suggest otherwise. A recent study by the Air Safety Foundation found that commercial-licensed pilots are more likely to fall victim to stall-spin accidents than are private or student pilots. When a stall-spin develops at pattern altitude, the chances of recovery or survival are essentially nil.

Several factors come into play when we leave the stabilized approach regime and begin maneuvering around the pattern for a landing. Not only does initiating a maneuver -- particularly one with significant angles of bank -- destabilize the approach, but it introduces further complications. A critical factor to consider when transitioning into, say, a circling approach is the relationship between Vref, bank angle, load factor and stall speed. For example, a Vref of 1.3 Vso might provide an adequate margin above stall as long as maneuvering is limited, but not so when making steep banking maneuvers required for collision avoidance or other unexpected situations. Remember that in a level turn, stall speed increases sharply once we exceed moderate bank angles.

Up to 20 degrees of bank, the increased load factor and attendant increase in stall speed in a level turn are negligible. At 40 degrees of bank in a level turn, we've increased stall speed by 14 percent. Increase the bank to about 55 degrees in that level turn, and we're right at 1.3 Vso. It's for reasons such as this that some operators require the use of autopilots for circling maneuvers, particularly when the autopilots limit the bank angle.

In fact, a circling maneuver provides an excellent opportunity to make a fatal mistake with respect to the stall margin. With flaps and gear extended, lots of additional power is required when leveling off to maintain airspeed. Distractions outside the cockpit can easily divert the pilot's attention, compromising airspeed control and eroding the margin above stall speed.

I recall clearly the sunny September day when a Piper Cheyenne returning from a maintenance test flight at Jefferson County Airport (BJC) in Colorado got caught in just that trap. While on a straight-in approach to Runway 11, the pilot was advised of traffic ahead, and was asked to slow his speed as much as possible. The pilot acknowledged, and about 15 seconds later was asked to make a 360-degree turn to the left for spacing. The pilot complied.

The aircraft began a steep turn, and was then seen entering a steep, spiraling, nose-down descent. The pilot and two mechanics aboard were killed in the crash. No mechanical failure or malfunctions were found in the investigation. The NTSB report concluded the probable cause to be "failure of the pilot to maintain adequate airspeed while maneuvering for spacing in the traffic pattern, which resulted in an inadvertent stall at low speed." Clearly, the pilot's straight-in approach had been destabilized, and the maneuvering load factor played a part in the inadvertent stall.

Further complications come when we attempt to maneuver with respect to ground references, terrain, or even clouds. When a pilot attempts to follow a desired ground track -- such as when turning to align with the runway -- or maneuver around terrain or clouds, the focus shifts from cockpit references to outside references, and airspeed control can suffer.

Put birds, another airplane or wind gusts into the picture, and the problem only worsens.

At an airport like Groton, these factors are likely. Just to the south of Runway 5-23 lies Bluff Point, a bird sanctuary and recreation area bordering the Poquonnock River. Rich in shellfish, the brackish waters of the Poquonnock provide a haven for a multitude of migratory waterfowl and numerous birds that reside year-round. Especially at low tide when the shellfish are more readily available, flocks of gulls and other sea birds can be found in the area. The airport is also home to two flight schools and serves as a base for an active fleet of light general aviation aircraft. Hazards in the pattern are not uncommon.

In the case of the Learjet 35 at Groton, it appears the pilot was maneuvering around low, scuddy clouds, as well as making a tight turn to final. But yet another factor may have come into play. Remember that when maneuvering with reference to the ground, the highest bank angle occurs when the ground speed is highest -- i.e., with a tailwind. In this case, the winds were 150 at 10 knots, meaning the pilots had a 10-knot tailwind on base, and thus would have required a slightly steeper bank angle maneuvering from base to final. This factor is a minor one at best, but perhaps represents another nail in the proverbial coffin.

Final Blow

The preliminary NTSB report on the accident contains one final item of information that's particularly telling. It says the captain's airspeed bug was set 20 knots higher than that of the first officer. It's unclear exactly who was flying at the time, or where the airspeed bugs should have been set, but it's clear enough that a 20-knot error in airspeed combined with even a moderate bank angle could have put the pilots squarely in the danger zone. If nothing else, such an error indicates a breakdown in CRM and suggests that the pilots may have been rushed.

In fact, "get-there-itis" and time pressure on the part of pilots and controllers alike may be the real culprits behind many unstabilized approaches. As one flight manager points out, "I've seen this too many times, with pilots rushing to get somewhere, complying with ATC instructions, or taking a night visual approach when an instrument approach would add only a few extra minutes to the flight."

In its investigation of the Southwest 737 overrun accident at Burbank-Glendale-Pasadena Airport (BUR) in March 2000, the NTSB said a contributing factor in that career-ending mishap was "the controller's positioning of the airplane in such a manner as to leave no safe options for the flight crew other than a go-around maneuver." Of course the onus was on the flight crew's "failure to abort the approach when stabilized-approach criteria were not met."

The May 2003 Callback, from NASA's Aviation Safety Reporting System, reports three cases where flight crews were backed into a corner by ATC, causing them to violate their stabilized approach mandate.

A circling approach -- or standard traffic pattern -- is often the maneuver required to complete a flight, but it's not always as simple as it appears. Only by developing and strictly following a clear set of SOPs can we limit the risks that can turn a simple maneuver to landing into a truly final approach. B/CA

STABILIZED APPROACH CRITERIA

The following is provided by the Flight Safety Foundation as part of its Approach and Landing Accident Reduction (ALAR) Risk Awareness Tools. More information is available at www.flightsafety.org.

Recommended Elements of a Stabilized Approach

All flights must be stabilized by 1,000 feet above airport elevation in IMC and by 500 feet above airport elevation in VMC. An approach is stabilized when all of the following criteria are met:

(1) The aircraft is on the correct flight path;

(2) Only small changes in heading/pitch are required to maintain the correct flight path;

(3) The aircraft speed is not more than Vref + 20 knots indicated and not less than Vref;

(4) The aircraft is in the landing configuration;

(5) Sink rate is no greater than 1,000 fpm; if an approach requires a sink rate greater than 1,000 fpm, a special briefing should be conducted;

(6) Power setting is appropriate for the aircraft configuration and is not below the minimum power for approach as defined by the aircraft operating manual;

(7) All briefings and checklist have been conducted;

(8) Specific types of approaches are stabilized if they also fulfill the following: Instrument landing system (ILS) approaches must be flown within one dot of the glideslope and localizer; a Category II or Category III ILS must be flown within the expanded localizer band; during a circling approach, wings should be level on final when the aircraft reaches 300 feet above airport elevation; and,

(9) Unique approach procedures or abnormal conditions requiring a deviation from the above elements of a stabilized approach require a special briefing.

An approach that becomes unstabilized below 1,000 feet above airport elevation in IMC or below 500 feet above the airport in VMC requires an immediate go-around.

SAFE APPROACH STRATEGIES

(1) Keep safety a priority, even when schedules are tight.

(2) When available, fly an instrument approach rather than a visual, especially for a night arrival. Doing so improves instrument proficiency (especially for non-precision approaches) and heightens safety.

(3) When making a VFR approach, abide by the IFR circling minimums and procedures.

(4) When possible, choose a straight-in instrument approach rather than a circling maneuver.

(5) Take steps to avoid situations that destabilize an approach, and use an approach speed that allows any anticipated maneuvering.

(6) Consider stabilized approach procedures as limits, not targets. If an approach becomes unstabilized below the minimum altitude (1,000 feet agl for IFR, 500 feet agl for VFR), go around.

(7) Advise ATC and decline clearances if they violate your stabilized approach requirements.

Reprinted from the November 2003 issue of Business & Commercial Aviation magazine